Compare commits

..
10 Commits
Author SHA1 Message Date
swung0x48 4322427e78 [Fix] (DirectGLES): lower gl_ClipDistance for Adreno's ESSL compiler - shadow the builtin in a Private array with constant-index flushes before EmitVertex/return, loop-copy gl_in clip distances through dynamic indices (whole-array reads segfault the Qualcomm compiler, constant-index element reads miscompile), strip the SPIRV-Cross redeclaration Adreno rejects, and split const struct-array LUT initializers so they stay dynamically indexable; quirk-gated to Qualcomm with MOBILEGL_QUIRK_CLIP_DISTANCE override 2026-07-26 19:56:45 -04:00
swung0x48 203d4bce5e [Fix] (DirectGLES): piglit fixes batch 1 - keep enabled-but-unsourceable vertex attribs disabled on the backend VAO (Adreno memcpy-from-NULL SIGSEGV on gl-3.1-vao-broken-attrib), content-sync READ-framebuffer texture attachments before blits, clamp out-of-bounds access-chain indices via GraphicsRobustAccessPass before ESSL transpile (Adreno poisons whole-shader output on constant OOB), and fold ConstOffset into the coordinate for 1D texelFetch (SPIRV-Cross emulates 1D as 2D but leaves the scalar offset, which ESSL rejects) 2026-07-26 19:17:25 -04:00
swung0x48 761114d022 Merge remote-tracking branch 'origin/cts-gl33' into dev 2026-07-26 18:29:02 -04:00
swung0x48 9caf34d5b1 [Fix] (Logging): rank WARN/ERROR above INFO so release builds keep them - the old ordering (DEBUG=0, WARN=1, ERROR=2, INFO=3) compiled every MGLOG_W/MGLOG_E out of the default MOBILEGL_LOG_LEVEL_INFO build, silently hiding backend shader-compile failures, unsupported-path skips, and enum-conversion fallbacks during the piglit runs 2026-07-26 18:14:56 -04:00
swung0x48 5e676b338b [Fix] (DirectVulkan): back legacy low-bit formats (RGB565/RGB5A1/RGBA4/R3G3B2/RGB4/RGBA2/RGB10/12) with their UNorm8/16 canonical shadow layouts and add capability fallbacks - they mapped to VK_FORMAT_UNDEFINED and crashed or wedged the GPU on upload; also admit 2DMSArray/CubeMap/3D color attachment targets in the render pass 2026-07-26 13:28:30 -04:00
swung0x48 db01bfa3e8 [Fix] (DirectVulkan): general (format,type) readback conversion - hoist the CTS-verified StoreWideRowsToClient into shared ReadbackImpl and decode any color VkFormat to wide RGBA rows; readback previously supported only RGB/BGR/RGBA/BGRA x UNSIGNED_BYTE/FLOAT and silently returned zeros for everything else 2026-07-26 13:28:30 -04:00
swung0x48 cc3dcfd80e [Fix] (DirectVulkan): support UBO instance arrays as arrayed descriptors - uniform Block{...}b[N] reflected as one binding with descriptorCount=N, per-element GL block mapping, per-element buffer infos and dynamic offsets; non-UBO descriptor arrays now fail program creation cleanly instead of continuing corrupt 2026-07-26 12:25:24 -04:00
swung0x48 d9556ff041 [Fix] (DirectVulkan): implement color renderbuffer attachments - render pass/pipeline/blit/copy/readback/clear paths treated color renderbuffers as absent (writes masked to VK_ATTACHMENT_UNUSED, glClear dropped, readback zeros) 2026-07-26 11:30:54 -04:00
swung0x48 39f21e52ea [Test] (CTS): isolate the DirectVulkan renderbuffer-FBO readback defect so the rest of KHR-GL33 can be measured 2026-07-26 10:43:48 -04:00
swung0x48 0e933b8f2f [Test] (CTS): run VK-GL-CTS KHR-GL33 against MobileGL on Android via a standalone glcts binary 2026-07-26 08:05:02 -04:00
149 changed files with 2140 additions and 17634 deletions
-2
View File
@@ -25,5 +25,3 @@ MobileGL/MG*/cmake-build*
/android-plugin/app/src/trace/jniLibs
/android-plugin/local.properties
tools/trace_replay/work/
__pycache__/
*.py[cod]
+3 -15
View File
@@ -191,10 +191,12 @@ set(SOURCE_FILES
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/DecoratePositionInvariantPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/LowerDrawParametersPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/RebaseInstanceIndexPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/NormalizeRectCoordinatesPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/StripUboMemberRelaxedPrecisionPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/StripNoPerspectivePass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/EmulateNoPerspectivePass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/FoldConstOffsetFor1DFetchPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/LowerClipDistanceForEsslPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/DefeatConstStructArrayLutPass.cpp
MobileGL/MG_Util/BackendLoaders/OpenGL/Loader.cpp
MobileGL/MG_Util/BackendLoaders/Vulkan/Loader.cpp
@@ -456,21 +458,8 @@ if (ANDROID)
endif()
if (APPLE AND NOT MOBILEGL_IOS)
# MobileGL statically embeds glslang, SPIRV-Tools, and SPIRV-Cross. When
# this dylib is injected with DYLD_INSERT_LIBRARIES, exporting those C++
# symbols interposes incompatible copies embedded by host libraries such
# as shaderc. Keep only the public GL/EGL/CGL loader surface globally
# visible; GetProcAddress can still return pointers to hidden internals.
set(MOBILEGL_MACOS_EXPORTED_SYMBOLS
"${CMAKE_CURRENT_SOURCE_DIR}/MobileGL/MG_Impl/DyldInterpose/ExportedSymbols.txt")
target_link_options(${CMAKE_PROJECT_NAME} PRIVATE
"LINKER:-exported_symbols_list,${MOBILEGL_MACOS_EXPORTED_SYMBOLS}")
set_property(TARGET ${CMAKE_PROJECT_NAME} APPEND PROPERTY
LINK_DEPENDS "${MOBILEGL_MACOS_EXPORTED_SYMBOLS}")
target_link_libraries(${CMAKE_PROJECT_NAME} PUBLIC
"-framework Cocoa"
"-framework CoreVideo"
"-framework QuartzCore"
"-framework Foundation"
"-framework OpenGL"
@@ -478,7 +467,6 @@ if (APPLE AND NOT MOBILEGL_IOS)
if(TARGET ${CMAKE_PROJECT_NAME}_s)
target_link_libraries(${CMAKE_PROJECT_NAME}_s PUBLIC
"-framework Cocoa"
"-framework CoreVideo"
"-framework QuartzCore"
"-framework Foundation"
"-framework OpenGL"
+6 -1
View File
@@ -14,7 +14,7 @@ namespace MobileGL::MG_Config {
inline const String ProjectName = "MobileGL";
inline const String CoreName = "MobileGL Core";
inline const String CoreVendor = "MobileGL-Dev (BZLZHH, Swung0x48, Tungsten)";
inline const Version CoreVersion = {26, 8, 0, "-dev", VersionType::Development};
inline const Version CoreVersion = {26, 7, 0, "-dev", VersionType::Development};
inline const VersionStringFormatAttrib DefaultVersionStringFormatAttrib = {2, 2, 0, true, true};
inline const Uint64 CacheVersion = 0;
@@ -80,6 +80,11 @@ namespace MobileGL::MG_Config {
// rewrites the recognized workgroup prefix-scan template on Qualcomm devices with
// subgroups wider than 32 lanes (see ShaderSourceProcessor's quirk registry).
QuirkOverride SubgroupPrefixScanQuirk = QuirkOverride::Auto;
// MOBILEGL_QUIRK_CLIP_DISTANCE: overrides the DirectGLES quirk that lowers
// gl_ClipDistance for Adreno's ESSL compiler (shadow Private arrays with
// constant-index builtin flushes, dynamic-index gl_in copy loop, redeclaration
// strip, and const struct-array LUT splitting). Auto detects Qualcomm.
QuirkOverride ClipDistanceQuirk = QuirkOverride::Auto;
// MOBILEGL_MAGMA_DISABLE_BLENDED_DEPTH_WRITE: overrides the DirectVulkan quirk that
// strips depth writes from accumulation-blended pipelines (MIN/MAX or additive
// ONE+ONE - the multi-pass depth-equality signature) on drivers without
+1
View File
@@ -135,6 +135,7 @@ namespace MobileGL::MG_ConfigLoader {
features.DisableUboRing = QueryEnvFlag("MOBILEGL_DISABLE_UBO_RING");
features.RelaxedSemantics = QueryEnvFlag("MOBILEGL_RELAXED_SEMANTICS");
features.SubgroupPrefixScanQuirk = QueryEnvQuirkOverride("MOBILEGL_QUIRK_SUBGROUP_PREFIX_SCAN");
features.ClipDistanceQuirk = QueryEnvQuirkOverride("MOBILEGL_QUIRK_CLIP_DISTANCE");
features.MagmaDisableBlendedDepthWriteQuirk =
QueryEnvQuirkOverride("MOBILEGL_MAGMA_DISABLE_BLENDED_DEPTH_WRITE");
features.DisableRobustBufferAccess = QueryEnvFlag("MOBILEGL_DISABLE_ROBUST_BUFFER_ACCESS");
+4 -13
View File
@@ -14,7 +14,6 @@
#include <MG_State/EGLState/Core.h>
#include <MG_Impl/GLImpl/Texture/ProxyTexture.h>
#include <MG_Impl/GLImpl/Framebuffer/GL_Framebuffer.h>
#include <MG_Impl/GLImpl/Sync/GL_Sync.h>
#include <atomic>
#include <mutex>
@@ -38,12 +37,6 @@ namespace MobileGL {
MGLOG_I("MobileGL closing...");
}
glslang::FinalizeProcess();
// GL syncs die with their contexts, and every context is gone by the
// time full teardown runs: drain the live-sync registry while the
// backend function table can still release the backend handles (and
// before a re-initialized library could pair them with the wrong
// backend's DeleteSync).
MG_Impl::GLImpl::DestroyAllSyncObjects();
MG_Backend::pActiveBackendObject.reset();
MG_State::pGLContext.reset();
MG_State::pEGLContext.reset();
@@ -107,11 +100,9 @@ namespace MobileGL {
// (EGL/WGL/CGL): initialization happens lazily on the first entry point
// via EnsureInitialized(), and full teardown happens deterministically
// when the last EGL display is terminated with nothing current (EGLImpl
// calls Destroy()). There is intentionally no backend-initializing static
// constructor, no static destructor, and no DllMain: the global singletons
// use leak-at-exit storage (see GlobalObjects.cpp), so a process that exits
// calls Destroy()). There is intentionally no static constructor, no
// static destructor, and no DllMain: the global singletons use
// leak-at-exit storage (see GlobalObjects.cpp), so a process that exits
// without eglTerminate simply leaks them to the OS instead of running
// backend destructors during static teardown. macOS has a lightweight
// dyld constructor that installs NSOpenGL dispatch hooks only; full backend
// initialization still enters here from the first hooked CGL context.
// backend destructors during static teardown.
} // namespace MobileGL
+3 -4
View File
@@ -13,10 +13,9 @@ namespace MobileGL {
void Initialize();
// Thread-safe, idempotent, and re-entrant wrapper around Initialize().
// Host layers (EGL/WGL/CGL entry points) call this lazily on first use so
// full backend initialization never depends on ELF/DLL static constructors,
// and so a fresh init can follow a full Destroy() (e.g. after the last
// eglTerminate). The macOS dyld bootstrap installs only lightweight
// NSOpenGL method hooks.
// MobileGL's lifecycle never depends on ELF/DLL static constructors, and
// so a fresh init can follow a full Destroy() (e.g. after the last
// eglTerminate).
void EnsureInitialized();
void Destroy();
-48
View File
@@ -220,31 +220,6 @@ namespace MobileGL {
// and leave the query readable later.
Bool (*GetQueryResult64)(BackendQueryHandle query, Bool wait, Uint64* outNanoseconds);
void (*DeleteBackendQuery)(BackendQueryHandle query);
// GL_SAMPLES_PASSED occlusion queries (optional; null = unsupported,
// the frontend then rejects the target). Results/deletion flow through
// GetQueryResult64 / DeleteBackendQuery like timer queries.
BackendQueryHandle (*BeginOcclusionQuery)();
void (*EndOcclusionQuery)(BackendQueryHandle query);
// Transform feedback primitive queries backed by real GPU query pools
// (optional; null = frontend falls back to CPU accounting).
BackendQueryHandle (*BeginXfbPrimitivesQuery)(Bool generated);
void (*EndXfbPrimitivesQuery)(BackendQueryHandle query);
// Transform feedback capture spans, for backends whose own GL/ES driver
// performs the capture (DirectGLES). Both optional; null means the backend
// drives capture from its draw recording instead (DirectVulkan). End is
// called while the frontend capture state is still active, so the backend
// can still see the capture program and buffer bindings.
// GL_PATCH_VERTICES; ES 3.2 spells it the same way.
void (*PatchParameteri)(GLenum pname, GLint value);
void (*BeginTransformFeedback)(GLenum primitiveMode);
void (*EndTransformFeedback)();
// ARB_transform_feedback2. A backend that leaves these null keeps the single
// implicit capture span the frontend has always modelled; the frontend state
// (paused flag, per-object bindings) is tracked either way.
void (*PauseTransformFeedback)();
void (*ResumeTransformFeedback)();
void (*BindTransformFeedback)(GLuint name);
void (*DeleteTransformFeedback)(GLuint name);
Int64 (*GetGpuTimestampNs)(); // glGetInteger64v(GL_TIMESTAMP); 0 if unsupported
};
struct GlobalBackendFunctionsTable {
@@ -297,13 +272,6 @@ namespace MobileGL {
Int MaxIntegerSamples = 1;
Int MaxSamples = 1;
Int MaxSampleMaskWords = 1;
// Tessellation limits; defaults are the GL 4.0 core minimums.
Int MaxPatchVertices = 32;
Int MaxTessGenLevel = 64;
// GL_MIN/MAX_PROGRAM_TEXTURE_GATHER_OFFSET. Defaults are the GL 4.0 core
// minimums, which every ES 3.1 driver also guarantees.
Int MinProgramTextureGatherOffset = -8;
Int MaxProgramTextureGatherOffset = 7;
Int MaxTextureImageUnits = 32;
Int MaxVertexTextureImageUnits = 32;
Int MaxComputeTextureImageUnits = 32;
@@ -315,8 +283,6 @@ namespace MobileGL {
Int MaxComputeWorkGroupInvocations = 128;
Int MaxShaderStorageBufferBindings = 8;
Int MaxTextureBufferSize = 65536;
// GL_TEXTURE_BUFFER_OFFSET_ALIGNMENT; 1 means the offset is unconstrained.
Int TextureBufferOffsetAlignment = 1;
Int MaxUniformBufferBindings = 24;
Int MaxUniformBlockSize = 16384;
Int MaxImageUnits = 8;
@@ -334,21 +300,7 @@ namespace MobileGL {
Float ViewportBoundsRangeMin = 0.0f;
Float ViewportBoundsRangeMax = 0.0f;
Int ViewportSubpixelBits = 0;
// GL 4.x fragment-interpolation offset limits. These defaults are the
// core minimums and are replaced by live GLES/Vulkan device limits.
Float MinFragmentInterpolationOffset = -0.5f;
// For four fractional bits the greatest required legal offset is
// 0.5 - 2^-4 = 0.4375 (GL 4.6 table 23.70).
Float MaxFragmentInterpolationOffset = 0.4375f;
Int FragmentInterpolationOffsetBits = 4;
Bool SupportsWideLines = false;
// Whether a framebuffer whose depth and stencil attachments are distinct
// images can be rendered to. GL only requires support when both refer to the
// same image and lets an implementation answer GL_FRAMEBUFFER_UNSUPPORTED
// otherwise, which is what DirectVulkan (one combined attachment) and the
// real ES drivers behind DirectGLES both do. Defaults to true so a backend
// that never sets it keeps the permissive behaviour.
Bool SupportsDistinctDepthStencilAttachments = true;
SizeT MaxShaderStorageBlockSize = 128 * 1024 * 1024;
Uint32 SubgroupSize = 0;
Uint32 SubgroupSupportedStages = 0;
@@ -20,7 +20,6 @@
#include <MG_Util/Texture/TextureFormatProcessor.h>
#include <Config.h>
#include <algorithm>
#include <cmath>
#include <format>
namespace MobileGL::MG_Backend::DirectGLES {
@@ -210,12 +209,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
if (options & PixelFormatNormalizeOptionBit::NoDepthComponent32) {
reasons.push_back("GL_DEPTH_COMPONENT32 native probe failed on OpenGL ES");
}
if (options & PixelFormatNormalizeOptionBit::NoThreeChannelRenderTarget) {
reasons.push_back("no three-channel multisample storage format on OpenGL ES");
}
if (options & PixelFormatNormalizeOptionBit::NoSnorm16RenderTarget) {
reasons.push_back("EXT_render_snorm not supported");
}
String reason;
for (SizeT i = 0; i < reasons.size(); ++i) {
@@ -363,42 +356,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
return complete;
}
// Whether the driver renders to a framebuffer whose depth and stencil come from
// two different renderbuffers. GL only requires support when both attachments are
// the same image, and ES drivers commonly answer GL_FRAMEBUFFER_UNSUPPORTED here;
// reporting COMPLETE from the frontend and then rendering into a framebuffer the
// driver refuses leaves the results silently empty.
Bool ProbeDistinctDepthStencilAttachments(const MG_External::GLESFunctionsTable& gl) {
if (!gl.glGenFramebuffers || !gl.glBindFramebuffer || !gl.glFramebufferRenderbuffer ||
!gl.glCheckFramebufferStatus || !gl.glDeleteFramebuffers || !gl.glGenRenderbuffers ||
!gl.glBindRenderbuffer || !gl.glRenderbufferStorage || !gl.glDeleteRenderbuffers) {
return true;
}
GLint prevFramebuffer = 0, prevRenderbuffer = 0;
gl.glGetIntegerv(GL_FRAMEBUFFER_BINDING, &prevFramebuffer);
gl.glGetIntegerv(GL_RENDERBUFFER_BINDING, &prevRenderbuffer);
GLuint framebuffer = 0;
GLuint renderbuffers[2] = {0, 0};
gl.glGenFramebuffers(1, &framebuffer);
gl.glGenRenderbuffers(2, renderbuffers);
gl.glBindRenderbuffer(GL_RENDERBUFFER, renderbuffers[0]);
gl.glRenderbufferStorage(GL_RENDERBUFFER, GL_DEPTH_COMPONENT16, 4, 4);
gl.glBindRenderbuffer(GL_RENDERBUFFER, renderbuffers[1]);
gl.glRenderbufferStorage(GL_RENDERBUFFER, GL_STENCIL_INDEX8, 4, 4);
gl.glBindFramebuffer(GL_FRAMEBUFFER, framebuffer);
gl.glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_DEPTH_ATTACHMENT, GL_RENDERBUFFER, renderbuffers[0]);
gl.glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_STENCIL_ATTACHMENT, GL_RENDERBUFFER, renderbuffers[1]);
const Bool supported = gl.glCheckFramebufferStatus(GL_FRAMEBUFFER) == GL_FRAMEBUFFER_COMPLETE;
gl.glBindFramebuffer(GL_FRAMEBUFFER, static_cast<GLuint>(prevFramebuffer));
gl.glBindRenderbuffer(GL_RENDERBUFFER, static_cast<GLuint>(prevRenderbuffer));
gl.glDeleteFramebuffers(1, &framebuffer);
gl.glDeleteRenderbuffers(2, renderbuffers);
return supported;
}
Bool ProbeFramebufferCompletenessForRenderbuffer(const MG_External::GLESFunctionsTable& gl,
GLuint renderbuffer,
TextureInternalFormat format) {
@@ -562,50 +519,20 @@ namespace MobileGL::MG_Backend::DirectGLES {
}
const GLESProbeFormatInfo nativeInfo = BuildNativeProbeFormatInfo(requestedInternalFormat);
GLESProbeFormatInfo outerFallbackInfo;
const Bool outerHasForcedFallback =
BuildFallbackProbeFormatInfo(requestedInternalFormat, forcedOptions, true, outerFallbackInfo);
if (!outerHasForcedFallback) {
BuildFallbackProbeFormatInfo(requestedInternalFormat, driverOptions, false, outerFallbackInfo);
GLESProbeFormatInfo fallbackInfo;
const Bool hasForcedFallback =
BuildFallbackProbeFormatInfo(requestedInternalFormat, forcedOptions, true, fallbackInfo);
if (!hasForcedFallback) {
BuildFallbackProbeFormatInfo(requestedInternalFormat, driverOptions, false, fallbackInfo);
}
for (SizeT targetIndex = 0; targetIndex < kFormatCapabilityTextureTargetCount; ++targetIndex) {
const auto target = static_cast<TextureTarget>(targetIndex);
// A multisample texture can only ever be rendered into, so its storage format
// has to stay colour-renderable; the ordinary fallback for a three-channel
// format is a three-channel one, which ES accepts as a texture but rejects as
// multisample storage. Recompute the fallback per target so those formats get
// widened here and nowhere else.
Flags<PixelFormatNormalizeOptionBit> targetOptions;
if (IsGLESProbeMultisampleTarget(target)) {
targetOptions |= PixelFormatNormalizeOptionBit::NoThreeChannelRenderTarget;
if (!capabilities.SupportsRenderSnorm || !capabilities.SupportsNorm16Texture) {
targetOptions |= PixelFormatNormalizeOptionBit::NoSnorm16RenderTarget;
}
}
GLESProbeFormatInfo fallbackInfo = outerFallbackInfo;
Bool hasForcedFallback = outerHasForcedFallback;
if (targetOptions) {
hasForcedFallback = BuildFallbackProbeFormatInfo(
requestedInternalFormat, forcedOptions | targetOptions, true, fallbackInfo);
if (!hasForcedFallback) {
BuildFallbackProbeFormatInfo(requestedInternalFormat, driverOptions | targetOptions,
false, fallbackInfo);
}
}
// 1D, 1D-array and rectangle textures live on an ES target (see
// TextureImpl::MapToBackendTextureTarget), so they have to be probed there too -
// probing the desktop-only target itself always failed, which left those slots
// of the cache empty and stopped any fallback format from being selected for
// them (a GL_DEPTH_COMPONENT32 1D texture then got no storage at all).
const TextureTarget probeTarget = TextureImpl::MapToBackendTextureTarget(target);
Bool shouldProbeFallback = hasForcedFallback;
if (!hasForcedFallback) {
Bool nativeRenderable = false;
const Bool nativeCreated =
ProbeTexture(gl, probeTarget, nativeInfo.InternalFormat, nativeInfo.ImageFormat,
ProbeTexture(gl, target, nativeInfo.InternalFormat, nativeInfo.ImageFormat,
nativeInfo.ImageType, logicalFormat, &nativeRenderable);
if (nativeCreated) {
AddFullFormatCaps(cache, targetIndex, formatIndex,
@@ -620,7 +547,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
if (shouldProbeFallback && fallbackInfo.InternalFormat != GL_UNKNOWN_MGL) {
Bool fallbackRenderable = false;
const Bool fallbackCreated =
ProbeTexture(gl, probeTarget, fallbackInfo.InternalFormat, fallbackInfo.ImageFormat,
ProbeTexture(gl, target, fallbackInfo.InternalFormat, fallbackInfo.ImageFormat,
fallbackInfo.ImageType, logicalFormat, &fallbackRenderable);
if (fallbackCreated) {
if (AddCaveatFormatCaps(cache, targetIndex, formatIndex,
@@ -636,8 +563,8 @@ namespace MobileGL::MG_Backend::DirectGLES {
}
const SizeT renderbufferTargetIndex = GetRenderbufferFormatCapabilityTargetIndex();
Bool shouldProbeFallbackRenderbuffer = outerHasForcedFallback;
if (!outerHasForcedFallback) {
Bool shouldProbeFallbackRenderbuffer = hasForcedFallback;
if (!hasForcedFallback) {
const Bool nativeRenderbufferComplete =
ProbeRenderbuffer(gl, nativeInfo.InternalFormat, logicalFormat, false, 1);
if (nativeRenderbufferComplete) {
@@ -651,16 +578,16 @@ namespace MobileGL::MG_Backend::DirectGLES {
shouldProbeFallbackRenderbuffer = true;
}
}
if (shouldProbeFallbackRenderbuffer && outerFallbackInfo.InternalFormat != GL_UNKNOWN_MGL &&
ProbeRenderbuffer(gl, outerFallbackInfo.InternalFormat, logicalFormat, false, 1)) {
if (shouldProbeFallbackRenderbuffer && fallbackInfo.InternalFormat != GL_UNKNOWN_MGL &&
ProbeRenderbuffer(gl, fallbackInfo.InternalFormat, logicalFormat, false, 1)) {
if (AddCaveatFormatCaps(cache, renderbufferTargetIndex, formatIndex,
GetRenderbufferFeatureCaps(logicalFormat))) {
LogGLESFormatCaveat(logicalFormat, renderbufferTargetIndex, outerFallbackInfo);
LogGLESFormatCaveat(logicalFormat, renderbufferTargetIndex, fallbackInfo);
}
const Int maxSamples =
GetGLESFormatMaxSamples(capabilities, logicalFormat, outerFallbackInfo.ImageFormat);
GetGLESFormatMaxSamples(capabilities, logicalFormat, fallbackInfo.ImageFormat);
cache.SampleCounts[renderbufferTargetIndex][formatIndex] =
ProbeRenderbufferSampleCounts(gl, outerFallbackInfo.InternalFormat, logicalFormat, maxSamples);
ProbeRenderbufferSampleCounts(gl, fallbackInfo.InternalFormat, logicalFormat, maxSamples);
}
}
}
@@ -676,7 +603,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
.ExtraVendor = Nullopt, // Extra vendor
.RendererGLInfo =
{
.TargetGLVersion = {3, 3, 0}, // GL target version
.TargetGLVersion = {3, 3, 0}, // Target OpenGL Version
.TargetGLSLVersion = {4, 6, 0}, // Target Shading Language Version
// Baseline advertisement (no timer queries / anisotropy yet); reconciled
// once the ES capabilities exist, see UpdateAdvertisedCapabilityExtensions.
@@ -904,17 +831,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
E_GL_ARB_multi_draw_indirect, E_GL_ARB_indirect_parameters,
E_GL_ARB_shader_draw_parameters, E_GL_ARB_gpu_shader5, E_GL_ARB_multi_bind,
E_GL_ARB_shading_language_420pack, E_GL_ARB_vertex_attrib_binding,
// Both are core from GL 3.2/3.3 on and implemented here for
// every advertised version, but an app targeting 3.0/3.1
// only reaches them through the extension string - the CTS
// picks a whole different shader for draw_buffers without
// explicit_attrib_location. DirectVulkan advertises both.
E_GL_ARB_explicit_attrib_location, E_GL_ARB_texture_multisample,
E_GL_ARB_shader_image_size,
// Advertised with GL_NUM_PROGRAM_BINARY_FORMATS = 0, which the
// extension explicitly permits. It is also the only thing that
// exposes glProgramParameteri before GL 4.1.
E_GL_ARB_get_program_binary};
E_GL_ARB_shader_image_size};
// Only advertised when the device driver actually has usable timer queries
// (GL_EXT_disjoint_timer_query plus its entry points) and the
// MOBILEGL_DISABLE_TIMERQUERY escape hatch is off.
@@ -1017,30 +934,11 @@ namespace MobileGL::MG_Backend::DirectGLES {
funcsTable.GL.BeginTimeElapsedQuery = BeginTimeElapsedQuery;
funcsTable.GL.EndTimeElapsedQuery = EndTimeElapsedQuery;
funcsTable.GL.QueryCounterTimestamp = QueryCounterTimestamp;
funcsTable.GL.IsQueryResultAvailable = IsQueryResultAvailable;
funcsTable.GL.GetQueryResult64 = GetQueryResult64;
funcsTable.GL.DeleteBackendQuery = DeleteBackendQuery;
funcsTable.GL.GetGpuTimestampNs = GetGpuTimestampNs;
}
// Occlusion queries are core ES3 (independent of MOBILEGL_DISABLE_TIMERQUERY)
// and share the handle-based result/delete entries, which must exist even
// when the timer-query group above is disabled.
funcsTable.GL.BeginOcclusionQuery = BeginOcclusionQuery;
funcsTable.GL.EndOcclusionQuery = EndOcclusionQuery;
// Real driver primitive counters: the frontend's CPU accounting cannot see a
// geometry shader's amplification.
funcsTable.GL.BeginXfbPrimitivesQuery = BeginXfbPrimitivesQuery;
funcsTable.GL.EndXfbPrimitivesQuery = EndXfbPrimitivesQuery;
funcsTable.GL.IsQueryResultAvailable = IsQueryResultAvailable;
funcsTable.GL.GetQueryResult64 = GetQueryResult64;
funcsTable.GL.DeleteBackendQuery = DeleteBackendQuery;
// Transform feedback is captured by the real ES driver rather than
// reconstructed from the draw recording, so the frontend has to hand the
// span boundaries over.
funcsTable.GL.PatchParameteri = DirectGLES::PatchParameteri;
funcsTable.GL.BeginTransformFeedback = XfbImpl::BeginTransformFeedback;
funcsTable.GL.EndTransformFeedback = XfbImpl::EndTransformFeedback;
funcsTable.GL.PauseTransformFeedback = XfbImpl::PauseTransformFeedback;
funcsTable.GL.ResumeTransformFeedback = XfbImpl::ResumeTransformFeedback;
funcsTable.GL.BindTransformFeedback = XfbImpl::BindTransformFeedback;
funcsTable.GL.DeleteTransformFeedback = XfbImpl::DeleteTransformFeedback;
funcsTableInitialized = true;
}
return funcsTable;
@@ -1081,10 +979,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
m_dynamicParameters.MaxIntegerSamples = m_GLESCapabilities.MaxIntegerSamples;
m_dynamicParameters.MaxSamples = m_GLESCapabilities.MaxSamples;
m_dynamicParameters.MaxSampleMaskWords = m_GLESCapabilities.MaxSampleMaskWords;
m_dynamicParameters.MaxPatchVertices = m_GLESCapabilities.MaxPatchVertices;
m_dynamicParameters.MaxTessGenLevel = m_GLESCapabilities.MaxTessGenLevel;
m_dynamicParameters.MinProgramTextureGatherOffset = m_GLESCapabilities.MinProgramTextureGatherOffset;
m_dynamicParameters.MaxProgramTextureGatherOffset = m_GLESCapabilities.MaxProgramTextureGatherOffset;
// Clamp the advertised sampler limits the same way the DirectVulkan backend does: per-stage
// GL_MAX_TEXTURE_IMAGE_UNITS must never exceed host-side fixed arrays sized off it (e.g.
// Minecraft's 128-entry Blaze3D GlStateManager.TEXTURES[], iterated by Iris), and the combined
@@ -1112,7 +1006,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
m_dynamicParameters.MaxComputeWorkGroupInvocations = m_GLESCapabilities.MaxComputeWorkGroupInvocations;
m_dynamicParameters.MaxShaderStorageBufferBindings = m_GLESCapabilities.MaxShaderStorageBufferBindings;
m_dynamicParameters.MaxTextureBufferSize = m_GLESCapabilities.MaxTextureBufferSize;
m_dynamicParameters.TextureBufferOffsetAlignment = m_GLESCapabilities.TextureBufferOffsetAlignment;
m_dynamicParameters.MaxUniformBufferBindings = m_GLESCapabilities.MaxUniformBufferBindings;
m_dynamicParameters.MaxUniformBlockSize = m_GLESCapabilities.MaxUniformBlockSize;
const Int maxSupportedTextureUnits =
@@ -1132,8 +1025,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
clampStageImageUniforms(m_GLESCapabilities.MaxFragmentImageUniforms);
m_dynamicParameters.MaxComputeImageUniforms =
clampStageImageUniforms(m_GLESCapabilities.MaxComputeImageUniforms);
m_dynamicParameters.SupportsDistinctDepthStencilAttachments =
ProbeDistinctDepthStencilAttachments(DirectGLES::g_GLESFuncs);
m_dynamicParameters.MaxDrawBuffers = m_GLESCapabilities.MaxDrawBuffers;
m_dynamicParameters.MaxColorAttachments = m_GLESCapabilities.MaxColorAttachments;
m_dynamicParameters.MaxClipDistances = m_GLESCapabilities.MaxClipDistances;
@@ -1143,24 +1034,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
m_dynamicParameters.ViewportBoundsRangeMin = m_GLESCapabilities.ViewportBoundsRangeMin;
m_dynamicParameters.ViewportBoundsRangeMax = m_GLESCapabilities.ViewportBoundsRangeMax;
m_dynamicParameters.ViewportSubpixelBits = m_GLESCapabilities.ViewportSubpixelBits;
m_dynamicParameters.MinFragmentInterpolationOffset =
std::isfinite(m_GLESCapabilities.MinFragmentInterpolationOffset) &&
m_GLESCapabilities.MinFragmentInterpolationOffset <= -0.5f
? m_GLESCapabilities.MinFragmentInterpolationOffset
: -0.5f;
m_dynamicParameters.MaxFragmentInterpolationOffset = 0.4375f;
m_dynamicParameters.FragmentInterpolationOffsetBits = 4;
if (m_GLESCapabilities.FragmentInterpolationOffsetBits >= 4 &&
std::isfinite(m_GLESCapabilities.MaxFragmentInterpolationOffset)) {
const Float requiredMaxOffset =
0.5f - std::ldexp(1.0f, -m_GLESCapabilities.FragmentInterpolationOffsetBits);
if (m_GLESCapabilities.MaxFragmentInterpolationOffset >= requiredMaxOffset) {
m_dynamicParameters.MaxFragmentInterpolationOffset =
m_GLESCapabilities.MaxFragmentInterpolationOffset;
m_dynamicParameters.FragmentInterpolationOffsetBits =
m_GLESCapabilities.FragmentInterpolationOffsetBits;
}
}
m_dynamicParameters.SupportsWideLines =
m_GLESCapabilities.AliasedLineWidthRangeMax > 1.0f || m_GLESCapabilities.SmoothLineWidthRangeMax > 1.0f;
File diff suppressed because it is too large Load Diff
@@ -130,16 +130,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
BackendQueryHandle BeginTimeElapsedQuery();
void EndTimeElapsedQuery(BackendQueryHandle query);
BackendQueryHandle QueryCounterTimestamp();
// GL_ANY_SAMPLES_PASSED(_CONSERVATIVE) occlusion queries: core ES3, independent of
// GL_EXT_disjoint_timer_query and of MOBILEGL_DISABLE_TIMERQUERY. Results/deletion
// flow through GetQueryResult64/DeleteBackendQuery like the timer queries above.
BackendQueryHandle BeginOcclusionQuery();
void EndOcclusionQuery(BackendQueryHandle query);
// GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN / GL_PRIMITIVES_GENERATED, also core ES
// (GL_PRIMITIVES_GENERATED from ES 3.2 on). Null when the target is unavailable, in
// which case the frontend falls back to counting primitives from the draw calls.
BackendQueryHandle BeginXfbPrimitivesQuery(Bool generated);
void EndXfbPrimitivesQuery(BackendQueryHandle query);
Bool IsQueryResultAvailable(BackendQueryHandle query);
// Returns true when a final value landed in *outNanoseconds (a zero for
// null or stale-generation handles IS final: the frontend may cache it
@@ -164,24 +154,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
void SetGLESCapabilities(const MG_External::GLESCapabilities& capabilities);
void DestroyEGLContext();
// Transform feedback capture spans, performed by the real ES driver. The
// capture set is declared on the backend program at link time; the driver-side
// begin is deferred to the first draw of the span (ES needs the capturing
// program current and the capture buffers bound), and the end also mirrors the
// captured bytes back into the frontend buffer shadows.
void PatchParameteri(GLenum pname, GLint value);
namespace XfbImpl {
Bool AreTransformFeedbacksSupported();
void BeginTransformFeedback(GLenum primitiveMode);
void EndTransformFeedback();
void PauseTransformFeedback();
void ResumeTransformFeedback();
void BindTransformFeedback(GLuint name);
void DeleteTransformFeedback(GLuint name);
void OnBackendContextDestroyed();
} // namespace XfbImpl
extern MG_External::EGLFunctionsTable g_EGLFuncs;
extern MG_External::GLESFunctionsTable g_GLESFuncs;
extern MG_External::GLESCapabilities g_GLESCapabilities;
+110 -138
View File
@@ -630,32 +630,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
resource->syncedChangeSerial = bufferObject.GetChangeSerial();
}
// A shader wrote this buffer through a storage/atomic-counter binding, so the ES
// driver's copy is ahead of the frontend shadow. Pull the whole thing back so
// MapBuffer/GetBufferSubData/CopyBufferSubData see the real results.
void Ops_ReadbackFromGpu(BufferObject& bufferObject) {
auto* resource = ResourceOf(bufferObject);
if (!resource || resource->id == 0 || !resource->storageInitialized) return;
if (resource->persistentMapped) return; // shadow already IS the GPU storage
if (!CanTouchGLNow() || resource->contextGeneration != g_bufferContextGeneration) return;
if (!g_GLESFuncs.glMapBufferRange || !g_GLESFuncs.glUnmapBuffer) return;
const SizeT size = std::min<SizeT>(bufferObject.GetSize(), resource->storageSize);
if (size == 0) return;
BindBufferId(TempBufferTarget, resource->id);
void* mapped = g_GLESFuncs.glMapBufferRange(TempBufferTarget, 0, static_cast<GLsizeiptr>(size),
GL_MAP_READ_BIT);
if (mapped == nullptr) {
MGLOG_E("Ops_ReadbackFromGpu: glMapBufferRange(read) failed for buffer %u", resource->id);
return;
}
bufferObject.WritebackFromBackend({mapped, size}, 0);
g_GLESFuncs.glUnmapBuffer(TempBufferTarget);
// The shadow now matches the backend byte for byte; without this the next
// draw would see a newer change serial and re-upload the readback over it.
resource->syncedChangeSerial = bufferObject.GetChangeSerial();
}
void Ops_OnDestroy(SharedPtr<BackendBufferResource>&& resource) {
if (!resource) return;
auto* glesResource = static_cast<GLESBufferResource*>(resource.get());
@@ -688,7 +662,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
.FlushMappedRange = Ops_FlushMappedRange,
.OnDestroy = Ops_OnDestroy,
.AcquirePersistentMap = Ops_AcquirePersistentMap,
.ReadbackFromGpu = Ops_ReadbackFromGpu,
};
} // namespace
@@ -1289,15 +1262,31 @@ namespace MobileGL::MG_Backend::DirectGLES {
const auto& allAttributes = stateVAOObject->GetAllAttributes();
for (Uint attribIndex = 0; attribIndex < allAttributes.size(); ++attribIndex) {
const auto& attrib = allAttributes[attribIndex];
const Uint32 attribBit = 1u << attribIndex;
// An enabled attrib with neither a buffer object nor a client pointer has no
// source; GL tolerates the state (only draws consuming it are undefined), but
// Adreno's ES driver memcpys the "client array" from address 0 at draw time
// (SIGSEGV). Keep such attribs disabled on the backend VAO and re-enable them
// the moment they gain a source - the mask-vs-current compare below triggers
// the enable even when only the Buffer/Format versions changed.
const Bool unsourceable = attrib.Enabled && !attrib.Buffer && attrib.Offset == 0;
const Bool wasForceDisabled = (m_forceDisabledAttribsMask & attribBit) != 0;
Bool needsSyncSwitch = allAttributeVersions[attribIndex].SwitchVersion !=
m_syncedAttributeVersions[attribIndex].SwitchVersion;
if (needsSyncSwitch) {
if (attrib.Enabled) {
if (needsSyncSwitch || unsourceable != wasForceDisabled) {
if (attrib.Enabled && !unsourceable) {
g_GLESFuncs.glEnableVertexAttribArray(attribIndex);
} else {
g_GLESFuncs.glDisableVertexAttribArray(attribIndex);
}
}
if (unsourceable) {
m_forceDisabledAttribsMask |= attribBit;
} else {
m_forceDisabledAttribsMask &= ~attribBit;
}
Bool needsSyncFormat = allAttributeVersions[attribIndex].FormatVersion !=
m_syncedAttributeVersions[attribIndex].FormatVersion;
@@ -1305,7 +1294,17 @@ namespace MobileGL::MG_Backend::DirectGLES {
m_syncedAttributeVersions[attribIndex].BufferVersion;
if (!needsSyncFormat && !needsSyncBuffer) continue;
if (unsourceable) continue;
// Client-side array with a non-null pointer: the pointer is uploaded and applied
// per draw by SyncClientSideAttributesForDrawArrays.
if (!attrib.Buffer) continue;
if (!BindAttributeBuffer(attrib)) {
if (attrib.Enabled) {
g_GLESFuncs.glDisableVertexAttribArray(attribIndex);
m_forceDisabledAttribsMask |= attribBit;
}
continue;
}
@@ -2232,24 +2231,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
"ID: %u, buffer ID: %u, buffer size: %zu, format: %s",
m_backendTextureId, backendId, buffer->GetSize(),
MG_Util::ConvertGLEnumToString(glInternalFormat).c_str());
// A texture that names a window of the buffer needs the range form; the
// whole-buffer forms report offset 0 and the buffer's current size, which
// glTexBuffer expresses more directly (and works where the range entry point
// is absent).
const SizeT rangeOffset = textureBufferObject->GetBufferRangeOffset();
const SizeT rangeSize = textureBufferObject->GetBufferRangeSizeInBytes();
if (rangeOffset == 0 && rangeSize == buffer->GetSize()) {
g_GLESFuncs.glTexBuffer(GL_TEXTURE_BUFFER, glInternalFormat, backendId);
} else if (g_GLESFuncs.glTexBufferRange != nullptr) {
g_GLESFuncs.glTexBufferRange(GL_TEXTURE_BUFFER, glInternalFormat, backendId,
static_cast<GLintptr>(rangeOffset),
static_cast<GLsizeiptr>(rangeSize));
} else {
MGLOG_E("Texture buffer %u names a sub-range but the driver has no "
"glTexBufferRange; binding the whole buffer instead",
stateTextureObject->GetExternalIndex());
g_GLESFuncs.glTexBuffer(GL_TEXTURE_BUFFER, glInternalFormat, backendId);
}
g_GLESFuncs.glTexBuffer(GL_TEXTURE_BUFFER, glInternalFormat, backendId);
DebugImpl::ErrorLopper::Loop(
[file = __FILE__, line = __LINE__, func = __func__, glInternalFormat, backendId](GLenum err) {
MGLOG_D("%s(%s:%d) glTexBuffer(format=%s, buffer=%u) ES error: %s",
@@ -2449,19 +2431,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
MGLOG_D("%s(%s:%d) ES error %s", func, file, line, MG_Util::ConvertGLEnumToString(err).c_str());
});
// A three-channel format widened to four for a multisample target (see
// NormalizePixelFormat) gains an alpha channel the frontend format does not have, and
// whatever the draw that filled it wrote there is not what GL would report: a format
// without alpha reads back as 1.0. Answer the ALPHA swizzle source with ONE so the
// promotion stays invisible, composed with the swizzle the application asked for.
Vec4<TextureSwizzleParam> swizzleParams = stateTextureObject->GetAllSwizzleParams();
if (TextureImpl::BackendTextureFormatAddsAlpha(stateTextureObject->GetFormat(), targetInternal)) {
for (SizeT channel = 0; channel < 4; ++channel) {
if (swizzleParams[channel] == TextureSwizzleParam::Alpha) {
swizzleParams[channel] = TextureSwizzleParam::One;
}
}
}
const auto& swizzleParams = stateTextureObject->GetAllSwizzleParams();
if (swizzleParams != m_cacheSwizzleParams) {
#define SYNC_TEX_SWIZZLE_PARAM_IF_CHANGED(func, glEnum) \
if (m_cacheSwizzleParams.func != swizzleParams.func) { \
@@ -2735,31 +2705,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
return false;
}
Bool IsFixedPointFallbackReadAttachment() {
const auto& readFBO =
MG_State::pGLContext->GetFramebufferBindingSlot(FramebufferTarget::Read).GetBoundObject();
if (!readFBO) {
return false;
}
const auto readBuffer = readFBO->GetReadBuffer();
if (readBuffer < FramebufferAttachmentType::Color0 || readBuffer > FramebufferAttachmentType::Color31) {
return false;
}
// Any signed-normalized attachment, not just the ones currently substituted:
// ES has no GL_CLAMP_READ_COLOR at all, so even a natively stored SNORM buffer
// hands back the negative half that desktop GL clamps away.
const auto& attachmentObject = readFBO->GetAttachment(readBuffer);
if (attachmentObject.IsTexture()) {
const auto& textureObject = attachmentObject.GetTexture();
return textureObject && IsSnormFormat(textureObject->GetFormat());
}
if (attachmentObject.IsRenderbuffer()) {
const auto& renderbufferObject = attachmentObject.GetRenderbuffer();
return renderbufferObject && IsSnormFormat(renderbufferObject->GetInternalFormat());
}
return false;
}
void BackendFramebufferObject::SyncReadBufferToBackend(
const SharedPtr<MG_State::GLState::FramebufferObject>& stateFBOObject) {
if (!stateFBOObject) {
@@ -3262,7 +3207,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
namespace PrgramImpl {
Uint32 g_snormFallbackClampOutputMask = 0;
Uint g_fragColorBroadcastCount = 1;
Uint32 g_unormFallbackClampOutputMask = 0;
Uint g_lastUsedBackendProgramId = 0;
StateBackendObjectRegistry<MG_State::GLState::ProgramObject, BackendProgramObjectImpl> g_backendProgramObjects;
@@ -3314,10 +3258,8 @@ namespace MobileGL::MG_Backend::DirectGLES {
MGLOG_D("Syncing program to backend. State program ID: %u, Backend ID: %u",
stateProgramObject->GetExternalIndex(), m_backendProgramId);
m_backendProgramUsable = true;
m_snormFallbackClampOutputMask = g_snormFallbackClampOutputMask;
m_unormFallbackClampOutputMask = g_unormFallbackClampOutputMask;
m_fragColorBroadcastCount = g_fragColorBroadcastCount;
// Detach all existing shaders
GLint attachedCount = 0;
@@ -3349,6 +3291,19 @@ namespace MobileGL::MG_Backend::DirectGLES {
}
auto& shaderSpirvs = stateProgramObject->GetGeneratedSpirv();
// Adreno's ESSL compiler mishandles gl_ClipDistance (rejects redeclarations,
// miscompiles non-constant-index writes and constant-index gl_in element reads,
// crashes on whole-array gl_in reads) and cannot dynamically index the global
// const struct[] LUTs SPIRV-Cross likes to emit. Gate the workarounds to
// Qualcomm; MOBILEGL_QUIRK_CLIP_DISTANCE overrides the device detection.
const MG_Config::QuirkOverride clipDistanceQuirkOverride =
MG_Config::Features.ClipDistanceQuirk;
const Bool applyClipDistanceQuirk =
clipDistanceQuirkOverride == MG_Config::QuirkOverride::ForceOn ||
(clipDistanceQuirkOverride == MG_Config::QuirkOverride::Auto &&
pActiveBackendObject &&
pActiveBackendObject->GetDynamicParameters().GpuVendor == GpuVendorKind::Qualcomm);
for (int index = 0; index < attachedShaders.size(); ++index) {
auto& shader = attachedShaders[index];
GLenum glShaderType = MG_Util::ConvertShaderStageToGLEnum(shader->GetShaderStage());
@@ -3371,6 +3326,63 @@ namespace MobileGL::MG_Backend::DirectGLES {
effectiveSpirv = &loweredSpirv;
}
// GL 3.3 only promises undefined *values* for out-of-bounds array indexing, but
// Adreno's ESSL compiler constant-folds a provably out-of-bounds local-array
// index into poison that corrupts the whole shader's output. Clamp every
// access-chain index to its declared bounds before transpiling.
Vector<unsigned int> clampedSpirv;
if (MG_Util::ShaderTranspiler::ShaderCompiler::ClampAccessChainIndicesForEssl(*effectiveSpirv,
clampedSpirv) &&
!clampedSpirv.empty()) {
effectiveSpirv = &clampedSpirv;
} else {
MGLOG_W("ClampAccessChainIndicesForEssl failed, continuing with unclamped SPIR-V.");
}
// SPIRV-Cross emulates 1D samplers as 2D for ES: it widens texelFetch coordinates
// to ivec2 but keeps the ConstOffset operand scalar, which is not a valid ESSL
// texelFetchOffset overload (Adreno rejects it). Fold the constant offset into the
// coordinate instead (texelFetchOffset(t,P,l,o) == texelFetch(t,P+o,l)).
Vector<unsigned int> foldedOffsetSpirv;
if (MG_Util::ShaderTranspiler::ShaderCompiler::FoldConstOffsetFor1DFetchForEssl(
*effectiveSpirv, foldedOffsetSpirv) &&
!foldedOffsetSpirv.empty()) {
effectiveSpirv = &foldedOffsetSpirv;
} else {
MGLOG_W("FoldConstOffsetFor1DFetchForEssl failed, continuing with unfolded SPIR-V.");
}
// Adreno quirk: shadow gl_ClipDistance in Private arrays so the transpiled
// ESSL only writes the builtin with literal constant indices (flushed before
// EmitVertex/return) and only reads gl_in clip distances through dynamic loop
// indices - the shapes this driver compiles correctly. Must run after the
// access-chain clamp above so the flush indices stay literal constants.
Vector<unsigned int> clipDistanceSpirv;
if (applyClipDistanceQuirk &&
(glShaderType == GL_VERTEX_SHADER || glShaderType == GL_GEOMETRY_SHADER)) {
if (MG_Util::ShaderTranspiler::ShaderCompiler::LowerClipDistanceForEssl(
*effectiveSpirv, clipDistanceSpirv) &&
!clipDistanceSpirv.empty()) {
effectiveSpirv = &clipDistanceSpirv;
} else {
MGLOG_W("LowerClipDistanceForEssl failed, continuing with unlowered SPIR-V.");
}
}
// Adreno quirk: split single constant-composite stores of struct arrays so
// SPIRV-Cross does not promote them to global const struct[] LUTs, which this
// driver cannot dynamically index ("Cannot offset into the structure").
Vector<unsigned int> structLutSpirv;
if (applyClipDistanceQuirk) {
if (MG_Util::ShaderTranspiler::ShaderCompiler::DefeatConstStructArrayLutForEssl(
*effectiveSpirv, structLutSpirv) &&
!structLutSpirv.empty()) {
effectiveSpirv = &structLutSpirv;
} else {
MGLOG_W("DefeatConstStructArrayLutForEssl failed, continuing with unsplit SPIR-V.");
}
}
// ESSL stage-matches uniform blocks by member precision, but SPIRV-Cross prints
// a RelaxedPrecision member as explicit "mediump" in the vertex stage and as
// UNQUALIFIED (mediump-by-default) in the fragment stage; after
@@ -3400,16 +3412,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
effectiveSpirv = &noperspectiveSpirv;
}
// ES has no rectangle sampler, and SPIRV-Cross refuses the whole module rather
// than approximating one. The shared pass turns the type into the 2D one and
// divides the coordinate of every normalized-coordinate lookup by the texture
// size, which is the whole of the difference between the two.
Vector<unsigned int> rectLoweredSpirv;
if (MG_Util::ShaderTranspiler::ShaderCompiler::LowerRectImages(*effectiveSpirv, rectLoweredSpirv) &&
!rectLoweredSpirv.empty()) {
effectiveSpirv = &rectLoweredSpirv;
}
MG_Util::ShaderTranspiler::SpvcSession spvcSession(*effectiveSpirv,
MG_Util::ShaderTranspiler::SessionUsageBit::Transpile);
@@ -3432,7 +3434,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
r.log += spvcSession.GetLastErrorString();
r.errc = -5;
MGLOG_E("%s", r.log.c_str());
m_backendProgramUsable = false;
continue;
}
@@ -3440,10 +3441,14 @@ namespace MobileGL::MG_Backend::DirectGLES {
source = RebindImageUniformsToFrontendUnits(std::move(source), stateProgramObject);
source = RemoveLayoutBinding(source);
if (applyClipDistanceQuirk) {
// Adreno rejects the gl_ClipDistance redeclaration SPIRV-Cross still emits
// ("reserved built-in name") but accepts plain usage with
// GL_EXT_clip_cull_distance required; drop the line, keep the #extension.
source = RemoveClipDistanceRedeclaration(source);
}
source = ProcessOutColorLocations(source);
source = ForceFlatIntegerVaryings(source, glShaderType);
source = BroadcastLegacyFragColor(std::move(source), glShaderType, m_fragColorBroadcastCount);
source = EmulateTextureLodBias(source);
source = EmulateBaseInstanceInVertexShader(std::move(source), glShaderType);
source = PromoteDrawParameterGlobalsToUniforms(std::move(source), glShaderType);
source = ForceSupporterOutput(source);
@@ -3474,7 +3479,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
Vector<GLchar> log(logLength);
g_GLESFuncs.glGetShaderInfoLog(backendShaderId, logLength, nullptr, log.data());
MGLOG_E("Shader compilation failed for backend ID %u: %s", backendShaderId, log.data());
m_backendProgramUsable = false;
continue;
}
@@ -3484,33 +3488,12 @@ namespace MobileGL::MG_Backend::DirectGLES {
MGLOG_D("Processed shader source length: %zu", source.length());
}
// Transform feedback capture runs on the real driver (see XfbImpl in
// DirectGLES.cpp), so the capture set has to be declared on the backend
// program before it links. SPIRV-Cross keeps user output names verbatim in
// the transpiled ESSL (`out vec4 result_0;` stays `result_0`), so the
// frontend's requested names carry over unchanged.
if (stateProgramObject->GetTransformFeedbackVaryingCount() > 0 &&
g_GLESFuncs.glTransformFeedbackVaryings != nullptr) {
const auto& xfbVaryings = stateProgramObject->GetTransformFeedbackVaryings();
Vector<const GLchar*> xfbNames;
xfbNames.reserve(xfbVaryings.size());
for (const auto& xfbVarying : xfbVaryings) {
xfbNames.push_back(xfbVarying.name.c_str());
}
MGLOG_D("Declaring %zu transform feedback varyings on program %u", xfbNames.size(),
m_backendProgramId);
g_GLESFuncs.glTransformFeedbackVaryings(m_backendProgramId, static_cast<GLsizei>(xfbNames.size()),
xfbNames.data(),
stateProgramObject->GetTransformFeedbackBufferMode());
}
// Link program
MGLOG_D("Linking program %u", m_backendProgramId);
g_GLESFuncs.glLinkProgram(m_backendProgramId);
GLint linkStatus;
g_GLESFuncs.glGetProgramiv(m_backendProgramId, GL_LINK_STATUS, &linkStatus);
m_backendProgramUsable = m_backendProgramUsable && linkStatus == GL_TRUE;
if (linkStatus != GL_TRUE) {
GLint logLength;
g_GLESFuncs.glGetProgramiv(m_backendProgramId, GL_INFO_LOG_LENGTH, &logLength);
@@ -3623,11 +3606,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
binding.backendLocation = backendLoc;
binding.uniformType = uniformType;
binding.lastAssignedUnit = -1;
// Present only for the samplers EmulateTextureLodBias actually rewrote; the
// pass names it after the sampler, which SPIRV-Cross preserves verbatim.
binding.lodBiasLocation =
g_GLESFuncs.glGetUniformLocation(m_backendProgramId, (String(LOD_BIAS_UNIFORM_PREFIX) + name).c_str());
binding.lastAssignedLodBias = 0.0f;
m_samplerUniformBindings.push_back(binding);
}
}
@@ -3636,18 +3614,12 @@ namespace MobileGL::MG_Backend::DirectGLES {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
#endif
// glUseProgram on a program that did not link is an INVALID_OPERATION and
// leaves the *previous* program current, so the draw would silently render
// with an unrelated shader (KHR-GL3x.texture_size_promotion read another
// test case's alpha that way once a sampler2DRect stage failed to
// transpile). Bind nothing instead: the draw is then a visible no-op.
const Uint programToBind = m_backendProgramUsable ? m_backendProgramId : 0;
if (g_lastUsedBackendProgramId == programToBind) {
if (g_lastUsedBackendProgramId == m_backendProgramId) {
return;
}
MGLOG_D("Using program %u", programToBind);
g_GLESFuncs.glUseProgram(programToBind);
g_lastUsedBackendProgramId = programToBind;
MGLOG_D("Using program %u", m_backendProgramId);
g_GLESFuncs.glUseProgram(m_backendProgramId);
g_lastUsedBackendProgramId = m_backendProgramId;
}
void BackendProgramObjectImpl::SetBaseInstance(Uint32 baseInstance) const {
+12 -37
View File
@@ -258,6 +258,11 @@ namespace MobileGL::MG_Backend::DirectGLES {
private:
Uint m_backendVAOId = 0;
Array<Uint, MG_State::GLState::VertexArrayObject::MAX_VERTEX_ATTRIBS> m_clientAttributeBufferIds;
// Attribs the frontend has Enabled but that have no source at all (no buffer object
// and NULL client pointer). GL keeps such attribs latently enabled, but Adreno's ES
// driver treats them as client arrays and memcpys from address 0 at draw time
// (SIGSEGV), so they are kept disabled on the backend VAO until they gain a source.
Uint32 m_forceDisabledAttribsMask = 0;
Bool m_isInitialized = false;
Uint16 m_syncedIndexBufferVersion = 0;
Array<MG_State::GLState::VertexAttributeVersion, MG_State::GLState::VertexArrayObject::MAX_VERTEX_ATTRIBS>
@@ -270,21 +275,18 @@ namespace MobileGL::MG_Backend::DirectGLES {
namespace TextureImpl {
inline Bool IsSupportedTextureTarget(TextureTarget target) {
// Every desktop-only target is stored on an ES one; see MapToBackendTextureTarget.
(void)target;
return true;
// Rectangle textures need non-normalized sampling ES cannot express; everything else is
// either native or emulated (1D -> 2D with height 1, 1D array -> 2D array, see
// MapToBackendTextureTarget). SPIRV-Cross already emits the matching ESSL samplers and
// coordinate padding for 1D/1D-array shaders.
return target != TextureTarget::TextureRectangle;
}
// ES has none of the desktop-only targets: 1D textures are stored as 2D (height 1), 1D
// arrays as 2D arrays (height 1, layers in depth), and rectangle textures as plain 2D -
// they are single-level and already clamp, so only the non-normalized coordinates differ.
// Must match the shader-side emulation: SPIRV-Cross handles 1D/1D-array itself, and
// ShaderCompiler::LowerRectImages rewrites rectangle images (declining any module
// whose lookups are not integer-coordinate, which SPIRV-Cross then still rejects).
// ES has no 1D targets: 1D textures are stored as 2D (height 1) and 1D arrays as 2D arrays
// (height 1, layers in depth). Must match SPIRV-Cross's ES 1D-as-2D shader emulation.
inline TextureTarget MapToBackendTextureTarget(TextureTarget target) {
switch (target) {
case TextureTarget::Texture1D:
case TextureTarget::TextureRectangle:
return TextureTarget::Texture2D;
case TextureTarget::Texture1DArray:
return TextureTarget::Texture2DArray;
@@ -300,7 +302,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
inline GLenum ConvertTextureUploadTargetToBackendGLEnum(TextureUploadTarget uploadTarget) {
switch (uploadTarget) {
case TextureUploadTarget::Texture1D:
case TextureUploadTarget::TextureRectangle:
return GL_TEXTURE_2D;
case TextureUploadTarget::Texture1DArray:
return GL_TEXTURE_2D_ARRAY;
@@ -442,13 +443,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
extern StateBackendObjectRegistry<MG_State::GLState::FramebufferObject, BackendFramebufferObject>
g_backendFramebufferObjects;
// True when the read buffer names a fixed-point (norm/snorm) attachment that the
// backend actually stores in a floating-point format. GL clamps a read from a
// fixed-point colour buffer to [0,1] (GL_CLAMP_READ_COLOR defaults to
// GL_FIXED_ONLY); the substituted float storage would not, so the readback path
// has to apply the clamp itself.
Bool IsFixedPointFallbackReadAttachment();
extern Array<Uint16, SizeT(FramebufferTarget::FramebufferTargetCount)> g_fboBindVersions;
// Tracks the bound FBO's object version (bumped on any attachment/drawbuffer change)
// per target: re-attaching textures or changing draw buffers on an already-bound FBO
@@ -587,12 +581,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
Int backendLocation = -1;
GLenum uniformType = 0;
Int lastAssignedUnit = -1;
// Location of this sampler's emulated GL_TEXTURE_LOD_BIAS uniform
// (PrgramImpl::EmulateTextureLodBias), -1 when the shader has none.
// lastAssignedLodBias mirrors the value the program currently holds,
// so an unbiased shader issues no per-draw glUniform1f at all.
Int lodBiasLocation = -1;
Float lastAssignedLodBias = 0.0f;
};
BackendProgramObjectImpl();
@@ -604,14 +592,9 @@ namespace MobileGL::MG_Backend::DirectGLES {
void SetDrawID(Uint32 drawId) const;
Int GetIndirectParamsBinding() const { return m_indirectParamsBinding; }
Uint GetBackendProgramId() const { return m_backendProgramId; }
// False when the last SyncToBackend could not produce a usable program (a
// shader failed to transpile or compile, or the link itself failed). Use()
// must not leave the previously bound program current in that case.
Bool IsBackendProgramUsable() const { return m_backendProgramUsable; }
Uint GetBackendGlobalUBOId() const { return m_backendGlobalUBOId; }
Uint32 GetSnormFallbackClampOutputMask() const { return m_snormFallbackClampOutputMask; }
Uint32 GetUnormFallbackClampOutputMask() const { return m_unormFallbackClampOutputMask; }
Uint GetFragColorBroadcastCount() const { return m_fragColorBroadcastCount; }
Bool HasGlobalUboBlock() const { return m_globalUboBackendBlockIndex >= 0; }
const Vector<Int>& GetUniformBlockBackendIndices() const { return m_uniformBlockBackendIndices; }
@@ -638,11 +621,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
Int m_indirectParamsBinding = -1;
Uint32 m_snormFallbackClampOutputMask = 0;
Uint32 m_unormFallbackClampOutputMask = 0;
// Draw buffers a legacy gl_FragColor write has to reach (see
// PrgramImpl::BroadcastLegacyFragColor); 1 keeps the plain single-output shader.
Uint m_fragColorBroadcastCount = 1;
Bool m_isInitialized = false;
Bool m_backendProgramUsable = false;
Int m_globalUboBackendBlockIndex = -1;
Int m_globalUboBackendBlockSize = 0;
@@ -655,10 +634,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
extern Uint32 g_snormFallbackClampOutputMask;
extern Uint32 g_unormFallbackClampOutputMask;
// Draw buffers the current draw framebuffer enables. Like the clamp masks above it
// is framebuffer state that the shader has to be compiled against, so a program
// whose snapshot no longer matches is relinked.
extern Uint g_fragColorBroadcastCount;
// Backend id of the last glUseProgram issued through this backend; lets Use()
// skip redundant rebinds. Reset to 0 wherever glUseProgram(0) is issued or the
// ES context is recreated.
+30 -250
View File
@@ -22,9 +22,6 @@
#include <MG_Util/Math/SmallFloat.h>
#include <cmath>
#include <cctype>
#include <cstring>
#include <regex>
namespace MobileGL::MG_Backend::DirectGLES {
namespace {
@@ -48,40 +45,15 @@ namespace MobileGL::MG_Backend::DirectGLES {
return options;
}
Flags<PixelFormatNormalizeOptionBit>
GetRuntimeFallbackNormalizeOptions(GLenum requestedInternalFormat,
Flags<PixelFormatNormalizeOptionBit> extraOptions) {
Flags<PixelFormatNormalizeOptionBit> GetRuntimeFallbackNormalizeOptions(GLenum requestedInternalFormat) {
using namespace MG_Util::TextureFormatProcessor;
const Flags<PixelFormatNormalizeOptionBit> forcedOptions = GetApplicablePixelFormatNormalizeOptions(
requestedInternalFormat, GetForcedPixelFormatNormalizeOptions() | extraOptions);
const Flags<PixelFormatNormalizeOptionBit> forcedOptions =
GetApplicablePixelFormatNormalizeOptions(requestedInternalFormat, GetForcedPixelFormatNormalizeOptions());
if (forcedOptions) {
return forcedOptions;
}
return GetApplicablePixelFormatNormalizeOptions(
requestedInternalFormat, GetDriverPixelFormatNormalizeOptions() | extraOptions);
}
// Multisample textures can only ever be rendered into, never uploaded to, so a fallback
// format for them has to stay colour-renderable - a three-channel float fallback is a legal
// ES texture format but not a legal multisample storage format. Widening to four channels
// is safe here precisely because there is no transfer path that would have to expand
// three-channel client data, and the alpha the draw writes for a three-channel source is
// already the 1.0 the frontend format implies.
Bool TargetRequiresRenderableFormat(SizeT targetIndex) {
return targetIndex == static_cast<SizeT>(TextureTarget::Texture2DMultisample) ||
targetIndex == static_cast<SizeT>(TextureTarget::Texture2DMultisampleArray);
}
Flags<PixelFormatNormalizeOptionBit> GetRenderTargetNormalizeOptions(SizeT targetIndex) {
Flags<PixelFormatNormalizeOptionBit> options;
if (!TargetRequiresRenderableFormat(targetIndex)) {
return options;
}
options |= PixelFormatNormalizeOptionBit::NoThreeChannelRenderTarget;
if (!g_GLESCapabilities.SupportsRenderSnorm || !g_GLESCapabilities.SupportsNorm16Texture) {
options |= PixelFormatNormalizeOptionBit::NoSnorm16RenderTarget;
}
return options;
return GetApplicablePixelFormatNormalizeOptions(requestedInternalFormat,
GetDriverPixelFormatNormalizeOptions());
}
Bool HasCachedFormatCapability(TextureInternalFormat internalFormat,
@@ -141,8 +113,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
const GLenum requestedInternalFormat = MG_Util::ConvertTextureInternalFormatToGLEnum(internalFormat);
Flags<PixelFormatNormalizeOptionBit> options;
if (!pActiveBackendObject || ShouldUseCaveatFormat(internalFormat, targetIndex)) {
options = GetRuntimeFallbackNormalizeOptions(requestedInternalFormat,
GetRenderTargetNormalizeOptions(targetIndex));
options = GetRuntimeFallbackNormalizeOptions(requestedInternalFormat);
}
NormalizePixelFormat(requestedInternalFormat, options, outInternalFormat, outFormat, outType);
}
@@ -177,22 +148,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
Bool ShouldUseCaveatRenderbufferFormat(TextureInternalFormat internalFormat) {
return ShouldUseCaveatFormat(internalFormat, GetRenderbufferFormatCapabilityTargetIndex());
}
Bool BackendTextureFormatAddsAlpha(TextureInternalFormat internalFormat, TextureTarget target) {
const SizeT targetIndex =
target == TextureTarget::Unknown ? kFormatCapabilityTargetCount : GetFormatCapabilityTargetIndex(target);
if (!TargetRequiresRenderableFormat(targetIndex)) {
return false;
}
if (pActiveBackendObject && !ShouldUseCaveatFormat(internalFormat, targetIndex)) {
return false;
}
const GLenum requestedInternalFormat = MG_Util::ConvertTextureInternalFormatToGLEnum(internalFormat);
const Flags<PixelFormatNormalizeOptionBit> options =
GetRuntimeFallbackNormalizeOptions(requestedInternalFormat,
GetRenderTargetNormalizeOptions(targetIndex));
return static_cast<Bool>(options & PixelFormatNormalizeOptionBit::NoThreeChannelRenderTarget);
}
} // namespace TextureImpl
namespace PrgramImpl {
String ProcessOutColorLocations(const String& glslCode) {
@@ -321,55 +276,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
return glslCode;
}
String BroadcastLegacyFragColor(String glslCode, GLenum shaderType, Uint drawBufferCount) {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
#endif
// The name is the marker: ShaderSourceProcessor only emits it when the source
// wrote gl_FragColor, and such a shader can have no other output.
static const char* const kLoweredName = "mg_FragColor";
if (shaderType != GL_FRAGMENT_SHADER || drawBufferCount <= 1) {
return glslCode;
}
static const std::regex declRegex(
R"(layout\s*\(\s*location\s*=\s*0\s*\)\s*out\s+((?:lowp|mediump|highp)\s+)?vec4\s+mg_FragColor\s*;)");
std::smatch declMatch;
if (!std::regex_search(glslCode, declMatch, declRegex)) {
return glslCode;
}
const String precision = declMatch[1].matched ? declMatch[1].str() : String();
String replicaDecls;
String replicaCopies;
for (Uint location = 1; location < drawBufferCount; ++location) {
const String name = String(kLoweredName) + "_" + std::to_string(location);
replicaDecls += "\nlayout(location = " + std::to_string(location) + ") out " + precision + "vec4 " +
name + ";";
replicaCopies += "\n " + name + " = " + kLoweredName + ";";
}
static const std::regex mainRegex(R"(void\s+main\s*\([^)]*\)\s*\{)");
std::smatch mainMatch;
if (!std::regex_search(glslCode, mainMatch, mainRegex)) {
return glslCode;
}
SizeT bracePos = static_cast<SizeT>(mainMatch.position(0) + mainMatch.length(0) - 1);
Int depth = 0;
for (SizeT pos = bracePos; pos < glslCode.size(); ++pos) {
if (glslCode[pos] == '{') {
++depth;
} else if (glslCode[pos] == '}') {
--depth;
if (depth == 0) {
glslCode.insert(pos, replicaCopies + "\n");
break;
}
}
}
glslCode.insert(static_cast<SizeT>(declMatch.position(0)) + declMatch[0].str().size(), replicaDecls);
return glslCode;
}
String ForceFlatIntegerVaryings(const String& glslCode, GLenum shaderType) {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
@@ -437,166 +343,40 @@ namespace MobileGL::MG_Backend::DirectGLES {
return result;
}
namespace {
// How a lookup carries its level of detail, and how many arguments it takes
// before the optional bias.
struct LodLookupForm {
const char* name;
Int requiredArgs; // arguments before the optional bias (implicit form)
Int explicitLodArg; // index of the explicit LOD argument, -1 for implicit
};
// texelFetch* is deliberately absent: an integer fetch names its level directly
// and takes no LOD bias. textureGather has no bias either. textureGrad* derives
// the LOD from gradients and offers no argument to fold a bias into, so it is
// left alone rather than rewritten incorrectly.
constexpr LodLookupForm LOD_LOOKUP_FORMS[] = {
{"textureProjLodOffset", 0, 2}, {"textureProjOffset", 4, -1}, {"textureProjLod", 0, 2},
{"textureLodOffset", 0, 2}, {"textureOffset", 3, -1}, {"textureProj", 2, -1},
{"textureLod", 0, 2}, {"texture", 2, -1},
};
// Sampler types with no mip chain, or whose GLSL lookups have no bias overload
// at all (the array-shadow forms), so nothing can or should be folded in.
Bool IsBiasableSamplerType(const String& samplerType) {
if (samplerType.find("MS") != String::npos) return false; // multisample
if (samplerType.find("Buffer") != String::npos) return false; // texture buffer
if (samplerType.find("Rect") != String::npos) return false; // rectangle: no mips
if (samplerType == "sampler2DArrayShadow") return false;
if (samplerType == "samplerCubeArrayShadow") return false;
return true;
}
Bool IsIdentifierChar(char c) { return std::isalnum(static_cast<unsigned char>(c)) || c == '_'; }
// Byte offsets of the top-level argument separators and of the closing paren,
// starting from the '(' at openParen. Empty when the parentheses do not balance.
Vector<SizeT> SplitCallArguments(const String& code, SizeT openParen) {
Vector<SizeT> marks;
Int depth = 0;
for (SizeT i = openParen; i < code.size(); ++i) {
const char c = code[i];
if (c == '(' || c == '[') {
++depth;
} else if (c == ']') {
--depth;
} else if (c == ')') {
--depth;
if (depth == 0) {
marks.push_back(i);
return marks;
}
} else if (c == ',' && depth == 1) {
marks.push_back(i);
}
}
return {};
}
} // namespace
String EmulateTextureLodBias(const String& glslCode) {
String RemoveClipDistanceRedeclaration(const String& glslCode) {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
#endif
if (glslCode.find("sampler") == String::npos || glslCode.find("texture") == String::npos) {
return glslCode;
}
// Adreno rejects any redeclaration of gl_ClipDistance/gl_CullDistance ("reserved
// built-in name") even with GL_EXT_clip_cull_distance required, but accepts plain
// usage of the builtin. Drop the desktop-style redeclaration line SPIRV-Cross
// prints; the "#extension GL_EXT_clip_cull_distance : require" line stays.
static const std::regex redeclarationRegex(
R"(^\s*(?:out|in)\s+(?:(?:high|medium|low)p\s+)?float\s+gl_(?:Clip|Cull)Distance\[[0-9]+\];\s*$)");
// Collect the mip-capable sampler uniforms this shader declares.
static const std::regex samplerDeclRegex(
R"(uniform\s+(?:(?:highp|mediump|lowp)\s+)?([iu]?sampler[A-Za-z0-9]*)\s+([A-Za-z_][A-Za-z0-9_]*)\s*;)");
UnorderedMap<String, String> samplerNames; // name -> bias uniform name
for (std::sregex_iterator it(glslCode.begin(), glslCode.end(), samplerDeclRegex), end; it != end; ++it) {
const String samplerType = (*it)[1].str();
if (!IsBiasableSamplerType(samplerType)) continue;
const String name = (*it)[2].str();
samplerNames.emplace(name, String(LOD_BIAS_UNIFORM_PREFIX) + name);
}
if (samplerNames.empty()) {
return glslCode;
}
String result;
result.reserve(glslCode.size());
SizeT lineStart = 0;
Bool firstLine = true;
while (lineStart <= glslCode.size()) {
SizeT lineEnd = glslCode.find('\n', lineStart);
const Bool lastLine = lineEnd == String::npos;
String line = glslCode.substr(lineStart, lastLine ? String::npos : lineEnd - lineStart);
// Rewrite the lookups. Right-to-left so earlier offsets stay valid, and only for
// samplers named directly as the first argument (SPIRV-Cross never produces an
// expression there for ES output, which has no separate sampler objects).
String result = glslCode;
Vector<String> usedSamplers;
for (SizeT scan = result.size(); scan-- > 0;) {
if (result[scan] != 't') continue;
if (scan > 0 && IsIdentifierChar(result[scan - 1])) continue;
const LodLookupForm* form = nullptr;
SizeT openParen = 0;
for (const auto& candidate : LOD_LOOKUP_FORMS) {
const SizeT nameLength = std::strlen(candidate.name);
if (result.compare(scan, nameLength, candidate.name) != 0) continue;
SizeT after = result.find_first_not_of(" \t", scan + nameLength);
if (after == String::npos || result[after] != '(') continue;
form = &candidate;
openParen = after;
if (!std::regex_match(line, redeclarationRegex)) {
if (!firstLine) {
result += '\n';
}
result += line;
firstLine = false;
}
if (lastLine) {
break;
}
if (form == nullptr) continue;
const Vector<SizeT> marks = SplitCallArguments(result, openParen);
if (marks.empty()) continue;
const SizeT argCount = marks.size();
const SizeT closeParen = marks.back();
// First argument must be one of our samplers.
const SizeT firstArgStart = result.find_first_not_of(" \t", openParen + 1);
SizeT firstArgEnd = marks.front();
while (firstArgEnd > firstArgStart && (result[firstArgEnd - 1] == ' ' || result[firstArgEnd - 1] == '\t')) {
--firstArgEnd;
}
if (firstArgStart == String::npos || firstArgEnd <= firstArgStart) continue;
const String samplerName = result.substr(firstArgStart, firstArgEnd - firstArgStart);
const auto samplerIt = samplerNames.find(samplerName);
if (samplerIt == samplerNames.end()) continue;
const String& biasName = samplerIt->second;
if (form->explicitLodArg >= 0) {
// Explicit LOD: the bias adds to it, as Vulkan does for
// OpImageSampleExplicitLod and as the CTS reference expects.
const SizeT lodIndex = static_cast<SizeT>(form->explicitLodArg);
if (argCount <= lodIndex) continue;
const SizeT lodStart = marks[lodIndex - 1] + 1;
const SizeT lodEnd = marks[lodIndex];
result.insert(lodEnd, String(") + ") + biasName + ")");
result.insert(lodStart, "((");
} else {
const SizeT required = static_cast<SizeT>(form->requiredArgs);
if (argCount == required) {
result.insert(closeParen, String(", ") + biasName);
} else if (argCount == required + 1) {
const SizeT biasStart = marks[argCount - 2] + 1;
result.insert(closeParen, String(") + ") + biasName + ")");
result.insert(biasStart, "((");
} else {
continue;
}
}
usedSamplers.push_back(samplerName);
}
if (usedSamplers.empty()) {
return glslCode;
}
// Declare the bias uniforms that were actually referenced, right after the
// sampler declaration line they belong to.
for (const auto& samplerName : usedSamplers) {
const String& biasName = samplerNames[samplerName];
if (result.find(String("float ") + biasName + ";") != String::npos) continue;
const std::regex declRegex(
R"(uniform\s+(?:(?:highp|mediump|lowp)\s+)?[iu]?sampler[A-Za-z0-9]*\s+)" + samplerName + R"(\s*;)");
std::smatch match;
if (!std::regex_search(result, match, declRegex)) continue;
const SizeT declEnd = static_cast<SizeT>(match.position(0)) + match[0].str().size();
result.insert(declEnd, String("\nuniform highp float ") + biasName + ";");
lineStart = lineEnd + 1;
}
return result;
}
} // namespace PrgramImpl
namespace Utils {
+1 -24
View File
@@ -40,11 +40,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
void GenerateRenderbufferFormatInfo(TextureInternalFormat internalFormat, GLenum* outInternalFormat,
GLenum* outFormat, GLenum* outType);
Bool ShouldUseCaveatTextureFormat(TextureInternalFormat internalFormat, TextureTarget target);
// True when the format the texture is actually created with has an alpha channel the
// frontend format does not (the three-channel multisample widening). GL reads such a
// channel back as 1.0, so any swizzle source of ALPHA has to be answered with ONE.
Bool BackendTextureFormatAddsAlpha(TextureInternalFormat internalFormat, TextureTarget target);
Bool ShouldUseCaveatRenderbufferFormat(TextureInternalFormat internalFormat);
} // namespace TextureImpl
@@ -109,26 +104,8 @@ namespace MobileGL::MG_Backend::DirectGLES {
String ClampNormFallbackOutputs(String glslCode, GLenum shaderType, Uint32 snormOutputMask,
Uint32 unormOutputMask);
String ForceFlatIntegerVaryings(const String& glslCode, GLenum shaderType);
// Legacy GLSL's gl_FragColor is broadcast to every enabled draw buffer (GL 4.6
// 15.2.3), but ShaderSourceProcessor lowers it to the single output mg_FragColor,
// which only ever reaches draw buffer 0. Replicates it across `drawBufferCount`
// outputs and copies the value into them at the end of main. A no-op for
// drawBufferCount <= 1, i.e. for everything but a framebuffer that actually
// enables several draw buffers, so the ordinary single-target shader is untouched.
String BroadcastLegacyFragColor(String glslCode, GLenum shaderType, Uint drawBufferCount);
String RemoveLayoutBinding(const String& glslCode);
// Prefix of the per-sampler float uniform that carries GL_TEXTURE_LOD_BIAS into
// the shader (see EmulateTextureLodBias); the suffix is the sampler's own name.
constexpr const char* LOD_BIAS_UNIFORM_PREFIX = "mg_lodBias_";
// ES has no per-texture/sampler LOD bias at all (GL_TEXTURE_LOD_BIAS is desktop
// only; Vulkan spells it VkSamplerCreateInfo::mipLodBias), so it has to reach the
// shader as a uniform and be folded into every lookup's level of detail. Declares
// one `uniform highp float mg_lodBias_<sampler>;` per mip-capable sampler and adds
// it to the bias / explicit-LOD argument of every lookup that takes one. Draws push
// the bound texture's (or sampler object's) value into it; a shader whose samplers
// all have a zero bias is therefore unaffected. Returns the source unchanged when
// there is nothing to rewrite.
String EmulateTextureLodBias(const String& glslCode);
String RemoveClipDistanceRedeclaration(const String& glslCode);
} // namespace PrgramImpl
namespace Utils {
@@ -18,7 +18,6 @@
#include "MG_Util/Texture/TextureFormatProcessor.h"
#include <Config.h>
#include <cmath>
#include <cstdlib>
#include <cstring>
@@ -470,9 +469,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// treat them as signaled/available with zero results from here on.
BumpRendererGeneration();
pVulkanRenderer.reset();
// The reflection cache is file-scope, not renderer-owned; without this the
// deleted programs' reflection strings survive full context teardown.
ClearProgramResourceCaches();
BackendObject::ReleaseEGLResources();
}
@@ -482,9 +478,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// treat them as signaled/available with zero results from here on.
BumpRendererGeneration();
pVulkanRenderer.reset();
// The reflection cache is file-scope, not renderer-owned; without this the
// deleted programs' reflection strings survive full context teardown.
ClearProgramResourceCaches();
}
const RendererInfo& BackendObject_DirectVulkan::GetRendererInfo() const {
@@ -526,15 +519,10 @@ namespace MobileGL::MG_Backend::DirectVulkan {
E_GL_ARB_multi_draw_indirect, E_GL_ARB_indirect_parameters,
E_GL_EXT_framebuffer_object, E_GL_ARB_depth_texture, E_GL_ARB_buffer_storage,
E_GL_ARB_texture_storage, E_GL_ARB_texture_storage_multisample,
E_GL_ARB_texture_multisample, E_GL_ARB_clear_texture, E_GL_ARB_direct_state_access,
E_GL_ARB_clear_texture, E_GL_ARB_direct_state_access,
E_GL_ARB_shader_draw_parameters, E_GL_ARB_gpu_shader_int64, E_GL_KHR_debug,
E_GL_ARB_gpu_shader5, E_GL_ARB_multi_bind, E_GL_ARB_shading_language_420pack,
E_GL_ARB_vertex_attrib_binding, E_GL_ARB_shader_image_size,
E_GL_ARB_explicit_attrib_location,
// Advertised with GL_NUM_PROGRAM_BINARY_FORMATS = 0, which the
// extension explicitly permits. It is also the only thing that
// exposes glProgramParameteri before GL 4.1.
E_GL_ARB_get_program_binary};
E_GL_ARB_vertex_attrib_binding, E_GL_ARB_shader_image_size};
if (shaderSubgroupSupported && !MG_Config::Features.DisableSubgroup) {
extensions.push_back(E_GL_KHR_shader_subgroup);
}
@@ -640,15 +628,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
funcsTable.GL.DeleteBackendQuery = DeleteBackendQuery;
funcsTable.GL.GetGpuTimestampNs = GetGpuTimestampNs;
}
// Occlusion queries share the handle-based result/delete entries, which must
// exist even when timer queries are disabled.
funcsTable.GL.BeginOcclusionQuery = BeginOcclusionQuery;
funcsTable.GL.EndOcclusionQuery = EndOcclusionQuery;
funcsTable.GL.BeginXfbPrimitivesQuery = BeginXfbPrimitivesQuery;
funcsTable.GL.EndXfbPrimitivesQuery = EndXfbPrimitivesQuery;
funcsTable.GL.IsQueryResultAvailable = IsQueryResultAvailable;
funcsTable.GL.GetQueryResult64 = GetQueryResult64;
funcsTable.GL.DeleteBackendQuery = DeleteBackendQuery;
funcsTableInitialized = true;
}
return funcsTable;
@@ -780,7 +759,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_dynamicParameters.MaxComputeWorkGroupInvocations = m_vulkanCaps.MaxComputeWorkGroupInvocations;
m_dynamicParameters.MaxShaderStorageBufferBindings = m_vulkanCaps.MaxShaderStorageBufferBindings;
m_dynamicParameters.MaxTextureBufferSize = m_vulkanCaps.MaxTextureBufferSize;
m_dynamicParameters.TextureBufferOffsetAlignment = m_vulkanCaps.TextureBufferOffsetAlignment;
m_dynamicParameters.MaxUniformBufferBindings = m_vulkanCaps.MaxUniformBufferBindings;
m_dynamicParameters.MaxUniformBlockSize = m_vulkanCaps.MaxUniformBlockSize;
m_dynamicParameters.MaxImageUnits =
@@ -812,23 +790,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_dynamicParameters.ViewportBoundsRangeMin = m_vulkanCaps.ViewportBoundsRangeMin;
m_dynamicParameters.ViewportBoundsRangeMax = m_vulkanCaps.ViewportBoundsRangeMax;
m_dynamicParameters.ViewportSubpixelBits = m_vulkanCaps.ViewportSubpixelBits;
m_dynamicParameters.MinFragmentInterpolationOffset =
std::isfinite(m_vulkanCaps.MinFragmentInterpolationOffset) &&
m_vulkanCaps.MinFragmentInterpolationOffset <= -0.5f
? m_vulkanCaps.MinFragmentInterpolationOffset
: -0.5f;
m_dynamicParameters.MaxFragmentInterpolationOffset = 0.4375f;
m_dynamicParameters.FragmentInterpolationOffsetBits = 4;
if (m_vulkanCaps.FragmentInterpolationOffsetBits >= 4 &&
std::isfinite(m_vulkanCaps.MaxFragmentInterpolationOffset)) {
const Float requiredMaxOffset =
0.5f - std::ldexp(1.0f, -m_vulkanCaps.FragmentInterpolationOffsetBits);
if (m_vulkanCaps.MaxFragmentInterpolationOffset >= requiredMaxOffset) {
m_dynamicParameters.MaxFragmentInterpolationOffset = m_vulkanCaps.MaxFragmentInterpolationOffset;
m_dynamicParameters.FragmentInterpolationOffsetBits =
m_vulkanCaps.FragmentInterpolationOffsetBits;
}
}
m_dynamicParameters.SupportsWideLines = m_vulkanCaps.SupportsWideLines;
m_dynamicParameters.MaxShaderStorageBlockSize =
std::min(m_vulkanCaps.MaxShaderStorageBlockSize, kMaxAdvertisedShaderStorageBlockSize);
@@ -61,12 +61,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
};
struct ProgramResourceCache {
// Lifetime id of the program the cached reflection belongs to. GL names are
// recycled (IndexGenerator hands freed indices straight back), and a
// recreated program's backendStateVersion restarts at the same small values,
// so the version alone can collide; the never-reused lifetime id makes the
// slot's ownership unambiguous.
Uint64 programLifetimeId = 0;
Uint32 backendStateVersion = 0;
Vector<StorageBlockResource> storageBlocks;
Vector<BufferVariableResource> bufferVariables;
@@ -88,11 +82,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Uint32 baseInstance = 0;
};
// Keyed by GL program name so the freed-name reuse in IndexGenerator bounds the
// map at the peak-simultaneous-program high-water mark; each slot's ownership is
// checked against the program's lifetime id before it is served (see
// GetProgramResourceCache). Cleared wholesale at EGL teardown via
// ClearProgramResourceCaches.
UnorderedMap<GLuint, ProgramResourceCache> g_programResourceCaches;
void ClearReadPixelsOutput(GLsizei width, GLsizei height, GLenum format, GLenum type, void* pixels) {
@@ -153,19 +142,13 @@ namespace MobileGL::MG_Backend::DirectVulkan {
ProgramResourceCache& GetProgramResourceCache(const MG_State::GLState::ProgramObject& program) {
auto& cache = g_programResourceCaches[program.GetExternalIndex()];
const Uint64 programLifetimeId = program.GetLifetimeId();
const Uint32 backendStateVersion = program.GetBackendStateVersion();
// The lifetime id must match too: a new program that reuses a deleted
// program's name and happens to land on the same backendStateVersion (both
// count from zero) would otherwise be served the dead program's reflection.
if (cache.programLifetimeId == programLifetimeId &&
cache.backendStateVersion == backendStateVersion &&
if (cache.backendStateVersion == backendStateVersion &&
(!cache.storageBlocks.empty() || !cache.bufferVariables.empty())) {
return cache;
}
cache = {};
cache.programLifetimeId = programLifetimeId;
cache.backendStateVersion = backendStateVersion;
Vector<SpvReflectShaderModule> modules;
@@ -383,15 +366,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
} // namespace
void ClearProgramResourceCaches() {
// Called from EGL teardown while the backend's m_eglStateMutex is held; GL
// calls are serialized in this codebase (contexts migrate threads but never
// run concurrently), so no other thread can be inside the unsynchronized map.
// Live programs in another context self-heal: their entry rebuilds from the
// retained generated SPIR-V on the next resource query.
g_programResourceCaches.clear();
}
GLuint GetShaderStorageBlockIndex(const MG_State::GLState::ProgramObject& program, const String& name) {
auto& cache = GetProgramResourceCache(program);
const auto it = std::find_if(cache.storageBlocks.begin(), cache.storageBlocks.end(),
@@ -1250,76 +1224,10 @@ namespace MobileGL::MG_Backend::DirectVulkan {
pVulkanRenderer->Clear(mask);
}
// Vulkan has no LINE_LOOP topology; rewrite the draw as an indexed LINE_STRIP
// whose synthesized index list revisits the first vertex at the end.
static void DrawLineLoopAsIndexedStrip(const Vector<Uint32>& closedIndices, GLint basevertex) {
DrawIndexedCmd payload{};
payload.mode = GL_LINE_STRIP;
payload.indexBufferView.indexType = GL_UNSIGNED_INT;
payload.indexBufferView.indexByteOffset = reinterpret_cast<SizeT>(closedIndices.data());
payload.indexBufferView.indexByteSize = closedIndices.size() * sizeof(Uint32);
payload.indexBufferView.forceClientMemory = true;
payload.params.indexCount = static_cast<Uint32>(closedIndices.size());
payload.params.instanceCount = 1;
payload.params.vertexOffset = basevertex;
pVulkanRenderer->DrawElements(payload);
}
// Resolve a DrawElements index list (bound element-array buffer or client
// memory) into uint32 values with the loop-closing first index appended.
static Bool BuildClosedLineLoopIndices(GLsizei count, GLenum type, const void* indices,
Vector<Uint32>& outIndices) {
const SizeT indexSize = MG_Util::GetGLTypeSize(type);
if (indexSize == 0 || count < 2) {
return false;
}
const Uint8* indexBytes = nullptr;
const auto& vao = *MG_State::pGLContext->GetBoundVertexArray();
const auto& indexBufferShared = vao.GetIndexBufferBindingSlot().GetBoundObject();
if (indexBufferShared != nullptr) {
const SizeT offset = reinterpret_cast<SizeT>(indices);
const SizeT bufferSize = indexBufferShared->GetSize();
if (indexBufferShared->MappedData() == nullptr || offset > bufferSize ||
static_cast<SizeT>(count) * indexSize > bufferSize - offset) {
return false;
}
indexBufferShared->SyncPersistentMappedRange();
indexBytes = indexBufferShared->MappedData() + offset;
} else {
indexBytes = static_cast<const Uint8*>(indices);
if (indexBytes == nullptr) {
return false;
}
}
outIndices.resize(static_cast<SizeT>(count) + 1);
for (GLsizei i = 0; i < count; ++i) {
switch (indexSize) {
case 1: outIndices[i] = indexBytes[i]; break;
case 2: outIndices[i] = reinterpret_cast<const Uint16*>(indexBytes)[i]; break;
default: outIndices[i] = reinterpret_cast<const Uint32*>(indexBytes)[i]; break;
}
}
outIndices[count] = outIndices[0];
return true;
}
void DrawArrays(GLenum mode, GLint first, GLsizei count) {
MOBILEGL_ASSERT(pVulkanRenderer, "DirectVulkan::DrawArrays called with null VulkanRenderer");
MOBILEGL_ASSERT(MG_State::pGLContext, "DirectVulkan::DrawArrays called with null GL context");
if (mode == GL_LINE_LOOP) {
if (count < 2) {
return;
}
Vector<Uint32> closedIndices(static_cast<SizeT>(count) + 1);
for (GLsizei i = 0; i < count; ++i) {
closedIndices[i] = static_cast<Uint32>(first + i);
}
closedIndices[count] = static_cast<Uint32>(first);
DrawLineLoopAsIndexedStrip(closedIndices, 0);
return;
}
DrawCmd payload{};
payload.mode = mode;
payload.params.firstVertex = first;
@@ -1332,14 +1240,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
MOBILEGL_ASSERT(pVulkanRenderer, "DirectVulkan::DrawElements called with null VulkanRenderer");
MOBILEGL_ASSERT(MG_State::pGLContext, "DirectVulkan::DrawElements called with null GL context");
if (mode == GL_LINE_LOOP) {
Vector<Uint32> closedIndices;
if (BuildClosedLineLoopIndices(count, type, indices, closedIndices)) {
DrawLineLoopAsIndexedStrip(closedIndices, 0);
}
return;
}
DrawIndexedCmd payload{};
payload.mode = mode;
payload.indexBufferView.indexType = type;
@@ -1408,13 +1308,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
void DrawElementsBaseVertex(GLenum mode, GLsizei count, GLenum type, const GLvoid* indices, GLint basevertex) {
MOBILEGL_ASSERT(pVulkanRenderer, "DirectVulkan::DrawElementsBaseVertex called with null VulkanRenderer");
MOBILEGL_ASSERT(MG_State::pGLContext, "DirectVulkan::DrawElementsBaseVertex called with null GL context");
if (mode == GL_LINE_LOOP) {
Vector<Uint32> closedIndices;
if (BuildClosedLineLoopIndices(count, type, indices, closedIndices)) {
DrawLineLoopAsIndexedStrip(closedIndices, basevertex);
}
return;
}
DrawIndexedCmd payload{};
payload.mode = mode;
payload.indexBufferView.indexType = type;
@@ -1564,12 +1457,8 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// records are shared (SharedPtr) with the owning pool's pending list,
// so deleting the query while results are still in flight is safe.
struct VulkanTimerQuery {
enum class Kind : Uint8 { Timer, Occlusion, XfbWritten, XfbGenerated };
Kind kind = Kind::Timer;
SharedPtr<VkTimerQueryManager::TimestampRecord> begin;
SharedPtr<VkTimerQueryManager::TimestampRecord> end;
// Kind::Occlusion - pool slots recorded between Begin/End; summed at result time.
Vector<Uint32> occlusionSlots;
// Renderer generation the records were written under (see
// g_rendererGeneration). A stale generation resolves as available
// with a final zero result: the records' pool indices and frame
@@ -1578,12 +1467,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// (and, via the SharedPtrs, the records), never pool slots, so
// stale queries are always safe to delete.
Uint64 rendererGeneration = 0;
// Kind::XfbGenerated - the frontend's paused-draw primitive counter when the
// query began. VK_QUERY_TYPE_TRANSFORM_FEEDBACK_STREAM_EXT counts only what the
// capture saw, so a draw made while the span was paused is invisible to it -
// but GL_PRIMITIVES_GENERATED counts what the last vertex processing stage
// emitted regardless. The delta closes that gap at result time.
Uint64 pausedPrimitiveSnapshot = 0;
};
} // namespace
@@ -1665,30 +1548,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// ever be produced, so resolve with a final 0.
return true;
}
if (query->kind == VulkanTimerQuery::Kind::Occlusion) {
Uint64 samples = 0;
if (!pVulkanRenderer->ResolveOcclusionQueryResult(query->occlusionSlots, samples)) {
return false;
}
query->occlusionSlots.clear(); // slots are recycled by the resolve
*outNanoseconds = samples;
return true;
}
if (query->kind == VulkanTimerQuery::Kind::XfbWritten ||
query->kind == VulkanTimerQuery::Kind::XfbGenerated) {
Uint64 primitives = 0;
if (!pVulkanRenderer->ResolveXfbQueryResult(query->occlusionSlots,
query->kind == VulkanTimerQuery::Kind::XfbGenerated,
primitives)) {
return false;
}
if (query->kind == VulkanTimerQuery::Kind::XfbGenerated && MG_State::pGLContext != nullptr) {
primitives += MG_State::pGLContext->GetTransformFeedbackPausedPrimitiveCounter() -
query->pausedPrimitiveSnapshot;
}
*outNanoseconds = primitives;
return true;
}
// With wait, mirrors ClientWaitSync: a query ended this frame cannot
// complete until Present submits the commands, so the wait refuses to
// block on the current unsubmitted serial. Returning false keeps the
@@ -1721,49 +1580,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
delete static_cast<VulkanTimerQuery*>(handle);
}
BackendQueryHandle BeginXfbPrimitivesQuery(Bool generated) {
MOBILEGL_ASSERT(pVulkanRenderer, "DirectVulkan::BeginXfbPrimitivesQuery called with null VulkanRenderer");
if (!pVulkanRenderer->StartXfbQueryCapture(generated ? 1u : 0u)) {
return nullptr;
}
auto* query = new VulkanTimerQuery{};
query->kind = generated ? VulkanTimerQuery::Kind::XfbGenerated : VulkanTimerQuery::Kind::XfbWritten;
query->rendererGeneration = GetRendererGeneration();
query->pausedPrimitiveSnapshot =
MG_State::pGLContext ? MG_State::pGLContext->GetTransformFeedbackPausedPrimitiveCounter() : 0;
return query;
}
void EndXfbPrimitivesQuery(BackendQueryHandle handle) {
auto* query = static_cast<VulkanTimerQuery*>(handle);
MOBILEGL_ASSERT(pVulkanRenderer, "DirectVulkan::EndXfbPrimitivesQuery called with null VulkanRenderer");
if (query == nullptr || query->rendererGeneration != GetRendererGeneration()) {
return;
}
pVulkanRenderer->StopXfbQueryCapture(
query->kind == VulkanTimerQuery::Kind::XfbGenerated ? 1u : 0u, query->occlusionSlots);
}
BackendQueryHandle BeginOcclusionQuery() {
MOBILEGL_ASSERT(pVulkanRenderer, "DirectVulkan::BeginOcclusionQuery called with null VulkanRenderer");
if (!pVulkanRenderer->StartOcclusionQueryCapture()) {
return nullptr;
}
auto* query = new VulkanTimerQuery{};
query->kind = VulkanTimerQuery::Kind::Occlusion;
query->rendererGeneration = GetRendererGeneration();
return query;
}
void EndOcclusionQuery(BackendQueryHandle handle) {
auto* query = static_cast<VulkanTimerQuery*>(handle);
MOBILEGL_ASSERT(pVulkanRenderer, "DirectVulkan::EndOcclusionQuery called with null VulkanRenderer");
if (query == nullptr || query->rendererGeneration != GetRendererGeneration()) {
return;
}
pVulkanRenderer->StopOcclusionQueryCapture(query->occlusionSlots);
}
Int64 GetGpuTimestampNs() {
// Vulkan cannot synchronously sample the GPU clock: timestamps only
// exist as vkCmdWriteTimestamp results read back later, and
@@ -23,12 +23,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Uint64 GetRendererGeneration();
void BumpRendererGeneration();
// Drops every cached program-resource reflection entry (CPU-side strings/vectors
// only, no Vulkan handles). Called at EGL teardown next to the renderer reset;
// safe because GL calls are serialized in this codebase, and any still-live
// program rebuilds its entry from the retained generated SPIR-V on demand.
void ClearProgramResourceCaches();
void ClearBufferfi(GLenum buffer, GLint drawbuffer, GLfloat depth, GLint stencil);
void ClearBufferfv(GLenum buffer, GLint drawbuffer, const GLfloat* value);
void ClearBufferuiv(GLenum buffer, GLint drawbuffer, const GLuint* value);
@@ -123,10 +117,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// only while a live renderer exists whose device can actually time.
Bool IsTimerQuerySupported();
BackendQueryHandle BeginTimeElapsedQuery();
BackendQueryHandle BeginXfbPrimitivesQuery(Bool generated);
void EndXfbPrimitivesQuery(BackendQueryHandle query);
BackendQueryHandle BeginOcclusionQuery();
void EndOcclusionQuery(BackendQueryHandle query);
void EndTimeElapsedQuery(BackendQueryHandle query);
BackendQueryHandle QueryCounterTimestamp();
Bool IsQueryResultAvailable(BackendQueryHandle query);
@@ -16,19 +16,18 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_device = device;
m_commandPool = commandPool;
Vector<VkCommandBuffer> commandBuffers(frameCount * 2, VK_NULL_HANDLE);
Vector<VkCommandBuffer> commandBuffers(frameCount, VK_NULL_HANDLE);
VkCommandBufferAllocateInfo allocInfo{};
allocInfo.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_ALLOCATE_INFO;
allocInfo.commandPool = commandPool;
allocInfo.level = VK_COMMAND_BUFFER_LEVEL_PRIMARY;
allocInfo.commandBufferCount = frameCount * 2;
allocInfo.commandBufferCount = frameCount;
VkResult result = vkAllocateCommandBuffers(device, &allocInfo, commandBuffers.data());
if (result != VK_SUCCESS) {
return result;
}
for (Uint32 i = 0; i < frameCount; ++i) {
m_frames[i].commandBuffer = commandBuffers[i];
m_frames[i].preCommandBuffer = commandBuffers[frameCount + i];
}
VkSemaphoreCreateInfo semaphoreInfo{VK_STRUCTURE_TYPE_SEMAPHORE_CREATE_INFO};
@@ -48,10 +47,9 @@ namespace MobileGL::MG_Backend::DirectVulkan {
void FrameContext::Destroy(VkDevice device, VkCommandPool commandPool) {
const Uint32 frameCount = static_cast<Uint32>(m_frames.size());
Vector<VkCommandBuffer> commandBuffers(frameCount * 2, VK_NULL_HANDLE);
Vector<VkCommandBuffer> commandBuffers(frameCount, VK_NULL_HANDLE);
for (Uint32 i = 0; i < frameCount; ++i) {
commandBuffers[i] = m_frames[i].commandBuffer;
commandBuffers[frameCount + i] = m_frames[i].preCommandBuffer;
}
for (Uint32 i = 0; i < frameCount; ++i) {
@@ -62,7 +60,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
for (auto& frame : m_frames) {
FreeRetiredCommandBuffers(frame);
}
vkFreeCommandBuffers(device, commandPool, frameCount * 2, commandBuffers.data());
vkFreeCommandBuffers(device, commandPool, frameCount, commandBuffers.data());
}
m_frames.clear();
currentFrameIndex = 0;
@@ -89,8 +87,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
currentFrameIndex = (currentFrameIndex + 1) % static_cast<Uint32>(m_frames.size());
GetCurrent().isCommandRecording = false;
GetCurrent().hasCommandBufferRecorded = false;
GetCurrent().isPreCommandRecording = false;
GetCurrent().hasPreCommandBufferRecorded = false;
}
VkCommandBuffer& FrameContext::BeginCommandRecording(VkCommandBufferUsageFlags flags,
@@ -122,41 +118,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
frame.hasCommandBufferRecorded = true;
}
VkCommandBuffer FrameContext::BeginPreCommandRecording() {
auto& frame = GetCurrent();
if (frame.isPreCommandRecording) {
return frame.preCommandBuffer;
}
MOBILEGL_ASSERT(!frame.hasPreCommandBufferRecorded,
"BeginPreCommandRecording: a recorded pre stream is still awaiting submission");
VK_VERIFY(vkResetCommandBuffer(frame.preCommandBuffer, 0), "BeginPreCommandRecording, vkResetCommandBuffer");
VkCommandBufferBeginInfo beginInfo{};
beginInfo.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_BEGIN_INFO;
VK_VERIFY(vkBeginCommandBuffer(frame.preCommandBuffer, &beginInfo),
"BeginPreCommandRecording, vkBeginCommandBuffer");
frame.isPreCommandRecording = true;
return frame.preCommandBuffer;
}
void FrameContext::EndPreCommandRecordingIfOpen() {
auto& frame = GetCurrent();
if (!frame.isPreCommandRecording) {
return;
}
VK_VERIFY(vkEndCommandBuffer(frame.preCommandBuffer), "EndPreCommandRecordingIfOpen, vkEndCommandBuffer");
frame.isPreCommandRecording = false;
frame.hasPreCommandBufferRecorded = true;
}
void FrameContext::AbandonPreCommandRecording() {
auto& frame = GetCurrent();
if (frame.isPreCommandRecording) {
VK_VERIFY(vkEndCommandBuffer(frame.preCommandBuffer), "AbandonPreCommandRecording, vkEndCommandBuffer");
}
frame.isPreCommandRecording = false;
frame.hasPreCommandBufferRecorded = false;
}
VkResult FrameContext::InitializeSwapchainSemaphores(VkDevice device, Uint32 swapchainImageCount) {
DestroySwapchainSemaphores(device);
if (swapchainImageCount == 0) {
@@ -189,30 +150,12 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Bool FrameContext::TransitionToPresent(VkImage image, VkImageLayout oldLayout, VkImageLayout presentLayout) {
auto& frame = GetCurrent();
if (oldLayout == presentLayout || oldLayout == VK_IMAGE_LAYOUT_SHARED_PRESENT_KHR) {
if (frame.hasCommandBufferRecorded || frame.isCommandRecording || oldLayout == presentLayout ||
oldLayout == VK_IMAGE_LAYOUT_SHARED_PRESENT_KHR) {
return false;
}
// The barrier belongs in the frame's own recording. Bailing out because
// something was already recorded (the previous behaviour) dropped the
// transition entirely for every frame that never ran a default-framebuffer
// render pass - the only other thing that carries the image to
// PRESENT_SRC_KHR, via that pass's finalLayout - so the swapchain image was
// handed to the WSI still in the layout it was acquired in.
// A closed-but-unsubmitted buffer can only come from a submit that already
// failed (SubmitPendingCommandBuffer leaves the flag set on error), and
// appending to it is illegal while reopening would reset the frame's own
// commands away. The device is gone on that path anyway - stay silent-safe
// rather than trade a lost device for a barrier into a closed buffer.
if (frame.hasCommandBufferRecorded) {
MGLOG_E("TransitionToPresent: command buffer already closed; skipping the present barrier");
return false;
}
// Reopening a recording here would vkResetCommandBuffer this frame's own
// commands away, so append to the open one and let the caller close it.
const Bool openedRecording = !frame.isCommandRecording;
VkCommandBuffer commandBuffer = openedRecording ? BeginCommandRecording() : frame.commandBuffer;
auto& commandBuffer = BeginCommandRecording();
VkImageMemoryBarrier presentBarrier{};
presentBarrier.sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER;
@@ -231,9 +174,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
vkCmdPipelineBarrier(commandBuffer, VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT, VK_PIPELINE_STAGE_BOTTOM_OF_PIPE_BIT, 0, 0,
nullptr, 0, nullptr, 1, &presentBarrier);
if (openedRecording) {
EndCommandRecording();
}
EndCommandRecording();
return true;
}
@@ -241,27 +182,17 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Uint32 swapchainImageIndex) const {
const auto& frame = GetCurrent();
MOBILEGL_ASSERT(!frame.isCommandRecording, "GetSubmitInfo called while command buffer recording is still active");
MOBILEGL_ASSERT(!frame.isPreCommandRecording,
"GetSubmitInfo called while the pre-pass stream is still recording");
AssertValidSwapchainImageIndex(swapchainImageIndex);
SubmitInfoPacket packet{};
packet.waitSemaphore = frame.imageAvailableSemaphore;
packet.signalSemaphore = m_swapchainImageRenderFinishedSemaphores[swapchainImageIndex];
Uint32 commandBufferCount = 0;
// The pre-pass stream executes strictly before the frame's commands.
if (frame.hasPreCommandBufferRecorded) {
packet.commandBuffers[commandBufferCount++] = frame.preCommandBuffer;
}
if (shouldSubmitCommandBuffer) {
packet.commandBuffers[commandBufferCount++] = frame.commandBuffer;
}
packet.commandBuffer = frame.commandBuffer;
packet.submitInfo.waitSemaphoreCount = frame.imageAvailableSemaphoreConsumed ? 0U : 1U;
packet.submitInfo.pWaitSemaphores = frame.imageAvailableSemaphoreConsumed ? nullptr : &packet.waitSemaphore;
packet.submitInfo.pWaitDstStageMask = frame.imageAvailableSemaphoreConsumed ? nullptr : &packet.waitDstStageMask;
packet.submitInfo.commandBufferCount = commandBufferCount;
packet.submitInfo.pCommandBuffers = commandBufferCount > 0 ? packet.commandBuffers : nullptr;
packet.submitInfo.commandBufferCount = shouldSubmitCommandBuffer ? 1U : 0U;
packet.submitInfo.pCommandBuffers = shouldSubmitCommandBuffer ? &packet.commandBuffer : nullptr;
packet.submitInfo.signalSemaphoreCount = 1;
packet.submitInfo.pSignalSemaphores = &packet.signalSemaphore;
return packet;
@@ -296,21 +227,12 @@ namespace MobileGL::MG_Backend::DirectVulkan {
result = vkAcquireNextImageKHR(device, swapchain, timeout, frame.imageAvailableSemaphore, acquireFence,
&outImageIndex);
// VK_SUBOPTIMAL_KHR is a success code: an image *was* acquired and
// imageAvailableSemaphore *will* be signaled. Bailing out on it skipped both
// the consumed-flag reset (leaving a stale "already consumed", so the next
// submit never waited on the pending signal) and the fence reset (leaving
// the slot's fence signaled for the next submit to reuse). Only a genuine
// failure - VK_ERROR_OUT_OF_DATE_KHR and friends, where nothing is acquired
// and nothing is signaled - skips the bookkeeping.
if (result != VK_SUCCESS && result != VK_SUBOPTIMAL_KHR) {
if (result != VK_SUCCESS) {
return result;
}
frame.imageAvailableSemaphoreConsumed = false;
const VkResult resetResult = vkResetFences(device, 1, &frame.imageInFlightFence);
// Hand the acquire's own code back so the caller can schedule a rebuild.
return resetResult == VK_SUCCESS ? result : resetResult;
return vkResetFences(device, 1, &frame.imageInFlightFence);
}
Uint32 FrameContext::GetCurrentFrameIndex() const {
@@ -325,14 +247,12 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_recordingObserver = observer;
}
VkResult FrameContext::RetireCurrentCommandBuffer(Bool retirePreCommandBuffer) {
VkResult FrameContext::RetireCurrentCommandBuffer() {
MOBILEGL_ASSERT(m_device != VK_NULL_HANDLE && m_commandPool != VK_NULL_HANDLE,
"RetireCurrentCommandBuffer requires an initialized FrameContext");
auto& frame = GetCurrent();
MOBILEGL_ASSERT(!frame.isCommandRecording,
"RetireCurrentCommandBuffer called while the command buffer is still recording");
MOBILEGL_ASSERT(!frame.isPreCommandRecording,
"RetireCurrentCommandBuffer called while the pre-pass stream is still recording");
VkCommandBufferAllocateInfo allocInfo{};
allocInfo.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_ALLOCATE_INFO;
@@ -340,23 +260,11 @@ namespace MobileGL::MG_Backend::DirectVulkan {
allocInfo.level = VK_COMMAND_BUFFER_LEVEL_PRIMARY;
allocInfo.commandBufferCount = 1;
VkCommandBuffer replacement = VK_NULL_HANDLE;
VkResult result = vkAllocateCommandBuffers(m_device, &allocInfo, &replacement);
const VkResult result = vkAllocateCommandBuffers(m_device, &allocInfo, &replacement);
if (result != VK_SUCCESS) {
return result;
}
if (retirePreCommandBuffer) {
VkCommandBuffer preReplacement = VK_NULL_HANDLE;
result = vkAllocateCommandBuffers(m_device, &allocInfo, &preReplacement);
if (result != VK_SUCCESS) {
vkFreeCommandBuffers(m_device, m_commandPool, 1, &replacement);
return result;
}
frame.retiredCommandBuffers.push_back({frame.preCommandBuffer, frame.lastSubmitIndex});
frame.preCommandBuffer = preReplacement;
}
// lastSubmitIndex was just written by the renderer for the submission
// that carried this command buffer.
frame.retiredCommandBuffers.push_back({frame.commandBuffer, frame.lastSubmitIndex});
frame.retiredCommandBuffers.push_back(frame.commandBuffer);
frame.commandBuffer = replacement;
return VK_SUCCESS;
}
@@ -366,40 +274,12 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return;
}
if (m_device != VK_NULL_HANDLE && m_commandPool != VK_NULL_HANDLE) {
for (const auto& retired : frame.retiredCommandBuffers) {
vkFreeCommandBuffers(m_device, m_commandPool, 1, &retired.commandBuffer);
}
vkFreeCommandBuffers(m_device, m_commandPool, static_cast<Uint32>(frame.retiredCommandBuffers.size()),
frame.retiredCommandBuffers.data());
}
frame.retiredCommandBuffers.clear();
}
void FrameContext::FreeRetiredCommandBuffersCompletedUpTo(Uint64 completedSubmitIndex) {
if (m_device == VK_NULL_HANDLE || m_commandPool == VK_NULL_HANDLE) {
return;
}
for (auto& frame : m_frames) {
// Retired buffers are appended in submit order, so the completed
// ones form a prefix.
SizeT completedCount = 0;
while (completedCount < frame.retiredCommandBuffers.size() &&
frame.retiredCommandBuffers[completedCount].submitIndex <= completedSubmitIndex) {
vkFreeCommandBuffers(m_device, m_commandPool, 1,
&frame.retiredCommandBuffers[completedCount].commandBuffer);
++completedCount;
}
if (completedCount > 0) {
frame.retiredCommandBuffers.erase(frame.retiredCommandBuffers.begin(),
frame.retiredCommandBuffers.begin() + completedCount);
}
}
}
void FrameContext::FreeAllRetiredCommandBuffers() {
for (auto& frame : m_frames) {
FreeRetiredCommandBuffers(frame);
}
}
void FrameContext::AssertValidFrameIndex(Uint32 frameIndex) const {
MOBILEGL_ASSERT(frameIndex < m_frames.size(), "FrameContext index out of range");
}
@@ -29,9 +29,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkPipelineStageFlags waitDstStageMask = VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT;
VkSemaphore waitSemaphore = VK_NULL_HANDLE;
VkSemaphore signalSemaphore = VK_NULL_HANDLE;
// [0] = pre-pass command buffer (when recorded), then the frame
// command buffer; submitInfo.pCommandBuffers points here.
VkCommandBuffer commandBuffers[2] = {VK_NULL_HANDLE, VK_NULL_HANDLE};
VkCommandBuffer commandBuffer = VK_NULL_HANDLE;
VkSubmitInfo submitInfo{VK_STRUCTURE_TYPE_SUBMIT_INFO};
};
@@ -42,35 +40,17 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkPresentInfoKHR presentInfo{VK_STRUCTURE_TYPE_PRESENT_INFO_KHR};
};
// A command buffer submitted mid-frame (FlushPendingCommands), tagged
// with the submit-tracker index it was submitted under so it can be
// freed as soon as that submission is observed complete - without
// waiting for the slot's fence to be waited again (present-less flush
// loops never wait it).
struct RetiredCommandBuffer {
VkCommandBuffer commandBuffer = VK_NULL_HANDLE;
Uint64 submitIndex = 0;
};
struct FrameData {
VkCommandBuffer commandBuffer = VK_NULL_HANDLE;
// Pre-pass work stream: out-of-pass commands (deferred clear
// materialization, sampled-layout transitions) for resources the
// frame's recording has not touched yet. Submitted immediately
// BEFORE commandBuffer in the same vkQueueSubmit, so recording
// into it never has to split the frame's active render pass.
VkCommandBuffer preCommandBuffer = VK_NULL_HANDLE;
VkSemaphore imageAvailableSemaphore = VK_NULL_HANDLE;
VkFence imageInFlightFence = VK_NULL_HANDLE;
Bool isCommandRecording = false;
Bool hasCommandBufferRecorded = false;
Bool isPreCommandRecording = false;
Bool hasPreCommandBufferRecorded = false;
Bool imageAvailableSemaphoreConsumed = false;
// Command buffers submitted mid-frame (FlushPendingCommands),
// appended in submit order; freed once their submission is known
// complete (fence wait or completion poll).
Vector<RetiredCommandBuffer> retiredCommandBuffers;
// Command buffers submitted mid-frame (FlushPendingCommands) whose
// execution is only known complete once this slot's fence has been
// waited again; freed at that point.
Vector<VkCommandBuffer> retiredCommandBuffers;
// Submit-tracker index of this slot's most recent queue submission
// (written by the renderer at submit time).
Uint64 lastSubmitIndex = 0;
@@ -87,14 +67,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkCommandBuffer& BeginCommandRecording(VkCommandBufferUsageFlags flags = 0,
const VkCommandBufferInheritanceInfo* pInheritanceInfo = nullptr);
void EndCommandRecording();
// Lazily opens the pre-pass work stream (see FrameData::preCommandBuffer).
VkCommandBuffer BeginPreCommandRecording();
// Closes the pre stream if open, marking it for submission ahead of the
// frame command buffer. Safe to call when it never opened.
void EndPreCommandRecordingIfOpen();
// Drops an in-progress or recorded-but-unsubmitted pre stream (dropped
// frame recordings, swapchain recreation).
void AbandonPreCommandRecording();
VkResult InitializeSwapchainSemaphores(VkDevice device, Uint32 swapchainImageCount);
void DestroySwapchainSemaphores(VkDevice device);
Bool TransitionToPresent(VkImage image, VkImageLayout oldLayout,
@@ -107,18 +79,8 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// Parks the current (already ended and submitted) command buffer on the
// slot's retired list and installs a freshly allocated one, so recording
// can restart while the submitted buffer is still executing. Retired
// buffers are freed after the slot's fence is next waited, or as soon
// as their submission is observed complete.
VkResult RetireCurrentCommandBuffer(Bool retirePreCommandBuffer = false);
// Frees every retired command buffer whose tagged submission index is
// known complete. Driven by the renderer's submit tracker on completion
// events (fence waits and non-blocking polls), so present-less flush
// loops reclaim their buffers without any extra wait.
void FreeRetiredCommandBuffersCompletedUpTo(Uint64 completedSubmitIndex);
// Frees every slot's retired command buffers. Only valid when the
// caller has proven every queue submission complete.
void FreeAllRetiredCommandBuffers();
// buffers are freed after the slot's fence is next waited.
VkResult RetireCurrentCommandBuffer();
Uint32 GetCurrentFrameIndex() const;
Uint32 GetFrameCount() const;
@@ -8,7 +8,6 @@
#include "PipelineFactory.h"
#include <algorithm>
namespace MobileGL::MG_Backend::DirectVulkan {
static const char* PrimitiveTopologyToString(VkPrimitiveTopology topology) {
@@ -205,7 +204,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.topology, sizeof(payload.topology)));
XXHASH_VERIFY(
XXH64_update(m_hashState, &payload.primitiveRestartEnable, sizeof(payload.primitiveRestartEnable)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.patchControlPoints, sizeof(payload.patchControlPoints)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.polygonMode, sizeof(payload.polygonMode)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.cullMode, sizeof(payload.cullMode)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.frontFace, sizeof(payload.frontFace)));
@@ -245,108 +243,23 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const HashType hash = ComputeHash(payload);
auto it = m_cache.find(hash);
if (it != m_cache.end()) {
it->second.lastUsedFrame = m_frameCounter;
return it->second.pipeline;
return it->second;
}
VkPipeline pipeline = CreatePipeline(payload);
m_cache.emplace(hash, PipelineCacheEntry{pipeline, payload.programHash, payload.renderPass,
m_frameCounter});
m_cache.emplace(hash, pipeline);
return pipeline;
}
void PipelineFactory::DestroyAll() {
for (auto& pair : m_cache) {
if (pair.second.pipeline != VK_NULL_HANDLE) {
vkDestroyPipeline(m_device, pair.second.pipeline, nullptr);
if (pair.second != VK_NULL_HANDLE) {
vkDestroyPipeline(m_device, pair.second, nullptr);
}
}
m_cache.clear();
}
Uint32 PipelineFactory::OnFrameBoundary() {
++m_frameCounter;
// Sweep cadence and retire age mirror VkRenderPassManager::OnPresent: an entry
// idle for more than kRetireAgeFrames frame boundaries cannot be referenced by
// any in-flight command buffer (frames-in-flight <= MOBILEGL_MAGMA_FRAMESINFLIGHT),
// so immediate vkDestroyPipeline is safe. The caller must drop its "last
// pipeline" memo when this returns non-zero: the memo can return a cached
// handle without touching this cache, so an evicted pipeline may still be
// memoized (present-less flush loops never reset the memo per frame).
constexpr Uint64 kSweepInterval = 256;
constexpr Uint64 kRetireAgeFrames = 1024;
if ((m_frameCounter % kSweepInterval) != 0) {
return 0;
}
Uint32 evicted = 0;
for (auto it = m_cache.begin(); it != m_cache.end();) {
if (m_frameCounter - it->second.lastUsedFrame > kRetireAgeFrames) {
if (it->second.pipeline != VK_NULL_HANDLE) {
vkDestroyPipeline(m_device, it->second.pipeline, nullptr);
}
it = m_cache.erase(it);
++evicted;
} else {
++it;
}
}
if (evicted > 0) {
MGLOG_D("PipelineFactory::OnFrameBoundary: evicted %u idle pipelines (%zu remain)", evicted,
m_cache.size());
}
return evicted;
}
Uint32 PipelineFactory::EvictByRenderPasses(const Vector<VkRenderPass>& renderPasses) {
if (renderPasses.empty() || m_cache.empty()) {
return 0;
}
// Sorted-batch membership test keeps a mass eviction (shader-pack switch,
// dimension exit) at one O(cache * log batch) scan instead of one full scan
// per dying pass.
Vector<VkRenderPass> sortedPasses = renderPasses;
std::sort(sortedPasses.begin(), sortedPasses.end());
Uint32 evicted = 0;
for (auto it = m_cache.begin(); it != m_cache.end();) {
if (std::binary_search(sortedPasses.begin(), sortedPasses.end(), it->second.renderPass)) {
if (it->second.pipeline != VK_NULL_HANDLE) {
vkDestroyPipeline(m_device, it->second.pipeline, nullptr);
}
it = m_cache.erase(it);
++evicted;
} else {
++it;
}
}
if (evicted > 0) {
MGLOG_D("PipelineFactory::EvictByRenderPasses: evicted %u pipelines for %zu destroyed render passes",
evicted, sortedPasses.size());
}
return evicted;
}
Uint32 PipelineFactory::EvictByProgramHash(HashType programHash) {
Uint32 evicted = 0;
for (auto it = m_cache.begin(); it != m_cache.end();) {
if (it->second.programHash == programHash) {
if (it->second.pipeline != VK_NULL_HANDLE) {
vkDestroyPipeline(m_device, it->second.pipeline, nullptr);
}
it = m_cache.erase(it);
++evicted;
} else {
++it;
}
}
if (evicted > 0) {
MGLOG_D("PipelineFactory::EvictByProgramHash: evicted %u pipelines for program hash 0x%llx",
evicted, static_cast<unsigned long long>(programHash));
}
return evicted;
}
VkPipeline PipelineFactory::CreatePipeline(const PipelineCreatePayload& payload) const {
MOBILEGL_ASSERT(payload.stages != nullptr && !payload.stages->empty(), "PipelineFactory: stages are empty");
MOBILEGL_ASSERT(payload.vertexInputState != nullptr, "PipelineFactory: vertexInputState is null");
@@ -381,11 +294,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
ia.topology = payload.topology;
ia.primitiveRestartEnable = payload.primitiveRestartEnable ? VK_TRUE : VK_FALSE;
// Only a patch topology has a tessellation stage to configure; leaving the pointer null
// otherwise is what the spec expects.
VkPipelineTessellationStateCreateInfo tessellation{VK_STRUCTURE_TYPE_PIPELINE_TESSELLATION_STATE_CREATE_INFO};
tessellation.patchControlPoints = payload.patchControlPoints;
VkPipelineViewportStateCreateInfo vpci{VK_STRUCTURE_TYPE_PIPELINE_VIEWPORT_STATE_CREATE_INFO};
vpci.viewportCount = 1;
vpci.scissorCount = 1;
@@ -450,8 +358,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
gpi.pStages = payload.stages->data();
gpi.pVertexInputState = payload.vertexInputState;
gpi.pInputAssemblyState = &ia;
gpi.pTessellationState =
payload.topology == VK_PRIMITIVE_TOPOLOGY_PATCH_LIST ? &tessellation : nullptr;
gpi.pViewportState = &vpci;
gpi.pRasterizationState = &raster;
gpi.pMultisampleState = &ms;
@@ -30,8 +30,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Uint32 subpass = 0;
VkPrimitiveTopology topology = VK_PRIMITIVE_TOPOLOGY_TRIANGLE_LIST;
Bool primitiveRestartEnable = false;
// GL_PATCH_VERTICES; only read for a PATCH_LIST topology.
Uint32 patchControlPoints = 3;
VkPolygonMode polygonMode = VK_POLYGON_MODE_FILL;
VkCullModeFlags cullMode = VK_CULL_MODE_BACK_BIT;
VkFrontFace frontFace = VK_FRONT_FACE_CLOCKWISE;
@@ -67,26 +65,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkPipeline GetOrCreatePipeline(const PipelineCreatePayload& payload);
void DestroyAll();
// Frame boundary hook: ages the pipeline cache and destroys long-unused entries
// (their command buffers retired many frames ago), mirroring
// VkRenderPassManager::OnPresent's sweep. Returns the number of pipelines
// destroyed so the caller can drop any memoized VkPipeline handle.
Uint32 OnFrameBoundary();
// Destroys every cached pipeline hashed on one of `renderPasses`. Only safe
// when the caller guarantees GPU idleness for them - the render-pass manager
// calls this (via the renderer) for passes its own >1024-boundary-idle sweep
// just evicted, and a pipeline hashed on those handles is only ever bound by
// draws that also hit the render-pass entries. Also closes the handle-recycling
// hazard: a recycled VkRenderPass value must never serve a stale pipeline.
// Batched: one cache scan regardless of how many passes died in the sweep.
// Returns the number destroyed (callers invalidate memos when non-zero).
Uint32 EvictByRenderPasses(const Vector<VkRenderPass>& renderPasses);
// Destroys every cached pipeline built from the program with content hash
// `programHash`. Called from the ProgramFactory eviction path, which proves the
// same >1024-boundary idleness (the program's pipelines are only bound by draws
// that stamp its factory entry). Returns the number destroyed.
Uint32 EvictByProgramHash(HashType programHash);
// Driver quirk: suppress depth writes on accumulation-blended pipelines. Multi-pass
// depth-equality rendering (a blended prepass writes depth that later passes re-test
// with an equality-inclusive compare on the re-rasterized geometry) requires
@@ -109,26 +87,12 @@ namespace MobileGL::MG_Backend::DirectVulkan {
static Bool ShouldSuppressDepthWrite(const PipelineCreatePayload& payload);
private:
struct PipelineCacheEntry {
VkPipeline pipeline = VK_NULL_HANDLE;
// The hashed inputs the eviction paths key on: programHash ties the entry to
// its ProgramFactory entry, renderPass records the exact handle the hash
// folded in (the hash is one-way, so targeted eviction needs them verbatim).
HashType programHash = 0;
VkRenderPass renderPass = VK_NULL_HANDLE;
// Frame-boundary counter value of the last GetOrCreatePipeline hit; drives
// cache eviction (see OnFrameBoundary).
Uint64 lastUsedFrame = 0;
};
VkPipeline CreatePipeline(const PipelineCreatePayload& payload) const;
VkDevice m_device = VK_NULL_HANDLE;
const VulkanRendererConfig& m_config;
VkPipelineCache m_pipelineCache = VK_NULL_HANDLE;
UnorderedMap<HashType, PipelineCacheEntry> m_cache;
// Monotonic frame-boundary counter (bumped in OnFrameBoundary) for cache aging.
Uint64 m_frameCounter = 0;
UnorderedMap<HashType, VkPipeline> m_cache;
static inline XXH64_state_t* m_hashState = XXH64_createState();
static inline Bool s_suppressBlendedDepthWrite = false;
};
@@ -923,403 +923,11 @@ namespace MobileGL::MG_Backend::DirectVulkan {
ProgramFactory::CompileOptionFlags m_transformFlags;
};
// Decorates the module's captured varyings for VK_EXT_transform_feedback:
// user outputs get XfbBuffer/XfbStride/Offset directly; a captured
// gl_Position (a gl_PerVertex member) is mirrored into a dedicated output
// variable copied before every OpReturn, BEFORE the position fixup runs,
// so the captured value is the shader's own (pre-remap) gl_Position.
class XfbCaptureDecoratePass final : public spvtools::opt::Pass {
public:
struct CapturedVarying {
std::string name;
Uint32 bufferIndex = 0;
Uint32 offsetBytes = 0;
};
const char* name() const override { return "mobilegl-xfb-capture-decorate"; }
XfbCaptureDecoratePass(Vector<CapturedVarying> varyings, Vector<Uint32> strides)
: m_varyings(Move(varyings)), m_strides(Move(strides)) {}
Status Process() override {
using namespace spvtools::opt;
if (m_varyings.empty()) return Status::SuccessWithoutChange;
auto entryPointIter = get_module()->entry_points().begin();
if (entryPointIter == get_module()->entry_points().end()) return Status::SuccessWithoutChange;
spvtools::opt::Instruction* entryPoint = &*entryPointIter;
const Uint32 entryFunctionId = entryPoint->GetSingleWordInOperand(1);
// Name -> result id map from the debug section.
std::unordered_map<std::string, Uint32> idsByName;
for (auto& debugInst : get_module()->debugs2()) {
if (debugInst.opcode() != spv::Op::OpName) continue;
idsByName[debugInst.GetInOperand(1).AsString()] = debugInst.GetSingleWordInOperand(0);
}
auto* decorationManager = context()->get_decoration_mgr();
const auto decorateForXfb = [&](Uint32 targetId, Uint32 bufferIndex, Uint32 offsetBytes) {
const Uint32 stride = bufferIndex < m_strides.size() ? m_strides[bufferIndex] : 0;
decorationManager->AddDecorationVal(targetId, static_cast<Uint32>(spv::Decoration::XfbBuffer),
bufferIndex);
decorationManager->AddDecorationVal(targetId, static_cast<Uint32>(spv::Decoration::XfbStride),
stride);
decorationManager->AddDecorationVal(targetId, static_cast<Uint32>(spv::Decoration::Offset),
offsetBytes);
};
Bool modified = false;
Bool needsPositionMirror = false;
Uint32 positionBufferIndex = 0;
Uint32 positionOffset = 0;
for (const auto& varying : m_varyings) {
if (varying.name == "gl_Position") {
needsPositionMirror = true;
positionBufferIndex = varying.bufferIndex;
positionOffset = varying.offsetBytes;
continue;
}
const auto idIt = idsByName.find(varying.name);
if (idIt == idsByName.end()) {
MGLOG_E("XfbCaptureDecoratePass: no SPIR-V variable named '%s'", varying.name.c_str());
continue;
}
decorateForXfb(idIt->second, varying.bufferIndex, varying.offsetBytes);
modified = true;
}
if (needsPositionMirror) {
modified |= MirrorPositionForCapture(entryFunctionId, *entryPoint, positionBufferIndex,
positionOffset, decorateForXfb);
}
if (!modified) return Status::SuccessWithoutChange;
context()->AddCapability(spv::Capability::TransformFeedback);
{
auto executionMode = MakeUnique<spvtools::opt::Instruction>(
context(), spv::Op::OpExecutionMode, 0, 0,
std::initializer_list<spvtools::opt::Operand>{
{SPV_OPERAND_TYPE_ID, {entryPoint->GetSingleWordInOperand(1)}},
{SPV_OPERAND_TYPE_EXECUTION_MODE, {static_cast<Uint32>(spv::ExecutionMode::Xfb)}}});
get_module()->AddExecutionMode(Move(executionMode));
}
context()->InvalidateAnalysesExceptFor(spvtools::opt::IRContext::kAnalysisNone);
return Status::SuccessWithChange;
}
private:
template <typename DecorateFn>
Bool MirrorPositionForCapture(Uint32 entryFunctionId, spvtools::opt::Instruction& entryPoint,
Uint32 bufferIndex, Uint32 offsetBytes, const DecorateFn& decorateForXfb) {
const Uint32 entryPointModel = entryPoint.GetSingleWordInOperand(0);
using namespace spvtools::opt;
PositionTargetInfo target{};
if (!FindPositionTarget(context(), &target)) {
MGLOG_E("XfbCaptureDecoratePass: gl_Position capture requested but no position output found");
return false;
}
if (!target.isMember) {
// Standalone gl_Position variable: decorate it directly.
decorateForXfb(target.variableId, bufferIndex, offsetBytes);
return true;
}
auto* typeManager = context()->get_type_mgr();
const Uint32 mirrorPointerTypeId =
typeManager->FindPointerToType(target.vectorTypeId, spv::StorageClass::Output);
if (mirrorPointerTypeId == 0) return false;
const Uint32 mirrorVariableId = context()->TakeNextId();
auto mirrorVariable = MakeUnique<Instruction>(
context(), spv::Op::OpVariable, mirrorPointerTypeId, mirrorVariableId,
std::initializer_list<Operand>{
{SPV_OPERAND_TYPE_STORAGE_CLASS, {static_cast<Uint32>(spv::StorageClass::Output)}}});
get_module()->AddGlobalValue(Move(mirrorVariable));
// A free output location: past every explicitly decorated output.
Uint32 mirrorLocation = 0;
for (auto& annotation : get_module()->annotations()) {
if (annotation.opcode() != spv::Op::OpDecorate ||
annotation.GetSingleWordInOperand(1) != static_cast<Uint32>(spv::Decoration::Location)) {
continue;
}
mirrorLocation = std::max(mirrorLocation, annotation.GetSingleWordInOperand(2) + 1);
}
auto* decorationManager = context()->get_decoration_mgr();
decorationManager->AddDecorationVal(mirrorVariableId,
static_cast<Uint32>(spv::Decoration::Location), mirrorLocation);
decorateForXfb(mirrorVariableId, bufferIndex, offsetBytes);
entryPoint.AddOperand({SPV_OPERAND_TYPE_ID, {mirrorVariableId}});
auto* function = context()->GetFunction(entryFunctionId);
if (function == nullptr) return false;
const auto model = static_cast<spv::ExecutionModel>(entryPointModel);
Bool injected = false;
for (auto& block : *function) {
for (auto instIter = block.begin(); instIter != block.end(); ++instIter) {
// Geometry stages capture per emitted vertex; other stages at return.
const Bool isInjectionSite =
model == spv::ExecutionModel::Geometry
? instIter->opcode() == spv::Op::OpEmitVertex
: instIter->opcode() == spv::Op::OpReturn;
if (!isInjectionSite) continue;
InstructionBuilder builder(context(), &*instIter, IRContext::kAnalysisNone);
const Uint32 memberIndexId = builder.GetUintConstantId(target.memberIndex);
auto* access =
builder.AddAccessChain(target.vectorPtrTypeId, target.variableId, {memberIndexId});
if (access == nullptr) return injected;
auto* value = builder.AddLoad(target.vectorTypeId, access->result_id());
if (value == nullptr) return injected;
builder.AddStore(mirrorVariableId, value->result_id());
injected = true;
}
}
return injected;
}
Vector<CapturedVarying> m_varyings;
Vector<Uint32> m_strides;
};
// Adreno 650 (driver 512.502) faults the GPU on an implicit-LOD sample of a full-screen
// colour render target: the texture unit's derivative path reads outside the image's
// allocation even though the sampler clamps LOD to 0 and the mapping is 1:1. MobileGL's
// own default-framebuffer blit shader works around it with textureLod, but an
// application's shader (Minecraft's blit.fsh is `texture(InSampler, texCoord)`) cannot be
// edited - so rewrite the sample at the SPIR-V level instead.
//
// The rewrite is only requested for draws whose every sampler binding is clamped to one
// mip level, where explicit LOD 0 is exactly what the implicit form must already produce:
// lambda' = clamp(lambda + bias, minLod, maxLod) with minLod = maxLod = 0. Bias and MinLod
// operands are therefore dropped rather than translated.
class ForceExplicitLod0SamplePass final : public spvtools::opt::Pass {
public:
const char* name() const override { return "force-explicit-lod0-sample"; }
Status Process() override {
Bool isFragment = false;
for (auto& entryPoint : get_module()->entry_points()) {
if (entryPoint.opcode() != spv::Op::OpEntryPoint) continue;
if (static_cast<spv::ExecutionModel>(entryPoint.GetSingleWordInOperand(0)) ==
spv::ExecutionModel::Fragment) {
isFragment = true;
break;
}
}
if (!isFragment) return Status::SuccessWithoutChange;
// Plan first, mutate second. Materializing the LOD constant is itself a module
// change, so it must not happen unless at least one rewrite is going to follow -
// otherwise the pass would grow the binary while reporting SuccessWithoutChange.
Vector<RewritePlan> plans;
for (auto& function : *get_module()) {
for (auto& block : function) {
for (auto& inst : block) {
RewritePlan plan{};
if (PlanRewrite(&inst, plan)) plans.push_back(Move(plan));
}
}
}
if (plans.empty()) return Status::SuccessWithoutChange;
const Uint32 zeroId = GetFloatZeroId();
if (zeroId == 0) return Status::SuccessWithoutChange;
for (auto& plan : plans) {
plan.operands.push_back({SPV_OPERAND_TYPE_ID, {zeroId}});
for (auto& operand : plan.trailingOperands) {
plan.operands.push_back(operand);
}
plan.instruction->SetOpcode(plan.opcode);
plan.instruction->SetInOperands(Move(plan.operands));
}
// Opcodes and operand lists changed underneath every cached analysis.
context()->InvalidateAnalysesExceptFor(spvtools::opt::IRContext::kAnalysisNone);
return Status::SuccessWithChange;
}
private:
struct RewritePlan {
spvtools::opt::Instruction* instruction = nullptr;
spv::Op opcode = spv::Op::OpNop;
// Everything up to and including the Image Operands mask; the Lod id and the
// trailing operand values are appended once the constant exists.
Vector<spvtools::opt::Operand> operands;
Vector<spvtools::opt::Operand> trailingOperands;
};
// Image Operands bits that may accompany an implicit-LOD sample, in the canonical
// ascending order SPIR-V requires the operand values to appear in.
static constexpr Uint32 kBias = 0x1;
static constexpr Uint32 kLod = 0x2;
static constexpr Uint32 kGrad = 0x4;
static constexpr Uint32 kConstOffset = 0x8;
static constexpr Uint32 kOffset = 0x10;
static constexpr Uint32 kConstOffsets = 0x20;
static constexpr Uint32 kSample = 0x40;
static constexpr Uint32 kMinLod = 0x80;
static constexpr Uint32 kKnownMask = 0xFF;
Uint32 GetFloatZeroId() {
// Reuse a 32-bit float type already in the module; a shader that samples always has
// one, and looking it up avoids depending on type-creation API details.
Uint32 floatTypeId = 0;
for (auto& inst : get_module()->types_values()) {
if (inst.opcode() == spv::Op::OpTypeFloat && inst.NumInOperands() >= 1 &&
inst.GetSingleWordInOperand(0) == 32) {
floatTypeId = inst.result_id();
break;
}
}
if (floatTypeId == 0) return 0;
const auto* floatType = context()->get_type_mgr()->GetType(floatTypeId);
if (floatType == nullptr) return 0;
const auto zeroBits = std::bit_cast<Uint32>(0.0f);
const auto* zeroConst = context()->get_constant_mgr()->GetConstant(floatType, {zeroBits});
if (zeroConst == nullptr) return 0;
auto* zeroInst = context()->get_constant_mgr()->GetDefiningInstruction(zeroConst);
return zeroInst != nullptr ? zeroInst->result_id() : 0;
}
static Bool MapOpcode(spv::Op op, spv::Op& outOpcode, Uint32& outFixedOperandCount) {
switch (op) {
case spv::Op::OpImageSampleImplicitLod:
outOpcode = spv::Op::OpImageSampleExplicitLod;
outFixedOperandCount = 2; // sampled image, coordinate
return true;
case spv::Op::OpImageSampleProjImplicitLod:
outOpcode = spv::Op::OpImageSampleProjExplicitLod;
outFixedOperandCount = 2;
return true;
case spv::Op::OpImageSampleDrefImplicitLod:
outOpcode = spv::Op::OpImageSampleDrefExplicitLod;
outFixedOperandCount = 3; // sampled image, coordinate, Dref
return true;
case spv::Op::OpImageSampleProjDrefImplicitLod:
outOpcode = spv::Op::OpImageSampleProjDrefExplicitLod;
outFixedOperandCount = 3;
return true;
default:
return false;
}
}
static Bool PlanRewrite(spvtools::opt::Instruction* inst, RewritePlan& outPlan) {
spv::Op newOpcode = spv::Op::OpNop;
Uint32 fixedCount = 0;
if (!MapOpcode(inst->opcode(), newOpcode, fixedCount)) return false;
if (inst->NumInOperands() < fixedCount) return false;
Uint32 mask = 0;
Uint32 next = fixedCount;
if (inst->NumInOperands() > fixedCount) {
mask = inst->GetSingleWordInOperand(fixedCount);
next = fixedCount + 1;
}
// An operand this pass does not model would be silently reordered or dropped, and
// Grad cannot legally accompany an implicit-LOD sample: leave such an instruction be.
if ((mask & ~kKnownMask) != 0 || (mask & kGrad) != 0) return false;
Vector<spvtools::opt::Operand> fixedOperands;
fixedOperands.reserve(fixedCount + 1);
for (Uint32 i = 0; i < fixedCount; ++i) {
fixedOperands.push_back(inst->GetInOperand(i));
}
// Collect the surviving operand values in the same ascending-bit order they were
// encoded in, so the rebuilt list stays canonical.
Uint32 keptMask = kLod;
Vector<spvtools::opt::Operand> keptOperands;
static constexpr Uint32 kOrderedBits[] = {kBias, kLod, kGrad, kConstOffset,
kOffset, kConstOffsets, kSample, kMinLod};
for (const Uint32 bit : kOrderedBits) {
if ((mask & bit) == 0) continue;
if (next >= inst->NumInOperands()) return false;
const spvtools::opt::Operand value = inst->GetInOperand(next++);
// Bias and MinLod only shift a lambda that is already clamped to 0, and any
// original Lod is replaced by the constant the caller appends.
if (bit == kBias || bit == kMinLod || bit == kLod) continue;
keptMask |= bit;
keptOperands.push_back(value);
}
fixedOperands.push_back({SPV_OPERAND_TYPE_IMAGE, {keptMask}});
outPlan.instruction = inst;
outPlan.opcode = newOpcode;
outPlan.operands = Move(fixedOperands);
outPlan.trailingOperands = Move(keptOperands);
return true;
}
};
spvtools::Optimizer::PassToken CreateForceExplicitLod0SamplePass() {
return spvtools::Optimizer::PassToken(MakeUnique<ForceExplicitLod0SamplePass>());
}
Bool TransformSpirvForExplicitLod0Sampling(const Vector<Uint>& input, Vector<Uint>& output) {
if (input.empty()) {
output.clear();
return true;
}
spvtools::Optimizer optimizer(SPV_ENV_VULKAN_1_3);
spvtools::OptimizerOptions options;
// Matches the position-fix pass: this build of spirv-tools asserts rather than
// reporting, so validation stays off in the shipping path.
options.set_run_validator(false);
optimizer.SetMessageConsumer([](spv_message_level_t, const char*, const spv_position_t&,
const char* message) {
MGLOG_E("Vulkan: explicit-LOD0 pass: %s", message != nullptr ? message : "");
});
optimizer.RegisterPass(CreateForceExplicitLod0SamplePass());
const Bool success = optimizer.Run(input.data(), input.size(), &output, options);
if (!success) {
MGLOG_E("Vulkan: explicit-LOD0 sampling pass failed; keeping the original module");
output = input;
}
return success;
}
spvtools::Optimizer::PassToken CreateGlToVulkanPositionFixPass(
ProgramFactory::CompileOptionFlags transformFlags) {
return spvtools::Optimizer::PassToken(MakeUnique<GlToVulkanPositionFixPass>(transformFlags));
}
Bool TransformSpirvForXfbCapture(const Vector<Uint>& input, Vector<Uint>& output,
const MG_State::GLState::ProgramObject& program) {
if (input.empty()) {
output.clear();
return true;
}
Vector<XfbCaptureDecoratePass::CapturedVarying> varyings;
varyings.reserve(program.GetTransformFeedbackVaryingCount());
for (const auto& varying : program.GetTransformFeedbackVaryings()) {
varyings.push_back({varying.name, varying.bufferIndex, varying.offsetBytes});
}
Vector<Uint32> strides;
strides.reserve(program.GetTransformFeedbackBufferCount());
for (SizeT i = 0; i < program.GetTransformFeedbackBufferCount(); ++i) {
strides.push_back(program.GetTransformFeedbackStride(static_cast<Uint32>(i)));
}
spvtools::Optimizer optimizer(SPV_ENV_VULKAN_1_3);
spvtools::OptimizerOptions options;
options.set_run_validator(false);
optimizer.SetMessageConsumer([](spv_message_level_t, const char*, const spv_position_t&,
const char* message) {
MGLOG_E("Vulkan: xfb capture pass: %s", message != nullptr ? message : "");
});
optimizer.RegisterPass(spvtools::Optimizer::PassToken(
MakeUnique<XfbCaptureDecoratePass>(Move(varyings), Move(strides))));
const Bool success = optimizer.Run(input.data(), input.size(), &output, options);
if (!success) {
MGLOG_E("Vulkan: xfb capture decoration pass failed; keeping the original module");
output = input;
}
return success;
}
Bool TransformSpirvForVulkanPositionFix(const Vector<Uint>& input, Vector<Uint>& output,
ProgramFactory::CompileOptionFlags transformFlags) {
if (input.empty()) {
@@ -1761,25 +1369,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
XXHASH_VERIFY(XXH64_update(m_hashState, &binding, sizeof(binding)));
}
// The transform feedback capture layout is baked into the modules by
// XfbCaptureDecoratePass rather than coming from the SPIR-V, so it has to be part of
// the key: two programs can share every shader and still capture differently, which
// is exactly what changing the buffer mode does (glTransformFeedbackVaryings with the
// same varyings but GL_SEPARATE_ATTRIBS instead of GL_INTERLEAVED_ATTRIBS). Only
// hashed for a capturing compile, so nothing else changes key.
if (flags & CompileOptionBit::XfbCapture) {
for (const auto& varying : program.GetTransformFeedbackVaryings()) {
XXHASH_VERIFY(XXH64_update(m_hashState, varying.name.data(), varying.name.size()));
XXHASH_VERIFY(XXH64_update(m_hashState, &varying.bufferIndex, sizeof(varying.bufferIndex)));
XXHASH_VERIFY(XXH64_update(m_hashState, &varying.offsetBytes, sizeof(varying.offsetBytes)));
}
const SizeT bufferCount = program.GetTransformFeedbackBufferCount();
for (SizeT i = 0; i < bufferCount; ++i) {
const Uint32 stride = program.GetTransformFeedbackStride(static_cast<Uint32>(i));
XXHASH_VERIFY(XXH64_update(m_hashState, &stride, sizeof(stride)));
}
}
HashType hash = XXH64_digest(m_hashState);
return hash;
}
@@ -2361,17 +1950,11 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
auto it = m_cache.find(hash);
if (it != m_cache.end()) {
// Every draw/dispatch funnels through this lookup (the renderer memos only
// skip re-hashing, never the factory lookup), so an actively-used entry is
// stamped at least once per frame boundary and can never be aged out while
// any in-flight command buffer still references it.
it->second.lastUsedFrame = m_frameCounter;
return it->second;
}
auto& entry = m_cache[hash];
entry.hash = hash;
entry.lastUsedFrame = m_frameCounter;
auto& shaders = program.GetAttachedShaders();
auto& spirv = program.GetGeneratedSpirv();
Vector<Vector<Uint>> moduleSpirvs(spirv.size());
@@ -2384,40 +1967,11 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// Apply position fixup if needed
if (fixupStage != ShaderStage::Unknown && shaders[i] && shaders[i]->GetShaderStage() == fixupStage) {
const Vector<Uint>* fixupInput = &spv;
Vector<Uint> xfbSpirv;
if ((flags & ProgramFactory::CompileOptionBit::XfbCapture) &&
program.GetTransformFeedbackVaryingCount() > 0) {
// Decorate BEFORE the position fixup so a captured gl_Position
// mirror copies the shader's own (pre-remap) value.
if (TransformSpirvForXfbCapture(spv, xfbSpirv, program)) {
fixupInput = &xfbSpirv;
}
}
TransformSpirvForVulkanPositionFix(*fixupInput, moduleSpirvs[i], flags);
TransformSpirvForVulkanPositionFix(spv, moduleSpirvs[i], flags);
} else {
moduleSpirvs[i] = spv;
}
if ((flags & ProgramFactory::CompileOptionBit::ExplicitLod0Sampling) && shaders[i] &&
shaders[i]->GetShaderStage() == ShaderStage::Fragment) {
Vector<Uint> explicitLodSpirv;
if (TransformSpirvForExplicitLod0Sampling(moduleSpirvs[i], explicitLodSpirv)) {
moduleSpirvs[i] = Move(explicitLodSpirv);
}
}
// Vulkan's SPIR-V environment has no rectangle image dimension, so a
// GL_TEXTURE_RECTANGLE lookup has to become the 2D one the texture is really
// stored as - which addresses [0,1] where the application addressed texels.
{
Vector<Uint> rectLoweredSpirv;
if (MG_Util::ShaderTranspiler::ShaderCompiler::LowerRectImages(moduleSpirvs[i], rectLoweredSpirv) &&
!rectLoweredSpirv.empty()) {
moduleSpirvs[i] = Move(rectLoweredSpirv);
}
}
// GL apps depend on cross-program position invariance for multi-pass equality
// depth tests (MC 26.3's OIT re-draws the cloud geometry with GEQUAL against the
// depth its own first pass wrote); decorate Position outputs Invariant so
@@ -2514,42 +2068,4 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return entry;
}
void ProgramFactory::OnFrameBoundary() {
++m_frameCounter;
// Sweep cadence and retire age mirror VkRenderPassManager::OnPresent: an entry
// idle for more than kRetireAgeFrames frame boundaries cannot be referenced by
// any in-flight command buffer (frames-in-flight <= MOBILEGL_MAGMA_FRAMESINFLIGHT),
// so its shader modules and layouts are destroyed immediately - no deferred-
// destroy machinery needed. Eviction is content-based, never tied to
// glDeleteProgram: the cache is content-hash-shared across GL programs, so a
// delete-driven erase could free an entry another live program still resolves.
// An evicted entry self-heals - the frontend program keeps its generated
// SPIR-V, so the next GetOrCreateProgram rebuilds it (this also covers the
// renderer's internal blit/depth-mipmap programs).
constexpr Uint64 kSweepInterval = 256;
constexpr Uint64 kRetireAgeFrames = 1024;
if ((m_frameCounter % kSweepInterval) != 0) {
return;
}
for (auto it = m_cache.begin(); it != m_cache.end();) {
if (m_frameCounter - it->second.lastUsedFrame > kRetireAgeFrames) {
const HashType hash = it->first;
const VkDescriptorSetLayout descriptorSetLayout = it->second.descriptorSetLayout;
MGLOG_D("ProgramFactory::OnFrameBoundary: evicting idle program entry hash=0x%llx",
static_cast<unsigned long long>(hash));
// erase runs ~VkProgramObject (modules/layouts destroyed); notify after
// so an observer never observes a half-destroyed entry through a lookup.
// Observers only need the handle values to purge their keyed caches.
it = m_cache.erase(it);
if (m_evictionObserver != nullptr) {
m_evictionObserver->OnProgramEvicted(hash, descriptorSetLayout);
}
} else {
++it;
}
}
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -42,16 +42,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
SurfaceRotate90 = 1 << 2,
SurfaceRotate180 = 1 << 3,
SurfaceRotate270 = 1 << 4,
// Rewrites the fragment stage's implicit-LOD image samples to explicit LOD 0.
// Only ever set for a draw whose every sampler binding is clamped to a single mip
// level, which makes the two forms produce identical texels (the implicit lambda is
// clamped into [minLod, maxLod] = [0, 0] regardless of derivatives or bias).
ExplicitLod0Sampling = 1 << 5,
// Decorates the last vertex-processing stage's captured varyings with
// XfbBuffer/XfbStride/Offset (VK_EXT_transform_feedback). Set only for draws
// recorded while GL transform feedback is active, so plain draws keep the
// undecorated variant.
XfbCapture = 1 << 6,
};
using CompileOptionFlags = Flags<CompileOptionBit>;
using HashType = Uint64;
@@ -98,9 +88,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// gl_FragDepth); shader-computed depth is immune to the cross-pipeline
// position-invariance quirk (see PipelineFactory::ShouldSuppressDepthWrite).
Bool fragmentReplacesDepth = false;
// Frame-boundary counter value of the last GetOrCreateProgram hit; drives
// cache eviction (see OnFrameBoundary).
Uint64 lastUsedFrame = 0;
static inline VkDevice s_device = VK_NULL_HANDLE;
@@ -137,7 +124,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
producerOutputComponentCount = other.producerOutputComponentCount;
fragmentInputComponentCount = other.fragmentInputComponentCount;
fragmentReplacesDepth = other.fragmentReplacesDepth;
lastUsedFrame = other.lastUsedFrame;
other.hash = 0;
other.descriptorSetLayout = VK_NULL_HANDLE;
other.pipelineLayout = VK_NULL_HANDLE;
@@ -149,7 +135,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
other.producerOutputComponentCount = 0;
other.fragmentInputComponentCount = 0;
other.fragmentReplacesDepth = false;
other.lastUsedFrame = 0;
}
VkProgramObject& operator=(VkProgramObject&& other) noexcept {
if (this == &other) {
@@ -185,7 +170,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
producerOutputComponentCount = other.producerOutputComponentCount;
fragmentInputComponentCount = other.fragmentInputComponentCount;
fragmentReplacesDepth = other.fragmentReplacesDepth;
lastUsedFrame = other.lastUsedFrame;
other.hash = 0;
other.descriptorSetLayout = VK_NULL_HANDLE;
other.pipelineLayout = VK_NULL_HANDLE;
@@ -197,7 +181,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
other.producerOutputComponentCount = 0;
other.fragmentInputComponentCount = 0;
other.fragmentReplacesDepth = false;
other.lastUsedFrame = 0;
return *this;
}
@@ -227,18 +210,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
};
// Notified when the OnFrameBoundary sweep destroys an aged-out cache entry,
// carrying the entry's content hash and the VkDescriptorSetLayout it owned.
// Dependent caches (compute pipelines, PipelineFactory entries, UniformManager's
// per-layout descriptor sets) must purge in the same step: after vkDestroy the
// layout handle value may be recycled for an unrelated layout, and the program
// hash may be re-inserted by a later rebuild of the same content.
class IEvictionObserver {
public:
virtual ~IEvictionObserver() = default;
virtual void OnProgramEvicted(HashType programHash, VkDescriptorSetLayout descriptorSetLayout) = 0;
};
explicit ProgramFactory(VkDevice device, const VulkanRendererConfig& config, Uint32 maxBindings = 16,
Bool shaderDrawParametersEnabled = false,
Bool unformattedFloatStorageImagesEnabled = false)
@@ -254,13 +225,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const VkProgramObject& GetOrCreateProgram(
const MG_State::GLState::ProgramObject& program, CompileOptionFlags flags);
// Observer may be null (no notifications). Not owned.
void SetEvictionObserver(IEvictionObserver* observer) { m_evictionObserver = observer; }
// Frame boundary hook: ages the program cache and evicts long-unused entries
// (their command buffers retired many frames ago), mirroring
// VkRenderPassManager::OnPresent's sweep.
void OnFrameBoundary();
static VkShaderStageFlagBits ToVkStage(ShaderStage stage);
static VkFormat ConvertSpirvImageFormatToVkFormat(SpvImageFormat format);
static SamplerNumericDomain UniformTypeToSamplerNumericDomain(GLenum glType);
@@ -302,9 +266,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// shaderStorageImageReadWithoutFormat and shaderStorageImageWriteWithoutFormat.
Bool m_unformattedFloatStorageImagesEnabled = false;
mutable ProgramLookupCache m_lastLookup;
// Monotonic frame-boundary counter (bumped in OnFrameBoundary) for cache aging.
Uint64 m_frameCounter = 0;
IEvictionObserver* m_evictionObserver = nullptr;
static inline XXH64_state_t* m_hashState = XXH64_createState();
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -247,11 +247,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_surfaceFormat = {createInfo.imageFormat, createInfo.imageColorSpace};
m_extent = createInfo.imageExtent;
// The surface-space extent this swapchain was built from, i.e. before the
// quarter-turn swap above. Out-of-date checks must compare in THIS space: comparing a
// freshly queried currentExtent against the swapped m_extent flips axes every rotation
// and makes the comparison alternate forever.
m_surfaceExtent = defaultFramebufferExtent;
m_preTransform = createInfo.preTransform;
VK_VERIFY(vkCreateSwapchainKHR(device, &createInfo, nullptr, &m_swapchain));
@@ -262,9 +257,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_images.resize(imageCount, VK_NULL_HANDLE);
VK_VERIFY(vkGetSwapchainImagesKHR(device, m_swapchain, &imageCount, m_images.data()));
m_imageLayouts.assign(imageCount, VK_IMAGE_LAYOUT_UNDEFINED);
// Fresh swapchain images hold garbage until a render pass stores into them.
m_imageContentDefined.assign(imageCount, false);
m_depthStencilContentDefined.assign(imageCount, false);
CreateImageViews(device);
CreateDepthStencilResources(device, physicalDevice);
@@ -436,39 +428,9 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_images.clear();
m_imageLayouts.clear();
m_imageContentDefined.clear();
m_depthStencilContentDefined.clear();
m_preTransform = VK_SURFACE_TRANSFORM_IDENTITY_BIT_KHR;
}
Bool SwapchainObject::IsImageContentDefined(Uint32 index) const {
MOBILEGL_ASSERT(index < m_imageContentDefined.size(), "Swapchain image content index out of range");
return m_imageContentDefined[index];
}
void SwapchainObject::SetImageContentDefined(Uint32 index, Bool defined) {
MOBILEGL_ASSERT(index < m_imageContentDefined.size(), "Swapchain image content index out of range");
m_imageContentDefined[index] = defined;
}
Bool SwapchainObject::IsDepthStencilContentDefined(Uint32 index) const {
MOBILEGL_ASSERT(index < m_depthStencilContentDefined.size(),
"Swapchain depth/stencil content index out of range");
return m_depthStencilContentDefined[index];
}
void SwapchainObject::SetDepthStencilContentDefined(Uint32 index, Bool defined) {
MOBILEGL_ASSERT(index < m_depthStencilContentDefined.size(),
"Swapchain depth/stencil content index out of range");
m_depthStencilContentDefined[index] = defined;
}
void SwapchainObject::SetAllDepthStencilContentUndefined() {
for (SizeT i = 0; i < m_depthStencilContentDefined.size(); ++i) {
m_depthStencilContentDefined[i] = false;
}
}
VkImage SwapchainObject::GetImage(Uint32 index) const {
MOBILEGL_ASSERT(index < m_images.size(), "Swapchain image index out of range");
return m_images[index];
@@ -35,9 +35,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkSwapchainKHR GetHandle() const { return m_swapchain; }
const VkSurfaceFormatKHR& GetSurfaceFormat() const { return m_surfaceFormat; }
VkExtent2D GetExtent() const { return m_extent; }
// Surface-space extent (before the pre-rotation quarter-turn swap) this swapchain was
// created from - the value to compare a freshly queried currentExtent against.
VkExtent2D GetSurfaceExtent() const { return m_surfaceExtent; }
VkSurfaceTransformFlagBitsKHR GetPreTransform() const { return m_preTransform; }
const Vector<VkImage>& GetImages() const { return m_images; }
const Vector<VkImageView>& GetImageViews() const { return m_imageViews; }
@@ -52,21 +49,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
void SetImageLayout(Uint32 index, VkImageLayout layout);
SizeT GetImageCount() const { return m_images.size(); }
// EGL content-validity tracking for the default framebuffer. A color
// buffer's content is undefined once its image has been presented
// (EGL_BUFFER_DESTROYED swap behaviour, the implementation default),
// and every ancillary (depth/stencil) buffer's content is undefined
// after ANY swap regardless of swap behaviour (EGL 1.5 §3.10.1). The
// render-pass manager turns an undefined attachment's tile load into
// LOAD_OP_DONT_CARE. Flags start false (a fresh swapchain image holds
// garbage) and a render pass storing into an attachment sets it back
// to defined.
Bool IsImageContentDefined(Uint32 index) const;
void SetImageContentDefined(Uint32 index, Bool defined);
Bool IsDepthStencilContentDefined(Uint32 index) const;
void SetDepthStencilContentDefined(Uint32 index, Bool defined);
void SetAllDepthStencilContentUndefined();
private:
void CreateImageViews(VkDevice device);
void CreateDepthStencilResources(VkDevice device, VkPhysicalDevice physicalDevice);
@@ -81,7 +63,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkSwapchainKHR m_swapchain = VK_NULL_HANDLE;
VkSurfaceFormatKHR m_surfaceFormat{};
VkExtent2D m_extent{};
VkExtent2D m_surfaceExtent{};
VkSurfaceTransformFlagBitsKHR m_preTransform = VK_SURFACE_TRANSFORM_IDENTITY_BIT_KHR;
Vector<VkImage> m_images;
Vector<VkImageView> m_imageViews;
@@ -92,7 +73,5 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Vector<VkDeviceMemory> m_depthStencilImageMemories;
Vector<VkImageView> m_depthStencilImageViews;
Vector<VkImageLayout> m_depthStencilImageLayouts;
Vector<Bool> m_imageContentDefined;
Vector<Bool> m_depthStencilContentDefined;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -18,7 +18,6 @@
#include "MG_Util/Converters/MGToVk/TextureEnumConverter.h"
#include "MG_Util/Metrics/TextureMetrics.h"
#include <Config.h>
#include <algorithm>
#include <cstdio>
#include <cstdlib>
#include <cstring>
@@ -205,7 +204,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// The frame's descriptor sets are recycled above, so last frame's reuse target
// is gone: start the per-draw descriptor-reuse cache fresh this frame.
m_hasLastDescriptor = false;
m_lastBindValid = false;
// Re-fingerprint the bound sampler set fresh this frame so any GL object address
// reuse cannot outlive a single frame (see SamplerResolveMemo).
for (auto& memo : m_samplerResolveMemo) {
@@ -213,40 +211,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
}
void UniformManager::OnDescriptorSetLayoutDestroyed(VkDescriptorSetLayout descriptorSetLayout) {
SizeT purgedSets = 0;
for (auto& frame : m_frames) {
const auto it = frame.descriptorSetCacheByLayout.find(descriptorSetLayout);
if (it == frame.descriptorSetCacheByLayout.end()) {
continue;
}
// Free the sets back to their pools and credit the bucket accounting, so
// program churn recycles pool capacity instead of abandoning the slots.
// GPU-safe: the layout only dies after >1024 idle frame boundaries, so no
// in-flight command buffer references these sets.
for (const auto& cached : it->second.sets) {
if (cached.set == VK_NULL_HANDLE) {
continue;
}
vkFreeDescriptorSets(m_device, cached.pool, 1, &cached.set);
const auto bucket = std::find_if(
frame.descriptorPools.begin(), frame.descriptorPools.end(),
[&cached](const DescriptorPoolBucket& candidate) { return candidate.handle == cached.pool; });
if (bucket != frame.descriptorPools.end() && bucket->allocatedSets > 0) {
--bucket->allocatedSets;
}
}
purgedSets += it->second.sets.size();
frame.descriptorSetCacheByLayout.erase(it);
}
if (purgedSets > 0) {
// The per-draw reuse memo folds the layout handle into its signature; drop
// it so a recycled handle value cannot revive a purged set mid-frame.
m_hasLastDescriptor = false;
MGLOG_D("UniformDescriptorBinder: freed %zu descriptor sets for destroyed layout", purgedSets);
}
}
Bool UniformManager::ResolveSamplerDescriptor(VkCommandBuffer commandBuffer,
const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj,
@@ -266,24 +230,12 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const auto& samplerOverride = textureUnit.GetSamplerObject();
const auto preferredTarget = programObj.samplerTextureTargetByBinding[binding];
SharedPtr<MG_State::GLState::ITextureObject> fallbackHolder;
// A texture that fails the completeness rules for the filter in effect reads
// (0, 0, 0, 1), which is exactly what the fallback texture holds - so it takes the
// same route as a sampler with nothing bound.
if (texture != nullptr &&
MG_State::GLState::SamplesAsIncompleteTexture(
texture, samplerOverride ? samplerOverride.get() : texture->GetSamplerObject().get())) {
texture = nullptr;
}
if (texture == nullptr) {
fallbackHolder = GetFallbackTexture(preferredTarget);
texture = fallbackHolder.get();
if (texture == nullptr) {
MGLOG_E("ResolveSamplerDescriptor: no fallback texture available for binding=%u ('%s') "
"location=%d unit=%d target=%d",
binding, programObj.samplerNameByBinding[binding].c_str(), location, unit,
static_cast<Int>(preferredTarget));
return false;
}
MOBILEGL_ASSERT(texture != nullptr,
"ResolveSamplerDescriptor: no fallback texture available for binding=%u location=%d unit=%d target=%d",
binding, location, unit, static_cast<Int>(preferredTarget));
MGLOG_W(
"ResolveSamplerDescriptor: using fallback texture for unbound sampler binding=%u ('%s') location=%d unit=%d target=%d",
binding, programObj.samplerNameByBinding[binding].c_str(), location, unit,
@@ -398,28 +350,24 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const Uint16 samplerVersion = samplerToUse->GetVersion();
const Uint64 textureLifetimeId = texture->GetLifetimeId();
const Uint16 textureParamsVersion = texture->GetTextureParamsVersion();
// The sampler's LOD clamp depends on how many levels the sampled view exposes, and that
// follows uploads as well as GL parameters - so it belongs in the memo key too.
const Uint32 viewLevelCount = resource->sampledLevelCount;
if (memo.valid && memo.samplerLifetimeId == samplerLifetimeId && memo.samplerVersion == samplerVersion &&
memo.textureLifetimeId == textureLifetimeId && memo.textureParamsVersion == textureParamsVersion &&
memo.forceNearestFiltering == forceNearestFiltering && memo.viewLevelCount == viewLevelCount) {
memo.forceNearestFiltering == forceNearestFiltering) {
resolvedSampler = memo.sampler;
} else {
resolvedSampler = m_samplerManager->GetOrCreateSampler(*samplerToUse, *texture,
forceNearestFiltering, viewLevelCount);
resolvedSampler =
m_samplerManager->GetOrCreateSampler(*samplerToUse, *texture, forceNearestFiltering);
memo.samplerLifetimeId = samplerLifetimeId;
memo.samplerVersion = samplerVersion;
memo.textureLifetimeId = textureLifetimeId;
memo.textureParamsVersion = textureParamsVersion;
memo.forceNearestFiltering = forceNearestFiltering;
memo.viewLevelCount = viewLevelCount;
memo.sampler = resolvedSampler;
memo.valid = true;
}
} else {
resolvedSampler = m_samplerManager->GetOrCreateSampler(*samplerToUse, *texture, forceNearestFiltering,
resource->sampledLevelCount);
resolvedSampler =
m_samplerManager->GetOrCreateSampler(*samplerToUse, *texture, forceNearestFiltering);
}
outImageInfo = {
.sampler = resolvedSampler,
@@ -460,48 +408,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return outImageInfo.sampler != VK_NULL_HANDLE;
}
Bool UniformManager::ProgramSamplesOnlySingleLevelTextures(
const MG_State::GLState::ProgramObject& program, const ProgramFactory::VkProgramObject& programObj) {
Bool sawSampler = false;
for (Uint32 binding = 0; binding < programObj.bindingKinds.size(); ++binding) {
if (programObj.bindingKinds[binding] != ProgramFactory::DescriptorBindingKind::CombinedImageSampler) {
continue;
}
const auto* texture = ResolveSamplerTextureRaw(program, programObj, binding);
if (texture == nullptr) return false;
const auto& levelRange = texture->GetLevelRange();
if (levelRange.x() != levelRange.y()) return false;
// An explicit-LOD sample is a single filtered tap, so it also gives up anisotropic
// filtering - which a single-level view can still have. Resolve the sampler exactly
// the way ResolveSamplerDescriptor does and bail if anisotropy would apply.
const Int location = programObj.samplerUniformLocationByBinding[binding];
const Int unit = ResolveSamplerUnitIndex(program, location, binding);
const auto& samplerOverride = MG_State::pGLContext->GetTextureUnitObject(unit).GetSamplerObject();
const auto* effectiveSampler =
samplerOverride ? samplerOverride.get() : texture->GetSamplerObject().get();
if (effectiveSampler == nullptr) return false;
if (effectiveSampler->GetMaxAnisotropy() > 1.0f &&
effectiveSampler->GetMinFilter() == SamplerFilterMode::Linear &&
effectiveSampler->GetMagFilter() == SamplerFilterMode::Linear) {
return false;
}
// An explicit LOD 0 makes lambda exactly 0, which is the magnification side of the
// min/mag decision. That only matches the implicit form when lambda could not have been
// positive anyway (the LOD clamp already pins it at or below 0), or when the two
// filters are the same and the choice cannot be observed.
const Float effectiveMaxLod = effectiveSampler->GetMipmapMode() == SamplerMipmapMode::None
? 0.0f
: effectiveSampler->GetMaxLod();
if (effectiveMaxLod > 0.0f && effectiveSampler->GetMinFilter() != effectiveSampler->GetMagFilter()) {
return false;
}
sawSampler = true;
}
return sawSampler;
}
Bool UniformManager::ResolveSamplerTexture(const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj, Uint32 binding,
SharedPtr<MG_State::GLState::ITextureObject>& outTexture) {
@@ -604,11 +510,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const VkDeviceSize texelSize =
static_cast<VkDeviceSize>(MG_Util::GetSizedInternalFormatSizeInBytes(internalFormat));
// glTextureBufferRange addresses a window of the buffer, not all of it; the whole-buffer
// forms report the buffer's current size here, so both go through the same clamp.
const VkDeviceSize rangeOffset = static_cast<VkDeviceSize>(textureBuffer->GetBufferRangeOffset());
const VkDeviceSize rangeSize = static_cast<VkDeviceSize>(textureBuffer->GetBufferRangeSizeInBytes());
VkDeviceSize viewRange = std::min(rangeSize, slice.size > rangeOffset ? slice.size - rangeOffset : 0);
VkDeviceSize viewRange = slice.size;
if (texelSize > 0) {
viewRange = (viewRange / texelSize) * texelSize;
}
@@ -621,7 +523,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
viewInfo.sType = VK_STRUCTURE_TYPE_BUFFER_VIEW_CREATE_INFO;
viewInfo.buffer = slice.buffer;
viewInfo.format = vkFormat;
viewInfo.offset = slice.offset + rangeOffset;
viewInfo.offset = slice.offset;
viewInfo.range = viewRange;
VkBufferView bufferView = VK_NULL_HANDLE;
@@ -666,14 +568,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return false;
}
// The shader may write this buffer, and those writes land in GPU memory behind the
// frontend's CPU shadow - which is what MapBuffer and GetBufferSubData read.
// Host-visible coherent GPU residency makes the shadow BE that memory, so the
// results are visible without a readback path, exactly as for a capture buffer.
bufferObject->EnsureGpuResidentStorage();
// ... and the read that follows has to wait for this draw or dispatch to retire.
bufferObject->MarkGpuWritten();
BufferSlice slice{};
if (!m_bufferManager->AcquireResidentSlice(BufferKind::ShaderStorage, bufferObject, slice) || !slice.IsValid()) {
MGLOG_E("ResolveStorageBufferDescriptor: failed to sync GL buffer %u for block '%s'",
@@ -776,27 +670,15 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
SharedPtr<MG_State::GLState::ITextureObject> UniformManager::GetFallbackTexture(TextureTarget target) const {
// The fallback is a single-sampled 2D image, so it can only stand in for a sampler that
// would accept one. A multisample sampler in particular cannot: its descriptor demands a
// multisample view, and handing it this one is invalid Vulkan, not a degraded picture.
// Report that there is no fallback and let the caller decline the draw - aborting the
// process over an unbound sampler is never the right answer.
if (target != TextureTarget::Texture2D && target != TextureTarget::TextureRectangle) {
MGLOG_E("UniformManager::GetFallbackTexture: no fallback exists for target=%d",
static_cast<Int>(target));
return nullptr;
}
MOBILEGL_ASSERT(target == TextureTarget::Texture2D || target == TextureTarget::TextureRectangle,
"UniformManager::GetFallbackTexture: unsupported fallback target=%d",
static_cast<Int>(target));
if (m_fallbackTexture2D == nullptr) {
auto fallbackTexture = MakeShared<MG_State::GLState::TextureObject2D>(kFallbackTexture2DExternalIndex);
fallbackTexture->SetInternalFormat(TextureInternalFormat::RGBA8);
fallbackTexture->AllocateStorage(TextureUploadTarget::Texture2D, 0,
{.texelSize = {1, 1, 1}, .byteSize = 4});
// (0, 0, 0, 1): what GL reads from a texture that is not complete, and the only
// sensible answer for a sampler with nothing bound.
static Uint8 kOpaqueBlackTexel[4] = {0, 0, 0, 255};
fallbackTexture->UpdateMipmapSubData(TextureUploadTarget::Texture2D, 0,
{kOpaqueBlackTexel, sizeof(kOpaqueBlackTexel)});
fallbackTexture->MarkStorageDirty(TextureUploadTarget::Texture2D, 0, true);
m_fallbackTexture2D = fallbackTexture;
}
@@ -1029,11 +911,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkDescriptorPoolCreateInfo poolInfo{};
poolInfo.sType = VK_STRUCTURE_TYPE_DESCRIPTOR_POOL_CREATE_INFO;
// FREE_DESCRIPTOR_SET_BIT lets a destroyed layout's cached sets be freed back
// (OnDescriptorSetLayoutDestroyed) so program churn recycles pool capacity.
// The cost is on set allocation only, which happens when a layout's per-frame
// cache grows - never on the per-draw reuse path.
poolInfo.flags = VK_DESCRIPTOR_POOL_CREATE_FREE_DESCRIPTOR_SET_BIT;
poolInfo.maxSets = maxSets;
poolInfo.poolSizeCount = static_cast<Uint32>(std::size(poolSizes));
poolInfo.pPoolSizes = poolSizes;
@@ -1113,7 +990,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
auto& frame = m_frames[frameIndex];
auto& cache = frame.descriptorSetCacheByLayout[programObj.descriptorSetLayout];
if (cache.cursor < cache.sets.size()) {
outDescriptorSet = cache.sets[cache.cursor++].set;
outDescriptorSet = cache.sets[cache.cursor++];
} else {
VkResult allocResult = AllocateDescriptorSetsFromActivePool(frameIndex, programObj, outDescriptorSet);
if (allocResult == VK_ERROR_OUT_OF_POOL_MEMORY || allocResult == VK_ERROR_FRAGMENTED_POOL) {
@@ -1127,9 +1004,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return allocResult;
}
// The successful allocation came from the bucket the alloc helper left
// active; record it so a layout-destroyed purge can free the set back.
cache.sets.push_back({outDescriptorSet, frame.descriptorPools[frame.activeDescriptorPoolIndex].handle});
cache.sets.push_back(outDescriptorSet);
++cache.cursor;
MGLOG_D("UniformDescriptorBinder: cached descriptor set count for frame=%u grew to %zu", frameIndex,
cache.sets.size());
@@ -1227,47 +1102,16 @@ namespace MobileGL::MG_Backend::DirectVulkan {
bufferInfo.range = ubo.range;
dynOffset = static_cast<Uint32>(ubo.dynamicOffset);
} else {
// Global-UBO slice reuse (see GlobalUboSliceMemo): unchanged
// uniform bytes re-use the slice already uploaded this frame.
const Bool isGlobalUbo =
programObj.globalUboBinding == static_cast<Int>(binding) && element == 0;
const Uint64 uboFrameSerial = m_bufferManager->GetFrameSerial();
const Uint64 uboProgramLifetimeId = program.GetLifetimeId();
const Uint32 uboContentVersion = program.GetUBOContentVersion();
Bool reusedSlice = false;
if (isGlobalUbo) {
for (const auto& memo : m_globalUboMemo) {
if (memo.buffer != VK_NULL_HANDLE &&
memo.programLifetimeId == uboProgramLifetimeId &&
memo.frameSerial == uboFrameSerial &&
memo.uboContentVersion == uboContentVersion &&
memo.range == static_cast<VkDeviceSize>(ubo.payloadSize)) {
bufferInfo.buffer = memo.buffer;
bufferInfo.range = memo.range;
dynOffset = static_cast<Uint32>(memo.offset);
reusedSlice = true;
break;
}
}
}
if (!reusedSlice) {
BufferSlice slice{};
if (!m_bufferManager->UploadTransient(BufferKind::Uniform, frameIndex, ubo.payload,
ubo.payloadSize, m_minDynamicOffsetAlignment, slice)) {
MOBILEGL_ASSERT(false, "UniformDescriptorBinder::BindProgramUniformBuffers failed: UBO upload failed on binding %u element %u",
binding, element);
return false;
}
bufferInfo.buffer = slice.buffer;
bufferInfo.range = ubo.payloadSize;
dynOffset = static_cast<Uint32>(slice.offset);
if (isGlobalUbo) {
m_globalUboMemo[m_globalUboMemoNext] = GlobalUboSliceMemo{
uboProgramLifetimeId, uboFrameSerial, uboContentVersion,
slice.buffer, slice.offset, static_cast<VkDeviceSize>(ubo.payloadSize)};
m_globalUboMemoNext = (m_globalUboMemoNext + 1) % kGlobalUboMemoSize;
}
BufferSlice slice{};
if (!m_bufferManager->UploadTransient(BufferKind::Uniform, frameIndex, ubo.payload,
ubo.payloadSize, m_minDynamicOffsetAlignment, slice)) {
MOBILEGL_ASSERT(false, "UniformDescriptorBinder::BindProgramUniformBuffers failed: UBO upload failed on binding %u element %u",
binding, element);
return false;
}
bufferInfo.buffer = slice.buffer;
bufferInfo.range = ubo.payloadSize;
dynOffset = static_cast<Uint32>(slice.offset);
}
bufferInfos.push_back(bufferInfo);
// Dynamic offsets are consumed in binding order, then array element order,
@@ -1406,34 +1250,8 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_hasLastDescriptor = cacheable;
}
// Skip the driver call when this exact binding is already live on the
// command buffer (see the bind-dedup shadow in the header).
const Uint32 offsetCount = static_cast<Uint32>(dynamicOffsets.size());
Bool identicalBind = m_lastBindValid && m_lastBindSet == descriptorSet &&
m_lastBindLayout == programObj.pipelineLayout && m_lastBindPoint == bindPoint &&
m_lastBindOffsetCount == offsetCount && offsetCount <= kMaxShadowedDynamicOffsets;
if (identicalBind) {
for (Uint32 i = 0; i < offsetCount; ++i) {
if (m_lastBindOffsets[i] != dynamicOffsets[i]) {
identicalBind = false;
break;
}
}
}
if (!identicalBind) {
vkCmdBindDescriptorSets(commandBuffer, bindPoint, programObj.pipelineLayout, 0, 1,
&descriptorSet, offsetCount, dynamicOffsets.data());
if (offsetCount <= kMaxShadowedDynamicOffsets) {
m_lastBindValid = true;
m_lastBindSet = descriptorSet;
m_lastBindLayout = programObj.pipelineLayout;
m_lastBindPoint = bindPoint;
m_lastBindOffsetCount = offsetCount;
std::copy_n(dynamicOffsets.data(), offsetCount, m_lastBindOffsets);
} else {
m_lastBindValid = false;
}
}
vkCmdBindDescriptorSets(commandBuffer, bindPoint, programObj.pipelineLayout, 0, 1,
&descriptorSet, static_cast<Uint32>(dynamicOffsets.size()), dynamicOffsets.data());
return true;
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -39,20 +39,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
void Shutdown();
void BeginFrame(Uint32 frameIndex);
// A command buffer (re)began recording: descriptor bindings recorded into
// the previous buffer do not carry over, so drop the bind-dedup shadow.
void OnCommandBufferBoundary() { m_lastBindValid = false; }
// A ProgramFactory eviction just destroyed this layout: purge every frame
// slot's cached descriptor sets for it, so a recycled handle value can never
// stale-hit sets written for the dead layout's bindings. The sets are
// vkFreeDescriptorSets'd back to their pools (created with
// FREE_DESCRIPTOR_SET_BIT) and the pool accounting is credited, so program
// churn recycles pool capacity instead of abandoning it. GPU-safe: the layout
// only dies after >1024 idle frame boundaries, so no in-flight command buffer
// references its sets. This is the only eviction path for the per-layout
// caches - a live layout's entry must never be purged (its sets would be
// unreachable pool slots), so there is deliberately no age-based sweep here.
void OnDescriptorSetLayoutDestroyed(VkDescriptorSetLayout descriptorSetLayout);
Bool CollectSampledTextures(const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj,
Vector<MG_State::GLState::ITextureObject*>& outTextures);
@@ -72,16 +58,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
static VkFormat ResolveStorageImageViewFormat(VkFormat reflectedFormat, GLenum bindingFormat,
VkFormat resourceFormat, Bool useBindingFormat);
// True when the program reads at least one sampler and every one of them is bound to a
// texture whose GL level range is a single level. Such a sampler resolves to
// minLod = maxLod = 0 (see VkSamplerManager::GetOrCreateSampler), so an implicit-LOD sample
// and an explicit LOD 0 sample must read the same texel - which is what makes the
// ExplicitLod0Sampling SPIR-V rewrite safe to request. Deliberately conservative: it reads
// only GL state, so a texture that ends up single-level for another reason (one uploaded
// level under a wide level range) merely misses the rewrite.
static Bool ProgramSamplesOnlySingleLevelTextures(const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj);
private:
struct DescriptorPoolBucket {
VkDescriptorPool handle = VK_NULL_HANDLE;
@@ -89,16 +65,8 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Uint32 allocatedSets = 0;
};
// A cached descriptor set together with the pool it was allocated from, so a
// layout-destroyed purge can vkFreeDescriptorSets it back and credit the
// owning bucket's accounting.
struct CachedDescriptorSet {
VkDescriptorSet set = VK_NULL_HANDLE;
VkDescriptorPool pool = VK_NULL_HANDLE;
};
struct DescriptorSetCacheEntry {
Vector<CachedDescriptorSet> sets;
Vector<VkDescriptorSet> sets;
Uint32 cursor = 0;
};
@@ -188,35 +156,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Uint64 m_lastDescriptorSignature = 0;
Bool m_hasLastDescriptor = false;
// vkCmdBindDescriptorSets dedup: consecutive draws with a static uniform
// block resolve to the same set AND the same dynamic offsets, so the
// driver call can be skipped outright. Command-buffer-scope state; reset
// via OnCommandBufferBoundary whenever a recording (re)begins. Keyed on
// layout+bind point, so a pipeline-layout switch always rebinds.
static constexpr Uint32 kMaxShadowedDynamicOffsets = 8;
Bool m_lastBindValid = false;
VkDescriptorSet m_lastBindSet = VK_NULL_HANDLE;
VkPipelineLayout m_lastBindLayout = VK_NULL_HANDLE;
VkPipelineBindPoint m_lastBindPoint = VK_PIPELINE_BIND_POINT_GRAPHICS;
Uint32 m_lastBindOffsetCount = 0;
Uint32 m_lastBindOffsets[kMaxShadowedDynamicOffsets] = {};
// Global-UBO transient-slice reuse: MC leaves the default uniform block
// untouched across long GUI/terrain runs, so the per-draw re-upload of
// the same bytes can reuse the slice uploaded earlier THIS frame (frame
// serial guards arena recycling; the content version guards writes).
struct GlobalUboSliceMemo {
Uint64 programLifetimeId = 0;
Uint64 frameSerial = 0;
Uint32 uboContentVersion = 0;
VkBuffer buffer = VK_NULL_HANDLE;
VkDeviceSize offset = 0;
VkDeviceSize range = 0;
};
static constexpr Uint32 kGlobalUboMemoSize = 4;
GlobalUboSliceMemo m_globalUboMemo[kGlobalUboMemoSize];
Uint32 m_globalUboMemoNext = 0;
// Per-binding fast path over VkSamplerManager's content-hashed sampler cache, which
// stays the source of truth: its key hashes all sampler+texture state, so two distinct
// sampler objects with identical state still resolve to one VkSampler. This memo only
@@ -233,7 +172,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Uint64 samplerLifetimeId = 0;
Uint64 textureLifetimeId = 0;
VkSampler sampler = VK_NULL_HANDLE;
Uint32 viewLevelCount = 0;
Uint16 samplerVersion = 0;
Uint16 textureParamsVersion = 0;
Bool forceNearestFiltering = false;
@@ -32,13 +32,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.IsBgra, sizeof(attr.IsBgra)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.Divisor, sizeof(attr.Divisor)));
// The buffer's heap address is an identity component of the key: a freed
// buffer's reused address can alias an old cache entry, but only under a
// byte-identical attribute layout - and the entry payload is a pure function
// of the hashed inputs, with the draw path re-resolving bindingBufferKeys
// against the live VAO attribute pointers, so an aliased hit returns exactly
// what a rebuild would. Address drift only grows the map; the OnFrameBoundary
// aging sweep bounds that.
const SizeT bufferKey = reinterpret_cast<SizeT>(attr.Buffer.get());
XXHASH_VERIFY(XXH64_update(m_hashState, &bufferKey, sizeof(bufferKey)));
}
@@ -58,27 +51,14 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const VertexInputStateFactory::BackendVertexInputState& VertexInputStateFactory::GetOrCreateVertexInputState(
const MG_State::GLState::VertexArrayObject& vao) {
// Per-draw fast path: the VAO carries a pointer to its resolved entry,
// valid while its config version and the cache's eviction epoch both
// match - no re-hash, no map lookup.
const void* memoState = nullptr;
Uint64 memoEpoch = 0;
if (vao.GetBackendStateMemo(memoState, memoEpoch) && memoEpoch == m_evictionEpoch) {
const auto* entry = static_cast<const BackendVertexInputState*>(memoState);
entry->lastUsedFrameBoundary = m_frameBoundaryCounter;
return *entry;
}
const BackendVertexInputState& entry = GetOrCreateVertexInputState(vao, GetOrComputeHash(vao));
vao.SetBackendStateMemo(&entry, m_evictionEpoch);
return entry;
return GetOrCreateVertexInputState(vao, GetOrComputeHash(vao));
}
const VertexInputStateFactory::BackendVertexInputState& VertexInputStateFactory::GetOrCreateVertexInputState(
const MG_State::GLState::VertexArrayObject& vao, HashType hash) {
auto it = m_cache.find(hash);
if (it != m_cache.end()) {
it->second->lastUsedFrameBoundary = m_frameBoundaryCounter;
return *it->second;
return it->second;
}
VertexInputStateBuilder builder;
@@ -87,7 +67,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Vector<Uint32> bindingAttributeLocations;
Vector<Bool> bindingUsesClientMemory;
Vector<VertexStreamConversion> bindingConversions;
Vector<VkVertexInputBindingDivisorDescriptionEXT> bindingDivisors;
Uint32 unsupportedAttribMask = 0;
for (Uint32 location = 0; location < MG_State::GLState::VertexArrayObject::MAX_VERTEX_ATTRIBS; ++location) {
@@ -181,51 +160,14 @@ namespace MobileGL::MG_Backend::DirectVulkan {
bindingConversions.push_back(conversion);
builder.AddBinding(binding, stride, inputRate);
builder.AddAttribute(location, binding, vkFormat, 0);
// Divisor 1 is what VK_VERTEX_INPUT_RATE_INSTANCE already means; only anything
// else needs the extension to say it.
if (inputRate == VK_VERTEX_INPUT_RATE_INSTANCE && attr.Divisor != 1) {
bindingDivisors.push_back({binding, static_cast<Uint32>(attr.Divisor)});
}
}
const auto& state = builder.Build();
auto& slot = m_cache[hash];
if (!slot) {
slot = MakeUnique<BackendVertexInputState>();
}
BackendVertexInputState& entry = *slot;
auto& entry = m_cache[hash];
entry.hash = hash;
entry.lastUsedFrameBoundary = m_frameBoundaryCounter;
entry.bindingDivisors = Move(bindingDivisors);
entry.bindings = builder.GetBindings();
entry.attributes = builder.GetAttributes();
// See the layoutHash declaration: hash only the resolved layout, never
// buffer identities, so identical layouts across VAOs/buffers agree.
XXHASH_VERIFY(XXH64_reset(m_hashState, 0));
for (const auto& binding : entry.bindings) {
XXHASH_VERIFY(XXH64_update(m_hashState, &binding.binding, sizeof(binding.binding)));
XXHASH_VERIFY(XXH64_update(m_hashState, &binding.stride, sizeof(binding.stride)));
XXHASH_VERIFY(XXH64_update(m_hashState, &binding.inputRate, sizeof(binding.inputRate)));
}
for (const auto& attribute : entry.attributes) {
XXHASH_VERIFY(XXH64_update(m_hashState, &attribute.location, sizeof(attribute.location)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attribute.binding, sizeof(attribute.binding)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attribute.format, sizeof(attribute.format)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attribute.offset, sizeof(attribute.offset)));
}
for (const auto& divisor : entry.bindingDivisors) {
XXHASH_VERIFY(XXH64_update(m_hashState, &divisor.binding, sizeof(divisor.binding)));
XXHASH_VERIFY(XXH64_update(m_hashState, &divisor.divisor, sizeof(divisor.divisor)));
}
XXHASH_VERIFY(XXH64_update(m_hashState, &unsupportedAttribMask, sizeof(unsupportedAttribMask)));
entry.layoutHash = XXH64_digest(m_hashState);
entry.attributeLocationMask = 0;
for (const auto& attribute : entry.attributes) {
if (attribute.location < 32u) {
entry.attributeLocationMask |= (1u << attribute.location);
}
}
entry.bindingBufferKeys = std::move(bindingBufferKeys);
entry.bindingBaseOffsets = std::move(bindingBaseOffsets);
entry.bindingAttributeLocations = std::move(bindingAttributeLocations);
@@ -235,43 +177,9 @@ namespace MobileGL::MG_Backend::DirectVulkan {
entry.state = state;
entry.state.pVertexBindingDescriptions = entry.bindings.empty() ? nullptr : entry.bindings.data();
entry.state.pVertexAttributeDescriptions = entry.attributes.empty() ? nullptr : entry.attributes.data();
if (!entry.bindingDivisors.empty()) {
entry.divisorState.vertexBindingDivisorCount = static_cast<Uint32>(entry.bindingDivisors.size());
entry.divisorState.pVertexBindingDivisors = entry.bindingDivisors.data();
entry.state.pNext = &entry.divisorState;
} else {
entry.state.pNext = nullptr;
}
return entry;
}
void VertexInputStateFactory::OnFrameBoundary() {
++m_frameBoundaryCounter;
// Sweep occasionally; evict entries whose last hit is far in the past.
// Erasure happens only here, never mid-frame: the draw path holds a
// reference into the current entry across its setup, and unordered_map
// erase would invalidate it. Entries are CPU-side only, so no GPU-idle
// proof is needed; an evicted entry that is used again is simply rebuilt
// from the VAO state (same hash, same content).
constexpr Uint64 kSweepInterval = 256;
constexpr Uint64 kRetireAgeBoundaries = 1024;
if ((m_frameBoundaryCounter % kSweepInterval) != 0) {
return;
}
for (auto it = m_cache.begin(); it != m_cache.end();) {
if (m_frameBoundaryCounter - it->second->lastUsedFrameBoundary > kRetireAgeBoundaries) {
it = m_cache.erase(it);
// Invalidate every VAO's state-pointer memo: the erased node's
// address may be reused by a future insert.
++m_evictionEpoch;
} else {
++it;
}
}
}
VkFormat VertexInputStateFactory::ToVkVertexFormat(DataType type, Int size, Bool normalized, Bool isInteger,
Bool isBgra) {
if (isBgra) {
@@ -27,18 +27,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
struct BackendVertexInputState {
HashType hash = 0;
// Hash of the resolved Vulkan vertex layout only (bindings, attributes,
// unsupported mask) - NO buffer identities. `hash` mixes buffer heap
// addresses so per-chunk VBOs mint a fresh identity per buffer; keying
// pipelines on that minted one VkPipeline per chunk section for an
// identical layout, defeating pipeline reuse and the per-draw memo.
// Pipelines depend only on the layout, so they key on this instead.
HashType layoutHash = 0;
// Frame boundary of the last cache hit; entries idle past the
// OnFrameBoundary retirement age are evicted (CPU heap only).
// Mutable: the VAO's state-pointer memo fast path stamps it through
// a const entry reference.
mutable Uint64 lastUsedFrameBoundary = 0;
Vector<VkVertexInputBindingDescription> bindings;
Vector<VkVertexInputAttributeDescription> attributes;
Vector<SizeT> bindingBufferKeys;
@@ -50,17 +38,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// absent from `attributes`, so without this mask the draw path cannot tell them apart from
// a genuinely disabled array and would silently feed the shader the current attribute value.
Uint32 unsupportedAttribMask = 0;
// Bitmask of `attributes[i].location` - the draw path needs it up to
// three times per draw, so it is baked once at build time.
Uint32 attributeLocationMask = 0;
// Per-binding glVertexAttribDivisor values other than 1. Vulkan's instance input
// rate advances once per instance and nothing else, so anything else has to be
// stated through VK_EXT_vertex_attribute_divisor. Empty when every instanced
// binding uses divisor 1, which is what the plain input rate already means.
Vector<VkVertexInputBindingDivisorDescriptionEXT> bindingDivisors;
VkPipelineVertexInputDivisorStateCreateInfoEXT divisorState{
VK_STRUCTURE_TYPE_PIPELINE_VERTEX_INPUT_DIVISOR_STATE_CREATE_INFO_EXT
};
VkPipelineVertexInputStateCreateInfo state{
VK_STRUCTURE_TYPE_PIPELINE_VERTEX_INPUT_STATE_CREATE_INFO
};
@@ -78,14 +55,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const BackendVertexInputState& GetOrCreateVertexInputState(
const MG_State::GLState::VertexArrayObject& vao, HashType hash);
const BackendVertexInputState& GetOrCreateVertexInputState(const MG_State::GLState::VertexArrayObject& vao);
// Frame boundary hook: ages the cache and evicts entries not hit for many
// frames. The key mixes buffer heap addresses, so buffer/VAO churn keeps
// minting fresh keys; without eviction the map grows for the whole session.
// Entries hold no Vulkan handles (pipeline creation copies the descriptions)
// and the draw path's entry reference never spans a frame boundary, so
// eviction here needs no GPU-idle proof. Self-gated: one counter bump and
// compare except on sweep boundaries.
void OnFrameBoundary();
static SizeT GetComponentSize(DataType type);
// Tightly-packed byte size of one vertex element for this attribute: componentSize * size for
// normal types, and 4 (one packed word) for the 2_10_10_10 types and GL_BGRA. Returns 0 for
@@ -100,19 +69,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const VulkanRendererConfig& m_config;
VkPhysicalDevice m_physicalDevice = VK_NULL_HANDLE;
// Values are heap-allocated: FastSTL::unordered_map is open-addressing,
// so INSERT invalidates references to stored values. The draw path (and
// the VAOs' state-pointer memos) hold entry pointers across inserts;
// only the unique_ptr cell moves, never the pointee.
UnorderedMap<HashType, UniquePtr<BackendVertexInputState>> m_cache;
// Monotonic frame-boundary counter (bumped in OnFrameBoundary) for cache aging.
Uint64 m_frameBoundaryCounter = 0;
// Bumped whenever any cache entry is erased. VAOs memo a raw pointer to
// their heap-allocated entry (stable across map insert/rehash by
// construction); a memo is honored only while its recorded epoch
// matches, so an evicted entry can never be dereferenced through a
// stale memo.
Uint64 m_evictionEpoch = 1;
UnorderedMap<HashType, BackendVertexInputState> m_cache;
static inline XXH64_state_t* m_hashState = XXH64_createState();
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -7,8 +7,6 @@
// End of Source File Header
#include "VkBufferManager.h"
#include "../DirectVulkan.h"
#include "VulkanRenderer.h"
namespace MobileGL::MG_Backend::DirectVulkan {
namespace {
@@ -24,10 +22,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT | VK_BUFFER_USAGE_STORAGE_BUFFER_BIT |
VK_BUFFER_USAGE_INDIRECT_BUFFER_BIT | VK_BUFFER_USAGE_UNIFORM_TEXEL_BUFFER_BIT |
VK_BUFFER_USAGE_TRANSFER_SRC_BIT;
// Appended to kPersistentBackedUsage when VK_EXT_transform_feedback is enabled
// (see VkBufferManagerInitInfo::transformFeedbackUsageEnabled).
constexpr VkBufferUsageFlags kTransformFeedbackUsage =
VK_BUFFER_USAGE_TRANSFORM_FEEDBACK_BUFFER_BIT_EXT;
// The app writes into the persistent map with no explicit flush, so its memory must
// be host-coherent (Adreno host-visible memory is; requiring it keeps us portable).
constexpr VkMemoryPropertyFlags kPersistentBackedRequiredFlags =
@@ -59,18 +53,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
}
// The CPU is about to read a buffer a shader wrote. Its bytes live in coherent
// host-visible GPU storage (EnsureGpuResidentStorage adopts it when the buffer is
// bound as a shader storage buffer), so nothing needs copying - but coherence only
// says the writes are visible once they have happened, so the work has to retire
// first.
void Ops_ReadbackFromGpu(BufferObject& bufferObject) {
(void)bufferObject;
if (pVulkanRenderer) {
pVulkanRenderer->FinishPendingGpuWork();
}
}
void* Ops_AcquirePersistentMap(BufferObject& bufferObject) {
if (g_activeBufferManager) {
return g_activeBufferManager->AcquirePersistentMap(bufferObject);
@@ -94,7 +76,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
.FlushMappedRange = Ops_FlushMappedRange,
.OnDestroy = Ops_OnDestroy,
.AcquirePersistentMap = Ops_AcquirePersistentMap,
.ReadbackFromGpu = Ops_ReadbackFromGpu,
};
} // namespace
@@ -160,15 +141,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_transientUploadArena.BeginFrame(frameIndex);
}
void VkBufferManager::CollectAllDeferredReleases() {
for (Uint32 frameIndex = 0; frameIndex < m_deferredBufferReleases.size(); ++frameIndex) {
CollectDeferredReleases(frameIndex);
}
for (Uint32 frameIndex = 0; frameIndex < m_transientUploadArena.GetFrameCount(); ++frameIndex) {
m_transientUploadArena.CollectDeferredReleases(frameIndex);
}
}
void VkBufferManager::NotifyDeviceIdle() {
// Everything submitted so far has completed. Work recorded for the
// current frame has not been submitted yet, so the current serial
@@ -484,10 +456,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// it from the current shadow - MappedData() is still the shadow here because the
// frontend adopts (and drops) the shadow only after this returns.
DeferRelease(std::move(resource->buffer));
const VkBufferUsageFlags persistentUsage =
kPersistentBackedUsage |
(m_initInfo.transformFeedbackUsageEnabled ? kTransformFeedbackUsage : 0);
if (!CreateResidentStorage(*resource, size, persistentUsage, kPersistentBackedRequiredFlags)) {
if (!CreateResidentStorage(*resource, size, kPersistentBackedUsage, kPersistentBackedRequiredFlags)) {
resource->persistentMapped = false;
resource->storageSize = 0;
resource->usageFlags = 0;
@@ -561,16 +530,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
auto resource = GetOrCreateResource(bufferObject);
bufferObject->SyncPersistentMappedRange();
// A persistently mapped resource's storage IS the application's copy of the bytes -
// the frontend adopted it in place of the shadow and hands out pointers into it, and
// a shader can have written bytes the shadow never saw (a transform feedback
// capture). Streaming a second copy would feed this draw the stale shadow, and the
// downgrade below would release the storage the application still points at,
// breaking the "never recreated" promise AcquirePersistentMap makes.
if (resource->persistentMapped) {
return AcquireResidentSlice(kind, bufferObject, outSlice);
}
const VkDeviceSize size = static_cast<VkDeviceSize>(bufferObject->GetSize());
if (size == 0) {
MGLOG_E("VkBufferManager::AcquireStreamedSlice failed: buffer size is zero");
@@ -31,9 +31,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VmaMemoryUsage transientMemoryUsage = VMA_MEMORY_USAGE_AUTO;
VmaAllocationCreateFlags transientAllocationFlags = VMA_ALLOCATION_CREATE_HOST_ACCESS_SEQUENTIAL_WRITE_BIT;
Bool transientPersistentMapping = false;
// VK_EXT_transform_feedback is enabled: persistent-map storage additionally
// carries the transform feedback usage so capture targets can bind directly.
Bool transformFeedbackUsageEnabled = false;
};
// The DirectVulkan storage behind one frontend buffer (pipe_resource analogue).
@@ -80,11 +77,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// Recreate all per-frame transient arenas
Bool RecreateTransientArenas(Uint32 frameCount);
void BeginFrame(Uint32 frameIndex);
// Drains every frame slot's deferred buffer/resource releases (and the
// transient arena's parked superseded blocks). Only valid when the
// caller has proven every queue submission complete; used by the
// present-less frame-boundary drain.
void CollectAllDeferredReleases();
// All previously submitted GPU work has completed (vkDeviceWaitIdle).
void NotifyDeviceIdle();
// A frame slot's submission fence has been waited: every serial up to
@@ -93,7 +93,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const std::lock_guard<std::mutex> lock(m_mutex);
m_pendingClears.clear();
m_aliveObjects.clear();
m_pendingCount.store(static_cast<Uint32>(m_pendingClears.size()), std::memory_order_relaxed);
}
TextureIdentity VkClearManager::MakeTextureIdentity(MG_State::GLState::ITextureObject* texture) {
@@ -128,7 +127,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_pendingClears.erase(key);
}
m_aliveObjects.erase(identity);
m_pendingCount.store(static_cast<Uint32>(m_pendingClears.size()), std::memory_order_relaxed);
}
Bool VkClearManager::LockTextureIdentityLocked(const TextureIdentity& identity,
@@ -223,7 +221,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_aliveObjects[MakeTextureIdentity(texture.get())] = texture;
auto& pending = m_pendingClears[key];
MergeClearPayload(pending, clearPayload);
m_pendingCount.store(static_cast<Uint32>(m_pendingClears.size()), std::memory_order_relaxed);
}
void VkClearManager::QueueClear(const ClearAttachmentPayload& clearPayload,
@@ -241,7 +238,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_aliveObjects[MakeTextureIdentity(texture.get())] = texture;
auto& pending = m_pendingClears[key];
MergeClearPayload(pending, clearPayload);
m_pendingCount.store(static_cast<Uint32>(m_pendingClears.size()), std::memory_order_relaxed);
}
Bool VkClearManager::HasPendingClear(MG_State::GLState::ITextureObject* texture) {
@@ -249,10 +245,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return false;
}
if (m_pendingCount.load(std::memory_order_relaxed) == 0) {
return false; // per-draw hot path: nothing pending anywhere
}
const Uint64 lifetimeId = texture->GetLifetimeId();
const std::lock_guard<std::mutex> lock(m_mutex);
for (auto it = m_pendingClears.begin(); it != m_pendingClears.end(); ++it) {
@@ -268,9 +260,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
if (key.texture == nullptr) {
return false;
}
if (m_pendingCount.load(std::memory_order_relaxed) == 0) {
return false; // per-draw hot path: nothing pending anywhere
}
const std::lock_guard<std::mutex> lock(m_mutex);
if (m_pendingClears.find(key) == m_pendingClears.end()) {
@@ -298,9 +287,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
if (key.texture == nullptr) {
return false;
}
if (m_pendingCount.load(std::memory_order_relaxed) == 0) {
return false; // per-draw hot path: nothing pending anywhere
}
const std::lock_guard<std::mutex> lock(m_mutex);
if (!LockTextureLocked(key, outTexture)) {
@@ -339,9 +325,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
if (texture == nullptr) {
return false;
}
if (m_pendingCount.load(std::memory_order_relaxed) == 0) {
return false; // per-draw hot path: nothing pending anywhere
}
const Uint64 lifetimeId = texture->GetLifetimeId();
const std::lock_guard<std::mutex> lock(m_mutex);
@@ -362,9 +345,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return;
}
if (m_pendingCount.load(std::memory_order_relaxed) == 0) {
return; // per-draw hot path: nothing pending anywhere
}
const TextureIdentity identity = MakeTextureIdentity(texture);
MGLOG_D("%s: Pop all pending clears for texture %d", __func__, texture->GetExternalIndex());
const std::lock_guard<std::mutex> lock(m_mutex);
@@ -381,7 +361,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
auto it = m_pendingClears.find(key);
if (it != m_pendingClears.end()) {
m_pendingClears.erase(it);
m_pendingCount.store(static_cast<Uint32>(m_pendingClears.size()), std::memory_order_relaxed);
}
}
@@ -14,7 +14,6 @@
#include "MG_Util/Math/VectorTypes.h"
#include <Includes.h>
#include <atomic>
#include <unordered_map>
namespace MobileGL::MG_Backend::DirectVulkan {
@@ -121,19 +120,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
SharedPtr<MG_State::GLState::ITextureObject>& outTexture);
Uint8 m_gcCounter = 0;
public:
// Lock-free probe for the consecutive-draw fast path: any pending clear
// forces the full SetupDraw path (which materializes/consumes it).
Bool HasAnyPendingClears() const { return m_pendingCount.load(std::memory_order_relaxed) != 0; }
private:
mutable std::mutex m_mutex;
// Lock-free mirror of m_pendingClears.size(), maintained under m_mutex
// by every mutation. The per-draw probes (HasPendingClear/GetPending*)
// read it before taking the lock: during draw batches the pending set
// is almost always empty, so this turns several locked map probes per
// draw into one relaxed load.
std::atomic<Uint32> m_pendingCount{0};
std::unordered_map<PendingClearKey, ClearAttachmentPayload, PendingClearKeyHash> m_pendingClears;
std::unordered_map<TextureIdentity, WeakPtr<MG_State::GLState::ITextureObject>, TextureIdentityHash> m_aliveObjects;
};
@@ -16,21 +16,31 @@
namespace MobileGL::MG_Backend::DirectVulkan {
static Bool TryResolveSampleCountFlagBits(Int requestedSamples, VkSampleCountFlagBits& outSampleCount) {
// GL promises "at least the requested samples", so a non-power-of-two
// request (legal in GL, e.g. 3) rounds up to the next Vulkan bit.
if (requestedSamples <= 1) {
switch (requestedSamples <= 0 ? 1 : requestedSamples) {
case 1:
outSampleCount = VK_SAMPLE_COUNT_1_BIT;
return true;
}
if (requestedSamples > 64) {
case 2:
outSampleCount = VK_SAMPLE_COUNT_2_BIT;
return true;
case 4:
outSampleCount = VK_SAMPLE_COUNT_4_BIT;
return true;
case 8:
outSampleCount = VK_SAMPLE_COUNT_8_BIT;
return true;
case 16:
outSampleCount = VK_SAMPLE_COUNT_16_BIT;
return true;
case 32:
outSampleCount = VK_SAMPLE_COUNT_32_BIT;
return true;
case 64:
outSampleCount = VK_SAMPLE_COUNT_64_BIT;
return true;
default:
return false;
}
Uint32 bit = 1;
while (bit < static_cast<Uint32>(requestedSamples)) {
bit <<= 1;
}
outSampleCount = static_cast<VkSampleCountFlagBits>(bit);
return true;
}
static VkImageAspectFlags ResolveImageAspectMaskForFormat(VkFormat format) {
@@ -156,9 +166,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
if (view != VK_NULL_HANDLE) {
vkDestroyImageView(device, view, nullptr);
}
if (unormTwinView != VK_NULL_HANDLE) {
vkDestroyImageView(device, unormTwinView, nullptr);
}
if (image != VK_NULL_HANDLE && allocation != nullptr) {
vmaDestroyImage(allocator, image, allocation);
}
@@ -166,7 +173,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
image = VK_NULL_HANDLE;
allocation = nullptr;
view = VK_NULL_HANDLE;
unormTwinView = VK_NULL_HANDLE;
layout = VK_IMAGE_LAYOUT_UNDEFINED;
format = VK_FORMAT_UNDEFINED;
aspect = VK_IMAGE_ASPECT_NONE;
@@ -174,7 +180,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
sampleCount = VK_SAMPLE_COUNT_1_BIT;
internalFormat = TextureInternalFormat::Unknown;
samples = 0;
deadSinceFrame = kNeverObservedDead;
}
VkRenderPassManager::VkRenderPassManager(VkDevice device,
@@ -201,7 +206,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
resource.Destroy(m_device, m_allocator);
}
m_renderbufferResources.clear();
CollectDeferredRenderbufferReleases(/*destroyAll=*/true); // caller guarantees device idle
m_pendingRenderbufferClears.clear();
RenderPassEntry::s_textureResourcesScratch.clear();
s_activeRenderPass = {};
@@ -209,80 +213,22 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_rpFastValid = false;
}
Uint64 VkRenderPassManager::RetireAgeFrames() const {
// MaxFramesInFlight + 2 covers the frame ring plus one boundary for the
// recording-to-submit gap and one because OnPresent runs ahead of Present's
// fence wait; the floor of 8 keeps a margin over the default ring of 3 while
// still releasing multi-MB attachment memory promptly (the render-pass cache's
// 1024-frame retirement would pin it for no additional safety).
return std::max<Uint64>(8, static_cast<Uint64>(m_config.MaxFramesInFlight) + 2);
}
void VkRenderPassManager::DeferRenderbufferBackingRelease(RenderbufferResource& resource) {
// The superseded backing may still be referenced by in-flight command buffers
// (glRenderbufferStorage can respecify a renderbuffer drawn this very frame),
// so it is parked and destroyed only after RetireAgeFrames() boundaries.
if (resource.image == VK_NULL_HANDLE && resource.view == VK_NULL_HANDLE) {
return;
}
m_deferredRenderbufferReleases.push_back(
{resource.image, resource.allocation, resource.view, resource.unormTwinView, m_frameCounter});
resource.image = VK_NULL_HANDLE;
resource.allocation = nullptr;
resource.view = VK_NULL_HANDLE;
resource.unormTwinView = VK_NULL_HANDLE;
}
void VkRenderPassManager::CollectDeferredRenderbufferReleases(Bool destroyAll) {
if (m_deferredRenderbufferReleases.empty()) {
return;
}
const Uint64 retireAgeFrames = RetireAgeFrames();
std::erase_if(m_deferredRenderbufferReleases, [&](DeferredRenderbufferRelease& release) {
if (!destroyAll && m_frameCounter - release.deferredAtFrame < retireAgeFrames) {
return false;
}
if (release.view != VK_NULL_HANDLE) {
vkDestroyImageView(m_device, release.view, nullptr);
}
if (release.unormTwinView != VK_NULL_HANDLE) {
vkDestroyImageView(m_device, release.unormTwinView, nullptr);
}
if (release.image != VK_NULL_HANDLE) {
vmaDestroyImage(m_allocator, release.image, release.allocation);
}
return true;
});
}
void VkRenderPassManager::CollectRenderbufferGarbage() {
// Two-phase reclamation: a dead renderbuffer's VkImage may still be referenced by
// command buffers submitted up to frames-in-flight frames ago (it was legally
// attached and drawn right up to its deletion), so the first observation of an
// expired weak reference only stamps the current frame counter; Destroy runs once
// enough frame boundaries have passed that the stamping frame's submission fence
// has provably been waited (see RetireAgeFrames).
const Uint64 retireAgeFrames = RetireAgeFrames();
for (auto it = m_renderbufferResources.begin(); it != m_renderbufferResources.end();) {
auto& resource = it->second;
Vector<MG_State::GLState::RenderbufferObject*> deadRenderbuffers;
deadRenderbuffers.reserve(m_renderbufferResources.size());
for (auto& [renderbuffer, resource] : m_renderbufferResources) {
const auto liveRenderbuffer = resource.renderbuffer.lock();
if (liveRenderbuffer && liveRenderbuffer.get() == it->first) {
resource.deadSinceFrame = RenderbufferResource::kNeverObservedDead;
++it;
continue;
if (!liveRenderbuffer || liveRenderbuffer.get() != renderbuffer) {
deadRenderbuffers.emplace_back(renderbuffer);
}
if (resource.deadSinceFrame == RenderbufferResource::kNeverObservedDead) {
resource.deadSinceFrame = m_frameCounter;
++it;
continue;
}
for (auto* renderbuffer : deadRenderbuffers) {
auto resourceIt = m_renderbufferResources.find(renderbuffer);
if (resourceIt != m_renderbufferResources.end()) {
resourceIt->second.Destroy(m_device, m_allocator);
m_renderbufferResources.erase(resourceIt);
}
if (m_frameCounter - resource.deadSinceFrame < retireAgeFrames) {
++it;
continue;
}
m_pendingRenderbufferClears.erase(it->first);
resource.Destroy(m_device, m_allocator);
it = m_renderbufferResources.erase(it);
m_pendingRenderbufferClears.erase(renderbuffer);
}
}
@@ -303,47 +249,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
const auto internalFormat = renderbuffer->GetInternalFormat();
// Three-channel color formats widen to their RGBA twin exactly like textures do
// (VkTextureManager::ResolveTextureFormatInfo): blits/resolves between a
// renderbuffer and a texture of the same GL format then see one VkFormat.
const VkFormat format = [&]() -> VkFormat {
switch (internalFormat) {
case TextureInternalFormat::RGB:
case TextureInternalFormat::RGB8:
case TextureInternalFormat::R3G3B2:
case TextureInternalFormat::RGB4:
case TextureInternalFormat::RGB5:
return VK_FORMAT_R8G8B8A8_UNORM;
case TextureInternalFormat::SRGB8:
return VK_FORMAT_R8G8B8A8_SRGB;
case TextureInternalFormat::RGB8Snorm:
return VK_FORMAT_R8G8B8A8_SNORM;
case TextureInternalFormat::RGB10:
case TextureInternalFormat::RGB12:
case TextureInternalFormat::RGB16:
return VK_FORMAT_R16G16B16A16_UNORM;
case TextureInternalFormat::RGB16Snorm:
return VK_FORMAT_R16G16B16A16_SNORM;
case TextureInternalFormat::RGB16F:
return VK_FORMAT_R16G16B16A16_SFLOAT;
case TextureInternalFormat::RGB32F:
return VK_FORMAT_R32G32B32A32_SFLOAT;
case TextureInternalFormat::RGB8I:
return VK_FORMAT_R8G8B8A8_SINT;
case TextureInternalFormat::RGB8UI:
return VK_FORMAT_R8G8B8A8_UINT;
case TextureInternalFormat::RGB16I:
return VK_FORMAT_R16G16B16A16_SINT;
case TextureInternalFormat::RGB16UI:
return VK_FORMAT_R16G16B16A16_UINT;
case TextureInternalFormat::RGB32I:
return VK_FORMAT_R32G32B32A32_SINT;
case TextureInternalFormat::RGB32UI:
return VK_FORMAT_R32G32B32A32_UINT;
default:
return MG_Util::ConvertTextureInternalFormatToVkEnum(internalFormat);
}
}();
const VkFormat format = MG_Util::ConvertTextureInternalFormatToVkEnum(internalFormat);
const VkImageAspectFlags aspect = ResolveImageAspectMaskForFormat(format);
// Renderbuffers are never sampled (GL has no way to bind one to a sampler), so the
// usage set is attachment + transfer: transfer covers readback (vkCmdCopyImageToBuffer),
@@ -353,46 +259,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
: VK_IMAGE_USAGE_DEPTH_STENCIL_ATTACHMENT_BIT) |
VK_IMAGE_USAGE_TRANSFER_SRC_BIT | VK_IMAGE_USAGE_TRANSFER_DST_BIT;
// GL allows the implementation to allocate more samples than requested
// (glRenderbufferStorageMultisample only promises "at least"), and devices
// like llvmpipe expose 1x/4x but not 2x. Round the request up to the
// nearest supported count for this format.
if (renderbuffer->GetSamples() > 0) {
auto supportedIt = m_attachmentSampleCountsByFormat.find(format);
if (supportedIt == m_attachmentSampleCountsByFormat.end()) {
VkImageFormatProperties formatProperties{};
VkSampleCountFlags supported = VK_SAMPLE_COUNT_1_BIT;
if (vkGetPhysicalDeviceImageFormatProperties(m_physicalDevice, format, VK_IMAGE_TYPE_2D,
VK_IMAGE_TILING_OPTIMAL, imageUsage, 0,
&formatProperties) == VK_SUCCESS) {
supported = formatProperties.sampleCounts;
}
supportedIt = m_attachmentSampleCountsByFormat.emplace(format, supported).first;
}
const VkSampleCountFlags supported = supportedIt->second;
if ((supported & sampleCount) == 0) {
// Smallest supported count above the request, else the largest below it.
Uint32 rounded = 0;
for (Uint32 bit = static_cast<Uint32>(sampleCount) << 1; bit <= VK_SAMPLE_COUNT_64_BIT; bit <<= 1) {
if ((supported & bit) != 0) {
rounded = bit;
break;
}
}
if (rounded == 0) {
for (Uint32 bit = static_cast<Uint32>(sampleCount) >> 1; bit != 0; bit >>= 1) {
if ((supported & bit) != 0) {
rounded = bit;
break;
}
}
}
if (rounded != 0) {
sampleCount = static_cast<VkSampleCountFlagBits>(rounded);
}
}
}
auto& resource = m_renderbufferResources[renderbuffer.get()];
const Bool needsCreate =
resource.image == VK_NULL_HANDLE ||
@@ -404,15 +270,9 @@ namespace MobileGL::MG_Backend::DirectVulkan {
resource.samples != renderbuffer->GetSamples();
if (!needsCreate) {
resource.renderbuffer = renderbuffer;
// A new renderbuffer at a recycled address may adopt a compatible entry that
// was already stamped dead; it is alive again, so cancel the aging.
resource.deadSinceFrame = RenderbufferResource::kNeverObservedDead;
return &resource;
}
// Respecify: park the old backing for aged destruction instead of destroying
// inline - it may still be referenced by in-flight command buffers.
DeferRenderbufferBackingRelease(resource);
resource.Destroy(m_device, m_allocator);
resource.renderbuffer = renderbuffer;
@@ -430,12 +290,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
imageInfo.usage = imageUsage;
imageInfo.samples = sampleCount;
imageInfo.sharingMode = VK_SHARING_MODE_EXCLUSIVE;
// sRGB renderbuffers attach through their UNORM twin while GL_FRAMEBUFFER_SRGB
// is disabled, which needs a format-reinterpreting second view.
const Bool hasUnormTwin = ResolveSrgbAttachmentWriteFormat(format, false) != format;
if (hasUnormTwin) {
imageInfo.flags |= VK_IMAGE_CREATE_MUTABLE_FORMAT_BIT;
}
VkImageFormatProperties imageFormatProperties{};
const VkResult imageFormatResult = vkGetPhysicalDeviceImageFormatProperties(
@@ -470,11 +324,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
viewInfo.subresourceRange.layerCount = 1;
VK_VERIFY(vkCreateImageView(m_device, &viewInfo, nullptr, &resource.view),
"vkCreateImageView(renderbuffer)");
if (hasUnormTwin) {
viewInfo.format = ResolveSrgbAttachmentWriteFormat(format, false);
VK_VERIFY(vkCreateImageView(m_device, &viewInfo, nullptr, &resource.unormTwinView),
"vkCreateImageView(renderbuffer unorm twin)");
}
resource.layout = VK_IMAGE_LAYOUT_UNDEFINED;
resource.format = format;
@@ -571,18 +420,12 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
VkRenderPassManager::HashType VkRenderPassManager::ComputeHash(
const MG_State::GLState::FramebufferObject& fbo, Uint32 swapchainImageIndex, Bool includePendingClear,
Bool includeDefaultFboDepthStencil) {
const MG_State::GLState::FramebufferObject& fbo, Uint32 swapchainImageIndex, Bool includePendingClear) {
XXHASH_VERIFY(XXH64_reset(m_hashState, m_config.CacheVersion));
const Bool isDefaultFbo = fbo.IsDefaultFramebuffer();
if (isDefaultFbo) {
XXHASH_VERIFY(XXH64_update(m_hashState, &swapchainImageIndex, sizeof(swapchainImageIndex)));
}
// sRGB attachments switch between their sRGB and UNORM-twin views with this
// capability (ResolveSrgbAttachmentWriteFormat), changing the render pass formats.
const Bool framebufferSrgbEnabled =
MG_State::pGLContext->IsCapabilityEnabled(MobileGL::CapabilityInput::FramebufferSrgb);
XXHASH_VERIFY(XXH64_update(m_hashState, &framebufferSrgbEnabled, sizeof(framebufferSrgbEnabled)));
auto& drawBuffers = fbo.GetDrawBuffers();
XXHASH_VERIFY(XXH64_update(m_hashState, drawBuffers.data(), drawBuffers.size() * sizeof(drawBuffers[0])));
auto readBuffer = fbo.GetReadBuffer();
@@ -656,17 +499,9 @@ namespace MobileGL::MG_Backend::DirectVulkan {
attachment <= FramebufferAttachmentType::BackRight);
if (isDefaultColorAttachment) {
currentLayout = m_swapchainObject.GetImageLayout(swapchainImageIndex);
// Content validity feeds the attachment's loadOp (see the
// creation path), so it must key the cache as well.
if (!m_swapchainObject.IsImageContentDefined(swapchainImageIndex)) {
currentLayout = VK_IMAGE_LAYOUT_UNDEFINED;
}
} else if (attachment == FramebufferAttachmentType::Depth ||
attachment == FramebufferAttachmentType::Stencil) {
currentLayout = m_swapchainObject.GetDepthStencilImageLayout(swapchainImageIndex);
if (!m_swapchainObject.IsDepthStencilContentDefined(swapchainImageIndex)) {
currentLayout = VK_IMAGE_LAYOUT_UNDEFINED;
}
}
} else {
auto* textureResource = m_textureManager.SyncTextureAndGetDescriptor(*texture);
@@ -721,49 +556,14 @@ namespace MobileGL::MG_Backend::DirectVulkan {
combineFramebufferAttachmentObjHash(drawbuf);
}
// The depth-less default-FBO flavor omits the depth/stencil attachment
// entirely, so it must hash differently from the depth-full flavor.
const Bool depthStencilIncluded = !isDefaultFbo || includeDefaultFboDepthStencil;
XXHASH_VERIFY(XXH64_update(m_hashState, &depthStencilIncluded, sizeof(depthStencilIncluded)));
if (depthStencilIncluded) {
combineFramebufferAttachmentObjHash(FramebufferAttachmentType::Depth);
combineFramebufferAttachmentObjHash(FramebufferAttachmentType::Stencil);
}
combineFramebufferAttachmentObjHash(FramebufferAttachmentType::Depth);
combineFramebufferAttachmentObjHash(FramebufferAttachmentType::Stencil);
return XXH64_digest(m_hashState);
}
RenderPassEntry& VkRenderPassManager::GetOrCreateRenderPass(const MG_State::GLState::FramebufferObject& fbo,
Uint32 swapchainImageIndex,
Bool drawUsesDepthStencil) {
// Resolve the default-FBO depth flavor (see the header comment): keep the
// depth attachment when the caller needs it, when a depth/stencil clear is
// pending, or when the active pass already carries it (escalate-only, so
// alternating depth-less draws never split an established depth pass).
Bool includeDefaultFboDepthStencil = true;
if (fbo.IsDefaultFramebuffer()) {
Bool activeDefaultHasDepthStencil = false;
if (const auto* active = GetActiveRenderPass()) {
Bool activeIsSwapchainPass = false;
Bool activeHasSwapchainDepthStencil = false;
for (const auto& tracked : active->trackedAttachmentLayouts) {
activeIsSwapchainPass |= tracked.target == TrackedAttachmentTarget::SwapchainColor;
activeHasSwapchainDepthStencil |=
tracked.target == TrackedAttachmentTarget::SwapchainDepthStencil;
}
activeDefaultHasDepthStencil = activeIsSwapchainPass && activeHasSwapchainDepthStencil;
}
const auto& defaultDepthAtt = fbo.GetAttachment(FramebufferAttachmentType::Depth);
const auto& defaultStencilAtt = fbo.GetAttachment(FramebufferAttachmentType::Stencil);
const Bool pendingDepthStencilClear =
(defaultDepthAtt.IsTexture() && m_clearManager.HasPendingClear(defaultDepthAtt)) ||
HasPendingRenderbufferClear(defaultDepthAtt) ||
(defaultStencilAtt.IsTexture() && m_clearManager.HasPendingClear(defaultStencilAtt)) ||
HasPendingRenderbufferClear(defaultStencilAtt);
includeDefaultFboDepthStencil =
drawUsesDepthStencil || activeDefaultHasDepthStencil || pendingDepthStencilClear;
}
Uint32 swapchainImageIndex) {
auto hasPendingClearOnFramebuffer = [&]() -> Bool {
const auto& drawBuffers = fbo.GetDrawBuffers();
for (auto attachment : drawBuffers) {
@@ -813,7 +613,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_rpFastFboVersion == fbo.GetObjectVersion() && m_rpFastSwapchainIndex == swapchainImageIndex &&
m_rpFastTexEpoch == m_textureManager.GetTextureImageEpoch() &&
m_rpFastRbEpoch == m_renderbufferImageEpoch &&
(!fbo.IsDefaultFramebuffer() || m_rpFastHadDepthStencil == includeDefaultFboDepthStencil) &&
m_rpFastRenderPassHash == activeRenderPass->hash && !hasPendingClearOnFramebuffer()) {
auto activeIt = m_renderPasses.find(activeRenderPass->hash);
if (activeIt != m_renderPasses.end()) {
@@ -822,7 +621,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
}
auto compatibilityHash = ComputeHash(fbo, swapchainImageIndex, false, includeDefaultFboDepthStencil);
auto compatibilityHash = ComputeHash(fbo, swapchainImageIndex, false);
if (activeRenderPass != nullptr &&
activeRenderPass->CompatibleWith(compatibilityHash) &&
!hasPendingClearOnFramebuffer()) {
@@ -839,11 +638,10 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_rpFastTexEpoch = m_textureManager.GetTextureImageEpoch();
m_rpFastRbEpoch = m_renderbufferImageEpoch;
m_rpFastRenderPassHash = activeRenderPass->hash;
m_rpFastHadDepthStencil = activeIt->second.hasDepthStencilAttachment;
activeIt->second.lastUsedFrame = m_frameCounter;
return activeIt->second;
}
auto hash = ComputeHash(fbo, swapchainImageIndex, true, includeDefaultFboDepthStencil);
auto hash = ComputeHash(fbo, swapchainImageIndex, true);
auto it = m_renderPasses.find(hash);
if (it != m_renderPasses.end()) {
it->second.lastUsedFrame = m_frameCounter;
@@ -930,12 +728,8 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
const VkImageLayout trackedRbLayout = rbResource->layout;
const Bool rbFramebufferSrgb =
MG_State::pGLContext->IsCapabilityEnabled(MobileGL::CapabilityInput::FramebufferSrgb);
const VkFormat rbAttachmentFormat =
ResolveSrgbAttachmentWriteFormat(rbResource->format, rbFramebufferSrgb);
rbDesc.flags = 0;
rbDesc.format = rbAttachmentFormat;
rbDesc.format = rbResource->format;
rbDesc.samples = rbResource->sampleCount;
rbDesc.loadOp = rbHasClear ? VK_ATTACHMENT_LOAD_OP_CLEAR :
(trackedRbLayout == VK_IMAGE_LAYOUT_UNDEFINED ? VK_ATTACHMENT_LOAD_OP_DONT_CARE
@@ -970,8 +764,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
.finalLayout = rbDesc.finalLayout,
});
textureResources.emplace_back(nullptr);
attachmentViews.emplace_back(rbAttachmentFormat != rbResource->format ? rbResource->unormTwinView
: rbResource->view);
attachmentViews.emplace_back(rbResource->view);
MOBILEGL_ASSERT(attachmentViews.back() != VK_NULL_HANDLE,
"GetOrCreateRenderPass: renderbuffer view missing at color attachment %d", i);
@@ -1040,13 +833,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
MOBILEGL_ASSERT(swapchainImageIndex < swapchainViews.size(),
"GetOrCreateRenderPass: swapchain image index out of range");
trackedColorLayout = m_swapchainObject.GetImageLayout(swapchainImageIndex);
// EGL: a presented color buffer's content is undefined when its
// image comes back around (EGL_BUFFER_DESTROYED, the default
// swap behaviour) - skip the tile load instead of reloading
// stale pixels nobody may rely on.
if (!hasClear && !m_swapchainObject.IsImageContentDefined(swapchainImageIndex)) {
trackedColorLayout = VK_IMAGE_LAYOUT_UNDEFINED;
}
trackedAttachmentLayouts.emplace_back(TrackedAttachmentLayoutInfo {
.target = TrackedAttachmentTarget::SwapchainColor,
.swapchainImageIndex = swapchainImageIndex,
@@ -1060,15 +846,12 @@ namespace MobileGL::MG_Backend::DirectVulkan {
MOBILEGL_ASSERT(textureResource,
"GetOrCreateRenderPass: SyncTextureAndGetDescriptor failed at color attachment %d", i);
textureResources.emplace_back(textureResource);
desc.format = ResolveSrgbAttachmentWriteFormat(
textureResource->format,
MG_State::pGLContext->IsCapabilityEnabled(MobileGL::CapabilityInput::FramebufferSrgb));
desc.format = textureResource->format;
attachmentSampleCount = textureResource->sampleCount;
trackedColorLayout = textureResource->layout;
trackedAttachmentLayouts.emplace_back(TrackedAttachmentLayoutInfo {
.target = TrackedAttachmentTarget::Texture,
.texture = att.GetTexture(),
.textureRaw = att.GetTexture().get(),
.textureMipLevel = attachmentMipLevel,
.finalLayout = desc.finalLayout,
});
@@ -1132,12 +915,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
};
const auto* selectedDepthStencilAttachment = isUsableDepthStencilAttachment(depthAtt) ? &depthAtt :
(isUsableDepthStencilAttachment(stencilAtt) ? &stencilAtt : nullptr);
// Depth-less default-FBO flavor: nothing in this pass touches depth/stencil
// and their content is undefined anyway (EGL swap), so drop the attachment
// and its whole tile load + store.
if (isDefaultFbo && !includeDefaultFboDepthStencil) {
selectedDepthStencilAttachment = nullptr;
}
const Bool hasDistinctDepthAndStencilAttachments =
isUsableDepthStencilAttachment(depthAtt) && isUsableDepthStencilAttachment(stencilAtt) &&
!sameDepthStencilAttachmentObject(depthAtt, stencilAtt);
@@ -1156,12 +933,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkImageLayout trackedDepthLayout = isDefaultFbo ?
m_swapchainObject.GetDepthStencilImageLayout(swapchainImageIndex) :
VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL;
// EGL 1.5 §3.10.1: every ancillary (depth/stencil) buffer's content is
// undefined after a swap, so the first default-FBO pass of a frame can
// skip the depth/stencil tile load outright.
if (isDefaultFbo && !m_swapchainObject.IsDepthStencilContentDefined(swapchainImageIndex)) {
trackedDepthLayout = VK_IMAGE_LAYOUT_UNDEFINED;
}
depthAttachmentDescription.flags = 0;
VkSampleCountFlagBits depthAttachmentSampleCount = VK_SAMPLE_COUNT_1_BIT;
Int depthAttachmentId = 0;
@@ -1245,7 +1016,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
trackedAttachmentLayouts.emplace_back(TrackedAttachmentLayoutInfo {
.target = TrackedAttachmentTarget::Texture,
.texture = selectedDepthStencilAttachment->GetTexture(),
.textureRaw = selectedDepthStencilAttachment->GetTexture().get(),
.textureMipLevel = attachmentMipLevel,
.finalLayout = depthAttachmentDescription.finalLayout,
});
@@ -1291,22 +1061,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
const Bool hasDepthStencilAttachment = depthAttachmentRef.attachment != VK_ATTACHMENT_UNUSED;
// Declare only the used colour-reference span. The GL draw-buffer array
// always spans 8 slots, so passes used to declare colorAttachmentCount=8
// with trailing VK_ATTACHMENT_UNUSED holes - and Adreno configures its
// per-pixel render-backend/export path from the DECLARED count, so every
// fragment of every pass paid the 8-target export cost (measured on
// Adreno 650 / MC 26.2: 11.9 -> 7.5 ms of GPU time per frame, with the
// single-quad swapchain blit pass alone dropping 1.26 -> 0.40 ms).
// Interior GL_NONE holes keep their slots so fragment-output locations
// still line up; a fragment output at a location past the trimmed count
// is discarded, which is exactly GL's semantic for writing to a draw
// buffer set to GL_NONE.
while (!colorAttachmentRefs.empty() &&
colorAttachmentRefs.back().attachment == VK_ATTACHMENT_UNUSED) {
colorAttachmentRefs.pop_back();
}
// Subpass
VkSubpassDescription subpassDesc;
subpassDesc.flags = 0;
@@ -1424,14 +1178,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
void VkRenderPassManager::OnPresent() {
++m_frameCounter;
// Runs every frame boundary, ahead of the render-pass sweep gate below: the walk
// is O(#renderbuffer resources) — single digits in practice — and per-frame
// invocation keeps dead-resource reclaim latency at the aging bound instead of
// coupling it to renderbuffer *use* (the GetOrCreateRenderbufferResource call
// site never runs again once an app stops using renderbuffers).
CollectRenderbufferGarbage();
CollectDeferredRenderbufferReleases(/*destroyAll=*/false);
// Sweep occasionally; evict entries whose last use is far past every
// in-flight frame so their VkRenderPass/VkFramebuffer can be destroyed
// safely (RenderPassEntry's destructor releases the handles).
@@ -1441,12 +1187,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return;
}
// Collect the dying handles and notify once after the loop: pipelines hashed
// on them share the entries' >kRetireAgeFrames idleness (they are only bound
// by draws that hit those entries), so the observer may destroy them
// immediately - and a single batched notification costs one pipeline-cache
// scan instead of one per evicted pass.
Vector<VkRenderPass> destroyedRenderPasses;
const Uint64 activeHash = s_hasActiveRenderPass ? s_activeRenderPass.hash : 0;
for (auto it = m_renderPasses.begin(); it != m_renderPasses.end();) {
const Bool isActive = s_hasActiveRenderPass && it->first == activeHash;
@@ -1454,15 +1194,11 @@ namespace MobileGL::MG_Backend::DirectVulkan {
if (m_rpFastValid && m_rpFastRenderPassHash == it->first) {
m_rpFastValid = false;
}
destroyedRenderPasses.push_back(it->second.renderPass);
it = m_renderPasses.erase(it);
} else {
++it;
}
}
if (!destroyedRenderPasses.empty() && m_evictionObserver != nullptr) {
m_evictionObserver->OnRenderPassesDestroyed(destroyedRenderPasses);
}
}
Bool VkRenderPassManager::BeginRenderPass(VkCommandBuffer commandBuffer, RenderPassEntry& renderPassEntry) {
@@ -1515,17 +1251,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
renderPassBeginInfo.pClearValues = clearValues.data();
vkCmdBeginRenderPass(commandBuffer, &renderPassBeginInfo, VK_SUBPASS_CONTENTS_INLINE);
// Pre-pass stream bookkeeping: this pass's attachment images are now
// referenced by the open frame recording.
if (s_textureManager != nullptr) {
for (const auto& tracked : renderPassEntry.trackedAttachmentLayouts) {
if (tracked.target == TrackedAttachmentTarget::Texture) {
if (const auto texture = tracked.texture.lock()) {
s_textureManager->StampTextureRecordingUse(texture.get());
}
}
}
}
for (const auto& pending: renderPassEntry.pendingClearAttachments) {
if (pending.hasInlinePayload) {
if (s_renderPassManager != nullptr) {
@@ -1578,15 +1303,11 @@ namespace MobileGL::MG_Backend::DirectVulkan {
case TrackedAttachmentTarget::SwapchainColor:
MOBILEGL_ASSERT(s_swapchainObject != nullptr, "EndRenderPass: swapchain object is null");
s_swapchainObject->SetImageLayout(trackedAttachment.swapchainImageIndex, trackedAttachment.finalLayout);
// The pass stored into the attachment: its content is defined
// until the image is next presented.
s_swapchainObject->SetImageContentDefined(trackedAttachment.swapchainImageIndex, true);
break;
case TrackedAttachmentTarget::SwapchainDepthStencil:
MOBILEGL_ASSERT(s_swapchainObject != nullptr, "EndRenderPass: swapchain object is null");
s_swapchainObject->SetDepthStencilImageLayout(trackedAttachment.swapchainImageIndex,
trackedAttachment.finalLayout);
s_swapchainObject->SetDepthStencilContentDefined(trackedAttachment.swapchainImageIndex, true);
break;
default:
MOBILEGL_ASSERT(false, "EndRenderPass: unsupported tracked attachment target=%d",
@@ -42,11 +42,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
struct TrackedAttachmentLayoutInfo {
TrackedAttachmentTarget target = TrackedAttachmentTarget::Texture;
WeakPtr<MG_State::GLState::ITextureObject> texture;
// Identity-compare shortcut for the per-draw "does the active pass use
// this sampled texture" probe: comparing this against a LIVE texture's
// address needs no weak_ptr::lock (two refcount atomics per probe).
// May dangle once the texture dies - compare only, never dereference.
MG_State::GLState::ITextureObject* textureRaw = nullptr;
WeakPtr<MG_State::GLState::RenderbufferObject> renderbuffer;
Uint32 textureMipLevel = 0;
Uint32 swapchainImageIndex = 0;
@@ -162,53 +157,19 @@ namespace MobileGL::MG_Backend::DirectVulkan {
class VkRenderPassManager {
public:
using HashType = Uint64;
// Notified once per OnPresent sweep with every aged-out entry's VkRenderPass
// value: pipelines are hashed on the raw handle, and once destroyed the value
// may be recycled for an incompatible pass, so dependent caches must purge
// everything keyed on them before any new pass can be created (the sweep and
// the notification run back-to-back with no creation in between; observers
// compare the values, never dereference them). Batched so a mass-idle cohort
// (shader-pack switch, dimension exit) costs the observer one pipeline-cache
// scan, not one per dying pass. The wholesale paths
// (Shutdown/RecreateSwapchain) do not notify - their callers already drop
// every pipeline outright.
class IEvictionObserver {
public:
virtual ~IEvictionObserver() = default;
virtual void OnRenderPassesDestroyed(const Vector<VkRenderPass>& renderPasses) = 0;
};
VkRenderPassManager(VkDevice device,
VkPhysicalDevice physicalDevice, VmaAllocator allocator, const VulkanRendererConfig& config,
VkClearManager& clearManager, VkTextureManager& textureManager, SwapchainObject& swapchainObject);
~VkRenderPassManager();
// Observer may be null (no notifications). Not owned.
void SetEvictionObserver(IEvictionObserver* observer) { m_evictionObserver = observer; }
Bool Initialize();
void Shutdown();
HashType ComputeHash(
const MG_State::GLState::FramebufferObject& fbo,
Uint32 swapchainImageIndex,
Bool includePendingClear = true,
Bool includeDefaultFboDepthStencil = true);
// drawUsesDepthStencil: whether the operation about to run inside the pass
// reads or writes the depth/stencil buffer (depth test or stencil test
// enabled, or a depth/stencil clear). Only consulted for the DEFAULT
// framebuffer: EGL undefines its ancillary buffers at every swap, so a
// default-FBO pass whose draws provably never touch depth/stencil is
// created WITHOUT the depth attachment - on a tiler that skips the whole
// depth tile load AND store. The flavor only escalates: once a pass with
// depth is active, later depth-less draws keep using it, and a depth-using
// draw against a depth-less active pass resolves to a new (incompatible)
// entry, which the caller's compatibility check turns into a pass split;
// the new pass's depth loads DONT_CARE (content was undefined all along).
RenderPassEntry& GetOrCreateRenderPass(const MG_State::GLState::FramebufferObject& fbo,
Uint32 swapchainImageIndex,
Bool drawUsesDepthStencil = true);
Bool includePendingClear = true);
RenderPassEntry& GetOrCreateRenderPass(const MG_State::GLState::FramebufferObject& fbo, Uint32 swapchainImageIndex);
void QueueRenderbufferClear(GLbitfield mask, const ClearFramebufferPayload& clearPayload,
const MG_State::GLState::FramebufferObject& drawFbo);
void QueueRenderbufferClear(const ClearAttachmentPayload& clearPayload,
@@ -231,20 +192,12 @@ namespace MobileGL::MG_Backend::DirectVulkan {
UnorderedMap<Uint64, RenderPassEntry> m_renderPasses;
// Monotonic frame counter (bumped in OnPresent) for render-pass cache aging.
Uint64 m_frameCounter = 0;
IEvictionObserver* m_evictionObserver = nullptr;
// Bumped whenever a renderbuffer VkImage is (re)created; together with the texture
// manager's image epoch this invalidates the render-pass fast path on any attachment
// image recreation.
Uint64 m_renderbufferImageEpoch = 1;
public:
// Bumped whenever a renderbuffer backing is (re)created; consecutive-draw
// snapshots include it so an attachment respecify forces a re-resolve.
Uint64 GetRenderbufferImageEpoch() const { return m_renderbufferImageEpoch; }
private:
// Per-draw fast-path memo for GetOrCreateRenderPass (dirty-flag state tracking): when the
// framebuffer state is provably unchanged since the last resolution, the active render pass
// is reused WITHOUT recomputing the expensive per-draw hash. Invalidated by FBO switch /
@@ -257,26 +210,13 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Uint64 m_rpFastTexEpoch = 0;
Uint64 m_rpFastRbEpoch = 0;
Uint64 m_rpFastRenderPassHash = 0;
// Whether the memoized entry carries a depth/stencil attachment; a
// default-FBO resolution whose effective depth request differs must
// miss the memo (the depth-less/depth-full flavors hash differently).
Bool m_rpFastHadDepthStencil = false;
public:
struct RenderbufferResource {
// deadSinceFrame sentinel: the owning weak reference has not been observed
// expired. Dead resources age past every in-flight frame before Destroy
// (see CollectRenderbufferGarbage); the GPU may still reference the image
// for frames-in-flight frames after the GL object dies.
static constexpr Uint64 kNeverObservedDead = UINT64_MAX;
WeakPtr<MG_State::GLState::RenderbufferObject> renderbuffer;
VkImage image = VK_NULL_HANDLE;
VmaAllocation allocation = nullptr;
VkImageView view = VK_NULL_HANDLE;
// UNORM reinterpretation of an sRGB image, used as the attachment view while
// GL_FRAMEBUFFER_SRGB is disabled (raw writes). Null for non-sRGB formats.
VkImageView unormTwinView = VK_NULL_HANDLE;
VkImageLayout layout = VK_IMAGE_LAYOUT_UNDEFINED;
VkFormat format = VK_FORMAT_UNDEFINED;
VkImageAspectFlags aspect = VK_IMAGE_ASPECT_NONE;
@@ -284,8 +224,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkSampleCountFlagBits sampleCount = VK_SAMPLE_COUNT_1_BIT;
TextureInternalFormat internalFormat = TextureInternalFormat::Unknown;
Int samples = 0;
// m_frameCounter value at which the weak reference was first seen expired.
Uint64 deadSinceFrame = kNeverObservedDead;
void Destroy(VkDevice device, VmaAllocator allocator);
};
@@ -303,32 +241,12 @@ namespace MobileGL::MG_Backend::DirectVulkan {
ClearAttachmentPayload payload{};
};
// A superseded renderbuffer backing (glRenderbufferStorage respecify) parked
// until enough frame boundaries have passed that no in-flight command buffer
// can still reference it; destroyed in OnPresent (see RetireAgeFrames).
struct DeferredRenderbufferRelease {
VkImage image = VK_NULL_HANDLE;
VmaAllocation allocation = nullptr;
VkImageView view = VK_NULL_HANDLE;
VkImageView unormTwinView = VK_NULL_HANDLE;
Uint64 deferredAtFrame = 0;
};
UnorderedMap<MG_State::GLState::RenderbufferObject*, RenderbufferResource> m_renderbufferResources;
UnorderedMap<MG_State::GLState::RenderbufferObject*, PendingRenderbufferClear> m_pendingRenderbufferClears;
Vector<DeferredRenderbufferRelease> m_deferredRenderbufferReleases;
// Supported sample counts per attachment format, so per-draw resource lookups
// do not repeat vkGetPhysicalDeviceImageFormatProperties.
UnorderedMap<VkFormat, VkSampleCountFlags> m_attachmentSampleCountsByFormat;
Bool HasPendingRenderbufferClear(
const MG_State::GLState::FramebufferAttachmentObject& attachment) const;
void CollectRenderbufferGarbage();
// Frame-boundary margin after which a resource last referenced by a retired
// GL object (or superseded backing) is provably past every in-flight frame.
Uint64 RetireAgeFrames() const;
void DeferRenderbufferBackingRelease(RenderbufferResource& resource);
void CollectDeferredRenderbufferReleases(Bool destroyAll);
static inline XXH64_state_t* m_hashState = XXH64_createState();
static inline ActiveRenderPassInfo s_activeRenderPass{};
@@ -51,18 +51,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Float ResolveEffectiveMinLod(const MG_State::GLState::SamplerObject& sampler, Float effectiveMaxLod) {
return std::min(sampler.GetMinLod(), effectiveMaxLod);
}
// A single-level view can only ever deliver the base level, but the LOD clamp must not be
// collapsed to exactly 0: both GL and Vulkan pick magFilter over minFilter from the
// *clamped* lambda, so maxLod = 0 would make every fragment magnify and quietly retire the
// min filter. 0.25 is the value VkSamplerCreateInfo's own note prescribes for emulating
// GL's non-mipmapped minification - large enough for lambda to stay positive, small enough
// that a NEAREST mip mode still rounds down to level 0. Clamped rather than assigned, so a
// texture whose GL_TEXTURE_MAX_LOD really is 0 keeps magnifying as GL says it must.
Float ResolveSingleLevelMaxLod(const MG_State::GLState::SamplerObject& sampler, Bool singleLevelView) {
const Float maxLod = ResolveEffectiveMaxLod(sampler);
return singleLevelView ? std::min(maxLod, 0.25f) : maxLod;
}
} // namespace
Bool VkSamplerManager::Initialize(const InitInfo& initInfo) {
@@ -101,43 +89,15 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_device = VK_NULL_HANDLE;
m_config = nullptr;
m_frameBoundaryCounter = 0;
}
void VkSamplerManager::OnFrameBoundary() {
++m_frameBoundaryCounter;
// Sweep occasionally; destroy samplers whose last use is far past every
// in-flight frame. Destroy and erase must stay atomic, or Shutdown would
// double-free the handle; an evicted key that recurs simply re-creates
// its sampler on the next miss.
constexpr Uint64 kSweepInterval = 256;
constexpr Uint64 kRetireAgeBoundaries = 1024;
if ((m_frameBoundaryCounter % kSweepInterval) != 0) {
return;
}
for (auto it = m_samplers.begin(); it != m_samplers.end();) {
auto& entry = it->second;
if (m_frameBoundaryCounter - entry.lastUsedFrameBoundary > kRetireAgeBoundaries) {
if (m_device != VK_NULL_HANDLE && entry.handle != VK_NULL_HANDLE) {
vkDestroySampler(m_device, entry.handle, nullptr);
}
it = m_samplers.erase(it);
} else {
++it;
}
}
}
Uint64 VkSamplerManager::BuildSamplerKey(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture,
Bool forceNearestFiltering, Bool singleLevelView) const {
Bool forceNearestFiltering) const {
MOBILEGL_ASSERT(m_config != nullptr, "VkSamplerManager::BuildSamplerKey: m_config is null");
XXHASH_VERIFY(XXH64_reset(m_hashState, m_config->CacheVersion));
XXHASH_VERIFY(XXH64_update(m_hashState, &forceNearestFiltering, sizeof(forceNearestFiltering)));
XXHASH_VERIFY(XXH64_update(m_hashState, &singleLevelView, sizeof(singleLevelView)));
const auto minFilter = sampler.GetMinFilter();
XXHASH_VERIFY(XXH64_update(m_hashState, &minFilter, sizeof(minFilter)));
@@ -151,7 +111,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
XXHASH_VERIFY(XXH64_update(m_hashState, &wrapT, sizeof(wrapT)));
const auto wrapR = sampler.GetWrapR();
XXHASH_VERIFY(XXH64_update(m_hashState, &wrapR, sizeof(wrapR)));
const auto maxLod = ResolveSingleLevelMaxLod(sampler, singleLevelView);
const auto maxLod = ResolveEffectiveMaxLod(sampler);
const auto minLod = ResolveEffectiveMinLod(sampler, maxLod);
XXHASH_VERIFY(XXH64_update(m_hashState, &minLod, sizeof(minLod)));
XXHASH_VERIFY(XXH64_update(m_hashState, &maxLod, sizeof(maxLod)));
@@ -173,20 +133,10 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkSampler VkSamplerManager::GetOrCreateSampler(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture,
Bool forceNearestFiltering, Uint32 viewLevelCount) {
// A view that exposes a single mip level has no second level to blend with, so GL's
// *_MIPMAP_* minification filters degenerate to plain filtering on the base level -
// sampling is unchanged by pinning the Vulkan sampler to NEAREST mip mode at LOD 0.
// It is not cosmetic: MobileGL backs such a view with a fully allocated mip chain whose
// tail is never written, and a LINEAR mip mode lets the texture unit issue the level+1
// fetch anyway. On Adreno that fetch lands in uninitialized UBWC pages (or past the
// allocation for a genuinely single-level image) and faults the GPU - the same failure
// the default-framebuffer blit shader had to work around with an explicit-LOD sample.
const Bool singleLevelView = viewLevelCount == 1;
const Uint64 key = BuildSamplerKey(sampler, texture, forceNearestFiltering, singleLevelView);
Bool forceNearestFiltering) {
const Uint64 key = BuildSamplerKey(sampler, texture, forceNearestFiltering);
auto it = m_samplers.find(key);
if (it != m_samplers.end()) {
it->second.lastUsedFrameBoundary = m_frameBoundaryCounter;
return it->second.handle;
}
@@ -194,9 +144,8 @@ namespace MobileGL::MG_Backend::DirectVulkan {
samplerInfo.sType = VK_STRUCTURE_TYPE_SAMPLER_CREATE_INFO;
samplerInfo.magFilter = forceNearestFiltering ? VK_FILTER_NEAREST : ToVkFilter(sampler.GetMagFilter());
samplerInfo.minFilter = forceNearestFiltering ? VK_FILTER_NEAREST : ToVkFilter(sampler.GetMinFilter());
samplerInfo.mipmapMode = (forceNearestFiltering || singleLevelView)
? VK_SAMPLER_MIPMAP_MODE_NEAREST
: ToVkMipmapMode(sampler.GetMipmapMode());
samplerInfo.mipmapMode = forceNearestFiltering ? VK_SAMPLER_MIPMAP_MODE_NEAREST
: ToVkMipmapMode(sampler.GetMipmapMode());
samplerInfo.addressModeU = ToVkAddressMode(sampler.GetWrapS());
samplerInfo.addressModeV = ToVkAddressMode(sampler.GetWrapT());
samplerInfo.addressModeW = ToVkAddressMode(sampler.GetWrapR());
@@ -208,8 +157,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
samplerInfo.maxAnisotropy = maxAnisotropy;
samplerInfo.compareEnable = sampler.GetCompareMode() == SamplerCompareMode::CompareToTexture ? VK_TRUE : VK_FALSE;
samplerInfo.compareOp = ToVkCompareOp(ResolveCompareFunc(sampler, texture));
// Must match BuildSamplerKey's resolution exactly.
samplerInfo.maxLod = ResolveSingleLevelMaxLod(sampler, singleLevelView);
samplerInfo.maxLod = ResolveEffectiveMaxLod(sampler);
samplerInfo.minLod = ResolveEffectiveMinLod(sampler, samplerInfo.maxLod);
samplerInfo.borderColor = ResolveVkBorderColor(sampler, texture);
samplerInfo.unnormalizedCoordinates = VK_FALSE;
@@ -221,7 +169,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
entry.handle = vkSampler;
entry.externalIndex = sampler.GetExternalIndex();
entry.version = sampler.GetVersion();
entry.lastUsedFrameBoundary = m_frameBoundaryCounter;
m_samplers[key] = entry;
return vkSampler;
}
@@ -33,38 +33,20 @@ public:
Bool Initialize(const InitInfo& initInfo);
void Shutdown();
// viewLevelCount is the mip-level count of the image view this sampler will be paired
// with; 0 means "unknown, do not narrow". See GetOrCreateSampler for why it matters.
VkSampler GetOrCreateSampler(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture,
Bool forceNearestFiltering = false,
Uint32 viewLevelCount = 0);
// Frame boundary hook: ages the sampler cache and destroys samplers not used
// for many frames. The key hashes continuous float state (lodBias, LOD clamps,
// anisotropy), so an app animating those would otherwise mint an unbounded
// stream of never-destroyed VkSamplers and eventually exhaust the device's
// maxSamplerAllocationCount. A sampler idle for over a thousand frame
// boundaries cannot be referenced by any in-flight command buffer (frames in
// flight are single digits), and every descriptor set the GPU consumes is
// written that same frame with live handles (the per-binding resolve memo and
// descriptor-set reuse are both frame-reset), so destruction here needs no
// fence wait. Self-gated: one counter bump and compare except on sweep
// boundaries.
void OnFrameBoundary();
Bool forceNearestFiltering = false);
private:
struct SamplerCacheEntry {
VkSampler handle = VK_NULL_HANDLE;
Uint externalIndex = 0;
Uint16 version = 0;
// Frame boundary of the last cache hit; entries idle past the
// OnFrameBoundary retirement age have their VkSampler destroyed.
Uint64 lastUsedFrameBoundary = 0;
};
Uint64 BuildSamplerKey(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture,
Bool forceNearestFiltering, Bool singleLevelView) const;
Bool forceNearestFiltering) const;
static VkFilter ToVkFilter(SamplerFilterMode mode);
static VkSamplerMipmapMode ToVkMipmapMode(SamplerMipmapMode mode);
static VkSamplerAddressMode ToVkAddressMode(SamplerWrapMode mode);
@@ -85,8 +67,6 @@ private:
Bool m_samplerAnisotropySupported = false;
Float m_maxSamplerAnisotropy = 1.0f;
UnorderedMap<Uint64, SamplerCacheEntry> m_samplers;
// Monotonic frame-boundary counter (bumped in OnFrameBoundary) for cache aging.
Uint64 m_frameBoundaryCounter = 0;
static inline XXH64_state_t* m_hashState = XXH64_createState();
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -120,21 +120,31 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
static Bool TryResolveSampleCountFlagBits(Int requestedSamples, VkSampleCountFlagBits& outSampleCount) {
// GL promises "at least the requested samples", so a non-power-of-two
// request (legal in GL, e.g. 3) rounds up to the next Vulkan bit.
if (requestedSamples <= 1) {
switch (requestedSamples) {
case 1:
outSampleCount = VK_SAMPLE_COUNT_1_BIT;
return true;
}
if (requestedSamples > 64) {
case 2:
outSampleCount = VK_SAMPLE_COUNT_2_BIT;
return true;
case 4:
outSampleCount = VK_SAMPLE_COUNT_4_BIT;
return true;
case 8:
outSampleCount = VK_SAMPLE_COUNT_8_BIT;
return true;
case 16:
outSampleCount = VK_SAMPLE_COUNT_16_BIT;
return true;
case 32:
outSampleCount = VK_SAMPLE_COUNT_32_BIT;
return true;
case 64:
outSampleCount = VK_SAMPLE_COUNT_64_BIT;
return true;
default:
return false;
}
Uint32 bit = 1;
while (bit < static_cast<Uint32>(requestedSamples)) {
bit <<= 1;
}
outSampleCount = static_cast<VkSampleCountFlagBits>(bit);
return true;
}
static Bool IsCubeMapFaceUploadTarget(TextureUploadTarget target) {
@@ -577,7 +587,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_allocator = initInfo.allocator;
m_commandPool = initInfo.commandPool;
m_graphicsQueue = initInfo.graphicsQueue;
m_imageFormatListSupported = initInfo.imageFormatListSupported;
m_currentFrameIndex = 0;
m_deferredReleases.clear();
m_deferredReleases.resize(initInfo.frameCount);
@@ -597,14 +606,9 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
void VkTextureManager::Shutdown() {
if (m_device != VK_NULL_HANDLE) {
ReclaimCompletedUploads(/*waitAll=*/true);
}
DestroyDeferredReleases();
++m_resourceEraseEpoch; // every memoized resource pointer dies with the map
m_textureResources.clear();
m_aliveObjects.clear();
m_storageImageTextures.clear();
m_device = VK_NULL_HANDLE;
m_physicalDevice = VK_NULL_HANDLE;
@@ -623,27 +627,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
frameIndex, m_deferredViewReleases.size());
m_currentFrameIndex = frameIndex;
CollectDeferredReleases(frameIndex);
ReclaimCompletedUploads();
// Frame-boundary GC: every 64 frame boundaries (~1 s at 60 fps) bounds the reclaim
// latency for dead textures regardless of draw traffic — workloads that churn
// textures through clears/readbacks alone never reach the draw-gated
// CollectGarbage. Must run after CollectDeferredReleases above: the prune defers
// its releases into this frame's slot, which was just drained, so they are
// destroyed only after the slot's fence has been waited again one full frame-ring
// cycle from now (never while an in-flight frame may still reference them).
constexpr Uint32 kGcFrameInterval = 64;
++m_gcFrameCounter;
if (m_gcFrameCounter % kGcFrameInterval == 0) {
PruneDeadTextures();
}
}
void VkTextureManager::CollectAllDeferredReleases() {
const SizeT frameCount = std::min(m_deferredReleases.size(), m_deferredViewReleases.size());
for (SizeT frameIndex = 0; frameIndex < frameCount; ++frameIndex) {
CollectDeferredReleases(static_cast<Uint32>(frameIndex));
}
}
void VkTextureManager::EraseTrackedTexture(const TextureIdentity& identity) {
@@ -653,10 +636,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_textureResources.erase(resourceIt);
}
m_aliveObjects.erase(identity);
m_storageImageTextures.erase(identity);
// Invalidate every cross-draw sampled-texture memo: the erased
// resource's address may be reused by a future emplace.
++m_resourceEraseEpoch;
}
void VkTextureManager::PruneStaleTextureAliases(MG_State::GLState::ITextureObject* texture) {
@@ -712,63 +691,32 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
}
// Cross-draw memo probe (see SyncedTextureMemoEntry): skips both map
// lookups and the (re)registration path for repeat-bound textures.
TextureResource* resourcePtr = nullptr;
for (Uint32 i = 0; i < kSyncedTextureMemoSize; ++i) {
const SyncedTextureMemoEntry& memo = m_syncedTextureMemo[i];
if (memo.texture == &texture && memo.lifetimeId == identity.lifetimeId &&
memo.eraseEpoch == m_resourceEraseEpoch) {
resourcePtr = memo.resource;
break;
auto aliveIt = m_aliveObjects.find(identity);
if (aliveIt != m_aliveObjects.end() && aliveIt->second.expired()) {
EraseTrackedTexture(aliveIt->first);
aliveIt = m_aliveObjects.end();
}
// Only (re)register and prune when this (texture, lifetime) pair is new: stale
// aliases can only come into existence through an address reuse, which by
// construction introduces a new identity. Doing this unconditionally made every
// sampled-texture sync scan the entire alive-texture map per draw.
if (aliveIt == m_aliveObjects.end()) {
const auto& liveTexture = MG_State::pGLContext->GetTextureObject(texture.GetExternalIndex());
if (liveTexture && liveTexture.get() == &texture) {
m_aliveObjects[identity] = WeakPtr<MG_State::GLState::ITextureObject>(liveTexture);
PruneStaleTextureAliases(&texture);
}
}
if (resourcePtr == nullptr) {
auto aliveIt = m_aliveObjects.find(identity);
if (aliveIt != m_aliveObjects.end() && aliveIt->second.expired()) {
EraseTrackedTexture(aliveIt->first);
aliveIt = m_aliveObjects.end();
}
// Only (re)register and prune when this (texture, lifetime) pair is new: stale
// aliases can only come into existence through an address reuse, which by
// construction introduces a new identity. Doing this unconditionally made every
// sampled-texture sync scan the entire alive-texture map per draw.
if (aliveIt == m_aliveObjects.end()) {
WeakPtr<MG_State::GLState::ITextureObject> aliveTexture;
const auto& liveTexture = MG_State::pGLContext->GetTextureObject(texture.GetExternalIndex());
if (liveTexture && liveTexture.get() == &texture) {
aliveTexture = liveTexture;
} else {
// The name lookup legally fails while the object is alive: the name was
// deleted with the texture still attached to an FBO (the attachment's
// SharedPtr keeps it alive), or the name was reused by a new texture, or
// this is a default texture object (name 0 lives outside the name map).
// Register through the object's own control block so the resource created
// below still participates in weak-expiry GC instead of becoming an
// orphan no reclamation path can reach until Shutdown.
aliveTexture = texture.weak_from_this();
}
if (!aliveTexture.expired()) {
m_aliveObjects[identity] = Move(aliveTexture);
PruneStaleTextureAliases(&texture);
}
}
auto it = m_textureResources.find(identity);
if (it == m_textureResources.end()) {
TextureResource initial{};
auto [insertIt, _] = m_textureResources.emplace(identity, Move(initial));
it = insertIt;
}
resourcePtr = &(it->second);
m_syncedTextureMemo[m_syncedTextureMemoNext] =
SyncedTextureMemoEntry{&texture, identity.lifetimeId, m_resourceEraseEpoch, resourcePtr};
m_syncedTextureMemoNext = (m_syncedTextureMemoNext + 1) % kSyncedTextureMemoSize;
auto it = m_textureResources.find(identity);
if (it == m_textureResources.end()) {
TextureResource initial{};
auto [insertIt, _] = m_textureResources.emplace(identity, Move(initial));
it = insertIt;
}
if (!SyncTexture(texture, *resourcePtr)) {
if (!SyncTexture(texture, it->second)) {
MGLOG_D("%s: Syncing texture %d failed", __func__, texture.GetExternalIndex());
return nullptr;
}
@@ -782,11 +730,11 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
}
if (!recorded) {
m_drawSyncedThisDraw.push_back({identity, resourcePtr});
m_drawSyncedThisDraw.push_back({identity, &(it->second)});
}
}
return resourcePtr;
return &(it->second);
}
VkImageView VkTextureManager::GetOrCreateViewAtMipLevel(MG_State::GLState::ITextureObject& texture, Uint32 mipLevel) {
@@ -830,12 +778,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return VK_NULL_HANDLE;
}
const Bool framebufferSrgbEnabled =
MG_State::pGLContext->IsCapabilityEnabled(MobileGL::CapabilityInput::FramebufferSrgb);
const VkFormat attachmentFormat = ResolveSrgbAttachmentWriteFormat(resource->format, framebufferSrgbEnabled);
if (attachmentFormat == resource->format && baseArrayLayer == 0 && layerCount == resource->arrayLayers &&
viewType == resource->viewType) {
if (baseArrayLayer == 0 && layerCount == resource->arrayLayers && viewType == resource->viewType) {
return GetOrCreateViewAtMipLevel(texture, mipLevel);
}
@@ -844,7 +787,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
.baseArrayLayer = baseArrayLayer,
.layerCount = layerCount,
.viewType = viewType,
.viewFormat = attachmentFormat,
};
auto it = resource->attachmentViews.find(key);
if (it == resource->attachmentViews.end()) {
@@ -855,7 +797,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return attachmentView;
}
attachmentView = CreateImageView(resource->image, attachmentFormat, resource->aspect, viewType,
attachmentView = CreateImageView(resource->image, resource->format, resource->aspect, viewType,
mipLevel, 1, baseArrayLayer, layerCount);
if (attachmentView == VK_NULL_HANDLE) {
MGLOG_D("%s: CreateImageView failed for textureId=%d mipLevel=%u baseArrayLayer=%u layerCount=%u viewType=%d",
@@ -1071,16 +1013,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return view;
}
void VkTextureManager::StampTextureRecordingUse(MG_State::GLState::ITextureObject* texture) {
if (texture == nullptr) {
return;
}
auto it = m_textureResources.find(MakeTextureIdentity(texture));
if (it != m_textureResources.end()) {
it->second.lastRecordingGeneration = m_recordingGeneration;
}
}
void VkTextureManager::UpdateTrackedImageLayout(MG_State::GLState::ITextureObject* texture, VkImageLayout newLayout) {
MOBILEGL_ASSERT(texture != nullptr, "UpdateTrackedImageLayout: texture is null");
auto it = m_textureResources.find(MakeTextureIdentity(texture));
@@ -1108,8 +1040,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
MOBILEGL_ASSERT(writtenMipLevel < resource.mipLevels,
"UpdateTrackedImageLayoutAfterAttachmentWrite: textureId=%d mipLevel=%u out of range %u",
texture->GetExternalIndex(), writtenMipLevel, resource.mipLevels);
// Pre-pass stream bookkeeping: the render pass that just ended wrote this image.
StampResourceRecordingUse(resource);
if (resource.layout != newLayout && resource.mipLevels > 1) {
VkPipelineStageFlags srcStageMask = VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT;
@@ -1194,8 +1124,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VK_ACCESS_SHADER_READ_BIT, resource->aspect, 0, resource->mipLevels,
resource->arrayLayers);
MOBILEGL_ASSERT(ok, "TransitionTextureForSampling: transition failed for textureId=%d", texture.GetExternalIndex());
// Pre-pass stream bookkeeping: a command referencing the image was recorded.
StampResourceRecordingUse(*resource);
return ok;
}
@@ -1225,44 +1153,11 @@ namespace MobileGL::MG_Backend::DirectVulkan {
resource->aspect, 0, resource->mipLevels, resource->arrayLayers);
MOBILEGL_ASSERT(ok, "TransitionTextureForStorageImage: transition failed for textureId=%d",
texture.GetExternalIndex());
// Pre-pass stream bookkeeping: a command referencing the image was recorded.
StampResourceRecordingUse(*resource);
return ok;
}
void VkTextureManager::MarkStorageImageTexture(MG_State::GLState::ITextureObject& texture) {
m_storageImageTextures.insert(MakeTextureIdentity(&texture));
}
Bool VkTextureManager::NeedsStorageUsageUpgrade(MG_State::GLState::ITextureObject& texture) const {
const TextureIdentity identity = MakeTextureIdentity(&texture);
if (m_storageImageTextures.find(identity) == m_storageImageTextures.end()) {
return false;
}
const auto it = m_textureResources.find(identity);
// No image yet: the first sync creates it with STORAGE straight away, so there is nothing
// to preserve and nothing to order against.
return it != m_textureResources.end() && it->second.image != VK_NULL_HANDLE &&
!it->second.storageUsageResolved;
}
Bool VkTextureManager::NeedsMipChainGrowth(MG_State::GLState::ITextureObject& texture) const {
const TextureIdentity identity = MakeTextureIdentity(&texture);
const auto it = m_textureResources.find(identity);
// No image yet: the first sync sizes the chain from the levels the texture already
// defines, so nothing is recreated and there is nothing to order against.
if (it == m_textureResources.end() || it->second.image == VK_NULL_HANDLE) {
return false;
}
const TextureResource& resource = it->second;
const IntVec3 extent = {static_cast<Int>(resource.extent.width), static_cast<Int>(resource.extent.height),
static_cast<Int>(resource.depth)};
return resource.mipLevels < ComputeFullMipLevelCount(extent);
}
Bool VkTextureManager::NeedsStorageImagePreparation(MG_State::GLState::ITextureObject& texture) const {
const TextureIdentity identity = MakeTextureIdentity(&texture);
const auto it = m_textureResources.find(identity);
const auto it = m_textureResources.find(MakeTextureIdentity(&texture));
if (it == m_textureResources.end()) {
return true;
}
@@ -1270,12 +1165,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
if (resource.image == VK_NULL_HANDLE || resource.layout != VK_IMAGE_LAYOUT_GENERAL) {
return true;
}
// The image predates this texture's first image-unit binding, so it was created without
// STORAGE usage and has to be recreated - which is illegal inside a render pass.
if (!resource.storageUsageResolved &&
m_storageImageTextures.find(identity) != m_storageImageTextures.end()) {
return true;
}
// Mirror SyncTexture's cross-draw skip condition: any version drift means the sync
// path may upload or rebuild, both of which need the render pass ended first.
const auto* mipTexture = MG_State::GLState::AsMipmapTexture(&texture);
@@ -1323,22 +1212,10 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
SizeT VkTextureManager::CollectGarbage() {
// Draw-gated stagger (1 in 256 calls): keeps the per-draw cost at one counter
// bump. The guaranteed reclaim path is the frame-boundary prune in BeginFrame;
// this remains as a cheap assist so draw-heavy workloads reclaim sooner.
m_gcCounter++;
if (m_gcCounter != 0) {
return 0;
}
return PruneDeadTextures();
}
SizeT VkTextureManager::PruneDeadTextures() {
// Erasing entries would dangle the raw TextureResource pointers memoized for the
// current draw; every call path (BeginFrame, and CollectGarbage at the top of a
// freshly opened draw-sync scope) runs before any memo entry is recorded.
MOBILEGL_ASSERT(m_drawSyncedThisDraw.empty(),
"PruneDeadTextures: draw-sync memo holds raw resource pointers an erase would dangle");
Vector<MG_State::GLState::ITextureObject*> expiredTextures;
expiredTextures.reserve(m_aliveObjects.size());
@@ -1350,25 +1227,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
for (auto* texture : expiredTextures) {
PruneStaleTextureAliases(texture);
}
SizeT prunedCount = expiredTextures.size();
// Orphan sweep: after the pass above, m_aliveObjects holds only live entries.
// Registration in SyncTextureAndGetDescriptor cannot fail for a SharedPtr-owned
// texture (weak_from_this fallback), so a resource whose identity has no alive
// entry has no trackable owner: its GL-side object is gone, or was never
// shared-owned, in which case recreation on a later sync is the safe fallback.
// Destruction goes through the per-frame deferred queues, never immediate.
Vector<TextureIdentity> orphanIdentities;
for (auto it = m_textureResources.begin(); it != m_textureResources.end(); ++it) {
if (m_aliveObjects.find(it->first) == m_aliveObjects.end()) {
orphanIdentities.emplace_back(it->first);
}
}
for (const auto& identity : orphanIdentities) {
EraseTrackedTexture(identity);
}
prunedCount += orphanIdentities.size();
return prunedCount;
return expiredTextures.size();
}
Bool VkTextureManager::SyncTexture(MG_State::GLState::ITextureObject &texture,
@@ -1382,13 +1241,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const auto* syncingMipTexture = MG_State::GLState::AsMipmapTexture(&texture);
const Uint32 syncingMipLevelCount =
syncingMipTexture != nullptr ? syncingMipTexture->GetMipmapLevelCount() : 0u;
// A pending storage-usage upgrade also has to bust the skip: nothing about the texture's
// content or params changed, but the image itself must be recreated with STORAGE usage
// before it can back an image-unit descriptor.
const Bool storageUpgradePending =
!outResource.storageUsageResolved &&
m_storageImageTextures.find(MakeTextureIdentity(&texture)) != m_storageImageTextures.end();
if (outResource.image != VK_NULL_HANDLE && !storageUpgradePending &&
if (outResource.image != VK_NULL_HANDLE &&
outResource.syncedContentVersion == syncingContentVersion &&
outResource.syncedTextureParamsVersion == texture.GetTextureParamsVersion() &&
outResource.syncedMipLevelCount == syncingMipLevelCount) {
@@ -1458,23 +1311,11 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const IntVec3 &texelSize, SizeT byteSize, Uint32 mipLevels,
TextureResource &resource) {
const TextureFormatInfo formatInfo = ResolveTextureFormatInfo(texture.GetFormat());
VkFormat format = formatInfo.format;
const VkFormat format = formatInfo.format;
if (format == VK_FORMAT_UNDEFINED) {
MGLOG_D("%s: format == VK_FORMAT_UNDEFINED", __func__);
return false;
}
// X8_D24 lacks optimal-tiling support on several drivers (lavapipe included);
// D32_SFLOAT holds every 24-bit depth value exactly, and the upload path
// converts the shadow words to float (see the pure-depth branch below).
if (format == VK_FORMAT_X8_D24_UNORM_PACK32) {
VkFormatProperties formatProperties{};
vkGetPhysicalDeviceFormatProperties(m_physicalDevice, format, &formatProperties);
constexpr VkFormatFeatureFlags kDepthAttachmentAndSample =
VK_FORMAT_FEATURE_DEPTH_STENCIL_ATTACHMENT_BIT | VK_FORMAT_FEATURE_SAMPLED_IMAGE_BIT;
if ((formatProperties.optimalTilingFeatures & kDepthAttachmentAndSample) != kDepthAttachmentAndSample) {
format = VK_FORMAT_D32_SFLOAT;
}
}
if (texelSize.x() <= 0 || texelSize.y() <= 0 /*|| byteSize == 0*/) {
MGLOG_D("%s: texelSize or byteSize is zero", __func__);
return false;
@@ -1484,17 +1325,8 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return false;
}
const Bool isMultisampleTexture = IsMultisampleTextureUploadTarget(uploadTarget);
// A texture that has only ever defined level 0 gets a single-level backing
// (ANGLE's model). Preallocating the full chain put every render target
// onto Adreno's multi-mip image layout and grew each texture by a third
// for levels most textures never define. Once a second level is defined
// the backing is recreated ONE time with the full chain (the
// preserve-copy path below carries the pixels over), so sequentially-
// defined atlas mips do not recreate per level, and glGenerateMipmap -
// which defines every level before syncing - works unchanged.
const Uint32 backingMipLevels =
isMultisampleTexture ? 1u
: (mipLevels > 1 ? std::max(mipLevels, ComputeFullMipLevelCount(texelSize)) : 1u);
isMultisampleTexture ? 1u : std::max(mipLevels, ComputeFullMipLevelCount(texelSize));
TextureShapeInfo shapeInfo{};
const Bool supportedShape = TryResolveTextureShapeInfo(texture, uploadTarget, texelSize, shapeInfo);
MOBILEGL_ASSERT(supportedShape,
@@ -1520,89 +1352,15 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const VkImageAspectFlags aspect = GetAspectMaskForFormat(format);
VkFormatProperties formatProperties{};
vkGetPhysicalDeviceFormatProperties(m_physicalDevice, format, &formatProperties);
// Only textures that have actually been bound to a GL image unit get STORAGE usage (and
// the MUTABLE_FORMAT it drags in for format-reinterpreting image views). Requesting it
// for every storage-capable colour texture costs real bandwidth: Adreno cannot keep UBWC
// compression on an image that may be written through a storage descriptor, so the whole
// render target - MC's included - runs uncompressed. MarkStorageImageTexture upgrades a
// texture before its first image-unit draw, and the usage below feeds the compatibility
// check so the upgrade recreates the image.
const Bool markedAsStorageImage =
m_storageImageTextures.find(MakeTextureIdentity(
const_cast<MG_State::GLState::ITextureObject*>(&texture))) != m_storageImageTextures.end();
// Storage-image CAPABILITY (does the format allow it at all) is deliberately separate from
// whether this texture actually needs the usage. MUTABLE_FORMAT keys off capability, as
// before: format-reinterpreting views are not a storage-only concern - the SAMPLED path
// needs them too (GetOrCreateSampledImageView bails out without it, see ~line 892), so
// tying MUTABLE_FORMAT to the image-unit mark would break sampled format reinterpretation
// for every texture that never becomes a storage image.
const Bool storageImageCapable =
const Bool supportsStorageImage =
!isMultisampleTexture &&
(aspect & VK_IMAGE_ASPECT_COLOR_BIT) != 0 &&
(formatProperties.optimalTilingFeatures & VK_FORMAT_FEATURE_STORAGE_IMAGE_BIT) != 0;
const Bool supportsStorageImage = storageImageCapable && markedAsStorageImage;
VkImageCreateFlags imageCreateFlags = shapeInfo.imageFlags;
if (storageImageCapable && IsMutableStorageImageFormat(format) &&
if (supportsStorageImage && IsMutableStorageImageFormat(format) &&
m_mutableFormatUnsupported.find(format) == m_mutableFormatUnsupported.end()) {
imageCreateFlags |= VK_IMAGE_CREATE_MUTABLE_FORMAT_BIT;
}
// sRGB color images attach through their UNORM twin while GL_FRAMEBUFFER_SRGB is
// disabled (see ResolveSrgbAttachmentWriteFormat), which needs format-reinterpreting
// views - multisample sRGB render targets included.
if (ResolveSrgbAttachmentWriteFormat(format, false) != format &&
(aspect & VK_IMAGE_ASPECT_COLOR_BIT) != 0 &&
m_mutableFormatUnsupported.find(format) == m_mutableFormatUnsupported.end()) {
imageCreateFlags |= VK_IMAGE_CREATE_MUTABLE_FORMAT_BIT;
}
VkImageUsageFlags desiredUsage =
VK_IMAGE_USAGE_SAMPLED_BIT |
(supportsStorageImage ? VK_IMAGE_USAGE_STORAGE_BIT : 0) |
((aspect & VK_IMAGE_ASPECT_COLOR_BIT) ? VK_IMAGE_USAGE_COLOR_ATTACHMENT_BIT : 0) |
(((aspect & VK_IMAGE_ASPECT_DEPTH_BIT) || (aspect & VK_IMAGE_ASPECT_STENCIL_BIT)) ?
VK_IMAGE_USAGE_DEPTH_STENCIL_ATTACHMENT_BIT :
0);
if (!isMultisampleTexture) {
desiredUsage |= VK_IMAGE_USAGE_TRANSFER_DST_BIT | VK_IMAGE_USAGE_TRANSFER_SRC_BIT;
}
// Round a multisample request up to a count the device supports for this
// format (GL only promises "at least"), mirroring the renderbuffer path.
if (isMultisampleTexture && resolvedSampleCount != VK_SAMPLE_COUNT_1_BIT) {
auto supportedIt = m_multisampleCountsByFormat.find(format);
if (supportedIt == m_multisampleCountsByFormat.end()) {
VkImageFormatProperties imageFormatProperties{};
VkSampleCountFlags supported = VK_SAMPLE_COUNT_1_BIT;
if (vkGetPhysicalDeviceImageFormatProperties(m_physicalDevice, format, shapeInfo.imageType,
VK_IMAGE_TILING_OPTIMAL, desiredUsage, imageCreateFlags,
&imageFormatProperties) == VK_SUCCESS) {
supported = imageFormatProperties.sampleCounts;
}
supportedIt = m_multisampleCountsByFormat.emplace(format, supported).first;
}
const VkSampleCountFlags supported = supportedIt->second;
if ((supported & resolvedSampleCount) == 0) {
Uint32 rounded = 0;
for (Uint32 bit = static_cast<Uint32>(resolvedSampleCount) << 1; bit <= VK_SAMPLE_COUNT_64_BIT;
bit <<= 1) {
if ((supported & bit) != 0) {
rounded = bit;
break;
}
}
if (rounded == 0) {
for (Uint32 bit = static_cast<Uint32>(resolvedSampleCount) >> 1; bit != 0; bit >>= 1) {
if ((supported & bit) != 0) {
rounded = bit;
break;
}
}
}
if (rounded != 0) {
resolvedSampleCount = static_cast<VkSampleCountFlagBits>(rounded);
}
}
}
const Bool compatible = resource.image != VK_NULL_HANDLE && resource.format == format &&
resource.extent.width == static_cast<Uint32>(texelSize.x()) &&
@@ -1612,7 +1370,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
resource.viewType == shapeInfo.viewType &&
resource.sampleCount == resolvedSampleCount &&
resource.imageCreateFlags == imageCreateFlags &&
resource.usageFlags == desiredUsage &&
resource.mipLevels == backingMipLevels;
if (compatible) {
if (resource.perMipViews.size() != backingMipLevels) {
@@ -1621,10 +1378,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
if (resource.perMipSampledViews.size() != backingMipLevels) {
resource.perMipSampledViews.resize(backingMipLevels, VK_NULL_HANDLE);
}
// Keeping the image is itself the answer to the mark: either it already carries
// STORAGE, or this format can never carry it. Either way there is nothing left to
// recreate, so stop reporting the texture as needing preparation.
resource.storageUsageResolved = markedAsStorageImage;
return true;
}
@@ -1639,10 +1392,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
resource.sampleCount == resolvedSampleCount &&
resource.imageCreateFlags == imageCreateFlags &&
resolvedSampleCount == VK_SAMPLE_COUNT_1_BIT &&
// '<=' rather than '<': a storage-usage upgrade recreates the image with an
// unchanged mip count, and its contents (a render target's pixels live only on the
// GPU) still have to survive. The vkCmdCopyImage below copies min(mipLevels).
resource.mipLevels <= backingMipLevels &&
resource.mipLevels < backingMipLevels &&
resource.layout != VK_IMAGE_LAYOUT_UNDEFINED;
std::unique_ptr<TextureResource> preservedResource;
@@ -1664,37 +1414,16 @@ namespace MobileGL::MG_Backend::DirectVulkan {
imageInfo.format = format;
imageInfo.tiling = VK_IMAGE_TILING_OPTIMAL;
imageInfo.initialLayout = VK_IMAGE_LAYOUT_UNDEFINED;
imageInfo.usage = desiredUsage;
imageInfo.samples = resolvedSampleCount;
// Bound the mutability. A blindly-mutable image has to be laid out so that ANY format in
// its compatibility class can be viewed, which costs bandwidth compression on tilers;
// naming the exact set instead lets the driver keep it. Only safe when that set really is
// exhaustive, so it is restricted to textures that are not image-unit bound: sampled views
// can only ever ask for ResolveSampledImageViewFormat's output, whereas glBindImageTexture
// may name any compatible format, which nothing here can enumerate ahead of time.
Vector<VkFormat> viewFormats;
VkImageFormatListCreateInfo formatListInfo{};
if (m_imageFormatListSupported && !supportsStorageImage &&
(imageInfo.flags & VK_IMAGE_CREATE_MUTABLE_FORMAT_BIT) != 0) {
viewFormats.push_back(format);
for (const SamplerNumericDomain domain : {SamplerNumericDomain::Float,
SamplerNumericDomain::SignedInteger,
SamplerNumericDomain::UnsignedInteger}) {
const VkFormat viewFormat = ResolveSampledImageViewFormat(format, domain);
if (viewFormat == VK_FORMAT_UNDEFINED) {
continue;
}
if (std::find(viewFormats.begin(), viewFormats.end(), viewFormat) == viewFormats.end()) {
viewFormats.push_back(viewFormat);
}
}
formatListInfo.sType = VK_STRUCTURE_TYPE_IMAGE_FORMAT_LIST_CREATE_INFO;
formatListInfo.viewFormatCount = static_cast<Uint32>(viewFormats.size());
formatListInfo.pViewFormats = viewFormats.data();
imageInfo.pNext = &formatListInfo;
imageInfo.usage = VK_IMAGE_USAGE_SAMPLED_BIT |
(supportsStorageImage ? VK_IMAGE_USAGE_STORAGE_BIT : 0) |
((aspect & VK_IMAGE_ASPECT_COLOR_BIT) ? VK_IMAGE_USAGE_COLOR_ATTACHMENT_BIT : 0) |
(((aspect & VK_IMAGE_ASPECT_DEPTH_BIT) || (aspect & VK_IMAGE_ASPECT_STENCIL_BIT)) ?
VK_IMAGE_USAGE_DEPTH_STENCIL_ATTACHMENT_BIT :
0);
if (!isMultisampleTexture) {
imageInfo.usage |= VK_IMAGE_USAGE_TRANSFER_DST_BIT | VK_IMAGE_USAGE_TRANSFER_SRC_BIT;
}
imageInfo.samples = resolvedSampleCount;
if (isMultisampleTexture || (imageInfo.flags & VK_IMAGE_CREATE_MUTABLE_FORMAT_BIT) != 0) {
VkImageFormatProperties imageFormatProperties{};
VkResult imageFormatResult = vkGetPhysicalDeviceImageFormatProperties(
@@ -1733,21 +1462,8 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VmaAllocationCreateInfo allocationInfo{};
allocationInfo.usage = VMA_MEMORY_USAGE_AUTO_PREFER_DEVICE;
allocationInfo.requiredFlags = VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT;
// Soft failure like the unsupported-sample-count path above: a driver can pass the
// vkGetPhysicalDeviceImageFormatProperties pre-check yet still refuse the creation
// (e.g. multisampled depth on lavapipe); the texture simply stays unbacked.
const VkResult createImageResult =
vmaCreateImage(m_allocator, &imageInfo, &allocationInfo, &resource.image, &resource.allocation, nullptr);
if (createImageResult != VK_SUCCESS) {
MGLOG_F("SyncTextureResource: vmaCreateImage failed (%d) textureId=%d extent=%ux%u depth=%u layers=%u "
"mips=%u samples=%d format=%d",
createImageResult, texture.GetExternalIndex(), imageInfo.extent.width, imageInfo.extent.height,
imageInfo.extent.depth, imageInfo.arrayLayers, imageInfo.mipLevels,
static_cast<Int>(imageInfo.samples), static_cast<Int>(imageInfo.format));
resource.image = VK_NULL_HANDLE;
resource.allocation = nullptr;
return false;
}
VK_VERIFY(vmaCreateImage(m_allocator, &imageInfo, &allocationInfo, &resource.image, &resource.allocation, nullptr),
"vmaCreateImage(texture)");
++m_textureImageEpoch; // a new attachment image invalidates cached render passes
resource.layout = VK_IMAGE_LAYOUT_UNDEFINED;
@@ -1764,8 +1480,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
resource.viewType = shapeInfo.viewType;
resource.sampleCount = resolvedSampleCount;
resource.imageCreateFlags = imageCreateFlags;
resource.usageFlags = imageInfo.usage;
resource.storageUsageResolved = markedAsStorageImage;
resource.syncedTextureParamsVersion = 0;
if (preservedResource) {
@@ -1815,28 +1529,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
m_deferredViewReleases[frameIndex].clear();
}
void VkTextureManager::ReclaimCompletedUploads(Bool waitAll) {
if (m_pendingUploadReclaims.empty()) {
return;
}
SizeT completed = 0;
for (; completed < m_pendingUploadReclaims.size(); ++completed) {
PendingUploadReclaim& entry = m_pendingUploadReclaims[completed];
if (waitAll) {
VK_VERIFY(vkWaitForFences(m_device, 1, &entry.fence, VK_TRUE, UINT64_MAX),
"vkWaitForFences(texture upload reclaim)");
} else if (vkGetFenceStatus(m_device, entry.fence) != VK_SUCCESS) {
break;
}
vkDestroyFence(m_device, entry.fence, nullptr);
vkFreeCommandBuffers(m_device, m_commandPool, 1, &entry.commandBuffer);
vmaDestroyBuffer(m_allocator, entry.stagingBuffer, entry.stagingAllocation);
}
m_pendingUploadReclaims.erase(m_pendingUploadReclaims.begin(),
m_pendingUploadReclaims.begin() + static_cast<std::ptrdiff_t>(completed));
}
void VkTextureManager::DestroyDeferredReleases() {
for (auto& deferredReleases : m_deferredReleases) {
deferredReleases.clear();
@@ -2038,119 +1730,17 @@ namespace MobileGL::MG_Backend::DirectVulkan {
return true;
}
// Combined depth-stencil images need per-aspect copies (VkBufferImageCopy aspectMask
// must have exactly one bit set), so de-interleave the shadow's GL wire format into
// a depth plane followed by a stencil plane per upload item.
// Combined depth-stencil images need per-aspect de-interleaved copies (VkBufferImageCopy
// aspectMask must have exactly one bit set). Until that is implemented, skip the upload
// instead of recording an invalid command buffer that kills the process.
const VkImageAspectFlags uploadAspectMask = GetAspectMaskForFormat(outResource.format);
const Bool isCombinedDepthStencil =
(uploadAspectMask & VK_IMAGE_ASPECT_DEPTH_BIT) && (uploadAspectMask & VK_IMAGE_ASPECT_STENCIL_BIT);
if (isCombinedDepthStencil) {
const Bool srcIsD24S8 = outResource.format == VK_FORMAT_D24_UNORM_S8_UINT;
const Bool srcIsD32FS8 = outResource.format == VK_FORMAT_D32_SFLOAT_S8_UINT;
if (!srcIsD24S8 && !srcIsD32FS8) {
MGLOG_E("UploadDirtyMipLevels: unsupported combined depth-stencil format %d for textureId=%d",
static_cast<Int>(outResource.format), mipmapTexture.GetExternalIndex());
for (const auto& item : uploadItems) {
mipmapTexture.MarkStorageDirty(item.target, item.level, false);
}
return true;
}
stagingSize = 0;
for (auto& item : uploadItems) {
const SizeT texelCount = static_cast<SizeT>(item.texelSize.x()) *
static_cast<SizeT>(item.texelSize.y()) *
static_cast<SizeT>(std::max(item.texelSize.z(), 1));
const SizeT shadowTexelSize = item.uploadByteSize / std::max<SizeT>(texelCount, 1);
MOBILEGL_ASSERT(shadowTexelSize == 4 || shadowTexelSize == 8,
"UploadDirtyMipLevels: unexpected depth-stencil shadow texel size %zu for textureId=%d",
shadowTexelSize, mipmapTexture.GetExternalIndex());
// Depth plane as the aspect's buffer-copy format (32-bit word for
// D24: low 24 bits; float for D32F), then one stencil byte per texel.
Vector<Uint8> deinterleaved(texelCount * 4 + texelCount);
Uint8* depthPlane = deinterleaved.data();
Uint8* stencilPlane = deinterleaved.data() + texelCount * 4;
const Uint8* shadow = static_cast<const Uint8*>(item.source);
for (SizeT t = 0; t < texelCount; ++t) {
if (shadowTexelSize == 8) {
// GL_FLOAT_32_UNSIGNED_INT_24_8_REV: float depth, then a word
// with stencil in its low 8 bits.
float depthValue;
Uint32 stencilWord;
std::memcpy(&depthValue, shadow + t * 8, sizeof(depthValue));
std::memcpy(&stencilWord, shadow + t * 8 + 4, sizeof(stencilWord));
if (srcIsD32FS8) {
std::memcpy(depthPlane + t * 4, &depthValue, sizeof(depthValue));
} else {
const float clamped = std::min(std::max(depthValue, 0.0f), 1.0f);
const Uint32 depthWord = static_cast<Uint32>(clamped * 16777215.0f + 0.5f);
std::memcpy(depthPlane + t * 4, &depthWord, sizeof(depthWord));
}
stencilPlane[t] = static_cast<Uint8>(stencilWord & 0xFFu);
} else {
// GL_UNSIGNED_INT_24_8: depth in the high 24 bits, stencil low 8.
Uint32 packed;
std::memcpy(&packed, shadow + t * 4, sizeof(packed));
if (srcIsD24S8) {
const Uint32 depthWord = packed >> 8;
std::memcpy(depthPlane + t * 4, &depthWord, sizeof(depthWord));
} else {
const float depthValue = static_cast<float>(packed >> 8) / 16777215.0f;
std::memcpy(depthPlane + t * 4, &depthValue, sizeof(depthValue));
}
stencilPlane[t] = static_cast<Uint8>(packed & 0xFFu);
}
}
item.expandedData = Move(deinterleaved);
item.source = item.expandedData.data();
item.uploadByteSize = item.expandedData.size();
item.offset = stagingSize;
stagingSize += static_cast<VkDeviceSize>(item.uploadByteSize);
}
}
// Pure-depth images whose canonical shadow layout differs from the image texel
// layout (the shadow keeps a full-scale 16/32-bit unorm word or a float; the
// image may be X8_D24 or a D32_SFLOAT fallback) convert per texel here.
if (uploadAspectMask == VK_IMAGE_ASPECT_DEPTH_BIT) {
const TextureInternalFormat depthInternal = mipmapTexture.GetFormat();
const Bool shadowIsFloat = depthInternal == TextureInternalFormat::DepthComponent32F;
const Bool dstIsFloat = outResource.format == VK_FORMAT_D32_SFLOAT;
const Bool dstIsD24Word = outResource.format == VK_FORMAT_X8_D24_UNORM_PACK32;
stagingSize = 0;
for (auto& item : uploadItems) {
const SizeT texelCount = static_cast<SizeT>(item.texelSize.x()) *
static_cast<SizeT>(item.texelSize.y()) *
static_cast<SizeT>(std::max(item.texelSize.z(), 1));
const SizeT shadowTexelSize = item.uploadByteSize / std::max<SizeT>(texelCount, 1);
const Bool needsConversion =
(dstIsFloat && !shadowIsFloat) || (dstIsD24Word && shadowTexelSize == 4 && !shadowIsFloat);
if (needsConversion) {
Vector<Uint8> converted(texelCount * 4);
const Uint8* shadow = static_cast<const Uint8*>(item.source);
for (SizeT t = 0; t < texelCount; ++t) {
Uint32 wide = 0;
if (shadowTexelSize == 2) {
Uint16 raw = 0;
std::memcpy(&raw, shadow + t * 2, sizeof(raw));
wide = (static_cast<Uint32>(raw) << 16) | raw;
} else {
std::memcpy(&wide, shadow + t * 4, sizeof(wide));
}
if (dstIsFloat) {
const float value = static_cast<float>(static_cast<double>(wide) / 4294967295.0);
std::memcpy(converted.data() + t * 4, &value, sizeof(value));
} else { // X8_D24: depth in the low 24 bits of a 32-bit word
const Uint32 word = wide >> 8;
std::memcpy(converted.data() + t * 4, &word, sizeof(word));
}
}
item.expandedData = Move(converted);
item.source = item.expandedData.data();
item.uploadByteSize = item.expandedData.size();
}
item.offset = stagingSize;
stagingSize += static_cast<VkDeviceSize>(item.uploadByteSize);
if ((uploadAspectMask & VK_IMAGE_ASPECT_DEPTH_BIT) && (uploadAspectMask & VK_IMAGE_ASPECT_STENCIL_BIT)) {
MGLOG_E("UploadDirtyMipLevels: skipping unimplemented depth-stencil data upload for textureId=%d",
mipmapTexture.GetExternalIndex());
for (const auto& item : uploadItems) {
mipmapTexture.MarkStorageDirty(item.target, item.level, false);
}
return true;
}
VkBuffer stagingBuffer = VK_NULL_HANDLE;
@@ -2203,14 +1793,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
aspectMask, 0, outResource.mipLevels, outResource.arrayLayers);
MOBILEGL_ASSERT(ok, "TransitionImageLayout to VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL failed");
// Array textures keep their GL "depth" in VkImage array layers, so the
// copy must address layerCount, not imageExtent.depth (which is invalid
// for 2D images and silently dropped every layer past the first).
const Bool depthSelectsArrayLayer = outResource.viewType == VK_IMAGE_VIEW_TYPE_1D_ARRAY ||
outResource.viewType == VK_IMAGE_VIEW_TYPE_2D_ARRAY ||
outResource.viewType == VK_IMAGE_VIEW_TYPE_CUBE_ARRAY;
for (const auto& item : uploadItems) {
const Uint32 depthOrLayers = item.texelSize.z() > 0 ? static_cast<Uint32>(item.texelSize.z()) : 1u;
VkBufferImageCopy copy{};
copy.bufferOffset = item.offset;
copy.bufferRowLength = 0;
@@ -2218,24 +1801,10 @@ namespace MobileGL::MG_Backend::DirectVulkan {
copy.imageSubresource.aspectMask = aspectMask;
copy.imageSubresource.mipLevel = item.level;
copy.imageSubresource.baseArrayLayer = item.baseArrayLayer;
copy.imageSubresource.layerCount = depthSelectsArrayLayer ? depthOrLayers : 1;
copy.imageSubresource.layerCount = 1;
copy.imageOffset = {0, 0, 0};
copy.imageExtent = {static_cast<Uint32>(item.texelSize.x()), static_cast<Uint32>(item.texelSize.y()),
depthSelectsArrayLayer ? 1u : depthOrLayers};
if (isCombinedDepthStencil) {
const SizeT texelCount = static_cast<SizeT>(item.texelSize.x()) *
static_cast<SizeT>(item.texelSize.y()) *
static_cast<SizeT>(std::max(item.texelSize.z(), 1));
VkBufferImageCopy depthCopy = copy;
depthCopy.imageSubresource.aspectMask = VK_IMAGE_ASPECT_DEPTH_BIT;
VkBufferImageCopy stencilCopy = copy;
stencilCopy.imageSubresource.aspectMask = VK_IMAGE_ASPECT_STENCIL_BIT;
stencilCopy.bufferOffset = item.offset + static_cast<VkDeviceSize>(texelCount) * 4;
const VkBufferImageCopy copies[2] = {depthCopy, stencilCopy};
vkCmdCopyBufferToImage(commandBuffer, stagingBuffer, outResource.image,
VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL, 2, copies);
continue;
}
item.texelSize.z() > 0 ? static_cast<Uint32>(item.texelSize.z()) : 1u};
vkCmdCopyBufferToImage(commandBuffer, stagingBuffer, outResource.image, VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL,
1, &copy);
}
@@ -2266,23 +1835,11 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VK_VERIFY(vkCreateFence(m_device, &fenceInfo, nullptr, &uploadFence), "vkCreateFence(texture upload)");
VK_VERIFY(vkQueueSubmit(m_graphicsQueue, 1, &submitInfo, uploadFence), "vkQueueSubmit(texture)");
// Do NOT wait the fence here: this submit sits behind the previous
// frame's rendering on the queue, so a synchronous wait stalls the CPU
// until the GPU drains - a per-frame vkQueueWaitIdle for any workload
// with animated textures. Ordering against the current frame's draws is
// already guaranteed (its command buffer is submitted later, at
// present), so only the transient objects need to survive execution;
// park them until the fence signals.
m_pendingUploadReclaims.push_back({uploadFence, commandBuffer, stagingBuffer, stagingAllocation});
ReclaimCompletedUploads();
// Backstop for pathological upload storms: bound in-flight staging
// memory by blocking on the oldest upload only once the list is deep.
constexpr SizeT kMaxPendingTextureUploads = 16;
if (m_pendingUploadReclaims.size() > kMaxPendingTextureUploads) {
VK_VERIFY(vkWaitForFences(m_device, 1, &m_pendingUploadReclaims.front().fence, VK_TRUE, UINT64_MAX),
"vkWaitForFences(texture upload backstop)");
ReclaimCompletedUploads();
}
VK_VERIFY(vkWaitForFences(m_device, 1, &uploadFence, VK_TRUE, UINT64_MAX), "vkWaitForFences(texture upload)");
vkDestroyFence(m_device, uploadFence, nullptr);
vkFreeCommandBuffers(m_device, m_commandPool, 1, &commandBuffer);
vmaDestroyBuffer(m_allocator, stagingBuffer, stagingAllocation);
if (!ok) {
MGLOG_D("%s: texture upload cmd failed", __func__);
@@ -28,9 +28,6 @@ public:
// manager keys its per-draw fast path on this so an attachment's image recreation
// invalidates the cached render pass (dirty-flag tracking; portable to Vulkan 1.1).
Uint64 GetTextureImageEpoch() const { return m_textureImageEpoch; }
// Bumped whenever any tracked texture resource is erased; cached
// TextureResource pointers are valid only while this is unchanged.
Uint64 GetResourceEraseEpoch() const { return m_resourceEraseEpoch; }
struct TextureIdentity {
MG_State::GLState::ITextureObject* texture = nullptr;
@@ -56,9 +53,6 @@ public:
VkCommandPool commandPool = VK_NULL_HANDLE;
VkQueue graphicsQueue = VK_NULL_HANDLE;
Uint32 frameCount = 0;
// VK_KHR_image_format_list is enabled: MUTABLE_FORMAT images can name the exact set of
// formats they will be viewed as, which is what lets a tiler keep them compressed.
Bool imageFormatListSupported = false;
};
struct TextureResource {
@@ -67,16 +61,12 @@ public:
Uint32 baseArrayLayer = 0;
Uint32 layerCount = 1;
VkImageViewType viewType = VK_IMAGE_VIEW_TYPE_2D;
// May differ from the image format: sRGB images attach through their UNORM
// twin while GL_FRAMEBUFFER_SRGB is disabled.
VkFormat viewFormat = VK_FORMAT_UNDEFINED;
Bool operator==(const AttachmentViewKey& other) const {
return mipLevel == other.mipLevel &&
baseArrayLayer == other.baseArrayLayer &&
layerCount == other.layerCount &&
viewType == other.viewType &&
viewFormat == other.viewFormat;
viewType == other.viewType;
}
};
@@ -87,8 +77,6 @@ public:
hash ^= std::hash<Uint32>{}(key.layerCount) + 0x9e3779b9u + (hash << 6) + (hash >> 2);
hash ^= std::hash<Uint32>{}(static_cast<Uint32>(key.viewType)) +
0x9e3779b9u + (hash << 6) + (hash >> 2);
hash ^= std::hash<Uint32>{}(static_cast<Uint32>(key.viewFormat)) +
0x9e3779b9u + (hash << 6) + (hash >> 2);
return hash;
}
};
@@ -169,25 +157,7 @@ public:
VkImageViewType viewType = VK_IMAGE_VIEW_TYPE_2D;
VkSampleCountFlagBits sampleCount = VK_SAMPLE_COUNT_1_BIT;
VkImageCreateFlags imageCreateFlags = 0;
// Usage the live image was created with. STORAGE is only requested for textures that
// have actually been bound to a GL image unit, because on Adreno a storage-capable
// image loses UBWC bandwidth compression; a later image binding upgrades the usage
// and recreates the image, so the resolved usage has to be part of the compatibility
// check that decides whether the existing image can be kept.
VkImageUsageFlags usageFlags = 0;
// True once this image was (re)resolved while the texture was already marked as an
// image-unit texture. Distinguishes "not upgraded yet" from "cannot be upgraded"
// (a format whose optimalTilingFeatures lack STORAGE_IMAGE never gains the bit), so
// NeedsStorageImagePreparation cannot ask for a recreate that will never happen.
Bool storageUsageResolved = false;
Uint16 syncedTextureParamsVersion = 0;
// Recording generation (VkTextureManager::GetRecordingGeneration) of the last
// command referencing this image that was recorded into the CURRENT frame
// command buffer. An image untouched by the open recording may have its
// out-of-pass work (deferred clears, sampled-layout transitions) recorded
// into the frame's PRE command buffer - which executes strictly before the
// frame's commands - instead of splitting the active render pass.
Uint64 lastRecordingGeneration = 0;
// Snapshot of ITextureObject::GetContentVersion() at the last successful sync;
// lets SyncTexture skip the whole re-check/re-upload when content is unchanged.
Uint64 syncedContentVersion = 0;
@@ -220,10 +190,7 @@ public:
std::swap(this->viewType, that.viewType);
std::swap(this->sampleCount, that.sampleCount);
std::swap(this->imageCreateFlags, that.imageCreateFlags);
std::swap(this->usageFlags, that.usageFlags);
std::swap(this->storageUsageResolved, that.storageUsageResolved);
std::swap(this->syncedTextureParamsVersion, that.syncedTextureParamsVersion);
std::swap(this->lastRecordingGeneration, that.lastRecordingGeneration);
std::swap(this->syncedContentVersion, that.syncedContentVersion);
std::swap(this->syncedMipLevelCount, that.syncedMipLevelCount);
}
@@ -284,8 +251,6 @@ public:
viewType = VK_IMAGE_VIEW_TYPE_2D;
sampleCount = VK_SAMPLE_COUNT_1_BIT;
imageCreateFlags = 0;
usageFlags = 0;
storageUsageResolved = false;
syncedTextureParamsVersion = 0;
syncedContentVersion = 0;
syncedMipLevelCount = 0;
@@ -302,10 +267,6 @@ public:
Bool Initialize(const InitInfo& initInfo);
void Shutdown();
void BeginFrame(Uint32 frameIndex);
// Drains every frame slot's deferred image/view releases. Only valid when
// the caller has proven every queue submission complete; used by the
// present-less frame-boundary drain.
void CollectAllDeferredReleases();
TextureResource* SyncTextureAndGetDescriptor(
MG_State::GLState::ITextureObject& texture);
@@ -324,36 +285,6 @@ public:
VkImageLayout newLayout);
Bool TransitionTextureForSampling(VkCommandBuffer commandBuffer, MG_State::GLState::ITextureObject& texture);
Bool TransitionTextureForStorageImage(VkCommandBuffer commandBuffer, MG_State::GLState::ITextureObject& texture);
// Recording-generation bookkeeping for the pre-pass command stream. The
// generation advances every time the frame command buffer (re)begins
// recording; a resource whose stamp does not match was not referenced by
// any command in the open recording, so its out-of-pass work may safely
// execute ahead of the whole recording (in the pre command buffer).
void AdvanceRecordingGeneration() { ++m_recordingGeneration; }
void StampResourceRecordingUse(TextureResource& resource) const {
resource.lastRecordingGeneration = m_recordingGeneration;
}
// Map-lookup variant for callers that only hold the GL texture object.
void StampTextureRecordingUse(MG_State::GLState::ITextureObject* texture);
Bool WasTouchedThisRecording(const TextureResource& resource) const {
return resource.lastRecordingGeneration == m_recordingGeneration;
}
// Records that this texture is bound to a GL image unit, so its image must carry
// VK_IMAGE_USAGE_STORAGE_BIT. Must be called before NeedsStorageImagePreparation, and
// therefore before the render pass is committed: an image that has to be upgraded is
// recreated, which is illegal inside a render pass. Sticky for the texture's lifetime -
// GL lets an image binding come and go, and re-creating the image every time it does
// would cost far more than the compression it wins back.
void MarkStorageImageTexture(MG_State::GLState::ITextureObject& texture);
// True when this texture is marked but its live image predates the mark, i.e. the next sync
// will recreate it with STORAGE usage and copy the old contents forward. Callers use this to
// submit their pending recording first, so that copy cannot read pre-flush content.
Bool NeedsStorageUsageUpgrade(MG_State::GLState::ITextureObject& texture) const;
// The same ordering question for the other recreate-and-preserve trigger: true when this
// texture's live image carries a shorter mip chain than a full one, so defining the missing
// levels recreates it and copies the old contents forward.
Bool NeedsMipChainGrowth(MG_State::GLState::ITextureObject& texture) const;
// Non-mutating probe for the per-draw storage-image fast path: true when preparing this
// texture as a storage image may need work that is illegal inside a render pass (resource
// creation, dirty-content upload, or a layout transition to GENERAL). Unknown state reports
@@ -400,9 +331,6 @@ public:
private:
// Bumped in SyncTextureResource right after vmaCreateImage(texture). See GetTextureImageEpoch().
Uint64 m_textureImageEpoch = 1;
// See AdvanceRecordingGeneration. Starts above every resource's default
// stamp of 0 so a fresh resource counts as untouched.
Uint64 m_recordingGeneration = 1;
Bool SyncTexture(MG_State::GLState::ITextureObject &texture,
TextureResource &outResource);
@@ -433,28 +361,18 @@ private:
void DeferViewRelease(VkImageView view);
void CollectDeferredReleases(Uint32 frameIndex);
void DestroyDeferredReleases();
// Frees the fence/command buffer/staging buffer of every in-flight texture
// upload whose fence has signaled (submission order = completion order on
// the single queue, so the scan stops at the first still-pending entry).
// waitAll blocks on every entry - Shutdown's drain.
void ReclaimCompletedUploads(Bool waitAll = false);
static TextureIdentity MakeTextureIdentity(MG_State::GLState::ITextureObject* texture);
void EraseTrackedTexture(const TextureIdentity& identity);
void PruneStaleTextureAliases(MG_State::GLState::ITextureObject* texture);
SizeT PruneDeadTextures();
VkDevice m_device = VK_NULL_HANDLE;
VkPhysicalDevice m_physicalDevice = VK_NULL_HANDLE;
VmaAllocator m_allocator = nullptr;
VkCommandPool m_commandPool = VK_NULL_HANDLE;
VkQueue m_graphicsQueue = VK_NULL_HANDLE;
Bool m_imageFormatListSupported = false;
Uint32 m_currentFrameIndex = 0;
Uint8 m_gcCounter = 0;
// Frame-boundary GC gate: counts BeginFrame calls, not draws, so texture churn
// through non-draw paths (FBO clears, readbacks) still reaches the prune.
Uint32 m_gcFrameCounter = 0;
// Active only between BeginDrawSyncScope/EndDrawSyncScope; identities of
// textures already fully synced in the current draw (small N -> flat scan).
Bool m_drawSyncScopeActive = false;
@@ -467,45 +385,12 @@ private:
TextureResource* resource = nullptr;
};
Vector<DrawSyncedTexture> m_drawSyncedThisDraw;
// Cross-draw sampled-texture memo: the same few textures (atlas, lightmap)
// are resolved on every draw, so cache their resource pointers and skip the
// alive/resource map lookups. Node-based std::unordered_map keeps the
// pointees stable across inserts; erases bump m_resourceEraseEpoch, which
// every memo entry must match. SyncTexture still runs on memo hits, so
// content/param freshness is unaffected. A dead-then-reused texture address
// cannot false-hit: the new object carries a new lifetime id.
struct SyncedTextureMemoEntry {
const MG_State::GLState::ITextureObject* texture = nullptr;
Uint64 lifetimeId = 0;
Uint64 eraseEpoch = 0;
TextureResource* resource = nullptr;
};
static constexpr Uint32 kSyncedTextureMemoSize = 8;
SyncedTextureMemoEntry m_syncedTextureMemo[kSyncedTextureMemoSize];
Uint32 m_syncedTextureMemoNext = 0;
Uint64 m_resourceEraseEpoch = 1;
// Formats whose mutable-image probe failed on this device; their images are created
// without MUTABLE_FORMAT_BIT so repeat syncs neither re-probe nor flag-mismatch.
std::unordered_set<VkFormat> m_mutableFormatUnsupported;
std::unordered_map<TextureIdentity, WeakPtr<MG_State::GLState::ITextureObject>, TextureIdentityHash> m_aliveObjects;
std::unordered_map<TextureIdentity, TextureResource, TextureIdentityHash> m_textureResources;
// Textures that have been bound to a GL image unit (see MarkStorageImageTexture).
std::unordered_set<TextureIdentity, TextureIdentityHash> m_storageImageTextures;
// Supported multisample counts per format, so repeat texture syncs do not
// re-query vkGetPhysicalDeviceImageFormatProperties.
std::unordered_map<VkFormat, VkSampleCountFlags> m_multisampleCountsByFormat;
Vector<Vector<TextureResource>> m_deferredReleases;
Vector<Vector<VkImageView>> m_deferredViewReleases;
// Texture uploads are submitted out-of-band but NOT waited on (waiting
// behind the queue serialized the CPU against the previous frame's GPU
// work every time an animated atlas re-uploaded). Their transient objects
// are parked here and reclaimed once the upload fence signals.
struct PendingUploadReclaim {
VkFence fence = VK_NULL_HANDLE;
VkCommandBuffer commandBuffer = VK_NULL_HANDLE;
VkBuffer stagingBuffer = VK_NULL_HANDLE;
VmaAllocation stagingAllocation = nullptr;
};
Vector<PendingUploadReclaim> m_pendingUploadReclaims;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
File diff suppressed because it is too large Load Diff
@@ -76,10 +76,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
GLenum indexType = GL_UNSIGNED_SHORT;
SizeT indexByteOffset = 0;
SizeT indexByteSize = 0;
// Interpret indexByteOffset as a raw client pointer even when an element
// array buffer is bound (backend-synthesized index lists, e.g. the
// GL_LINE_LOOP -> LINE_STRIP rewrite).
Bool forceClientMemory = false;
};
struct DrawIndexedCmd {
@@ -118,10 +114,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
};
class VulkanRenderer : public IBufferCopyCommandProvider,
public FrameContext::IRecordingObserver,
public VkRenderPassManager::IEvictionObserver,
public ProgramFactory::IEvictionObserver {
class VulkanRenderer : public IBufferCopyCommandProvider, public FrameContext::IRecordingObserver {
public:
VulkanRenderer(NativeWindowType window, const VulkanRendererConfig& cfg = {});
~VulkanRenderer();
@@ -138,31 +131,9 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// recording, before any render pass.
void OnFrameCommandRecordingBegan(VkCommandBuffer commandBuffer) override;
// VkRenderPassManager::IEvictionObserver: the render-pass aging sweep just
// destroyed these VkRenderPasses; evict every graphics pipeline hashed on a
// dying handle (they share its >1024-boundary idleness, so immediate
// destruction is safe) and drop the last-pipeline memo if any went.
void OnRenderPassesDestroyed(const Vector<VkRenderPass>& renderPasses) override;
// ProgramFactory::IEvictionObserver: an aged-out program entry was
// destroyed; evict its compute pipeline and graphics pipelines (same
// idleness guarantee - they are only bound through draws/dispatches that
// stamp the program entry) and purge the descriptor-set cache entries
// keyed by its now-recyclable VkDescriptorSetLayout handle.
void OnProgramEvicted(ProgramFactory::HashType programHash,
VkDescriptorSetLayout descriptorSetLayout) override;
Bool SetupDraw(FrameContext::FrameData& frame, GLenum mode, Flags<DrawSetupAspect> aspects,
const DrawCmdParam& drawParams,
const IndexBufferView* pIndexBufferView = nullptr);
// ANGLE-style consecutive-draw fast path: SetupDraw snapshots the fully
// resolved draw configuration; the next draw whose cheap version/identity
// checks all match skips the resolution half (LOD probe, sampled-set
// walk, render-pass and pipeline resolution) and jumps straight to the
// per-draw tail. Returns false (leaving no side effects that the full
// path cannot redo idempotently) whenever anything might have changed.
Bool TrySetupDrawFastPath(FrameContext::FrameData& frame, GLenum mode, Flags<DrawSetupAspect> aspects,
const DrawCmdParam& drawParams, const IndexBufferView* pIndexBufferView);
void ClearAttachmentsOnActiveRenderPass(VkCommandBuffer commandBuffer,
const RenderPassEntry& compatibleRenderPassEntry);
@@ -200,25 +171,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
GLsizei srcWidth, GLsizei srcHeight, GLsizei srcDepth);
void GenerateMipmap(GLenum target);
void ReadPixels(GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, void* pixels);
// GL_DEPTH_COMPONENT / GL_DEPTH_STENCIL / GL_STENCIL_INDEX readback from the
// read framebuffer's depth/stencil attachment (per-aspect buffer copies with
// CPU repacking into the requested client layout).
void ReadDepthStencilPixels(MG_State::GLState::FramebufferObject& readFbo, GLint x, GLint y, GLsizei width,
GLsizei height, GLenum format, GLenum type, void* pixels);
// Copy-and-repack core shared by depth-stencil ReadPixels and GetTexImage;
// expects command recording to be active and any render pass already ended.
void ReadDepthStencilImageToClient(VkImage image, VkFormat vkFormat, VkImageLayout* trackedLayout,
VkImageAspectFlags imageAspect, Uint32 mipLevel, Uint32 baseArrayLayer,
GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type,
void* pixels);
// Same-extent depth blit between images of different depth formats: host
// round-trip with a per-texel re-encode (see BlitNamedFramebuffer).
Bool BlitDepthAcrossFormats(FrameContext::FrameData& frame, VkImage srcImage, VkFormat srcFormat,
VkImageLayout* srcTrackedLayout, Uint32 srcMipLevel, Uint32 srcBaseArrayLayer,
VkImage dstImage, VkFormat dstFormat, VkImageLayout* dstTrackedLayout,
Uint32 dstMipLevel, Uint32 dstBaseArrayLayer, GLint srcX, GLint srcY, GLint dstX,
GLint dstY, GLint width, GLint height, VkImageLayout srcRestoreLayout,
VkImageLayout dstRestoreLayout, Bool stencilAspect);
static SizeT GetReadbackTexelSize(VkFormat sourceFormat);
static Bool ConvertReadbackPixels(const Uint8* sourcePixels, VkFormat sourceFormat,
GLsizei width, GLsizei height, GLenum destinationFormat,
@@ -304,20 +256,7 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const VkTimerQueryManager::TimestampRecord& end) const;
Uint64 GetTimerQueryTimestampNs(const VkTimerQueryManager::TimestampRecord& record) const;
// GL_SAMPLES_PASSED occlusion queries: every app draw between Start and Stop is
// wrapped in a Vulkan occlusion query slot; the result is the slot sum. Requires
// hostQueryReset for slot recycling - Start fails (frontend keeps the query
// unsupported) when the device lacks it.
Bool StartOcclusionQueryCapture();
void StopOcclusionQueryCapture(Vector<Uint32>& outSlots);
// Flushes pending commands, waits, sums the slots, and recycles them.
Bool ResolveOcclusionQueryResult(const Vector<Uint32>& slots, Uint64& outSamples);
void RequestSwapchainResize(Uint32 width, Uint32 height);
// Re-query the surface and report whether the live swapchain no longer matches it
// (size or orientation). This - not a VK_SUBOPTIMAL_KHR result - is what decides a
// rebuild, so a surface the driver merely considers suboptimal cannot thrash.
Bool SwapchainIsOutOfDate();
// Returns false when the surface is zero-area (minimized/hidden window):
// no new swapchain is installed and presentation must stay suspended.
Bool RecreateSwapchain();
@@ -406,40 +345,16 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkFence AcquirePooledSubmitFence();
void DestroySubmitFencePool();
Bool HasPendingRecordedWork() const;
// Frame-boundary housekeeping for paths that never reach Present's
// tail (present-less readback loops, suspended presentation, blocking
// sync waits): runs the same per-frame drains Present performs, but
// only when every queue submission has been observed complete AND no
// recorded-but-unsubmitted commands exist - i.e. when CPU-GPU overlap
// is provably already zero. Never blocks (non-blocking fence poll
// only), so the presenting path's frames-in-flight pipelining is
// untouched. Returns true when the drain ran.
Bool TryDrainFrameTransients();
Vector<SubmitRecord> m_inFlightSubmits;
Vector<VkFence> m_freeSubmitFences;
Uint64 m_submitCounter = 0;
Uint64 m_completedSubmitCounter = 0;
// Drains since the last Present, gating the drain's frame-boundary-equivalent
// work (arena rewind + cache aging): a presenting app's mid-frame
// readbacks/waits must neither churn the transient caches nor accelerate the
// aging clocks, while present-less loops still cross a boundary every few
// iterations. Reset in Present.
Uint32 m_drainsSinceLastPresent = 0;
NativeWindowType m_window = 0;
void* m_platformDisplay = nullptr;
void* m_platformLibrary = nullptr;
void* m_platformCloseDisplay = nullptr;
// Some real ICDs (e.g. NVIDIA's proprietary Linux driver) don't implement
// VK_EXT_headless_surface at all. Detected once in CreateInstance() from the
// enumerated instance extensions; when false, CreateSurface() falls back to a
// hidden Xlib window instead of vkCreateHeadlessSurfaceEXT.
Bool m_headlessSurfaceSupported = true;
// Set when CreateSurface() had to create its own Xlib window for the fallback
// above (rather than being handed one by the caller), so Shutdown() knows it
// owns that window and must destroy it.
Bool m_ownsFallbackXlibWindow = false;
VulkanRendererConfig m_config;
Bool m_swapchainResizeRequested = false;
// Presentation is suspended while the window is zero-area (minimized): the
@@ -452,9 +367,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Vector<VkExtensionProperties> m_extensions;
VkInstance m_instance = VK_NULL_HANDLE;
VkDebugUtilsMessengerEXT m_debugMessenger = VK_NULL_HANDLE;
// Fallback reporting channel for drivers that ship the validation layers but
// only expose the older VK_EXT_debug_report (Adreno 650 / Vulkan 1.1.128).
VkDebugReportCallbackEXT m_debugReportCallback = VK_NULL_HANDLE;
PhysicalDevice m_physicalDevice;
VkDevice m_device = VK_NULL_HANDLE;
VmaAllocator m_allocator = nullptr;
@@ -492,74 +404,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
Uint32 stride);
static inline PFNDrawIndexedIndirectCountFunc s_vkCmdDrawIndexedIndirectCount = nullptr;
// VK_EXT_transform_feedback (GL transform feedback capture)
Bool m_transformFeedbackFeatureEnabled = false;
// VK_EXT_vertex_attribute_divisor: without it every non-zero glVertexAttribDivisor
// behaves as 1, because that is all Vulkan's instance input rate can express.
Bool m_vertexAttributeDivisorEnabled = false;
static inline PFN_vkCmdBindTransformFeedbackBuffersEXT s_vkCmdBindTransformFeedbackBuffersEXT = nullptr;
static inline PFN_vkCmdBeginTransformFeedbackEXT s_vkCmdBeginTransformFeedbackEXT = nullptr;
static inline PFN_vkCmdEndTransformFeedbackEXT s_vkCmdEndTransformFeedbackEXT = nullptr;
// Counter buffers (one 4-byte slot per capture binding) let consecutive
// draws within one glBeginTransformFeedback append GL-style. Transform feedback
// objects can each hold an open, paused span at the same time, so the counters are
// per object: one group of four slots each, handed out on first use.
static constexpr SizeT kXfbCounterObjectSlots = 16;
VkBufferObject m_xfbCounterBuffer;
UnorderedMap<Uint, Uint32> m_xfbCounterSlotByObject;
Uint32 m_xfbNextCounterSlot = 0;
// Set for a slot once a captured draw has been recorded into its span; selects
// counter-buffer resume on the next captured draw of the same span.
Array<Bool, kXfbCounterObjectSlots> m_xfbCountersValid{};
Array<Uint64, kXfbCounterObjectSlots> m_xfbLastSeenGeneration{};
// Counter slot group of the bound transform feedback object.
Uint32 CurrentXfbCounterSlot();
// Wraps a recorded draw with BeginTransformFeedbackEXT/EndTransformFeedbackEXT
// when GL transform feedback is active; binds capture buffers on demand.
Bool BeginXfbCaptureForDraw(FrameContext::FrameData& frame);
void EndXfbCaptureForDraw(FrameContext::FrameData& frame, Bool began);
// Makes the captured bytes visible to whatever reads them next. Deferred rather than
// recorded next to the capture, because the capturing draw runs inside a render pass
// that declares no self-dependency.
void MakeXfbWritesVisible();
Bool m_xfbWritesPendingVisibility = false;
// Wrap one app draw in an occlusion-query slot while a GL_SAMPLES_PASSED
// query is active. Returns whether a slot was begun (End must mirror it).
Bool BeginOcclusionForDraw(VkCommandBuffer commandBuffer);
void EndOcclusionForDraw(VkCommandBuffer commandBuffer, Bool began);
Bool m_occlusionQueryPreciseEnabled = false;
Bool m_hostQueryResetEnabled = false;
PFN_vkResetQueryPool s_vkResetQueryPool = nullptr;
VkQueryPool m_occlusionQueryPool = VK_NULL_HANDLE;
static constexpr Uint32 kOcclusionQuerySlots = 8192;
Uint32 m_occlusionSlotCursor = 0;
Bool m_occlusionCaptureActive = false;
Vector<Uint32> m_occlusionActiveSlots;
// Transform feedback primitive queries: one pool slot per captured draw yields
// the (written, needed) pair; GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN sums the
// first, GL_PRIMITIVES_GENERATED the second - exact with geometry shaders,
// unlike the CPU fallback accounting.
Bool m_xfbQueriesSupported = false;
PFN_vkCmdBeginQueryIndexedEXT s_vkCmdBeginQueryIndexedEXT = nullptr;
PFN_vkCmdEndQueryIndexedEXT s_vkCmdEndQueryIndexedEXT = nullptr;
VkQueryPool m_xfbQueryPool = VK_NULL_HANDLE;
static constexpr Uint32 kXfbQuerySlots = 8192;
Uint32 m_xfbQuerySlotCursor = 0;
Bool m_xfbQueryCaptureActive[2] = {false, false}; // [0]=written, [1]=generated
Vector<Uint32> m_xfbQueryActiveSlots[2];
Bool m_xfbQuerySlotOpen = false;
Uint32 m_xfbQueryOpenSlot = 0;
public:
// kind: 0 = PRIMITIVES_WRITTEN, 1 = PRIMITIVES_GENERATED.
Bool StartXfbQueryCapture(Uint32 kind);
void StopXfbQueryCapture(Uint32 kind, Vector<Uint32>& outSlots);
Bool ResolveXfbQueryResult(const Vector<Uint32>& slots, Bool wantGenerated, Uint64& outPrimitives);
private:
void BeginXfbQueryForDraw(VkCommandBuffer commandBuffer);
void EndXfbQueryForDraw(VkCommandBuffer commandBuffer);
VkCommandPool m_commandPool = VK_NULL_HANDLE;
VkBufferManager m_bufferManager;
@@ -572,30 +416,14 @@ namespace MobileGL::MG_Backend::DirectVulkan {
// gather + synthetic vertex-input rebuild + payload hash + lookup) when the full pipeline
// state is unchanged from the previous draw. The key provably covers every pipeline field.
// Reset per-frame and on pipeline destruction so the cached handle can never dangle.
// Small N-way pipeline-resolution memo (round-robin replacement). A
// single-entry memo thrashed on draw sequences that alternate a few
// pipelines (GUI text/quad program ping-pong), paying the full
// payload-hash lookup per draw; eight entries cover such working sets
// while keeping the hit path a trivial linear scan.
struct PipelineMemoEntry {
GLenum mode = 0;
Uint64 programHash = 0;
Uint64 vertexInputHash = 0;
Uint64 renderPassHash = 0;
Uint renderStateVersion = 0;
ProgramFactory::CompileOptionFlags transformFlags = {};
VkPipeline pipeline = VK_NULL_HANDLE;
};
static constexpr Uint32 kPipelineMemoSize = 8;
PipelineMemoEntry m_pipelineMemo[kPipelineMemoSize];
Uint32 m_pipelineMemoCount = 0;
Uint32 m_pipelineMemoNext = 0;
// Drops every memoized pipeline handle. Required at command-buffer
// boundaries and whenever any pipeline may have been destroyed.
void InvalidatePipelineMemo() {
m_pipelineMemoCount = 0;
m_pipelineMemoNext = 0;
}
Bool m_lastPipelineValid = false;
GLenum m_lastPipelineMode = 0;
Uint64 m_lastPipelineProgramHash = 0;
Uint64 m_lastPipelineVertexInputHash = 0;
Uint64 m_lastPipelineRenderPassHash = 0;
Uint m_lastPipelineRenderStateVersion = 0;
ProgramFactory::CompileOptionFlags m_lastPipelineTransformFlags = {};
VkPipeline m_lastPipelineResult = VK_NULL_HANDLE;
UnorderedMap<ProgramFactory::HashType, VkPipeline> m_computePipelines;
UniquePtr<ProgramFactory> m_programFactory;
UniquePtr<UniformManager> m_uniformManager;
@@ -624,61 +452,9 @@ namespace MobileGL::MG_Backend::DirectVulkan {
ProgramFactory::CompileOptionFlags m_lastSampledSetTransformFlags = {};
Uint64 m_lastSampledSetBindGeneration = 0;
// Memo for the per-draw explicit-LOD-0 eligibility probe
// (ProgramSamplesOnlySingleLevelTextures): same key family as the
// sampled-set memo, plus the sampled textures' params-version sum so a
// level-range or filter change re-probes. On a hit the resolved
// transform flags are reused, which also collapses the two
// GetOrCreateProgram lookups into one.
Bool m_lastLodDecisionValid = false;
Uint64 m_lastLodProgramLifetimeId = 0;
Uint32 m_lastLodProgramVersion = 0;
Uint64 m_lastLodBindGeneration = 0;
Uint64 m_lastLodParamsSum = 0;
ProgramFactory::CompileOptionFlags m_lastLodBaseFlags = {};
ProgramFactory::CompileOptionFlags m_lastLodResultFlags = {};
// Snapshot behind TrySetupDrawFastPath. Values only: the program and
// render-pass caches are open-addressing maps whose entries move on
// insert, so no pointers into them are cached; the pipeline handle is
// protected by the command-buffer-boundary reset plus the mid-frame
// pipeline-destruction resets, and monotonic epochs guard everything
// that can be destroyed or recreated between draws.
struct SetupDrawSnapshot {
Bool valid = false;
Uint8 aspects = 0;
GLenum mode = 0;
Uint64 programLifetimeId = 0;
Uint32 programVersion = 0;
const void* vao = nullptr;
Uint32 vaoConfigVersion = 0;
const void* drawFbo = nullptr;
Uint16 fboVersion = 0;
Bool drawFboIsDefault = false;
Uint renderStateVersion = 0;
Uint64 bindGeneration = 0;
Uint32 baseTransformFlags = 0;
Uint32 resolvedTransformFlags = 0;
Uint64 renderPassHash = 0;
Uint32 imageIndex = 0;
Uint64 textureEraseEpoch = 0;
Uint64 textureImageEpoch = 0;
Uint64 renderbufferImageEpoch = 0;
Uint64 sampledContentSum = 0;
Uint64 sampledParamsSum = 0;
IntVec2 renderPassExtent = {0, 0};
VkPipeline pipeline = VK_NULL_HANDLE;
};
SetupDrawSnapshot m_setupDrawSnapshot;
// Per-draw scratch buffers (clear keeps capacity) — these paths run for every
// draw call and must not allocate.
Vector<MG_State::GLState::ITextureObject*> m_sampledTexturesScratch;
// Parallel to m_sampledTexturesScratch, refilled by every SetupDraw's
// first sampled-texture loop: the resolved backend resources, so the
// post-transition loop can skip re-resolving textures whose layout is
// already sampleable.
Vector<VkTextureManager::TextureResource*> m_sampledResourcesScratch;
Vector<MG_State::GLState::ITextureObject*> m_storageImageTexturesScratch;
Vector<VkBuffer> m_vertexBuffersScratch;
Vector<VkDeviceSize> m_vertexOffsetsScratch;
@@ -741,8 +517,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
void CreateInstance();
VkResult SetupDebugMessenger();
VkResult DestroyDebugMessenger();
VkResult SetupDebugReportCallback();
void DestroyDebugReportCallback();
VkDebugUtilsMessengerCreateInfoEXT PopulateDebugMessengerCreateInfo();
void CreateSurface();
void PickPhysicalDevice();
@@ -761,10 +535,8 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const RenderPassEntry& renderPassEntry);
VkPipeline GetOrCreateComputePipeline(const ProgramFactory::VkProgramObject& programObj);
void DestroyComputePipelines();
// Takes the frame rather than a command buffer: a first-time storage-usage upgrade has to
// flush the pending recording (see the body), which retires the current command buffer.
Bool PrepareStorageImageTextures(
FrameContext::FrameData& frame,
VkCommandBuffer commandBuffer,
const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj);
@@ -803,14 +575,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
VkImageLayout finalLayout);
Bool SubmitReadbackCommandsAndWait(FrameContext::FrameData& frame);
public:
// Submits whatever is recorded and waits for it. The CPU is about to read memory
// a shader wrote (a mapped shader storage buffer), and coherent host-visible
// storage only guarantees visibility once the work that produced it has retired.
Bool FinishPendingGpuWork();
private:
void ShutdownSwapchain();
// Static functions
@@ -834,10 +598,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
const PhysicalDevice& compareWithDevice,
PhysicalDevice& outBetterDevice);
static constexpr const char* s_validationLayerNames[] = {"VK_LAYER_KHRONOS_validation"};
// VK_KHR_image_format_list: lets MUTABLE_FORMAT images declare their exact view-format
// set so the driver can keep bandwidth compression (see CreateLogicalDeviceAndQueues).
Bool m_imageFormatListExtensionEnabled = false;
static constexpr const char* s_deviceExtensionNames[] = {VK_KHR_SWAPCHAIN_EXTENSION_NAME};
static Bool CheckValidationLayerSupport();
+3 -32
View File
@@ -52,48 +52,19 @@ namespace MobileGL::MG_Backend::DirectVulkan {
}
} // namespace MobileGL::MG_Backend::DirectVulkan
namespace MobileGL::MG_Backend::DirectVulkan {
// GL renders into sRGB color attachments RAW while GL_FRAMEBUFFER_SRGB is disabled
// (the core-profile default); Vulkan sRGB attachments always encode on write. The
// attachment view (and render pass format) therefore drops to the UNORM twin
// whenever the capability is off. Sampled views keep the sRGB format (decode on
// sample is unconditional in GL).
inline VkFormat ResolveSrgbAttachmentWriteFormat(VkFormat format, bool framebufferSrgbEnabled) {
if (framebufferSrgbEnabled) return format;
switch (format) {
case VK_FORMAT_R8G8B8A8_SRGB:
return VK_FORMAT_R8G8B8A8_UNORM;
case VK_FORMAT_B8G8R8A8_SRGB:
return VK_FORMAT_B8G8R8A8_UNORM;
default:
return format;
}
}
} // namespace MobileGL::MG_Backend::DirectVulkan
// The context line (__VA_ARGS__ = its own format string + args) must be a SEPARATE log
// call: appending its format to the base format while its arguments precede the base
// arguments makes every conversion read the wrong slot (a %s pulling an int crashes).
#define VK_VERIFY(expr, ...) \
do { \
VkResult _vk_verify_result = (expr); \
if (_vk_verify_result != VK_SUCCESS) { \
__VA_OPT__(MGLOG_F(__VA_ARGS__);) \
MGLOG_F("Vulkan error %s (%d) at %s:%d", \
MGLOG_F("Vulkan error %s (%d) at %s:%d" __VA_OPT__(" - ") __VA_ARGS__, \
MobileGL::MG_Backend::DirectVulkan::VkResultToString(_vk_verify_result), \
_vk_verify_result, __FILE__, __LINE__); \
} \
MOBILEGL_ASSERT(_vk_verify_result == VK_SUCCESS, "Vulkan error %s (%d) at %s:%d", \
MobileGL::MG_Backend::DirectVulkan::VkResultToString(_vk_verify_result), \
_vk_verify_result, __FILE__, __LINE__); \
MOBILEGL_ASSERT(_vk_verify_result == VK_SUCCESS, "Vulkan error %s (%d) at %s:%d" __VA_OPT__(" - ") __VA_ARGS__, MobileGL::MG_Backend::DirectVulkan::VkResultToString(_vk_verify_result), _vk_verify_result, __FILE__, __LINE__); \
} while (0)
#define XXHASH_VERIFY(expr, ...) \
do { \
XXH_errorcode _xxh_verify_result = (expr); \
if (_xxh_verify_result != XXH_OK) { \
__VA_OPT__(MGLOG_F(__VA_ARGS__);) \
} \
MOBILEGL_ASSERT(_xxh_verify_result == XXH_OK, "XXHash error %d at %s:%d", _xxh_verify_result, __FILE__, \
__LINE__); \
MOBILEGL_ASSERT(_xxh_verify_result == XXH_OK, "XXHash error %d at %s:%d" __VA_OPT__(" - ") __VA_ARGS__, _xxh_verify_result, __FILE__, __LINE__); \
} while (0)
-33
View File
@@ -24,7 +24,6 @@ namespace MobileGL::MG_Impl::CGLImpl {
GLint Samples = 0;
GLint Profile = kCGLOGLPVersion_3_2_Core;
GLint RendererId = 0x4d474c;
GLint DisplayMask = 0;
};
struct ContextObject {
@@ -135,9 +134,6 @@ namespace MobileGL::MG_Impl::CGLImpl {
case kCGLPFARendererID:
pixelFormat.RendererId = value;
break;
case kCGLPFADisplayMask:
pixelFormat.DisplayMask = value;
break;
default:
break;
}
@@ -347,9 +343,6 @@ namespace MobileGL::MG_Impl::CGLImpl {
case kCGLPFARendererID:
*value = pixelFormat->RendererId;
return kCGLNoError;
case kCGLPFADisplayMask:
*value = pixelFormat->DisplayMask;
return kCGLNoError;
case kCGLPFAOpenGLProfile:
*value = pixelFormat->Profile;
return kCGLNoError;
@@ -488,32 +481,6 @@ namespace MobileGL::MG_Impl::CGLImpl {
return it == currentContexts.end() ? nullptr : it->second;
}
CGLError SetVirtualScreen(CGLContextObj ctx, GLint screen) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (!object) {
return kCGLBadContext;
}
if (screen != 0) {
return kCGLBadValue;
}
object->VirtualScreen = screen;
return kCGLNoError;
}
CGLError GetVirtualScreen(CGLContextObj ctx, GLint* screen) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (!object) {
return kCGLBadContext;
}
if (!screen) {
return kCGLBadAddress;
}
*screen = object->VirtualScreen;
return kCGLNoError;
}
CGLError SetParameter(CGLContextObj ctx, CGLContextParameter pname, const GLint* params) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
-2
View File
@@ -32,8 +32,6 @@ namespace MobileGL::MG_Impl::CGLImpl {
CGLError SetCurrentContext(CGLContextObj ctx);
CGLContextObj GetCurrentContext();
CGLError SetVirtualScreen(CGLContextObj ctx, GLint screen);
CGLError GetVirtualScreen(CGLContextObj ctx, GLint* screen);
CGLError SetParameter(CGLContextObj ctx, CGLContextParameter pname, const GLint* params);
CGLError GetParameter(CGLContextObj ctx, CGLContextParameter pname, GLint* params);
CGLError UpdateContext(CGLContextObj ctx);
@@ -71,14 +71,6 @@ MOBILEGL_CGL_API CGLContextObj CGLGetCurrentContext(void) {
return MobileGL::MG_Impl::CGLImpl::GetCurrentContext();
}
MOBILEGL_CGL_API CGLError CGLSetVirtualScreen(CGLContextObj ctx, GLint screen) {
return MobileGL::MG_Impl::CGLImpl::SetVirtualScreen(ctx, screen);
}
MOBILEGL_CGL_API CGLError CGLGetVirtualScreen(CGLContextObj ctx, GLint* screen) {
return MobileGL::MG_Impl::CGLImpl::GetVirtualScreen(ctx, screen);
}
MOBILEGL_CGL_API CGLError CGLSetParameter(CGLContextObj ctx, CGLContextParameter pname, const GLint* params) {
return MobileGL::MG_Impl::CGLImpl::SetParameter(ctx, pname, params);
}
@@ -10,12 +10,8 @@
#if defined(__APPLE__)
#include "MG_Impl/CGLImpl/CGLImpl.h"
#include "MG_Impl/GetProcAddress.h"
#include <CoreGraphics/CoreGraphics.h>
#include <CoreVideo/CVDisplayLink.h>
#include <cstdint>
#include <dlfcn.h>
namespace {
@@ -51,52 +47,10 @@ namespace {
return dlsym(handle, symbol);
}
CGDirectDisplayID DisplayForMask(GLint displayMask) {
constexpr std::uint32_t MaxDisplays = sizeof(CGOpenGLDisplayMask) * 8;
CGDirectDisplayID displays[MaxDisplays] = {};
std::uint32_t displayCount = 0;
if (displayMask != 0 &&
CGGetActiveDisplayList(MaxDisplays, displays, &displayCount) == kCGErrorSuccess) {
const auto mask = static_cast<CGOpenGLDisplayMask>(displayMask);
for (std::uint32_t i = 0; i < displayCount; ++i) {
if ((CGDisplayIDToOpenGLDisplayMask(displays[i]) & mask) != 0) {
return displays[i];
}
}
}
return CGMainDisplayID();
}
#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wdeprecated-declarations"
CVReturn MobileGLCVDisplayLinkSetCurrentCGDisplayFromOpenGLContext(
CVDisplayLinkRef displayLink,
CGLContextObj context,
CGLPixelFormatObj pixelFormat) {
GLint virtualScreen = 0;
if (MobileGL::MG_Impl::CGLImpl::GetVirtualScreen(context, &virtualScreen) == kCGLNoError) {
GLint displayMask = 0;
if (!displayLink ||
MobileGL::MG_Impl::CGLImpl::DescribePixelFormat(
pixelFormat, virtualScreen, kCGLPFADisplayMask, &displayMask) != kCGLNoError) {
return kCVReturnInvalidArgument;
}
return CVDisplayLinkSetCurrentCGDisplay(displayLink, DisplayForMask(displayMask));
}
using OriginalFunction = CVReturn (*)(CVDisplayLinkRef, CGLContextObj, CGLPixelFormatObj);
static const auto original = reinterpret_cast<OriginalFunction>(
dlsym(RTLD_NEXT, "CVDisplayLinkSetCurrentCGDisplayFromOpenGLContext"));
return original ? original(displayLink, context, pixelFormat) : kCVReturnError;
}
__attribute__((used)) static const DyldInterposeEntry kMobileGLDyldInterpose[]
__attribute__((section("__DATA,__interpose"))) = {
{reinterpret_cast<const void*>(MobileGLDlsym), reinterpret_cast<const void*>(dlsym)},
{reinterpret_cast<const void*>(MobileGLCVDisplayLinkSetCurrentCGDisplayFromOpenGLContext),
reinterpret_cast<const void*>(CVDisplayLinkSetCurrentCGDisplayFromOpenGLContext)},
};
#pragma clang diagnostic pop
} // namespace
#endif
@@ -1,10 +0,0 @@
# Public CGL entry points.
_CGL*
# Public EGL entry points.
_egl*
# Public OpenGL and GLX entry points. OpenGL function names always use an
# uppercase letter or digit after the "gl" prefix; excluding lowercase here
# deliberately prevents glslang_* from matching this pattern.
_gl[A-Z0-9]*
+2 -38
View File
@@ -330,21 +330,12 @@ namespace MobileGL::MG_Impl::GLImpl {
} else if (access & BufferMappingAccessBit::Write) {
*params = GL_WRITE_ONLY;
} else {
*params = GL_READ_WRITE;
*params = 0;
}
} else {
// Initial value, and what glUnmapBuffer restores (GL 4.6 core table 6.2).
*params = GL_READ_WRITE;
*params = 0;
}
break;
case GL_BUFFER_ACCESS_FLAGS:
// The MapBufferRange flags verbatim; glMapBuffer's access enum has already been
// normalised into the same bits. Zero while the buffer is not mapped.
*params = bufferObject->IsMapped()
? static_cast<GLint>(
MG_Util::ConvertBufferMappingAccessToGLEnum(bufferObject->GetMappingAccess()))
: 0;
break;
case GL_BUFFER_MAPPED:
*params = bufferObject->IsMapped() ? GL_TRUE : GL_FALSE;
break;
@@ -887,7 +878,6 @@ namespace MobileGL::MG_Impl::GLImpl {
return;
}
bufferObject->SyncGpuWrites();
bufferObject->DownloadSubData(data, static_cast<SizeT>(offset), static_cast<SizeT>(size));
}
@@ -1361,14 +1351,6 @@ namespace MobileGL::MG_Impl::GLImpl {
BufferTarget bufferTarget = MG_Util::ConvertGLEnumToBufferTarget(target);
if (!BufferImpl::ValidateBufferBindingPointTarget(bufferTarget)) return;
if (!BufferImpl::ValidateBufferBindingPointIndex(bufferTarget, pointIndex)) return;
if (bufferTarget == BufferTarget::TransformFeedback && MG_State::pGLContext->IsTransformFeedbackActive()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"Transform feedback buffer bindings cannot change while transform "
"feedback is active."));
return;
}
MG_State::pGLContext->TouchBufferBindingPoint(bufferTarget, pointIndex);
auto& point = MG_State::pGLContext->GetBufferBindingPoint(bufferTarget, pointIndex);
@@ -1376,7 +1358,6 @@ namespace MobileGL::MG_Impl::GLImpl {
if (buffer == 0) {
point.Bind(nullptr);
point.SetRange(Range1D(0, 0));
GetBufferBindingSlot(bufferTarget).Bind(nullptr);
return;
}
@@ -1395,12 +1376,6 @@ namespace MobileGL::MG_Impl::GLImpl {
} else {
point.ClearRange();
}
// The indexed bind also binds to the generic binding point of the same target
// (GL 4.6 core 6.1.1). Callers rely on it: the texture_gather tests set up their
// SSBO with BindBufferBase and then size it through glBufferData on the generic
// target alone, which would otherwise raise GL_INVALID_OPERATION and leave the
// buffer with no storage.
GetBufferBindingSlot(bufferTarget).Bind(bufferObject);
}
void BindBufferRange_State(GLenum target, GLuint index, GLuint buffer, GLintptr offset, GLsizeiptr size) {
@@ -1409,14 +1384,6 @@ namespace MobileGL::MG_Impl::GLImpl {
BufferTarget bufferTarget = MG_Util::ConvertGLEnumToBufferTarget(target);
if (!BufferImpl::ValidateBufferBindingPointTarget(bufferTarget)) return;
if (!BufferImpl::ValidateBufferBindingPointIndex(bufferTarget, index)) return;
if (bufferTarget == BufferTarget::TransformFeedback && MG_State::pGLContext->IsTransformFeedbackActive()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"Transform feedback buffer bindings cannot change while transform "
"feedback is active."));
return;
}
MG_State::pGLContext->TouchBufferBindingPoint(bufferTarget, index);
auto& point = MG_State::pGLContext->GetBufferBindingPoint(bufferTarget, index);
@@ -1424,7 +1391,6 @@ namespace MobileGL::MG_Impl::GLImpl {
if (buffer == 0) {
point.Bind(nullptr);
point.SetRange(Range1D(0, 0));
GetBufferBindingSlot(bufferTarget).Bind(nullptr);
return;
}
@@ -1442,8 +1408,6 @@ namespace MobileGL::MG_Impl::GLImpl {
} else {
point.ClearRange();
}
// Also the generic binding point, exactly as BindBufferBase (GL 4.6 core 6.1.1).
GetBufferBindingSlot(bufferTarget).Bind(bufferObject);
}
/* @INSERTION_POINT:FUNCTION_IMPLEMENTATION@ */
@@ -60,11 +60,6 @@ namespace MobileGL::MG_Impl::GLImpl::BufferImpl {
MG_Backend::pActiveBackendObject->GetDynamicParameters().MaxShaderStorageBufferBindings;
pointCount = std::min(pointCount, static_cast<SizeT>(std::max(backendCount, 0)));
}
if (target == BufferTarget::TransformFeedback) {
// GL_MAX_TRANSFORM_FEEDBACK_SEPARATE_ATTRIBS bounds the indexed capture
// binding points in GL 3.3 (no ARB_transform_feedback3).
pointCount = std::min<SizeT>(pointCount, 4);
}
if (index < pointCount) {
return true;
+9 -734
View File
@@ -11,7 +11,6 @@
#include <MG_State/GLState/Core.h>
#include <MG_State/EGLState/Core.h>
#include <MG_Backend/BackendObjects.h>
#include "../Getter/GL_Getter.h"
namespace MobileGL::MG_Impl::GLImpl {
static Bool ValidateCurrentProgramForExecution(const char* functionName) {
@@ -49,109 +48,7 @@ namespace MobileGL::MG_Impl::GLImpl {
return true;
}
// Primitives a draw of `count` vertices in `mode` assembles (0 for
// incomplete primitives). Used for the CPU-side transform feedback
// primitive accounting.
static Uint64 CountPrimitivesForDraw(GLenum mode, GLsizei count) {
if (count <= 0) return 0;
switch (mode) {
case GL_POINTS: return static_cast<Uint64>(count);
case GL_LINES: return static_cast<Uint64>(count / 2);
case GL_LINE_STRIP: return count >= 2 ? static_cast<Uint64>(count - 1) : 0;
case GL_LINE_LOOP: return count >= 2 ? static_cast<Uint64>(count) : 0;
case GL_TRIANGLES: return static_cast<Uint64>(count / 3);
case GL_TRIANGLE_STRIP:
case GL_TRIANGLE_FAN: return count >= 3 ? static_cast<Uint64>(count - 2) : 0;
default: return 0;
}
}
// Accumulate the transform feedback primitive counter for a captured draw.
// Draws without a geometry stage write exactly the primitives they assemble,
// clamped by the capture buffers' remaining capacity (a full buffer stops
// recording whole primitives, which is what PRIMITIVES_WRITTEN reports).
// Geometry amplification is not modelled here.
static void AccountTransformFeedbackPrimitives(GLenum mode, GLsizei count) {
if (!MG_State::pGLContext->IsTransformFeedbackActive()) return;
// A paused span captures nothing, so a draw made while paused contributes to
// PRIMITIVES_GENERATED but not to TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN.
if (MG_State::pGLContext->IsTransformFeedbackPaused()) {
MG_State::pGLContext->AddTransformFeedbackPausedPrimitives(CountPrimitivesForDraw(mode, count));
return;
}
Uint64 primitives = CountPrimitivesForDraw(mode, count);
if (primitives == 0) return;
MG_State::pGLContext->AddTransformFeedbackInputPrimitives(primitives);
Uint64 verticesPerPrimitive = 1;
switch (mode) {
case GL_LINES:
case GL_LINE_STRIP:
case GL_LINE_LOOP:
verticesPerPrimitive = 2;
break;
case GL_TRIANGLES:
case GL_TRIANGLE_STRIP:
case GL_TRIANGLE_FAN:
verticesPerPrimitive = 3;
break;
default:
break;
}
const auto& program = MG_State::pGLContext->GetTransformFeedbackProgram();
if (program != nullptr) {
// Capacity in captured vertices = the tightest bound buffer.
Uint64 capacityVertices = ~0ull;
for (SizeT i = 0; i < program->GetTransformFeedbackBufferCount(); ++i) {
const Uint32 stride = program->GetTransformFeedbackStride(static_cast<Uint32>(i));
if (stride == 0) continue;
const auto& point = MG_State::pGLContext->GetBufferBindingPoint(BufferTarget::TransformFeedback,
static_cast<Uint>(i));
const Range1D range = point.GetRange();
const Uint64 bytes = range.end > range.start ? static_cast<Uint64>(range.end - range.start) : 0;
capacityVertices = std::min<Uint64>(capacityVertices, bytes / stride);
}
if (capacityVertices != ~0ull) {
const Uint64 usedVertices = MG_State::pGLContext->GetTransformFeedbackCapturedVertices();
const Uint64 remainingVertices = capacityVertices > usedVertices ? capacityVertices - usedVertices : 0;
primitives = std::min<Uint64>(primitives, remainingVertices / verticesPerPrimitive);
}
}
MG_State::pGLContext->AddTransformFeedbackPrimitives(primitives);
MG_State::pGLContext->AddTransformFeedbackCapturedVertices(primitives * verticesPerPrimitive);
}
// Every primitive mode a draw command accepts (GL 4.6 core table 10.1, plus
// GL_PATCHES for the tessellation pipeline). Anything else is GL_INVALID_ENUM.
static Bool IsAcceptedPrimitiveMode(GLenum mode) {
switch (mode) {
case GL_POINTS:
case GL_LINES:
case GL_LINE_LOOP:
case GL_LINE_STRIP:
case GL_LINES_ADJACENCY:
case GL_LINE_STRIP_ADJACENCY:
case GL_TRIANGLES:
case GL_TRIANGLE_STRIP:
case GL_TRIANGLE_FAN:
case GL_TRIANGLES_ADJACENCY:
case GL_TRIANGLE_STRIP_ADJACENCY:
case GL_PATCHES:
return true;
default:
return false;
}
}
static Bool ValidatePrimitiveModeForBackend(const char* functionName, GLenum mode) {
if (!IsAcceptedPrimitiveMode(mode)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName, "mode is not an accepted primitive type."));
return false;
}
const auto& activeBackendObject = MG_Backend::pActiveBackendObject;
if (!activeBackendObject) {
MG_State::pGLContext->RecordError(
@@ -160,6 +57,15 @@ namespace MobileGL::MG_Impl::GLImpl {
return false;
}
if (activeBackendObject->GetBackendType() == BackendType::DirectVulkan && mode == GL_LINE_LOOP) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", functionName,
"Primitive mode GL_LINE_LOOP is not supported by the DirectVulkan backend."));
return false;
}
const auto& vao = MG_State::pGLContext->GetBoundVertexArray();
if (vao && vao->GetExternalIndex() == 0 && !MG_State::IsRelaxedSemanticsActive()) {
MG_State::pGLContext->RecordError(
@@ -169,133 +75,9 @@ namespace MobileGL::MG_Impl::GLImpl {
return false;
}
// A geometry stage only accepts the primitive types that decompose into its declared
// input primitive (GL 4.6 core 11.3.1); anything else is INVALID_OPERATION. GL_PATCHES
// is the tessellation pipeline's input and reaches the geometry stage already
// converted, so it is not constrained here.
const auto& currentProgram = MG_State::pGLContext->GetCurrentProgram();
const GLenum gsInput = currentProgram ? currentProgram->GetGeometryInputType() : GL_NONE;
if (gsInput != GL_NONE && mode != GL_PATCHES) {
Bool compatible = false;
switch (gsInput) {
case GL_POINTS:
compatible = mode == GL_POINTS;
break;
case GL_LINES:
compatible = mode == GL_LINES || mode == GL_LINE_STRIP || mode == GL_LINE_LOOP;
break;
case GL_LINES_ADJACENCY:
compatible = mode == GL_LINES_ADJACENCY || mode == GL_LINE_STRIP_ADJACENCY;
break;
case GL_TRIANGLES:
compatible = mode == GL_TRIANGLES || mode == GL_TRIANGLE_STRIP || mode == GL_TRIANGLE_FAN;
break;
case GL_TRIANGLES_ADJACENCY:
compatible = mode == GL_TRIANGLES_ADJACENCY || mode == GL_TRIANGLE_STRIP_ADJACENCY;
break;
default:
break;
}
if (!compatible) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", functionName,
"Primitive mode is incompatible with the geometry shader's input primitive type."));
return false;
}
}
// While transform feedback is active the draw's primitive type must match
// the feedback primitive mode (GL 3.3 core 13.2.2). With a geometry shader
// the constraint moves to the shader's output primitive type instead, so
// the draw mode itself is unconstrained here. A paused span is exempt: it
// captures nothing, so there is nothing for the mode to be incompatible with
// (GL 4.6 core 13.2.3).
if (MG_State::pGLContext->IsTransformFeedbackActive() &&
!MG_State::pGLContext->IsTransformFeedbackPaused() &&
!(MG_State::pGLContext->GetTransformFeedbackProgram() &&
MG_State::pGLContext->GetTransformFeedbackProgram()->GetShaderIndexByStage(ShaderStage::Geometry) >= 0)) {
const GLenum feedbackMode = MG_State::pGLContext->GetTransformFeedbackPrimitiveMode();
Bool compatible = false;
switch (feedbackMode) {
case GL_POINTS:
compatible = mode == GL_POINTS;
break;
case GL_LINES:
compatible = mode == GL_LINES || mode == GL_LINE_STRIP || mode == GL_LINE_LOOP;
break;
case GL_TRIANGLES:
compatible = mode == GL_TRIANGLES || mode == GL_TRIANGLE_STRIP || mode == GL_TRIANGLE_FAN;
break;
default:
break;
}
if (!compatible) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", functionName,
"Primitive mode is incompatible with the active transform feedback primitive mode."));
return false;
}
}
return true;
}
// Byte size of the command structures the indirect draws read (GL 4.6 core 10.3.10).
constexpr SizeT kDrawArraysIndirectCommandBytes = 4 * sizeof(Uint32);
constexpr SizeT kDrawElementsIndirectCommandBytes = 5 * sizeof(Uint32);
// Shared preconditions of every *Indirect draw: `indirect` is a byte offset into the
// buffer bound to GL_DRAW_INDIRECT_BUFFER, must be 4-byte aligned, and the whole
// command has to lie inside that buffer.
static Bool ValidateIndirectDrawSource(const char* functionName, const void* indirect, SizeT commandBytes) {
const auto offset = reinterpret_cast<uintptr_t>(indirect);
if (offset % 4 != 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName,
"indirect offset must be a multiple of 4."));
return false;
}
const auto& buffer = MG_State::pGLContext->GetBufferBindingSlot(BufferTarget::DrawIndirect).GetBoundObject();
if (!buffer) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName,
"No buffer is bound to GL_DRAW_INDIRECT_BUFFER."));
return false;
}
if (offset + commandBytes > buffer->GetSize()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName,
"The indirect command extends past the end of the bound "
"GL_DRAW_INDIRECT_BUFFER."));
return false;
}
return true;
}
// Index type accepted by the DrawElements family (GL 4.6 core 10.3.9).
static Bool ValidateDrawElementsIndexType(const char* functionName, GLenum type) {
switch (type) {
case GL_UNSIGNED_BYTE:
case GL_UNSIGNED_SHORT:
case GL_UNSIGNED_INT:
return true;
default:
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName, "type is not an accepted index type."));
return false;
}
}
void Clear_Backend(GLbitfield mask) {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
@@ -490,28 +272,6 @@ namespace MobileGL::MG_Impl::GLImpl {
dispatchComputeIndirect(indirect);
}
void PatchParameteri(GLenum pname, GLint value) {
if (pname != GL_PATCH_VERTICES) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "pname must be GL_PATCH_VERTICES."));
return;
}
GLint maxPatchVertices = 32;
GetIntegerv(GL_MAX_PATCH_VERTICES, &maxPatchVertices);
if (value <= 0 || value > maxPatchVertices) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"value must be in [1, GL_MAX_PATCH_VERTICES]."));
return;
}
MG_State::pGLContext->SetPatchVertices(static_cast<Uint>(value));
if (const auto patchParameteri = MG_Backend::gBackendFunctionsTable.GL.PatchParameteri) {
patchParameteri(pname, value);
}
}
void MemoryBarrier(GLbitfield barriers) {
auto memoryBarrier = MG_Backend::gBackendFunctionsTable.GL.MemoryBarrier;
if (!memoryBarrier) {
@@ -617,8 +377,6 @@ namespace MobileGL::MG_Impl::GLImpl {
void DrawElementsIndirect(GLenum mode, GLenum type, const void* indirect) {
if (!ValidateCurrentProgramForExecution(__func__)) return;
if (!ValidatePrimitiveModeForBackend(__func__, mode)) return;
if (!ValidateDrawElementsIndexType(__func__, type)) return;
if (!ValidateIndirectDrawSource(__func__, indirect, kDrawElementsIndirectCommandBytes)) return;
DrawElementsIndirect_Backend(mode, type, indirect);
}
@@ -638,21 +396,18 @@ namespace MobileGL::MG_Impl::GLImpl {
void DrawArraysIndirect(GLenum mode, const void* indirect) {
if (!ValidateCurrentProgramForExecution(__func__)) return;
if (!ValidatePrimitiveModeForBackend(__func__, mode)) return;
if (!ValidateIndirectDrawSource(__func__, indirect, kDrawArraysIndirectCommandBytes)) return;
DrawArraysIndirect_Backend(mode, indirect);
}
void DrawElementsBaseVertex(GLenum mode, GLsizei count, GLenum type, const void* indices, GLint basevertex) {
if (!ValidateCurrentProgramForExecution(__func__)) return;
if (!ValidatePrimitiveModeForBackend(__func__, mode)) return;
AccountTransformFeedbackPrimitives(mode, count);
DrawElementsBaseVertex_Backend(mode, count, type, indices, basevertex);
}
void DrawArrays(GLenum mode, GLint first, GLsizei count) {
if (!ValidateCurrentProgramForExecution(__func__)) return;
if (!ValidatePrimitiveModeForBackend(__func__, mode)) return;
AccountTransformFeedbackPrimitives(mode, count);
DrawArrays_Backend(mode, first, count);
}
@@ -689,487 +444,7 @@ namespace MobileGL::MG_Impl::GLImpl {
void DrawElements(GLenum mode, GLsizei count, GLenum type, const void* indices) {
if (!ValidateCurrentProgramForExecution(__func__)) return;
if (!ValidatePrimitiveModeForBackend(__func__, mode)) return;
AccountTransformFeedbackPrimitives(mode, count);
DrawElements_Backend(mode, count, type, indices);
}
void BeginTransformFeedback(GLenum primitiveMode) {
if (primitiveMode != GL_POINTS && primitiveMode != GL_LINES && primitiveMode != GL_TRIANGLES) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"primitiveMode must be GL_POINTS, GL_LINES or GL_TRIANGLES."));
return;
}
if (MG_State::pGLContext->IsTransformFeedbackActive()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "Transform feedback is already active."));
return;
}
const auto& program = MG_State::pGLContext->GetCurrentProgram();
if (!program || !program->GetLinkStatus() || program->GetTransformFeedbackVaryingCount() == 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", __func__,
"No program with transform feedback varyings is active."));
return;
}
// Every capture buffer slot the program's mode uses must have a buffer bound. A slot
// of stride 0 - two consecutive gl_NextBuffer entries - captures nothing and so needs
// no binding.
const SizeT usedBufferCount = program->GetTransformFeedbackBufferCount();
for (SizeT i = 0; i < usedBufferCount; ++i) {
if (program->GetTransformFeedbackStride(static_cast<Uint32>(i)) == 0) continue;
const auto& point = MG_State::pGLContext->GetBufferBindingPoint(BufferTarget::TransformFeedback,
static_cast<Uint>(i));
if (point.GetBoundObject() == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", __func__,
"Transform feedback buffer binding point " + std::to_string(i) + " has no buffer bound."));
return;
}
}
MG_State::pGLContext->BeginTransformFeedback(primitiveMode, program);
if (const auto beginXfb = MG_Backend::gBackendFunctionsTable.GL.BeginTransformFeedback) {
beginXfb(primitiveMode);
}
}
// Vulkan transform feedback captures triangle strips in plain (i, i+1, i+2)
// vertex order, but GL decomposes odd strip triangles as (i+1, i, i+2)
// (GL 4.6 table 10.1). With the geometry stage's statically-known strip
// lengths the captured records are reordered in place: swap the first two
// vertex records of every odd triangle within each emitted strip.
static void FixupGsStripCaptureOrder(const SharedPtr<MG_State::GLState::ProgramObject>& program,
Uint64 inputPrimitives) {
// Only Vulkan-order captures need this. A backend that runs the capture on its
// own GL/ES driver (it owns the span, hence the EndTransformFeedback entry) has
// already produced GL's vertex order, and reordering it again would corrupt it.
if (MG_Backend::gBackendFunctionsTable.GL.EndTransformFeedback != nullptr) {
return;
}
if (program == nullptr || !program->HasGsTriangleStripCaptureFixup() || inputPrimitives == 0) {
return;
}
const auto& stripTriangles = program->GetGsStripTriangles();
// Global triangle indices whose leading vertex pair must swap.
Vector<Uint64> swapTriangles;
Uint64 triangleBase = 0;
for (Uint64 input = 0; input < inputPrimitives; ++input) {
for (const Uint32 stripLength : stripTriangles) {
for (Uint32 t = 1; t < stripLength; t += 2) {
swapTriangles.push_back(triangleBase + t);
}
triangleBase += stripLength;
}
}
if (swapTriangles.empty()) {
return;
}
for (SizeT bufferIndex = 0; bufferIndex < program->GetTransformFeedbackBufferCount(); ++bufferIndex) {
const Uint32 stride = program->GetTransformFeedbackStride(static_cast<Uint32>(bufferIndex));
if (stride == 0) continue;
const auto& bindingPoint =
MG_State::pGLContext->GetBufferBindingPoint(BufferTarget::TransformFeedback,
static_cast<Uint>(bufferIndex));
const auto& buffer = bindingPoint.GetBoundObject();
if (buffer == nullptr) continue;
const Range1D range = bindingPoint.GetRange();
const Uint8* mapped = buffer->MappedData();
if (mapped == nullptr) continue;
// The geometry stage amplifies, so the CPU vertex counter does not bound
// the capture; the binding range's whole-triangle capacity does.
const Uint64 rangeBytes = range.end > range.start ? static_cast<Uint64>(range.end - range.start) : 0;
const Uint64 capturedTriangles = std::min<Uint64>(triangleBase, (rangeBytes / stride) / 3);
// Observed Vulkan capture order for odd strip triangles is (i, i+2, i+1)
// (winding preserved by swapping the trailing pair); GL wants
// (i+1, i, i+2), which is one rotation away: (a,b,c) -> (c,a,b).
Vector<Uint8> scratch(stride);
for (const Uint64 triangle : swapTriangles) {
if (triangle >= capturedTriangles) break;
const SizeT v0Offset = static_cast<SizeT>(range.start) + static_cast<SizeT>(triangle * 3) * stride;
const SizeT v1Offset = v0Offset + stride;
const SizeT v2Offset = v1Offset + stride;
Memcpy(scratch.data(), mapped + v2Offset, stride);
buffer->WritebackFromBackend({const_cast<Uint8*>(mapped) + v1Offset, stride}, v2Offset);
buffer->WritebackFromBackend({const_cast<Uint8*>(mapped) + v0Offset, stride}, v1Offset);
buffer->WritebackFromBackend({scratch.data(), stride}, v0Offset);
}
}
}
void EndTransformFeedback(void) {
if (!MG_State::pGLContext->IsTransformFeedbackActive()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "Transform feedback is not active."));
return;
}
const auto capturedProgram = MG_State::pGLContext->GetTransformFeedbackProgram();
const Uint64 inputPrimitives = MG_State::pGLContext->GetTransformFeedbackInputPrimitives();
// Closed while the capture state is still active: a backend that captures
// through its own driver reads the capture program and buffer bindings here.
if (const auto endXfb = MG_Backend::gBackendFunctionsTable.GL.EndTransformFeedback) {
endXfb();
}
MG_State::pGLContext->EndTransformFeedback();
// Captured results must be visible to MapBuffer/GetBufferSubData after
// End; the capture targets are host-coherent GPU memory, so completing
// the GPU work is all that is required.
auto& backendGL = MG_Backend::gBackendFunctionsTable.GL;
if (backendGL.FenceSync && backendGL.ClientWaitSync) {
if (auto sync = backendGL.FenceSync()) {
backendGL.ClientWaitSync(sync, GL_SYNC_FLUSH_COMMANDS_BIT, ~0ull);
if (backendGL.DeleteSync) {
backendGL.DeleteSync(sync);
}
}
}
FixupGsStripCaptureOrder(capturedProgram, inputPrimitives);
}
void PauseTransformFeedback(void) {
if (!MG_State::pGLContext->IsTransformFeedbackActive() ||
MG_State::pGLContext->IsTransformFeedbackPaused()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"Transform feedback is not active, or is already paused."));
return;
}
MG_State::pGLContext->SetTransformFeedbackPaused(true);
if (const auto pauseXfb = MG_Backend::gBackendFunctionsTable.GL.PauseTransformFeedback) {
pauseXfb();
}
}
void ResumeTransformFeedback(void) {
if (!MG_State::pGLContext->IsTransformFeedbackActive() ||
!MG_State::pGLContext->IsTransformFeedbackPaused()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "Transform feedback is not paused."));
return;
}
MG_State::pGLContext->SetTransformFeedbackPaused(false);
if (const auto resumeXfb = MG_Backend::gBackendFunctionsTable.GL.ResumeTransformFeedback) {
resumeXfb();
}
}
void GenTransformFeedbacks(GLsizei n, GLuint* ids) {
if (n < 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "n must be non-negative."));
return;
}
if (n == 0 || ids == nullptr) return;
Vector<Uint> names;
MG_State::pGLContext->GenTransformFeedbackNames(static_cast<Uint>(n), names);
Memcpy(ids, names.data(), static_cast<SizeT>(n) * sizeof(GLuint));
}
void CreateTransformFeedbacks(GLsizei n, GLuint* ids) {
if (n < 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "n must be non-negative."));
return;
}
if (n == 0 || ids == nullptr) return;
Vector<Uint> names;
MG_State::pGLContext->GenTransformFeedbackNames(static_cast<Uint>(n), names);
// Unlike glGenTransformFeedbacks, the names are objects immediately: there is no bind step
// to create them from (GL 4.6 core 13.2.1).
for (const Uint name : names) {
MG_State::pGLContext->CreateTransformFeedbackObject(name);
}
Memcpy(ids, names.data(), static_cast<SizeT>(n) * sizeof(GLuint));
}
namespace {
// Shared front half of the by-name transform feedback entry points: the object has to exist
// (INVALID_OPERATION otherwise) before anything else about the call is looked at.
Bool ValidateNamedTransformFeedback(GLuint xfb, const char* functionName) {
if (!MG_State::pGLContext->IsTransformFeedbackObject(xfb)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName,
std::to_string(xfb) + " is not a transform feedback object."));
return false;
}
return true;
}
Bool ValidateTransformFeedbackBufferIndex(GLuint index, const char* functionName) {
if (index >= MG_State::GLState::GLContext::MAX_TRANSFORM_FEEDBACK_BUFFERS) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName,
"index exceeds GL_MAX_TRANSFORM_FEEDBACK_BUFFERS."));
return false;
}
return true;
}
// A capture binding may not be changed while the object is capturing (GL 4.6 core 13.2.2).
Bool ValidateNamedTransformFeedbackNotActive(GLuint xfb, const char* functionName) {
if (MG_State::pGLContext->IsNamedTransformFeedbackActive(xfb)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName,
"The transform feedback object is capturing."));
return false;
}
return true;
}
SharedPtr<MG_State::GLState::BufferObject> ResolveTransformFeedbackBuffer(GLuint buffer,
const char* functionName) {
if (buffer == 0) return nullptr;
if (!MG_State::pGLContext->ValidateBufferName(buffer)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName,
std::to_string(buffer) + " is not a buffer object."));
return nullptr;
}
return MG_State::pGLContext->GetBufferObject(buffer);
}
} // namespace
void TransformFeedbackBufferBase(GLuint xfb, GLuint index, GLuint buffer) {
if (!ValidateNamedTransformFeedback(xfb, __func__)) return;
if (!ValidateTransformFeedbackBufferIndex(index, __func__)) return;
if (!ValidateNamedTransformFeedbackNotActive(xfb, __func__)) return;
if (buffer != 0 && !MG_State::pGLContext->ValidateBufferName(buffer)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
std::to_string(buffer) + " is not a buffer object."));
return;
}
MG_State::pGLContext->SetNamedTransformFeedbackBinding(xfb, index,
ResolveTransformFeedbackBuffer(buffer, __func__), {},
false);
}
void TransformFeedbackBufferRange(GLuint xfb, GLuint index, GLuint buffer, GLintptr offset, GLsizeiptr size) {
if (!ValidateNamedTransformFeedback(xfb, __func__)) return;
if (!ValidateTransformFeedbackBufferIndex(index, __func__)) return;
if (!ValidateNamedTransformFeedbackNotActive(xfb, __func__)) return;
if (offset < 0 || size <= 0 || (offset % 4) != 0 || (size % 4) != 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"offset and size must be non-negative multiples of 4."));
return;
}
if (buffer != 0 && !MG_State::pGLContext->ValidateBufferName(buffer)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
std::to_string(buffer) + " is not a buffer object."));
return;
}
auto bufferObject = ResolveTransformFeedbackBuffer(buffer, __func__);
const Range1D range{static_cast<SizeT>(offset), static_cast<SizeT>(offset) + static_cast<SizeT>(size)};
MG_State::pGLContext->SetNamedTransformFeedbackBinding(xfb, index, bufferObject, range,
bufferObject != nullptr);
}
void GetTransformFeedbackiv(GLuint xfb, GLenum pname, GLint* param) {
if (!ValidateNamedTransformFeedback(xfb, __func__)) return;
if (!param) return;
switch (pname) {
case GL_TRANSFORM_FEEDBACK_ACTIVE:
*param = MG_State::pGLContext->IsNamedTransformFeedbackActive(xfb) ? GL_TRUE : GL_FALSE;
return;
case GL_TRANSFORM_FEEDBACK_PAUSED:
*param = MG_State::pGLContext->IsNamedTransformFeedbackPaused(xfb) ? GL_TRUE : GL_FALSE;
return;
default:
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"pname must be GL_TRANSFORM_FEEDBACK_ACTIVE or _PAUSED."));
return;
}
}
void GetTransformFeedbacki_v(GLuint xfb, GLenum pname, GLuint index, GLint* param) {
if (!ValidateNamedTransformFeedback(xfb, __func__)) return;
if (pname != GL_TRANSFORM_FEEDBACK_BUFFER_BINDING) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"pname must be GL_TRANSFORM_FEEDBACK_BUFFER_BINDING."));
return;
}
if (!ValidateTransformFeedbackBufferIndex(index, __func__)) return;
if (!param) return;
const auto binding = MG_State::pGLContext->GetNamedTransformFeedbackBinding(xfb, index);
*param = binding.Buffer ? static_cast<GLint>(binding.Buffer->GetExternalIndex()) : 0;
}
void GetTransformFeedbacki64_v(GLuint xfb, GLenum pname, GLuint index, GLint64* param) {
if (!ValidateNamedTransformFeedback(xfb, __func__)) return;
if (pname != GL_TRANSFORM_FEEDBACK_BUFFER_START && pname != GL_TRANSFORM_FEEDBACK_BUFFER_SIZE) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"pname must be GL_TRANSFORM_FEEDBACK_BUFFER_START or _SIZE."));
return;
}
if (!ValidateTransformFeedbackBufferIndex(index, __func__)) return;
if (!param) return;
const auto binding = MG_State::pGLContext->GetNamedTransformFeedbackBinding(xfb, index);
// glTransformFeedbackBufferBase leaves both at zero; only the range form sets them
// (GL 4.6 core table 23.48).
if (!binding.Buffer || !binding.HasExplicitRange) {
*param = 0;
return;
}
*param = (pname == GL_TRANSFORM_FEEDBACK_BUFFER_START)
? static_cast<GLint64>(binding.Range.start)
: static_cast<GLint64>(binding.Range.end - binding.Range.start);
}
void DeleteTransformFeedbacks(GLsizei n, const GLuint* ids) {
if (n < 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "n must be non-negative."));
return;
}
if (ids == nullptr) return;
for (GLsizei i = 0; i < n; ++i) {
const GLuint id = ids[i];
// Unknown names and 0 are silently ignored; an object whose capture span is
// still open is not (GL 4.6 core 13.2.1).
if (id == 0 || !MG_State::pGLContext->ValidateTransformFeedbackName(id)) continue;
if (id == MG_State::pGLContext->GetBoundTransformFeedbackName() &&
MG_State::pGLContext->IsTransformFeedbackActive()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"Cannot delete a transform feedback object whose capture is active."));
continue;
}
if (const auto deleteXfb = MG_Backend::gBackendFunctionsTable.GL.DeleteTransformFeedback) {
deleteXfb(id);
}
MG_State::pGLContext->MarkTransformFeedbackObjectForDeletion(id);
}
}
void BindTransformFeedback(GLenum target, GLuint id) {
if (target != GL_TRANSFORM_FEEDBACK) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "target must be GL_TRANSFORM_FEEDBACK."));
return;
}
// A running capture pins its object; only a paused one may be swapped out.
if (MG_State::pGLContext->IsTransformFeedbackActive() &&
!MG_State::pGLContext->IsTransformFeedbackPaused()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"Transform feedback is active and not paused."));
return;
}
if (!MG_State::pGLContext->ValidateTransformFeedbackName(id)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
std::to_string(id) + " is not a transform feedback object name."));
return;
}
MG_State::pGLContext->BindTransformFeedbackObject(id);
if (const auto bindXfb = MG_Backend::gBackendFunctionsTable.GL.BindTransformFeedback) {
bindXfb(id);
}
}
GLboolean IsTransformFeedback(GLuint id) {
// Name 0 is the default object, and a name glGenTransformFeedbacks handed out only
// becomes the name of an object once it has been bound.
return MG_State::pGLContext->IsTransformFeedbackObject(id) ? GL_TRUE : GL_FALSE;
}
// glDrawTransformFeedback[Stream][Instanced]: replays the vertices the named object
// captured in its last completed span, as if by glDrawArraysInstanced with that count
// (GL 4.6 core 10.3.7).
static void DrawTransformFeedbackImpl(const char* functionName, GLenum mode, GLuint id, GLuint stream,
GLsizei instancecount) {
if (!ValidateCurrentProgramForExecution(functionName)) return;
if (!ValidatePrimitiveModeForBackend(functionName, mode)) return;
if (instancecount < 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName, "instancecount must be non-negative."));
return;
}
if (!MG_State::pGLContext->ValidateTransformFeedbackName(id)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName,
std::to_string(id) + " is not a transform feedback object name."));
return;
}
// GL_MAX_VERTEX_STREAMS is 1, so stream 0 is the only one that exists.
if (stream != 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName,
"stream must be less than GL_MAX_VERTEX_STREAMS."));
return;
}
// Drawing from an object whose capture is currently open is legal and deliberate:
// it is how a transform feedback result is fed straight back into the next span
// (ARB_transform_feedback2 lists no such restriction).
if (!MG_State::pGLContext->HasTransformFeedbackCompletedSpan(id)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName,
"glEndTransformFeedback has never been called for this object."));
return;
}
const Uint64 vertices = MG_State::pGLContext->GetTransformFeedbackRecordedVertices(id);
if (vertices == 0) return;
const auto count = static_cast<GLsizei>(vertices);
AccountTransformFeedbackPrimitives(mode, count);
if (instancecount == 1) {
DrawArrays_Backend(mode, 0, count);
} else {
DrawArraysInstanced_Backend(mode, 0, count, instancecount);
}
}
void DrawTransformFeedback(GLenum mode, GLuint id) {
DrawTransformFeedbackImpl(__func__, mode, id, 0, 1);
}
void DrawTransformFeedbackInstanced(GLenum mode, GLuint id, GLsizei instancecount) {
DrawTransformFeedbackImpl(__func__, mode, id, 0, instancecount);
}
void DrawTransformFeedbackStream(GLenum mode, GLuint id, GLuint stream) {
DrawTransformFeedbackImpl(__func__, mode, id, stream, 1);
}
void DrawTransformFeedbackStreamInstanced(GLenum mode, GLuint id, GLuint stream, GLsizei instancecount) {
DrawTransformFeedbackImpl(__func__, mode, id, stream, instancecount);
}
} // namespace MobileGL::MG_Impl::GLImpl
@@ -11,27 +11,8 @@
namespace MobileGL::MG_Impl::GLImpl {
/* @INSERTION_POINT:FUNCTION_DECLARATION@ */
void BeginTransformFeedback(GLenum primitiveMode);
void EndTransformFeedback(void);
void PauseTransformFeedback(void);
void ResumeTransformFeedback(void);
void GenTransformFeedbacks(GLsizei n, GLuint* ids);
void CreateTransformFeedbacks(GLsizei n, GLuint* ids);
void DeleteTransformFeedbacks(GLsizei n, const GLuint* ids);
void TransformFeedbackBufferBase(GLuint xfb, GLuint index, GLuint buffer);
void TransformFeedbackBufferRange(GLuint xfb, GLuint index, GLuint buffer, GLintptr offset, GLsizeiptr size);
void GetTransformFeedbackiv(GLuint xfb, GLenum pname, GLint* param);
void GetTransformFeedbacki_v(GLuint xfb, GLenum pname, GLuint index, GLint* param);
void GetTransformFeedbacki64_v(GLuint xfb, GLenum pname, GLuint index, GLint64* param);
void BindTransformFeedback(GLenum target, GLuint id);
GLboolean IsTransformFeedback(GLuint id);
void DrawTransformFeedback(GLenum mode, GLuint id);
void DrawTransformFeedbackInstanced(GLenum mode, GLuint id, GLsizei instancecount);
void DrawTransformFeedbackStream(GLenum mode, GLuint id, GLuint stream);
void DrawTransformFeedbackStreamInstanced(GLenum mode, GLuint id, GLuint stream, GLsizei instancecount);
void DispatchCompute(GLuint numGroupsX, GLuint numGroupsY, GLuint numGroupsZ);
void DispatchComputeIndirect(GLintptr indirect);
void PatchParameteri(GLenum pname, GLint value);
void MemoryBarrier(GLbitfield barriers);
void MemoryBarrierByRegion(GLbitfield barriers);
void MultiDrawElementsIndirect(GLenum mode, GLenum type, const void* indirect, GLsizei drawcount, GLsizei stride);
@@ -236,12 +236,12 @@ DECLARE_GL_FUNCTION_HEAD(void, DeleteVertexArrays, GLsizei n, const GLuint* arra
DECLARE_GL_FUNCTION_HEAD(void, GenVertexArrays, GLsizei n, GLuint* arrays) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GenVertexArrays, n, arrays)
DECLARE_GL_FUNCTION_HEAD(GLboolean, IsVertexArray, GLuint array) DECLARE_GL_FUNCTION_END(GLboolean, IsVertexArray, array)
DECLARE_GL_FUNCTION_HEAD(void, GetIntegeri_v, GLenum target, GLuint index, GLint* data) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetIntegeri_v, target, index, data)
DECLARE_GL_FUNCTION_HEAD(void, BeginTransformFeedback, GLenum primitiveMode) DECLARE_GL_FUNCTION_END_NO_RETURN(void, BeginTransformFeedback, primitiveMode)
DECLARE_GL_FUNCTION_HEAD(void, EndTransformFeedback) DECLARE_GL_FUNCTION_END_NO_RETURN(void, EndTransformFeedback)
DECLARE_GL_FUNCTION_STUB_HEAD(void, BeginTransformFeedback, GLenum primitiveMode) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, BeginTransformFeedback, primitiveMode)
DECLARE_GL_FUNCTION_STUB_HEAD(void, EndTransformFeedback) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, EndTransformFeedback)
DECLARE_GL_FUNCTION_HEAD(void, BindBufferRange, GLenum target, GLuint index, GLuint buffer, GLintptr offset, GLsizeiptr size) DECLARE_GL_FUNCTION_END_NO_RETURN(void, BindBufferRange, target, index, buffer, offset, size)
DECLARE_GL_FUNCTION_HEAD(void, BindBufferBase, GLenum target, GLuint index, GLuint buffer) DECLARE_GL_FUNCTION_END_NO_RETURN(void, BindBufferBase, target, index, buffer)
DECLARE_GL_FUNCTION_HEAD(void, TransformFeedbackVaryings, GLuint program, GLsizei count, const GLchar* const* varyings, GLenum bufferMode) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TransformFeedbackVaryings, program, count, varyings, bufferMode)
DECLARE_GL_FUNCTION_HEAD(void, GetTransformFeedbackVarying, GLuint program, GLuint index, GLsizei bufSize, GLsizei* length, GLsizei* size, GLenum* type, GLchar* name) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetTransformFeedbackVarying, program, index, bufSize, length, size, type, name)
DECLARE_GL_FUNCTION_STUB_HEAD(void, TransformFeedbackVaryings, GLuint program, GLsizei count, const GLchar* const* varyings, GLenum bufferMode) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, TransformFeedbackVaryings, program, count, varyings, bufferMode)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetTransformFeedbackVarying, GLuint program, GLuint index, GLsizei bufSize, GLsizei* length, GLsizei* size, GLenum* type, GLchar* name) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetTransformFeedbackVarying, program, index, bufSize, length, size, type, name)
DECLARE_GL_FUNCTION_HEAD(void, VertexAttribIPointer, GLuint index, GLint size, GLenum type, GLsizei stride, const void* pointer) DECLARE_GL_FUNCTION_END_NO_RETURN(void, VertexAttribIPointer, index, size, type, stride, pointer)
DECLARE_GL_FUNCTION_HEAD(void, GetVertexAttribIiv, GLuint index, GLenum pname, GLint* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetVertexAttribIiv, index, pname, params)
DECLARE_GL_FUNCTION_HEAD(void, GetVertexAttribIuiv, GLuint index, GLenum pname, GLuint* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetVertexAttribIuiv, index, pname, params)
@@ -292,15 +292,21 @@ DECLARE_GL_FUNCTION_HEAD(void, SamplerParameterfv, GLuint sampler, GLenum pname,
DECLARE_GL_FUNCTION_HEAD(void, GetSamplerParameteriv, GLuint sampler, GLenum pname, GLint* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetSamplerParameteriv, sampler, pname, params)
DECLARE_GL_FUNCTION_HEAD(void, GetSamplerParameterfv, GLuint sampler, GLenum pname, GLfloat* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetSamplerParameterfv, sampler, pname, params)
DECLARE_GL_FUNCTION_HEAD(void, VertexAttribDivisor, GLuint index, GLuint divisor) DECLARE_GL_FUNCTION_END_NO_RETURN(void, VertexAttribDivisor, index, divisor)
DECLARE_GL_FUNCTION_HEAD(void, BindTransformFeedback, GLenum target, GLuint id) DECLARE_GL_FUNCTION_END_NO_RETURN(void, BindTransformFeedback, target, id)
DECLARE_GL_FUNCTION_HEAD(void, DeleteTransformFeedbacks, GLsizei n, const GLuint* ids) DECLARE_GL_FUNCTION_END_NO_RETURN(void, DeleteTransformFeedbacks, n, ids)
DECLARE_GL_FUNCTION_HEAD(void, GenTransformFeedbacks, GLsizei n, GLuint* ids) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GenTransformFeedbacks, n, ids)
DECLARE_GL_FUNCTION_HEAD(GLboolean, IsTransformFeedback, GLuint id) DECLARE_GL_FUNCTION_END(GLboolean, IsTransformFeedback, id)
DECLARE_GL_FUNCTION_HEAD(void, PauseTransformFeedback) DECLARE_GL_FUNCTION_END_NO_RETURN(void, PauseTransformFeedback)
DECLARE_GL_FUNCTION_HEAD(void, ResumeTransformFeedback) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ResumeTransformFeedback)
DECLARE_GL_FUNCTION_HEAD(void, GetProgramBinary, GLuint program, GLsizei bufSize, GLsizei* length, GLenum* binaryFormat, void* binary) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetProgramBinary, program, bufSize, length, binaryFormat, binary)
DECLARE_GL_FUNCTION_HEAD(void, ProgramBinary, GLuint program, GLenum binaryFormat, const void* binary, GLsizei length) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramBinary, program, binaryFormat, binary, length)
DECLARE_GL_FUNCTION_HEAD(void, ProgramParameteri, GLuint program, GLenum pname, GLint value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramParameteri, program, pname, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, BindTransformFeedback, GLenum target, GLuint id) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, BindTransformFeedback, target, id)
DECLARE_GL_FUNCTION_STUB_HEAD(void, DeleteTransformFeedbacks, GLsizei n, const GLuint* ids) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, DeleteTransformFeedbacks, n, ids)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GenTransformFeedbacks, GLsizei n, GLuint* ids) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GenTransformFeedbacks, n, ids)
// Transform feedback objects are not implemented, so no name is ever a live object. The shared
// stub returns (type)1, telling a probing caller that every id it invents already exists; GL_FALSE
// is both truthful and what the spec requires for a name that was never generated.
MOBILEGL_GL_API GLboolean glIsTransformFeedback(GLuint id) {
MGLOG_W("Stub function: %s(...)", __FUNCTION__);
return GL_FALSE;
}
DECLARE_GL_FUNCTION_STUB_HEAD(void, PauseTransformFeedback) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, PauseTransformFeedback)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ResumeTransformFeedback) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ResumeTransformFeedback)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetProgramBinary, GLuint program, GLsizei bufSize, GLsizei* length, GLenum* binaryFormat, void* binary) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetProgramBinary, program, bufSize, length, binaryFormat, binary)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramBinary, GLuint program, GLenum binaryFormat, const void* binary, GLsizei length) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramBinary, program, binaryFormat, binary, length)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramParameteri, GLuint program, GLenum pname, GLint value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramParameteri, program, pname, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, InvalidateFramebuffer, GLenum target, GLsizei numAttachments, const GLenum* attachments) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, InvalidateFramebuffer, target, numAttachments, attachments)
DECLARE_GL_FUNCTION_STUB_HEAD(void, InvalidateSubFramebuffer, GLenum target, GLsizei numAttachments, const GLenum* attachments, GLint x, GLint y, GLsizei width, GLsizei height) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, InvalidateSubFramebuffer, target, numAttachments, attachments, x, y, width, height)
DECLARE_GL_FUNCTION_HEAD(void, TexStorage2D, GLenum target, GLsizei levels, GLenum internalformat, GLsizei width, GLsizei height) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TexStorage2D, target, levels, internalformat, width, height)
@@ -419,12 +425,12 @@ DECLARE_GL_FUNCTION_HEAD(void, DrawElementsInstancedBaseVertex, GLenum mode, GLs
DECLARE_GL_FUNCTION_HEAD(void, FramebufferTexture, GLenum target, GLenum attachment, GLuint texture, GLint level) DECLARE_GL_FUNCTION_END_NO_RETURN(void, FramebufferTexture, target, attachment, texture, level)
DECLARE_GL_FUNCTION_STUB_HEAD(void, PrimitiveBoundingBox, GLfloat minX, GLfloat minY, GLfloat minZ, GLfloat minW, GLfloat maxX, GLfloat maxY, GLfloat maxZ, GLfloat maxW) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, PrimitiveBoundingBox, minX, minY, minZ, minW, maxX, maxY, maxZ, maxW)
DECLARE_GL_FUNCTION_HEAD(GLenum, GetGraphicsResetStatus) DECLARE_GL_FUNCTION_END(GLenum, GetGraphicsResetStatus)
DECLARE_GL_FUNCTION_HEAD(void, ReadnPixels, GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, GLsizei bufSize, void* data) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ReadnPixels, x, y, width, height, format, type, bufSize, data)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ReadnPixels, GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, GLsizei bufSize, void* data) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ReadnPixels, x, y, width, height, format, type, bufSize, data)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetnUniformfv, GLuint program, GLint location, GLsizei bufSize, GLfloat* params) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetnUniformfv, program, location, bufSize, params)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetnUniformiv, GLuint program, GLint location, GLsizei bufSize, GLint* params) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetnUniformiv, program, location, bufSize, params)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetnUniformuiv, GLuint program, GLint location, GLsizei bufSize, GLuint* params) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetnUniformuiv, program, location, bufSize, params)
DECLARE_GL_FUNCTION_STUB_HEAD(void, MinSampleShading, GLfloat value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, MinSampleShading, value)
DECLARE_GL_FUNCTION_HEAD(void, PatchParameteri, GLenum pname, GLint value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, PatchParameteri, pname, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, PatchParameteri, GLenum pname, GLint value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, PatchParameteri, pname, value)
DECLARE_GL_FUNCTION_HEAD(void, TexParameterIiv, GLenum target, GLenum pname, const GLint* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TexParameterIiv, target, pname, params)
DECLARE_GL_FUNCTION_HEAD(void, TexParameterIuiv, GLenum target, GLenum pname, const GLuint* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TexParameterIuiv, target, pname, params)
DECLARE_GL_FUNCTION_HEAD(void, GetTexParameterIiv, GLenum target, GLenum pname, GLint* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetTexParameterIiv, target, pname, params)
@@ -434,7 +440,7 @@ DECLARE_GL_FUNCTION_HEAD(void, SamplerParameterIuiv, GLuint sampler, GLenum pnam
DECLARE_GL_FUNCTION_HEAD(void, GetSamplerParameterIiv, GLuint sampler, GLenum pname, GLint* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetSamplerParameterIiv, sampler, pname, params)
DECLARE_GL_FUNCTION_HEAD(void, GetSamplerParameterIuiv, GLuint sampler, GLenum pname, GLuint* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetSamplerParameterIuiv, sampler, pname, params)
DECLARE_GL_FUNCTION_HEAD(void, TexBuffer, GLenum target, GLenum internalformat, GLuint buffer) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TexBuffer, target, internalformat, buffer)
DECLARE_GL_FUNCTION_HEAD(void, TexBufferRange, GLenum target, GLenum internalformat, GLuint buffer, GLintptr offset, GLsizeiptr size) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TexBufferRange, target, internalformat, buffer, offset, size)
DECLARE_GL_FUNCTION_STUB_HEAD(void, TexBufferRange, GLenum target, GLenum internalformat, GLuint buffer, GLintptr offset, GLsizeiptr size) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, TexBufferRange, target, internalformat, buffer, offset, size)
DECLARE_GL_FUNCTION_HEAD(void, TexStorage3DMultisample, GLenum target, GLsizei samples, GLenum internalformat, GLsizei width, GLsizei height, GLsizei depth, GLboolean fixedsamplelocations) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TexStorage3DMultisample, target, samples, internalformat, width, height, depth, fixedsamplelocations)
DECLARE_GL_FUNCTION_HEAD(void*, MapBufferRange, GLenum target, GLintptr offset, GLsizeiptr length, GLbitfield access) DECLARE_GL_FUNCTION_END(void*, MapBufferRange, target, offset, length, access)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ClearIndex, GLfloat c) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ClearIndex, c)
@@ -909,24 +915,24 @@ DECLARE_GL_FUNCTION_STUB_HEAD(void, ColorP4ui, GLenum type, GLuint color) DECLAR
DECLARE_GL_FUNCTION_STUB_HEAD(void, ColorP4uiv, GLenum type, const GLuint* color) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ColorP4uiv, type, color)
DECLARE_GL_FUNCTION_STUB_HEAD(void, SecondaryColorP3ui, GLenum type, GLuint color) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, SecondaryColorP3ui, type, color)
DECLARE_GL_FUNCTION_STUB_HEAD(void, SecondaryColorP3uiv, GLenum type, const GLuint* color) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, SecondaryColorP3uiv, type, color)
DECLARE_GL_FUNCTION_HEAD(void, Uniform1d, GLint location, GLdouble x) DECLARE_GL_FUNCTION_END_NO_RETURN(void, Uniform1d, location, x)
DECLARE_GL_FUNCTION_HEAD(void, Uniform2d, GLint location, GLdouble x, GLdouble y) DECLARE_GL_FUNCTION_END_NO_RETURN(void, Uniform2d, location, x, y)
DECLARE_GL_FUNCTION_HEAD(void, Uniform3d, GLint location, GLdouble x, GLdouble y, GLdouble z) DECLARE_GL_FUNCTION_END_NO_RETURN(void, Uniform3d, location, x, y, z)
DECLARE_GL_FUNCTION_HEAD(void, Uniform4d, GLint location, GLdouble x, GLdouble y, GLdouble z, GLdouble w) DECLARE_GL_FUNCTION_END_NO_RETURN(void, Uniform4d, location, x, y, z, w)
DECLARE_GL_FUNCTION_HEAD(void, Uniform1dv, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, Uniform1dv, location, count, value)
DECLARE_GL_FUNCTION_HEAD(void, Uniform2dv, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, Uniform2dv, location, count, value)
DECLARE_GL_FUNCTION_HEAD(void, Uniform3dv, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, Uniform3dv, location, count, value)
DECLARE_GL_FUNCTION_HEAD(void, Uniform4dv, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, Uniform4dv, location, count, value)
DECLARE_GL_FUNCTION_HEAD(void, UniformMatrix2dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, UniformMatrix2dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, UniformMatrix3dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, UniformMatrix3dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, UniformMatrix4dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, UniformMatrix4dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, UniformMatrix2x3dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, UniformMatrix2x3dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, UniformMatrix2x4dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, UniformMatrix2x4dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, UniformMatrix3x2dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, UniformMatrix3x2dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, UniformMatrix3x4dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, UniformMatrix3x4dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, UniformMatrix4x2dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, UniformMatrix4x2dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, UniformMatrix4x3dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, UniformMatrix4x3dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, GetUniformdv, GLuint program, GLint location, GLdouble* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetUniformdv, program, location, params)
DECLARE_GL_FUNCTION_STUB_HEAD(void, Uniform1d, GLint location, GLdouble x) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, Uniform1d, location, x)
DECLARE_GL_FUNCTION_STUB_HEAD(void, Uniform2d, GLint location, GLdouble x, GLdouble y) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, Uniform2d, location, x, y)
DECLARE_GL_FUNCTION_STUB_HEAD(void, Uniform3d, GLint location, GLdouble x, GLdouble y, GLdouble z) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, Uniform3d, location, x, y, z)
DECLARE_GL_FUNCTION_STUB_HEAD(void, Uniform4d, GLint location, GLdouble x, GLdouble y, GLdouble z, GLdouble w) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, Uniform4d, location, x, y, z, w)
DECLARE_GL_FUNCTION_STUB_HEAD(void, Uniform1dv, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, Uniform1dv, location, count, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, Uniform2dv, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, Uniform2dv, location, count, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, Uniform3dv, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, Uniform3dv, location, count, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, Uniform4dv, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, Uniform4dv, location, count, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, UniformMatrix2dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, UniformMatrix2dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, UniformMatrix3dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, UniformMatrix3dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, UniformMatrix4dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, UniformMatrix4dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, UniformMatrix2x3dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, UniformMatrix2x3dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, UniformMatrix2x4dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, UniformMatrix2x4dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, UniformMatrix3x2dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, UniformMatrix3x2dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, UniformMatrix3x4dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, UniformMatrix3x4dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, UniformMatrix4x2dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, UniformMatrix4x2dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, UniformMatrix4x3dv, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, UniformMatrix4x3dv, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetUniformdv, GLuint program, GLint location, GLdouble* params) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetUniformdv, program, location, params)
DECLARE_GL_FUNCTION_STUB_HEAD(GLint, GetSubroutineUniformLocation, GLuint program, GLenum shadertype, const GLchar* name) DECLARE_GL_FUNCTION_STUB_END(GLint, GetSubroutineUniformLocation, program, shadertype, name)
DECLARE_GL_FUNCTION_STUB_HEAD(GLuint, GetSubroutineIndex, GLuint program, GLenum shadertype, const GLchar* name) DECLARE_GL_FUNCTION_STUB_END(GLuint, GetSubroutineIndex, program, shadertype, name)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetActiveSubroutineUniformiv, GLuint program, GLenum shadertype, GLuint index, GLenum pname, GLint* values) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetActiveSubroutineUniformiv, program, shadertype, index, pname, values)
@@ -936,28 +942,28 @@ DECLARE_GL_FUNCTION_STUB_HEAD(void, UniformSubroutinesuiv, GLenum shadertype, GL
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetUniformSubroutineuiv, GLenum shadertype, GLint location, GLuint* params) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetUniformSubroutineuiv, shadertype, location, params)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetProgramStageiv, GLuint program, GLenum shadertype, GLenum pname, GLint* values) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetProgramStageiv, program, shadertype, pname, values)
DECLARE_GL_FUNCTION_STUB_HEAD(void, PatchParameterfv, GLenum pname, const GLfloat* values) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, PatchParameterfv, pname, values)
DECLARE_GL_FUNCTION_HEAD(void, DrawTransformFeedback, GLenum mode, GLuint id) DECLARE_GL_FUNCTION_END_NO_RETURN(void, DrawTransformFeedback, mode, id)
DECLARE_GL_FUNCTION_HEAD(void, DrawTransformFeedbackStream, GLenum mode, GLuint id, GLuint stream) DECLARE_GL_FUNCTION_END_NO_RETURN(void, DrawTransformFeedbackStream, mode, id, stream)
DECLARE_GL_FUNCTION_HEAD(void, BeginQueryIndexed, GLenum target, GLuint index, GLuint id) DECLARE_GL_FUNCTION_END_NO_RETURN(void, BeginQueryIndexed, target, index, id)
DECLARE_GL_FUNCTION_HEAD(void, EndQueryIndexed, GLenum target, GLuint index) DECLARE_GL_FUNCTION_END_NO_RETURN(void, EndQueryIndexed, target, index)
DECLARE_GL_FUNCTION_HEAD(void, GetQueryIndexediv, GLenum target, GLuint index, GLenum pname, GLint* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetQueryIndexediv, target, index, pname, params)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniform1d, GLuint program, GLint location, GLdouble v0) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniform1d, program, location, v0)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniform1dv, GLuint program, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniform1dv, program, location, count, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniform2d, GLuint program, GLint location, GLdouble v0, GLdouble v1) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniform2d, program, location, v0, v1)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniform2dv, GLuint program, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniform2dv, program, location, count, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniform3d, GLuint program, GLint location, GLdouble v0, GLdouble v1, GLdouble v2) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniform3d, program, location, v0, v1, v2)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniform3dv, GLuint program, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniform3dv, program, location, count, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniform4d, GLuint program, GLint location, GLdouble v0, GLdouble v1, GLdouble v2, GLdouble v3) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniform4d, program, location, v0, v1, v2, v3)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniform4dv, GLuint program, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniform4dv, program, location, count, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniformMatrix2dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniformMatrix2dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniformMatrix3dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniformMatrix3dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniformMatrix4dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniformMatrix4dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniformMatrix2x3dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniformMatrix2x3dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniformMatrix3x2dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniformMatrix3x2dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniformMatrix2x4dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniformMatrix2x4dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniformMatrix4x2dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniformMatrix4x2dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniformMatrix3x4dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniformMatrix3x4dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_HEAD(void, ProgramUniformMatrix4x3dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ProgramUniformMatrix4x3dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, DrawTransformFeedback, GLenum mode, GLuint id) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, DrawTransformFeedback, mode, id)
DECLARE_GL_FUNCTION_STUB_HEAD(void, DrawTransformFeedbackStream, GLenum mode, GLuint id, GLuint stream) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, DrawTransformFeedbackStream, mode, id, stream)
DECLARE_GL_FUNCTION_STUB_HEAD(void, BeginQueryIndexed, GLenum target, GLuint index, GLuint id) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, BeginQueryIndexed, target, index, id)
DECLARE_GL_FUNCTION_STUB_HEAD(void, EndQueryIndexed, GLenum target, GLuint index) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, EndQueryIndexed, target, index)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetQueryIndexediv, GLenum target, GLuint index, GLenum pname, GLint* params) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetQueryIndexediv, target, index, pname, params)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniform1d, GLuint program, GLint location, GLdouble v0) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniform1d, program, location, v0)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniform1dv, GLuint program, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniform1dv, program, location, count, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniform2d, GLuint program, GLint location, GLdouble v0, GLdouble v1) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniform2d, program, location, v0, v1)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniform2dv, GLuint program, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniform2dv, program, location, count, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniform3d, GLuint program, GLint location, GLdouble v0, GLdouble v1, GLdouble v2) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniform3d, program, location, v0, v1, v2)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniform3dv, GLuint program, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniform3dv, program, location, count, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniform4d, GLuint program, GLint location, GLdouble v0, GLdouble v1, GLdouble v2, GLdouble v3) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniform4d, program, location, v0, v1, v2, v3)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniform4dv, GLuint program, GLint location, GLsizei count, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniform4dv, program, location, count, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniformMatrix2dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniformMatrix2dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniformMatrix3dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniformMatrix3dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniformMatrix4dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniformMatrix4dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniformMatrix2x3dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniformMatrix2x3dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniformMatrix3x2dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniformMatrix3x2dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniformMatrix2x4dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniformMatrix2x4dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniformMatrix4x2dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniformMatrix4x2dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniformMatrix3x4dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniformMatrix3x4dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ProgramUniformMatrix4x3dv, GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ProgramUniformMatrix4x3dv, program, location, count, transpose, value)
DECLARE_GL_FUNCTION_STUB_HEAD(void, VertexAttribL1d, GLuint index, GLdouble x) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, VertexAttribL1d, index, x)
DECLARE_GL_FUNCTION_STUB_HEAD(void, VertexAttribL2d, GLuint index, GLdouble x, GLdouble y) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, VertexAttribL2d, index, x, y)
DECLARE_GL_FUNCTION_STUB_HEAD(void, VertexAttribL3d, GLuint index, GLdouble x, GLdouble y, GLdouble z) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, VertexAttribL3d, index, x, y, z)
@@ -982,8 +988,8 @@ DECLARE_GL_FUNCTION_HEAD(void, DrawArraysInstancedBaseInstance, GLenum mode, GLi
DECLARE_GL_FUNCTION_HEAD(void, DrawElementsInstancedBaseInstance, GLenum mode, GLsizei count, GLenum type, const void* indices, GLsizei instancecount, GLuint baseinstance) DECLARE_GL_FUNCTION_END_NO_RETURN(void, DrawElementsInstancedBaseInstance, mode, count, type, indices, instancecount, baseinstance)
DECLARE_GL_FUNCTION_HEAD(void, DrawElementsInstancedBaseVertexBaseInstance, GLenum mode, GLsizei count, GLenum type, const void* indices, GLsizei instancecount, GLint basevertex, GLuint baseinstance) DECLARE_GL_FUNCTION_END_NO_RETURN(void, DrawElementsInstancedBaseVertexBaseInstance, mode, count, type, indices, instancecount, basevertex, baseinstance)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetActiveAtomicCounterBufferiv, GLuint program, GLuint bufferIndex, GLenum pname, GLint* params) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetActiveAtomicCounterBufferiv, program, bufferIndex, pname, params)
DECLARE_GL_FUNCTION_HEAD(void, DrawTransformFeedbackInstanced, GLenum mode, GLuint id, GLsizei instancecount) DECLARE_GL_FUNCTION_END_NO_RETURN(void, DrawTransformFeedbackInstanced, mode, id, instancecount)
DECLARE_GL_FUNCTION_HEAD(void, DrawTransformFeedbackStreamInstanced, GLenum mode, GLuint id, GLuint stream, GLsizei instancecount) DECLARE_GL_FUNCTION_END_NO_RETURN(void, DrawTransformFeedbackStreamInstanced, mode, id, stream, instancecount)
DECLARE_GL_FUNCTION_STUB_HEAD(void, DrawTransformFeedbackInstanced, GLenum mode, GLuint id, GLsizei instancecount) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, DrawTransformFeedbackInstanced, mode, id, instancecount)
DECLARE_GL_FUNCTION_STUB_HEAD(void, DrawTransformFeedbackStreamInstanced, GLenum mode, GLuint id, GLuint stream, GLsizei instancecount) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, DrawTransformFeedbackStreamInstanced, mode, id, stream, instancecount)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ClearBufferData, GLenum target, GLenum internalformat, GLenum format, GLenum type, const void* data) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ClearBufferData, target, internalformat, format, type, data)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ClearBufferSubData, GLenum target, GLenum internalformat, GLintptr offset, GLsizeiptr size, GLenum format, GLenum type, const void* data) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ClearBufferSubData, target, internalformat, offset, size, format, type, data)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetInternalformati64v, GLenum target, GLenum internalformat, GLenum pname, GLsizei count, GLint64* params) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetInternalformati64v, target, internalformat, pname, count, params)
@@ -996,7 +1002,7 @@ DECLARE_GL_FUNCTION_HEAD(void, MultiDrawElementsIndirect, GLenum mode, GLenum ty
DECLARE_GL_FUNCTION_HEAD(GLint, GetProgramResourceLocationIndex, GLuint program, GLenum programInterface, const GLchar* name) DECLARE_GL_FUNCTION_END(GLint, GetProgramResourceLocationIndex, program, programInterface, name)
DECLARE_GL_FUNCTION_HEAD(void, ShaderStorageBlockBinding, GLuint program, GLuint storageBlockIndex, GLuint storageBlockBinding) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ShaderStorageBlockBinding, program, storageBlockIndex, storageBlockBinding)
DECLARE_GL_FUNCTION_STUB_HEAD(void, TextureView, GLuint texture, GLenum target, GLuint origtexture, GLenum internalformat, GLuint minlevel, GLuint numlevels, GLuint minlayer, GLuint numlayers) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, TextureView, texture, target, origtexture, internalformat, minlevel, numlevels, minlayer, numlayers)
DECLARE_GL_FUNCTION_HEAD(void, VertexAttribLFormat, GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset) DECLARE_GL_FUNCTION_END_NO_RETURN(void, VertexAttribLFormat, attribindex, size, type, relativeoffset)
DECLARE_GL_FUNCTION_STUB_HEAD(void, VertexAttribLFormat, GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, VertexAttribLFormat, attribindex, size, type, relativeoffset)
DECLARE_GL_FUNCTION_HEAD(void, BufferStorage, GLenum target, GLsizeiptr size, const void* data, GLbitfield flags) DECLARE_GL_FUNCTION_END_NO_RETURN(void, BufferStorage, target, size, data, flags)
DECLARE_GL_FUNCTION_HEAD(void, ClearTexImage, GLuint texture, GLint level, GLenum format, GLenum type, const void* data) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ClearTexImage, texture, level, format, type, data)
DECLARE_GL_FUNCTION_HEAD(void, ClearTexSubImage, GLuint texture, GLint level, GLint xoffset, GLint yoffset, GLint zoffset, GLsizei width, GLsizei height, GLsizei depth, GLenum format, GLenum type, const void* data) DECLARE_GL_FUNCTION_END_NO_RETURN(void, ClearTexSubImage, texture, level, xoffset, yoffset, zoffset, width, height, depth, format, type, data)
@@ -1007,12 +1013,12 @@ DECLARE_GL_FUNCTION_HEAD(void, BindSamplers, GLuint first, GLsizei count, const
DECLARE_GL_FUNCTION_STUB_HEAD(void, BindImageTextures, GLuint first, GLsizei count, const GLuint* textures) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, BindImageTextures, first, count, textures)
DECLARE_GL_FUNCTION_HEAD(void, BindVertexBuffers, GLuint first, GLsizei count, const GLuint* buffers, const GLintptr* offsets, const GLsizei* strides) DECLARE_GL_FUNCTION_END_NO_RETURN(void, BindVertexBuffers, first, count, buffers, offsets, strides)
DECLARE_GL_FUNCTION_STUB_HEAD(void, ClipControl, GLenum origin, GLenum depth) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, ClipControl, origin, depth)
DECLARE_GL_FUNCTION_HEAD(void, CreateTransformFeedbacks, GLsizei n, GLuint* ids) DECLARE_GL_FUNCTION_END_NO_RETURN(void, CreateTransformFeedbacks, n, ids)
DECLARE_GL_FUNCTION_HEAD(void, TransformFeedbackBufferBase, GLuint xfb, GLuint index, GLuint buffer) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TransformFeedbackBufferBase, xfb, index, buffer)
DECLARE_GL_FUNCTION_HEAD(void, TransformFeedbackBufferRange, GLuint xfb, GLuint index, GLuint buffer, GLintptr offset, GLsizeiptr size) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TransformFeedbackBufferRange, xfb, index, buffer, offset, size)
DECLARE_GL_FUNCTION_HEAD(void, GetTransformFeedbackiv, GLuint xfb, GLenum pname, GLint* param) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetTransformFeedbackiv, xfb, pname, param)
DECLARE_GL_FUNCTION_HEAD(void, GetTransformFeedbacki_v, GLuint xfb, GLenum pname, GLuint index, GLint* param) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetTransformFeedbacki_v, xfb, pname, index, param)
DECLARE_GL_FUNCTION_HEAD(void, GetTransformFeedbacki64_v, GLuint xfb, GLenum pname, GLuint index, GLint64* param) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetTransformFeedbacki64_v, xfb, pname, index, param)
DECLARE_GL_FUNCTION_STUB_HEAD(void, CreateTransformFeedbacks, GLsizei n, GLuint* ids) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, CreateTransformFeedbacks, n, ids)
DECLARE_GL_FUNCTION_STUB_HEAD(void, TransformFeedbackBufferBase, GLuint xfb, GLuint index, GLuint buffer) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, TransformFeedbackBufferBase, xfb, index, buffer)
DECLARE_GL_FUNCTION_STUB_HEAD(void, TransformFeedbackBufferRange, GLuint xfb, GLuint index, GLuint buffer, GLintptr offset, GLsizeiptr size) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, TransformFeedbackBufferRange, xfb, index, buffer, offset, size)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetTransformFeedbackiv, GLuint xfb, GLenum pname, GLint* param) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetTransformFeedbackiv, xfb, pname, param)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetTransformFeedbacki_v, GLuint xfb, GLenum pname, GLuint index, GLint* param) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetTransformFeedbacki_v, xfb, pname, index, param)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetTransformFeedbacki64_v, GLuint xfb, GLenum pname, GLuint index, GLint64* param) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetTransformFeedbacki64_v, xfb, pname, index, param)
DECLARE_GL_FUNCTION_HEAD(void, CreateBuffers, GLsizei n, GLuint* buffers) DECLARE_GL_FUNCTION_END_NO_RETURN(void, CreateBuffers, n, buffers)
DECLARE_GL_FUNCTION_HEAD(void, NamedBufferStorage, GLuint buffer, GLsizeiptr size, const void* data, GLbitfield flags) DECLARE_GL_FUNCTION_END_NO_RETURN(void, NamedBufferStorage, buffer, size, data, flags)
DECLARE_GL_FUNCTION_HEAD(void, NamedBufferData, GLuint buffer, GLsizeiptr size, const void* data, GLenum usage) DECLARE_GL_FUNCTION_END_NO_RETURN(void, NamedBufferData, buffer, size, data, usage)
@@ -1049,8 +1055,8 @@ DECLARE_GL_FUNCTION_HEAD(void, NamedRenderbufferStorage, GLuint renderbuffer, GL
DECLARE_GL_FUNCTION_HEAD(void, NamedRenderbufferStorageMultisample, GLuint renderbuffer, GLsizei samples, GLenum internalformat, GLsizei width, GLsizei height) DECLARE_GL_FUNCTION_END_NO_RETURN(void, NamedRenderbufferStorageMultisample, renderbuffer, samples, internalformat, width, height)
DECLARE_GL_FUNCTION_HEAD(void, GetNamedRenderbufferParameteriv, GLuint renderbuffer, GLenum pname, GLint* params) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetNamedRenderbufferParameteriv, renderbuffer, pname, params)
DECLARE_GL_FUNCTION_HEAD(void, CreateTextures, GLenum target, GLsizei n, GLuint* textures) DECLARE_GL_FUNCTION_END_NO_RETURN(void, CreateTextures, target, n, textures)
DECLARE_GL_FUNCTION_HEAD(void, TextureBuffer, GLuint texture, GLenum internalformat, GLuint buffer) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TextureBuffer, texture, internalformat, buffer)
DECLARE_GL_FUNCTION_HEAD(void, TextureBufferRange, GLuint texture, GLenum internalformat, GLuint buffer, GLintptr offset, GLsizeiptr size) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TextureBufferRange, texture, internalformat, buffer, offset, size)
DECLARE_GL_FUNCTION_STUB_HEAD(void, TextureBuffer, GLuint texture, GLenum internalformat, GLuint buffer) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, TextureBuffer, texture, internalformat, buffer)
DECLARE_GL_FUNCTION_STUB_HEAD(void, TextureBufferRange, GLuint texture, GLenum internalformat, GLuint buffer, GLintptr offset, GLsizeiptr size) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, TextureBufferRange, texture, internalformat, buffer, offset, size)
DECLARE_GL_FUNCTION_HEAD(void, TextureStorage1D, GLuint texture, GLsizei levels, GLenum internalformat, GLsizei width) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TextureStorage1D, texture, levels, internalformat, width)
DECLARE_GL_FUNCTION_HEAD(void, TextureStorage2D, GLuint texture, GLsizei levels, GLenum internalformat, GLsizei width, GLsizei height) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TextureStorage2D, texture, levels, internalformat, width, height)
DECLARE_GL_FUNCTION_HEAD(void, TextureStorage3D, GLuint texture, GLsizei levels, GLenum internalformat, GLsizei width, GLsizei height, GLsizei depth) DECLARE_GL_FUNCTION_END_NO_RETURN(void, TextureStorage3D, texture, levels, internalformat, width, height, depth)
@@ -1090,14 +1096,14 @@ DECLARE_GL_FUNCTION_HEAD(void, VertexArrayVertexBuffers, GLuint vaobj, GLuint fi
DECLARE_GL_FUNCTION_HEAD(void, VertexArrayAttribBinding, GLuint vaobj, GLuint attribindex, GLuint bindingindex) DECLARE_GL_FUNCTION_END_NO_RETURN(void, VertexArrayAttribBinding, vaobj, attribindex, bindingindex)
DECLARE_GL_FUNCTION_HEAD(void, VertexArrayAttribFormat, GLuint vaobj, GLuint attribindex, GLint size, GLenum type, GLboolean normalized, GLuint relativeoffset) DECLARE_GL_FUNCTION_END_NO_RETURN(void, VertexArrayAttribFormat, vaobj, attribindex, size, type, normalized, relativeoffset)
DECLARE_GL_FUNCTION_HEAD(void, VertexArrayAttribIFormat, GLuint vaobj, GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset) DECLARE_GL_FUNCTION_END_NO_RETURN(void, VertexArrayAttribIFormat, vaobj, attribindex, size, type, relativeoffset)
DECLARE_GL_FUNCTION_HEAD(void, VertexArrayAttribLFormat, GLuint vaobj, GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset) DECLARE_GL_FUNCTION_END_NO_RETURN(void, VertexArrayAttribLFormat, vaobj, attribindex, size, type, relativeoffset)
DECLARE_GL_FUNCTION_STUB_HEAD(void, VertexArrayAttribLFormat, GLuint vaobj, GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, VertexArrayAttribLFormat, vaobj, attribindex, size, type, relativeoffset)
DECLARE_GL_FUNCTION_HEAD(void, VertexArrayBindingDivisor, GLuint vaobj, GLuint bindingindex, GLuint divisor) DECLARE_GL_FUNCTION_END_NO_RETURN(void, VertexArrayBindingDivisor, vaobj, bindingindex, divisor)
DECLARE_GL_FUNCTION_HEAD(void, GetVertexArrayiv, GLuint vaobj, GLenum pname, GLint* param) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetVertexArrayiv, vaobj, pname, param)
DECLARE_GL_FUNCTION_HEAD(void, GetVertexArrayIndexediv, GLuint vaobj, GLuint index, GLenum pname, GLint* param) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetVertexArrayIndexediv, vaobj, index, pname, param)
DECLARE_GL_FUNCTION_HEAD(void, GetVertexArrayIndexed64iv, GLuint vaobj, GLuint index, GLenum pname, GLint64* param) DECLARE_GL_FUNCTION_END_NO_RETURN(void, GetVertexArrayIndexed64iv, vaobj, index, pname, param)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetVertexArrayiv, GLuint vaobj, GLenum pname, GLint* param) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetVertexArrayiv, vaobj, pname, param)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetVertexArrayIndexediv, GLuint vaobj, GLuint index, GLenum pname, GLint* param) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetVertexArrayIndexediv, vaobj, index, pname, param)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetVertexArrayIndexed64iv, GLuint vaobj, GLuint index, GLenum pname, GLint64* param) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetVertexArrayIndexed64iv, vaobj, index, pname, param)
DECLARE_GL_FUNCTION_HEAD(void, CreateSamplers, GLsizei n, GLuint* samplers) DECLARE_GL_FUNCTION_END_NO_RETURN(void, CreateSamplers, n, samplers)
DECLARE_GL_FUNCTION_STUB_HEAD(void, CreateProgramPipelines, GLsizei n, GLuint* pipelines) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, CreateProgramPipelines, n, pipelines)
DECLARE_GL_FUNCTION_HEAD(void, CreateQueries, GLenum target, GLsizei n, GLuint* ids) DECLARE_GL_FUNCTION_END_NO_RETURN(void, CreateQueries, target, n, ids)
DECLARE_GL_FUNCTION_STUB_HEAD(void, CreateQueries, GLenum target, GLsizei n, GLuint* ids) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, CreateQueries, target, n, ids)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetQueryBufferObjecti64v, GLuint id, GLuint buffer, GLenum pname, GLintptr offset) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetQueryBufferObjecti64v, id, buffer, pname, offset)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetQueryBufferObjectiv, GLuint id, GLuint buffer, GLenum pname, GLintptr offset) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetQueryBufferObjectiv, id, buffer, pname, offset)
DECLARE_GL_FUNCTION_STUB_HEAD(void, GetQueryBufferObjectui64v, GLuint id, GLuint buffer, GLenum pname, GLintptr offset) DECLARE_GL_FUNCTION_STUB_END_NO_RETURN(void, GetQueryBufferObjectui64v, id, buffer, pname, offset)
@@ -22,21 +22,9 @@
namespace MobileGL::MG_Impl::GLImpl {
namespace {
// GL only requires support for framebuffers whose depth and stencil attachments
// are the same image; anything else may be reported GL_FRAMEBUFFER_UNSUPPORTED.
// DirectVulkan cannot form two separate attachments at all, and the real ES
// drivers behind DirectGLES answer UNSUPPORTED for it too - so saying COMPLETE
// and then rendering into a framebuffer the driver refuses produced silently
// empty results (KHR-GL3x.packed_depth_stencil.verify_mixed_attachments).
Bool ActiveBackendRejectsDistinctDepthStencil() {
Bool IsActiveBackendDirectVulkan() {
auto* activeBackend = MG_Backend::pActiveBackendObject.get();
if (activeBackend == nullptr) {
return false;
}
if (activeBackend->GetBackendType() == BackendType::DirectVulkan) {
return true;
}
return !activeBackend->GetDynamicParameters().SupportsDistinctDepthStencilAttachments;
return activeBackend != nullptr && activeBackend->GetBackendType() == BackendType::DirectVulkan;
}
Bool HasDistinctCompleteDepthStencilTextureAttachments(
@@ -57,32 +45,10 @@ namespace MobileGL::MG_Impl::GLImpl {
depthAttachment.GetTextureLevel() != stencilAttachment.GetTextureLevel();
}
// Mirrors the renderer-side gate: distinct depth/stencil renderbuffers (or a
// renderbuffer paired with a texture) cannot form one Vulkan depth-stencil
// attachment, and GL permits reporting such framebuffers as UNSUPPORTED.
Bool HasDistinctCompleteDepthStencilRenderbufferAttachments(
const MG_State::GLState::FramebufferObject& framebufferObject) {
if (framebufferObject.GetExternalIndex() == 0) {
return false;
}
const auto& depthAttachment = framebufferObject.GetAttachment(FramebufferAttachmentType::Depth);
const auto& stencilAttachment = framebufferObject.GetAttachment(FramebufferAttachmentType::Stencil);
if (!depthAttachment.IsComplete() || !stencilAttachment.IsComplete()) {
return false;
}
if (depthAttachment.IsRenderbuffer() && stencilAttachment.IsRenderbuffer()) {
return depthAttachment.GetRenderbuffer().get() != stencilAttachment.GetRenderbuffer().get();
}
return (depthAttachment.IsRenderbuffer() || stencilAttachment.IsRenderbuffer()) &&
(depthAttachment.IsTexture() || stencilAttachment.IsTexture());
}
Bool HasUnsupportedDistinctDepthStencilAttachments(
Bool IsUnsupportedFramebufferForDirectVulkan(
const MG_State::GLState::FramebufferObject& framebufferObject) {
// TODO: Keep this in sync with DirectVulkan renderbuffer support as color renderbuffer rendering lands.
return HasDistinctCompleteDepthStencilTextureAttachments(framebufferObject) ||
HasDistinctCompleteDepthStencilRenderbufferAttachments(framebufferObject);
return HasDistinctCompleteDepthStencilTextureAttachments(framebufferObject);
}
Bool HasDefinedAttachment(const MG_State::GLState::FramebufferObject& framebufferObject) {
@@ -101,14 +67,7 @@ namespace MobileGL::MG_Impl::GLImpl {
// shared-exponent, SNORM, three-channel norm16/float32/sRGB and three-channel integer formats.
// Desktop GL treats those as texture-only too (not in the GL 3.3 required-renderable list), so
// reporting GL_FRAMEBUFFER_UNSUPPORTED for them is legal.
//
// `capabilityTargetIndex` is the row of the cache the attachment actually lives in;
// kFormatCapabilityTargetCount asks about the format in general. Asking per target matters
// because a capability recorded for one of them says nothing about the others: DirectGLES
// widens three-channel formats to four channels to keep them renderable as *multisample*
// storage, and a format that survives only through that substitution is still texture-only
// on every ordinary target.
Bool IsColorInternalFormatRenderable(TextureInternalFormat format, SizeT capabilityTargetIndex) {
Bool IsColorInternalFormatRenderable(TextureInternalFormat format) {
const SizeT formatIndex = static_cast<SizeT>(format);
if (MG_Backend::pActiveBackendObject && formatIndex < MG_Backend::kFormatCapabilityFormatCount) {
const auto& cache = MG_Backend::pActiveBackendObject->GetFormatCapabilities();
@@ -120,11 +79,8 @@ namespace MobileGL::MG_Impl::GLImpl {
MG_Backend::FormatCapability::Creatable);
}
if (cachePopulated) {
const Bool singleTarget = capabilityTargetIndex < MG_Backend::kFormatCapabilityTargetCount;
const SizeT firstTarget = singleTarget ? capabilityTargetIndex : 0;
const SizeT lastTarget =
singleTarget ? capabilityTargetIndex + 1 : MG_Backend::kFormatCapabilityTargetCount;
for (SizeT targetIndex = firstTarget; targetIndex < lastTarget; ++targetIndex) {
for (SizeT targetIndex = 0; targetIndex < MG_Backend::kFormatCapabilityTargetCount;
++targetIndex) {
if (MG_Backend::HasFormatCapability(cache.FullCaps[targetIndex][formatIndex],
MG_Backend::FormatCapability::FramebufferRenderable) ||
MG_Backend::HasFormatCapability(cache.CaveatCaps[targetIndex][formatIndex],
@@ -173,17 +129,12 @@ namespace MobileGL::MG_Impl::GLImpl {
const auto& attachment = attachments[i];
if (!attachment.IsValid()) continue;
TextureInternalFormat format = TextureInternalFormat::Unknown;
SizeT capabilityTargetIndex = MG_Backend::kFormatCapabilityTargetCount;
if (attachment.IsTexture() && attachment.GetTexture()) {
format = attachment.GetTexture()->GetFormat();
capabilityTargetIndex =
MG_Backend::GetFormatCapabilityTargetIndex(attachment.GetTexture()->GetTarget());
} else if (attachment.IsRenderbuffer() && attachment.GetRenderbuffer()) {
format = attachment.GetRenderbuffer()->GetInternalFormat();
capabilityTargetIndex = MG_Backend::GetRenderbufferFormatCapabilityTargetIndex();
}
if (format != TextureInternalFormat::Unknown &&
!IsColorInternalFormatRenderable(format, capabilityTargetIndex)) {
if (format != TextureInternalFormat::Unknown && !IsColorInternalFormatRenderable(format)) {
return true;
}
}
@@ -196,148 +147,6 @@ namespace MobileGL::MG_Impl::GLImpl {
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", functionName, detail));
}
GLint ClassifyAttachmentComponentType(TextureInternalFormat internalFormat,
FramebufferAttachmentType attachmentType) {
// A stencil value is an unsigned integer index regardless of the depth half
// of a packed format.
if (attachmentType == FramebufferAttachmentType::Stencil) return GL_UNSIGNED_INT;
switch (internalFormat) {
case TextureInternalFormat::R16F:
case TextureInternalFormat::RG16F:
case TextureInternalFormat::RGB16F:
case TextureInternalFormat::RGBA16F:
case TextureInternalFormat::R32F:
case TextureInternalFormat::RG32F:
case TextureInternalFormat::RGB32F:
case TextureInternalFormat::RGBA32F:
case TextureInternalFormat::R11FG11FB10F:
case TextureInternalFormat::RGB9E5:
case TextureInternalFormat::DepthComponent32F:
case TextureInternalFormat::Depth32FStencil8:
return GL_FLOAT;
case TextureInternalFormat::R8I:
case TextureInternalFormat::R16I:
case TextureInternalFormat::R32I:
case TextureInternalFormat::RG8I:
case TextureInternalFormat::RG16I:
case TextureInternalFormat::RG32I:
case TextureInternalFormat::RGB8I:
case TextureInternalFormat::RGB16I:
case TextureInternalFormat::RGB32I:
case TextureInternalFormat::RGBA8I:
case TextureInternalFormat::RGBA16I:
case TextureInternalFormat::RGBA32I:
return GL_INT;
case TextureInternalFormat::R8UI:
case TextureInternalFormat::R16UI:
case TextureInternalFormat::R32UI:
case TextureInternalFormat::RG8UI:
case TextureInternalFormat::RG16UI:
case TextureInternalFormat::RG32UI:
case TextureInternalFormat::RGB8UI:
case TextureInternalFormat::RGB16UI:
case TextureInternalFormat::RGB32UI:
case TextureInternalFormat::RGBA8UI:
case TextureInternalFormat::RGBA16UI:
case TextureInternalFormat::RGBA32UI:
case TextureInternalFormat::RGB10A2UI:
return GL_UNSIGNED_INT;
case TextureInternalFormat::R8Snorm:
case TextureInternalFormat::R16Snorm:
case TextureInternalFormat::RG8Snorm:
case TextureInternalFormat::RG16Snorm:
case TextureInternalFormat::RGB8Snorm:
case TextureInternalFormat::RGB16Snorm:
case TextureInternalFormat::RGBA8Snorm:
case TextureInternalFormat::RGBA16Snorm:
return GL_SIGNED_NORMALIZED;
case TextureInternalFormat::Unknown:
return GL_NONE;
default:
return GL_UNSIGNED_NORMALIZED;
}
}
// Handles the format-derived pnames shared by GetFramebufferAttachmentParameteriv
// and its DSA variant. Returns true when pname was one of them.
Bool TryAnswerAttachmentFormatQuery(const MG_State::GLState::FramebufferAttachmentObject* attachmentObject,
FramebufferAttachmentType attachmentType, Bool depthStencilAlias,
GLenum pname, GLint* params, const char* caller) {
switch (pname) {
case GL_FRAMEBUFFER_ATTACHMENT_RED_SIZE:
case GL_FRAMEBUFFER_ATTACHMENT_GREEN_SIZE:
case GL_FRAMEBUFFER_ATTACHMENT_BLUE_SIZE:
case GL_FRAMEBUFFER_ATTACHMENT_ALPHA_SIZE:
case GL_FRAMEBUFFER_ATTACHMENT_DEPTH_SIZE:
case GL_FRAMEBUFFER_ATTACHMENT_STENCIL_SIZE:
case GL_FRAMEBUFFER_ATTACHMENT_COMPONENT_TYPE:
case GL_FRAMEBUFFER_ATTACHMENT_COLOR_ENCODING:
break;
default:
return false;
}
if (pname == GL_FRAMEBUFFER_ATTACHMENT_COMPONENT_TYPE && depthStencilAlias) {
// The depth and stencil components have different types, so the combined
// attachment name has no single answer.
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", caller,
"GL_FRAMEBUFFER_ATTACHMENT_COMPONENT_TYPE cannot be queried on "
"GL_DEPTH_STENCIL_ATTACHMENT."));
return true;
}
if (attachmentObject == nullptr || attachmentObject->IsEmpty() || !attachmentObject->IsValid()) {
// With OBJECT_TYPE == GL_NONE only OBJECT_TYPE and OBJECT_NAME may be queried.
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller,
"No image is attached to the queried attachment point."));
return true;
}
TextureInternalFormat internalFormat = TextureInternalFormat::Unknown;
if (attachmentObject->IsTexture() && attachmentObject->GetTexture()) {
internalFormat = attachmentObject->GetTexture()->GetFormat();
} else if (attachmentObject->IsRenderbuffer() && attachmentObject->GetRenderbuffer()) {
internalFormat = attachmentObject->GetRenderbuffer()->GetInternalFormat();
}
const auto sizes = MG_Util::GetComponentSizesForInternalFormat(internalFormat);
switch (pname) {
case GL_FRAMEBUFFER_ATTACHMENT_RED_SIZE:
*params = sizes.Red;
break;
case GL_FRAMEBUFFER_ATTACHMENT_GREEN_SIZE:
*params = sizes.Green;
break;
case GL_FRAMEBUFFER_ATTACHMENT_BLUE_SIZE:
*params = sizes.Blue;
break;
case GL_FRAMEBUFFER_ATTACHMENT_ALPHA_SIZE:
*params = sizes.Alpha;
break;
case GL_FRAMEBUFFER_ATTACHMENT_DEPTH_SIZE:
*params = sizes.Depth;
break;
case GL_FRAMEBUFFER_ATTACHMENT_STENCIL_SIZE:
*params = sizes.Stencil;
break;
case GL_FRAMEBUFFER_ATTACHMENT_COLOR_ENCODING:
*params = (internalFormat == TextureInternalFormat::SRGB8 ||
internalFormat == TextureInternalFormat::SRGB8Alpha8)
? GL_SRGB
: GL_LINEAR;
break;
case GL_FRAMEBUFFER_ATTACHMENT_COMPONENT_TYPE:
*params = ClassifyAttachmentComponentType(internalFormat, attachmentType);
break;
}
return true;
}
Bool ResolveRepresentableFramebufferTextureUploadTarget(const MG_State::GLState::ITextureObject& textureObject,
TextureUploadTarget& outUploadTarget,
Bool& outLayered) {
@@ -603,27 +412,6 @@ namespace MobileGL::MG_Impl::GLImpl {
FramebufferTarget framebufferTarget = MG_Util::ConvertGLEnumToFramebufferTarget(target);
if (!FramebufferImpl::ValidateFramebufferTarget(framebufferTarget)) return;
// Default-framebuffer attachment names (GL_DEPTH, GL_STENCIL, GL_FRONT/GL_BACK
// variants) alias onto the equivalent attachment points.
switch (attachment) {
case GL_DEPTH:
attachment = GL_DEPTH_ATTACHMENT;
break;
case GL_STENCIL:
attachment = GL_STENCIL_ATTACHMENT;
break;
case GL_FRONT:
case GL_FRONT_LEFT:
case GL_FRONT_RIGHT:
case GL_BACK:
case GL_BACK_LEFT:
case GL_BACK_RIGHT:
attachment = GL_COLOR_ATTACHMENT0;
break;
default:
break;
}
const Bool depthStencilAlias = attachment == GL_DEPTH_STENCIL_ATTACHMENT;
FramebufferAttachmentType attachmentType = depthStencilAlias
? FramebufferAttachmentType::Depth
@@ -640,40 +428,20 @@ namespace MobileGL::MG_Impl::GLImpl {
return;
}
Bool depthStencilMismatch = false;
const auto* attachmentObject = [&]() -> const MG_State::GLState::FramebufferAttachmentObject* {
if (!depthStencilAlias) {
return &framebufferObject->GetAttachment(attachmentType);
}
const auto& depthAttachment = framebufferObject->GetAttachment(FramebufferAttachmentType::Depth);
if (depthAttachment.IsValid() && !depthAttachment.IsEmpty()) return &depthAttachment;
const auto& stencilAttachment = framebufferObject->GetAttachment(FramebufferAttachmentType::Stencil);
const Bool depthLive = depthAttachment.IsValid() && !depthAttachment.IsEmpty();
const Bool stencilLive = stencilAttachment.IsValid() && !stencilAttachment.IsEmpty();
if (depthLive && stencilLive) {
const Bool sameObject = depthAttachment.IsTexture() == stencilAttachment.IsTexture() &&
(!depthAttachment.IsTexture() || depthAttachment.GetTexture() == stencilAttachment.GetTexture()) &&
(!depthAttachment.IsRenderbuffer() ||
depthAttachment.GetRenderbuffer() == stencilAttachment.GetRenderbuffer());
depthStencilMismatch = !sameObject;
} else {
// GL_DEPTH_STENCIL_ATTACHMENT means "both halves"; a lone half does not answer it.
depthStencilMismatch = depthLive != stencilLive;
}
if (depthLive) return &depthAttachment;
if (stencilLive) return &stencilAttachment;
if (stencilAttachment.IsValid() && !stencilAttachment.IsEmpty()) return &stencilAttachment;
return nullptr;
}();
if (depthStencilMismatch) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", "GetFramebufferAttachmentParameteriv_State",
"GL_DEPTH_STENCIL_ATTACHMENT query with different depth and stencil "
"attachment images."));
return;
}
switch (pname) {
case GL_FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE:
if (attachmentObject == nullptr || attachmentObject->IsEmpty() || !attachmentObject->IsValid()) {
@@ -737,10 +505,6 @@ namespace MobileGL::MG_Impl::GLImpl {
: GL_FALSE;
break;
default:
if (TryAnswerAttachmentFormatQuery(attachmentObject, attachmentType, depthStencilAlias, pname, params,
"GetFramebufferAttachmentParameteriv_State")) {
return;
}
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>(
@@ -1143,7 +907,6 @@ namespace MobileGL::MG_Impl::GLImpl {
const FramebufferAttachmentType attachmentType = MG_Util::ConvertGLEnumToFramebufferAttachmentType(attachment);
if (!FramebufferImpl::ValidateFramebufferAttachmentType(attachmentType)) return;
if (!FramebufferImpl::ValidateColorAttachmentInRange(attachmentType, functionName)) return;
if (!TextureImpl::ValidateTextureName(texture, true)) return;
if (texture == 0) {
@@ -1240,7 +1003,6 @@ namespace MobileGL::MG_Impl::GLImpl {
FramebufferTarget framebufferTarget = MG_Util::ConvertGLEnumToFramebufferTarget(target);
RenderbufferTarget rbTarget = MG_Util::ConvertGLEnumToRenderbufferTarget(renderbuffertarget);
if (!FramebufferImpl::ValidateFramebufferAttachmentType(attachmentType)) return;
if (!FramebufferImpl::ValidateColorAttachmentInRange(attachmentType, "FramebufferRenderbuffer_State")) return;
if (!FramebufferImpl::ValidateFramebufferTarget(framebufferTarget)) return;
if (!FramebufferImpl::ValidateRenderbufferTarget(rbTarget)) return;
auto& bindingSlot = MG_State::pGLContext->GetFramebufferBindingSlot(framebufferTarget);
@@ -1285,8 +1047,6 @@ namespace MobileGL::MG_Impl::GLImpl {
FramebufferAttachmentType attachmentType = MG_Util::ConvertGLEnumToFramebufferAttachmentType(attachment);
RenderbufferTarget rbTarget = MG_Util::ConvertGLEnumToRenderbufferTarget(renderbuffertarget);
if (!FramebufferImpl::ValidateFramebufferAttachmentType(attachmentType)) return;
if (!FramebufferImpl::ValidateColorAttachmentInRange(attachmentType, "NamedFramebufferRenderbuffer_State"))
return;
if (!FramebufferImpl::ValidateRenderbufferTarget(rbTarget)) return;
if (renderbuffer == 0) {
@@ -1662,8 +1422,8 @@ namespace MobileGL::MG_Impl::GLImpl {
if (HasNonRenderableColorAttachment(*framebufferObject)) {
return GL_FRAMEBUFFER_UNSUPPORTED;
}
if (ActiveBackendRejectsDistinctDepthStencil() &&
HasUnsupportedDistinctDepthStencilAttachments(*framebufferObject)) {
if (IsActiveBackendDirectVulkan() &&
IsUnsupportedFramebufferForDirectVulkan(*framebufferObject)) {
return GL_FRAMEBUFFER_UNSUPPORTED;
}
return GL_FRAMEBUFFER_COMPLETE;
@@ -1690,8 +1450,8 @@ namespace MobileGL::MG_Impl::GLImpl {
if (HasNonRenderableColorAttachment(*framebufferObject)) {
return GL_FRAMEBUFFER_UNSUPPORTED;
}
if (ActiveBackendRejectsDistinctDepthStencil() &&
HasUnsupportedDistinctDepthStencilAttachments(*framebufferObject)) {
if (IsActiveBackendDirectVulkan() &&
IsUnsupportedFramebufferForDirectVulkan(*framebufferObject)) {
return GL_FRAMEBUFFER_UNSUPPORTED;
}
return GL_FRAMEBUFFER_COMPLETE;
@@ -1708,40 +1468,20 @@ namespace MobileGL::MG_Impl::GLImpl {
: MG_Util::ConvertGLEnumToFramebufferAttachmentType(attachment);
if (!FramebufferImpl::ValidateFramebufferAttachmentType(attachmentType)) return;
Bool depthStencilMismatch = false;
const auto* attachmentObject = [&]() -> const MG_State::GLState::FramebufferAttachmentObject* {
if (!depthStencilAlias) {
return &framebufferObject->GetAttachment(attachmentType);
}
const auto& depthAttachment = framebufferObject->GetAttachment(FramebufferAttachmentType::Depth);
if (depthAttachment.IsValid() && !depthAttachment.IsEmpty()) return &depthAttachment;
const auto& stencilAttachment = framebufferObject->GetAttachment(FramebufferAttachmentType::Stencil);
const Bool depthLive = depthAttachment.IsValid() && !depthAttachment.IsEmpty();
const Bool stencilLive = stencilAttachment.IsValid() && !stencilAttachment.IsEmpty();
if (depthLive && stencilLive) {
const Bool sameObject = depthAttachment.IsTexture() == stencilAttachment.IsTexture() &&
(!depthAttachment.IsTexture() || depthAttachment.GetTexture() == stencilAttachment.GetTexture()) &&
(!depthAttachment.IsRenderbuffer() ||
depthAttachment.GetRenderbuffer() == stencilAttachment.GetRenderbuffer());
depthStencilMismatch = !sameObject;
} else {
// GL_DEPTH_STENCIL_ATTACHMENT means "both halves"; a lone half does not answer it.
depthStencilMismatch = depthLive != stencilLive;
}
if (depthLive) return &depthAttachment;
if (stencilLive) return &stencilAttachment;
if (stencilAttachment.IsValid() && !stencilAttachment.IsEmpty()) return &stencilAttachment;
return nullptr;
}();
if (depthStencilMismatch) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller,
"GL_DEPTH_STENCIL_ATTACHMENT query with different depth and stencil "
"attachment images."));
return;
}
switch (pname) {
case GL_FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE:
if (attachmentObject == nullptr || attachmentObject->IsEmpty() || !attachmentObject->IsValid()) {
@@ -1787,10 +1527,6 @@ namespace MobileGL::MG_Impl::GLImpl {
: GL_FALSE;
break;
default:
if (TryAnswerAttachmentFormatQuery(attachmentObject, attachmentType, depthStencilAlias, pname, params,
caller)) {
return;
}
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>(
@@ -2203,56 +1939,6 @@ namespace MobileGL::MG_Impl::GLImpl {
ReadPixels_Backend(x, y, width, height, format, type, pixels);
}
// Bytes glReadPixels would write for this rectangle under the current GL_PACK_* state
// (GL 4.6 core 18.2.8): rows are padded to GL_PACK_ALIGNMENT and laid out GL_PACK_ROW_LENGTH
// wide, and the skip parameters offset the first texel. The last row is not padded - nothing
// follows it to align - which is what makes a tightly-sized destination legal.
static SizeT ComputePackedReadSizeInBytes(GLsizei width, GLsizei height, GLenum format, GLenum type) {
const SizeT bytesPerPixel =
MG_Util::GetInputBytesPerPixel(MG_Util::ConvertGLEnumToTextureInputFormat(format),
MG_Util::ConvertGLEnumToTexturePixelDataType(type));
if (bytesPerPixel == 0 || width <= 0 || height <= 0) return 0;
const auto packParam = [](PixelStoreParam param) {
return static_cast<SizeT>(std::max(0, MG_State::pGLContext->GetPixelStoreParam(param)));
};
const SizeT rowLengthInPixels =
packParam(PixelStoreParam::PackRowLength) != 0
? packParam(PixelStoreParam::PackRowLength)
: static_cast<SizeT>(width);
const SizeT alignment = std::max<SizeT>(1, packParam(PixelStoreParam::PackAlignment));
const SizeT unalignedRowBytes = rowLengthInPixels * bytesPerPixel;
const SizeT paddedRowBytes = ((unalignedRowBytes + alignment - 1) / alignment) * alignment;
const SizeT skipBytes = packParam(PixelStoreParam::PackSkipRows) * paddedRowBytes +
packParam(PixelStoreParam::PackSkipPixels) * bytesPerPixel;
return skipBytes + paddedRowBytes * (static_cast<SizeT>(height) - 1) +
static_cast<SizeT>(width) * bytesPerPixel;
}
// glReadnPixels is glReadPixels with a bound on how much it may write (GL 4.6 core 18.2.8,
// originally GL_ARB_robustness). It is identical in every other respect, so it validates and
// reads through exactly the same path once the destination is known to be big enough.
void ReadnPixels(GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, GLsizei bufSize,
void* data) {
if (bufSize < 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "bufSize must be non-negative."));
return;
}
if (!ReadPixels_State(x, y, width, height, format, type, data)) return;
if (ComputePackedReadSizeInBytes(width, height, format, type) > static_cast<SizeT>(bufSize)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"the data required for this read does not fit in bufSize."));
return;
}
ReadPixels_Backend(x, y, width, height, format, type, data);
}
void ClearBufferfi(GLenum buffer, GLint drawbuffer, GLfloat depth, GLint stencil) {
if (!ValidateClearBufferfi_State(buffer, drawbuffer)) return;
ClearBufferfi_Backend(buffer, drawbuffer, depth, stencil);
@@ -14,8 +14,6 @@
namespace MobileGL::MG_Impl::GLImpl {
/* @INSERTION_POINT:FUNCTION_DECLARATION@ */
void ReadPixels(GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, void* pixels);
void ReadnPixels(GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, GLsizei bufSize,
void* data);
void ClearBufferfi(GLenum buffer, GLint drawbuffer, GLfloat depth, GLint stencil);
void ClearBufferfv(GLenum buffer, GLint drawbuffer, const GLfloat* value);
void ClearBufferuiv(GLenum buffer, GLint drawbuffer, const GLuint* value);
@@ -7,7 +7,6 @@
// End of Source File Header
#include "Validators.h"
#include <MG_Backend/BackendObjects.h>
#include <MG_State/GLState/Core.h>
#include <MG_State/GLState/ErrorState/Error.h>
#include <MG_Util/Converters/GLToStr/GLEnumConverter.h>
@@ -61,26 +60,6 @@ namespace MobileGL::MG_Impl::GLImpl::FramebufferImpl {
return true;
}
Bool ValidateColorAttachmentInRange(FramebufferAttachmentType attachment, const char* caller) {
const auto first = static_cast<SizeT>(FramebufferAttachmentType::Color0);
const auto index = static_cast<SizeT>(attachment);
if (index < first) return true;
const auto colorIndex = index - first;
const auto limit = static_cast<SizeT>(
MG_Backend::pActiveBackendObject ? MG_Backend::pActiveBackendObject->GetDynamicParameters()
.MaxColorAttachments
: static_cast<Int>(MG_State::GLState::FramebufferObject::MAX_DRAW_BUFFERS));
if (colorIndex >= limit) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl/FramebufferImpl", caller,
std::format("Colour attachment {} is beyond GL_MAX_COLOR_ATTACHMENTS ({}).", colorIndex, limit)));
return false;
}
return true;
}
Bool ValidateRenderbufferTarget(RenderbufferTarget target) {
if (target == RenderbufferTarget::Unknown) {
using namespace MG_Util;
@@ -14,10 +14,6 @@ namespace MobileGL::MG_Impl::GLImpl::FramebufferImpl {
Bool ValidateFramebufferTarget(FramebufferTarget target);
Bool ValidateFramebufferName(Uint index, Bool allowZero = true);
Bool ValidateFramebufferAttachmentType(FramebufferAttachmentType attachment);
// GL_COLOR_ATTACHMENTn is a token per n up to 31, but only the first GL_MAX_COLOR_ATTACHMENTS of
// them name an attachment point of a framebuffer object; the rest are INVALID_OPERATION for the
// attaching entry points (GL 4.6 core 9.2.7). Non-colour attachments pass through unchanged.
Bool ValidateColorAttachmentInRange(FramebufferAttachmentType attachment, const char* caller);
Bool ValidateRenderbufferTarget(RenderbufferTarget target);
Bool ValidateRenderbufferName(Uint index, Bool allowZero = true);
} // namespace MobileGL::MG_Impl::GLImpl::FramebufferImpl
+10 -120
View File
@@ -213,13 +213,8 @@ namespace MobileGL::MG_Impl::GLImpl {
GLint maxSamples = 0;
for (const auto& attachment : drawFbo->GetAllAttachmentObjects()) {
if (attachment.IsRenderbuffer() && attachment.GetRenderbuffer()) {
maxSamples = std::max(maxSamples, static_cast<GLint>(attachment.GetRenderbuffer()->GetSamples()));
} else if (attachment.IsTexture() && attachment.GetTexture()) {
// Multisample texture attachments count too (GL_SAMPLE_BUFFERS must
// report 1 for any multisampled draw framebuffer).
maxSamples = std::max(maxSamples, static_cast<GLint>(attachment.GetTexture()->GetSamples()));
}
if (!attachment.IsRenderbuffer() || !attachment.GetRenderbuffer()) continue;
maxSamples = std::max(maxSamples, static_cast<GLint>(attachment.GetRenderbuffer()->GetSamples()));
}
return maxSamples;
}
@@ -470,14 +465,6 @@ namespace MobileGL::MG_Impl::GLImpl {
case GL_STENCIL_TEST:
*params = MG_State::pGLContext->IsCapabilityEnabled(CapabilityInput::StencilTest) ? GL_TRUE : GL_FALSE;
return;
case GL_MIN_FRAGMENT_INTERPOLATION_OFFSET:
case GL_MAX_FRAGMENT_INTERPOLATION_OFFSET:
case GL_FRAGMENT_INTERPOLATION_OFFSET_BITS: {
GLfloat value = 0.0f;
GetFloatv(pname, &value);
*params = value != 0.0f ? GL_TRUE : GL_FALSE;
return;
}
default:
break;
}
@@ -533,19 +520,6 @@ namespace MobileGL::MG_Impl::GLImpl {
params[1] = dynamicParameters.ViewportBoundsRangeMax;
return;
}
case GL_MIN_FRAGMENT_INTERPOLATION_OFFSET:
case GL_MAX_FRAGMENT_INTERPOLATION_OFFSET:
case GL_FRAGMENT_INTERPOLATION_OFFSET_BITS: {
const auto& dynamicParameters = MG_Backend::pActiveBackendObject->GetDynamicParameters();
if (pname == GL_MIN_FRAGMENT_INTERPOLATION_OFFSET) {
params[0] = dynamicParameters.MinFragmentInterpolationOffset;
} else if (pname == GL_MAX_FRAGMENT_INTERPOLATION_OFFSET) {
params[0] = dynamicParameters.MaxFragmentInterpolationOffset;
} else {
params[0] = static_cast<GLfloat>(dynamicParameters.FragmentInterpolationOffsetBits);
}
return;
}
case GL_DEPTH_CLEAR_VALUE:
params[0] = MG_State::pGLContext->GetClearDepth();
return;
@@ -671,40 +645,6 @@ namespace MobileGL::MG_Impl::GLImpl {
}
switch (target) {
// The vertex buffer binding points of the vertex array object that is bound. Indexed by
// binding point, not by attribute (GL 4.6 core 10.3.1).
case GL_VERTEX_BINDING_BUFFER:
case GL_VERTEX_BINDING_DIVISOR:
case GL_VERTEX_BINDING_OFFSET:
case GL_VERTEX_BINDING_STRIDE: {
if (index >= VertexArrayImpl::GetMaxVertexAttribBindings()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"Vertex buffer binding index is out of range."));
return;
}
const auto& vao = MG_State::pGLContext->GetBoundVertexArray();
if (!vao) {
*data = 0;
return;
}
const auto& binding = vao->GetBindingPoint(index);
switch (target) {
case GL_VERTEX_BINDING_BUFFER:
*data = binding.Buffer ? static_cast<GLint>(binding.Buffer->GetExternalIndex()) : 0;
return;
case GL_VERTEX_BINDING_DIVISOR:
*data = static_cast<GLint>(binding.Divisor);
return;
case GL_VERTEX_BINDING_OFFSET:
*data = static_cast<GLint>(binding.Offset);
return;
default:
*data = static_cast<GLint>(binding.Stride);
return;
}
}
case GL_IMAGE_BINDING_NAME:
case GL_IMAGE_BINDING_LEVEL:
case GL_IMAGE_BINDING_LAYERED:
@@ -1064,11 +1004,6 @@ namespace MobileGL::MG_Impl::GLImpl {
*params = obj ? static_cast<GLint>(obj->GetExternalIndex()) : 0;
return;
}
case GL_DRAW_INDIRECT_BUFFER_BINDING: {
auto& obj = MG_State::pGLContext->GetBufferBindingSlot(BufferTarget::DrawIndirect).GetBoundObject();
*params = obj ? static_cast<GLint>(obj->GetExternalIndex()) : 0;
return;
}
case GL_MAX_DEBUG_GROUP_STACK_DEPTH:
*params = 0; // debug-group entrypoints are stubbed
return;
@@ -1703,7 +1638,7 @@ namespace MobileGL::MG_Impl::GLImpl {
*params = static_cast<GLint>(MG_State::pGLContext->GetHint(pname));
return;
case GL_TEXTURE_BUFFER_OFFSET_ALIGNMENT:
*params = MG_Backend::pActiveBackendObject->GetDynamicParameters().TextureBufferOffsetAlignment;
*params = 0; // texture-buffer range entrypoints are stubbed
return;
case GL_TIMESTAMP: {
Int64 timestamp = 0;
@@ -1772,22 +1707,20 @@ namespace MobileGL::MG_Impl::GLImpl {
*params = vao ? static_cast<GLint>(vao->GetExternalIndex()) : 0;
return;
}
// The vertex buffer binding points are per-binding-index state, so the non-indexed getter
// has nothing to answer with (GL 4.6 core table 23.4).
case GL_VERTEX_BINDING_BUFFER:
case GL_VERTEX_BINDING_DIVISOR:
*params = 0; // vertex-binding entrypoints are stubbed
return;
case GL_VERTEX_BINDING_OFFSET:
*params = 0; // vertex-binding entrypoints are stubbed
return;
case GL_VERTEX_BINDING_STRIDE:
RecordIndexedOnlyGetterError(__func__, pname);
*params = 0; // vertex-binding entrypoints are stubbed
return;
case GL_MAX_VERTEX_ATTRIB_RELATIVE_OFFSET:
*params = static_cast<GLint>(VertexArrayImpl::GetMaxVertexAttribRelativeOffset());
*params = 0; // vertex-binding entrypoints are stubbed
return;
case GL_MAX_VERTEX_ATTRIB_BINDINGS:
*params = static_cast<GLint>(VertexArrayImpl::GetMaxVertexAttribBindings());
return;
case GL_MAX_VERTEX_ATTRIB_STRIDE:
*params = static_cast<GLint>(VertexArrayImpl::GetMaxVertexAttribStride());
*params = 0; // vertex-binding entrypoints are stubbed
return;
case GL_VIEWPORT: {
const auto& vp = MG_State::pGLContext->GetViewport();
@@ -1953,21 +1886,6 @@ namespace MobileGL::MG_Impl::GLImpl {
case GL_MAX_SAMPLE_MASK_WORDS:
*params = dynamicParameters.MaxSampleMaskWords;
break;
case GL_PATCH_VERTICES:
*params = static_cast<GLint>(MG_State::pGLContext->GetPatchVertices());
break;
case GL_MAX_PATCH_VERTICES:
*params = dynamicParameters.MaxPatchVertices;
break;
case GL_MAX_TESS_GEN_LEVEL:
*params = dynamicParameters.MaxTessGenLevel;
break;
case GL_MIN_PROGRAM_TEXTURE_GATHER_OFFSET:
*params = dynamicParameters.MinProgramTextureGatherOffset;
break;
case GL_MAX_PROGRAM_TEXTURE_GATHER_OFFSET:
*params = dynamicParameters.MaxProgramTextureGatherOffset;
break;
case GL_MAX_SHADER_STORAGE_BUFFER_BINDINGS:
*params = static_cast<GLint>(GetIndexedBufferQueryPointCount(BufferTarget::ShaderStorage));
break;
@@ -1983,25 +1901,6 @@ namespace MobileGL::MG_Impl::GLImpl {
case GL_MAX_TRANSFORM_FEEDBACK_SEPARATE_COMPONENTS:
*params = kFrontendMaxTransformFeedbackSeparateComponents;
break;
// ARB_transform_feedback3 limits. The GL CTS queries these before checking
// whether the extension is advertised and requires no GL error; desktop
// drivers all accept them, so answer with the separate-attrib capacity and
// the single vertex stream the backends provide.
case GL_MAX_TRANSFORM_FEEDBACK_BUFFERS:
*params = kFrontendMaxTransformFeedbackSeparateAttribs;
break;
case GL_MAX_VERTEX_STREAMS:
*params = 1;
break;
case GL_TRANSFORM_FEEDBACK_ACTIVE:
*params = MG_State::pGLContext->IsTransformFeedbackActive() ? 1 : 0;
break;
case GL_TRANSFORM_FEEDBACK_PAUSED:
*params = MG_State::pGLContext->IsTransformFeedbackPaused() ? 1 : 0;
break;
case GL_TRANSFORM_FEEDBACK_BINDING:
*params = static_cast<GLint>(MG_State::pGLContext->GetBoundTransformFeedbackName());
break;
case GL_MAX_TEXTURE_IMAGE_UNITS:
*params = dynamicParameters.MaxTextureImageUnits;
break;
@@ -2058,15 +1957,6 @@ namespace MobileGL::MG_Impl::GLImpl {
case GL_SUBPIXEL_BITS:
*params = std::max(dynamicParameters.ViewportSubpixelBits, kFrontendSubpixelBits);
break;
case GL_MIN_FRAGMENT_INTERPOLATION_OFFSET:
*params = static_cast<GLint>(std::lround(dynamicParameters.MinFragmentInterpolationOffset));
break;
case GL_MAX_FRAGMENT_INTERPOLATION_OFFSET:
*params = static_cast<GLint>(std::lround(dynamicParameters.MaxFragmentInterpolationOffset));
break;
case GL_FRAGMENT_INTERPOLATION_OFFSET_BITS:
*params = dynamicParameters.FragmentInterpolationOffsetBits;
break;
case GL_UNIFORM_BUFFER_OFFSET_ALIGNMENT:
*params = static_cast<Int>(dynamicParameters.UniformBufferOffsetAlignment);
break;
+14 -681
View File
@@ -8,8 +8,6 @@
#include "GL_Program.h"
#include "Config.h"
#include <cmath>
#include <limits>
#include <MG_Impl/GLImpl/VertexArray/Validators.h>
#include <MG_State/GLState/Core.h>
#include <MG_Util/Converters/GLToStr/GLEnumConverter.h>
@@ -51,18 +49,10 @@ namespace MobileGL::MG_Impl::GLImpl {
static bool CheckProgramNameValidity(GLuint program) {
if (!MG_State::pGLContext->ValidateProgramName(program)) {
// Programs and shaders share one name space: a name that exists but
// belongs to a shader is INVALID_OPERATION, a name GL never handed
// out is INVALID_VALUE (GL 3.3 core 2.11.x).
const ErrorCode error = MG_State::pGLContext->ValidateShaderName(program)
? ErrorCode::InvalidOperation
: ErrorCode::InvalidValue;
MG_State::pGLContext->RecordError(
error,
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
std::to_string(program) +
(error == ErrorCode::InvalidOperation ? " is not a program object."
: " is not a valid name.")));
std::to_string(program) + " is not a valid name."));
return false;
}
return true;
@@ -205,50 +195,6 @@ namespace MobileGL::MG_Impl::GLImpl {
}
}
// The GL_UNIFORM interface and glGetActiveUniform(s)iv are the same query in two
// spellings, so they answer from the same place - the frontend reflection. The backend
// program is not that place: it does not exist at all for a program whose types its
// shading language cannot express (a double-precision uniform has no ESSL form), and
// the interface queries would then describe a program with no uniforms.
//
// Writes the GL_UNIFORM value of `prop` for active uniform `index`; false for a prop
// the reflection does not model, which the caller forwards to the backend instead.
Bool GetUniformResourceProp(const SharedPtr<MG_State::GLState::ProgramObject>& programObject, Uint index,
GLenum prop, GLint* out) {
switch (prop) {
case GL_TYPE:
*out = static_cast<GLint>(programObject->GetActiveUniformType(index));
return true;
case GL_ARRAY_SIZE:
*out = programObject->GetActiveUniformArraySize(index);
return true;
case GL_NAME_LENGTH:
*out = static_cast<GLint>(programObject->GetActiveUniformName(index).length() + 1);
return true;
case GL_BLOCK_INDEX:
*out = programObject->GetActiveUniformBlockIndex(index);
return true;
case GL_OFFSET:
*out = programObject->GetActiveUniformOffset(index);
return true;
case GL_ARRAY_STRIDE:
*out = programObject->GetActiveUniformArrayStride(index);
return true;
case GL_MATRIX_STRIDE:
*out = programObject->GetActiveUniformMatrixStride(index);
return true;
case GL_IS_ROW_MAJOR:
*out = programObject->GetActiveUniformIsRowMajor(index);
return true;
case GL_LOCATION:
// A block member has no location; GetUniformLocation already reports -1 for one.
*out = programObject->GetUniformLocation(programObject->GetActiveUniformName(index));
return true;
default:
return false;
}
}
void CopyStr(GLsizei bufSize, GLsizei* length, GLchar* dst, const char* src, GLsizei srcLength) {
if (bufSize <= 0) {
if (length) *length = 0;
@@ -371,16 +317,11 @@ namespace MobileGL::MG_Impl::GLImpl {
}
void DeleteProgram_State(GLuint program) {
// "If program is zero, it is silently ignored" (GL 4.6 core 7.3) - unlike every
// other program entry point, where 0 is a name GL never handed out.
if (program == 0) return;
if (!CheckProgramNameValidity(program)) return;
MG_State::pGLContext->MarkProgramForDeletion(program);
}
void DeleteShader_State(GLuint shader) {
// Same silent-zero rule as glDeleteProgram (GL 4.6 core 7.1).
if (shader == 0) return;
if (!CheckShaderNameValidity(shader)) return;
MG_State::pGLContext->MarkShaderForDeletion(shader);
}
@@ -664,18 +605,6 @@ namespace MobileGL::MG_Impl::GLImpl {
*params = programObject->GetActiveUniformBlocksMaxNameLength() + 1;
MGLOG_D("%s: %s = %d", __func__, MG_Util::ConvertGLEnumToString(pname).c_str(), *params);
break;
case GL_TRANSFORM_FEEDBACK_VARYINGS:
*params = static_cast<GLint>(programObject->GetTransformFeedbackVaryingCount());
MGLOG_D("%s: %s = %d", __func__, MG_Util::ConvertGLEnumToString(pname).c_str(), *params);
break;
case GL_TRANSFORM_FEEDBACK_BUFFER_MODE:
*params = static_cast<GLint>(programObject->GetTransformFeedbackBufferMode());
MGLOG_D("%s: %s = %d", __func__, MG_Util::ConvertGLEnumToString(pname).c_str(), *params);
break;
case GL_TRANSFORM_FEEDBACK_VARYING_MAX_LENGTH:
*params = programObject->GetTransformFeedbackVaryingMaxLength();
MGLOG_D("%s: %s = %d", __func__, MG_Util::ConvertGLEnumToString(pname).c_str(), *params);
break;
case GL_COMPUTE_WORK_GROUP_SIZE: { // GL >= 4.3
if (!programObject->GetLinkStatus() || programObject->GetShaderIndexByStage(ShaderStage::Compute) < 0) {
MG_State::pGLContext->RecordError(
@@ -694,14 +623,10 @@ namespace MobileGL::MG_Impl::GLImpl {
}
case GL_PROGRAM_BINARY_LENGTH:
// No program binary format is exposed, so a program never has a retrievable
// binary and its length is zero (ARB_get_program_binary).
*params = 0;
break;
case GL_PROGRAM_BINARY_RETRIEVABLE_HINT:
*params = programObject->GetBinaryRetrievableHint() ? GL_TRUE : GL_FALSE;
break;
case GL_TRANSFORM_FEEDBACK_BUFFER_MODE:
case GL_TRANSFORM_FEEDBACK_VARYINGS:
case GL_TRANSFORM_FEEDBACK_VARYING_MAX_LENGTH:
case GL_GEOMETRY_VERTICES_OUT:
case GL_GEOMETRY_INPUT_TYPE:
case GL_GEOMETRY_OUTPUT_TYPE:
@@ -876,8 +801,7 @@ namespace MobileGL::MG_Impl::GLImpl {
}
if constexpr (std::is_same_v<T, GLfloat>) {
if (ttype->getBasicType() != glslang::EbtDouble && ttype->isMatrix() &&
ttype->getMatrixCols() == 3) {
if (ttype->isMatrix() && ttype->getMatrixCols() == 3) {
auto* pBase = pUBO + offset;
for (int i = 0; i < ttype->getMatrixRows(); i++) {
Memcpy(reinterpret_cast<char*>(params) + ttype->getMatrixCols() * sizeof(GLfloat) * i,
@@ -887,46 +811,9 @@ namespace MobileGL::MG_Impl::GLImpl {
}
}
// A double-precision uniform is the one case where the stored component type can
// differ from the queried one for a non-opaque uniform, and the difference is not
// just a reinterpretation: it is twice as wide, so a raw copy would overrun the
// caller's buffer as well as return nonsense. Read component by component and let
// GL's conversion rules (7.6: round to nearest for the integer queries) apply.
if (ttype->getBasicType() == glslang::EbtDouble) {
const Int columns = ttype->isMatrix() ? ttype->getMatrixCols() : 1;
const Int rows = ttype->isMatrix() ? ttype->getMatrixRows()
: (ttype->isVector() ? ttype->getVectorSize() : 1);
// The slot the linker handed out is exactly `columns` columns wide, so it also
// states the column stride - which for a double matrix is not a float's 16 bytes.
const SizeT columnStride = columns > 0 ? size / static_cast<SizeT>(columns) : size;
for (Int column = 0; column < columns; ++column) {
for (Int row = 0; row < rows; ++row) {
GLdouble component = 0.0;
Memcpy(&component, pUBO + offset + column * columnStride + row * sizeof(GLdouble),
sizeof(component));
if constexpr (std::is_integral_v<T>) {
// Rounded to the nearest integer and clamped into the queried type's
// range, so a negative double read through glGetUniformuiv is 0
// rather than its two's complement.
const GLdouble rounded = std::nearbyint(component);
const GLdouble lowest = static_cast<GLdouble>(std::numeric_limits<T>::lowest());
const GLdouble highest = static_cast<GLdouble>(std::numeric_limits<T>::max());
params[column * rows + row] = static_cast<T>(std::clamp(rounded, lowest, highest));
} else {
params[column * rows + row] = static_cast<T>(component);
}
}
}
return;
}
Memcpy(params, pUBO + offset, size);
}
void GetUniformdv_State(GLuint program, GLint location, GLdouble* params) {
GetUniformScalar_State(program, location, params);
}
void GetUniformfv_State(GLuint program, GLint location, GLfloat* params) {
GetUniformScalar_State(program, location, params);
}
@@ -940,13 +827,19 @@ namespace MobileGL::MG_Impl::GLImpl {
}
GLboolean IsProgram_State(GLuint program) {
// Deletion-flagged names stay valid while the object is still GL-visible (program in
// use, shader attached), so name validity is exactly the Is* answer.
/* FIXME: Handle situations that:
* A program object marked for deletion with glDeleteProgram but still in use as part of current
* rendering state is still considered a program object and glIsProgram will return GL_TRUE.
*/
if (program == 0) return GL_FALSE;
return MG_State::pGLContext->ValidateProgramName(program) ? GL_TRUE : GL_FALSE;
}
GLboolean IsShader_State(GLuint shader) {
/* FIXME: Handle situations that:
* A shader object marked for deletion with glDeleteShader but still attached to a program object is still
* considered a shader object and glIsShader will return GL_TRUE.
*/
if (shader == 0) return GL_FALSE;
return MG_State::pGLContext->ValidateShaderName(shader) ? GL_TRUE : GL_FALSE;
}
@@ -956,18 +849,6 @@ namespace MobileGL::MG_Impl::GLImpl {
if (!programObject) return;
MGLOG_D("%s: linking program %d", __func__, program);
// Relinking the program an active transform feedback captures from would
// invalidate its varyings mid-capture (GL 3.3 core 2.11.3).
if (MG_State::pGLContext->IsTransformFeedbackActive() &&
MG_State::pGLContext->GetTransformFeedbackProgram().get() == programObject.get()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", __func__,
"The program used by active transform feedback cannot be relinked."));
return;
}
static Bool allowVSOnlyPrograms;
static Bool initialized = false;
if (!initialized) {
@@ -1011,18 +892,6 @@ namespace MobileGL::MG_Impl::GLImpl {
void UseProgram_State(GLuint program) {
MGLOG_D("UseProgram_State: program=%u", program);
// The program in use may not change while transform feedback is active - unless
// the capture is paused, which is exactly what ARB_transform_feedback2 added the
// pause for (GL 4.6 core 7.3).
if (MG_State::pGLContext->IsTransformFeedbackActive() &&
!MG_State::pGLContext->IsTransformFeedbackPaused()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"The current program cannot change while transform feedback is active."));
return;
}
if (program == 0) {
MG_State::pGLContext->UseProgram(0);
return;
@@ -1145,38 +1014,6 @@ namespace MobileGL::MG_Impl::GLImpl {
}
}
// glUniform*d / glUniformMatrix*dv. The vector forms need nothing beyond the shared
// upload template - it is already typed on the component - but a matrix does: the
// column stride the linker used for a double matrix is not the 16 bytes a float one
// gets. It is not guessed here; the slot the uniform was given is exactly `columns`
// columns wide, so dividing states the stride the rest of the pipeline agreed on.
template <typename Program>
void UniformMatrixdv_Object(Program& programObject, GLint location, GLsizei count, GLboolean transpose,
const GLdouble* value, Int columns, Int rows) {
const SizeT slotSize = programObject.GetUniformSizesInBytes(location);
const SizeT columnStride = columns > 0 ? slotSize / static_cast<SizeT>(columns) : slotSize;
const SizeT componentCount = static_cast<SizeT>(columns) * static_cast<SizeT>(rows);
Vector<GLdouble> column(static_cast<SizeT>(rows));
for (GLint matrix = 0; matrix < count; ++matrix) {
if (matrix > 0 && !programObject.UniformLocationsAliasSameUniform(location, location + matrix)) break;
if (!programObject.IsValidUniformLocation(location + matrix)) {
RecordInvalidUniformLocationError(__func__, location + matrix, "the current program object");
return;
}
const GLdouble* source = value + matrix * componentCount;
for (Int c = 0; c < columns; ++c) {
for (Int r = 0; r < rows; ++r) {
column[r] = transpose == GL_TRUE ? source[r * columns + c] : source[c * rows + r];
}
Uniform_State<1>(programObject, location + matrix, column.data(), c * columnStride);
for (Int r = 1; r < rows; ++r) {
Uniform_State<1>(programObject, location + matrix, column.data() + r,
c * columnStride + r * sizeof(GLdouble));
}
}
}
}
// Helper function to transpose a 2x2 matrix
void TransposeMatrix2x2(const GLfloat* input, GLfloat* output) {
// Input matrix is in column-major order (OpenGL default)
@@ -1985,312 +1822,6 @@ namespace MobileGL::MG_Impl::GLImpl {
GLuint v[] = {v0, v1, v2, v3};
Uniform4uiv(location, 1, v);
}
void Uniform1d(GLint location, GLdouble v0) {
const GLdouble v[] = {v0};
Uniformv_State<1>(location, 1, v);
}
void Uniform1dv(GLint location, GLsizei count, const GLdouble* value) {
Uniformv_State<1>(location, count, value);
}
void ProgramUniform1d(GLuint program, GLint location, GLdouble v0) {
const GLdouble v[] = {v0};
ProgramUniformv_State<1>(program, location, 1, v);
}
void ProgramUniform1dv(GLuint program, GLint location, GLsizei count, const GLdouble* value) {
ProgramUniformv_State<1>(program, location, count, value);
}
void Uniform2d(GLint location, GLdouble v0, GLdouble v1) {
const GLdouble v[] = {v0, v1};
Uniformv_State<2>(location, 1, v);
}
void Uniform2dv(GLint location, GLsizei count, const GLdouble* value) {
Uniformv_State<2>(location, count, value);
}
void ProgramUniform2d(GLuint program, GLint location, GLdouble v0, GLdouble v1) {
const GLdouble v[] = {v0, v1};
ProgramUniformv_State<2>(program, location, 1, v);
}
void ProgramUniform2dv(GLuint program, GLint location, GLsizei count, const GLdouble* value) {
ProgramUniformv_State<2>(program, location, count, value);
}
void Uniform3d(GLint location, GLdouble v0, GLdouble v1, GLdouble v2) {
const GLdouble v[] = {v0, v1, v2};
Uniformv_State<3>(location, 1, v);
}
void Uniform3dv(GLint location, GLsizei count, const GLdouble* value) {
Uniformv_State<3>(location, count, value);
}
void ProgramUniform3d(GLuint program, GLint location, GLdouble v0, GLdouble v1, GLdouble v2) {
const GLdouble v[] = {v0, v1, v2};
ProgramUniformv_State<3>(program, location, 1, v);
}
void ProgramUniform3dv(GLuint program, GLint location, GLsizei count, const GLdouble* value) {
ProgramUniformv_State<3>(program, location, count, value);
}
void Uniform4d(GLint location, GLdouble v0, GLdouble v1, GLdouble v2, GLdouble v3) {
const GLdouble v[] = {v0, v1, v2, v3};
Uniformv_State<4>(location, 1, v);
}
void Uniform4dv(GLint location, GLsizei count, const GLdouble* value) {
Uniformv_State<4>(location, count, value);
}
void ProgramUniform4d(GLuint program, GLint location, GLdouble v0, GLdouble v1, GLdouble v2, GLdouble v3) {
const GLdouble v[] = {v0, v1, v2, v3};
ProgramUniformv_State<4>(program, location, 1, v);
}
void ProgramUniform4dv(GLuint program, GLint location, GLsizei count, const GLdouble* value) {
ProgramUniformv_State<4>(program, location, count, value);
}
void UniformMatrix2dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) {
if (location == -1) return;
auto& programObject = MG_State::pGLContext->GetCurrentProgram();
if (programObject == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "There is no current program object."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 2, 2);
}
void ProgramUniformMatrix2dv(GLuint program, GLint location, GLsizei count, GLboolean transpose,
const GLdouble* value) {
if (location == -1) return;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (!programObject->GetLinkStatus()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"program " + std::to_string(program) + " is not linked."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 2, 2);
}
void UniformMatrix3dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) {
if (location == -1) return;
auto& programObject = MG_State::pGLContext->GetCurrentProgram();
if (programObject == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "There is no current program object."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 3, 3);
}
void ProgramUniformMatrix3dv(GLuint program, GLint location, GLsizei count, GLboolean transpose,
const GLdouble* value) {
if (location == -1) return;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (!programObject->GetLinkStatus()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"program " + std::to_string(program) + " is not linked."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 3, 3);
}
void UniformMatrix4dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) {
if (location == -1) return;
auto& programObject = MG_State::pGLContext->GetCurrentProgram();
if (programObject == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "There is no current program object."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 4, 4);
}
void ProgramUniformMatrix4dv(GLuint program, GLint location, GLsizei count, GLboolean transpose,
const GLdouble* value) {
if (location == -1) return;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (!programObject->GetLinkStatus()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"program " + std::to_string(program) + " is not linked."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 4, 4);
}
void UniformMatrix2x3dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) {
if (location == -1) return;
auto& programObject = MG_State::pGLContext->GetCurrentProgram();
if (programObject == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "There is no current program object."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 2, 3);
}
void ProgramUniformMatrix2x3dv(GLuint program, GLint location, GLsizei count, GLboolean transpose,
const GLdouble* value) {
if (location == -1) return;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (!programObject->GetLinkStatus()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"program " + std::to_string(program) + " is not linked."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 2, 3);
}
void UniformMatrix2x4dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) {
if (location == -1) return;
auto& programObject = MG_State::pGLContext->GetCurrentProgram();
if (programObject == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "There is no current program object."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 2, 4);
}
void ProgramUniformMatrix2x4dv(GLuint program, GLint location, GLsizei count, GLboolean transpose,
const GLdouble* value) {
if (location == -1) return;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (!programObject->GetLinkStatus()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"program " + std::to_string(program) + " is not linked."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 2, 4);
}
void UniformMatrix3x2dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) {
if (location == -1) return;
auto& programObject = MG_State::pGLContext->GetCurrentProgram();
if (programObject == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "There is no current program object."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 3, 2);
}
void ProgramUniformMatrix3x2dv(GLuint program, GLint location, GLsizei count, GLboolean transpose,
const GLdouble* value) {
if (location == -1) return;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (!programObject->GetLinkStatus()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"program " + std::to_string(program) + " is not linked."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 3, 2);
}
void UniformMatrix3x4dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) {
if (location == -1) return;
auto& programObject = MG_State::pGLContext->GetCurrentProgram();
if (programObject == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "There is no current program object."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 3, 4);
}
void ProgramUniformMatrix3x4dv(GLuint program, GLint location, GLsizei count, GLboolean transpose,
const GLdouble* value) {
if (location == -1) return;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (!programObject->GetLinkStatus()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"program " + std::to_string(program) + " is not linked."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 3, 4);
}
void UniformMatrix4x2dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) {
if (location == -1) return;
auto& programObject = MG_State::pGLContext->GetCurrentProgram();
if (programObject == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "There is no current program object."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 4, 2);
}
void ProgramUniformMatrix4x2dv(GLuint program, GLint location, GLsizei count, GLboolean transpose,
const GLdouble* value) {
if (location == -1) return;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (!programObject->GetLinkStatus()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"program " + std::to_string(program) + " is not linked."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 4, 2);
}
void UniformMatrix4x3dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value) {
if (location == -1) return;
auto& programObject = MG_State::pGLContext->GetCurrentProgram();
if (programObject == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "There is no current program object."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 4, 3);
}
void ProgramUniformMatrix4x3dv(GLuint program, GLint location, GLsizei count, GLboolean transpose,
const GLdouble* value) {
if (location == -1) return;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (!programObject->GetLinkStatus()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"program " + std::to_string(program) + " is not linked."));
return;
}
UniformMatrixdv_Object(*programObject, location, count, transpose, value, 4, 3);
}
void GetUniformdv(GLuint program, GLint location, GLdouble* params) {
GetUniformdv_State(program, location, params);
}
void Uniform1fv(GLint location, GLsizei count, const GLfloat* value) {
Uniform1fv_State(location, count, value);
}
@@ -2570,17 +2101,6 @@ namespace MobileGL::MG_Impl::GLImpl {
"Backend does not support program interface queries."));
return;
}
if (programInterface == GL_UNIFORM) {
if (pname == GL_ACTIVE_RESOURCES) {
*params = static_cast<GLint>(programObject->GetUniformCount());
return;
}
if (pname == GL_MAX_NAME_LENGTH) {
// Stored as the bare length; GL_MAX_NAME_LENGTH counts the terminator.
*params = programObject->GetUniformMaxLength() + 1;
return;
}
}
getProgramInterfaceiv(program, programInterface, pname, params);
}
@@ -2589,10 +2109,6 @@ namespace MobileGL::MG_Impl::GLImpl {
if (!programObject) return GL_INVALID_INDEX;
if (!ValidateNamedProgramResourceInterface(programInterface, __func__)) return GL_INVALID_INDEX;
if (!name) return GL_INVALID_INDEX;
if (programInterface == GL_UNIFORM) {
const Int uniformIndex = programObject->GetActiveUniformIndex(name);
return uniformIndex < 0 ? GL_INVALID_INDEX : static_cast<GLuint>(uniformIndex);
}
auto getProgramResourceIndex = MG_Backend::gBackendFunctionsTable.GL.GetProgramResourceIndex;
if (!getProgramResourceIndex) {
MG_State::pGLContext->RecordError(
@@ -2629,13 +2145,6 @@ namespace MobileGL::MG_Impl::GLImpl {
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "bufSize must be non-negative."));
return;
}
if (programInterface == GL_UNIFORM) {
// Same index space GetProgramResourceIndex answers in, and the range check above
// already used it.
const String& uniformName = programObject->GetActiveUniformName(index);
CopyStr(bufSize, length, name, uniformName.c_str(), static_cast<GLsizei>(uniformName.length()));
return;
}
auto getProgramResourceName = MG_Backend::gBackendFunctionsTable.GL.GetProgramResourceName;
if (!getProgramResourceName) {
MG_State::pGLContext->RecordError(
@@ -2657,36 +2166,6 @@ namespace MobileGL::MG_Impl::GLImpl {
"propCount and bufSize must be non-negative."));
return;
}
if (programInterface == GL_UNIFORM) {
if (index >= programObject->GetUniformCount()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "index is out of range."));
return;
}
if (props == nullptr || params == nullptr) return;
GLsizei written = 0;
for (GLsizei i = 0; i < propCount && written < bufSize; ++i) {
GLint value = 0;
if (!GetUniformResourceProp(programObject, index, props[i], &value)) {
// GL_ATOMIC_COUNTER_BUFFER_INDEX and the GL_REFERENCED_BY_* stage props are
// not modelled here; ask the backend, which indexes resources by name.
auto backendGetIndex = MG_Backend::gBackendFunctionsTable.GL.GetProgramResourceIndex;
auto backendGetiv = MG_Backend::gBackendFunctionsTable.GL.GetProgramResourceiv;
if (backendGetIndex && backendGetiv) {
const GLuint backendIndex = backendGetIndex(program, GL_UNIFORM,
programObject->GetActiveUniformName(index).c_str());
if (backendIndex != GL_INVALID_INDEX) {
GLsizei one = 0;
backendGetiv(program, GL_UNIFORM, backendIndex, 1, &props[i], 1, &one, &value);
}
}
}
params[written++] = value;
}
if (length) *length = written;
return;
}
auto getProgramResourceiv = MG_Backend::gBackendFunctionsTable.GL.GetProgramResourceiv;
if (!getProgramResourceiv) {
MG_State::pGLContext->RecordError(
@@ -2738,150 +2217,4 @@ namespace MobileGL::MG_Impl::GLImpl {
void ValidateProgram(GLuint program) {
ValidateProgram_State(program);
}
// ARB_get_program_binary with no supported binary format (GL_NUM_PROGRAM_BINARY_FORMATS
// is 0, which the extension explicitly allows). The three entry points below are what an
// application - and dEQP's function loader - reach through the extension; without it
// glProgramParameteri is not exposed in a 4.0 context at all.
void ProgramParameteri(GLuint program, GLenum pname, GLint value) {
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (pname != GL_PROGRAM_BINARY_RETRIEVABLE_HINT) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "pname is not an accepted value."));
return;
}
if (value != GL_TRUE && value != GL_FALSE) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "value must be GL_TRUE or GL_FALSE."));
return;
}
programObject->SetBinaryRetrievableHint(value == GL_TRUE);
}
void GetProgramBinary(GLuint program, GLsizei bufSize, GLsizei* length, GLenum* binaryFormat, void* binary) {
(void)binaryFormat;
(void)binary;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (bufSize < 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "bufSize must be non-negative."));
return;
}
if (length) *length = 0;
// GL_PROGRAM_BINARY_LENGTH is always zero here, which the spec makes an error to ask for.
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "The program has no retrievable binary."));
}
void ProgramBinary(GLuint program, GLenum binaryFormat, const void* binary, GLsizei length) {
(void)binaryFormat;
(void)binary;
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (length < 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "length must be non-negative."));
return;
}
// No format is supported, so every binary is rejected - and the program's link status
// has to read FALSE afterwards.
programObject->MarkLinkFailedByProgramBinary();
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "binaryFormat is not a supported format."));
}
void TransformFeedbackVaryings(GLuint program, GLsizei count, const GLchar* const* varyings, GLenum bufferMode) {
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (bufferMode != GL_INTERLEAVED_ATTRIBS && bufferMode != GL_SEPARATE_ATTRIBS) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "bufferMode is not a valid capture mode."));
return;
}
if (count < 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "count must be non-negative."));
return;
}
// GL 3.3 core: SEPARATE_ATTRIBS count may not exceed the separate-attrib limit.
if (bufferMode == GL_SEPARATE_ATTRIBS && count > 4) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"count exceeds GL_MAX_TRANSFORM_FEEDBACK_SEPARATE_ATTRIBS."));
return;
}
Vector<String> names;
names.reserve(static_cast<SizeT>(count));
for (GLsizei i = 0; i < count; ++i) {
names.emplace_back(varyings != nullptr && varyings[i] != nullptr ? varyings[i] : "");
}
// ARB_transform_feedback3's special names only mean anything in an interleaved
// capture, and gl_NextBuffer cannot advance past the last capture buffer.
constexpr Uint maxTransformFeedbackBuffers = 4;
Uint nextBufferCount = 0;
for (const String& name : names) {
const Bool isNextBuffer = name == "gl_NextBuffer";
const Bool isSkipComponents = name.size() == 18 && name.compare(0, 17, "gl_SkipComponents") == 0 &&
name[17] >= '1' && name[17] <= '4';
if (!isNextBuffer && !isSkipComponents) continue;
if (bufferMode != GL_INTERLEAVED_ATTRIBS) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"'" + name + "' requires GL_INTERLEAVED_ATTRIBS."));
return;
}
if (isNextBuffer && ++nextBufferCount >= maxTransformFeedbackBuffers) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"More gl_NextBuffer entries than "
"GL_MAX_TRANSFORM_FEEDBACK_BUFFERS allows."));
return;
}
}
programObject->SetTransformFeedbackVaryings(Move(names), bufferMode);
}
void GetTransformFeedbackVarying(GLuint program, GLuint index, GLsizei bufSize, GLsizei* length, GLsizei* size,
GLenum* type, GLchar* name) {
auto& programObject = TryToGetProgramObject(program);
if (!programObject) return;
if (!programObject->GetLinkStatus()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
std::to_string(program) + " has not been successfully linked."));
return;
}
const auto* varying = programObject->GetTransformFeedbackVarying(index);
if (varying == nullptr) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", __func__,
"index is not an active transform feedback varying of the program."));
return;
}
if (size != nullptr) *size = varying->size;
if (type != nullptr) *type = varying->type;
GLsizei written = 0;
if (name != nullptr && bufSize > 0) {
written = std::min<GLsizei>(bufSize - 1, static_cast<GLsizei>(varying->name.size()));
Memcpy(name, varying->name.data(), static_cast<SizeT>(written));
name[written] = '\0';
}
if (length != nullptr) *length = written;
}
} // namespace MobileGL::MG_Impl::GLImpl
@@ -137,46 +137,5 @@ namespace MobileGL::MG_Impl::GLImpl {
GLint GetProgramResourceLocation(GLuint program, GLenum programInterface, const GLchar* name);
GLint GetProgramResourceLocationIndex(GLuint program, GLenum programInterface, const GLchar* name);
void ShaderStorageBlockBinding(GLuint program, GLuint storageBlockIndex, GLuint storageBlockBinding);
void Uniform1d(GLint location, GLdouble v0);
void Uniform1dv(GLint location, GLsizei count, const GLdouble* value);
void ProgramUniform1d(GLuint program, GLint location, GLdouble v0);
void ProgramUniform1dv(GLuint program, GLint location, GLsizei count, const GLdouble* value);
void Uniform2d(GLint location, GLdouble v0, GLdouble v1);
void Uniform2dv(GLint location, GLsizei count, const GLdouble* value);
void ProgramUniform2d(GLuint program, GLint location, GLdouble v0, GLdouble v1);
void ProgramUniform2dv(GLuint program, GLint location, GLsizei count, const GLdouble* value);
void Uniform3d(GLint location, GLdouble v0, GLdouble v1, GLdouble v2);
void Uniform3dv(GLint location, GLsizei count, const GLdouble* value);
void ProgramUniform3d(GLuint program, GLint location, GLdouble v0, GLdouble v1, GLdouble v2);
void ProgramUniform3dv(GLuint program, GLint location, GLsizei count, const GLdouble* value);
void Uniform4d(GLint location, GLdouble v0, GLdouble v1, GLdouble v2, GLdouble v3);
void Uniform4dv(GLint location, GLsizei count, const GLdouble* value);
void ProgramUniform4d(GLuint program, GLint location, GLdouble v0, GLdouble v1, GLdouble v2, GLdouble v3);
void ProgramUniform4dv(GLuint program, GLint location, GLsizei count, const GLdouble* value);
void UniformMatrix2dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void ProgramUniformMatrix2dv(GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void UniformMatrix3dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void ProgramUniformMatrix3dv(GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void UniformMatrix4dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void ProgramUniformMatrix4dv(GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void UniformMatrix2x3dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void ProgramUniformMatrix2x3dv(GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void UniformMatrix2x4dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void ProgramUniformMatrix2x4dv(GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void UniformMatrix3x2dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void ProgramUniformMatrix3x2dv(GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void UniformMatrix3x4dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void ProgramUniformMatrix3x4dv(GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void UniformMatrix4x2dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void ProgramUniformMatrix4x2dv(GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void UniformMatrix4x3dv(GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void ProgramUniformMatrix4x3dv(GLuint program, GLint location, GLsizei count, GLboolean transpose, const GLdouble* value);
void GetUniformdv(GLuint program, GLint location, GLdouble* params);
void ValidateProgram(GLuint program);
void ProgramParameteri(GLuint program, GLenum pname, GLint value);
void GetProgramBinary(GLuint program, GLsizei bufSize, GLsizei* length, GLenum* binaryFormat, void* binary);
void ProgramBinary(GLuint program, GLenum binaryFormat, const void* binary, GLsizei length);
void TransformFeedbackVaryings(GLuint program, GLsizei count, const GLchar* const* varyings, GLenum bufferMode);
void GetTransformFeedbackVarying(GLuint program, GLuint index, GLsizei bufSize, GLsizei* length, GLsizei* size,
GLenum* type, GLchar* name);
} // namespace MobileGL::MG_Impl::GLImpl
+23 -209
View File
@@ -7,7 +7,6 @@
// End of Source File Header
#include "GL_Query.h"
#include "../Getter/GL_Getter.h"
#include <Config.h>
#include <MG_Backend/BackendObjects.h>
#include <MG_State/GLState/Core.h>
@@ -23,16 +22,11 @@ namespace MobileGL::MG_Impl::GLImpl {
struct QueryObject {
GLuint id = 0;
GLenum target = 0; // 0 = gen'd but never used with BeginQuery/QueryCounter
// glCreateQueries makes the object outright; glGenQueries only reserves the name,
// and the object appears when the name is first used (GL 4.6 core 4.2.1).
Bool created = false;
MG_Backend::BackendQueryHandle backendHandle = nullptr;
Bool active = false;
Bool ended = false;
Bool resultCached = false;
Uint64 cachedResult = 0;
// Transform feedback primitive counter at BeginQuery time.
Uint64 counterSnapshot = 0;
};
// Query calls may arrive from any thread (launchers migrate the context
@@ -47,11 +41,6 @@ namespace MobileGL::MG_Impl::GLImpl {
GLuint g_nextQueryId = 1;
// Id of the query currently active on GL_TIME_ELAPSED (0 = none).
GLuint g_activeTimeElapsedQueryId = 0;
// Ids of the queries active on the transform feedback targets (0 = none).
GLuint g_activePrimitivesWrittenQueryId = 0;
GLuint g_activePrimitivesGeneratedQueryId = 0;
// Id of the query active on GL_SAMPLES_PASSED (0 = none).
GLuint g_activeSamplesPassedQueryId = 0;
Bool TimerQueryDisabled() {
return MG_Config::Features.DisableTimerQuery;
@@ -137,11 +126,6 @@ namespace MobileGL::MG_Impl::GLImpl {
outValue = 0;
return true;
}
// ANY_SAMPLES_PASSED* report a boolean.
if (queryObject->target == GL_ANY_SAMPLES_PASSED ||
queryObject->target == GL_ANY_SAMPLES_PASSED_CONSERVATIVE) {
result = result != 0 ? 1 : 0;
}
// Final value produced (or no GetQueryResult64 hook: the
// query degrades to a zero result); the backend handle is
// consumed and the value cached for later reads.
@@ -180,41 +164,6 @@ namespace MobileGL::MG_Impl::GLImpl {
}
}
// glCreateQueries differs from glGenQueries in creating the objects outright, with their
// target already fixed and the rest of their state at the defaults (GL 4.6 core 4.2.1).
void CreateQueries(GLenum target, GLsizei n, GLuint* ids) {
switch (target) {
case GL_SAMPLES_PASSED:
case GL_ANY_SAMPLES_PASSED:
case GL_ANY_SAMPLES_PASSED_CONSERVATIVE:
case GL_TIME_ELAPSED:
case GL_TIMESTAMP:
case GL_PRIMITIVES_GENERATED:
case GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN:
break;
default:
RecordQueryError(ErrorCode::InvalidEnum, __FUNCTION__, "Query target is not accepted.");
return;
}
if (n < 0) {
RecordQueryError(ErrorCode::InvalidValue, __FUNCTION__, "n cannot be negative.");
return;
}
if (!ids) {
return;
}
const std::lock_guard<std::mutex> lock(g_queryObjectsMutex);
for (GLsizei i = 0; i < n; ++i) {
const GLuint id = g_nextQueryId++;
auto* queryObject = new QueryObject;
queryObject->id = id;
queryObject->target = target;
queryObject->created = true;
g_liveQueryObjects[id] = queryObject;
ids[i] = id;
}
}
void DeleteQueries(GLsizei n, const GLuint* ids) {
if (n < 0) {
RecordQueryError(ErrorCode::InvalidValue, __FUNCTION__, "n cannot be negative.");
@@ -231,24 +180,7 @@ namespace MobileGL::MG_Impl::GLImpl {
}
QueryObject* queryObject = it->second;
if (queryObject->active) {
// Implicitly end before deletion, releasing the matching active slot.
if (queryObject->target == GL_SAMPLES_PASSED || queryObject->target == GL_ANY_SAMPLES_PASSED ||
queryObject->target == GL_ANY_SAMPLES_PASSED_CONSERVATIVE) {
if (const auto endOcclusionQuery = MG_Backend::gBackendFunctionsTable.GL.EndOcclusionQuery;
endOcclusionQuery && queryObject->backendHandle) {
endOcclusionQuery(queryObject->backendHandle);
}
queryObject->active = false;
g_activeSamplesPassedQueryId = 0;
} else if (queryObject->target == GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN ||
queryObject->target == GL_PRIMITIVES_GENERATED) {
queryObject->active = false;
(queryObject->target == GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN
? g_activePrimitivesWrittenQueryId
: g_activePrimitivesGeneratedQueryId) = 0;
} else {
EndTimeElapsedQueryLocked(queryObject);
}
EndTimeElapsedQueryLocked(queryObject); // implicitly end before deletion
}
if (queryObject->backendHandle) {
if (const auto deleteBackendQuery = MG_Backend::gBackendFunctionsTable.GL.DeleteBackendQuery) {
@@ -266,23 +198,16 @@ namespace MobileGL::MG_Impl::GLImpl {
return GL_FALSE;
}
const std::lock_guard<std::mutex> lock(g_queryObjectsMutex);
// A name from glGenQueries is not yet a query object: it becomes one when it is first
// used with BeginQuery/QueryCounter (which is what a non-zero target records), or
// immediately if it came from glCreateQueries.
const auto* queryObject = FindQueryObjectLocked(id);
return (queryObject != nullptr && (queryObject->created || queryObject->target != 0)) ? GL_TRUE : GL_FALSE;
// Gen'd ids count as query objects here: the registry creates live
// objects at GenQueries time.
return FindQueryObjectLocked(id) != nullptr ? GL_TRUE : GL_FALSE;
}
void BeginQuery(GLenum target, GLuint id) {
const Bool isTransformFeedbackQuery =
target == GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN || target == GL_PRIMITIVES_GENERATED;
const Bool isOcclusionQuery =
(target == GL_SAMPLES_PASSED || target == GL_ANY_SAMPLES_PASSED ||
target == GL_ANY_SAMPLES_PASSED_CONSERVATIVE) &&
MG_Backend::gBackendFunctionsTable.GL.BeginOcclusionQuery != nullptr;
if (target != GL_TIME_ELAPSED && !isTransformFeedbackQuery && !isOcclusionQuery) {
// GL_TIMESTAMP is not a valid BeginQuery target; the occlusion targets
// need backend support.
if (target != GL_TIME_ELAPSED) {
// Only GL_TIME_ELAPSED timer queries are implemented (occlusion and
// primitive queries remain stubs); GL_TIMESTAMP is not a valid
// BeginQuery target either.
RecordQueryError(ErrorCode::InvalidEnum, __FUNCTION__, "Query target is not supported.");
return;
}
@@ -296,13 +221,9 @@ namespace MobileGL::MG_Impl::GLImpl {
RecordQueryError(ErrorCode::InvalidOperation, __FUNCTION__, "Query object does not exist.");
return;
}
GLuint& activeQueryId = isTransformFeedbackQuery
? (target == GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN ? g_activePrimitivesWrittenQueryId
: g_activePrimitivesGeneratedQueryId)
: (isOcclusionQuery ? g_activeSamplesPassedQueryId : g_activeTimeElapsedQueryId);
if (activeQueryId != 0) {
if (g_activeTimeElapsedQueryId != 0) {
RecordQueryError(ErrorCode::InvalidOperation, __FUNCTION__,
"A query is already active on this target.");
"A query is already active on GL_TIME_ELAPSED.");
return;
}
if (queryObject->active) {
@@ -318,72 +239,25 @@ namespace MobileGL::MG_Impl::GLImpl {
ResetQueryObjectLocked(queryObject); // discard any previous result
queryObject->target = target;
queryObject->active = true;
if (isTransformFeedbackQuery) {
// Prefer real GPU transform-feedback queries (exact with geometry shaders);
// the CPU accounting delta stays as the fallback when the backend lacks them.
const auto beginXfbPrimitivesQuery = MG_Backend::gBackendFunctionsTable.GL.BeginXfbPrimitivesQuery;
queryObject->backendHandle =
beginXfbPrimitivesQuery ? beginXfbPrimitivesQuery(target == GL_PRIMITIVES_GENERATED) : nullptr;
queryObject->counterSnapshot = MG_State::pGLContext->GetTransformFeedbackPrimitiveCounter();
} else if (isOcclusionQuery) {
queryObject->backendHandle = MG_Backend::gBackendFunctionsTable.GL.BeginOcclusionQuery();
} else {
const auto beginTimeElapsedQuery = MG_Backend::gBackendFunctionsTable.GL.BeginTimeElapsedQuery;
queryObject->backendHandle =
(!TimerQueryDisabled() && beginTimeElapsedQuery) ? beginTimeElapsedQuery() : nullptr;
}
activeQueryId = id;
const auto beginTimeElapsedQuery = MG_Backend::gBackendFunctionsTable.GL.BeginTimeElapsedQuery;
queryObject->backendHandle =
(!TimerQueryDisabled() && beginTimeElapsedQuery) ? beginTimeElapsedQuery() : nullptr;
g_activeTimeElapsedQueryId = id;
}
void EndQuery(GLenum target) {
const Bool isTransformFeedbackQuery =
target == GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN || target == GL_PRIMITIVES_GENERATED;
const Bool isOcclusionQuery =
(target == GL_SAMPLES_PASSED || target == GL_ANY_SAMPLES_PASSED ||
target == GL_ANY_SAMPLES_PASSED_CONSERVATIVE) &&
MG_Backend::gBackendFunctionsTable.GL.BeginOcclusionQuery != nullptr;
if (target != GL_TIME_ELAPSED && !isTransformFeedbackQuery && !isOcclusionQuery) {
if (target != GL_TIME_ELAPSED) {
RecordQueryError(ErrorCode::InvalidEnum, __FUNCTION__, "Query target is not supported.");
return;
}
const std::lock_guard<std::mutex> lock(g_queryObjectsMutex);
GLuint& activeQueryId = isTransformFeedbackQuery
? (target == GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN ? g_activePrimitivesWrittenQueryId
: g_activePrimitivesGeneratedQueryId)
: (isOcclusionQuery ? g_activeSamplesPassedQueryId : g_activeTimeElapsedQueryId);
if (activeQueryId == 0) {
RecordQueryError(ErrorCode::InvalidOperation, __FUNCTION__, "No query is active on this target.");
if (g_activeTimeElapsedQueryId == 0) {
RecordQueryError(ErrorCode::InvalidOperation, __FUNCTION__, "No query is active on GL_TIME_ELAPSED.");
return;
}
auto* queryObject = FindQueryObjectLocked(activeQueryId);
auto* queryObject = FindQueryObjectLocked(g_activeTimeElapsedQueryId);
if (!queryObject) {
activeQueryId = 0; // should not happen; keep state consistent
return;
}
if (isTransformFeedbackQuery) {
if (queryObject->backendHandle) {
if (const auto endXfbPrimitivesQuery = MG_Backend::gBackendFunctionsTable.GL.EndXfbPrimitivesQuery) {
endXfbPrimitivesQuery(queryObject->backendHandle);
}
// Result comes from the GPU query at read time.
} else {
queryObject->cachedResult =
MG_State::pGLContext->GetTransformFeedbackPrimitiveCounter() - queryObject->counterSnapshot;
queryObject->resultCached = true;
}
queryObject->active = false;
queryObject->ended = true;
activeQueryId = 0;
return;
}
if (isOcclusionQuery) {
if (const auto endOcclusionQuery = MG_Backend::gBackendFunctionsTable.GL.EndOcclusionQuery;
endOcclusionQuery && queryObject->backendHandle) {
endOcclusionQuery(queryObject->backendHandle);
}
queryObject->active = false;
queryObject->ended = true;
activeQueryId = 0;
g_activeTimeElapsedQueryId = 0; // should not happen; keep state consistent
return;
}
EndTimeElapsedQueryLocked(queryObject);
@@ -429,25 +303,9 @@ namespace MobileGL::MG_Impl::GLImpl {
switch (pname) {
case GL_CURRENT_QUERY: {
const std::lock_guard<std::mutex> lock(g_queryObjectsMutex);
switch (target) {
case GL_TIME_ELAPSED:
*params = static_cast<GLint>(g_activeTimeElapsedQueryId);
break;
case GL_SAMPLES_PASSED:
case GL_ANY_SAMPLES_PASSED:
case GL_ANY_SAMPLES_PASSED_CONSERVATIVE:
*params = static_cast<GLint>(g_activeSamplesPassedQueryId);
break;
case GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN:
*params = static_cast<GLint>(g_activePrimitivesWrittenQueryId);
break;
case GL_PRIMITIVES_GENERATED:
*params = static_cast<GLint>(g_activePrimitivesGeneratedQueryId);
break;
default:
*params = 0;
break;
}
// Only GL_TIME_ELAPSED queries can be active; GL_TIMESTAMP queries
// never are, and other targets remain unimplemented.
*params = target == GL_TIME_ELAPSED ? static_cast<GLint>(g_activeTimeElapsedQueryId) : 0;
return;
}
case GL_QUERY_COUNTER_BITS: {
@@ -455,13 +313,7 @@ namespace MobileGL::MG_Impl::GLImpl {
// time: IsTimerQuerySupported is the dynamic truth (extension /
// entry points / timestamp valid bits at call time, not at table
// init), and the MOBILEGL_DISABLE_TIMERQUERY kill switch always
// wins.
if (target == GL_SAMPLES_PASSED || target == GL_ANY_SAMPLES_PASSED ||
target == GL_ANY_SAMPLES_PASSED_CONSERVATIVE) {
const Bool occlusionSupported = MG_Backend::gBackendFunctionsTable.GL.BeginOcclusionQuery != nullptr;
*params = occlusionSupported ? (target == GL_SAMPLES_PASSED ? 32 : 1) : 0;
return;
}
// wins. Non-timer targets remain unimplemented and report 0.
const Bool timerTarget = target == GL_TIME_ELAPSED || target == GL_TIMESTAMP;
const auto isTimerQuerySupported = MG_Backend::gBackendFunctionsTable.GL.IsTimerQuerySupported;
const Bool supported =
@@ -507,42 +359,4 @@ namespace MobileGL::MG_Impl::GLImpl {
}
*params = static_cast<GLuint64>(value);
}
namespace {
// The indexed query entry points differ from the plain ones only in the vertex
// stream they address (GL 4.6 core 4.2.1): index must be below GL_MAX_VERTEX_STREAMS
// for the two transform feedback targets and zero for every other target. With a
// single vertex stream both bounds are 1, so a valid call is always index 0 and
// forwards to the unindexed implementation.
Bool ValidateQueryStreamIndex(const char* function, GLenum target, GLuint index) {
const Bool perStreamTarget =
target == GL_PRIMITIVES_GENERATED || target == GL_TRANSFORM_FEEDBACK_PRIMITIVES_WRITTEN;
GLint maxVertexStreams = 1;
if (perStreamTarget) {
GetIntegerv(GL_MAX_VERTEX_STREAMS, &maxVertexStreams);
}
if (index < static_cast<GLuint>(std::max(maxVertexStreams, 1))) {
return true;
}
RecordQueryError(ErrorCode::InvalidValue, function,
perStreamTarget ? "index is not less than GL_MAX_VERTEX_STREAMS."
: "index must be zero for this query target.");
return false;
}
} // namespace
void BeginQueryIndexed(GLenum target, GLuint index, GLuint id) {
if (!ValidateQueryStreamIndex(__FUNCTION__, target, index)) return;
BeginQuery(target, id);
}
void EndQueryIndexed(GLenum target, GLuint index) {
if (!ValidateQueryStreamIndex(__FUNCTION__, target, index)) return;
EndQuery(target);
}
void GetQueryIndexediv(GLenum target, GLuint index, GLenum pname, GLint* params) {
if (!ValidateQueryStreamIndex(__FUNCTION__, target, index)) return;
GetQueryiv(target, pname, params);
}
} // namespace MobileGL::MG_Impl::GLImpl
-4
View File
@@ -11,15 +11,11 @@
namespace MobileGL::MG_Impl::GLImpl {
void GenQueries(GLsizei n, GLuint* ids);
void CreateQueries(GLenum target, GLsizei n, GLuint* ids);
void DeleteQueries(GLsizei n, const GLuint* ids);
GLboolean IsQuery(GLuint id);
void BeginQuery(GLenum target, GLuint id);
void EndQuery(GLenum target);
void GetQueryiv(GLenum target, GLenum pname, GLint* params);
void BeginQueryIndexed(GLenum target, GLuint index, GLuint id);
void EndQueryIndexed(GLenum target, GLuint index);
void GetQueryIndexediv(GLenum target, GLuint index, GLenum pname, GLint* params);
void GetQueryObjectiv(GLuint id, GLenum pname, GLint* params);
void GetQueryObjectuiv(GLuint id, GLenum pname, GLuint* params);
void GetQueryObjecti64v(GLuint id, GLenum pname, GLint64* params);
@@ -185,11 +185,6 @@ namespace MobileGL::MG_Impl::GLImpl {
static thread_local Vector<GLuint> names;
MG_State::pGLContext->GenSamplerNames(count, names);
Memcpy(samplers, names.data(), count * sizeof(GLuint));
// Unlike textures/buffers, glGenSamplers CREATES the sampler objects: each name
// is immediately a sampler (glIsSampler == GL_TRUE before any bind).
for (GLsizei i = 0; i < count; ++i) {
MG_State::pGLContext->CreateSamplerObject(names[i]);
}
}
void DeleteSamplers_State(GLsizei count, const GLuint* samplers) {
-29
View File
@@ -133,33 +133,4 @@ namespace MobileGL::MG_Impl::GLImpl {
values[0] = value;
}
}
void DestroyAllSyncObjects() {
// Detach the registry under the lock, release outside it. Entries the app
// already deleted were erased by DeleteSync, so nothing here double-frees;
// a DeleteSync racing this sweep finds an empty registry and returns. A
// thread still blocked inside ClientWaitSync/GetSynciv during teardown
// holds a raw SyncObject* these deletes invalidate - the same undefined
// race an app-driven DeleteSync already has.
UnorderedMap<GLsync, SyncObject*> orphans;
{
const std::lock_guard<std::mutex> lock(g_syncObjectsMutex);
orphans.swap(g_liveSyncObjects);
}
if (orphans.empty()) {
return;
}
// Both backends' DeleteSync only free the heap wrapper once their GL
// context/renderer is gone (generation/current-thread guards), so this is
// safe after the backend has released its EGL resources - but not after
// the function table itself is cleared.
const auto backendDeleteSync = MG_Backend::gBackendFunctionsTable.GL.DeleteSync;
for (const auto& [_, syncObject] : orphans) {
if (backendDeleteSync && syncObject->backendHandle) {
backendDeleteSync(syncObject->backendHandle);
}
delete syncObject;
}
MGLOG_D("DestroyAllSyncObjects: reclaimed %zu sync object(s) the app left undeleted", orphans.size());
}
} // namespace MobileGL::MG_Impl::GLImpl
-8
View File
@@ -16,12 +16,4 @@ namespace MobileGL::MG_Impl::GLImpl {
void WaitSync(GLsync sync, GLbitfield flags, GLuint64 timeout);
void DeleteSync(GLsync sync);
void GetSynciv(GLsync sync, GLenum pname, GLsizei bufSize, GLsizei* length, GLint* values);
// Destroys every still-registered sync object exactly as DeleteSync would.
// GL requires syncs to die with their context; called only from full library
// teardown (DestroyImpl), where no context survives on any thread, so the
// process-global registry can be drained wholesale. Must run while the
// backend function table is still populated: each backend handle has to be
// released by the backend that created it, never by a later re-initialized
// one.
void DestroyAllSyncObjects();
} // namespace MobileGL::MG_Impl::GLImpl
+1 -186
View File
@@ -2035,128 +2035,6 @@ namespace MobileGL::MG_Impl::GLImpl {
MaybeAutoGenerateMipmap(target, textureObject, isProxy, level);
}
// The work glTexBuffer[Range] and glTextureBuffer[Range] all share once the texture has been
// resolved - by binding for the target forms, by name for the DSA ones. `size` is
// kWholeBuffer for the non-Range entry points, which attach the buffer as it grows rather
// than freezing the size it happens to have now.
// The sized internal formats a buffer texture accepts (GL 4.6 core table 8.16). This is a much
// shorter list than the renderable or texturable formats, so it cannot be inferred from either.
static Bool IsBufferTextureInternalFormat(GLenum internalformat) {
switch (internalformat) {
case GL_R8:
case GL_R16:
case GL_R16F:
case GL_R32F:
case GL_R8I:
case GL_R16I:
case GL_R32I:
case GL_R8UI:
case GL_R16UI:
case GL_R32UI:
case GL_RG8:
case GL_RG16:
case GL_RG16F:
case GL_RG32F:
case GL_RG8I:
case GL_RG16I:
case GL_RG32I:
case GL_RG8UI:
case GL_RG16UI:
case GL_RG32UI:
case GL_RGB32F:
case GL_RGB32I:
case GL_RGB32UI:
case GL_RGBA8:
case GL_RGBA16:
case GL_RGBA16F:
case GL_RGBA32F:
case GL_RGBA8I:
case GL_RGBA16I:
case GL_RGBA32I:
case GL_RGBA8UI:
case GL_RGBA16UI:
case GL_RGBA32UI:
return true;
default:
return false;
}
}
static void AttachBufferToTexture(const SharedPtr<MG_State::GLState::ITextureObject>& textureObject,
GLenum internalformat, GLuint buffer, GLintptr offset, SizeT size,
const char* caller) {
using MG_State::GLState::TextureObjectBuffer;
TextureInternalFormat textureInternalFormat = MG_Util::ConvertGLEnumToTextureInternalFormat(internalformat);
if (!IsBufferTextureInternalFormat(internalformat)) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", caller,
std::format("internalformat 0x{:X} is not one of the sized formats a buffer texture accepts.",
internalformat)));
return;
}
if (!TextureImpl::ValidateTextureInternalFormat(textureInternalFormat)) return;
auto& bufferObject = MG_State::pGLContext->GetBufferObject(buffer);
if (buffer != 0 && !bufferObject) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller,
"`buffer` is not zero and is not the name of an existing buffer object."));
return;
}
if (!TextureImpl::ValidateTextureObject(textureObject)) return;
if (textureObject->GetStorageType() != TextureStorageType::Buffer) {
// A texture whose target is something else is a wrong object, not a wrong token
// (GL 4.6 core 8.9).
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller,
"The effective target of `texture` is not `GL_TEXTURE_BUFFER`."));
return;
}
if (size != TextureObjectBuffer::kWholeBuffer) {
// GL 4.6 core 8.9: offset must be non-negative and aligned to
// GL_TEXTURE_BUFFER_OFFSET_ALIGNMENT, and size must be positive.
if (offset < 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller, "offset must be non-negative."));
return;
}
if (size == 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller, "size must be greater than zero."));
return;
}
const Int alignment = std::max(
1, MG_Backend::pActiveBackendObject->GetDynamicParameters().TextureBufferOffsetAlignment);
if (offset % alignment != 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller,
"offset is not a multiple of GL_TEXTURE_BUFFER_OFFSET_ALIGNMENT."));
return;
}
// The range has to lie inside the buffer that is being attached. Detaching (buffer
// zero) carries no range to check.
if (bufferObject && static_cast<SizeT>(offset) + size > bufferObject->GetSize()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller,
"offset + size is greater than the buffer object's GL_BUFFER_SIZE."));
return;
}
}
auto* texBufferObject = static_cast<TextureObjectBuffer*>(textureObject.get());
texBufferObject->GetBufferBindingSlot().Bind(bufferObject);
texBufferObject->SetBufferRange(static_cast<SizeT>(offset < 0 ? 0 : offset), size);
texBufferObject->SetInternalFormat(textureInternalFormat);
}
void TexBuffer_State(GLenum target, GLenum internalformat, GLuint buffer) {
// ======================= Converting ================================
TextureTarget textureTarget = MG_Util::ConvertGLEnumToTextureTarget(target);
@@ -2203,7 +2081,6 @@ namespace MobileGL::MG_Impl::GLImpl {
auto& bufferSlot = texBufferObject->GetBufferBindingSlot();
bufferSlot.Bind(bufferObject);
texBufferObject->SetBufferRange(0, MG_State::GLState::TextureObjectBuffer::kWholeBuffer);
texBufferObject->SetInternalFormat(textureInternalFormat);
}
@@ -2943,22 +2820,7 @@ namespace MobileGL::MG_Impl::GLImpl {
"The attachment specified by the read buffer is incomplete.")); \
return false; \
}
if (isDepth && isStencil) {
// A combined internalformat copies both halves, so the read framebuffer
// must populate both attachment points.
const auto& stencilAttachment = currentReadFBO->GetAttachment(FramebufferAttachmentType::Stencil);
const auto& depthAttachment = currentReadFBO->GetAttachment(FramebufferAttachmentType::Depth);
if (!depthAttachment.IsValid() || depthAttachment.IsEmpty() || !stencilAttachment.IsValid() ||
stencilAttachment.IsEmpty()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", "CopyTexImage2D_State",
"DEPTH_STENCIL copy requires both depth and stencil attachments in the read framebuffer."));
return false;
}
GET_SRC_INTERNAL_FORMAT(FramebufferAttachmentType::Depth);
} else if (isDepth) {
if (isDepth) {
GET_SRC_INTERNAL_FORMAT(FramebufferAttachmentType::Depth);
} else if (isStencil) {
GET_SRC_INTERNAL_FORMAT(FramebufferAttachmentType::Stencil);
@@ -3793,43 +3655,23 @@ namespace MobileGL::MG_Impl::GLImpl {
GetTextureImage(texture, level, format, type, bufSize, pixels);
}
// A buffer texture carries none of the sampler or level state these queries report. Reached by
// name there is no target token to blame, so the wrong object is INVALID_OPERATION rather than
// the INVALID_ENUM the target forms report for an unaccepted target (GL 4.6 core 8.11).
static Bool ValidateNamedTextureHasParameters(const SharedPtr<MG_State::GLState::ITextureObject>& textureObject,
const char* caller) {
if (!textureObject) return false;
if (textureObject->GetStorageType() == TextureStorageType::Buffer) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller,
"The effective target of `texture` has no texture parameters."));
return false;
}
return true;
}
void GetTextureParameteriv(GLuint texture, GLenum pname, GLint* params) {
auto textureObject = GetTextureObjectByName(texture, __func__);
if (!ValidateNamedTextureHasParameters(textureObject, __func__)) return;
WithTemporarilyBoundNamedTexture(textureObject, [&](GLenum target) { GetTexParameteriv_State(target, pname, params); });
}
void GetTextureParameterfv(GLuint texture, GLenum pname, GLfloat* params) {
auto textureObject = GetTextureObjectByName(texture, __func__);
if (!ValidateNamedTextureHasParameters(textureObject, __func__)) return;
WithTemporarilyBoundNamedTexture(textureObject, [&](GLenum target) { GetTexParameterfv_State(target, pname, params); });
}
void GetTextureParameterIiv(GLuint texture, GLenum pname, GLint* params) {
auto textureObject = GetTextureObjectByName(texture, __func__);
if (!ValidateNamedTextureHasParameters(textureObject, __func__)) return;
WithTemporarilyBoundNamedTexture(textureObject, [&](GLenum target) { GetTexParameterIiv_State(target, pname, params); });
}
void GetTextureParameterIuiv(GLuint texture, GLenum pname, GLuint* params) {
auto textureObject = GetTextureObjectByName(texture, __func__);
if (!ValidateNamedTextureHasParameters(textureObject, __func__)) return;
WithTemporarilyBoundNamedTexture(textureObject, [&](GLenum target) { GetTexParameterIuiv_State(target, pname, params); });
}
@@ -4231,33 +4073,6 @@ namespace MobileGL::MG_Impl::GLImpl {
TexBuffer_State(target, internalformat, buffer);
}
// The buffer texture bound to `target` on the active unit - what the non-DSA range form
// operates on. Kept separate from TexBuffer_State because that one resolves the target
// itself and attaches the whole buffer.
static const SharedPtr<MG_State::GLState::ITextureObject>& GetBoundBufferTexture(GLenum target,
const char* caller) {
TextureUploadTarget uploadTarget = MG_Util::ConvertGLEnumToTextureUploadTarget(target);
if (!TextureImpl::ValidateTextureUploadTarget(uploadTarget)) return nullTextureObject;
(void)caller;
auto& activeUnit = MG_State::pGLContext->GetTextureUnitObject(MG_State::pGLContext->GetActiveTextureUnit());
return activeUnit.GetBindingSlot(MG_Util::ConvertGLEnumToTextureTarget(target)).GetBoundObject();
}
void TexBufferRange(GLenum target, GLenum internalformat, GLuint buffer, GLintptr offset, GLsizeiptr size) {
AttachBufferToTexture(GetBoundBufferTexture(target, __func__), internalformat, buffer, offset,
static_cast<SizeT>(size < 0 ? 0 : size), __func__);
}
void TextureBuffer(GLuint texture, GLenum internalformat, GLuint buffer) {
AttachBufferToTexture(GetTextureObjectByName(texture, __func__), internalformat, buffer, 0,
MG_State::GLState::TextureObjectBuffer::kWholeBuffer, __func__);
}
void TextureBufferRange(GLuint texture, GLenum internalformat, GLuint buffer, GLintptr offset, GLsizeiptr size) {
AttachBufferToTexture(GetTextureObjectByName(texture, __func__), internalformat, buffer, offset,
static_cast<SizeT>(size < 0 ? 0 : size), __func__);
}
GLboolean IsTexture(GLuint texture) {
return IsTexture_State(texture);
}
@@ -42,9 +42,6 @@ namespace MobileGL::MG_Impl::GLImpl {
void GenerateTextureMipmap(GLuint texture);
void BindTextureUnit(GLuint unit, GLuint texture);
void GetTextureImage(GLuint texture, GLint level, GLenum format, GLenum type, GLsizei bufSize, void* pixels);
void TexBufferRange(GLenum target, GLenum internalformat, GLuint buffer, GLintptr offset, GLsizeiptr size);
void TextureBuffer(GLuint texture, GLenum internalformat, GLuint buffer);
void TextureBufferRange(GLuint texture, GLenum internalformat, GLuint buffer, GLintptr offset, GLsizeiptr size);
void GetTextureSubImage(GLuint texture, GLint level, GLint xoffset, GLint yoffset, GLint zoffset, GLsizei width,
GLsizei height, GLsizei depth, GLenum format, GLenum type, GLsizei bufSize, void* pixels);
void GetTextureParameterfv(GLuint texture, GLenum pname, GLfloat* params);
@@ -105,45 +105,12 @@ namespace MobileGL::MG_Impl::GLImpl {
return pname == GL_CURRENT_VERTEX_ATTRIB;
}
// The stride a pointer-style call gives its binding point: the argument when it is non-zero,
// otherwise the tightly packed element size (GL 4.6 core 10.3.2). A packed 2_10_10_10 or
// 10F_11F_11F attribute is one 32-bit word regardless of its component count.
static int EffectiveVertexStride(GLsizei stride, GLint size, GLenum type) {
if (stride != 0) return static_cast<int>(stride);
switch (type) {
case GL_INT_2_10_10_10_REV:
case GL_UNSIGNED_INT_2_10_10_10_REV:
case GL_UNSIGNED_INT_10F_11F_11F_REV:
return 4;
default:
break;
}
return static_cast<int>(size * MG_Util::GetGLTypeSize(type));
}
// glBindVertexBuffers / glVertexArrayVertexBuffers take a range of binding points, and a
// range that runs past the last one is INVALID_OPERATION rather than the INVALID_VALUE a
// single out-of-range index gets (GL 4.6 core 10.3.1).
static bool ValidateVertexBindingRange(GLuint first, GLsizei count, const char* funcName) {
if (count < 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", funcName, "count must be non-negative."));
return false;
}
if (static_cast<Uint64>(first) + static_cast<Uint64>(count) >
VertexArrayImpl::GetMaxVertexAttribBindings()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", funcName,
"first + count exceeds GL_MAX_VERTEX_ATTRIB_BINDINGS."));
return false;
}
return true;
}
static bool ValidateVertexBindingIndex(GLuint bindingindex, const char* funcName) {
if (bindingindex >= VertexArrayImpl::GetMaxVertexAttribBindings()) {
// Bound by the same dynamic limit as attribute indices: the default attribute -> binding
// mapping is the identity, so a binding point the backend cannot address as an attribute
// would resolve into an attribute the backend must then reject on every draw. Real drivers
// likewise report MAX_VERTEX_ATTRIB_BINDINGS == MAX_VERTEX_ATTRIBS.
if (bindingindex >= VertexArrayImpl::GetMaxVertexAttribs()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", funcName,
@@ -188,17 +155,6 @@ namespace MobileGL::MG_Impl::GLImpl {
SharedPtr<MG_State::GLState::VertexArrayObject> GetNamedVertexArrayObject_State(GLuint vaobj,
const char* caller) {
// Name zero is not a vertex array object in a core profile: it names the default vertex
// array, which the by-name (direct state access) entry points never accept. MobileGL keeps a
// real object at index 0 for the compatibility paths, so the generic name validation below
// would otherwise let it through (GL 4.6 core 10.3.1).
if (vaobj == 0) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller,
"Vertex array name 0 is not a vertex array object."));
return nullptr;
}
if (!VertexArrayImpl::ValidateVertexArrayName(vaobj)) return nullptr;
if (!VertexArrayImpl::ValidateVertexArrayObject(vaobj)) return nullptr;
return MG_State::pGLContext->GetVertexArrayObject(vaobj);
@@ -254,7 +210,7 @@ namespace MobileGL::MG_Impl::GLImpl {
DataType dataType = MG_Util::ConvertGLEnumToDataType(type);
// Integer path: never normalized, never BGRA/packed (the validator rejects those).
if (!VertexArrayImpl::ValidateVertexAttribFormat(index, size, type, dataType, false, stride, true)) return;
if (!VertexArrayImpl::ValidateVertexAttribFormat(index, size, dataType, false, stride, true)) return;
auto& vao = MG_State::pGLContext->GetBoundVertexArray();
if (!vao) {
@@ -271,7 +227,6 @@ namespace MobileGL::MG_Impl::GLImpl {
vao->SetAttributeFormat(index, size, dataType, false, stride, offset, true, false);
vao->BindAttributeBuffer(index, vbo);
vao->MirrorPointerIntoBinding(index, vbo, offset, EffectiveVertexStride(stride, size, type));
}
void VertexAttribPointer_State(GLuint index, GLint size, GLenum type, GLboolean normalized, GLsizei stride,
@@ -279,7 +234,7 @@ namespace MobileGL::MG_Impl::GLImpl {
if (!VertexArrayImpl::ValidateVertexAttributeIndex(index)) return;
DataType dataType = MG_Util::ConvertGLEnumToDataType(type);
if (!VertexArrayImpl::ValidateVertexAttribFormat(index, size, type, dataType, normalized == GL_TRUE, stride, false))
if (!VertexArrayImpl::ValidateVertexAttribFormat(index, size, dataType, normalized == GL_TRUE, stride, false))
return;
auto& vao = MG_State::pGLContext->GetBoundVertexArray();
@@ -301,7 +256,6 @@ namespace MobileGL::MG_Impl::GLImpl {
const int effectiveSize = isBgra ? 4 : size;
vao->SetAttributeFormat(index, effectiveSize, dataType, normalized, stride, offset, false, isBgra);
vao->BindAttributeBuffer(index, vbo);
vao->MirrorPointerIntoBinding(index, vbo, offset, EffectiveVertexStride(stride, effectiveSize, type));
}
void BindVertexArray_State(GLuint array) {
@@ -405,13 +359,6 @@ namespace MobileGL::MG_Impl::GLImpl {
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller, "offset and stride must be non-negative."));
return;
}
if (static_cast<Uint>(stride) > VertexArrayImpl::GetMaxVertexAttribStride()) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", caller,
"stride exceeds GL_MAX_VERTEX_ATTRIB_STRIDE."));
return;
}
auto bufferObject = GetVertexArrayBufferObject_State(buffer, caller);
if (buffer != 0 && !bufferObject) return;
@@ -429,7 +376,6 @@ namespace MobileGL::MG_Impl::GLImpl {
const GLintptr* offsets, const GLsizei* strides) {
auto vao = GetNamedVertexArrayObject_State(vaobj, "VertexArrayVertexBuffers_State");
if (!vao) return;
if (!ValidateVertexBindingRange(first, count, "VertexArrayVertexBuffers_State")) return;
for (GLsizei i = 0; i < count; ++i) {
if (!buffers) {
VertexBufferBinding_State(vao, first + i, 0, 0, 16, "VertexArrayVertexBuffers_State");
@@ -443,36 +389,12 @@ namespace MobileGL::MG_Impl::GLImpl {
static void VertexAttribFormatSeparate_State(const SharedPtr<MG_State::GLState::VertexArrayObject>& vao,
GLuint attribindex, GLint size, GLenum type, GLboolean normalized,
GLuint relativeoffset, Bool isInteger, const char* caller) {
static_cast<void>(caller);
if (!VertexArrayImpl::ValidateVertexAttributeIndex(attribindex)) return;
DataType dataType = MG_Util::ConvertGLEnumToDataType(type);
// The separate-format entry points take the same size/type rules as the pointer ones,
// GL_BGRA included, so they need the full format validation rather than the pointer-only
// subset - that one reports GL_BGRA as an out-of-range size.
if (!VertexArrayImpl::ValidateVertexAttribFormat(attribindex, size, type, dataType, normalized == GL_TRUE, 0,
isInteger))
return;
if (!VertexArrayImpl::ValidateVertexAttribRelativeOffset(relativeoffset)) return;
if (!VertexArrayImpl::ValidateVertexAttribPointerParams(attribindex, size, dataType, 0)) return;
const Bool isBgra = (size == static_cast<GLint>(GL_BGRA));
vao->SetAttributeFormatSeparate(attribindex, isBgra ? 4 : size, dataType, normalized, isInteger,
relativeoffset, isBgra);
}
// The long (64-bit) attribute format. MobileGL has no 64-bit vertex attributes, so nothing is
// recorded; what the entry point owes the application is the parameter validation, which is
// observable through glGetError regardless of whether the format could be used in a draw.
static void VertexAttribLFormatSeparate_State(GLuint attribindex, GLint size, GLenum type,
GLuint relativeoffset) {
if (!VertexArrayImpl::ValidateVertexAttributeIndex(attribindex)) return;
if (!VertexArrayImpl::ValidateVertexAttribLFormat(attribindex, size, type)) return;
if (!VertexArrayImpl::ValidateVertexAttribRelativeOffset(relativeoffset)) return;
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", "VertexAttribLFormat",
"64-bit vertex attributes are not supported."));
vao->SetAttributeFormatSeparate(attribindex, size, dataType, normalized, isInteger, relativeoffset);
}
void VertexArrayAttribFormat_State(GLuint vaobj, GLuint attribindex, GLint size, GLenum type,
@@ -1122,83 +1044,6 @@ namespace MobileGL::MG_Impl::GLImpl {
VertexArrayVertexBuffer_State(vaobj, bindingindex, buffer, offset, stride);
}
// glGetVertexArrayiv reports exactly one thing (GL 4.6 core table 23.4): which buffer the
// named vertex array takes its indices from. Everything else about a vertex array is
// per-attribute and belongs to the indexed queries below.
void GetVertexArrayiv(GLuint vaobj, GLenum pname, GLint* param) {
auto vao = GetNamedVertexArrayObject_State(vaobj, __func__);
if (!vao || !param) return;
if (pname != GL_ELEMENT_ARRAY_BUFFER_BINDING) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"pname must be GL_ELEMENT_ARRAY_BUFFER_BINDING."));
return;
}
const auto& indexBuffer = vao->GetIndexBufferBindingSlot().GetBoundObject();
*param = indexBuffer ? static_cast<GLint>(indexBuffer->GetExternalIndex()) : 0;
}
void GetVertexArrayIndexediv(GLuint vaobj, GLuint index, GLenum pname, GLint* param) {
auto vao = GetNamedVertexArrayObject_State(vaobj, __func__);
if (!vao || !param) return;
if (!VertexArrayImpl::ValidateVertexAttributeIndex(index)) return;
const auto& attr = vao->GetAttribute(index);
switch (pname) {
case GL_VERTEX_ATTRIB_ARRAY_ENABLED:
*param = attr.Enabled ? GL_TRUE : GL_FALSE;
return;
case GL_VERTEX_ATTRIB_ARRAY_SIZE:
*param = static_cast<GLint>(attr.Size);
return;
case GL_VERTEX_ATTRIB_ARRAY_STRIDE:
*param = static_cast<GLint>(attr.Stride);
return;
case GL_VERTEX_ATTRIB_ARRAY_TYPE:
*param = static_cast<GLint>(MG_Util::ConvertDataTypeToGLEnum(attr.Type));
return;
case GL_VERTEX_ATTRIB_ARRAY_NORMALIZED:
*param = attr.Normalized ? GL_TRUE : GL_FALSE;
return;
case GL_VERTEX_ATTRIB_ARRAY_INTEGER:
*param = attr.IsInteger ? GL_TRUE : GL_FALSE;
return;
case GL_VERTEX_ATTRIB_ARRAY_LONG:
// 64-bit attributes are not supported, so no attribute is ever a long one.
*param = GL_FALSE;
return;
case GL_VERTEX_ATTRIB_ARRAY_DIVISOR:
*param = static_cast<GLint>(attr.Divisor);
return;
case GL_VERTEX_ATTRIB_RELATIVE_OFFSET:
*param = static_cast<GLint>(vao->GetAttributeRelativeOffset(index));
return;
default:
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__,
"pname is not an accepted indexed vertex array query."));
return;
}
}
// Only GL_VERTEX_BINDING_OFFSET needs 64 bits. Its `index` names a vertex buffer binding
// point directly (GL 4.6 core 10.3.1), not an attribute - unlike every pname the 32-bit
// indexed query above accepts, which is why this one does not go through an attribute's
// binding index.
void GetVertexArrayIndexed64iv(GLuint vaobj, GLuint index, GLenum pname, GLint64* param) {
auto vao = GetNamedVertexArrayObject_State(vaobj, __func__);
if (!vao || !param) return;
if (!VertexArrayImpl::ValidateVertexAttributeIndex(index)) return;
if (pname != GL_VERTEX_BINDING_OFFSET) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>("MG_Impl/GLImpl", __func__, "pname must be GL_VERTEX_BINDING_OFFSET."));
return;
}
*param = static_cast<GLint64>(vao->GetBindingPoint(index).Offset);
}
void VertexArrayAttribFormat(GLuint vaobj, GLuint attribindex, GLint size, GLenum type, GLboolean normalized,
GLuint relativeoffset) {
VertexArrayAttribFormat_State(vaobj, attribindex, size, type, normalized, relativeoffset);
@@ -1231,7 +1076,6 @@ namespace MobileGL::MG_Impl::GLImpl {
const GLsizei* strides) {
auto vao = GetBoundVertexArrayOrError("BindVertexBuffers");
if (!vao) return;
if (!ValidateVertexBindingRange(first, count, "BindVertexBuffers")) return;
for (GLsizei i = 0; i < count; ++i) {
if (!buffers) {
VertexBufferBinding_State(vao, first + i, 0, 0, 16, "BindVertexBuffers");
@@ -1256,18 +1100,6 @@ namespace MobileGL::MG_Impl::GLImpl {
"VertexAttribIFormat");
}
void VertexAttribLFormat(GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset) {
auto vao = GetBoundVertexArrayOrError("VertexAttribLFormat");
if (!vao) return;
VertexAttribLFormatSeparate_State(attribindex, size, type, relativeoffset);
}
void VertexArrayAttribLFormat(GLuint vaobj, GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset) {
auto vao = GetNamedVertexArrayObject_State(vaobj, "VertexArrayAttribLFormat");
if (!vao) return;
VertexAttribLFormatSeparate_State(attribindex, size, type, relativeoffset);
}
void VertexAttribBinding(GLuint attribindex, GLuint bindingindex) {
auto vao = GetBoundVertexArrayOrError("VertexAttribBinding");
if (!vao) return;
@@ -92,13 +92,9 @@ namespace MobileGL::MG_Impl::GLImpl {
void EnableVertexArrayAttrib(GLuint vaobj, GLuint index);
void VertexArrayElementBuffer(GLuint vaobj, GLuint buffer);
void VertexArrayVertexBuffer(GLuint vaobj, GLuint bindingindex, GLuint buffer, GLintptr offset, GLsizei stride);
void GetVertexArrayiv(GLuint vaobj, GLenum pname, GLint* param);
void GetVertexArrayIndexediv(GLuint vaobj, GLuint index, GLenum pname, GLint* param);
void GetVertexArrayIndexed64iv(GLuint vaobj, GLuint index, GLenum pname, GLint64* param);
void VertexArrayAttribFormat(GLuint vaobj, GLuint attribindex, GLint size, GLenum type, GLboolean normalized,
GLuint relativeoffset);
void VertexArrayAttribIFormat(GLuint vaobj, GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset);
void VertexArrayAttribLFormat(GLuint vaobj, GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset);
void VertexArrayAttribBinding(GLuint vaobj, GLuint attribindex, GLuint bindingindex);
void VertexArrayBindingDivisor(GLuint vaobj, GLuint bindingindex, GLuint divisor);
void VertexArrayVertexBuffers(GLuint vaobj, GLuint first, GLsizei count, const GLuint* buffers,
@@ -108,7 +104,6 @@ namespace MobileGL::MG_Impl::GLImpl {
const GLsizei* strides);
void VertexAttribFormat(GLuint attribindex, GLint size, GLenum type, GLboolean normalized, GLuint relativeoffset);
void VertexAttribIFormat(GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset);
void VertexAttribLFormat(GLuint attribindex, GLint size, GLenum type, GLuint relativeoffset);
void VertexAttribBinding(GLuint attribindex, GLuint bindingindex);
void VertexBindingDivisor(GLuint bindingindex, GLuint divisor);
void VertexAttribDivisor(GLuint index, GLuint divisor);
@@ -23,18 +23,6 @@ namespace MobileGL::MG_Impl::GLImpl::VertexArrayImpl {
return std::min(static_cast<Uint>(backendLimit), capacity);
}
Uint GetMaxVertexAttribBindings() {
return GetMaxVertexAttribs();
}
Uint GetMaxVertexAttribRelativeOffset() {
return 2047;
}
Uint GetMaxVertexAttribStride() {
return 2048;
}
Bool ValidateVertexArrayName(Uint index) {
Bool isValid = MG_State::pGLContext->ValidateVertexArrayName(index);
if (!isValid) {
@@ -102,31 +90,9 @@ namespace MobileGL::MG_Impl::GLImpl::VertexArrayImpl {
return true;
}
Bool ValidateVertexAttribFormat(Uint index, GLint sizeRaw, GLenum glType, DataType type, Bool normalized,
Int stride, Bool integerPath) {
Bool ValidateVertexAttribFormat(Uint index, GLint sizeRaw, DataType type, Bool normalized, Int stride,
Bool integerPath) {
constexpr const char* fn = "ValidateVertexAttribFormat";
// GL_UNSIGNED_INT_10F_11F_11F_REV is a three-component float-path-only packing that has no
// DataType of its own, so it has to be recognised by name before the conversion below turns
// it into Unknown and reports the wrong error (GL 4.6 core 10.3.2).
if (glType == GL_UNSIGNED_INT_10F_11F_11F_REV) {
if (integerPath) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", fn,
std::format("GL_UNSIGNED_INT_10F_11F_11F_REV is not an integer-path type (attribute {}).",
index)));
return false;
}
if (sizeRaw != 3) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidOperation,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", fn,
std::format("GL_UNSIGNED_INT_10F_11F_11F_REV requires size 3 (attribute {}).", index)));
return false;
}
}
if (type == DataType::Unknown) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
@@ -204,40 +170,4 @@ namespace MobileGL::MG_Impl::GLImpl::VertexArrayImpl {
}
return true;
}
Bool ValidateVertexAttribLFormat(Uint index, GLint size, GLenum type) {
constexpr const char* fn = "ValidateVertexAttribLFormat";
if (size < 1 || size > 4) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", fn,
std::format("Invalid size {} for attribute {}. Must be 1-4.", size, index)));
return false;
}
// GL 4.6 core 10.3.2: the long form takes GL_DOUBLE and nothing else.
if (type != GL_DOUBLE) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidEnum,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", fn,
std::format("Type 0x{:X} is not GL_DOUBLE (attribute {}).", type, index)));
return false;
}
return true;
}
Bool ValidateVertexAttribRelativeOffset(Uint relativeOffset) {
const Uint limit = GetMaxVertexAttribRelativeOffset();
if (relativeOffset > limit) {
MG_State::pGLContext->RecordError(
ErrorCode::InvalidValue,
MakeUnique<GenericErrorInfo>(
"MG_Impl/GLImpl", "ValidateVertexAttribRelativeOffset",
std::format("relativeoffset {} exceeds GL_MAX_VERTEX_ATTRIB_RELATIVE_OFFSET ({}).", relativeOffset,
limit)));
return false;
}
return true;
}
} // namespace MobileGL::MG_Impl::GLImpl::VertexArrayImpl
@@ -15,20 +15,6 @@ namespace MobileGL::MG_Impl::GLImpl::VertexArrayImpl {
// capacity). Falls back to the capacity when no backend is active (unit tests).
Uint GetMaxVertexAttribs();
// GL_MAX_VERTEX_ATTRIB_BINDINGS. The default attribute -> binding mapping is the identity, so a
// binding point that cannot also be an attribute index would resolve into an attribute the
// backend has to reject on every draw; real drivers report the two limits equal as well.
Uint GetMaxVertexAttribBindings();
// GL_MAX_VERTEX_ATTRIB_RELATIVE_OFFSET. The relative offset is folded into the resolved
// attribute offset in the frontend and never reaches a backend limit, so this is the value the
// spec requires an implementation to support at minimum (GL 4.6 core table 23.63).
Uint GetMaxVertexAttribRelativeOffset();
// GL_MAX_VERTEX_ATTRIB_STRIDE. Like the relative offset above, the stride never reaches a
// backend limit of its own, so this is the spec minimum (GL 4.6 core table 23.63).
Uint GetMaxVertexAttribStride();
Bool ValidateVertexArrayName(Uint index);
Bool ValidateVertexArrayObject(Uint index);
Bool ValidateVertexAttributeIndex(Uint index);
@@ -36,13 +22,6 @@ namespace MobileGL::MG_Impl::GLImpl::VertexArrayImpl {
// Full glVertexAttribPointer / glVertexAttribIPointer format validation, including the packed
// 2_10_10_10 types and GL_BGRA size. sizeRaw is the untranslated GL size (possibly GL_BGRA);
// integerPath selects the glVertexAttribIPointer rules.
Bool ValidateVertexAttribFormat(Uint index, GLint sizeRaw, GLenum glType, DataType type, Bool normalized,
Int stride, Bool integerPath);
// glVertexAttribLFormat / glVertexArrayAttribLFormat: the only accepted type is GL_DOUBLE and
// the size range is 1-4 (GL_BGRA is a float-path size). Separate from the function above
// because the long path shares none of its type or size rules.
Bool ValidateVertexAttribLFormat(Uint index, GLint size, GLenum type);
// Shared by every *Format entry point: INVALID_VALUE once relativeoffset leaves the range the
// implementation advertises.
Bool ValidateVertexAttribRelativeOffset(Uint relativeOffset);
Bool ValidateVertexAttribFormat(Uint index, GLint sizeRaw, DataType type, Bool normalized, Int stride,
Bool integerPath);
} // namespace MobileGL::MG_Impl::GLImpl::VertexArrayImpl
-2
View File
@@ -85,8 +85,6 @@ namespace MobileGL::MG_Impl {
GETPROC(CGLGetPixelFormat, name);
GETPROC(CGLSetCurrentContext, name);
GETPROC(CGLGetCurrentContext, name);
GETPROC(CGLSetVirtualScreen, name);
GETPROC(CGLGetVirtualScreen, name);
GETPROC(CGLSetParameter, name);
GETPROC(CGLGetParameter, name);
GETPROC(CGLUpdateContext, name);
+4 -36
View File
@@ -29,19 +29,10 @@ namespace MobileGL::MG_Impl::NSOpenGLImpl {
char kContextViewKey;
char kContextLayerKey;
std::once_flag g_installOnce;
IMP g_pixelFormatDealloc = nullptr;
IMP g_contextDealloc = nullptr;
std::mutex& HookInstallMutex() {
static auto* mutex = new std::mutex();
return *mutex;
}
Bool& HooksInstalled() {
static auto* installed = new Bool(false);
return *installed;
}
template <typename Fn>
Fn ObjcMsgSend() {
return reinterpret_cast<Fn>(objc_msgSend);
@@ -440,12 +431,12 @@ namespace MobileGL::MG_Impl::NSOpenGLImpl {
method_setImplementation(method, replacement);
}
Bool InstallHooksOnce() {
void InstallHooksOnce() {
Class pixelFormatClass = objc_getClass("NSOpenGLPixelFormat");
Class contextClass = objc_getClass("NSOpenGLContext");
if (!pixelFormatClass || !contextClass) {
MGLOG_W("NSOpenGLImpl: NSOpenGL classes are not loaded; hooks not installed");
return false;
return;
}
ReplaceInstanceMethod(pixelFormatClass, "initWithAttributes:",
@@ -480,34 +471,11 @@ namespace MobileGL::MG_Impl::NSOpenGLImpl {
ReplaceInstanceMethod(contextClass, "dealloc", reinterpret_cast<IMP>(ContextDealloc), &g_contextDealloc);
MGLOG_I("NSOpenGLImpl hooks installed");
return true;
}
} // namespace
void InstallHooks() {
const std::lock_guard<std::mutex> lock(HookInstallMutex());
if (!HooksInstalled()) {
// Do not permanently consume the install attempt when the OpenGL
// framework has not registered its Objective-C classes yet. The
// dyld bootstrap normally runs after framework dependencies, but
// an explicitly loaded/static-linked MobileGL can arrive earlier.
HooksInstalled() = InstallHooksOnce();
}
std::call_once(g_installOnce, InstallHooksOnce);
}
} // namespace MobileGL::MG_Impl::NSOpenGLImpl
namespace {
// SDL's Cocoa backend creates NSOpenGLPixelFormat/NSOpenGLContext before
// its first dlsym("glGetString") or other MobileGL host-API call. Install
// only the lightweight Objective-C dispatch hooks while the injected dylib
// is loading so those first Cocoa objects are routed through CGLImpl. The
// hooked context constructor reaches EGLImpl::GetDisplay(), which performs
// the full, thread-safe MobileGL initialization outside this bootstrap.
//
// There is intentionally no matching destructor: backend teardown remains
// owned by the EGL lifecycle and process-exit globals remain leak-at-exit.
__attribute__((constructor)) void BootstrapNSOpenGLHooks() {
MobileGL::MG_Impl::NSOpenGLImpl::InstallHooks();
}
} // namespace
#endif
@@ -174,21 +174,6 @@ namespace MobileGL::MG_State::GLState {
++m_changeSerial;
}
void BufferObject::MarkGpuWritten() {
m_gpuWritePending = true;
}
void BufferObject::SyncGpuWrites() {
if (!m_gpuWritePending) return;
// Cleared unconditionally: without a readback op the shadow can never catch up,
// and retrying on every subsequent read would only repeat the same no-op.
m_gpuWritePending = false;
if (m_size == 0 || g_bufferBackendOps == nullptr || g_bufferBackendOps->ReadbackFromGpu == nullptr) {
return;
}
g_bufferBackendOps->ReadbackFromGpu(*this);
}
void BufferObject::UploadSubData(DataPtr data, SizeT atOffset) {
MOBILEGL_ASSERT(!m_isMapped || (m_mappingAccess & BufferMappingAccessBit::Persistent),
"Cannot upload sub data while buffer is non-persistently mapped.");
@@ -219,13 +204,11 @@ namespace MobileGL::MG_State::GLState {
"Destination buffer copy out of bounds: dstOffset (%zu) + size (%zu) > m_size (%zu)", dstOffset,
size, m_size);
src->SyncGpuWrites();
Memcpy(m_resource.Bytes() + dstOffset, src->m_resource.Bytes() + srcOffset, size);
NotifyContentWrite(dstOffset, size);
}
void* BufferObject::AcquireMemory(Bool markMapped, Bool read, Bool write) {
SyncGpuWrites();
if (markMapped) {
m_isMapped = true;
m_mappingAccess = (read ? BufferMappingAccessBit::Read : BufferMappingAccessBit::Null) |
@@ -248,29 +231,10 @@ namespace MobileGL::MG_State::GLState {
return m_resource.Bytes();
}
Bool BufferObject::EnsureGpuResidentStorage() {
if (m_resource.IsGpuResident()) {
return true;
}
if (m_size == 0 || g_bufferBackendOps == nullptr || g_bufferBackendOps->AcquirePersistentMap == nullptr) {
return false;
}
void* base = g_bufferBackendOps->AcquirePersistentMap(*this);
if (base == nullptr) {
return false;
}
m_resource.AdoptPersistentMap(base);
return true;
}
void* BufferObject::AcquireMemoryRange(Range1D range, Flags<BufferMappingAccessBit> access) {
MOBILEGL_ASSERT(range.end <= m_size && range.start <= range.end,
"AcquireMemoryRange out of bounds: range (%zu, %zu) exceeds m_size (%zu)", range.start,
range.end, m_size);
// The app is about to look at the bytes; a shader may have rewritten them since
// the shadow was last authoritative. Also needed for a write map without an
// invalidate bit, whose staging copy is seeded from the shadow.
SyncGpuWrites();
m_isMapped = true;
m_mappingAccess = access;
m_mappedRange = range;
@@ -99,13 +99,6 @@ namespace MobileGL {
// Must be idempotent: a second call for an already-backed buffer returns the
// same base pointer.
void* (*AcquirePersistentMap)(BufferObject& bufferObject) = nullptr;
// Pulls the backend's current contents for the whole buffer into the shadow
// (through WritebackFromBackend). Only ever called for a buffer the GPU may
// have written behind the frontend's back - a shader storage or atomic counter
// binding of a draw or dispatch - because nothing else can desynchronise the
// shadow. Backends that cannot read their storage back leave this null; the
// shadow then keeps its pre-dispatch bytes, which is the old behaviour.
void (*ReadbackFromGpu)(BufferObject& bufferObject) = nullptr;
};
// Registered by the active backend at init, cleared at shutdown.
@@ -140,11 +133,6 @@ namespace MobileGL {
void* AcquireMemory(Bool markMapped, Bool read, Bool write);
void* AcquireMemoryRange(Range1D range, Flags<BufferMappingAccessBit> access);
// Adopt backend host-visible coherent GPU storage as the source of truth
// (used for GPU-written targets like transform feedback capture, so
// MapBuffer/GetBufferSubData read real GPU results). No-op when already
// resident or when the backend declines.
Bool EnsureGpuResidentStorage();
void ReleaseMemory();
void FlushMemoryRange(SizeT offset, SizeT length);
@@ -156,16 +144,6 @@ namespace MobileGL {
// backend op: the backend storage already holds these bytes.
void WritebackFromBackend(DataPtr data, SizeT atOffset);
// A draw or dispatch just ran with this buffer bound where a shader can write
// it (shader storage / atomic counter). The next read has to reconcile with
// that: pull the bytes back, or - when the shadow already IS coherent GPU
// memory - wait for the work that wrote them to retire. Which of the two is
// the backend's business; the flag only says a GPU write is outstanding.
void MarkGpuWritten();
// Refreshes the shadow from the backend when a GPU write is outstanding. Called
// from every path that reads the shadow on the app's behalf.
void SyncGpuWrites();
Bool IsMapped() const;
Bool IsImmutableStorage() const;
SizeT GetSize() const;
@@ -213,8 +191,6 @@ namespace MobileGL {
Bool m_isImmutableStorage = false;
GLbitfield m_storageFlags = 0;
Uint64 m_changeSerial = 0;
// Set by MarkGpuWritten, cleared by SyncGpuWrites once the shadow is refreshed.
Bool m_gpuWritePending = false;
Range1D m_mappedRange;
Vector<Uint8> m_stagingData;
Bool m_ownsStagingData;
@@ -31,9 +31,6 @@ namespace MobileGL::MG_State::GLState {
BindingSlot<BufferObject>& GetBindingSlot(BufferTarget target);
// For glBindBufferBase / glBindBufferRange
BindingSlotRange1D<BufferObject>& GetBindingPoint(BufferTarget target, Uint index);
const BindingSlotRange1D<BufferObject>& GetBindingPoint(BufferTarget target, Uint index) const {
return const_cast<BufferState*>(this)->GetBindingPoint(target, index);
}
constexpr SizeT GetBindingPointCount(const BufferTarget target) const {
auto it = std::find(BufferBindPointTargets.begin(), BufferBindPointTargets.end(), target);
auto index = std::distance(BufferBindPointTargets.begin(), it);
-181
View File
@@ -241,32 +241,6 @@ namespace MobileGL::MG_State {
}
void GLContext::MarkTextureObjectForDeletion(Uint index) {
// GL 3.3 core 4.4.2: deleting a texture whose image is attached to the framebuffer
// that is currently bound acts as if FramebufferTexture* had been called with texture
// zero for every attachment point it occupied there. Framebuffers that are NOT bound
// keep the orphaned attachment, so only the bound ones are touched.
//
// Without this the framebuffer object goes on holding the deleted texture alive as its
// attachment, and a later read through that framebuffer returns the dead texture's
// contents rather than those of whatever the application put in its place - the name
// it deleted usually comes straight back from the next glGenTextures, so the two are
// indistinguishable from the outside (KHR-GL32.packed_pixels read a stale gradient).
if (const auto& textureObject = m_textureState.GetTextureObject(index)) {
for (SizeT targetIndex = 0; targetIndex < SizeT(FramebufferTarget::FramebufferTargetCount);
++targetIndex) {
const auto& framebuffer =
GetFramebufferBindingSlot(static_cast<FramebufferTarget>(targetIndex)).GetBoundObject();
if (!framebuffer || framebuffer->IsDefaultFramebuffer()) {
continue;
}
const auto& attachments = framebuffer->GetAllAttachmentObjects();
for (SizeT i = 0; i < attachments.size(); ++i) {
if (attachments[i].IsTexture() && attachments[i].GetTexture() == textureObject) {
framebuffer->Detach(static_cast<FramebufferAttachmentType>(i));
}
}
}
}
m_textureState.MarkTextureObjectForDeletion(index, IsRelaxedSemanticsActive());
}
@@ -424,14 +398,6 @@ namespace MobileGL::MG_State {
m_renderState.SetPointSize(size);
}
void GLContext::SetPatchVertices(Uint vertices) {
m_renderState.SetPatchVertices(vertices);
}
Uint GLContext::GetPatchVertices() const {
return m_renderState.GetPatchVertices();
}
Float GLContext::GetPointSize() const {
return m_renderState.GetPointSize();
}
@@ -753,153 +719,6 @@ namespace MobileGL::MG_State {
Bool GLContext::ValidateRenderbufferObject(Uint index) const {
return m_renderbufferState.ValidateRenderbufferObject(index);
}
void GLContext::SaveBoundTransformFeedbackState() {
auto& object = m_transformFeedbackObjects[m_boundTransformFeedback];
for (Uint i = 0; i < MAX_TRANSFORM_FEEDBACK_BUFFERS; ++i) {
const auto& point = m_bufferState.GetBindingPoint(BufferTarget::TransformFeedback, i);
object.bindings[i] = {point.GetBoundObject(), point.GetRange(), point.HasExplicitRange()};
}
object.active = m_transformFeedbackActive;
object.paused = m_transformFeedbackPaused;
object.primitiveMode = m_transformFeedbackPrimitiveMode;
object.program = m_transformFeedbackProgram;
object.generation = m_transformFeedbackGeneration;
object.capturedVertices = m_transformFeedbackCapturedVertices;
object.inputPrimitives = m_transformFeedbackInputPrimitives;
}
void GLContext::RestoreBoundTransformFeedbackState() {
const auto& object = m_transformFeedbackObjects[m_boundTransformFeedback];
for (Uint i = 0; i < MAX_TRANSFORM_FEEDBACK_BUFFERS; ++i) {
auto& point = m_bufferState.GetBindingPoint(BufferTarget::TransformFeedback, i);
point.Bind(object.bindings[i].buffer);
if (object.bindings[i].buffer) {
point.SetRange(object.bindings[i].range, object.bindings[i].hasExplicitRange);
} else {
point.ClearRange();
}
}
m_transformFeedbackActive = object.active;
m_transformFeedbackPaused = object.paused;
m_transformFeedbackPrimitiveMode = object.primitiveMode;
m_transformFeedbackProgram = object.program;
// The generation identifies one capture span, and a span belongs to the object
// that opened it - a backend keys its append state on it, so switching objects
// has to bring the right one back.
m_transformFeedbackGeneration = object.generation;
m_transformFeedbackCapturedVertices = object.capturedVertices;
m_transformFeedbackInputPrimitives = object.inputPrimitives;
}
void GLContext::GenTransformFeedbackNames(Uint number, Vector<Uint>& ids) {
ids.resize(number);
if (number == 0) return;
m_transformFeedbackNames.Generate(number, ids.data());
// A generated name already denotes an object with the default state, so that a
// bind never has to distinguish "first use" from any later one.
for (const Uint id : ids) {
m_transformFeedbackObjects[id] = {};
}
}
Bool GLContext::ValidateTransformFeedbackName(Uint index) const {
return index == 0 || m_transformFeedbackNames.IsValid(index);
}
void GLContext::BindTransformFeedbackObject(Uint index) {
if (index == m_boundTransformFeedback) return;
SaveBoundTransformFeedbackState();
m_boundTransformFeedback = index;
m_transformFeedbackObjects[index].everBound = true;
RestoreBoundTransformFeedbackState();
}
Bool GLContext::IsTransformFeedbackObject(Uint index) const {
if (index == 0 || !m_transformFeedbackNames.IsValid(index)) return false;
const auto it = m_transformFeedbackObjects.find(index);
return it != m_transformFeedbackObjects.end() && it->second.everBound;
}
void GLContext::MarkTransformFeedbackObjectForDeletion(Uint index) {
if (index == 0 || !m_transformFeedbackNames.IsValid(index)) return;
// Deleting the bound object reverts to the default one (GL 4.6 core 13.2.1);
// its state is dropped rather than saved back into the dying object.
if (index == m_boundTransformFeedback) {
m_boundTransformFeedback = 0;
RestoreBoundTransformFeedbackState();
}
m_transformFeedbackObjects.erase(index);
m_transformFeedbackNames.Delete(index);
}
Uint64 GLContext::GetTransformFeedbackRecordedVertices(Uint index) const {
const auto it = m_transformFeedbackObjects.find(index);
return it == m_transformFeedbackObjects.end() ? 0 : it->second.recordedVertices;
}
Bool GLContext::HasTransformFeedbackCompletedSpan(Uint index) const {
const auto it = m_transformFeedbackObjects.find(index);
return it != m_transformFeedbackObjects.end() && it->second.hasCompletedSpan;
}
void GLContext::CreateTransformFeedbackObject(Uint index) {
// glCreateTransformFeedbacks has no bind step to infer existence from, so the name it
// hands out is already the name of an object (GL 4.6 core 13.2.1).
m_transformFeedbackObjects[index] = {};
m_transformFeedbackObjects[index].everBound = true;
}
Bool GLContext::IsNamedTransformFeedbackActive(Uint index) const {
if (index == m_boundTransformFeedback) return m_transformFeedbackActive;
const auto it = m_transformFeedbackObjects.find(index);
return it != m_transformFeedbackObjects.end() && it->second.active;
}
Bool GLContext::IsNamedTransformFeedbackPaused(Uint index) const {
if (index == m_boundTransformFeedback) return m_transformFeedbackPaused;
const auto it = m_transformFeedbackObjects.find(index);
return it != m_transformFeedbackObjects.end() && it->second.paused;
}
NamedTransformFeedbackBinding GLContext::GetNamedTransformFeedbackBinding(Uint index, Uint bufferIndex) const {
NamedTransformFeedbackBinding result;
if (bufferIndex >= MAX_TRANSFORM_FEEDBACK_BUFFERS) return result;
// The bound object's capture bindings live in the context's own binding points, not in
// the saved copy - that one is only written when the object is swapped out.
if (index == m_boundTransformFeedback) {
const auto& point = m_bufferState.GetBindingPoint(BufferTarget::TransformFeedback, bufferIndex);
result.Buffer = point.GetBoundObject();
result.Range = point.GetRange();
result.HasExplicitRange = point.HasExplicitRange();
return result;
}
const auto it = m_transformFeedbackObjects.find(index);
if (it == m_transformFeedbackObjects.end()) return result;
const auto& saved = it->second.bindings[bufferIndex];
result.Buffer = saved.buffer;
result.Range = saved.range;
result.HasExplicitRange = saved.hasExplicitRange;
return result;
}
void GLContext::SetNamedTransformFeedbackBinding(Uint index, Uint bufferIndex,
const SharedPtr<BufferObject>& buffer, Range1D range,
Bool hasExplicitRange) {
if (bufferIndex >= MAX_TRANSFORM_FEEDBACK_BUFFERS) return;
if (index == m_boundTransformFeedback) {
auto& point = m_bufferState.GetBindingPoint(BufferTarget::TransformFeedback, bufferIndex);
point.Bind(buffer);
if (buffer && hasExplicitRange) {
point.SetRange(range, true);
} else {
point.ClearRange();
}
return;
}
auto& object = m_transformFeedbackObjects[index];
object.bindings[bufferIndex] = {buffer, range, hasExplicitRange};
}
} // namespace GLState
// Leak-at-exit storage; see GlobalObjects.cpp.
-147
View File
@@ -45,14 +45,6 @@ namespace MobileGL {
// translates the result into its own API call.
VertexAttribTypeInfo ClassifyVertexAttribType(GLenum glType);
// One indexed capture binding of a transform feedback object, as the by-name queries
// report it. An empty Buffer means the binding point is unbound.
struct NamedTransformFeedbackBinding {
SharedPtr<BufferObject> Buffer;
Range1D Range{};
Bool HasExplicitRange = false;
};
class GLContext {
public:
GLContext() = default;
@@ -146,8 +138,6 @@ namespace MobileGL {
Float GetLineWidth() const;
void SetPointSize(Float size);
Float GetPointSize() const;
void SetPatchVertices(Uint vertices);
Uint GetPatchVertices() const;
void SetPolygonOffset(Float factor, Float units);
Float GetPolygonOffsetFactor() const;
Float GetPolygonOffsetUnits() const;
@@ -221,103 +211,6 @@ namespace MobileGL {
void SetScissorBox(IntVec4 box); // x, y, width, height
const IntVec4& GetScissorBox() const; // x, y, width, height
// Transform feedback. The fields below are the state of the transform
// feedback object currently bound to GL_TRANSFORM_FEEDBACK; see the object
// block further down for how a bind swaps them.
void BeginTransformFeedback(GLenum primitiveMode, const SharedPtr<ProgramObject>& program) {
m_transformFeedbackActive = true;
m_transformFeedbackPaused = false;
m_transformFeedbackPrimitiveMode = primitiveMode;
m_transformFeedbackProgram = program;
m_transformFeedbackGeneration = ++m_transformFeedbackNextGeneration;
m_transformFeedbackCapturedVertices = 0;
m_transformFeedbackInputPrimitives = 0;
}
void EndTransformFeedback() {
m_transformFeedbackActive = false;
m_transformFeedbackPaused = false;
m_transformFeedbackProgram.reset();
// What glDrawTransformFeedback on this object replays from now on.
auto& object = m_transformFeedbackObjects[m_boundTransformFeedback];
object.recordedVertices = m_transformFeedbackCapturedVertices;
object.hasCompletedSpan = true;
}
Bool IsTransformFeedbackActive() const { return m_transformFeedbackActive; }
Bool IsTransformFeedbackPaused() const { return m_transformFeedbackPaused; }
void SetTransformFeedbackPaused(Bool paused) { m_transformFeedbackPaused = paused; }
GLenum GetTransformFeedbackPrimitiveMode() const { return m_transformFeedbackPrimitiveMode; }
const SharedPtr<ProgramObject>& GetTransformFeedbackProgram() const {
return m_transformFeedbackProgram;
}
// Bumped on every BeginTransformFeedback; the backend uses it to
// distinguish "resume appending" from "fresh capture".
Uint64 GetTransformFeedbackGeneration() const { return m_transformFeedbackGeneration; }
// CPU-side primitive accounting for the transform feedback queries:
// every captured draw adds its primitive count (draws without a
// geometry stage write exactly what they generate).
void AddTransformFeedbackPrimitives(Uint64 primitives) {
m_transformFeedbackPrimitiveCounter += primitives;
}
Uint64 GetTransformFeedbackPrimitiveCounter() const { return m_transformFeedbackPrimitiveCounter; }
// Primitives a draw assembled while the capture was paused. GL counts those in
// PRIMITIVES_GENERATED, but a backend that answers the query with its own
// transform feedback counter cannot see them - nothing was being captured.
void AddTransformFeedbackPausedPrimitives(Uint64 primitives) {
m_transformFeedbackPausedPrimitiveCounter += primitives;
}
Uint64 GetTransformFeedbackPausedPrimitiveCounter() const {
return m_transformFeedbackPausedPrimitiveCounter;
}
// Vertices already captured since BeginTransformFeedback (drives the
// buffer-capacity clamp on the primitives-written accounting).
void AddTransformFeedbackCapturedVertices(Uint64 vertices) {
m_transformFeedbackCapturedVertices += vertices;
}
Uint64 GetTransformFeedbackCapturedVertices() const { return m_transformFeedbackCapturedVertices; }
// Raw assembled input primitives fed to the capture stage since Begin
// (pre-clamp; drives the GS strip capture-order fixup at EndTF).
void AddTransformFeedbackInputPrimitives(Uint64 primitives) {
m_transformFeedbackInputPrimitives += primitives;
}
Uint64 GetTransformFeedbackInputPrimitives() const { return m_transformFeedbackInputPrimitives; }
// Transform feedback objects (ARB_transform_feedback2 / GL 4.0 core).
// The capture state above and the indexed GL_TRANSFORM_FEEDBACK_BUFFER
// binding points are object state, but the context keeps exactly one live
// copy of both so that every existing reader - the backends' per-draw sync,
// the drawing and getter paths - needs no notion of which object owns them.
// A bind therefore saves the live copy into the outgoing object and restores
// the incoming one's. Object 0 is the default object and always exists.
static constexpr Uint MAX_TRANSFORM_FEEDBACK_BUFFERS = 4;
void GenTransformFeedbackNames(Uint number, Vector<Uint>& ids);
// A name glGenTransformFeedbacks handed out and glDeleteTransformFeedbacks
// has not taken back. Name 0 is always valid.
Bool ValidateTransformFeedbackName(Uint index) const;
// What glIsTransformFeedback reports: a generated name only becomes the name
// of an object once it has been bound at least once (GL 4.6 core 13.2.1).
Bool IsTransformFeedbackObject(Uint index) const;
void BindTransformFeedbackObject(Uint index);
void MarkTransformFeedbackObjectForDeletion(Uint index);
Uint GetBoundTransformFeedbackName() const { return m_boundTransformFeedback; }
// Vertices the object captured in its last completed span; the vertex count
// glDrawTransformFeedback replays.
Uint64 GetTransformFeedbackRecordedVertices(Uint index) const;
// Whether the object has ever completed a capture span. glDrawTransformFeedback
// on an object that has not is INVALID_OPERATION, which a zero vertex count
// cannot express: an empty completed span is legal and draws nothing.
Bool HasTransformFeedbackCompletedSpan(Uint index) const;
// The by-name (direct state access) view. A named object that happens to be the
// bound one is answered from the live copy, since that is where its state actually
// is until a bind swaps it out.
void CreateTransformFeedbackObject(Uint index);
Bool IsNamedTransformFeedbackActive(Uint index) const;
Bool IsNamedTransformFeedbackPaused(Uint index) const;
NamedTransformFeedbackBinding GetNamedTransformFeedbackBinding(Uint index, Uint bufferIndex) const;
void SetNamedTransformFeedbackBinding(Uint index, Uint bufferIndex,
const SharedPtr<BufferObject>& buffer, Range1D range,
Bool hasExplicitRange);
// Framebuffer
void GenFramebufferNames(Uint number, Vector<Uint>& framebuffers);
const SharedPtr<FramebufferObject>& GetFramebufferObject(Uint index);
@@ -350,46 +243,6 @@ namespace MobileGL {
BufferState m_bufferState;
VertexArrayState m_vertexArrayState;
Array<CurrentVertexAttributeValue, VertexArrayObject::MAX_VERTEX_ATTRIBS> m_currentVertexAttributes{};
Bool m_transformFeedbackActive = false;
Bool m_transformFeedbackPaused = false;
GLenum m_transformFeedbackPrimitiveMode = GL_POINTS;
SharedPtr<ProgramObject> m_transformFeedbackProgram;
Uint64 m_transformFeedbackGeneration = 0;
// Source of the per-span ids above; never rolls back with an object switch.
Uint64 m_transformFeedbackNextGeneration = 0;
// Not object state: the transform feedback queries snapshot it at BeginQuery
// and take the delta at EndQuery, which spans whatever objects were used.
Uint64 m_transformFeedbackPrimitiveCounter = 0;
Uint64 m_transformFeedbackPausedPrimitiveCounter = 0;
Uint64 m_transformFeedbackCapturedVertices = 0;
Uint64 m_transformFeedbackInputPrimitives = 0;
// Everything a transform feedback object owns while it is NOT the bound one.
struct TransformFeedbackObjectState {
struct SavedBufferBinding {
SharedPtr<BufferObject> buffer;
Range1D range;
Bool hasExplicitRange = false;
};
Array<SavedBufferBinding, MAX_TRANSFORM_FEEDBACK_BUFFERS> bindings;
Bool active = false;
Bool paused = false;
GLenum primitiveMode = GL_POINTS;
SharedPtr<ProgramObject> program;
Uint64 generation = 0;
Uint64 capturedVertices = 0;
Uint64 inputPrimitives = 0;
Uint64 recordedVertices = 0;
Bool hasCompletedSpan = false;
Bool everBound = false;
};
void SaveBoundTransformFeedbackState();
void RestoreBoundTransformFeedbackState();
// operator[] materialises an entry with the default state on first touch, so
// the default object (name 0) needs no seeding here.
UnorderedMap<Uint, TransformFeedbackObjectState> m_transformFeedbackObjects;
IndexGenerator<Uint> m_transformFeedbackNames;
Uint m_boundTransformFeedback = 0;
TextureState m_textureState;
ProgramState m_programState;
RenderState m_renderState;
@@ -170,265 +170,9 @@ namespace MobileGL::MG_State::GLState {
m_uniformNameMaxLength = 0;
m_attribInNameMaxLength = 0;
m_uniformBlockNameMaxLength = 0;
m_xfbVaryings.clear();
m_xfbStrides.clear();
m_xfbBufferMode = GL_INTERLEAVED_ATTRIBS;
m_xfbVaryingNameMaxLength = 0;
m_xfbNeedsScatteredCapture = false;
m_xfbPackedStride = 0;
m_gsInputPrimitive = GL_NONE;
m_linkStatus = false;
}
namespace {
// GL type enum for a vertex-stage output symbol captured by transform
// feedback. Covers the scalar/vector/matrix float+integer types transform
// feedback may legally capture in GL 3.3.
Bool ResolveXfbSymbolType(const glslang::TType& type, GLenum& outType, GLint& outArraySize,
Uint32& outBytesPerElement) {
outArraySize = type.isArray() ? type.getOuterArraySize() : 1;
const Int columns = type.isMatrix() ? type.getMatrixCols() : 1;
const Int components = type.isMatrix() ? type.getMatrixRows()
: (type.isVector() ? type.getVectorSize() : 1);
const glslang::TBasicType basic = type.getBasicType();
static constexpr GLenum kFloatTypes[5] = {0, GL_FLOAT, GL_FLOAT_VEC2, GL_FLOAT_VEC3, GL_FLOAT_VEC4};
static constexpr GLenum kIntTypes[5] = {0, GL_INT, GL_INT_VEC2, GL_INT_VEC3, GL_INT_VEC4};
static constexpr GLenum kUintTypes[5] = {0, GL_UNSIGNED_INT, GL_UNSIGNED_INT_VEC2, GL_UNSIGNED_INT_VEC3,
GL_UNSIGNED_INT_VEC4};
if (type.isMatrix()) {
if (basic != glslang::EbtFloat) return false;
static constexpr GLenum kMatTypes[5][5] = {
{}, {},
{0, 0, GL_FLOAT_MAT2, GL_FLOAT_MAT2x3, GL_FLOAT_MAT2x4},
{0, 0, GL_FLOAT_MAT3x2, GL_FLOAT_MAT3, GL_FLOAT_MAT3x4},
{0, 0, GL_FLOAT_MAT4x2, GL_FLOAT_MAT4x3, GL_FLOAT_MAT4},
};
if (columns < 2 || columns > 4 || components < 2 || components > 4) return false;
outType = kMatTypes[columns][components];
} else if (components >= 1 && components <= 4) {
switch (basic) {
case glslang::EbtFloat: outType = kFloatTypes[components]; break;
case glslang::EbtInt: outType = kIntTypes[components]; break;
case glslang::EbtUint: outType = kUintTypes[components]; break;
default: return false;
}
} else {
return false;
}
outBytesPerElement = static_cast<Uint32>(columns * components) * 4u;
return true;
}
} // namespace
Bool ProgramObject::ResolveTransformFeedbackVaryings() {
m_xfbVaryings.clear();
m_xfbStrides.clear();
m_xfbBufferMode = m_requestedXfbBufferMode;
m_xfbVaryingNameMaxLength = 0;
m_xfbNeedsScatteredCapture = false;
m_xfbPackedStride = 0;
if (m_requestedXfbVaryings.empty()) {
return true;
}
// Capture happens at the last vertex-processing stage (geometry, then
// tessellation evaluation, then vertex).
const glslang::TIntermediate* captureIntermediate = nullptr;
for (EShLanguage stage : {EShLangGeometry, EShLangTessEvaluation, EShLangVertex}) {
captureIntermediate = m_program->getIntermediate(stage);
if (captureIntermediate != nullptr) {
break;
}
}
if (captureIntermediate == nullptr) {
m_infoLog = "Transform feedback varyings requested but the program has no vertex-processing stage.";
return false;
}
const glslang::TIntermAggregate* linkerObjects = captureIntermediate->findLinkerObjects();
const Bool interleaved = m_xfbBufferMode == GL_INTERLEAVED_ATTRIBS;
Uint32 interleavedOffset = 0;
// ARB_transform_feedback3 lets an interleaved capture leave holes (gl_SkipComponents1..4)
// and move on to the next buffer (gl_NextBuffer). Both only affect where the following
// varyings land, so they are consumed here and never become XfbVaryings of their own -
// which also keeps them out of the name list a backend declares on its own driver.
Uint32 interleavedBufferIndex = 0;
Vector<Uint32> interleavedStrides;
for (SizeT i = 0; i < m_requestedXfbVaryings.size(); ++i) {
const String& name = m_requestedXfbVaryings[i];
if (interleaved && name == "gl_NextBuffer") {
interleavedStrides.push_back(interleavedOffset);
interleavedOffset = 0;
++interleavedBufferIndex;
m_xfbNeedsScatteredCapture = true;
continue;
}
if (interleaved && name.size() == 18 && name.compare(0, 17, "gl_SkipComponents") == 0 &&
name[17] >= '1' && name[17] <= '4') {
interleavedOffset += static_cast<Uint32>(name[17] - '0') * 4;
m_xfbNeedsScatteredCapture = true;
continue;
}
for (SizeT j = 0; j < i; ++j) {
if (m_requestedXfbVaryings[j] == name) {
m_infoLog = "Transform feedback varying '" + name + "' is specified more than once.";
return false;
}
}
XfbVarying varying;
varying.name = name;
Uint32 bytesPerElement = 0;
Bool resolved = false;
if (name == "gl_Position") {
varying.type = GL_FLOAT_VEC4;
varying.size = 1;
bytesPerElement = 16;
resolved = true;
} else if (name == "gl_PointSize") {
varying.type = GL_FLOAT;
varying.size = 1;
bytesPerElement = 4;
resolved = true;
} else if (linkerObjects != nullptr) {
for (const auto* node : linkerObjects->getSequence()) {
const glslang::TIntermSymbol* symbol = node->getAsSymbolNode();
if (symbol == nullptr || symbol->getType().getQualifier().storage != glslang::EvqVaryingOut) {
continue;
}
if (symbol->getName() != name.c_str()) {
continue;
}
resolved = ResolveXfbSymbolType(symbol->getType(), varying.type, varying.size, bytesPerElement);
break;
}
}
if (!resolved) {
m_infoLog = "Transform feedback varying '" + name + "' is not an output of the vertex stage.";
return false;
}
varying.byteSize = bytesPerElement * static_cast<Uint32>(varying.size);
varying.packedOffsetBytes = m_xfbPackedStride;
m_xfbPackedStride += varying.byteSize;
if (interleaved) {
varying.bufferIndex = interleavedBufferIndex;
varying.offsetBytes = interleavedOffset;
interleavedOffset += varying.byteSize;
} else {
varying.bufferIndex = static_cast<Uint32>(m_xfbVaryings.size());
varying.offsetBytes = 0;
}
m_xfbVaryingNameMaxLength =
std::max(m_xfbVaryingNameMaxLength, static_cast<Int>(name.size()) + 1);
m_xfbVaryings.push_back(Move(varying));
}
constexpr Uint32 kMaxSeparateAttribs = 4;
constexpr Uint32 kMaxSeparateComponents = 4;
constexpr Uint32 kMaxInterleavedComponents = 64;
constexpr Uint32 kMaxTransformFeedbackBuffers = 4;
if (interleaved) {
interleavedStrides.push_back(interleavedOffset);
if (interleavedStrides.size() > kMaxTransformFeedbackBuffers) {
m_infoLog = "Transform feedback capture uses more buffers than "
"GL_MAX_TRANSFORM_FEEDBACK_BUFFERS.";
return false;
}
for (const Uint32 stride : interleavedStrides) {
if (stride > kMaxInterleavedComponents * 4) {
m_infoLog = "Transform feedback interleaved capture exceeds "
"GL_MAX_TRANSFORM_FEEDBACK_INTERLEAVED_COMPONENTS.";
return false;
}
}
m_xfbStrides = Move(interleavedStrides);
} else {
if (m_xfbVaryings.size() > kMaxSeparateAttribs) {
m_infoLog = "Transform feedback separate capture exceeds "
"GL_MAX_TRANSFORM_FEEDBACK_SEPARATE_ATTRIBS.";
return false;
}
m_xfbStrides.resize(m_xfbVaryings.size());
for (SizeT i = 0; i < m_xfbVaryings.size(); ++i) {
if (m_xfbVaryings[i].byteSize > kMaxSeparateComponents * 4) {
m_infoLog = "Transform feedback varying '" + m_xfbVaryings[i].name +
"' exceeds GL_MAX_TRANSFORM_FEEDBACK_SEPARATE_COMPONENTS.";
return false;
}
m_xfbStrides[i] = m_xfbVaryings[i].byteSize;
}
}
ResolveGsTriangleStripCapture(captureIntermediate);
return true;
}
namespace {
// Extracts a geometry shader's per-invocation EmitVertex/EndPrimitive sequence
// when it is statically knowable (no emit inside selection/loop/switch). Vulkan
// transform feedback captures triangle strips in plain (i, i+1, i+2) order while
// GL decomposes odd strip triangles as (i+1, i, i+2) (GL 4.6 table 10.1); with
// the static strip lengths the capture buffer can be reordered after EndTF.
class GsEmitSequenceTraverser final : public glslang::TIntermTraverser {
public:
bool visitAggregate(glslang::TVisit, glslang::TIntermAggregate* node) override {
if (node->getOp() == glslang::EOpEmitVertex) {
++emitCount;
hasEmit = true;
} else if (node->getOp() == glslang::EOpEndPrimitive) {
FlushStrip();
}
return true;
}
bool visitSelection(glslang::TVisit, glslang::TIntermSelection*) override {
inControlFlow = true;
return true;
}
bool visitLoop(glslang::TVisit, glslang::TIntermLoop*) override {
inControlFlow = true;
return true;
}
bool visitSwitch(glslang::TVisit, glslang::TIntermSwitch*) override {
inControlFlow = true;
return true;
}
void FlushStrip() {
if (emitCount >= 3) {
stripTriangles.push_back(static_cast<Uint32>(emitCount - 2));
}
emitCount = 0;
}
Vector<Uint32> stripTriangles;
Uint32 emitCount = 0;
Bool hasEmit = false;
Bool inControlFlow = false;
};
} // namespace
void ProgramObject::ResolveGsTriangleStripCapture(const glslang::TIntermediate* captureIntermediate) {
m_gsStripTriangles.clear();
m_gsStripCaptureFixup = false;
if (captureIntermediate == nullptr || m_program == nullptr) {
return;
}
if (m_program->getIntermediate(EShLangGeometry) != captureIntermediate) {
return;
}
if (captureIntermediate->getOutputPrimitive() != glslang::ElgTriangleStrip) {
return;
}
GsEmitSequenceTraverser traverser;
const_cast<glslang::TIntermediate*>(captureIntermediate)->getTreeRoot()->traverse(&traverser);
traverser.FlushStrip(); // the invocation end acts as an implicit EndPrimitive
if (!traverser.hasEmit || traverser.inControlFlow || traverser.stripTriangles.empty()) {
return;
}
m_gsStripTriangles = Move(traverser.stripTriangles);
m_gsStripCaptureFixup = true;
}
bool ProgramObject::ShaderIsAttached(const SharedPtr<ShaderObject>& shader) {
MGLOG_D("ProgramObject %u: ShaderIsAttached check for shader %p", m_externalIndex, shader.get());
auto it = std::find_if(m_shaders.begin(), m_shaders.end(),
@@ -577,32 +321,12 @@ namespace MobileGL::MG_State::GLState {
return;
}
// GL_GEOMETRY_INPUT_TYPE. A draw's primitive type has to be compatible with it
// (GL 4.6 core 11.3.1), so it is resolved for every link, not only a capturing one.
m_gsInputPrimitive = GL_NONE;
if (const glslang::TIntermediate* gs = m_program->getIntermediate(EShLangGeometry)) {
switch (gs->getInputPrimitive()) {
case glslang::ElgPoints: m_gsInputPrimitive = GL_POINTS; break;
case glslang::ElgLines: m_gsInputPrimitive = GL_LINES; break;
case glslang::ElgLinesAdjacency: m_gsInputPrimitive = GL_LINES_ADJACENCY; break;
case glslang::ElgTriangles: m_gsInputPrimitive = GL_TRIANGLES; break;
case glslang::ElgTrianglesAdjacency: m_gsInputPrimitive = GL_TRIANGLES_ADJACENCY; break;
default: break;
}
}
MGLOG_D("ProgramObject %u: Starting reflection", m_externalIndex);
DoReflection();
MGLOG_D("ProgramObject %u: Reflection done (linkStatus=%d)", m_externalIndex, (int)m_linkStatus);
if (!ValidateFragmentOutputLocations()) {
return;
}
if (!ResolveTransformFeedbackVaryings()) {
m_linkStatus = false;
MGLOG_E("ProgramObject %u: transform feedback varying resolution failed: %s", m_externalIndex,
m_infoLog.c_str());
return;
}
MGLOG_D("ProgramObject %u: Starting binary generation", m_externalIndex);
GenerateBinary();
@@ -334,32 +334,16 @@ namespace MobileGL::MG_State::GLState {
// draw. The memo is keyed by (backendStateVersion, flags); ResetLinkArtifacts and
// the binding setters below invalidate it by bumping m_backendStateVersion.
Bool GetBackendHashMemo(Uint flags, Uint64& outHash) const {
if (m_backendHashMemoVersion != m_backendStateVersion) return false;
for (const auto& slot : m_backendHashMemoSlots) {
if (slot.valid && slot.flags == flags) {
outHash = slot.hash;
return true;
}
if (m_backendHashMemoVersion != m_backendStateVersion || m_backendHashMemoFlags != flags) {
return false;
}
return false;
outHash = m_backendHashMemo;
return true;
}
void SetBackendHashMemo(Uint flags, Uint64 hash) const {
if (m_backendHashMemoVersion != m_backendStateVersion) {
for (auto& slot : m_backendHashMemoSlots) slot.valid = false;
m_backendHashMemoVersion = m_backendStateVersion;
m_backendHashMemoNextSlot = 0;
}
for (auto& slot : m_backendHashMemoSlots) {
if (slot.valid && slot.flags == flags) {
slot.hash = hash;
return;
}
}
auto& slot = m_backendHashMemoSlots[m_backendHashMemoNextSlot];
slot.flags = flags;
slot.hash = hash;
slot.valid = true;
m_backendHashMemoNextSlot = (m_backendHashMemoNextSlot + 1) % kBackendHashMemoSlotCount;
m_backendHashMemo = hash;
m_backendHashMemoVersion = m_backendStateVersion;
m_backendHashMemoFlags = flags;
}
void SetUniformSamplerOrImageUnitIndex(Uint location, Int unit) {
@@ -377,17 +361,6 @@ namespace MobileGL::MG_State::GLState {
Bool GetDeleteStatus() const { return m_deleteStatus; }
Bool GetLinkStatus() const { return m_linkStatus; }
// GL_PROGRAM_BINARY_RETRIEVABLE_HINT. MobileGL exposes no program binary format
// (GL_NUM_PROGRAM_BINARY_FORMATS is 0), so the hint is pure state - which is all
// ARB_get_program_binary requires of it.
Bool GetBinaryRetrievableHint() const { return m_binaryRetrievableHint; }
void SetBinaryRetrievableHint(Bool hint) { m_binaryRetrievableHint = hint; }
// glProgramBinary always fails here (there is no format it could accept) and the
// spec then requires the program's LINK_STATUS to read FALSE.
void MarkLinkFailedByProgramBinary() {
ResetLinkArtifacts();
m_infoLog = "No program binary format is supported.";
}
Bool GetValidateStatus() const { return m_validateStatus; }
Int GetActiveAtomicCounterCount() const { return m_program->getNumAtomicCounters(); }
Int GetActiveAttributesCount() const { return m_program->getNumPipeInputs(); }
@@ -476,55 +449,6 @@ namespace MobileGL::MG_State::GLState {
return it == m_shaders.end() ? -1 : (Int)std::distance(m_shaders.begin(), it);
}
// Transform feedback (GL 3.0 core: glTransformFeedbackVaryings applies on
// the NEXT link; the linked snapshot below is what draws and queries see).
struct XfbVarying {
String name;
GLenum type = GL_FLOAT;
GLint size = 1; // array element count
Uint32 bufferIndex = 0; // capture buffer slot
Uint32 offsetBytes = 0; // offset within the capture buffer
Uint32 byteSize = 0; // bytes captured per vertex for this varying
// Offset within the gap-free record a backend that cannot express the GL
// layout captures into; see NeedsScatteredTransformFeedbackCapture.
Uint32 packedOffsetBytes = 0;
};
void SetTransformFeedbackVaryings(Vector<String>&& names, GLenum bufferMode) {
m_requestedXfbVaryings = Move(names);
m_requestedXfbBufferMode = bufferMode;
}
GLenum GetTransformFeedbackBufferMode() const { return m_xfbBufferMode; }
SizeT GetTransformFeedbackVaryingCount() const { return m_xfbVaryings.size(); }
const XfbVarying* GetTransformFeedbackVarying(SizeT index) const {
return index < m_xfbVaryings.size() ? &m_xfbVaryings[index] : nullptr;
}
const Vector<XfbVarying>& GetTransformFeedbackVaryings() const { return m_xfbVaryings; }
// Stride of one captured vertex in the given capture buffer slot.
Uint32 GetTransformFeedbackStride(Uint32 bufferIndex) const {
return bufferIndex < m_xfbStrides.size() ? m_xfbStrides[bufferIndex] : 0;
}
SizeT GetTransformFeedbackBufferCount() const { return m_xfbStrides.size(); }
Int GetTransformFeedbackVaryingMaxLength() const { return m_xfbVaryingNameMaxLength; }
// True when the capture layout uses gl_SkipComponents / gl_NextBuffer
// (ARB_transform_feedback3), which no ES driver can express: it can only pack every
// captured varying into one record with no gaps. A backend that captures through
// such a driver has to capture into scratch storage and scatter the records into the
// application's buffers itself, using packedOffsetBytes as the source offset and
// (bufferIndex, offsetBytes, stride) as the destination.
Bool NeedsScatteredTransformFeedbackCapture() const { return m_xfbNeedsScatteredCapture; }
// Bytes one gap-free captured record occupies.
Uint32 GetTransformFeedbackPackedStride() const { return m_xfbPackedStride; }
// True when the capture stage is a triangle-strip geometry shader with a
// statically-known emit sequence: the Vulkan capture order then needs the GL
// odd-triangle vertex swap after EndTransformFeedback.
Bool HasGsTriangleStripCaptureFixup() const { return m_gsStripCaptureFixup; }
// Triangles per strip, in emission order, for ONE geometry invocation.
const Vector<Uint32>& GetGsStripTriangles() const { return m_gsStripTriangles; }
// GL_GEOMETRY_INPUT_TYPE of the linked geometry stage (GL_POINTS, GL_LINES,
// GL_LINES_ADJACENCY, GL_TRIANGLES or GL_TRIANGLES_ADJACENCY), or GL_NONE when the
// program has no geometry stage. Draws must present a compatible primitive type.
GLenum GetGeometryInputType() const { return m_gsInputPrimitive; }
Uint GetExternalIndex() const { return m_externalIndex; }
// Globally-unique, never-reused id for this program object's lifetime. Unlike the GL
// name (external index), which is freed to a LIFO list and immediately handed back by
@@ -535,11 +459,6 @@ namespace MobileGL::MG_State::GLState {
private:
void ResetLinkArtifacts();
void DoReflection();
// Resolves the requested transform feedback varyings against the linked
// vertex stage; fails the link (GL semantics) on unknown or duplicate
// names or exceeded capture limits.
Bool ResolveTransformFeedbackVaryings();
void ResolveGsTriangleStripCapture(const glslang::TIntermediate* captureIntermediate);
void GenerateBinary();
void WaitUntilGenerationCompleted() const;
void AddDefaultFragmentShaderIfMissing();
@@ -604,38 +523,15 @@ namespace MobileGL::MG_State::GLState {
String m_infoLog;
Bool m_deleteStatus = false;
Bool m_linkStatus = false;
Bool m_binaryRetrievableHint = false;
Bool m_validateStatus = true;
Uint32 m_backendStateVersion = 0;
// Backend-owned content-hash memo (see GetBackendHashMemo): valid only while
// m_backendStateVersion matches. Several slots, not one: a backend may resolve the same
// program under more than one compile-flag set within a frame (surface rotation, and the
// explicit-LOD sampling variant), and a single slot would then miss on every lookup and
// re-hash the program's whole SPIR-V once per draw.
static constexpr SizeT kBackendHashMemoSlotCount = 4;
struct BackendHashMemoSlot {
Uint64 hash = 0;
Uint flags = 0;
Bool valid = false;
};
mutable Array<BackendHashMemoSlot, kBackendHashMemoSlotCount> m_backendHashMemoSlots{};
mutable SizeT m_backendHashMemoNextSlot = 0;
// m_backendStateVersion and the compile flags match the recorded values.
mutable Uint64 m_backendHashMemo = 0;
mutable Uint32 m_backendHashMemoVersion = ~0u;
mutable Uint m_backendHashMemoFlags = 0;
Uint32 m_uboContentVersion = 0;
Uint32 m_linkVersion = 0;
// Transform feedback: request (applies at next link) and linked snapshot.
Vector<String> m_requestedXfbVaryings;
GLenum m_requestedXfbBufferMode = GL_INTERLEAVED_ATTRIBS;
Vector<XfbVarying> m_xfbVaryings;
Vector<Uint32> m_xfbStrides;
Vector<Uint32> m_gsStripTriangles;
Bool m_gsStripCaptureFixup = false;
GLenum m_gsInputPrimitive = GL_NONE;
GLenum m_xfbBufferMode = GL_INTERLEAVED_ATTRIBS;
Int m_xfbVaryingNameMaxLength = 0;
Bool m_xfbNeedsScatteredCapture = false;
Uint32 m_xfbPackedStride = 0;
};
} // namespace MobileGL::MG_State::GLState
@@ -29,25 +29,17 @@ namespace MobileGL::MG_State::GLState {
if (!CheckIndexAvail(program, m_programObjects)) return; // FIXME: add error reporting here
auto& programObject = m_programObjects[program];
if (programObject != nullptr) {
// Snapshot the attachments: deleting the program is a detach point for shaders
// that were flagged with glDeleteShader while still attached.
const Vector<SharedPtr<ShaderObject>> attachedShaders = programObject->GetAttachedShaders();
programObject->MarkAsDeleted();
// A program in use is only FLAGGED: its name (and every program query) stays
// valid until it stops being current, at which point UseProgram finishes the job.
if (programObject == m_currentProgram) return;
DestroyProgramSlot(program);
}
}
void ProgramState::DestroyProgramSlot(const Uint program) {
auto& programObject = m_programObjects[program];
// Snapshot the attachments: deleting the program is a detach point for shaders
// that were flagged with glDeleteShader while still attached.
const Vector<SharedPtr<ShaderObject>> attachedShaders = programObject->GetAttachedShaders();
programObject.reset();
m_programIndexGenerator.Delete(program);
for (const auto& shader : attachedShaders) {
const Uint shaderName = shader->GetExternalIndex();
if (CheckIndexAvail(shaderName, m_shaderObjects) && m_shaderObjects[shaderName] == shader) {
ReleaseShaderNameIfOrphaned(shaderName);
programObject.reset();
m_programIndexGenerator.Delete(program);
for (const auto& shader : attachedShaders) {
const Uint shaderName = shader->GetExternalIndex();
if (CheckIndexAvail(shaderName, m_shaderObjects) && m_shaderObjects[shaderName] == shader) {
ReleaseShaderNameIfOrphaned(shaderName);
}
}
}
}
@@ -57,22 +49,10 @@ namespace MobileGL::MG_State::GLState {
}
void ProgramState::UseProgram(Uint program) {
const SharedPtr<ProgramObject> previous = m_currentProgram;
if (program == 0) m_currentProgram.reset();
if (CheckIndexAvail(program, m_programObjects)) {
m_currentProgram = m_programObjects[program];
}
// A deletion flagged while the program was current takes effect the moment it
// stops being current.
if (previous != nullptr && previous != m_currentProgram && previous->GetDeleteStatus()) {
const Uint previousName = previous->GetExternalIndex();
if (CheckIndexAvail(previousName, m_programObjects) && m_programObjects[previousName] == previous) {
DestroyProgramSlot(previousName);
}
}
if (!CheckIndexAvail(program, m_programObjects)) return;
m_currentProgram = m_programObjects[program];
}
Uint ProgramState::CreateShader(ShaderStage stage) {
@@ -35,9 +35,6 @@ namespace MobileGL::MG_State::GLState {
private:
Bool ShaderHasGLVisibleAttachment(const SharedPtr<ShaderObject>& shaderObject) const;
// Frees the name slot and releases orphaned attached shaders; the immediate half
// of glDeleteProgram (deferred while the program is current).
void DestroyProgramSlot(Uint program);
template <typename T>
static Bool CheckIndexAvail(const SizeT idx, const Vector<T>& vec) {
@@ -169,15 +169,6 @@ namespace MobileGL::MG_State::GLState {
}
}
const std::optional<String> reservedError =
MG_Util::ShaderTranspiler::FindReservedIdentifierViolation(compileSource);
if (reservedError) {
m_compileStatus = false;
m_shader.reset();
m_infoLog = *reservedError;
return;
}
// Compile for OpenGL here, so that we can do validation and link
// like a real OpenGL driver at linking stage
// Will compile for other backends later.
@@ -154,17 +154,6 @@ namespace MobileGL {
return m_parameters.PointSize;
}
void RenderState::SetPatchVertices(Uint vertices) {
if (m_parameters.PatchVertices == vertices) return;
m_parameters.PatchVertices = vertices;
++m_version;
}
Uint RenderState::GetPatchVertices() const {
return m_parameters.PatchVertices;
}
void RenderState::SetPolygonOffset(Float factor, Float units) {
if (m_parameters.PolygonOffsetFactor == factor && m_parameters.PolygonOffsetUnits == units) return;
@@ -224,8 +224,6 @@ namespace MobileGL {
IntVec4 Viewport = IntVec4(0, 0, 0, 0); // x, y, width, height
Float LineWidth = 1.0f;
Float PointSize = 1.0f;
// GL_PATCH_VERTICES: how many vertices one tessellation patch consumes.
Uint PatchVertices = 3;
Float PolygonOffsetFactor = 0.0f;
Float PolygonOffsetUnits = 0.0f;
@@ -321,8 +319,6 @@ namespace MobileGL {
Float GetLineWidth() const;
void SetPointSize(Float size);
Float GetPointSize() const;
void SetPatchVertices(Uint vertices);
Uint GetPatchVertices() const;
void SetPolygonOffset(Float factor, Float units);
Float GetPolygonOffsetFactor() const;
Float GetPolygonOffsetUnits() const;
@@ -162,7 +162,6 @@ namespace MobileGL {
DepthComponent32F,
Depth24Stencil8,
Depth32FStencil8,
StencilIndex8,
DepthComponent,
DepthStencil,
@@ -24,19 +24,6 @@ namespace MobileGL {
TextureObjectBase::TextureObjectBase(TextureTarget target, Uint externalIndex)
: m_externalIndex(externalIndex), m_lifetimeId(AllocateLifetimeId()), m_target(target) {
m_sampler = MakeShared<SamplerObject>(0);
if (target == TextureTarget::TextureRectangle) {
// A rectangle texture has no mip chain, so its initial sampler state is not
// the shared one: TEXTURE_MIN_FILTER is LINEAR and TEXTURE_WRAP_S/T are
// CLAMP_TO_EDGE (GL 4.6 core table 23.15). Leaving the 2D default of
// NEAREST_MIPMAP_LINEAR in place makes the texture mipmap-incomplete from
// birth, and every lookup that the application never re-filtered reads
// (0, 0, 0, 1) instead of its contents.
m_sampler->SetMinFilter(SamplerFilterMode::Linear);
m_sampler->SetMipmapMode(SamplerMipmapMode::None);
m_sampler->SetWrapS(SamplerWrapMode::ClampToEdge);
m_sampler->SetWrapT(SamplerWrapMode::ClampToEdge);
m_sampler->SetWrapR(SamplerWrapMode::ClampToEdge);
}
}
TextureInternalFormat TextureObjectBase::GetFormat() const {
@@ -348,64 +335,6 @@ namespace MobileGL {
// TODO: add other texture types as needed
Bool SamplesAsIncompleteTexture(const ITextureObject* texture, const SamplerObject* effectiveSampler) {
const Bool mipmapped =
effectiveSampler != nullptr && effectiveSampler->GetMipmapMode() != SamplerMipmapMode::None;
return !IsMipmapCompleteForFilter(texture, mipmapped);
}
Bool IsMipmapCompleteForFilter(const ITextureObject* texture, Bool mipmapped) {
if (texture == nullptr) return true;
if (!texture->IsComplete()) return false;
if (!mipmapped) return true;
const auto* mipmapTexture = AsMipmapTexture(texture);
if (mipmapTexture == nullptr) return true; // no mip chain to be incomplete about
const UintVec2& levelRange = texture->GetLevelRange();
const Uint baseLevel = levelRange.x();
const Uint storedLevels = mipmapTexture->GetMipmapLevelCount();
if (baseLevel >= storedLevels) return false;
// An array texture's layer count is not a dimension of the image: it stays put all
// the way down the chain (GL 4.6 core 8.14.3). GetMipmapTexelSize reports it in the
// slot after the image's own dimensions.
const TextureTarget target = texture->GetTarget();
Int shrinkingComponents = 3;
if (target == TextureTarget::Texture1DArray) {
shrinkingComponents = 1;
} else if (target == TextureTarget::Texture2DArray || target == TextureTarget::TextureCubeMapArray) {
shrinkingComponents = 2;
}
for (const auto uploadTarget : texture->GetUploadTargets()) {
const IntVec3 baseSize = mipmapTexture->GetMipmapTexelSize(uploadTarget, baseLevel);
Int largest = 0;
for (Int component = 0; component < shrinkingComponents; ++component) {
largest = std::max(largest, baseSize[component]);
}
if (largest <= 0) return false;
// p = log2 of the largest base dimension: the last level the chain needs
// before every dimension has reached 1. TEXTURE_MAX_LEVEL can cut it short.
Uint p = 0;
for (Int extent = largest; extent > 1; extent >>= 1) ++p;
const Uint lastLevel = std::min(baseLevel + p, levelRange.y());
for (Uint level = baseLevel; level <= lastLevel; ++level) {
if (level >= storedLevels) return false;
const IntVec3 actual = mipmapTexture->GetMipmapTexelSize(uploadTarget, level);
for (Int component = 0; component < 3; ++component) {
const Int expected = component < shrinkingComponents
? std::max(1, baseSize[component] >> (level - baseLevel))
: baseSize[component];
if (actual[component] != expected) return false;
}
}
}
return true;
}
} // namespace GLState
} // namespace MG_State
} // namespace MobileGL
@@ -15,11 +15,7 @@
#include <MG_Util/Math/VectorTypes.h>
namespace MobileGL::MG_State::GLState {
// Texture objects are always SharedPtr-owned (TextureState creates every instance via
// MakeShared, including the per-target default objects). enable_shared_from_this lets
// backends that only receive a reference (e.g. syncing a name-deleted texture kept
// alive by an FBO attachment) still register a weak liveness reference for GC.
class ITextureObject : public std::enable_shared_from_this<ITextureObject> {
class ITextureObject {
public:
using TargetEnum = TextureTarget;
virtual ~ITextureObject() = default;
@@ -156,20 +152,6 @@ namespace MobileGL::MG_State::GLState {
? static_cast<TextureObjectMipmap*>(texture)
: nullptr;
}
// Whether the texture satisfies the mipmap-completeness rules a minification filter
// that samples the mip chain imposes (GL 4.6 core 8.17): every level from the base to
// the effective max must exist at exactly half the previous one's size. `mipmapped` is
// the effective sampler's answer to "does this filter read more than the base level" -
// when it is false only base-level completeness matters, which the ordinary
// IsComplete() already covers. Sampling an incomplete texture returns (0, 0, 0, 1).
Bool IsMipmapCompleteForFilter(const ITextureObject* texture, Bool mipmapped);
// The rule above asked as the backends need it: does a lookup on this texture read
// (0, 0, 0, 1) instead of its contents? `effectiveSampler` is the sampler object bound
// to the unit when there is one, otherwise the texture's own. A backend answers yes by
// routing the texture to whatever it already uses for "nothing is bound there".
Bool SamplesAsIncompleteTexture(const ITextureObject* texture, const SamplerObject* effectiveSampler);
inline const TextureObjectMipmap* AsMipmapTexture(const ITextureObject* texture) {
return (texture && texture->GetStorageType() == TextureStorageType::Mipmap)
? static_cast<const TextureObjectMipmap*>(texture)
@@ -21,31 +21,10 @@ namespace MobileGL {
BindingSlot<BufferObject>& GetBufferBindingSlot(
TextureUploadTarget target = TextureUploadTarget::TextureBuffer);
// The window of the attached buffer the texture addresses. glTexBuffer attaches
// the whole buffer, which is expressed here as an offset of 0 and a size of
// kWholeBuffer so a later respecify of the buffer keeps being followed - a stored
// size would freeze the texture at the size the buffer happened to have.
static constexpr SizeT kWholeBuffer = ~static_cast<SizeT>(0);
void SetBufferRange(SizeT offset, SizeT size) {
m_bufferRangeOffset = offset;
m_bufferRangeSize = size;
}
SizeT GetBufferRangeOffset() const { return m_bufferRangeOffset; }
// Resolved against the buffer's current size, so kWholeBuffer tracks it.
SizeT GetBufferRangeSizeInBytes() const {
const auto& buffer = m_bufferBindingSlot.GetBoundObject();
const SizeT bufferSize = buffer != nullptr ? buffer->GetSize() : 0;
if (m_bufferRangeSize == kWholeBuffer) return bufferSize;
const SizeT available = bufferSize > m_bufferRangeOffset ? bufferSize - m_bufferRangeOffset : 0;
return std::min(m_bufferRangeSize, available);
}
protected:
Uint GetIndexOfTextureUploadTarget(TextureUploadTarget target) const override;
BindingSlot<BufferObject> m_bufferBindingSlot = BindingSlot<BufferObject>(BufferTarget::Texture);
SizeT m_bufferRangeOffset = 0;
SizeT m_bufferRangeSize = kWholeBuffer;
const Vector<TextureUploadTarget> m_uploadTargets{TextureUploadTarget::TextureBuffer};
};
} // namespace GLState
@@ -77,26 +77,6 @@ namespace MobileGL::MG_State::GLState {
BumpAttributeFormatVersion(index);
}
void VertexArrayObject::MirrorPointerIntoBinding(Uint index, const SharedPtr<BufferObject>& buffer, SizeT offset,
int effectiveStride) {
if (index >= MAX_VERTEX_ATTRIBS || index >= MAX_VERTEX_ATTRIB_BINDINGS) return;
// glVertexAttribPointer is defined in terms of the binding model (GL 4.6 core 10.3.2): it
// also sets binding point `index` to the buffer, the pointer as the offset, and the
// *effective* stride, and points the attribute at that binding point with relative offset 0.
// The flat attribute view keeps the raw stride, because VERTEX_ATTRIB_ARRAY_STRIDE reports
// that argument verbatim, so the binding point is recorded alongside the resolved attribute
// rather than being resolved into it.
m_attributeBindingIndex[index] = index;
m_attributeRelativeOffset[index] = 0;
auto& binding = m_bindingPoints[index];
binding.Buffer = buffer;
binding.Offset = offset;
binding.Stride = effectiveStride;
binding.Divisor = m_attributes[index].Divisor;
}
void VertexArrayObject::BindAttributeBuffer(Uint index, const SharedPtr<BufferObject>& buffer) {
if (index >= MAX_VERTEX_ATTRIBS) return;
@@ -131,11 +111,6 @@ namespace MobileGL::MG_State::GLState {
void VertexArrayObject::SetAttributeDivisor(Uint index, Uint divisor) {
if (index >= MAX_VERTEX_ATTRIBS) return;
// glVertexAttribDivisor is VertexBindingDivisor on the attribute's own binding point
// (GL 4.6 core 10.3.2), so the binding-point view has to follow the resolved attribute.
if (index < MAX_VERTEX_ATTRIB_BINDINGS && m_attributeBindingIndex[index] == index) {
m_bindingPoints[index].Divisor = divisor;
}
if (m_attributes[index].Divisor == divisor) return;
m_attributes[index].Divisor = divisor;
BumpAttributeFormatVersion(index);
@@ -212,18 +187,18 @@ namespace MobileGL::MG_State::GLState {
}
void VertexArrayObject::SetAttributeFormatSeparate(Uint attribIndex, int size, DataType type, Bool normalized,
Bool isInteger, Uint relativeOffset, Bool isBgra) {
Bool isInteger, Uint relativeOffset) {
if (attribIndex >= MAX_VERTEX_ATTRIBS) return;
if (size < 1 || size > 4) return;
auto& attr = m_attributes[attribIndex];
if (attr.Size != size || attr.Type != type || attr.Normalized != normalized || attr.IsInteger != isInteger ||
attr.IsBgra != isBgra || m_attributeRelativeOffset[attribIndex] != relativeOffset) {
attr.IsBgra || m_attributeRelativeOffset[attribIndex] != relativeOffset) {
attr.Size = size;
attr.Type = type;
attr.Normalized = normalized;
attr.IsInteger = isInteger;
attr.IsBgra = isBgra;
attr.IsBgra = false; // the binding-format path (glVertexAttribFormat) does not carry BGRA
m_attributeRelativeOffset[attribIndex] = relativeOffset;
BumpAttributeFormatVersion(attribIndex);
}
@@ -64,12 +64,6 @@ namespace MobileGL {
void BindAttributeBuffer(Uint index, const SharedPtr<BufferObject>& buffer);
// Record what the pointer-style API implies for the binding-point view: attribute
// `index` bound to binding point `index` with relative offset 0, and that binding
// point carrying the buffer, the pointer offset and the effective stride.
void MirrorPointerIntoBinding(Uint index, const SharedPtr<BufferObject>& buffer, SizeT offset,
int effectiveStride);
BindingSlot<BufferObject>& GetIndexBufferBindingSlot();
const BindingSlot<BufferObject>& GetIndexBufferBindingSlot() const;
@@ -88,22 +82,7 @@ namespace MobileGL {
void SetBindingDivisor(Uint bindingIndex, Uint divisor);
void SetAttributeBinding(Uint attribIndex, Uint bindingIndex);
void SetAttributeFormatSeparate(Uint attribIndex, int size, DataType type, Bool normalized,
Bool isInteger, Uint relativeOffset, Bool isBgra = false);
// The binding-point view the attributes were resolved from. Kept queryable
// because glGetVertexArrayIndexed[64]iv reports it verbatim, and the resolved
// flat attribute cannot always be inverted back into it.
Uint GetAttributeRelativeOffset(Uint attribIndex) const {
return attribIndex < m_attributeRelativeOffset.size() ? m_attributeRelativeOffset[attribIndex] : 0;
}
Uint GetAttributeBindingIndex(Uint attribIndex) const {
return attribIndex < m_attributeBindingIndex.size() ? m_attributeBindingIndex[attribIndex]
: attribIndex;
}
const VertexBufferBindingPoint& GetBindingPoint(Uint bindingIndex) const {
static const VertexBufferBindingPoint kEmpty{};
return bindingIndex < m_bindingPoints.size() ? m_bindingPoints[bindingIndex] : kEmpty;
}
Bool isInteger, Uint relativeOffset);
const VertexAttributeVersion& GetAttributeVersion(Uint index) const;
const Array<VertexAttributeVersion, MAX_VERTEX_ATTRIBS>& GetAllAttributeVersions() const;
@@ -125,23 +104,6 @@ namespace MobileGL {
m_backendHashMemoVersion = m_configVersion;
}
// Backend-owned resolved-state memo: an opaque pointer into the
// backend's vertex-input-state cache plus the cache's eviction
// epoch, valid while the config version matches. Lets the
// per-draw path skip the content hash AND the cache lookup; the
// epoch guards against the cache evicting the pointee.
Bool GetBackendStateMemo(const void*& outState, Uint64& outEpoch) const {
if (m_backendStateMemoVersion != m_configVersion) return false;
outState = m_backendStateMemo;
outEpoch = m_backendStateMemoEpoch;
return true;
}
void SetBackendStateMemo(const void* state, Uint64 epoch) const {
m_backendStateMemo = state;
m_backendStateMemoEpoch = epoch;
m_backendStateMemoVersion = m_configVersion;
}
private:
void BumpAttributeFormatVersion(Uint index);
void BumpAttributeBufferVersion(Uint index);
@@ -175,9 +137,6 @@ namespace MobileGL {
Uint32 m_configVersion = 0;
mutable Uint64 m_backendHashMemo = 0;
mutable Uint32 m_backendHashMemoVersion = ~0u;
mutable const void* m_backendStateMemo = nullptr;
mutable Uint64 m_backendStateMemoEpoch = 0;
mutable Uint32 m_backendStateMemoVersion = ~0u;
};
} // namespace GLState
} // namespace MG_State
@@ -34,11 +34,6 @@ namespace {
GLint maxFragmentImageUniforms = 4;
GLint maxComputeImageUniforms = 5;
bool maxGeometryImageUniformsQueried = false;
GLfloat minFragmentInterpolationOffset = -0.75f;
GLfloat maxFragmentInterpolationOffset = 0.625f;
GLint fragmentInterpolationOffsetBits = 6;
bool fragmentInterpolationLimitsQueried = false;
bool fragmentInterpolationQueryRaisesError = false;
// Emulates ANGLE-on-Vulkan: the draw reads the indirect command's
// baseInstance word and exposes it through gl_InstanceID.
bool drawLeaksBaseInstanceWord = false;
@@ -113,14 +108,6 @@ namespace {
case GL_MAX_COMPUTE_IMAGE_UNIFORMS:
*data = g_fake.maxComputeImageUniforms;
break;
case GL_FRAGMENT_INTERPOLATION_OFFSET_BITS:
g_fake.fragmentInterpolationLimitsQueried = true;
if (g_fake.fragmentInterpolationQueryRaisesError) {
g_fake.pendingError = GL_INVALID_ENUM;
} else {
*data = g_fake.fragmentInterpolationOffsetBits;
}
break;
// FillInGLESCapabilities reads the context version before running the
// baseInstance probe, which requires ES >= 3.1.
case GL_MAJOR_VERSION:
@@ -168,22 +155,6 @@ namespace {
g_fake.maxTextureMaxAnisotropyQueried = true;
data[0] = g_fake.maxTextureMaxAnisotropy;
break;
case GL_MIN_FRAGMENT_INTERPOLATION_OFFSET:
g_fake.fragmentInterpolationLimitsQueried = true;
if (g_fake.fragmentInterpolationQueryRaisesError) {
g_fake.pendingError = GL_INVALID_ENUM;
} else {
data[0] = g_fake.minFragmentInterpolationOffset;
}
break;
case GL_MAX_FRAGMENT_INTERPOLATION_OFFSET:
g_fake.fragmentInterpolationLimitsQueried = true;
if (g_fake.fragmentInterpolationQueryRaisesError) {
g_fake.pendingError = GL_INVALID_ENUM;
} else {
data[0] = g_fake.maxFragmentInterpolationOffset;
}
break;
// Two-component range queries.
case GL_ALIASED_LINE_WIDTH_RANGE:
case GL_SMOOTH_LINE_WIDTH_RANGE:
@@ -491,52 +462,6 @@ TEST(ImageUniformCapabilities, QueriesRealPerStageLimitsAndConservativelyGatesGe
EXPECT_TRUE(g_fake.maxGeometryImageUniformsQueried);
}
TEST(FragmentInterpolationCapabilities, QueriesOnlyWhenSupportedAndPreservesDriverLimits) {
const auto funcs = MakeFakeGLESFunctions();
ResetFakeDriver();
g_fake.maxVertexSsboBlocks = 0;
MobileGL::MG_External::GLESCapabilities unsupportedCaps;
ASSERT_TRUE(MobileGL::MG_Util::BackendLoader::FillInGLESCapabilities(unsupportedCaps, funcs));
EXPECT_FALSE(unsupportedCaps.SupportsShaderMultisampleInterpolation);
EXPECT_FALSE(g_fake.fragmentInterpolationLimitsQueried);
EXPECT_FLOAT_EQ(unsupportedCaps.MinFragmentInterpolationOffset, -0.5f);
EXPECT_FLOAT_EQ(unsupportedCaps.MaxFragmentInterpolationOffset, 0.4375f);
EXPECT_EQ(unsupportedCaps.FragmentInterpolationOffsetBits, 4);
ResetFakeDriver();
g_fake.maxVertexSsboBlocks = 0;
g_fake.extensions.emplace_back("GL_OES_shader_multisample_interpolation");
// A stale error from an earlier capability probe must not make the optional
// interpolation query look like it failed.
g_fake.pendingError = GL_INVALID_OPERATION;
MobileGL::MG_External::GLESCapabilities supportedCaps;
ASSERT_TRUE(MobileGL::MG_Util::BackendLoader::FillInGLESCapabilities(supportedCaps, funcs));
EXPECT_TRUE(supportedCaps.SupportsShaderMultisampleInterpolation);
EXPECT_TRUE(g_fake.fragmentInterpolationLimitsQueried);
EXPECT_FLOAT_EQ(supportedCaps.MinFragmentInterpolationOffset, g_fake.minFragmentInterpolationOffset);
EXPECT_FLOAT_EQ(supportedCaps.MaxFragmentInterpolationOffset, g_fake.maxFragmentInterpolationOffset);
EXPECT_EQ(supportedCaps.FragmentInterpolationOffsetBits, g_fake.fragmentInterpolationOffsetBits);
EXPECT_EQ(funcs.glGetError(), GL_NO_ERROR);
}
TEST(FragmentInterpolationCapabilities, QueryErrorIsDrainedAndFallsBackToCoreMinimums) {
ResetFakeDriver();
g_fake.maxVertexSsboBlocks = 0;
g_fake.extensions.emplace_back("GL_OES_shader_multisample_interpolation");
g_fake.fragmentInterpolationQueryRaisesError = true;
const auto funcs = MakeFakeGLESFunctions();
MobileGL::MG_External::GLESCapabilities caps;
ASSERT_TRUE(MobileGL::MG_Util::BackendLoader::FillInGLESCapabilities(caps, funcs));
EXPECT_TRUE(g_fake.fragmentInterpolationLimitsQueried);
EXPECT_FLOAT_EQ(caps.MinFragmentInterpolationOffset, -0.5f);
EXPECT_FLOAT_EQ(caps.MaxFragmentInterpolationOffset, 0.4375f);
EXPECT_EQ(caps.FragmentInterpolationOffsetBits, 4);
EXPECT_EQ(funcs.glGetError(), GL_NO_ERROR);
}
// The extension string is what apps gate on (LWJGL builds GLCapabilities from it), so advertising
// it on a driver that cannot filter anisotropically would leave them silently on trilinear.
TEST(TextureAnisotropyCapabilities, ExtensionIsAdvertisedOnlyWhenTheHostDriverSupportsIt) {
+19 -35
View File
@@ -223,7 +223,7 @@ void main() {
PreprocessShaderSource(ShaderStage::Vertex, source);
EXPECT_EQ(source.find("#version 330 core "), 0);
EXPECT_EQ(source.find("#version 330 core\n"), 0);
EXPECT_NE(source.find("in vec3 position;"), String::npos);
EXPECT_NE(source.find("out vec2 uv;"), String::npos);
EXPECT_EQ(source.find("attribute"), String::npos);
@@ -323,7 +323,7 @@ void main() {
PreprocessShaderSource(ShaderStage::Fragment, source);
EXPECT_EQ(source.find("#version 330 core "), 0);
EXPECT_EQ(source.find("#version 330 core\n"), 0);
EXPECT_NE(source.find("out vec4 mg_FragColor;\n"), String::npos);
EXPECT_NE(source.find("in vec2 uv;"), String::npos);
EXPECT_NE(source.find("texture(texture0, uv)"), String::npos);
@@ -379,7 +379,7 @@ void main() {
PreprocessShaderSource(ShaderStage::Fragment, source);
EXPECT_EQ(source.find("#version 330 core "), 0);
EXPECT_EQ(source.find("#version 330 core\n"), 0);
EXPECT_NE(source.find("vec4 sample = texture(DiffuseSampler"), String::npos);
EXPECT_NE(source.find("totalAlpha = totalAlpha + sample.a;"), String::npos);
EXPECT_NE(source.find("float totalSamples = 0.0;"), String::npos);
@@ -405,7 +405,7 @@ void main() {
)";
PreprocessShaderSource(ShaderStage::Vertex, vertexSource);
EXPECT_EQ(vertexSource.find("#version 330 core "), 0);
EXPECT_EQ(vertexSource.find("#version 330 core\n"), 0);
EXPECT_NE(vertexSource.find("in vec3 sample;"), String::npos);
ShaderAttrib vertexAttrib{.shaderType = GL_VERTEX_SHADER, .sourceStr = vertexSource};
@@ -425,7 +425,7 @@ void main() {
)";
PreprocessShaderSource(ShaderStage::Fragment, fragmentSource);
EXPECT_EQ(fragmentSource.find("#version 330 core "), 0);
EXPECT_EQ(fragmentSource.find("#version 330 core\n"), 0);
EXPECT_NE(fragmentSource.find("uniform sampler2D sample;"), String::npos);
EXPECT_NE(fragmentSource.find("texture(sample, texCoord)"), String::npos);
@@ -483,9 +483,7 @@ void main() {
)";
PreprocessShaderSource(ShaderStage::Fragment, source);
// An explicitly declared modern core version keeps its number (see "keep declared modern
// GLSL versions strict"); only the BOM goes.
EXPECT_EQ(source.find(String(inputVersion) + "\n"), 0);
EXPECT_EQ(source.find("#version 460 core\n"), 0);
EXPECT_EQ(source.find("\xef\xbb\xbf"), String::npos);
ShaderAttrib attrib{.shaderType = GL_FRAGMENT_SHADER, .sourceStr = source};
@@ -570,12 +568,10 @@ void main() {
)";
PreprocessShaderSource(ShaderStage::Fragment, source);
const SizeT versionPos = source.find("#version 330 core ");
const SizeT versionPos = source.find("#version 330 core\n");
const SizeT outputPos = source.find("out vec4 mg_FragColor;\n");
EXPECT_NE(versionPos, String::npos);
// The normalized directive carries a marker recording that this 330 came from a legacy
// declaration, so measure the line rather than assuming its length.
EXPECT_EQ(outputPos, source.find('\n', versionPos) + 1);
EXPECT_EQ(outputPos, versionPos + std::strlen("#version 330 core\n"));
EXPECT_NE(source.find("// #version 460 core"), String::npos);
// This #line sits ahead of the version directive, where GLSL would never have honoured it, so
// it is still dropped. Directives that follow the version line are kept - see
@@ -688,7 +684,7 @@ void main() {
}
}
TEST_F(ProgramUtilTest, PreprocessModernSampleQualifierStaysAtItsDeclaredVersion) {
TEST_F(ProgramUtilTest, PreprocessModernSampleQualifierStaysAtVersion460) {
using namespace MG_Util::ShaderTranspiler;
String source = R"(#version 400 core
@@ -701,7 +697,7 @@ void main() {
)";
PreprocessShaderSource(ShaderStage::Fragment, source);
EXPECT_EQ(source.find("#version 400 core\n"), 0);
EXPECT_EQ(source.find("#version 460 core\n"), 0);
EXPECT_NE(source.find("sample in vec4 interpolatedColor;"), String::npos);
ShaderAttrib attrib{.shaderType = GL_FRAGMENT_SHADER, .sourceStr = source};
@@ -750,7 +746,7 @@ void main() {
PreprocessShaderSource(ShaderStage::Fragment, source);
EXPECT_EQ(source.find("#version 330 core "), 0);
EXPECT_EQ(source.find("#version 330 core\n"), 0);
EXPECT_NE(source.find("layout(location = 0) out vec4 mg_FragData[8];\n"), String::npos);
EXPECT_NE(source.find("mg_FragData[0] = vec4(1.0);"), String::npos);
EXPECT_NE(source.find("mg_FragData[1].a = 0.5;"), String::npos);
@@ -906,17 +902,16 @@ TEST_F(ProgramUtilTest, CompileSimpleVertexShader) {
}
// Legacy desktop sources are normalized to "#version 330 core", which is stricter than the 460 they
// used to be forced to. A legacy shader using 420-era syntax without the matching #extension line
// is accepted by real drivers, so CompileShader retries the normalized source at 460 rather than
// failing. Only MobileGL's own normalization is rescued this way - an application-declared
// "#version 330" keeps strict 3.30 semantics, which is what the CTS negative-compile cases need.
TEST_F(ProgramUtilTest, CompileShaderRetriesAt460WhenNormalizedLegacyVersionRejects420Syntax) {
// used to be forced to. A shader declaring 330 while using 420-era syntax without the matching
// #extension line is accepted by real drivers, so CompileShader retries it at 460 instead of failing.
TEST_F(ProgramUtilTest, CompileShaderRetriesAt460WhenLegacyVersionRejects420Syntax) {
using namespace MG_Util::ShaderTranspiler;
String source = R"(#version 130
String source = R"(#version 330
layout(binding = 0) uniform sampler2D InSampler;
varying vec2 texCoord;
in vec2 texCoord;
out vec4 fragColor;
void main() {
gl_FragColor = texture2D(InSampler, texCoord);
fragColor = texture(InSampler, texCoord);
})";
PreprocessShaderSource(ShaderStage::Fragment, source);
// The normal path still emits 330 - the retry must not become the default.
@@ -957,21 +952,10 @@ void main() {
TEST_F(ProgramUtilTest, RetargetLegacyVersionDirectiveOnlyTouchesNormalizedDesktopCore) {
using namespace MG_Util::ShaderTranspiler;
// Only MobileGL's own normalization is retargetable, and it is recognised by the marker the
// preprocessor leaves on the directive line - so normalize a legacy source rather than
// hand-writing the directive the marker belongs to.
String normalized = "#version 130\nvoid main() {}\n";
PreprocessShaderSource(ShaderStage::Vertex, normalized);
ASSERT_EQ(normalized.find("#version 330 core "), 0u);
String normalized = "#version 330 core\nvoid main() {}\n";
EXPECT_TRUE(RetargetLegacyVersionDirectiveTo460(normalized));
EXPECT_EQ(normalized.find("#version 460 core"), 0u);
// An application that declared 330 itself keeps strict 3.30 semantics: raising it would
// re-legalize the CTS negative-compile cases.
String declared330 = "#version 330 core\nvoid main() {}\n";
EXPECT_FALSE(RetargetLegacyVersionDirectiveTo460(declared330));
EXPECT_EQ(declared330.find("#version 330 core"), 0u);
// Already modern: nothing to retarget.
String modern = "#version 460 core\nvoid main() {}\n";
EXPECT_FALSE(RetargetLegacyVersionDirectiveTo460(modern));
+2 -150
View File
@@ -33,7 +33,6 @@
#include <MG_Util/ShaderTranspiler/ShaderCompiler.h>
#include <MG_Util/ShaderTranspiler/ShaderSourceProcessor.h>
#include <MG_Util/Debug/Log.h>
#include <FastSTL/UnorderedMap.h>
namespace {
class DynamicParameterBackend final : public MobileGL::MG_Backend::BackendObject {
@@ -197,9 +196,7 @@ TEST(DirectGLESSanity, AdvertisesDepthTextureForGlmarkShadowScenes) {
EXPECT_NE(std::find(extensions.begin(), extensions.end(), MobileGL::E_GL_ARB_depth_texture), extensions.end());
}
// The advertised target went back to 3.3 (see "restore target GL version to 3.3"); Voxy only ever
// needed the extensions, which stay advertised, so assert what the backend really reports.
TEST(DirectGLESSanity, AdvertisesVoxyRequiredRenderingExtensions) {
TEST(DirectGLESSanity, AdvertisesVoxyRequiredRenderingExtensionsWithoutRaisingGLVersion) {
MobileGL::MG_Backend::DirectGLES::BackendObject_DirectGLES backend;
const auto& rendererInfo = backend.GetRendererInfo().RendererGLInfo;
const auto& extensions = rendererInfo.Extensions;
@@ -399,8 +396,7 @@ TEST(DirectVulkanSanity, RenderPassExtentUsesSwapchainSizeOnlyForDefaultFramebuf
MobileGL::IntVec2(512, 512));
}
// See the DirectGLES twin above: the target version is 3.3 again, the extensions are what matter.
TEST(DirectVulkanSanity, AdvertisesVoxyRequiredRenderingExtensions) {
TEST(DirectVulkanSanity, AdvertisesVoxyRequiredRenderingExtensionsWithoutRaisingGLVersion) {
MobileGL::MG_Backend::DirectVulkan::BackendObject_DirectVulkan backend;
const auto& rendererInfo = backend.GetRendererInfo().RendererGLInfo;
const auto& extensions = rendererInfo.Extensions;
@@ -519,50 +515,6 @@ TEST(DirectGLESSanity, PreservesHostPerStageImageUniformLimits) {
EXPECT_EQ(params.MaxComputeImageUniforms, 5);
}
TEST(FragmentInterpolationCapabilities, PlumbsGLESAndBothVulkanPropertyPaths) {
using namespace MobileGL;
MG_External::GLESCapabilities glesCaps;
glesCaps.MinFragmentInterpolationOffset = -0.75f;
glesCaps.MaxFragmentInterpolationOffset = 0.625f;
glesCaps.FragmentInterpolationOffsetBits = 6;
MG_Backend::DirectGLES::BackendObject_DirectGLES glesBackend;
glesBackend.ApplyGLESCapabilitiesForTesting(glesCaps);
EXPECT_FLOAT_EQ(glesBackend.GetDynamicParameters().MinFragmentInterpolationOffset, -0.75f);
EXPECT_FLOAT_EQ(glesBackend.GetDynamicParameters().MaxFragmentInterpolationOffset, 0.625f);
EXPECT_EQ(glesBackend.GetDynamicParameters().FragmentInterpolationOffsetBits, 6);
VkPhysicalDeviceProperties properties{};
// A common Vulkan limit pair: max is one representable 4-bit step below 0.5.
properties.limits.minInterpolationOffset = -0.5f;
properties.limits.maxInterpolationOffset = 0.4375f;
properties.limits.subPixelInterpolationOffsetBits = 4;
MG_External::VulkanCapabilities vkCaps;
MG_Util::BackendLoader::FillInVulkanCapabilities(vkCaps, properties);
EXPECT_FLOAT_EQ(vkCaps.MinFragmentInterpolationOffset, -0.5f);
EXPECT_FLOAT_EQ(vkCaps.MaxFragmentInterpolationOffset, 0.4375f);
EXPECT_EQ(vkCaps.FragmentInterpolationOffsetBits, 4);
vkCaps.MinFragmentInterpolationOffset = -0.875f;
vkCaps.MaxFragmentInterpolationOffset = 0.75f;
vkCaps.FragmentInterpolationOffsetBits = 7;
MG_Backend::DirectVulkan::BackendObject_DirectVulkan vkBackend;
vkBackend.ApplyVulkanCapabilitiesForTesting(vkCaps);
EXPECT_FLOAT_EQ(vkBackend.GetDynamicParameters().MinFragmentInterpolationOffset, -0.875f);
EXPECT_FLOAT_EQ(vkBackend.GetDynamicParameters().MaxFragmentInterpolationOffset, 0.75f);
EXPECT_EQ(vkBackend.GetDynamicParameters().FragmentInterpolationOffsetBits, 7);
// Invalid/zero host data cannot under-advertise the OpenGL 4 minimums.
MG_External::VulkanCapabilities invalidCaps;
invalidCaps.MinFragmentInterpolationOffset = 0.0f;
invalidCaps.MaxFragmentInterpolationOffset = 0.0f;
invalidCaps.FragmentInterpolationOffsetBits = 0;
vkBackend.ApplyVulkanCapabilitiesForTesting(invalidCaps);
EXPECT_LE(vkBackend.GetDynamicParameters().MinFragmentInterpolationOffset, -0.5f);
EXPECT_FLOAT_EQ(vkBackend.GetDynamicParameters().MaxFragmentInterpolationOffset, 0.4375f);
EXPECT_EQ(vkBackend.GetDynamicParameters().FragmentInterpolationOffsetBits, 4);
}
TEST(DirectVulkanSanity, AdvertisesSubgroupOnlyWhenVulkanReportsUsableSupport) {
using namespace MobileGL;
@@ -685,48 +637,6 @@ TEST(GetterSanity, ClampsMaxVertexAttribsToCurrentValueStorageCapacity) {
MG_State::pGLContext.reset();
}
TEST(GetterSanity, ReportsFragmentInterpolationLimitsForFloatAndIntegerQueries) {
using namespace MobileGL;
auto previousContext = Move(MG_State::pGLContext);
auto previousBackend = Move(MG_Backend::pActiveBackendObject);
MG_State::pGLContext = MakeUnique<MG_State::GLState::GLContext>();
MG_Backend::DynamicBackendParameters params;
params.MinFragmentInterpolationOffset = -0.75f;
params.MaxFragmentInterpolationOffset = 0.4375f;
params.FragmentInterpolationOffsetBits = 6;
MG_Backend::pActiveBackendObject = MakeUnique<DynamicParameterBackend>(params);
GLfloat floatValue = 0.0f;
MG_Impl::GLImpl::GetFloatv(GL_MIN_FRAGMENT_INTERPOLATION_OFFSET, &floatValue);
EXPECT_FLOAT_EQ(floatValue, -0.75f);
MG_Impl::GLImpl::GetFloatv(GL_MAX_FRAGMENT_INTERPOLATION_OFFSET, &floatValue);
EXPECT_FLOAT_EQ(floatValue, 0.4375f);
MG_Impl::GLImpl::GetFloatv(GL_FRAGMENT_INTERPOLATION_OFFSET_BITS, &floatValue);
EXPECT_FLOAT_EQ(floatValue, 6.0f);
GLint intValue = 0;
MG_Impl::GLImpl::GetIntegerv(GL_MIN_FRAGMENT_INTERPOLATION_OFFSET, &intValue);
EXPECT_EQ(intValue, -1);
MG_Impl::GLImpl::GetIntegerv(GL_MAX_FRAGMENT_INTERPOLATION_OFFSET, &intValue);
EXPECT_EQ(intValue, 0);
MG_Impl::GLImpl::GetIntegerv(GL_FRAGMENT_INTERPOLATION_OFFSET_BITS, &intValue);
EXPECT_EQ(intValue, 6);
GLboolean boolValue = GL_FALSE;
MG_Impl::GLImpl::GetBooleanv(GL_MIN_FRAGMENT_INTERPOLATION_OFFSET, &boolValue);
EXPECT_EQ(boolValue, GL_TRUE);
MG_Impl::GLImpl::GetBooleanv(GL_MAX_FRAGMENT_INTERPOLATION_OFFSET, &boolValue);
EXPECT_EQ(boolValue, GL_TRUE);
MG_Impl::GLImpl::GetBooleanv(GL_FRAGMENT_INTERPOLATION_OFFSET_BITS, &boolValue);
EXPECT_EQ(boolValue, GL_TRUE);
EXPECT_EQ(MG_Impl::GLImpl::GetError(), GL_NO_ERROR);
MG_Backend::pActiveBackendObject = Move(previousBackend);
MG_State::pGLContext = Move(previousContext);
}
TEST(GetterSanity, PerStageImageUniformQueriesMatchShaderCompilerLimits) {
using namespace MobileGL;
@@ -1826,61 +1736,3 @@ TEST(DirectGLESStateGuards, DefaultFramebufferBindGoesThroughShadow) {
FramebufferImpl::BindFramebufferId(GL_DRAW_FRAMEBUFFER, 7); // must reach the driver again
EXPECT_EQ(mocks.log.Count("BindFramebuffer:"), 3u);
}
// FastSTL::unordered_map::erase(iterator) regression coverage. The open-addressing
// iterator constructor snaps forward from a tombstoned slot to the successor, so
// erase must NOT advance the rebuilt iterator again: the old double-advance skipped
// one live element per erase, and erasing the element in the highest occupied
// bucket pushed the returned index past bucket_count where it never compared equal
// to end() again - erase-while-iterating sweeps (pipeline/program cache eviction)
// then ran off the bucket array and fed garbage handles to vkDestroyPipeline
// (device crash on first mass eviction during world load).
TEST(FastSTLSanity, EraseWhileIteratingVisitsEveryElementExactlyOnce) {
FastSTL::unordered_map<MobileGL::Uint64, MobileGL::Uint64> map;
constexpr MobileGL::Uint64 kCount = 1000;
for (MobileGL::Uint64 key = 0; key < kCount; ++key) {
map.emplace(key * 0x9e3779b97f4a7c15ull, key);
}
ASSERT_EQ(map.size(), kCount);
MobileGL::SizeT visited = 0;
for (auto it = map.begin(); it != map.end();) {
it = map.erase(it);
++visited;
ASSERT_LE(visited, kCount); // old code: runaway past end / skipped entries
}
EXPECT_EQ(visited, kCount);
EXPECT_EQ(map.size(), 0u);
}
TEST(FastSTLSanity, EraseReturnsTheSuccessorElement) {
FastSTL::unordered_map<MobileGL::Uint32, MobileGL::Uint32> map;
for (MobileGL::Uint32 key = 1; key <= 64; ++key) {
map.emplace(key, key);
}
// Erasing every other visited element must still visit all 64 exactly once:
// the iterator returned by erase names the very next element, not one past it.
MobileGL::SizeT visited = 0;
MobileGL::SizeT erased = 0;
for (auto it = map.begin(); it != map.end();) {
++visited;
if ((visited & 1) != 0) {
it = map.erase(it);
++erased;
} else {
++it;
}
ASSERT_LE(visited, 64u);
}
EXPECT_EQ(visited, 64u);
EXPECT_EQ(map.size(), 64u - erased);
}
TEST(FastSTLSanity, ErasingTheOnlyElementReturnsEnd) {
FastSTL::unordered_map<MobileGL::Uint32, MobileGL::Uint32> map;
map.emplace(42u, 1u);
auto next = map.erase(map.begin());
EXPECT_EQ(next, map.end());
EXPECT_TRUE(map.empty());
}
+3 -5
View File
@@ -1867,13 +1867,11 @@ TEST_F(TextureTest, DirectGLESTreats2DArrayAsSupportedTextureTarget) {
EXPECT_TRUE(IsSupportedTextureTarget(TextureTarget::Texture2DArray));
EXPECT_TRUE(IsSupportedTextureTarget(TextureTarget::Texture3D));
EXPECT_TRUE(IsSupportedTextureTarget(TextureTarget::Texture2D));
// Every desktop-only target is stored on an ES one (MapToBackendTextureTarget): 1D and
// 1D-array as 2D / 2D-array, matching SPIRV-Cross's ES 1D-as-2D shader emission, and
// rectangle as a plain 2D - it is single-level and already clamps, so only the
// non-normalized coordinates differ and LowerRectImages handles those.
// 1D and 1D-array are emulated as 2D / 2D-array (MapToBackendTextureTarget), matching
// SPIRV-Cross's ES 1D-as-2D shader emission; only rectangle textures stay unsupported.
EXPECT_TRUE(IsSupportedTextureTarget(TextureTarget::Texture1D));
EXPECT_TRUE(IsSupportedTextureTarget(TextureTarget::Texture1DArray));
EXPECT_TRUE(IsSupportedTextureTarget(TextureTarget::TextureRectangle));
EXPECT_FALSE(IsSupportedTextureTarget(TextureTarget::TextureRectangle));
}
// 2D-array textures keep their layer count constant across mip levels (GL 3.3 §3.9);
@@ -9,7 +9,6 @@
#include "Loader.h"
#include "MG_Util/Types.h"
#include <Config.h>
#include <cmath>
#if defined(_WIN32)
#ifndef WIN32_LEAN_AND_MEAN
#define WIN32_LEAN_AND_MEAN 1
@@ -824,12 +823,6 @@ namespace MobileGL::MG_Util::BackendLoader {
if (std::strcmp(extension, "GL_EXT_texture_norm16") == 0) {
caps.SupportsNorm16Texture = true;
}
if (std::strcmp(extension, "GL_EXT_render_snorm") == 0) {
caps.SupportsRenderSnorm = true;
}
if (std::strcmp(extension, "GL_EXT_sRGB_write_control") == 0) {
caps.SupportsSrgbWriteControl = true;
}
if (std::strcmp(extension, "GL_EXT_texture_filter_anisotropic") == 0) {
caps.SupportsTextureFilterAnisotropy = true;
}
@@ -845,14 +838,8 @@ namespace MobileGL::MG_Util::BackendLoader {
if (std::strcmp(extension, "GL_NV_shader_noperspective_interpolation") == 0) {
caps.SupportsNoperspectiveInterpolation = true;
}
if (std::strcmp(extension, "GL_OES_shader_multisample_interpolation") == 0) {
caps.SupportsShaderMultisampleInterpolation = true;
}
}
}
caps.SupportsShaderMultisampleInterpolation =
caps.SupportsShaderMultisampleInterpolation || caps.GLESVersion.Major > 3 ||
(caps.GLESVersion.Major == 3 && caps.GLESVersion.Minor >= 2);
// Detect optional raster/color-mask entry points by whether they loaded. glColorMaski is GLES
// 3.2 core (no extension string), so pointer presence is the reliable signal for all of these.
@@ -913,15 +900,6 @@ namespace MobileGL::MG_Util::BackendLoader {
GLint maxColorAttachments = 8;
GLint maxClipDistances = 8;
GLint maxViewports = 16;
GLfloat minFragmentInterpolationOffset = -0.5f;
GLfloat maxFragmentInterpolationOffset = 0.4375f;
GLint fragmentInterpolationOffsetBits = 4;
// Core minimums of both APIs (GL 4.6 table 23.53, ES 3.1 table 20.40); the probe
// below only ever widens them.
GLint minProgramTextureGatherOffset = -8;
GLint maxProgramTextureGatherOffset = 7;
GLint maxPatchVertices = 32;
GLint maxTessGenLevel = 64;
glesFuncs.glGetFloatv(GL_ALIASED_LINE_WIDTH_RANGE, aliasedLineWidthRange);
glesFuncs.glGetFloatv(GL_SMOOTH_LINE_WIDTH_RANGE, smoothLineWidthRange);
glesFuncs.glGetFloatv(GL_SMOOTH_LINE_WIDTH_GRANULARITY, &smoothLineWidthGranularity);
@@ -941,23 +919,6 @@ namespace MobileGL::MG_Util::BackendLoader {
glesFuncs.glGetIntegerv(GL_MAX_INTEGER_SAMPLES, &maxIntegerSamples);
glesFuncs.glGetIntegerv(GL_MAX_SAMPLES, &maxSamples);
glesFuncs.glGetIntegerv(GL_MAX_SAMPLE_MASK_WORDS, &maxSampleMaskWords);
// MobileGL's sample-mask state only stores a single 32-bit word (see
// RenderState::SampleMaskValue) and SampleMaski_State() rejects any
// maskNumber other than 0. Advertising the real driver's value here (e.g.
// NVIDIA's GLES driver reports 2) makes glSampleMaski(1, ...) - which
// dEQP's per-case gluStateReset always issues up to GL_MAX_SAMPLE_MASK_WORDS -
// raise GL_INVALID_VALUE and aborts the whole glcts process after every
// single test case. 1 is a spec-legal value (the minimum required), so cap
// to what is actually implemented instead of forwarding the raw driver limit.
maxSampleMaskWords = std::min(maxSampleMaskWords, 1);
glesFuncs.glGetIntegerv(GL_MAX_PATCH_VERTICES, &maxPatchVertices);
glesFuncs.glGetIntegerv(GL_MAX_TESS_GEN_LEVEL, &maxTessGenLevel);
glesFuncs.glGetIntegerv(GL_MIN_PROGRAM_TEXTURE_GATHER_OFFSET, &minProgramTextureGatherOffset);
glesFuncs.glGetIntegerv(GL_MAX_PROGRAM_TEXTURE_GATHER_OFFSET, &maxProgramTextureGatherOffset);
// A driver that leaves the probe untouched (pre-ES 3.1, or an ignored enum) must not
// drag the advertised range below what GL 4.0 requires of us.
minProgramTextureGatherOffset = std::min(minProgramTextureGatherOffset, -8);
maxProgramTextureGatherOffset = std::max(maxProgramTextureGatherOffset, 7);
glesFuncs.glGetIntegerv(GL_MAX_TEXTURE_IMAGE_UNITS, &maxTextureImageUnits);
glesFuncs.glGetIntegerv(GL_MAX_VERTEX_TEXTURE_IMAGE_UNITS, &maxVertexTextureImageUnits);
glesFuncs.glGetIntegerv(GL_MAX_COMPUTE_TEXTURE_IMAGE_UNITS, &maxComputeTextureImageUnits);
@@ -989,29 +950,6 @@ namespace MobileGL::MG_Util::BackendLoader {
glesFuncs.glGetIntegerv(GL_MAX_VIEWPORTS, &maxViewports);
glesFuncs.glGetIntegerv(GL_MAX_VIEWPORT_DIMS, maxViewportDims);
glesFuncs.glGetIntegerv(GL_VIEWPORT_SUBPIXEL_BITS, &viewportSubpixelBits);
if (caps.SupportsShaderMultisampleInterpolation && glesFuncs.glGetFloatv) {
const auto drainErrors = [&glesFuncs]() {
Bool hadError = false;
if (glesFuncs.glGetError) {
while (glesFuncs.glGetError() != GL_NO_ERROR) hadError = true;
}
return hadError;
};
// Isolate these optional queries from errors raised by preceding capability
// probes, then consume any query error so initialization never leaks it into
// the application's first glGetError call.
drainErrors();
glesFuncs.glGetFloatv(GL_MIN_FRAGMENT_INTERPOLATION_OFFSET, &minFragmentInterpolationOffset);
glesFuncs.glGetFloatv(GL_MAX_FRAGMENT_INTERPOLATION_OFFSET, &maxFragmentInterpolationOffset);
glesFuncs.glGetIntegerv(GL_FRAGMENT_INTERPOLATION_OFFSET_BITS, &fragmentInterpolationOffsetBits);
if (drainErrors()) {
MGLOG_W("Fragment interpolation limit query failed; using OpenGL minimums");
minFragmentInterpolationOffset = -0.5f;
maxFragmentInterpolationOffset = 0.4375f;
fragmentInterpolationOffsetBits = 4;
}
}
// Only legal to query once the extension has been seen in the loop above, hence not batched
// with the unconditional probes: on a driver without it this raises GL_INVALID_ENUM.
if (caps.SupportsTextureFilterAnisotropy) {
@@ -1041,10 +979,6 @@ namespace MobileGL::MG_Util::BackendLoader {
caps.MaxIntegerSamples = maxIntegerSamples;
caps.MaxSamples = maxSamples;
caps.MaxSampleMaskWords = maxSampleMaskWords;
caps.MaxPatchVertices = maxPatchVertices;
caps.MaxTessGenLevel = maxTessGenLevel;
caps.MinProgramTextureGatherOffset = minProgramTextureGatherOffset;
caps.MaxProgramTextureGatherOffset = maxProgramTextureGatherOffset;
caps.MaxTextureImageUnits = maxTextureImageUnits;
caps.MaxVertexTextureImageUnits = maxVertexTextureImageUnits;
caps.MaxComputeTextureImageUnits = maxComputeTextureImageUnits;
@@ -1056,18 +990,6 @@ namespace MobileGL::MG_Util::BackendLoader {
caps.MaxComputeWorkGroupInvocations = maxComputeWorkGroupInvocations;
caps.MaxShaderStorageBufferBindings = maxShaderStorageBufferBindings;
caps.MaxTextureBufferSize = maxTextureBufferSize;
// Through glesFuncs, like every other capability query here: a bare glGetIntegerv resolves
// to MobileGL's own exported entry point, which answers this pname from the very
// capability table being filled in - so the driver's real alignment never arrived and the
// backend reported an unconstrained offset it cannot honour.
GLint textureBufferOffsetAlignment = 1;
glesFuncs.glGetIntegerv(GL_TEXTURE_BUFFER_OFFSET_ALIGNMENT, &textureBufferOffsetAlignment);
// Core in ES 3.2 and in EXT_texture_buffer; an older context rejects the pname and leaves
// the default in place.
if (glesFuncs.glGetError) {
while (glesFuncs.glGetError() != GL_NO_ERROR) {}
}
caps.TextureBufferOffsetAlignment = std::max(1, textureBufferOffsetAlignment);
caps.MaxUniformBufferBindings = maxUniformBufferBindings;
caps.MaxUniformBlockSize = maxUniformBlockSize;
caps.MaxImageUnits = maxImageUnits;
@@ -1085,19 +1007,6 @@ namespace MobileGL::MG_Util::BackendLoader {
caps.ViewportBoundsRangeMin = viewportBoundsRange[0];
caps.ViewportBoundsRangeMax = viewportBoundsRange[1];
caps.ViewportSubpixelBits = viewportSubpixelBits;
caps.MinFragmentInterpolationOffset =
std::isfinite(minFragmentInterpolationOffset) && minFragmentInterpolationOffset <= -0.5f
? minFragmentInterpolationOffset
: -0.5f;
caps.MaxFragmentInterpolationOffset = 0.4375f;
caps.FragmentInterpolationOffsetBits = 4;
if (fragmentInterpolationOffsetBits >= 4 && std::isfinite(maxFragmentInterpolationOffset)) {
const Float requiredMaxOffset = 0.5f - std::ldexp(1.0f, -fragmentInterpolationOffsetBits);
if (maxFragmentInterpolationOffset >= requiredMaxOffset) {
caps.MaxFragmentInterpolationOffset = maxFragmentInterpolationOffset;
caps.FragmentInterpolationOffsetBits = fragmentInterpolationOffsetBits;
}
}
MGLOG_I(" GL_ALIASED_LINE_WIDTH_RANGE: [%.3f, %.3f]", caps.AliasedLineWidthRangeMin,
caps.AliasedLineWidthRangeMax);
MGLOG_I(" GL_SMOOTH_LINE_WIDTH_RANGE: [%.3f, %.3f]", caps.SmoothLineWidthRangeMin,
@@ -1031,13 +1031,6 @@ namespace MobileGL {
String GLESShadingLanguageVersionString;
Bool SupportsPersistentMapping = false;
Bool SupportsNorm16Texture = false;
// GL_EXT_render_snorm is present, so the signed-normalized formats are colour-renderable
// (and usable as multisample texture storage) rather than texture-only.
Bool SupportsRenderSnorm = false;
// GL_EXT_sRGB_write_control is present, so GL_FRAMEBUFFER_SRGB can be turned off.
// GLES has no such switch in core: writes into an sRGB attachment are ALWAYS encoded,
// while desktop GL leaves GL_FRAMEBUFFER_SRGB disabled by default and writes raw.
Bool SupportsSrgbWriteControl = false;
// GL_EXT_texture_filter_anisotropic is present, so sampler/texture
// anisotropy may be forwarded without raising GL_INVALID_ENUM in GLES.
Bool SupportsTextureFilterAnisotropy = false;
@@ -1062,9 +1055,6 @@ namespace MobileGL {
// SPIRV-Cross's `#extension ... : require` would fail to compile and MobileGL falls back
// to stripping the NoPerspective decoration (smooth interpolation) via StripNoPerspectivePass.
Bool SupportsNoperspectiveInterpolation = false;
// GLES 3.2 core or GL_OES_shader_multisample_interpolation exposes
// interpolateAtOffset and the three fragment-offset limit queries.
Bool SupportsShaderMultisampleInterpolation = false;
// GL_RENDERER contains "ANGLE".
Bool IsAngleRenderer = false;
// GL_RENDERER contains both "ANGLE" and "llvmpipe".
@@ -1102,10 +1092,6 @@ namespace MobileGL {
Int MaxIntegerSamples = 1;
Int MaxSamples = 1;
Int MaxSampleMaskWords = 1;
Int MaxPatchVertices = 32;
Int MaxTessGenLevel = 64;
Int MinProgramTextureGatherOffset = -8;
Int MaxProgramTextureGatherOffset = 7;
Int MaxTextureImageUnits = 32;
Int MaxVertexTextureImageUnits = 32;
Int MaxComputeTextureImageUnits = 32;
@@ -1117,8 +1103,6 @@ namespace MobileGL {
Int MaxComputeWorkGroupInvocations = 128;
Int MaxShaderStorageBufferBindings = 8;
Int MaxTextureBufferSize = 65536;
// GL_TEXTURE_BUFFER_OFFSET_ALIGNMENT; 1 means the offset is unconstrained.
Int TextureBufferOffsetAlignment = 1;
Int MaxUniformBufferBindings = 24;
Int MaxUniformBlockSize = 16384;
Int MaxImageUnits = 8;
@@ -1136,9 +1120,6 @@ namespace MobileGL {
Float ViewportBoundsRangeMin = 0.0f;
Float ViewportBoundsRangeMax = 0.0f;
Int ViewportSubpixelBits = 0;
Float MinFragmentInterpolationOffset = -0.5f;
Float MaxFragmentInterpolationOffset = 0.4375f;
Int FragmentInterpolationOffsetBits = 4;
};
} // namespace MG_External
@@ -9,7 +9,6 @@
#include "Loader.h"
#include <Config.h>
#include <cmath>
namespace MobileGL::MG_Util::BackendLoader {
namespace {
@@ -41,25 +40,6 @@ namespace MobileGL::MG_Util::BackendLoader {
return MaxSampleCountFromFlags(commonFlags);
}
void FillFragmentInterpolationLimits(MG_External::VulkanCapabilities& caps,
const VkPhysicalDeviceLimits& limits) {
caps.MinFragmentInterpolationOffset =
std::isfinite(limits.minInterpolationOffset) && limits.minInterpolationOffset <= -0.5f
? limits.minInterpolationOffset
: -0.5f;
caps.MaxFragmentInterpolationOffset = 0.4375f;
caps.FragmentInterpolationOffsetBits = 4;
const Int bits = static_cast<Int>(limits.subPixelInterpolationOffsetBits);
if (bits >= 4 && std::isfinite(limits.maxInterpolationOffset)) {
const Float requiredMaxOffset = 0.5f - std::ldexp(1.0f, -bits);
if (limits.maxInterpolationOffset >= requiredMaxOffset) {
caps.MaxFragmentInterpolationOffset = limits.maxInterpolationOffset;
caps.FragmentInterpolationOffsetBits = bits;
}
}
}
VulkanDynamicFunctions LoadVulkanDynamicFunctions(VkInstance instance) {
VulkanDynamicFunctions loaded{};
if (instance == VK_NULL_HANDLE) {
@@ -177,8 +157,6 @@ namespace MobileGL::MG_Util::BackendLoader {
caps.MaxComputeWorkGroupInvocations = static_cast<Int>(p.limits.maxComputeWorkGroupInvocations);
caps.MaxShaderStorageBufferBindings = static_cast<Int>(p.limits.maxDescriptorSetStorageBuffers);
caps.MaxTextureBufferSize = static_cast<Int>(p.limits.maxTexelBufferElements);
caps.TextureBufferOffsetAlignment =
static_cast<Int>(std::max<VkDeviceSize>(1, p.limits.minTexelBufferOffsetAlignment));
caps.MaxUniformBufferBindings = static_cast<Int>(p.limits.maxDescriptorSetUniformBuffers);
caps.MaxUniformBlockSize = static_cast<Int>(p.limits.maxUniformBufferRange);
caps.MaxImageUnits = static_cast<Int>(p.limits.maxPerStageDescriptorStorageImages);
@@ -193,7 +171,6 @@ namespace MobileGL::MG_Util::BackendLoader {
caps.ViewportBoundsRangeMin = p.limits.viewportBoundsRange[0];
caps.ViewportBoundsRangeMax = p.limits.viewportBoundsRange[1];
caps.ViewportSubpixelBits = static_cast<Int>(p.limits.viewportSubPixelBits);
FillFragmentInterpolationLimits(caps, p.limits);
VkPhysicalDeviceFeatures supportedFeatures{};
vkGetPhysicalDeviceFeatures(physicalDevice, &supportedFeatures);
@@ -270,8 +247,6 @@ namespace MobileGL::MG_Util::BackendLoader {
caps.MaxComputeWorkGroupInvocations = static_cast<Int>(properties.limits.maxComputeWorkGroupInvocations);
caps.MaxShaderStorageBufferBindings = static_cast<Int>(properties.limits.maxDescriptorSetStorageBuffers);
caps.MaxTextureBufferSize = static_cast<Int>(properties.limits.maxTexelBufferElements);
caps.TextureBufferOffsetAlignment =
static_cast<Int>(std::max<VkDeviceSize>(1, properties.limits.minTexelBufferOffsetAlignment));
caps.MaxUniformBufferBindings = static_cast<Int>(properties.limits.maxDescriptorSetUniformBuffers);
caps.MaxUniformBlockSize = static_cast<Int>(properties.limits.maxUniformBufferRange);
caps.MaxImageUnits = static_cast<Int>(properties.limits.maxPerStageDescriptorStorageImages);
@@ -286,7 +261,6 @@ namespace MobileGL::MG_Util::BackendLoader {
caps.ViewportBoundsRangeMin = properties.limits.viewportBoundsRange[0];
caps.ViewportBoundsRangeMax = properties.limits.viewportBoundsRange[1];
caps.ViewportSubpixelBits = static_cast<Int>(properties.limits.viewportSubPixelBits);
FillFragmentInterpolationLimits(caps, properties.limits);
caps.SupportsWideLines = false;
// This helper only receives properties, not VkPhysicalDeviceFeatures. Leave optional
// stage writes disabled rather than inferring them from descriptor limits alone.
@@ -54,8 +54,6 @@ namespace MobileGL {
Int MaxComputeWorkGroupInvocations = 128;
Int MaxShaderStorageBufferBindings = 8;
Int MaxTextureBufferSize = 65536;
// GL_TEXTURE_BUFFER_OFFSET_ALIGNMENT; 1 means the offset is unconstrained.
Int TextureBufferOffsetAlignment = 1;
Int MaxUniformBufferBindings = 24;
Int MaxUniformBlockSize = 16384;
Int MaxImageUnits = 8;
@@ -70,9 +68,6 @@ namespace MobileGL {
Float ViewportBoundsRangeMin = 0.0f;
Float ViewportBoundsRangeMax = 0.0f;
Int ViewportSubpixelBits = 0;
Float MinFragmentInterpolationOffset = -0.5f;
Float MaxFragmentInterpolationOffset = 0.4375f;
Int FragmentInterpolationOffsetBits = 4;
Bool SupportsWideLines = false;
// Storage-image descriptors are limited per stage by
// maxPerStageDescriptorStorageImages, but writes/atomics outside compute additionally
@@ -28,7 +28,6 @@ namespace MobileGL {
bool IsStencilFormatInternalFormat(TextureInternalFormat internalformat) {
switch (internalformat) {
case TextureInternalFormat::StencilIndex8:
case TextureInternalFormat::Depth24Stencil8:
case TextureInternalFormat::Depth32FStencil8:
case TextureInternalFormat::DepthStencil:

Some files were not shown because too many files have changed in this diff Show More