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
121 changed files with 1874 additions and 11916 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 -14
View File
@@ -194,6 +194,9 @@ set(SOURCE_FILES
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
@@ -455,21 +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"
@@ -477,7 +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();
-16
View File
@@ -220,15 +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);
Int64 (*GetGpuTimestampNs)(); // glGetInteger64v(GL_TIMESTAMP); 0 if unsupported
};
struct GlobalBackendFunctionsTable {
@@ -309,13 +300,6 @@ 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;
SizeT MaxShaderStorageBlockSize = 128 * 1024 * 1024;
Uint32 SubgroupSize = 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 {
@@ -604,7 +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.
@@ -935,16 +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;
funcsTable.GL.IsQueryResultAvailable = IsQueryResultAvailable;
funcsTable.GL.GetQueryResult64 = GetQueryResult64;
funcsTable.GL.DeleteBackendQuery = DeleteBackendQuery;
funcsTableInitialized = true;
}
return funcsTable;
@@ -1040,24 +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;
+20 -66
View File
@@ -449,6 +449,20 @@ namespace MobileGL::MG_Backend::DirectGLES {
}
}
}
// Textures attached only to the READ framebuffer (blit / ReadPixels sources) need
// their content synced too, or the backend reads stale texel data.
const auto& readFBO =
MG_State::pGLContext->GetFramebufferBindingSlot(FramebufferTarget::Read).GetBoundObject();
if (readFBO && readFBO != currentFBO) {
for (const auto& attachment : readFBO->GetAllAttachmentObjects()) {
if (!attachment.IsTexture()) continue;
auto& textureObject = attachment.GetTexture();
if (textureObject) {
SyncTextureObjectToBackend(textureObject);
}
}
}
}
static Bool SupportsLayeredImageBinding(TextureTarget target) {
@@ -3632,16 +3646,12 @@ namespace MobileGL::MG_Backend::DirectGLES {
static Bool IsLegacyNativeReadPixelsFormat(GLenum format) {
return format == GL_RGBA || format == GL_RGBA_INTEGER || format == GL_RED || format == GL_RED_INTEGER ||
format == GL_DEPTH_COMPONENT || format == GL_STENCIL_INDEX || format == GL_DEPTH_STENCIL;
format == GL_DEPTH_COMPONENT || format == GL_STENCIL_INDEX;
}
static Bool IsLegacyNativeReadPixelsType(GLenum type) {
// GL_UNSIGNED_INT_24_8 / GL_FLOAT_32_UNSIGNED_INT_24_8_REV are only ever valid
// paired with GL_DEPTH_STENCIL (packed_depth_stencil.verify_read_pixels); the real
// driver already implements this readback natively.
return type == GL_UNSIGNED_BYTE || type == GL_UNSIGNED_INT || type == GL_UNSIGNED_INT_2_10_10_10_REV ||
type == GL_INT || type == GL_FLOAT || type == GL_UNSIGNED_INT_24_8 ||
type == GL_FLOAT_32_UNSIGNED_INT_24_8_REV;
type == GL_INT || type == GL_FLOAT;
}
void ReadPixels(GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, void* pixels) {
@@ -3786,9 +3796,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
if (format == GL_RGBA_INTEGER) {
return type == GL_INT || type == GL_UNSIGNED_INT || type == GL_UNSIGNED_INT_2_10_10_10_REV;
}
if (format == GL_DEPTH_STENCIL) {
return type == GL_UNSIGNED_INT_24_8 || type == GL_FLOAT_32_UNSIGNED_INT_24_8_REV;
}
return false;
}
@@ -3857,14 +3864,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
MGLOG_D("GetTexImage: attaching level %d to the scratch FBO", level);
const GLenum backendAttachTarget = TextureImpl::ConvertTextureUploadTargetToBackendGLEnum(
MG_Util::ConvertGLEnumToTextureUploadTarget(target));
// GL_DEPTH_STENCIL can't be attached as a color attachment (glCheckFramebufferStatus
// would report it incomplete); it has its own combined depth+stencil attachment point.
// glReadBuffer only selects among color attachments, so it does not apply here.
if (format == GL_DEPTH_STENCIL) {
ScratchFBOImpl::EnsureDepthAttachment2D(
tempFB, GL_READ_FRAMEBUFFER, backendTexId,
backendAttachTarget == GL_UNKNOWN_MGL ? target : backendAttachTarget, level, /*withStencil=*/true);
} else if (backendAttachTarget == GL_TEXTURE_3D || backendAttachTarget == GL_TEXTURE_2D_ARRAY) {
if (backendAttachTarget == GL_TEXTURE_3D || backendAttachTarget == GL_TEXTURE_2D_ARRAY) {
// ES cannot attach 3D/array textures through glFramebufferTexture2D; read layer 0. Reads
// of deeper slices are served from the CPU shadow instead (see the shadow-first branch).
ScratchFBOImpl::EnsureColorAttachmentLayer(tempFB, GL_READ_FRAMEBUFFER, backendTexId, level, 0);
@@ -3873,10 +3873,8 @@ namespace MobileGL::MG_Backend::DirectGLES {
tempFB, GL_READ_FRAMEBUFFER, backendTexId,
backendAttachTarget == GL_UNKNOWN_MGL ? target : backendAttachTarget, level);
}
if (format != GL_DEPTH_STENCIL) {
MGLOG_D("GetTexImage: glReadBuffer(GL_COLOR_ATTACHMENT0)");
ScratchFBOImpl::EnsureReadBuffer(tempFB, GL_COLOR_ATTACHMENT0);
}
MGLOG_D("GetTexImage: glReadBuffer(GL_COLOR_ATTACHMENT0)");
ScratchFBOImpl::EnsureReadBuffer(tempFB, GL_COLOR_ATTACHMENT0);
GLenum fbStatus = g_GLESFuncs.glCheckFramebufferStatus(GL_READ_FRAMEBUFFER);
MGLOG_D("GetTexImage: GL_READ_FRAMEBUFFER status = %s", MG_Util::ConvertGLEnumToString(fbStatus).c_str());
@@ -4360,11 +4358,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
struct GLESQueryObject {
GLuint queryId = 0;
Uint contextGeneration = 0;
// GL_ANY_SAMPLES_PASSED result is 0/1 and only ever reachable through the
// core (non-extension) glGetQueryObjectuiv getter - GL_EXT_disjoint_timer_query's
// 64-bit glGetQueryObjectui64vEXT is timer-specific and may be entirely absent
// on drivers that otherwise fully support core ES3 occlusion queries.
Bool isOcclusion = false;
};
}
@@ -4557,38 +4550,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
return new GLESQueryObject{queryId, g_syncContextGeneration};
}
// GL_SAMPLES_PASSED/GL_ANY_SAMPLES_PASSED(_CONSERVATIVE) occlusion queries. Unlike the
// timer queries above, these are core ES 3.0 (no GL_EXT_disjoint_timer_query needed).
Bool AreOcclusionQueriesSupported() {
return g_GLESFuncs.glGenQueries && g_GLESFuncs.glDeleteQueries && g_GLESFuncs.glBeginQuery &&
g_GLESFuncs.glEndQuery && g_GLESFuncs.glGetQueryObjectuiv;
}
BackendQueryHandle BeginOcclusionQuery() {
if (!IsBackendContextCurrentOnThisThread() || !AreOcclusionQueriesSupported()) {
return nullptr;
}
GLuint queryId = 0;
g_GLESFuncs.glGenQueries(1, &queryId);
if (queryId == 0) {
return nullptr;
}
// ES only implements the boolean ANY_SAMPLES_PASSED variant, not an exact
// GL_SAMPLES_PASSED count; the frontend already coerces ANY_SAMPLES_PASSED*
// targets to boolean, and desktop GL_SAMPLES_PASSED reads a 0/1 approximation.
g_GLESFuncs.glBeginQuery(GL_ANY_SAMPLES_PASSED, queryId);
return new GLESQueryObject{queryId, g_syncContextGeneration, /*isOcclusion=*/true};
}
void EndOcclusionQuery(BackendQueryHandle handle) {
const auto* query = static_cast<GLESQueryObject*>(handle);
if (query == nullptr || query->contextGeneration != g_syncContextGeneration ||
!IsBackendContextCurrentOnThisThread() || !g_GLESFuncs.glEndQuery) {
return;
}
g_GLESFuncs.glEndQuery(GL_ANY_SAMPLES_PASSED);
}
Bool IsQueryResultAvailable(BackendQueryHandle handle) {
const auto* query = static_cast<GLESQueryObject*>(handle);
// Null/stale handles report available so the frontend proceeds to
@@ -4615,8 +4576,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
// report it as produced and let the frontend cache it and release
// the handle.
if (query == nullptr || query->contextGeneration != g_syncContextGeneration ||
!g_GLESFuncs.glGetQueryObjectuiv ||
(!query->isOcclusion && !g_GLESFuncs.glGetQueryObjectui64vEXT)) {
!g_GLESFuncs.glGetQueryObjectuiv || !g_GLESFuncs.glGetQueryObjectui64vEXT) {
return true;
}
// A thread that does not own the ES context cannot issue GL calls,
@@ -4654,12 +4614,6 @@ namespace MobileGL::MG_Backend::DirectGLES {
// context switch) the result may be garbage, which is tolerable for
// an F3 GPU% readout, and consuming the latched flag here could hide
// the event from another observer.
if (query->isOcclusion) {
GLuint result32 = 0;
g_GLESFuncs.glGetQueryObjectuiv(query->queryId, GL_QUERY_RESULT, &result32);
*outNanoseconds = static_cast<Uint64>(result32);
return true;
}
GLuint64 result = 0;
g_GLESFuncs.glGetQueryObjectui64vEXT(query->queryId, GL_QUERY_RESULT, &result);
*outNanoseconds = static_cast<Uint64>(result);
@@ -130,11 +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);
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
+104 -2
View File
@@ -1262,15 +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;
@@ -1278,7 +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;
}
@@ -3265,6 +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());
@@ -3287,6 +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
@@ -3345,6 +3441,12 @@ 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 = EmulateBaseInstanceInVertexShader(std::move(source), glShaderType);
@@ -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>
+35
View File
@@ -342,6 +342,41 @@ namespace MobileGL::MG_Backend::DirectGLES {
}
return result;
}
String RemoveClipDistanceRedeclaration(const String& glslCode) {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
#endif
// 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*$)");
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);
if (!std::regex_match(line, redeclarationRegex)) {
if (!firstLine) {
result += '\n';
}
result += line;
firstLine = false;
}
if (lastLine) {
break;
}
lineStart = lineEnd + 1;
}
return result;
}
} // namespace PrgramImpl
namespace Utils {
+1
View File
@@ -105,6 +105,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
Uint32 unormOutputMask);
String ForceFlatIntegerVaryings(const String& glslCode, GLenum shaderType);
String RemoveLayoutBinding(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 {
@@ -526,11 +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};
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);
}
@@ -636,15 +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;
@@ -807,23 +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
@@ -1659,26 +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;
}
*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
@@ -1711,47 +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();
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) {
@@ -244,108 +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");
@@ -65,26 +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
@@ -107,26 +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()) {
@@ -2342,17 +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());
@@ -2365,29 +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);
}
}
// 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
@@ -2484,42 +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,
@@ -386,28 +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,
@@ -448,48 +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) {
@@ -993,11 +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;
@@ -1077,7 +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) {
@@ -1091,9 +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());
@@ -1191,47 +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,
@@ -1370,34 +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;
@@ -184,37 +164,10 @@ namespace MobileGL::MG_Backend::DirectVulkan {
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.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)));
}
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);
@@ -227,33 +180,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
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;
@@ -50,9 +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;
VkPipelineVertexInputStateCreateInfo state{
VK_STRUCTURE_TYPE_PIPELINE_VERTEX_INPUT_STATE_CREATE_INFO
};
@@ -70,14 +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
@@ -92,19 +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
@@ -22,10 +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 =
@@ -145,15 +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
@@ -469,10 +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;
@@ -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;
}
@@ -1225,30 +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::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;
}
@@ -1256,12 +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);
@@ -1309,22 +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());
@@ -1336,25 +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,
@@ -1368,13 +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) {
@@ -1444,23 +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;
@@ -1470,17 +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,
@@ -1506,89 +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()) &&
@@ -1598,7 +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) {
@@ -1607,10 +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;
}
@@ -1625,10 +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;
@@ -1650,37 +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(
@@ -1719,21 +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;
@@ -1750,8 +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) {
@@ -1801,28 +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();
@@ -2024,119 +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;
@@ -2189,14 +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;
@@ -2204,24 +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);
}
@@ -2252,23 +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);
@@ -324,32 +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;
// 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
@@ -396,9 +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);
@@ -429,28 +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;
@@ -463,45 +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;
@@ -492,59 +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;
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.
VkBufferObject m_xfbCounterBuffer;
// Non-zero while inside a GL Begin/End with at least one captured draw
// recorded; selects counter-buffer resume on the next captured draw.
Bool m_xfbCountersValid = false;
Uint64 m_xfbLastSeenGeneration = 0;
// 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);
// 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;
@@ -557,30 +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;
@@ -609,61 +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;
@@ -726,8 +517,6 @@ namespace MobileGL::MG_Backend::DirectVulkan {
void CreateInstance();
VkResult SetupDebugMessenger();
VkResult DestroyDebugMessenger();
VkResult SetupDebugReportCallback();
void DestroyDebugReportCallback();
VkDebugUtilsMessengerCreateInfoEXT PopulateDebugMessengerCreateInfo();
void CreateSurface();
void PickPhysicalDevice();
@@ -746,10 +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);
@@ -811,10 +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]*
@@ -1351,14 +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);
@@ -1392,14 +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);
@@ -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 -227
View File
@@ -48,73 +48,6 @@ 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;
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);
}
static Bool ValidatePrimitiveModeForBackend(const char* functionName, GLenum mode) {
const auto& activeBackendObject = MG_Backend::pActiveBackendObject;
if (!activeBackendObject) {
@@ -124,6 +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(
@@ -133,38 +75,6 @@ namespace MobileGL::MG_Impl::GLImpl {
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.
if (MG_State::pGLContext->IsTransformFeedbackActive() &&
!(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;
}
@@ -492,14 +402,12 @@ namespace MobileGL::MG_Impl::GLImpl {
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);
}
@@ -536,133 +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.
const SizeT usedBufferCount = program->GetTransformFeedbackBufferCount();
for (SizeT i = 0; i < usedBufferCount; ++i) {
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);
}
// 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) {
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();
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);
}
} // namespace MobileGL::MG_Impl::GLImpl
@@ -11,8 +11,6 @@
namespace MobileGL::MG_Impl::GLImpl {
/* @INSERTION_POINT:FUNCTION_DECLARATION@ */
void BeginTransformFeedback(GLenum primitiveMode);
void EndTransformFeedback(void);
void DispatchCompute(GLuint numGroupsX, GLuint numGroupsY, GLuint numGroupsZ);
void DispatchComputeIndirect(GLintptr indirect);
void MemoryBarrier(GLbitfield barriers);
@@ -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)
@@ -45,32 +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 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) {
@@ -169,148 +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) {
@@ -576,27 +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
@@ -613,40 +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()) {
@@ -710,10 +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>(
@@ -1677,40 +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()) {
@@ -1756,10 +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>(
+2 -53
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;
@@ -1927,22 +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 = 0;
break;
case GL_MAX_TEXTURE_IMAGE_UNITS:
*params = dynamicParameters.MaxTextureImageUnits;
break;
@@ -1999,15 +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;
+13 -108
View File
@@ -49,18 +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;
@@ -613,18 +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(
@@ -644,6 +624,9 @@ namespace MobileGL::MG_Impl::GLImpl {
case GL_PROGRAM_BINARY_LENGTH:
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:
@@ -844,13 +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;
}
@@ -860,18 +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) {
@@ -915,16 +892,6 @@ namespace MobileGL::MG_Impl::GLImpl {
void UseProgram_State(GLuint program) {
MGLOG_D("UseProgram_State: program=%u", program);
// GL 3.3 core 2.11.3: the program in use may not change while transform
// feedback is active (there is no pause in 3.3).
if (MG_State::pGLContext->IsTransformFeedbackActive()) {
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;
@@ -2250,66 +2217,4 @@ namespace MobileGL::MG_Impl::GLImpl {
void ValidateProgram(GLuint program) {
ValidateProgram_State(program);
}
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] : "");
}
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
@@ -138,7 +138,4 @@ namespace MobileGL::MG_Impl::GLImpl {
GLint GetProgramResourceLocationIndex(GLuint program, GLenum programInterface, const GLchar* name);
void ShaderStorageBlockBinding(GLuint program, GLuint storageBlockIndex, GLuint storageBlockBinding);
void ValidateProgram(GLuint program);
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
+20 -127
View File
@@ -27,8 +27,6 @@ namespace MobileGL::MG_Impl::GLImpl {
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
@@ -43,11 +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;
@@ -133,11 +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.
@@ -192,24 +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) {
@@ -233,15 +204,10 @@ namespace MobileGL::MG_Impl::GLImpl {
}
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;
}
@@ -255,13 +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) {
@@ -277,72 +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);
@@ -388,25 +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: {
@@ -414,13 +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 =
@@ -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 -16
View File
@@ -2820,22 +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);
-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
@@ -231,21 +231,6 @@ 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,
@@ -133,11 +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);
-48
View File
@@ -211,47 +211,6 @@ namespace MobileGL {
void SetScissorBox(IntVec4 box); // x, y, width, height
const IntVec4& GetScissorBox() const; // x, y, width, height
// Transform feedback (GL 3.0 core Begin/End; no feedback objects yet)
void BeginTransformFeedback(GLenum primitiveMode, const SharedPtr<ProgramObject>& program) {
m_transformFeedbackActive = true;
m_transformFeedbackPrimitiveMode = primitiveMode;
m_transformFeedbackProgram = program;
++m_transformFeedbackGeneration;
m_transformFeedbackCapturedVertices = 0;
m_transformFeedbackInputPrimitives = 0;
}
void EndTransformFeedback() {
m_transformFeedbackActive = false;
m_transformFeedbackProgram.reset();
}
Bool IsTransformFeedbackActive() const { return m_transformFeedbackActive; }
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; }
// 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; }
// Framebuffer
void GenFramebufferNames(Uint number, Vector<Uint>& framebuffers);
const SharedPtr<FramebufferObject>& GetFramebufferObject(Uint index);
@@ -284,13 +243,6 @@ namespace MobileGL {
BufferState m_bufferState;
VertexArrayState m_vertexArrayState;
Array<CurrentVertexAttributeValue, VertexArrayObject::MAX_VERTEX_ATTRIBS> m_currentVertexAttributes{};
Bool m_transformFeedbackActive = false;
GLenum m_transformFeedbackPrimitiveMode = GL_POINTS;
SharedPtr<ProgramObject> m_transformFeedbackProgram;
Uint64 m_transformFeedbackGeneration = 0;
Uint64 m_transformFeedbackPrimitiveCounter = 0;
Uint64 m_transformFeedbackCapturedVertices = 0;
Uint64 m_transformFeedbackInputPrimitives = 0;
TextureState m_textureState;
ProgramState m_programState;
RenderState m_renderState;
@@ -170,230 +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_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;
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;
for (SizeT i = 0; i < m_requestedXfbVaryings.size(); ++i) {
const String& name = m_requestedXfbVaryings[i];
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);
if (interleaved) {
varying.bufferIndex = 0;
varying.offsetBytes = interleavedOffset;
interleavedOffset += varying.byteSize;
} else {
varying.bufferIndex = static_cast<Uint32>(i);
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;
if (interleaved) {
if (interleavedOffset > kMaxInterleavedComponents * 4) {
m_infoLog = "Transform feedback interleaved capture exceeds "
"GL_MAX_TRANSFORM_FEEDBACK_INTERLEAVED_COMPONENTS.";
return false;
}
m_xfbStrides.assign(1, interleavedOffset);
} 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(),
@@ -548,12 +327,6 @@ namespace MobileGL::MG_State::GLState {
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) {
@@ -465,39 +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
};
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 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; }
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
@@ -508,11 +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();
@@ -581,30 +527,11 @@ namespace MobileGL::MG_State::GLState {
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_xfbBufferMode = GL_INTERLEAVED_ATTRIBS;
Int m_xfbVaryingNameMaxLength = 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.
@@ -162,7 +162,6 @@ namespace MobileGL {
DepthComponent32F,
Depth24Stencil8,
Depth32FStencil8,
StencilIndex8,
DepthComponent,
DepthStencil,
@@ -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;
@@ -104,23 +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);
@@ -154,9 +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) {
+6 -151
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 {
@@ -197,13 +196,13 @@ TEST(DirectGLESSanity, AdvertisesDepthTextureForGlmarkShadowScenes) {
EXPECT_NE(std::find(extensions.begin(), extensions.end(), MobileGL::E_GL_ARB_depth_texture), extensions.end());
}
TEST(DirectGLESSanity, AdvertisesVoxyRequiredRenderingExtensionsAtExperimentalCTSVersion) {
TEST(DirectGLESSanity, AdvertisesVoxyRequiredRenderingExtensionsWithoutRaisingGLVersion) {
MobileGL::MG_Backend::DirectGLES::BackendObject_DirectGLES backend;
const auto& rendererInfo = backend.GetRendererInfo().RendererGLInfo;
const auto& extensions = rendererInfo.Extensions;
EXPECT_EQ(rendererInfo.TargetGLVersion.Major, 4);
EXPECT_EQ(rendererInfo.TargetGLVersion.Minor, 6);
EXPECT_EQ(rendererInfo.TargetGLVersion.Major, 3);
EXPECT_EQ(rendererInfo.TargetGLVersion.Minor, 3);
EXPECT_EQ(rendererInfo.TargetGLVersion.Patch, 0);
EXPECT_NE(std::find(extensions.begin(), extensions.end(), MobileGL::E_GL_ARB_compute_shader),
@@ -397,13 +396,13 @@ TEST(DirectVulkanSanity, RenderPassExtentUsesSwapchainSizeOnlyForDefaultFramebuf
MobileGL::IntVec2(512, 512));
}
TEST(DirectVulkanSanity, AdvertisesVoxyRequiredRenderingExtensionsAtExperimentalCTSVersion) {
TEST(DirectVulkanSanity, AdvertisesVoxyRequiredRenderingExtensionsWithoutRaisingGLVersion) {
MobileGL::MG_Backend::DirectVulkan::BackendObject_DirectVulkan backend;
const auto& rendererInfo = backend.GetRendererInfo().RendererGLInfo;
const auto& extensions = rendererInfo.Extensions;
EXPECT_EQ(rendererInfo.TargetGLVersion.Major, 4);
EXPECT_EQ(rendererInfo.TargetGLVersion.Minor, 6);
EXPECT_EQ(rendererInfo.TargetGLVersion.Major, 3);
EXPECT_EQ(rendererInfo.TargetGLVersion.Minor, 3);
EXPECT_EQ(rendererInfo.TargetGLVersion.Patch, 0);
EXPECT_NE(std::find(extensions.begin(), extensions.end(), MobileGL::E_GL_ARB_compute_shader),
@@ -516,50 +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;
@@ -682,48 +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;
@@ -1823,61 +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());
}
@@ -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
@@ -839,14 +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.
@@ -907,9 +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;
glesFuncs.glGetFloatv(GL_ALIASED_LINE_WIDTH_RANGE, aliasedLineWidthRange);
glesFuncs.glGetFloatv(GL_SMOOTH_LINE_WIDTH_RANGE, smoothLineWidthRange);
glesFuncs.glGetFloatv(GL_SMOOTH_LINE_WIDTH_GRANULARITY, &smoothLineWidthGranularity);
@@ -929,15 +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_TEXTURE_IMAGE_UNITS, &maxTextureImageUnits);
glesFuncs.glGetIntegerv(GL_MAX_VERTEX_TEXTURE_IMAGE_UNITS, &maxVertexTextureImageUnits);
glesFuncs.glGetIntegerv(GL_MAX_COMPUTE_TEXTURE_IMAGE_UNITS, &maxComputeTextureImageUnits);
@@ -969,29 +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) {
@@ -1049,19 +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,
@@ -1055,9 +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".
@@ -1123,9 +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) {
@@ -191,7 +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);
@@ -282,7 +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.
@@ -68,9 +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:
@@ -251,8 +251,6 @@ namespace MobileGL {
return TextureInternalFormat::Depth24Stencil8;
case GL_DEPTH32F_STENCIL8:
return TextureInternalFormat::Depth32FStencil8;
case GL_STENCIL_INDEX8:
return TextureInternalFormat::StencilIndex8;
case GL_DEPTH_COMPONENT:
return TextureInternalFormat::DepthComponent;
case GL_DEPTH_STENCIL:
@@ -233,8 +233,6 @@ namespace MobileGL {
return GL_DEPTH24_STENCIL8;
case TextureInternalFormat::Depth32FStencil8:
return GL_DEPTH32F_STENCIL8;
case TextureInternalFormat::StencilIndex8:
return GL_STENCIL_INDEX8;
case TextureInternalFormat::DepthComponent32:
return GL_DEPTH_COMPONENT32;
case TextureInternalFormat::DepthStencil:
@@ -234,8 +234,6 @@ namespace MobileGL {
return "Depth24Stencil8";
case TextureInternalFormat::Depth32FStencil8:
return "Depth32FStencil8";
case TextureInternalFormat::StencilIndex8:
return "StencilIndex8";
case TextureInternalFormat::Red:
return "Red";
case TextureInternalFormat::RG:
@@ -25,19 +25,6 @@ namespace MobileGL {
case GL_TRIANGLE_FAN:
return VK_PRIMITIVE_TOPOLOGY_TRIANGLE_FAN;
case GL_LINE_LOOP:
// DrawArrays/DrawElements rewrite line loops into closed indexed
// strips; entry points without that rewrite (instanced/indirect)
// degrade to an open strip, which only misses the closing segment.
MGLOG_W("GL_LINE_LOOP without index rewrite; drawing as LINE_STRIP");
return VK_PRIMITIVE_TOPOLOGY_LINE_STRIP;
case GL_LINES_ADJACENCY:
return VK_PRIMITIVE_TOPOLOGY_LINE_LIST_WITH_ADJACENCY;
case GL_LINE_STRIP_ADJACENCY:
return VK_PRIMITIVE_TOPOLOGY_LINE_STRIP_WITH_ADJACENCY;
case GL_TRIANGLES_ADJACENCY:
return VK_PRIMITIVE_TOPOLOGY_TRIANGLE_LIST_WITH_ADJACENCY;
case GL_TRIANGLE_STRIP_ADJACENCY:
return VK_PRIMITIVE_TOPOLOGY_TRIANGLE_STRIP_WITH_ADJACENCY;
default:
MGLOG_W("Unrecognized primitive topology");
return VK_PRIMITIVE_TOPOLOGY_TRIANGLE_LIST;
@@ -133,11 +133,9 @@ namespace MobileGL {
case TextureInternalFormat::RGBA8Snorm:
return VK_FORMAT_R8G8B8A8_SNORM;
case TextureInternalFormat::RGB10A2:
// GL_UNSIGNED_INT_2_10_10_10_REV puts R in bits 0-9, which is Vulkan's
// A2B10G10R10 layout - A2R10G10B10 silently swaps R and B on upload.
return VK_FORMAT_A2B10G10R10_UNORM_PACK32;
return VK_FORMAT_A2R10G10B10_UNORM_PACK32;
case TextureInternalFormat::RGB10A2UI:
return VK_FORMAT_A2B10G10R10_UINT_PACK32;
return VK_FORMAT_A2R10G10B10_UINT_PACK32;
case TextureInternalFormat::RGBA16:
return VK_FORMAT_R16G16B16A16_UNORM;
case TextureInternalFormat::RGBA16Snorm:
@@ -226,8 +224,6 @@ namespace MobileGL {
return VK_FORMAT_D24_UNORM_S8_UINT;
case TextureInternalFormat::Depth32FStencil8:
return VK_FORMAT_D32_SFLOAT_S8_UINT;
case TextureInternalFormat::StencilIndex8:
return VK_FORMAT_S8_UINT;
case TextureInternalFormat::DepthComponent32:
return VK_FORMAT_D32_SFLOAT;
case TextureInternalFormat::DepthStencil:
+7 -3
View File
@@ -9,10 +9,14 @@
#pragma once
#include <Includes.h>
// Severity-ordered: a build compiled at level X keeps X and everything MORE
// severe. INFO builds must keep WARN/ERROR/FATAL — the old ordering
// (WARN=1/ERROR=2 below INFO=3) compiled every warning and error out of
// release builds and hid real backend failures.
#define MOBILEGL_LOG_LEVEL_DEBUG 0
#define MOBILEGL_LOG_LEVEL_WARN 1
#define MOBILEGL_LOG_LEVEL_ERROR 2
#define MOBILEGL_LOG_LEVEL_INFO 3
#define MOBILEGL_LOG_LEVEL_INFO 1
#define MOBILEGL_LOG_LEVEL_WARN 2
#define MOBILEGL_LOG_LEVEL_ERROR 3
#define MOBILEGL_LOG_LEVEL_FATAL 4
#define MOBILEGL_LOG_INTERNAL(levelTag, androidLogLevel, fmt, ...) \
+5 -20
View File
@@ -18,7 +18,6 @@ namespace MobileGL {
case TextureInternalFormat::Red: // UNorm8 shadow layout
case TextureInternalFormat::R8Snorm:
case TextureInternalFormat::R8I:
case TextureInternalFormat::StencilIndex8:
case TextureInternalFormat::R8UI:
return 1;
@@ -44,11 +43,8 @@ namespace MobileGL {
case TextureInternalFormat::SRGB8:
case TextureInternalFormat::RGB8I:
case TextureInternalFormat::RGB8UI:
return 3;
// Canonical depth shadow is a full 32-bit unorm word (see PixelStoreProcessor),
// converted at upload to the image's own 24/32-bit layout.
case TextureInternalFormat::DepthComponent24:
return 4;
return 3;
case TextureInternalFormat::RGBA2:
case TextureInternalFormat::RGBA4:
@@ -95,9 +91,6 @@ namespace MobileGL {
case TextureInternalFormat::RG32F:
case TextureInternalFormat::RG32I:
case TextureInternalFormat::RG32UI:
// Shadow bytes hold the GL_FLOAT_32_UNSIGNED_INT_24_8_REV wire format
// (float depth + a word whose low 8 bits are stencil), 8 bytes/texel.
case TextureInternalFormat::Depth32FStencil8:
return 8;
case TextureInternalFormat::RGB32F:
@@ -108,6 +101,7 @@ namespace MobileGL {
case TextureInternalFormat::RGBA32F:
case TextureInternalFormat::RGBA32I:
case TextureInternalFormat::RGBA32UI:
case TextureInternalFormat::Depth32FStencil8:
return 16;
case TextureInternalFormat::R11FG11FB10F:
@@ -259,10 +253,8 @@ namespace MobileGL {
case TexturePixelDataType::UnsignedInt101111Rev:
case TexturePixelDataType::UnsignedInt5999Rev:
case TexturePixelDataType::UnsignedInt248:
return 4;
case TexturePixelDataType::Float32UnsignedInt248Rev:
// A 32-bit float depth word followed by a 32-bit word holding stencil.
return 8;
return 4;
default:
return 0;
}
@@ -509,16 +501,9 @@ namespace MobileGL {
s.Depth = 32;
s.Stencil = 8;
break;
case TextureInternalFormat::StencilIndex8:
s.Stencil = 8;
break;
case TextureInternalFormat::Unknown:
// Queried for attachments that have no storage yet (e.g. framebuffer
// parameter queries on the initial state); every size stays 0.
break;
default:
MGLOG_W("Unimplemented internal format in GetComponentSizesForInternalFormat: %d",
static_cast<Int>(internal));
MOBILEGL_ASSERT(false, "Unimplemented internal format in GetComponentSizesForInternalFormat: %d",
static_cast<Int>(internal));
break;
}
@@ -22,6 +22,9 @@
#include "SpirvPasses/StripUboMemberRelaxedPrecisionPass.h"
#include "SpirvPasses/StripNoPerspectivePass.h"
#include "SpirvPasses/EmulateNoPerspectivePass.h"
#include "SpirvPasses/FoldConstOffsetFor1DFetchPass.h"
#include "SpirvPasses/LowerClipDistanceForEsslPass.h"
#include "SpirvPasses/DefeatConstStructArrayLutPass.h"
#include "spirv-tools/libspirv.h"
#include "spirv-tools/optimizer.hpp"
@@ -220,14 +223,11 @@ namespace MobileGL {
auto result = ParseShaderSource(lang, shaderType, source, attrib.flags);
if (result) return result;
// Legacy desktop sources are normalized to "#version 330 core" (with a marker on the
// directive), which parses under stricter rules than the 460 they used to be forced
// to: a shader declaring 110-150 while using e.g. layout(binding=...) without the
// matching #extension line compiles on real drivers but is rejected here. Retry once
// at 460 before reporting failure; a genuinely broken shader fails both attempts and
// keeps its original diagnostics. Application-declared 330+ sources carry no marker
// and keep their declared version's strict rules (the GL CTS negative-compile cases
// depend on that).
// Legacy desktop sources are normalized to "#version 330 core", which parses under
// stricter rules than the 460 they used to be forced to: a shader declaring 330 while
// using e.g. layout(binding=...) without the matching #extension line compiles on real
// drivers but is rejected here. Retry once at 460 before reporting failure; a genuinely
// broken shader fails both attempts and keeps its original diagnostics.
String retrySource = source;
if (!MG_Util::ShaderTranspiler::RetargetLegacyVersionDirectiveTo460(retrySource)) {
return result;
@@ -326,6 +326,55 @@ namespace MobileGL {
return optimizer.Run(inputBinary.data(), inputBinary.size(), &outputBinary, options);
}
bool ShaderCompiler::ClampAccessChainIndicesForEssl(const Vector<Uint32>& inputBinary,
Vector<uint32_t>& outputBinary) {
using namespace spvtools;
OptimizerOptions options;
options.set_run_validator(false);
Optimizer optimizer(SPV_ENV_VULKAN_1_1);
optimizer.RegisterPass(CreateGraphicsRobustAccessPass());
return optimizer.Run(inputBinary.data(), inputBinary.size(), &outputBinary, options);
}
bool ShaderCompiler::FoldConstOffsetFor1DFetchForEssl(const Vector<Uint32>& inputBinary,
Vector<uint32_t>& outputBinary) {
using namespace spvtools;
OptimizerOptions options;
options.set_run_validator(false);
Optimizer optimizer(SPV_ENV_VULKAN_1_1);
optimizer.RegisterPass(FoldConstOffsetFor1DFetchPass::CreateFoldConstOffsetFor1DFetchPass());
return optimizer.Run(inputBinary.data(), inputBinary.size(), &outputBinary, options);
}
bool ShaderCompiler::LowerClipDistanceForEssl(const Vector<Uint32>& inputBinary,
Vector<uint32_t>& outputBinary) {
using namespace spvtools;
OptimizerOptions options;
options.set_run_validator(false);
Optimizer optimizer(SPV_ENV_VULKAN_1_1);
optimizer.RegisterPass(LowerClipDistanceForEsslPass::CreateLowerClipDistanceForEsslPass());
return optimizer.Run(inputBinary.data(), inputBinary.size(), &outputBinary, options);
}
bool ShaderCompiler::DefeatConstStructArrayLutForEssl(const Vector<Uint32>& inputBinary,
Vector<uint32_t>& outputBinary) {
using namespace spvtools;
OptimizerOptions options;
options.set_run_validator(false);
Optimizer optimizer(SPV_ENV_VULKAN_1_1);
optimizer.RegisterPass(
DefeatConstStructArrayLutPass::CreateDefeatConstStructArrayLutPass());
return optimizer.Run(inputBinary.data(), inputBinary.size(), &outputBinary, options);
}
bool ShaderCompiler::StripUboMemberRelaxedPrecisionForEssl(const Vector<Uint32>& inputBinary,
Vector<uint32_t>& outputBinary) {
using namespace spvtools;
@@ -27,6 +27,36 @@ namespace MobileGL {
// Only for backends without native draw-parameter support (DirectGLES).
static bool LowerDrawParametersForEssl(const Vector<Uint32>& inputBinary,
Vector<uint32_t>& outputBinary);
// Clamps every access-chain index to its declared bounds (spirv-tools
// GraphicsRobustAccessPass). GL 3.3 only promises undefined *values* for
// out-of-bounds 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; clamping restores the "some value from the array" contract. Only for
// the DirectGLES transpile path.
static bool ClampAccessChainIndicesForEssl(const Vector<Uint32>& inputBinary,
Vector<uint32_t>& outputBinary);
// Folds the ConstOffset image operand of Dim1D OpImageFetch into the integer
// coordinate (texelFetchOffset(t,P,l,o) == texelFetch(t,P+o,l)). SPIRV-Cross
// emulates 1D samplers as 2D for ES: it widens the coordinate to ivec2 but keeps
// the scalar offset, and ESSL has no texelFetchOffset(sampler2D, ivec2, int,
// scalar) overload, so Adreno rejects the shader. Only for the DirectGLES
// transpile path.
static bool FoldConstOffsetFor1DFetchForEssl(const Vector<Uint32>& inputBinary,
Vector<uint32_t>& outputBinary);
// Shadows gl_ClipDistance in Private mg_ClipDistance/mg_ClipDistanceIn arrays so
// the decompiled ESSL only writes the builtin with literal constant indices
// (flush before EmitVertex/return) and only reads gl_in clip distances with
// dynamic loop indices (copy loop): the other shapes miscompile or crash
// Adreno's ESSL compiler. Vertex/geometry stages; DirectGLES transpile path on
// Qualcomm only (quirk-gated). See LowerClipDistanceForEsslPass.
static bool LowerClipDistanceForEssl(const Vector<Uint32>& inputBinary,
Vector<uint32_t>& outputBinary);
// Splits a Function-storage array-of-structs variable's single constant-composite
// store into per-element stores so SPIRV-Cross does not hoist it into a global
// const struct[] LUT, which Adreno cannot dynamically index. DirectGLES
// transpile path on Qualcomm only (quirk-gated). See DefeatConstStructArrayLutPass.
static bool DefeatConstStructArrayLutForEssl(const Vector<Uint32>& inputBinary,
Vector<uint32_t>& outputBinary);
// Drops RelaxedPrecision member decorations from uniform-block structs so
// SPIRV-Cross prints the same (highp) member precision in every stage; ES
// drivers reject cross-stage uniform blocks whose member precisions differ.
@@ -688,10 +688,6 @@ namespace {
return info;
}
// Stamped onto the normalized directive when a legacy (or absent) desktop
// version was rewritten to 330; consumed by RetargetLegacyVersionDirectiveTo460.
constexpr const char* kNormalizedLegacyMarker = "/*mobilegl-normalized-legacy*/";
MobileGL::String GetNormalizedVersionDirective(const ShaderLanguageInfo& info) {
if (info.profile == MobileGL::ShaderProfile::ES) {
// Preserve the pre-existing behavior for standard lowercase "es" directives. MobileGL's Vulkan
@@ -706,23 +702,9 @@ namespace {
return "#version 460 compatibility\n";
}
// An explicitly declared modern core version keeps its number: the GL CTS
// negative-compile cases (reserved names, layout-qualifier forms, missing
// overloads) rely on the declared version's rules, and raising it would
// silently legalize them. gpu_shader5 opt-ins keep the 460 escalation -
// Vulkan glslang's ARB_gpu_shader5 support is not complete enough alone.
if (info.hasValidVersionDirective && info.version >= 330 && !info.enablesGpuShader5) {
return "#version " + std::to_string(info.version) + " core\n";
}
const bool useLegacyDesktopVersion =
info.version < 400 && !info.enablesGpuShader5;
// The trailing marker records that this 330 came from a legacy declaration
// (or none at all), so the compile-failure retry may re-raise it to 460.
// An application's own "#version 330" never carries it and keeps strict
// 3.30 semantics.
return useLegacyDesktopVersion ? MobileGL::String("#version 330 core ") + kNormalizedLegacyMarker + "\n"
: "#version 460 core\n";
return useLegacyDesktopVersion ? "#version 330 core\n" : "#version 460 core\n";
}
void NormalizeVersionDirective(MobileGL::String& source, const ShaderLanguageInfo& info) {
@@ -1324,123 +1306,12 @@ namespace MobileGL {
// Only the set NormalizeVersionDirective downgraded: desktop core below 400. ES and
// compatibility shaders keep whatever they declared.
if (info.profile != ShaderProfile::Core || info.version >= 400) return false;
// Only rescue MobileGL's own legacy normalization (marked on the directive line).
// An application-declared "#version 330" keeps strict 3.30 semantics: raising it
// would re-legalize the CTS negative-compile cases (reserved names, arrays of
// arrays, missing overloads).
SizeT lineEnd = source.find('\n', info.versionDirectiveStart);
if (lineEnd == MobileGL::String::npos) {
lineEnd = source.size();
}
const SizeT markerPos = source.find(kNormalizedLegacyMarker, info.versionDirectiveStart);
if (markerPos == MobileGL::String::npos || markerPos > lineEnd) {
return false;
}
source.replace(info.versionDirectiveStart, info.versionDirectiveEnd - info.versionDirectiveStart,
"#version 460 core\n");
return true;
}
std::optional<String> FindReservedIdentifierViolation(const String& source) {
// Reserved anywhere; glslang accepts them as plain identifiers.
static constexpr const char* kAlwaysReserved[] = {
"image1DShadow",
"image2DShadow",
"image1DArrayShadow",
"image2DArrayShadow",
};
// Keywords legal only inside a layout(...) qualifier list.
static constexpr const char* kLayoutOnlyKeywords[] = {
"packed",
"row_major",
};
const auto isIdentChar = [](char c) {
return (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') || (c >= '0' && c <= '9') || c == '_';
};
const SizeT length = source.size();
SizeT i = 0;
Int layoutParenDepth = 0; // >0 while inside layout(...)
Bool pendingLayoutParen = false; // saw "layout", awaiting its '('
while (i < length) {
const char c = source[i];
// Comments.
if (c == '/' && i + 1 < length && source[i + 1] == '/') {
while (i < length && source[i] != '\n') ++i;
continue;
}
if (c == '/' && i + 1 < length && source[i + 1] == '*') {
i += 2;
while (i + 1 < length && !(source[i] == '*' && source[i + 1] == '/')) ++i;
i = (i + 1 < length) ? i + 2 : length;
continue;
}
// Preprocessor lines stay out of scope (macro names may shadow anything).
if (c == '#' && (i == 0 || source[i - 1] == '\n' ||
source.find_last_not_of(" \t", i - 1) == MobileGL::String::npos ||
source[source.find_last_not_of(" \t", i - 1)] == '\n')) {
while (i < length && source[i] != '\n') {
if (source[i] == '\\' && i + 1 < length && source[i + 1] == '\n') ++i;
++i;
}
continue;
}
if (c == '(') {
if (pendingLayoutParen) {
layoutParenDepth = 1;
pendingLayoutParen = false;
} else if (layoutParenDepth > 0) {
++layoutParenDepth;
}
++i;
continue;
}
if (c == ')') {
if (layoutParenDepth > 0) --layoutParenDepth;
++i;
continue;
}
if (c == ' ' || c == '\t' || c == '\r' || c == '\n') {
++i;
continue;
}
if (isIdentChar(c) && !(c >= '0' && c <= '9')) {
const SizeT start = i;
while (i < length && isIdentChar(source[i])) ++i;
const StringView word(source.data() + start, i - start);
if (word == "layout") {
pendingLayoutParen = true;
continue;
}
pendingLayoutParen = false;
for (const char* reserved : kAlwaysReserved) {
if (word == reserved) {
return String("ERROR: reserved identifier '") + reserved + "' may not be used.";
}
}
if (layoutParenDepth == 0) {
for (const char* keyword : kLayoutOnlyKeywords) {
if (word == keyword) {
return String("ERROR: '") + keyword +
"' is a keyword and may not be used as an identifier.";
}
}
}
continue;
}
if (isIdentChar(c)) { // digit-led token: skip the whole number/identifier tail
while (i < length && isIdentChar(source[i])) ++i;
pendingLayoutParen = false;
continue;
}
pendingLayoutParen = false;
++i;
}
return std::nullopt;
}
} // namespace ShaderTranspiler
} // namespace MG_Util
} // namespace MobileGL
@@ -40,11 +40,6 @@ namespace MobileGL {
// uses 420-era syntax without the matching #extension line, which real drivers tend to
// accept - can be retried instead of failing to compile.
Bool RetargetLegacyVersionDirectiveTo460(String& source);
// GLSL reserves a few names glslang happily accepts as identifiers ("packed",
// "row_major" outside a layout(...) list, the image*Shadow family). Returns the
// compile-error text for the first violation, or nullopt for a clean source.
std::optional<String> FindReservedIdentifierViolation(const String& source);
} // namespace ShaderTranspiler
} // namespace MG_Util
} // namespace MobileGL
@@ -0,0 +1,175 @@
// MobileGL - MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/DefeatConstStructArrayLutPass.cpp
// Copyright (c) 2025-2026 MobileGL-Dev
// Licensed under the GNU Lesser General Public License v3.0:
// https://www.gnu.org/licenses/gpl-3.0.txt
// https://www.gnu.org/licenses/lgpl-3.0.txt
// SPDX-License-Identifier: LGPL-3.0-only
// End of Source File Header
#include "DefeatConstStructArrayLutPass.h"
#include "spirv.hpp"
#include "source/opt/constants.h"
#include "source/opt/def_use_manager.h"
#include "source/opt/instruction.h"
#include "source/opt/ir_context.h"
#include "source/opt/module.h"
#include "source/opt/type_manager.h"
#include "source/opt/types.h"
#include "source/util/make_unique.h"
#include <vector>
namespace MobileGL {
namespace MG_Util {
namespace ShaderTranspiler {
namespace {
using spvtools::opt::BasicBlock;
using spvtools::opt::Function;
using spvtools::opt::Instruction;
using spvtools::opt::IRContext;
using spvtools::opt::Operand;
namespace analysis = spvtools::opt::analysis;
uint32_t PointerTypeTo(IRContext* ctx, uint32_t pointeeId, spv::StorageClass sc) {
analysis::Type* pointee = ctx->get_type_mgr()->GetType(pointeeId);
analysis::Pointer ptr(pointee, sc);
return ctx->get_type_mgr()->GetTypeInstruction(&ptr);
}
uint32_t SignedIntConstant(IRContext* ctx, uint32_t value) {
analysis::Integer i(32, true);
analysis::Type* reg = ctx->get_type_mgr()->GetRegisteredType(&i);
const analysis::Constant* c = ctx->get_constant_mgr()->GetConstant(reg, {value});
return ctx->get_constant_mgr()->GetDefiningInstruction(c)->result_id();
}
// True when |var| (a Function-storage OpVariable) points to an array of structs.
// Reports the struct type id on success.
bool IsArrayOfStructsVariable(IRContext* ctx, Instruction* var, uint32_t& structTypeId) {
auto* defUse = ctx->get_def_use_mgr();
Instruction* ptrType = defUse->GetDef(var->type_id());
if (ptrType == nullptr || ptrType->opcode() != spv::Op::OpTypePointer) return false;
Instruction* pointee = defUse->GetDef(ptrType->GetSingleWordInOperand(1));
if (pointee == nullptr || pointee->opcode() != spv::Op::OpTypeArray) return false;
Instruction* element = defUse->GetDef(pointee->GetSingleWordInOperand(0));
if (element == nullptr || element->opcode() != spv::Op::OpTypeStruct) return false;
structTypeId = element->result_id();
return true;
}
} // namespace
spvtools::opt::Pass::Status DefeatConstStructArrayLutPass::Process() {
auto* ctx = context();
auto* defUse = ctx->get_def_use_mgr();
bool modified = false;
for (Function& function : *get_module()) {
if (function.begin() == function.end()) continue;
BasicBlock* entryBlock = &*function.begin();
// Candidate variables: Function-storage arrays of structs declared in this
// function's entry block (where OpVariables must live).
struct Candidate {
Instruction* var;
uint32_t structTypeId;
};
std::vector<Candidate> candidates;
for (Instruction& inst : *entryBlock) {
if (inst.opcode() != spv::Op::OpVariable) break;
// Variables with initializers keep SPIRV-Cross's initializer path; the
// glslang pattern under attack is initializer-free with one OpStore.
if (inst.NumInOperands() > 1) continue;
uint32_t structTypeId = 0;
if (IsArrayOfStructsVariable(ctx, &inst, structTypeId)) {
candidates.push_back({&inst, structTypeId});
}
}
for (const Candidate& candidate : candidates) {
Instruction* var = candidate.var;
// The variable qualifies only when its single write is one direct
// OpStore of an OpConstantComposite; any other write shape already
// defeats SPIRV-Cross's LUT promotion, so it is left untouched.
Instruction* singleStore = nullptr;
bool disqualified = false;
defUse->ForEachUser(var, [&](Instruction* user) {
if (user->opcode() == spv::Op::OpStore &&
user->GetSingleWordInOperand(0) == var->result_id()) {
if (singleStore != nullptr) {
disqualified = true;
} else {
singleStore = user;
}
} else if (user->opcode() == spv::Op::OpCopyMemory) {
disqualified = true;
} else if (user->opcode() == spv::Op::OpAccessChain ||
user->opcode() == spv::Op::OpInBoundsAccessChain) {
defUse->ForEachUser(user, [&](Instruction* chainUser) {
if (chainUser->opcode() == spv::Op::OpStore ||
chainUser->opcode() == spv::Op::OpCopyMemory) {
disqualified = true;
}
});
}
});
if (disqualified || singleStore == nullptr) continue;
Instruction* composite = defUse->GetDef(singleStore->GetSingleWordInOperand(1));
if (composite == nullptr ||
composite->opcode() != spv::Op::OpConstantComposite) {
continue;
}
// The store must sit in the entry block: that is the only placement
// SPIRV-Cross treats as a LUT initializer.
bool storeInEntryBlock = false;
for (Instruction& inst : *entryBlock) {
if (&inst == singleStore) {
storeInEntryBlock = true;
break;
}
}
if (!storeInEntryBlock) continue;
// Split the composite store into one constant-index store per element.
const uint32_t ptrFnStruct =
PointerTypeTo(ctx, candidate.structTypeId, spv::StorageClass::Function);
for (uint32_t element = 0; element < composite->NumInOperands(); ++element) {
const uint32_t elementConstId = composite->GetSingleWordInOperand(element);
const uint32_t chainId = ctx->TakeNextId();
Instruction* chain =
singleStore->InsertBefore(spvtools::MakeUnique<Instruction>(
ctx, spv::Op::OpAccessChain, ptrFnStruct, chainId,
std::initializer_list<Operand>{
{SPV_OPERAND_TYPE_ID, {var->result_id()}},
{SPV_OPERAND_TYPE_ID, {SignedIntConstant(ctx, element)}}}));
ctx->AnalyzeDefUse(chain);
Instruction* store =
singleStore->InsertBefore(spvtools::MakeUnique<Instruction>(
ctx, spv::Op::OpStore, 0, 0,
std::initializer_list<Operand>{
{SPV_OPERAND_TYPE_ID, {chainId}},
{SPV_OPERAND_TYPE_ID, {elementConstId}}}));
ctx->AnalyzeDefUse(store);
}
ctx->KillInst(singleStore);
modified = true;
}
}
if (!modified) {
return Status::SuccessWithoutChange;
}
ctx->InvalidateAnalysesExceptFor(spvtools::opt::IRContext::kAnalysisNone);
return Status::SuccessWithChange;
}
spvtools::Optimizer::PassToken
DefeatConstStructArrayLutPass::CreateDefeatConstStructArrayLutPass() {
return spvtools::Optimizer::PassToken(MakeUnique<DefeatConstStructArrayLutPass>());
}
} // namespace ShaderTranspiler
} // namespace MG_Util
} // namespace MobileGL
@@ -0,0 +1,36 @@
// MobileGL - MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/DefeatConstStructArrayLutPass.h
// Copyright (c) 2025-2026 MobileGL-Dev
// Licensed under the GNU Lesser General Public License v3.0:
// https://www.gnu.org/licenses/gpl-3.0.txt
// https://www.gnu.org/licenses/lgpl-3.0.txt
// SPDX-License-Identifier: LGPL-3.0-only
// End of Source File Header
#pragma once
#include "source/opt/pass.h"
#include "spirv-tools/optimizer.hpp"
#include <Includes.h>
namespace MobileGL {
namespace MG_Util {
namespace ShaderTranspiler {
// SPIRV-Cross hoists a Function-storage array variable whose only write is a single
// constant-composite store into a global `const struct[]` LUT (variable_is_lut).
// Adreno's ESSL compiler cannot dynamically index such a global const struct array
// ("Cannot offset into the structure" - device-verified on Adreno 750). Splitting
// the one composite store into per-element constant-index stores makes
// variable_is_lut fail, so SPIRV-Cross keeps the array as an ordinary local that
// Adreno indexes fine. Scalar/vector const arrays are unaffected on Adreno and are
// left alone - only arrays OF STRUCTS are rewritten. Only meant for the DirectGLES
// transpile path on Qualcomm devices.
class DefeatConstStructArrayLutPass : public spvtools::opt::Pass {
public:
const char* name() const override { return "defeat-const-struct-array-lut"; }
Status Process() override;
static spvtools::Optimizer::PassToken CreateDefeatConstStructArrayLutPass();
};
} // namespace ShaderTranspiler
} // namespace MG_Util
} // namespace MobileGL
@@ -0,0 +1,131 @@
// MobileGL - MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/FoldConstOffsetFor1DFetchPass.cpp
// Copyright (c) 2025-2026 MobileGL-Dev
// Licensed under the GNU Lesser General Public License v3.0:
// https://www.gnu.org/licenses/gpl-3.0.txt
// https://www.gnu.org/licenses/lgpl-3.0.txt
// SPDX-License-Identifier: LGPL-3.0-only
// End of Source File Header
#include "FoldConstOffsetFor1DFetchPass.h"
#include "spirv.hpp"
#include "source/opt/def_use_manager.h"
#include "source/opt/instruction.h"
#include "source/opt/ir_builder.h"
#include "source/opt/ir_context.h"
#include "source/opt/module.h"
#include "source/util/make_unique.h"
#include <vector>
namespace MobileGL {
namespace MG_Util {
namespace ShaderTranspiler {
namespace {
using spvtools::opt::Instruction;
using spvtools::opt::InstructionBuilder;
using spvtools::opt::IRContext;
using spvtools::opt::Operand;
// Number of ImageOperands ids that precede the ConstOffset id: one per
// lower-order bit set in the mask, except Grad which carries two ids.
uint32_t CountIdsBeforeConstOffset(uint32_t mask) {
uint32_t count = 0;
if (mask & static_cast<uint32_t>(spv::ImageOperandsMask::Bias)) count += 1;
if (mask & static_cast<uint32_t>(spv::ImageOperandsMask::Lod)) count += 1;
if (mask & static_cast<uint32_t>(spv::ImageOperandsMask::Grad)) count += 2;
return count;
}
} // namespace
spvtools::opt::Pass::Status FoldConstOffsetFor1DFetchPass::Process() {
auto* irContext = context();
auto* defUseMgr = irContext->get_def_use_mgr();
Bool modified = false;
constexpr uint32_t kConstOffsetBit =
static_cast<uint32_t>(spv::ImageOperandsMask::ConstOffset);
for (auto& function : *get_module()) {
for (auto& block : function) {
for (auto& inst : block) {
if (inst.opcode() != spv::Op::OpImageFetch) continue;
// In-operands: image, coordinate, [ImageOperands mask, ids...].
if (inst.NumInOperands() < 3) continue;
const uint32_t operandsMask = inst.GetSingleWordInOperand(2);
if ((operandsMask & kConstOffsetBit) == 0) continue;
Instruction* imageInst = defUseMgr->GetDef(inst.GetSingleWordInOperand(0));
if (imageInst == nullptr) continue;
Instruction* imageType = defUseMgr->GetDef(imageInst->type_id());
if (imageType == nullptr || imageType->opcode() != spv::Op::OpTypeImage ||
static_cast<spv::Dim>(imageType->GetSingleWordInOperand(1)) != spv::Dim::Dim1D) {
continue;
}
const uint32_t offsetOperandIndex = 3 + CountIdsBeforeConstOffset(operandsMask);
const uint32_t offsetId = inst.GetSingleWordInOperand(offsetOperandIndex);
const uint32_t coordId = inst.GetSingleWordInOperand(1);
Instruction* coordType = defUseMgr->GetDef(defUseMgr->GetDef(coordId)->type_id());
InstructionBuilder builder(
irContext, &inst,
IRContext::kAnalysisDefUse | IRContext::kAnalysisInstrToBlockMapping);
uint32_t newCoordId = 0;
if (coordType->opcode() == spv::Op::OpTypeVector) {
// Arrayed 1D fetch: component 0 is the texel coordinate,
// component 1 the layer - only component 0 takes the offset.
const uint32_t componentTypeId = coordType->GetSingleWordInOperand(0);
Instruction* extracted = builder.AddCompositeExtract(componentTypeId, coordId, {0});
Instruction* sum =
builder.AddIAdd(componentTypeId, extracted->result_id(), offsetId);
Instruction* inserted = builder.AddInstruction(spvtools::MakeUnique<Instruction>(
irContext, spv::Op::OpCompositeInsert, coordType->result_id(),
irContext->TakeNextId(),
std::initializer_list<Operand>{
{SPV_OPERAND_TYPE_ID, {sum->result_id()}},
{SPV_OPERAND_TYPE_ID, {coordId}},
{SPV_OPERAND_TYPE_LITERAL_INTEGER, {0}}}));
newCoordId = inserted->result_id();
} else {
Instruction* sum = builder.AddIAdd(coordType->result_id(), coordId, offsetId);
newCoordId = sum->result_id();
}
const uint32_t newMask = operandsMask & ~kConstOffsetBit;
// 3 fixed operands + the offset id: anything beyond that is another
// image-operand id that must keep the mask word alive.
const Bool otherOperandIdsRemain = inst.NumInOperands() > 4;
irContext->ForgetUses(&inst);
std::vector<Operand> newOperands;
newOperands.push_back(inst.GetInOperand(0));
newOperands.push_back({SPV_OPERAND_TYPE_ID, {newCoordId}});
if (newMask != 0 || otherOperandIdsRemain) {
Operand maskOperand = inst.GetInOperand(2);
maskOperand.words[0] = newMask;
newOperands.push_back(maskOperand);
for (uint32_t i = 3; i < inst.NumInOperands(); ++i) {
if (i == offsetOperandIndex) continue;
newOperands.push_back(inst.GetInOperand(i));
}
}
inst.SetInOperands(std::move(newOperands));
irContext->AnalyzeUses(&inst);
modified = true;
}
}
}
return modified ? Status::SuccessWithChange : Status::SuccessWithoutChange;
}
spvtools::Optimizer::PassToken FoldConstOffsetFor1DFetchPass::CreateFoldConstOffsetFor1DFetchPass() {
return spvtools::Optimizer::PassToken(MakeUnique<FoldConstOffsetFor1DFetchPass>());
}
} // namespace ShaderTranspiler
} // namespace MG_Util
} // namespace MobileGL
@@ -0,0 +1,36 @@
// MobileGL - MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/FoldConstOffsetFor1DFetchPass.h
// Copyright (c) 2025-2026 MobileGL-Dev
// Licensed under the GNU Lesser General Public License v3.0:
// https://www.gnu.org/licenses/gpl-3.0.txt
// https://www.gnu.org/licenses/lgpl-3.0.txt
// SPDX-License-Identifier: LGPL-3.0-only
// End of Source File Header
#pragma once
#include "source/opt/pass.h"
#include "spirv-tools/optimizer.hpp"
#include <Includes.h>
namespace MobileGL {
namespace MG_Util {
namespace ShaderTranspiler {
// SPIRV-Cross emulates 1D textures as 2D for ES targets: it widens the texelFetch
// coordinate to ivec2 but keeps the ConstOffset image operand scalar, and ESSL has
// no texelFetchOffset(sampler2D, ivec2, int, scalar-offset) overload, so drivers
// (Adreno) reject the transpiled shader. This pass folds the constant offset into
// the integer coordinate before the fetch - texelFetchOffset(t, P, l, o) ==
// texelFetch(t, P + o, l) per the GLSL spec - and drops the ConstOffset operand,
// so SPIRV-Cross emits a plain texelFetch. For arrayed 1D fetches only coordinate
// component 0 is offset (component 1 is the layer). Only meant for the DirectGLES
// transpile path.
class FoldConstOffsetFor1DFetchPass : public spvtools::opt::Pass {
public:
const char* name() const override { return "fold-const-offset-for-1d-fetch"; }
Status Process() override;
static spvtools::Optimizer::PassToken CreateFoldConstOffsetFor1DFetchPass();
};
} // namespace ShaderTranspiler
} // namespace MG_Util
} // namespace MobileGL
@@ -0,0 +1,613 @@
// MobileGL - MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/LowerClipDistanceForEsslPass.cpp
// Copyright (c) 2025-2026 MobileGL-Dev
// Licensed under the GNU Lesser General Public License v3.0:
// https://www.gnu.org/licenses/gpl-3.0.txt
// https://www.gnu.org/licenses/lgpl-3.0.txt
// SPDX-License-Identifier: LGPL-3.0-only
// End of Source File Header
#include "LowerClipDistanceForEsslPass.h"
#include "spirv.hpp"
#include "source/opt/basic_block.h"
#include "source/opt/constants.h"
#include "source/opt/def_use_manager.h"
#include "source/opt/instruction.h"
#include "source/opt/ir_context.h"
#include "source/opt/module.h"
#include "source/opt/type_manager.h"
#include "source/opt/types.h"
#include "source/util/make_unique.h"
#include <memory>
#include <vector>
namespace MobileGL {
namespace MG_Util {
namespace ShaderTranspiler {
namespace {
using spvtools::opt::BasicBlock;
using spvtools::opt::Function;
using spvtools::opt::Instruction;
using spvtools::opt::IRContext;
using spvtools::opt::Operand;
namespace analysis = spvtools::opt::analysis;
spv::ExecutionModel EntryExecutionModel(IRContext* ctx) {
for (Instruction& ep : ctx->module()->entry_points()) {
return static_cast<spv::ExecutionModel>(ep.GetSingleWordInOperand(0));
}
return spv::ExecutionModel::Max;
}
uint32_t EntryFunctionId(IRContext* ctx) {
for (Instruction& ep : ctx->module()->entry_points()) {
// OpEntryPoint <model> <function> "name" <interface...>
return ep.GetSingleWordInOperand(1);
}
return 0;
}
uint32_t VariablePointeeType(IRContext* ctx, Instruction* var) {
Instruction* ptrType = ctx->get_def_use_mgr()->GetDef(var->type_id());
// OpTypePointer <storage-class> <pointee>
return ptrType->GetSingleWordInOperand(1);
}
uint32_t PointerTypeTo(IRContext* ctx, uint32_t pointeeId, spv::StorageClass sc) {
analysis::Type* pointee = ctx->get_type_mgr()->GetType(pointeeId);
analysis::Pointer ptr(pointee, sc);
return ctx->get_type_mgr()->GetTypeInstruction(&ptr);
}
uint32_t IntConstant(IRContext* ctx, bool isSigned, uint32_t value) {
analysis::Integer i(32, isSigned);
analysis::Type* reg = ctx->get_type_mgr()->GetRegisteredType(&i);
const analysis::Constant* c = ctx->get_constant_mgr()->GetConstant(reg, {value});
return ctx->get_constant_mgr()->GetDefiningInstruction(c)->result_id();
}
uint32_t UintType(IRContext* ctx) {
analysis::Integer i(32, false);
return ctx->get_type_mgr()->GetTypeInstruction(&i);
}
uint32_t BoolType(IRContext* ctx) {
analysis::Bool b;
return ctx->get_type_mgr()->GetTypeInstruction(&b);
}
// Constant length of OpTypeArray |arrayTypeId| (0 when not a sized constant).
uint32_t ArrayLength(IRContext* ctx, uint32_t arrayTypeId) {
Instruction* arrayType = ctx->get_def_use_mgr()->GetDef(arrayTypeId);
if (arrayType == nullptr || arrayType->opcode() != spv::Op::OpTypeArray) {
return 0;
}
Instruction* length = ctx->get_def_use_mgr()->GetDef(arrayType->GetSingleWordInOperand(1));
if (length == nullptr || length->opcode() != spv::Op::OpConstant) {
return 0;
}
return length->GetSingleWordInOperand(0);
}
bool IsConstantWithValue(IRContext* ctx, uint32_t id, uint32_t value) {
Instruction* def = ctx->get_def_use_mgr()->GetDef(id);
return def != nullptr && def->opcode() == spv::Op::OpConstant &&
def->GetSingleWordInOperand(0) == value;
}
bool IsAccessChain(const Instruction* inst) {
return inst->opcode() == spv::Op::OpAccessChain ||
inst->opcode() == spv::Op::OpInBoundsAccessChain;
}
Instruction* AddPrivateVariable(IRContext* ctx, uint32_t pointeeTypeId, const char* name) {
const uint32_t ptrType = PointerTypeTo(ctx, pointeeTypeId, spv::StorageClass::Private);
const uint32_t varId = ctx->TakeNextId();
ctx->AddGlobalValue(spvtools::MakeUnique<Instruction>(
ctx, spv::Op::OpVariable, ptrType, varId,
std::initializer_list<Operand>{
{SPV_OPERAND_TYPE_STORAGE_CLASS,
{static_cast<uint32_t>(spv::StorageClass::Private)}}}));
ctx->AddDebug2Inst(spvtools::MakeUnique<Instruction>(
ctx, spv::Op::OpName, 0, 0,
std::initializer_list<Operand>{
{SPV_OPERAND_TYPE_ID, {varId}},
{SPV_OPERAND_TYPE_LITERAL_STRING, spvtools::utils::MakeVector(name)}}));
return ctx->get_def_use_mgr()->GetDef(varId);
}
// Retargets |chain| onto |newBaseId|, dropping the first |dropIndexCount| index
// operands and switching the result pointer's storage class to Private.
void RetargetChainToPrivate(IRContext* ctx, Instruction* chain, uint32_t newBaseId,
uint32_t dropIndexCount) {
Instruction* chainPtrType = ctx->get_def_use_mgr()->GetDef(chain->type_id());
const uint32_t pointeeId = chainPtrType->GetSingleWordInOperand(1);
const uint32_t newPtrType = PointerTypeTo(ctx, pointeeId, spv::StorageClass::Private);
ctx->ForgetUses(chain);
std::vector<Operand> newOperands;
newOperands.push_back({SPV_OPERAND_TYPE_ID, {newBaseId}});
for (uint32_t i = 1 + dropIndexCount; i < chain->NumInOperands(); ++i) {
newOperands.push_back(chain->GetInOperand(i));
}
chain->SetResultType(newPtrType);
chain->SetInOperands(std::move(newOperands));
ctx->AnalyzeUses(chain);
}
// ---- Output side --------------------------------------------------------------
struct OutputTarget {
Instruction* var = nullptr; // Output gl_PerVertex block or standalone builtin
bool isBlockMember = false;
uint32_t memberIndex = 0; // valid when isBlockMember
uint32_t arrayTypeId = 0; // float[N]
uint32_t elemTypeId = 0; // float
uint32_t arrayLen = 0; // N
};
// Inserts "gl_ClipDistance[k] = mg_ClipDistance[k]" for every literal k before
// |before|. Constant-index writes are the only write shape Adreno links correctly.
void InsertFlushBefore(IRContext* ctx, Instruction* before, const OutputTarget& target,
uint32_t mgVarId) {
const uint32_t ptrPrivElem =
PointerTypeTo(ctx, target.elemTypeId, spv::StorageClass::Private);
const uint32_t ptrOutElem =
PointerTypeTo(ctx, target.elemTypeId, spv::StorageClass::Output);
for (uint32_t k = 0; k < target.arrayLen; ++k) {
const uint32_t kConst = IntConstant(ctx, true, k);
const uint32_t srcChainId = ctx->TakeNextId();
before->InsertBefore(spvtools::MakeUnique<Instruction>(
ctx, spv::Op::OpAccessChain, ptrPrivElem, srcChainId,
std::initializer_list<Operand>{{SPV_OPERAND_TYPE_ID, {mgVarId}},
{SPV_OPERAND_TYPE_ID, {kConst}}}));
const uint32_t valId = ctx->TakeNextId();
before->InsertBefore(spvtools::MakeUnique<Instruction>(
ctx, spv::Op::OpLoad, target.elemTypeId, valId,
std::initializer_list<Operand>{{SPV_OPERAND_TYPE_ID, {srcChainId}}}));
const uint32_t dstChainId = ctx->TakeNextId();
std::vector<Operand> dstOperands;
dstOperands.push_back({SPV_OPERAND_TYPE_ID, {target.var->result_id()}});
if (target.isBlockMember) {
dstOperands.push_back(
{SPV_OPERAND_TYPE_ID, {IntConstant(ctx, true, target.memberIndex)}});
}
dstOperands.push_back({SPV_OPERAND_TYPE_ID, {kConst}});
before->InsertBefore(spvtools::MakeUnique<Instruction>(
ctx, spv::Op::OpAccessChain, ptrOutElem, dstChainId, dstOperands));
before->InsertBefore(spvtools::MakeUnique<Instruction>(
ctx, spv::Op::OpStore, 0, 0,
std::initializer_list<Operand>{{SPV_OPERAND_TYPE_ID, {dstChainId}},
{SPV_OPERAND_TYPE_ID, {valId}}}));
}
}
bool LowerOutputClipDistance(IRContext* ctx, bool isGeometry) {
auto* defUse = ctx->get_def_use_mgr();
// Collect (struct type, member) pairs decorated BuiltIn ClipDistance and
// standalone variables decorated BuiltIn ClipDistance.
std::vector<std::pair<uint32_t, uint32_t>> memberTargets; // (structId, member)
std::vector<uint32_t> plainTargets; // variable ids
for (Instruction& ann : ctx->annotations()) {
if (ann.opcode() == spv::Op::OpMemberDecorate && ann.NumInOperands() >= 4 &&
static_cast<spv::Decoration>(ann.GetSingleWordInOperand(2)) ==
spv::Decoration::BuiltIn &&
static_cast<spv::BuiltIn>(ann.GetSingleWordInOperand(3)) ==
spv::BuiltIn::ClipDistance) {
memberTargets.emplace_back(ann.GetSingleWordInOperand(0),
ann.GetSingleWordInOperand(1));
} else if (ann.opcode() == spv::Op::OpDecorate && ann.NumInOperands() >= 3 &&
static_cast<spv::Decoration>(ann.GetSingleWordInOperand(1)) ==
spv::Decoration::BuiltIn &&
static_cast<spv::BuiltIn>(ann.GetSingleWordInOperand(2)) ==
spv::BuiltIn::ClipDistance) {
plainTargets.push_back(ann.GetSingleWordInOperand(0));
}
}
std::vector<OutputTarget> targets;
for (Instruction& inst : ctx->module()->types_values()) {
if (inst.opcode() != spv::Op::OpVariable ||
static_cast<spv::StorageClass>(inst.GetSingleWordInOperand(0)) !=
spv::StorageClass::Output) {
continue;
}
const uint32_t pointee = VariablePointeeType(ctx, &inst);
for (const auto& [structId, member] : memberTargets) {
if (pointee != structId) continue;
Instruction* structType = defUse->GetDef(structId);
if (structType == nullptr || member >= structType->NumInOperands()) continue;
OutputTarget target;
target.var = &inst;
target.isBlockMember = true;
target.memberIndex = member;
target.arrayTypeId = structType->GetSingleWordInOperand(member);
target.arrayLen = ArrayLength(ctx, target.arrayTypeId);
targets.push_back(target);
}
for (const uint32_t varId : plainTargets) {
if (inst.result_id() != varId) continue;
OutputTarget target;
target.var = &inst;
target.isBlockMember = false;
target.arrayTypeId = pointee;
target.arrayLen = ArrayLength(ctx, target.arrayTypeId);
targets.push_back(target);
}
}
bool changed = false;
for (OutputTarget& target : targets) {
if (target.arrayLen == 0) continue;
Instruction* arrayType = defUse->GetDef(target.arrayTypeId);
target.elemTypeId = arrayType->GetSingleWordInOperand(0);
// Collect the accesses to redirect. For the block form only chains whose
// leading index selects the ClipDistance member count; for the standalone
// form every chain plus whole-variable loads/stores.
std::vector<Instruction*> chains;
std::vector<Instruction*> directAccesses;
bool unsupportedUse = false;
defUse->ForEachUser(target.var, [&](Instruction* user) {
if (IsAccessChain(user) &&
user->GetSingleWordInOperand(0) == target.var->result_id()) {
if (target.isBlockMember) {
if (user->NumInOperands() >= 2 &&
IsConstantWithValue(ctx, user->GetSingleWordInOperand(1),
target.memberIndex)) {
chains.push_back(user);
}
} else {
chains.push_back(user);
}
} else if (!target.isBlockMember) {
if (user->opcode() == spv::Op::OpLoad ||
(user->opcode() == spv::Op::OpStore &&
user->GetSingleWordInOperand(0) == target.var->result_id())) {
directAccesses.push_back(user);
} else if (user->opcode() == spv::Op::OpCopyMemory) {
unsupportedUse = true;
}
}
});
if (unsupportedUse || (chains.empty() && directAccesses.empty())) {
continue;
}
Instruction* mgVar = AddPrivateVariable(ctx, target.arrayTypeId, "mg_ClipDistance");
const uint32_t mgVarId = mgVar->result_id();
for (Instruction* chain : chains) {
const uint32_t dropCount = target.isBlockMember ? 1u : 0u;
if (chain->NumInOperands() == 1 + dropCount) {
// Pointer to the whole float[N]: reuse the private variable itself.
ctx->ReplaceAllUsesWith(chain->result_id(), mgVarId);
ctx->KillInst(chain);
} else {
RetargetChainToPrivate(ctx, chain, mgVarId, dropCount);
}
}
for (Instruction* access : directAccesses) {
ctx->ForgetUses(access);
access->SetInOperand(0, {mgVarId});
ctx->AnalyzeUses(access);
}
// Flush the shadow into the real builtin: geometry right before every
// EmitVertex, vertex before every return of the entry point. The flush is
// also what keeps the builtin statically used for cross-stage IO matching.
std::vector<Instruction*> flushSites;
if (isGeometry) {
for (Function& function : *ctx->module()) {
function.ForEachInst([&](Instruction* inst) {
if (inst->opcode() == spv::Op::OpEmitVertex) {
flushSites.push_back(inst);
}
});
}
} else {
const uint32_t entryFuncId = EntryFunctionId(ctx);
for (Function& function : *ctx->module()) {
if (function.result_id() != entryFuncId) continue;
function.ForEachInst([&](Instruction* inst) {
if (inst->opcode() == spv::Op::OpReturn ||
inst->opcode() == spv::Op::OpReturnValue) {
flushSites.push_back(inst);
}
});
}
}
for (Instruction* site : flushSites) {
InsertFlushBefore(ctx, site, target, mgVarId);
}
changed = true;
}
return changed;
}
// ---- Input side (geometry gl_in) ----------------------------------------------
bool LowerInputClipDistance(IRContext* ctx) {
auto* defUse = ctx->get_def_use_mgr();
auto* typeMgr = ctx->get_type_mgr();
// Locate the gl_in block member decorated ClipDistance.
Instruction* glInVar = nullptr;
uint32_t memberIndex = 0;
uint32_t arrayTypeId = 0; // float[N]
for (Instruction& ann : ctx->annotations()) {
if (ann.opcode() != spv::Op::OpMemberDecorate || ann.NumInOperands() < 4 ||
static_cast<spv::Decoration>(ann.GetSingleWordInOperand(2)) !=
spv::Decoration::BuiltIn ||
static_cast<spv::BuiltIn>(ann.GetSingleWordInOperand(3)) !=
spv::BuiltIn::ClipDistance) {
continue;
}
const uint32_t structId = ann.GetSingleWordInOperand(0);
const uint32_t member = ann.GetSingleWordInOperand(1);
for (Instruction& inst : ctx->module()->types_values()) {
if (inst.opcode() != spv::Op::OpVariable ||
static_cast<spv::StorageClass>(inst.GetSingleWordInOperand(0)) !=
spv::StorageClass::Input) {
continue;
}
const uint32_t pointee = VariablePointeeType(ctx, &inst);
Instruction* pointeeType = defUse->GetDef(pointee);
if (pointeeType == nullptr || pointeeType->opcode() != spv::Op::OpTypeArray ||
pointeeType->GetSingleWordInOperand(0) != structId) {
continue;
}
Instruction* structType = defUse->GetDef(structId);
if (structType == nullptr || member >= structType->NumInOperands()) continue;
glInVar = &inst;
memberIndex = member;
arrayTypeId = structType->GetSingleWordInOperand(member);
break;
}
if (glInVar != nullptr) break;
}
if (glInVar == nullptr) {
return false;
}
const uint32_t clipCount = ArrayLength(ctx, arrayTypeId);
const uint32_t vertexCount = ArrayLength(ctx, VariablePointeeType(ctx, glInVar));
if (clipCount == 0 || vertexCount == 0) {
return false;
}
// Every gl_in chain that selects the ClipDistance member:
// (vertex, member) yields a whole float[N], (vertex, member, k) an element.
std::vector<Instruction*> chains;
defUse->ForEachUser(glInVar, [&](Instruction* user) {
if (IsAccessChain(user) && user->GetSingleWordInOperand(0) == glInVar->result_id() &&
user->NumInOperands() >= 3 &&
IsConstantWithValue(ctx, user->GetSingleWordInOperand(2), memberIndex)) {
chains.push_back(user);
}
});
if (chains.empty()) {
return false;
}
Instruction* arrayTypeInst = defUse->GetDef(arrayTypeId);
const uint32_t elemTypeId = arrayTypeInst->GetSingleWordInOperand(0);
// Private mg_ClipDistanceIn = float[vertexCount][clipCount].
const uint32_t vertexCountConst = IntConstant(ctx, false, vertexCount);
analysis::Type* innerType = typeMgr->GetType(arrayTypeId);
analysis::Array outerArray(
innerType, analysis::Array::LengthInfo{
vertexCountConst,
{analysis::Array::LengthInfo::kConstant, vertexCount}});
const uint32_t outerArrayTypeId = typeMgr->GetTypeInstruction(&outerArray);
Instruction* mgInVar = AddPrivateVariable(ctx, outerArrayTypeId, "mg_ClipDistanceIn");
const uint32_t mgInVarId = mgInVar->result_id();
// Copy loop at the top of the entry point:
// for (uint t = 0; t < vertexCount * clipCount; ++t)
// mg_ClipDistanceIn[t / clipCount][t % clipCount] =
// gl_in[t / clipCount].gl_ClipDistance[t % clipCount];
// Both gl_in indices are loop-derived (dynamic): constant-index element reads
// miscompile and whole-array reads crash the Adreno compiler.
const uint32_t entryFuncId = EntryFunctionId(ctx);
Function* entryFn = nullptr;
for (Function& function : *ctx->module()) {
if (function.result_id() == entryFuncId) {
entryFn = &function;
break;
}
}
if (entryFn == nullptr || entryFn->begin() == entryFn->end()) {
return false;
}
const uint32_t uintTypeId = UintType(ctx);
const uint32_t boolTypeId = BoolType(ctx);
const uint32_t ptrFnUint = PointerTypeTo(ctx, uintTypeId, spv::StorageClass::Function);
const uint32_t ptrInElem = PointerTypeTo(ctx, elemTypeId, spv::StorageClass::Input);
const uint32_t ptrPrivElem = PointerTypeTo(ctx, elemTypeId, spv::StorageClass::Private);
const uint32_t uint0 = IntConstant(ctx, false, 0);
const uint32_t uint1 = IntConstant(ctx, false, 1);
const uint32_t uintN = IntConstant(ctx, false, clipCount);
const uint32_t uintTotal = IntConstant(ctx, false, vertexCount * clipCount);
const uint32_t memberConst = IntConstant(ctx, true, memberIndex);
BasicBlock* entryBlock = &*entryFn->begin();
auto splitPoint = entryBlock->begin();
while (splitPoint != entryBlock->end() &&
splitPoint->opcode() == spv::Op::OpVariable) {
++splitPoint;
}
// Loop counter lives with the other function-local variables.
const uint32_t counterVarId = ctx->TakeNextId();
splitPoint->InsertBefore(spvtools::MakeUnique<Instruction>(
ctx, spv::Op::OpVariable, ptrFnUint, counterVarId,
std::initializer_list<Operand>{
{SPV_OPERAND_TYPE_STORAGE_CLASS,
{static_cast<uint32_t>(spv::StorageClass::Function)}}}));
const uint32_t restLabelId = ctx->TakeNextId();
BasicBlock* restBlock = entryBlock->SplitBasicBlock(ctx, restLabelId, splitPoint);
const uint32_t headerLabelId = ctx->TakeNextId();
const uint32_t checkLabelId = ctx->TakeNextId();
const uint32_t bodyLabelId = ctx->TakeNextId();
const uint32_t continueLabelId = ctx->TakeNextId();
auto makeBlock = [&](uint32_t labelId) {
return spvtools::MakeUnique<BasicBlock>(spvtools::MakeUnique<Instruction>(
ctx, spv::Op::OpLabel, 0, labelId, std::initializer_list<Operand>{}));
};
auto addInst = [&](BasicBlock* block, spv::Op opcode, uint32_t typeId,
uint32_t resultId, std::vector<Operand> operands) {
block->AddInstruction(spvtools::MakeUnique<Instruction>(
ctx, opcode, typeId, resultId, std::move(operands)));
};
// entry: t = 0; branch header
addInst(entryBlock, spv::Op::OpStore, 0, 0,
{{SPV_OPERAND_TYPE_ID, {counterVarId}}, {SPV_OPERAND_TYPE_ID, {uint0}}});
addInst(entryBlock, spv::Op::OpBranch, 0, 0, {{SPV_OPERAND_TYPE_ID, {headerLabelId}}});
// header: structured loop header
auto headerBlock = makeBlock(headerLabelId);
addInst(headerBlock.get(), spv::Op::OpLoopMerge, 0, 0,
{{SPV_OPERAND_TYPE_ID, {restLabelId}},
{SPV_OPERAND_TYPE_ID, {continueLabelId}},
{SPV_OPERAND_TYPE_LOOP_CONTROL,
{static_cast<uint32_t>(spv::LoopControlMask::MaskNone)}}});
addInst(headerBlock.get(), spv::Op::OpBranch, 0, 0,
{{SPV_OPERAND_TYPE_ID, {checkLabelId}}});
// check: t < vertexCount * clipCount ?
auto checkBlock = makeBlock(checkLabelId);
const uint32_t tCheckId = ctx->TakeNextId();
addInst(checkBlock.get(), spv::Op::OpLoad, uintTypeId, tCheckId,
{{SPV_OPERAND_TYPE_ID, {counterVarId}}});
const uint32_t condId = ctx->TakeNextId();
addInst(checkBlock.get(), spv::Op::OpULessThan, boolTypeId, condId,
{{SPV_OPERAND_TYPE_ID, {tCheckId}}, {SPV_OPERAND_TYPE_ID, {uintTotal}}});
addInst(checkBlock.get(), spv::Op::OpBranchConditional, 0, 0,
{{SPV_OPERAND_TYPE_ID, {condId}},
{SPV_OPERAND_TYPE_ID, {bodyLabelId}},
{SPV_OPERAND_TYPE_ID, {restLabelId}}});
// body: mg_ClipDistanceIn[t / N][t % N] = gl_in[t / N].gl_ClipDistance[t % N]
auto bodyBlock = makeBlock(bodyLabelId);
const uint32_t tBodyId = ctx->TakeNextId();
addInst(bodyBlock.get(), spv::Op::OpLoad, uintTypeId, tBodyId,
{{SPV_OPERAND_TYPE_ID, {counterVarId}}});
const uint32_t vertexIdxId = ctx->TakeNextId();
addInst(bodyBlock.get(), spv::Op::OpUDiv, uintTypeId, vertexIdxId,
{{SPV_OPERAND_TYPE_ID, {tBodyId}}, {SPV_OPERAND_TYPE_ID, {uintN}}});
const uint32_t clipIdxId = ctx->TakeNextId();
addInst(bodyBlock.get(), spv::Op::OpUMod, uintTypeId, clipIdxId,
{{SPV_OPERAND_TYPE_ID, {tBodyId}}, {SPV_OPERAND_TYPE_ID, {uintN}}});
const uint32_t srcChainId = ctx->TakeNextId();
addInst(bodyBlock.get(), spv::Op::OpAccessChain, ptrInElem, srcChainId,
{{SPV_OPERAND_TYPE_ID, {glInVar->result_id()}},
{SPV_OPERAND_TYPE_ID, {vertexIdxId}},
{SPV_OPERAND_TYPE_ID, {memberConst}},
{SPV_OPERAND_TYPE_ID, {clipIdxId}}});
const uint32_t valId = ctx->TakeNextId();
addInst(bodyBlock.get(), spv::Op::OpLoad, elemTypeId, valId,
{{SPV_OPERAND_TYPE_ID, {srcChainId}}});
const uint32_t dstChainId = ctx->TakeNextId();
addInst(bodyBlock.get(), spv::Op::OpAccessChain, ptrPrivElem, dstChainId,
{{SPV_OPERAND_TYPE_ID, {mgInVarId}},
{SPV_OPERAND_TYPE_ID, {vertexIdxId}},
{SPV_OPERAND_TYPE_ID, {clipIdxId}}});
addInst(bodyBlock.get(), spv::Op::OpStore, 0, 0,
{{SPV_OPERAND_TYPE_ID, {dstChainId}}, {SPV_OPERAND_TYPE_ID, {valId}}});
addInst(bodyBlock.get(), spv::Op::OpBranch, 0, 0,
{{SPV_OPERAND_TYPE_ID, {continueLabelId}}});
// continue: ++t
auto continueBlock = makeBlock(continueLabelId);
const uint32_t tContinueId = ctx->TakeNextId();
addInst(continueBlock.get(), spv::Op::OpLoad, uintTypeId, tContinueId,
{{SPV_OPERAND_TYPE_ID, {counterVarId}}});
const uint32_t tIncId = ctx->TakeNextId();
addInst(continueBlock.get(), spv::Op::OpIAdd, uintTypeId, tIncId,
{{SPV_OPERAND_TYPE_ID, {tContinueId}}, {SPV_OPERAND_TYPE_ID, {uint1}}});
addInst(continueBlock.get(), spv::Op::OpStore, 0, 0,
{{SPV_OPERAND_TYPE_ID, {counterVarId}}, {SPV_OPERAND_TYPE_ID, {tIncId}}});
addInst(continueBlock.get(), spv::Op::OpBranch, 0, 0,
{{SPV_OPERAND_TYPE_ID, {headerLabelId}}});
BasicBlock* headerPtr = entryFn->InsertBasicBlockBefore(std::move(headerBlock), restBlock);
BasicBlock* checkPtr = entryFn->InsertBasicBlockAfter(std::move(checkBlock), headerPtr);
BasicBlock* bodyPtr = entryFn->InsertBasicBlockAfter(std::move(bodyBlock), checkPtr);
entryFn->InsertBasicBlockAfter(std::move(continueBlock), bodyPtr);
// Redirect the pre-existing accesses to the shadow copy.
for (Instruction* chain : chains) {
if (chain->NumInOperands() == 3) {
// (vertex, member): whole float[N] of one vertex.
Instruction* chainPtrType = defUse->GetDef(chain->type_id());
const uint32_t pointeeId = chainPtrType->GetSingleWordInOperand(1);
const uint32_t newPtrType =
PointerTypeTo(ctx, pointeeId, spv::StorageClass::Private);
ctx->ForgetUses(chain);
std::vector<Operand> newOperands;
newOperands.push_back({SPV_OPERAND_TYPE_ID, {mgInVarId}});
newOperands.push_back(chain->GetInOperand(1));
chain->SetResultType(newPtrType);
chain->SetInOperands(std::move(newOperands));
ctx->AnalyzeUses(chain);
} else {
// (vertex, member, k, ...): drop the member index.
Instruction* chainPtrType = defUse->GetDef(chain->type_id());
const uint32_t pointeeId = chainPtrType->GetSingleWordInOperand(1);
const uint32_t newPtrType =
PointerTypeTo(ctx, pointeeId, spv::StorageClass::Private);
ctx->ForgetUses(chain);
std::vector<Operand> newOperands;
newOperands.push_back({SPV_OPERAND_TYPE_ID, {mgInVarId}});
newOperands.push_back(chain->GetInOperand(1));
for (uint32_t i = 3; i < chain->NumInOperands(); ++i) {
newOperands.push_back(chain->GetInOperand(i));
}
chain->SetResultType(newPtrType);
chain->SetInOperands(std::move(newOperands));
ctx->AnalyzeUses(chain);
}
}
return true;
}
} // namespace
spvtools::opt::Pass::Status LowerClipDistanceForEsslPass::Process() {
auto* ctx = context();
const spv::ExecutionModel model = EntryExecutionModel(ctx);
const bool isVertex = model == spv::ExecutionModel::Vertex;
const bool isGeometry = model == spv::ExecutionModel::Geometry;
if (!isVertex && !isGeometry) {
return Status::SuccessWithoutChange;
}
bool changed = LowerOutputClipDistance(ctx, isGeometry);
if (isGeometry) {
changed |= LowerInputClipDistance(ctx);
}
if (!changed) {
return Status::SuccessWithoutChange;
}
ctx->InvalidateAnalysesExceptFor(spvtools::opt::IRContext::kAnalysisNone);
return Status::SuccessWithChange;
}
spvtools::Optimizer::PassToken
LowerClipDistanceForEsslPass::CreateLowerClipDistanceForEsslPass() {
return spvtools::Optimizer::PassToken(MakeUnique<LowerClipDistanceForEsslPass>());
}
} // namespace ShaderTranspiler
} // namespace MG_Util
} // namespace MobileGL
@@ -0,0 +1,44 @@
// MobileGL - MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/LowerClipDistanceForEsslPass.h
// Copyright (c) 2025-2026 MobileGL-Dev
// Licensed under the GNU Lesser General Public License v3.0:
// https://www.gnu.org/licenses/gpl-3.0.txt
// https://www.gnu.org/licenses/lgpl-3.0.txt
// SPDX-License-Identifier: LGPL-3.0-only
// End of Source File Header
#pragma once
#include "source/opt/pass.h"
#include "spirv-tools/optimizer.hpp"
#include <Includes.h>
namespace MobileGL {
namespace MG_Util {
namespace ShaderTranspiler {
// Adreno's ESSL compiler mishandles gl_ClipDistance (device-verified on Adreno 750):
// - writes through non-constant indices silently fail to link,
// - reads of gl_in[i].gl_ClipDistance[k] with a CONSTANT k >= 1 fail to compile
// ("array indexing out of boundary") while dynamic-index reads work,
// - compiling a whole-array read of gl_in[i].gl_ClipDistance segfaults the
// compiler backend (libllvm-qgl.so).
// This pass shadows the builtin so the decompiled ESSL only ever touches it in the
// shapes Adreno accepts. Output side (vertex + geometry): all accesses to the
// Output ClipDistance (gl_PerVertex member or standalone variable) are redirected
// to a Private mg_ClipDistance array, and a flush writing the real builtin with
// literal constant indices is inserted before every OpEmitVertex (geometry) or
// every return of the entry point (vertex). Input side (geometry): accesses to
// gl_in[...].gl_ClipDistance are redirected to a Private mg_ClipDistanceIn
// array-of-arrays filled once at the top of the entry point by a structured loop
// whose gl_in reads use dynamic (loop-variable) indices. The builtin members stay
// statically referenced by the flush/copy so cross-stage IO matching is intact.
// Only meant for the DirectGLES transpile path on Qualcomm devices.
class LowerClipDistanceForEsslPass : public spvtools::opt::Pass {
public:
const char* name() const override { return "lower-clip-distance-for-essl"; }
Status Process() override;
static spvtools::Optimizer::PassToken CreateLowerClipDistanceForEsslPass();
};
} // namespace ShaderTranspiler
} // namespace MG_Util
} // namespace MobileGL
@@ -95,7 +95,6 @@ namespace MobileGL::MG_Util::PixelStoreProcessor {
Int32,
Half,
Float32,
UNorm32, // 32-bit fixed-point depth shadow
};
struct InternalShadowLayout {
@@ -124,21 +123,6 @@ namespace MobileGL::MG_Util::PixelStoreProcessor {
Bool GetInternalShadowLayout(TextureInternalFormat internal, InternalShadowLayout& out) {
switch (internal) {
// Depth shadows follow TextureFormatProcessor::NormalizePixelFormat: 16-bit
// unorm for DEPTH_COMPONENT16, 32-bit unorm for the 24/32-bit fixed-point
// depths, float for DEPTH_COMPONENT32F.
case TextureInternalFormat::DepthComponent16:
out = {1, ShadowComponent::UNorm16, false};
return true;
case TextureInternalFormat::DepthComponent24:
case TextureInternalFormat::DepthComponent32:
case TextureInternalFormat::DepthComponent:
out = {1, ShadowComponent::UNorm32, false};
return true;
case TextureInternalFormat::DepthComponent32F:
out = {1, ShadowComponent::Float32, false};
return true;
case TextureInternalFormat::R8:
case TextureInternalFormat::Red: out = {1, ShadowComponent::UNorm8, false}; return true;
case TextureInternalFormat::RG8:
@@ -315,10 +299,8 @@ namespace MobileGL::MG_Util::PixelStoreProcessor {
case TextureInputFormat::RGBAInteger: out = {{0, 1, 2, 3}, 4, true}; return true;
case TextureInputFormat::BGRA: out = {{2, 1, 0, 3}, 4, false}; return true;
case TextureInputFormat::BGRAInteger: out = {{2, 1, 0, 3}, 4, true}; return true;
// A depth value converts like a single normalized/float channel.
case TextureInputFormat::DepthComponent: out = {{0, -1, -1, -1}, 1, false}; return true;
default:
return false; // stencil / packed depth-stencil / unknown
return false; // depth / stencil / unknown
}
}
@@ -364,7 +346,8 @@ namespace MobileGL::MG_Util::PixelStoreProcessor {
out = isInteger ? ShadowComponent::Int16 : ShadowComponent::SNorm16;
return true;
case TexturePixelDataType::UnsignedInt:
out = isInteger ? ShadowComponent::UInt32 : ShadowComponent::UNorm32;
if (!isInteger) return false; // no 32-bit normalized shadow layout
out = ShadowComponent::UInt32;
return true;
case TexturePixelDataType::Int:
if (!isInteger) return false;
@@ -634,12 +617,6 @@ namespace MobileGL::MG_Util::PixelStoreProcessor {
case ShadowComponent::Float32:
Memcpy(dst, &v, sizeof(v));
break;
case ShadowComponent::UNorm32: {
const auto out = static_cast<Uint32>(
std::llround(static_cast<double>(std::clamp(v, 0.0f, 1.0f)) * 4294967295.0));
Memcpy(dst, &out, sizeof(out));
break;
}
default:
break; // integer components never reach the float encoder
}
@@ -794,64 +771,6 @@ namespace MobileGL::MG_Util::PixelStoreProcessor {
const Int effectiveHeight = (params.ImageHeight > 0) ? params.ImageHeight : height;
const SizeT inputRowStride = CalculateRowStride(effectiveWidth, pixelSize, params.Alignment);
// GL_DEPTH_COMPONENT client data may populate a packed depth-stencil internal
// format (the stencil half becomes zero); the generic channel converter cannot
// express the packed shadow words, so convert here.
const Bool packedDepthStencilInternal = targetInternalFormat == TextureInternalFormat::Depth24Stencil8 ||
targetInternalFormat == TextureInternalFormat::DepthStencil ||
targetInternalFormat == TextureInternalFormat::Depth32FStencil8;
if (!isBitmap && packedDepthStencilInternal && textureInputFormat == TextureInputFormat::DepthComponent &&
(inputDataType == TexturePixelDataType::Float || inputDataType == TexturePixelDataType::UnsignedInt ||
inputDataType == TexturePixelDataType::UnsignedShort)) {
const Bool floatShadow = targetInternalFormat == TextureInternalFormat::Depth32FStencil8;
const SizeT outPixelSize = floatShadow ? 8 : 4;
outSize = static_cast<SizeT>(width) * height * std::max(depth, 1) * outPixelSize;
Uint8* outputPixels = static_cast<Uint8*>(malloc(outSize));
if (!outputPixels) {
outSize = 0;
return nullptr;
}
const Uint8* srcBase = static_cast<const Uint8*>(inputPixels) +
static_cast<SizeT>(params.SkipImages) * static_cast<SizeT>(effectiveHeight) * inputRowStride +
static_cast<SizeT>(params.SkipRows) * inputRowStride +
static_cast<SizeT>(params.SkipPixels) * pixelSize;
Uint8* dst = outputPixels;
for (Int z = 0; z < std::max(depth, 1); ++z) {
for (Int y = 0; y < height; ++y) {
const Uint8* srcRow = srcBase +
static_cast<SizeT>(z) * static_cast<SizeT>(effectiveHeight) * inputRowStride +
static_cast<SizeT>(y) * inputRowStride;
for (Int x = 0; x < width; ++x) {
Float depthValue = 0.0f;
if (inputDataType == TexturePixelDataType::Float) {
Memcpy(&depthValue, srcRow + static_cast<SizeT>(x) * 4, sizeof(depthValue));
} else if (inputDataType == TexturePixelDataType::UnsignedInt) {
Uint32 raw = 0;
Memcpy(&raw, srcRow + static_cast<SizeT>(x) * 4, sizeof(raw));
depthValue = static_cast<Float>(static_cast<double>(raw) / 4294967295.0);
} else {
Uint16 raw = 0;
Memcpy(&raw, srcRow + static_cast<SizeT>(x) * 2, sizeof(raw));
depthValue = static_cast<Float>(raw) / 65535.0f;
}
if (floatShadow) {
const Uint32 stencilWord = 0;
Memcpy(dst, &depthValue, sizeof(depthValue));
Memcpy(dst + 4, &stencilWord, sizeof(stencilWord));
dst += 8;
} else {
const Uint32 depth24 = static_cast<Uint32>(
std::llround(static_cast<double>(std::clamp(depthValue, 0.0f, 1.0f)) * 16777215.0));
const Uint32 word = depth24 << 8;
Memcpy(dst, &word, sizeof(word));
dst += 4;
}
}
}
}
return outputPixels;
}
UnpackConversionSpec conversion{};
const Bool needConversion =
!isBitmap && GetUnpackConversionSpec(targetInternalFormat, textureInputFormat, inputDataType, conversion);
@@ -1053,11 +972,6 @@ namespace MobileGL::MG_Util::PixelStoreProcessor {
Memcpy(&v, p, sizeof(v));
return v;
}
case ShadowComponent::UNorm32: {
Uint32 v;
Memcpy(&v, p, sizeof(v));
return static_cast<Float>(static_cast<double>(v) / 4294967295.0);
}
default:
return 0.0f;
}

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