Compare commits

..
Author SHA1 Message Date
BZLZHH 4bf9c1cae6 [Perf] (MG_Backend/DirectVulkanTMP): Optimize buffer uploads with staging ring + batched copies. 2026-02-21 14:04:05 +08:00
BZLZHH c121328c98 [Feat] (MG_Backend/DirectVulkanTMP): Migrate to Vulkan 1.3 and use dynamic rendering. 2026-02-21 13:17:05 +08:00
BZLZHH 5fe5aae138 [Perf] (MG_Backend/DirectVulkan): Batch layout transitions and Reuse staging buffers. 2026-02-20 22:04:50 +08:00
BZLZHH d2485a3fbe [Improvement] (MG_Backend/DirectVulkanTMP): Rename current DirectVulkan to DirectVulkanTMP. 2026-02-20 16:33:26 +08:00
BZLZHH 3fe0999550 [Misc] (MG_Backend/DirectVulkan): Remove "VulkanRenerer". 2026-02-20 16:27:02 +08:00
BZLZHH 39eb6a9026 [Feat] (MG_Backend/DirectVulkan|MG_Util): Get Vulkan device info. 2026-02-20 16:21:28 +08:00
BZLZHH 99c1697fba Merge branch 'dev' into Feat/Backend-Direct-Vulkan-TMP 2026-02-20 15:27:21 +08:00
BZLZHH 5db5545b33 Merge branch 'dev' into Feat/Backend-Direct-Vulkan-TMP 2026-02-19 02:04:57 +08:00
BZLZHH e63eebf425 [Fix] (MG_Backend/DirectVulkan): Correct viewport setting. 2026-02-16 18:29:03 +08:00
BZLZHH 03d3fcdc66 [Feat|Fix] (MG_Backend): Implement BlitFramebuffer & DrawElementsBaseVertex. 2026-02-16 17:12:33 +08:00
BZLZHH 40fc317f69 [Fix] (MG_Backend/DirectVulkan): Add shader position transforms & correct namespace name. 2026-02-16 14:06:34 +08:00
BZLZHH 3bafdee45d [Feat] (MG_Backend/DirectVulkan): Implement DV::TmpImpl and use. 2026-02-11 02:01:08 +08:00
BZLZHH 188275c004 Merge branch 'Feat/Backend-Direct-Vulkan' of github.com:MobileGL-Dev/MobileGL into Feat/Backend-Direct-Vulkan 2026-02-09 21:03:10 +08:00
472 changed files with 15934 additions and 100330 deletions
+2
View File
@@ -14,6 +14,7 @@ bugprone-forwarding-reference-overload,
bugprone-inaccurate-erase,
bugprone-incorrect-roundings,
bugprone-integer-division,
bugprone-lambda-function-name,
bugprone-macro-parentheses,
bugprone-macro-repeated-side-effects,
bugprone-misplaced-operator-in-strlen-in-alloc,
@@ -62,6 +63,7 @@ cert-str34-c,
cppcoreguidelines-interfaces-global-init,
cppcoreguidelines-narrowing-conversions,
cppcoreguidelines-pro-type-member-init,
cppcoreguidelines-pro-type-static-cast-downcast,
cppcoreguidelines-slicing,
google-default-arguments,
google-runtime-operator,
-4
View File
@@ -1,4 +0,0 @@
tools/trace_replay/fixtures/*.tgz filter=lfs diff=lfs merge=lfs -text
tools/trace_replay/fixtures/*.png filter=lfs diff=lfs merge=lfs -text
tools/trace_replay/fixtures/openra.tgz -filter -diff -merge -text
tools/trace_replay/fixtures/openra.0000031249.png -filter -diff -merge -text
-239
View File
@@ -1,239 +0,0 @@
#!/usr/bin/env bash
set -euo pipefail
if [ "$#" -lt 1 ] || [ "$#" -gt 2 ]; then
echo "usage: $0 <trace-case> [fixture-dir]" >&2
exit 2
fi
case_name="$1"
fixture_dir="${2:-tools/trace_replay/fixtures}"
python_bin="${PYTHON:-python3}"
# Fixture mirrors, tried in order before falling back to Git LFS. Override the
# whole list with MOBILEGL_TRACE_FIXTURE_MIRROR_BASES (whitespace separated);
# MOBILEGL_TRACE_FIXTURE_MIRROR_BASE still works and is tried first.
default_mirror_bases=(
"https://git.hit.moe/swung0x48/MobileGL/media/branch/dev/tools/trace_replay/fixtures"
"https://repo.miawa.cn/mgl/tools/trace_replay/fixtures"
)
if [ -n "${MOBILEGL_TRACE_FIXTURE_MIRROR_BASES:-}" ]; then
read -r -a mirror_bases <<< "${MOBILEGL_TRACE_FIXTURE_MIRROR_BASES}"
else
mirror_bases=("${default_mirror_bases[@]}")
fi
if [ -n "${MOBILEGL_TRACE_FIXTURE_MIRROR_BASE:-}" ]; then
mirror_bases=("${MOBILEGL_TRACE_FIXTURE_MIRROR_BASE}" "${mirror_bases[@]}")
fi
# Optional bearer token for mirrors that require authentication (private Gitea).
mirror_token="${MOBILEGL_TRACE_FIXTURE_MIRROR_TOKEN:-}"
download_attempts="${MOBILEGL_TRACE_FIXTURE_DOWNLOAD_ATTEMPTS:-5}"
retry_delay="${MOBILEGL_TRACE_FIXTURE_RETRY_DELAY:-2}"
if ! command -v "${python_bin}" >/dev/null 2>&1 && command -v python >/dev/null 2>&1; then
python_bin=python
fi
if ! [[ "${download_attempts}" =~ ^[1-9][0-9]*$ ]]; then
echo "MOBILEGL_TRACE_FIXTURE_DOWNLOAD_ATTEMPTS must be a positive integer: ${download_attempts}" >&2
exit 2
fi
if ! [[ "${retry_delay}" =~ ^[0-9]+$ ]]; then
echo "MOBILEGL_TRACE_FIXTURE_RETRY_DELAY must be a non-negative integer: ${retry_delay}" >&2
exit 2
fi
fixture_list="$("${python_bin}" tools/trace_replay/trace_cases.py \
--format fixture-files \
--case "${case_name}" \
--fixture-root "${fixture_dir}")"
# Strip CR so the script also works when python emits CRLF (Git Bash on Windows).
mapfile -t files < <(printf '%s\n' "${fixture_list}" | tr -d '\r')
include="$(IFS=,; echo "${files[*]}")"
if [ "${case_name}" = "OpenRA" ]; then
echo "Fixture files for ${case_name} are stored in Git: ${include}"
for file in "${files[@]}"; do
test -s "${file}"
if head -n 1 "${file}" | grep -q "version https://git-lfs.github.com/spec/v1"; then
echo "fixture should not be stored as an LFS pointer: ${file}" >&2
exit 1
fi
done
exit 0
fi
get_lfs_metadata() {
local file="$1"
local pointer
local expected_oid
local expected_size
if ! pointer="$(git show "HEAD:${file}" 2>/dev/null)"; then
echo "failed to read tracked fixture metadata: ${file}" >&2
return 1
fi
if ! grep -q '^version https://git-lfs.github.com/spec/v1$' <<< "${pointer}"; then
echo "tracked fixture is not a Git LFS pointer: ${file}" >&2
return 1
fi
expected_oid="$(awk '$1 == "oid" && $2 ~ /^sha256:/ { sub(/^sha256:/, "", $2); print $2 }' <<< "${pointer}")"
expected_size="$(awk '$1 == "size" { print $2 }' <<< "${pointer}")"
if ! [[ "${expected_oid}" =~ ^[0-9a-f]{64}$ ]] || ! [[ "${expected_size}" =~ ^[0-9]+$ ]]; then
echo "invalid Git LFS pointer metadata: ${file}" >&2
return 1
fi
printf '%s %s\n' "${expected_oid}" "${expected_size}"
}
verify_fixture_file() {
local downloaded_file="$1"
local display_name="$2"
local expected_oid="$3"
local expected_size="$4"
local actual_oid
local actual_size
if [ ! -f "${downloaded_file}" ]; then
echo "fixture file is missing: ${display_name}" >&2
return 1
fi
actual_size="$(wc -c < "${downloaded_file}" | tr -d '[:space:]')"
if [ "${actual_size}" != "${expected_size}" ]; then
echo "fixture size mismatch for ${display_name}: expected ${expected_size}, got ${actual_size}" >&2
return 1
fi
actual_oid="$(sha256sum "${downloaded_file}" | awk '{ print $1 }')"
if [ "${actual_oid}" != "${expected_oid}" ]; then
echo "fixture SHA-256 mismatch for ${display_name}: expected ${expected_oid}, got ${actual_oid}" >&2
return 1
fi
}
fetch_file_from_mirror() {
local file="$1"
local url="$2"
local metadata
local expected_oid
local expected_size
local tmp_file="${file}.tmp"
local attempt
local partial_size
local curl_status
local curl_auth
metadata="$(get_lfs_metadata "${file}")" || return 1
read -r expected_oid expected_size <<< "${metadata}"
if [ -f "${tmp_file}" ]; then
partial_size="$(wc -c < "${tmp_file}" | tr -d '[:space:]')"
if [ "${partial_size}" -gt "${expected_size}" ]; then
echo "Discarding oversized partial fixture ${tmp_file}: ${partial_size} > ${expected_size}" >&2
rm -f "${tmp_file}"
elif [ "${partial_size}" = "${expected_size}" ]; then
if verify_fixture_file "${tmp_file}" "${file}" "${expected_oid}" "${expected_size}"; then
mv "${tmp_file}" "${file}"
return 0
fi
rm -f "${tmp_file}"
fi
fi
for ((attempt = 1; attempt <= download_attempts; attempt++)); do
partial_size=0
if [ -f "${tmp_file}" ]; then
partial_size="$(wc -c < "${tmp_file}" | tr -d '[:space:]')"
fi
if [ "${partial_size}" -gt 0 ]; then
echo "Resuming mirror download for ${file} at byte ${partial_size} (attempt ${attempt}/${download_attempts})"
else
echo "Starting mirror download for ${file} (attempt ${attempt}/${download_attempts})"
fi
curl_auth=()
if [ -n "${mirror_token}" ]; then
curl_auth=(--header "Authorization: token ${mirror_token}")
fi
if curl -L --fail --show-error --continue-at - "${curl_auth[@]}" --output "${tmp_file}" "${url}"; then
if verify_fixture_file "${tmp_file}" "${file}" "${expected_oid}" "${expected_size}"; then
mv "${tmp_file}" "${file}"
return 0
fi
echo "Mirror download failed integrity verification; retrying from the beginning: ${file}" >&2
rm -f "${tmp_file}"
else
curl_status=$?
partial_size=0
if [ -f "${tmp_file}" ]; then
partial_size="$(wc -c < "${tmp_file}" | tr -d '[:space:]')"
fi
if [ "${partial_size}" = "${expected_size}" ]; then
if verify_fixture_file "${tmp_file}" "${file}" "${expected_oid}" "${expected_size}"; then
mv "${tmp_file}" "${file}"
return 0
fi
rm -f "${tmp_file}"
partial_size=0
elif [ "${partial_size}" -gt "${expected_size}" ]; then
echo "Discarding oversized partial fixture ${tmp_file}: ${partial_size} > ${expected_size}" >&2
rm -f "${tmp_file}"
partial_size=0
elif [ "${curl_status}" -eq 33 ]; then
echo "Mirror refused the resume request; retrying from the beginning: ${file}" >&2
rm -f "${tmp_file}"
partial_size=0
fi
echo "Mirror download attempt ${attempt}/${download_attempts} failed with curl exit ${curl_status}; retained ${partial_size} bytes for resume: ${file}" >&2
fi
if [ "${attempt}" -lt "${download_attempts}" ]; then
sleep "${retry_delay}"
fi
done
rm -f "${tmp_file}"
return 1
}
fetch_from_mirror() {
mkdir -p "${fixture_dir}"
for file in "${files[@]}"; do
local name
local url
local base
local fetched=0
name="$(basename "${file}")"
for base in "${mirror_bases[@]}"; do
url="${base%/}/${name}"
echo "Fetching trace fixture from mirror: ${url}"
if fetch_file_from_mirror "${file}" "${url}"; then
fetched=1
break
fi
echo "Mirror did not serve ${name}; trying the next mirror" >&2
done
if [ "${fetched}" -ne 1 ]; then
return 1
fi
done
}
if fetch_from_mirror; then
echo "Fetched trace fixture files for ${case_name} from mirror: ${include}"
else
echo "All mirrors failed for ${case_name}; falling back to Git LFS: ${include}"
git lfs install --local
git lfs pull --include="${include}" --exclude=""
fi
for file in "${files[@]}"; do
metadata="$(get_lfs_metadata "${file}")"
read -r expected_oid expected_size <<< "${metadata}"
verify_fixture_file "${file}" "${file}" "${expected_oid}" "${expected_size}"
done
-75
View File
@@ -1,75 +0,0 @@
#!/usr/bin/env bash
set -euo pipefail
if [[ $# -ne 3 ]]; then
echo "Usage: $0 <aapt2> <plugin-apk> <trace-apk>" >&2
exit 64
fi
aapt2=$1
plugin_apk=$2
trace_apk=$3
require() {
local needle=$1
local content=$2
local description=$3
if ! grep -Fq -- "$needle" <<<"$content"; then
echo "::error::Missing ${description}: ${needle}" >&2
exit 1
fi
}
for apk in "$plugin_apk" "$trace_apk"; do
[[ -f "$apk" ]] || { echo "::error::APK not found: $apk" >&2; exit 1; }
done
plugin_manifest=$("$aapt2" dump xmltree --file AndroidManifest.xml "$plugin_apk")
plugin_resources=$("$aapt2" dump resources "$plugin_apk")
plugin_resource_text=$(tr -d '"' <<<"$plugin_resources")
trace_manifest=$("$aapt2" dump xmltree --file AndroidManifest.xml "$trace_apk")
plugin_contents=$(unzip -Z1 "$plugin_apk")
require 'top.mobilegl.plugin' "$plugin_manifest" 'plugin package name'
require 'MobileGL' "$plugin_manifest" 'plugin label'
require 'fclPlugin' "$plugin_manifest" 'legacy plugin marker'
require 'fclPlugin_V2' "$plugin_manifest" 'V2 plugin marker'
require 'LIBGL_ES=3:POJAV_RENDERER=opengles3:MOBILEGL_BACKEND_TYPE=DirectGLES' "$plugin_manifest" 'V1 DirectGLES fallback'
require 'string/config' "$plugin_resources" 'V2 renderer configuration resource'
require '{displayName:MobileGL,rendererId:opengles3' "$plugin_resource_text" 'V2 MobileGL entry and renderer ID'
require 'rendererGLPath:**|libMobileGL.so' "$plugin_resource_text" 'V2 GL library path'
require 'rendererEGLPath:**|libMobileGL.so' "$plugin_resource_text" 'V2 EGL library path'
require 'key:LIBGL_ES,value:3' "$plugin_resource_text" 'V2 fixed LIBGL_ES variable'
require 'key:MOBILEGL_BACKEND_TYPE' "$plugin_resource_text" 'V2 backend variable'
require 'defaultValue:DirectGLES' "$plugin_resource_text" 'V2 DirectGLES default'
require 'DirectVulkan' "$plugin_resource_text" 'V2 DirectVulkan option'
require 'key:MOBILEGL_DISABLE_TIMERQUERY' "$plugin_resource_text" 'V2 timer-query toggle'
require 'key:MOBILEGL_DISABLE_SUBGROUP' "$plugin_resource_text" 'V2 Vulkan subgroup toggle'
require 'key:MOBILEGL_MAGMA_R11G11B10F_FALLBACK' "$plugin_resource_text" 'V2 Magma format fallback toggle'
require 'key:MOBILEGL_MAGMA_FRAMESINFLIGHT' "$plugin_resource_text" 'V2 Magma frames-in-flight setting'
require 'key:MOBILEGL_AVOID_SAMPLER_MIPMAP_MIN_FILTER' "$plugin_resource_text" 'V2 sampler workaround toggle'
require 'key:MOBILEGL_COHERENT_AS_FLUSH' "$plugin_resource_text" 'V2 coherent-as-flush toggle'
require 'key:MOBILEGL_USE_ANGLE' "$plugin_resource_text" 'V2 ANGLE toggle'
if [[ $(grep -Fc 'fclPlugin_V2' <<<"$plugin_manifest") -ne 1 ]]; then
echo '::error::Plugin manifest must expose exactly one V2 descriptor' >&2
exit 1
fi
if ! grep -Eq '^lib/[^/]+/libMobileGL\.so$' <<<"$plugin_contents"; then
echo '::error::Plugin APK does not contain libMobileGL.so' >&2
exit 1
fi
require 'top.mobilegl.plugin.trace' "$trace_manifest" 'trace package name'
require 'top.mobilegl.plugin.TRACE_REPLAY' "$trace_manifest" 'trace replay action'
if grep -Fq 'fclPlugin' <<<"$trace_manifest"; then
echo '::error::Trace APK must not advertise renderer-plugin metadata' >&2
exit 1
fi
if grep -Fq 'android.intent.action.MAIN' <<<"$trace_manifest"; then
echo '::error::Trace APK must not expose a launcher activity' >&2
exit 1
fi
echo 'Validated unified MobileGL plugin APK and isolated trace APK.'
-560
View File
@@ -1,560 +0,0 @@
name: MobileGL APK
on:
push:
branches:
- dev
- Feat/Backend-Direct-GLES
- Feat/Backend-Direct-Vulkan
workflow_dispatch:
jobs:
build:
runs-on: ubuntu-latest
env:
CCACHE_BASEDIR: ${{ github.workspace }}
CCACHE_COMPRESS: "true"
CCACHE_DIR: ${{ github.workspace }}/.ccache
CCACHE_MAXSIZE: 4G
CCACHE_NOHASHDIR: "true"
MOBILEGL_CMAKE_COMPILER_LAUNCHER: ccache
steps:
- name: Checkout repo
uses: actions/checkout@v6
with:
submodules: recursive
- name: Set artifact metadata
run: |
echo "date_today=$(date +'%Y-%m-%d')" >> "$GITHUB_ENV"
- name: Set up JDK
uses: actions/setup-java@v5
with:
distribution: zulu
java-version: '17'
- name: Setup Gradle
uses: gradle/actions/setup-gradle@v6
with:
gradle-version: 8.10.2
- name: Restore ccache
uses: actions/cache@v5
with:
path: .ccache
key: ${{ runner.os }}-apk-${{ github.job }}-ccache-${{ github.ref_name }}-${{ github.run_id }}
restore-keys: |
${{ runner.os }}-apk-${{ github.job }}-ccache-${{ github.ref_name }}-
${{ runner.os }}-apk-${{ github.job }}-ccache-
- name: Install ccache
run: |
sudo apt-get update
sudo apt-get install -y ccache
ccache --version
- name: Setup Android SDK
uses: android-actions/setup-android@v4
with:
accept-android-sdk-licenses: false
- name: Accept Android SDK licenses
run: yes | sdkmanager --licenses >/dev/null
- name: Install Android NDK
run: |
sdkmanager "ndk;27.3.13750724"
echo "ndk.dir=$ANDROID_HOME/ndk/27.3.13750724" >> android-plugin/local.properties
- name: Update glslang external sources
working-directory: 3rdparty/glslang
run: python update_glslang_sources.py
- name: Build plugin APK
run: gradle --no-daemon -p android-plugin :app:assemblePluginRelease -Pmobilegl.apkSuffix="${GITHUB_SHA}" -Pmobilegl.logLevel=MOBILEGL_LOG_LEVEL_INFO --parallel --max-workers "$(nproc)"
env:
SIGNING_STORE_PASSWORD: ${{ secrets.SIGNING_STORE_PASSWORD }}
SIGNING_KEY_ALIAS: ${{ secrets.SIGNING_KEY_ALIAS }}
SIGNING_KEY_PASSWORD: ${{ secrets.SIGNING_KEY_PASSWORD }}
- name: Download ANGLE x86_64 libraries
run: |
angle_dir="android-plugin/app/src/trace/jniLibs/x86_64"
rm -rf "${angle_dir}"
mkdir -p "${angle_dir}"
package_angle_variant() {
variant="$1"
commit="$2"
egl_sha="$3"
gles_sha="$4"
source_dir="${RUNNER_TEMP}/mobilegl-angle-${variant}"
base="https://raw.githubusercontent.com/FCL-Team/FoldCraftLauncher/${commit}/FCLauncher/src/main/jniLibs/x86_64"
mkdir -p "${source_dir}"
curl -L --fail --retry 3 -o "${source_dir}/libEGL_angle.so" "${base}/libEGL_angle.so"
curl -L --fail --retry 3 -o "${source_dir}/libGLESv2_angle.so" "${base}/libGLESv2_angle.so"
echo "${egl_sha} ${source_dir}/libEGL_angle.so" | sha256sum -c -
echo "${gles_sha} ${source_dir}/libGLESv2_angle.so" | sha256sum -c -
for library in libEGL_angle libGLESv2_angle; do
filename="${library}_${variant}.so"
cp "${source_dir}/${library}.so" "${angle_dir}/${filename}"
done
}
package_angle_variant \
ec889e6ea831 \
f2a3d510dffd8f6540a52e1a7d0c5787d151075b \
c41828768d089899fa058ec0bee711a91be88347f29bdb935223da6be1149c40 \
e4f820d99f94365c66df868c7740fef142fe5c0cd7c941790a9e30638857ca4d
package_angle_variant \
90a62123d794 \
bdcc96ac11c79001018ae4375eb73cb54a9f682f \
d0f4298ccc770cc801fc52e21733521646161e8a4adb3bd0052d9a1b57ee0ca8 \
66fdc867e552192d553d59095ea2e3cef4829de65c356f1fd826027b1905972e
- name: Build retrace APK
run: gradle --no-daemon -p android-plugin :app:assembleTraceRelease -Pmobilegl.apkSuffix="${GITHUB_SHA}" -Pmobilegl.abis=all -Pmobilegl.debuggableRelease=true -Pmobilegl.logLevel=MOBILEGL_LOG_LEVEL_INFO --parallel --max-workers "$(nproc)"
env:
SIGNING_STORE_PASSWORD: ${{ secrets.SIGNING_STORE_PASSWORD }}
SIGNING_KEY_ALIAS: ${{ secrets.SIGNING_KEY_ALIAS }}
SIGNING_KEY_PASSWORD: ${{ secrets.SIGNING_KEY_PASSWORD }}
- name: Show ccache stats
if: always()
run: ccache --show-stats
- name: Verify APK metadata and packaging
run: |
AAPT2="$(find "$ANDROID_HOME/build-tools" -name aapt2 -type f | sort -V | tail -n 1)"
plugin_apk="android-plugin/app/build/outputs/apk/plugin/release/MobileGL-plugin-release-${GITHUB_SHA}.apk"
trace_apk="android-plugin/app/build/outputs/apk/trace/release/MobileGL-plugin-trace-release-${GITHUB_SHA}.apk"
test -f "${plugin_apk}"
test -f "${trace_apk}"
bash .github/scripts/validate-plugin-apks.sh "$AAPT2" "$plugin_apk" "$trace_apk"
- name: Verify signed APKs
run: |
APKSIGNER="$(find "$ANDROID_HOME/build-tools" -name apksigner -type f | sort -V | tail -n 1)"
mapfile -t APKS < <(printf '%s\n' \
"android-plugin/app/build/outputs/apk/plugin/release/MobileGL-plugin-release-${GITHUB_SHA}.apk" \
"android-plugin/app/build/outputs/apk/trace/release/MobileGL-plugin-trace-release-${GITHUB_SHA}.apk")
for APK in "${APKS[@]}"; do
if [[ ! -f "$APK" ]]; then
echo "::error::Expected release APK was not produced: $APK"
exit 1
fi
done
for APK in "${APKS[@]}"; do
if [[ "$APK" == *-unsigned.apk ]]; then
echo "::error::Unsigned release APK produced: $APK"
exit 1
fi
"$APKSIGNER" verify --verbose "$APK"
done
- name: Upload plugin APK
uses: actions/upload-artifact@v7
with:
name: MobileGL-plugin-${{ env.date_today }}-${{ github.sha }}
path: android-plugin/app/build/outputs/apk/plugin/release/MobileGL-plugin-release-${{ github.sha }}.apk
archive: false
if-no-files-found: error
- name: Upload retrace APK
uses: actions/upload-artifact@v7
with:
name: MobileGL-retrace-apk-${{ env.date_today }}-${{ github.sha }}
path: android-plugin/app/build/outputs/apk/trace/release/MobileGL-plugin-trace-release-${{ github.sha }}.apk
archive: false
if-no-files-found: error
trace-cases:
name: trace case matrix
runs-on: ubuntu-latest
needs: build
outputs:
android: ${{ steps.trace-cases.outputs.android }}
names: ${{ steps.trace-cases.outputs.names }}
steps:
- name: Checkout repo
uses: actions/checkout@v6
- name: Load trace cases
id: trace-cases
run: |
echo "android=$(python3 tools/trace_replay/trace_cases.py --ci --format github-apk)" >> "$GITHUB_OUTPUT"
echo "names=$(python3 tools/trace_replay/trace_cases.py --ci --format names)" >> "$GITHUB_OUTPUT"
trace-fixtures:
name: trace fixture (${{ matrix.case }})
runs-on: ubuntu-latest
needs: trace-cases
strategy:
fail-fast: false
max-parallel: 4
matrix:
case: ${{ fromJSON(needs.trace-cases.outputs.names) }}
steps:
- name: Checkout repo
uses: actions/checkout@v6
- name: Fetch trace fixture
run: bash .github/scripts/fetch-trace-fixture-lfs.sh '${{ matrix.case }}'
- name: Stage trace fixture
run: |
safe_case="$(printf '%s' '${{ matrix.case }}' | sed 's/[^A-Za-z0-9._-]/_/g')"
stage_dir="trace-fixtures/${safe_case}"
mkdir -p "${stage_dir}"
python3 tools/trace_replay/trace_cases.py --format fixture-files --case '${{ matrix.case }}' |
while IFS= read -r file; do
cp "${file}" "${stage_dir}/"
done
- name: Upload trace fixture
uses: actions/upload-artifact@v7
with:
name: MobileGL-trace-fixture-${{ matrix.case }}
path: trace-fixtures/**
if-no-files-found: error
android-avd:
name: android avd image
runs-on: ubuntu-latest
env:
AVD_NAME: mobilegl-ci
ANDROID_AVD_HOME: ${{ github.workspace }}/.android/avd
ANDROID_HOME: ${{ github.workspace }}/.android/sdk
ANDROID_SDK_ROOT: ${{ github.workspace }}/.android/sdk
steps:
- name: Checkout repo
uses: actions/checkout@v6
- name: Setup Android SDK
uses: android-actions/setup-android@v4
with:
accept-android-sdk-licenses: false
- name: Accept Android SDK licenses
run: yes | sdkmanager --licenses >/dev/null
- name: Restore Android AVD cache
id: android-avd-cache
uses: actions/cache@v5
with:
path: |
${{ env.ANDROID_AVD_HOME }}
${{ env.ANDROID_SDK_ROOT }}/emulator
${{ env.ANDROID_SDK_ROOT }}/platform-tools
${{ env.ANDROID_SDK_ROOT }}/platforms/android-35
${{ env.ANDROID_SDK_ROOT }}/system-images/android-35/google_apis/x86_64
key: ${{ runner.os }}-mobilegl-avd-api35-google_apis-x86_64-pixel_6-v2-${{ hashFiles('android-plugin/run-avd-ci.sh') }}
- name: Create AVD
if: steps.android-avd-cache.outputs.cache-hit != 'true'
run: |
sh android-plugin/run-avd-ci.sh create \
--api-level 35 \
--target google_apis \
--arch x86_64 \
--profile pixel_6 \
--avd-name "${AVD_NAME}"
retrace:
name: retrace (${{ matrix.backend.name }}, ${{ matrix.case.name }})
runs-on: ubuntu-latest
needs:
- build
- android-avd
- trace-cases
- trace-fixtures
if: ${{ always() && needs.build.result == 'success' && needs.android-avd.result == 'success' && needs.trace-cases.result == 'success' }}
timeout-minutes: 75
env:
AVD_NAME: mobilegl-ci
ANDROID_AVD_HOME: ${{ github.workspace }}/.android/avd
ANDROID_HOME: ${{ github.workspace }}/.android/sdk
ANDROID_SDK_ROOT: ${{ github.workspace }}/.android/sdk
strategy:
fail-fast: false
max-parallel: 4
matrix:
backend:
- name: DirectGLES
gpu: software
- name: DirectVulkan
gpu: lavapipe
case: ${{ fromJSON(needs.trace-cases.outputs.android) }}
steps:
- name: Set Swap Space
uses: pierotofy/set-swap-space@v1.0
with:
swap-size-gb: 8
- name: Checkout repo
uses: actions/checkout@v6
- name: Download trace fixture
uses: actions/download-artifact@v8
with:
name: MobileGL-trace-fixture-${{ matrix.case.name }}
path: trace-fixture-download
- name: Install trace fixture
run: |
mkdir -p tools/trace_replay/fixtures
find trace-fixture-download -type f -exec cp {} tools/trace_replay/fixtures/ \;
- name: Set artifact metadata
run: |
echo "date_today=$(date +'%Y-%m-%d')" >> "$GITHUB_ENV"
echo "EMULATOR_LOG=${RUNNER_TEMP}/mobilegl-emulator.log" >> "$GITHUB_ENV"
echo "EMULATOR_PID_FILE=${RUNNER_TEMP}/mobilegl-emulator.pid" >> "$GITHUB_ENV"
- name: Setup Android SDK
uses: android-actions/setup-android@v4
with:
accept-android-sdk-licenses: false
- name: Accept Android SDK licenses
run: yes | sdkmanager --licenses >/dev/null
- name: Restore Android AVD cache
id: android-avd-cache
uses: actions/cache/restore@v5
with:
path: |
${{ env.ANDROID_AVD_HOME }}
${{ env.ANDROID_SDK_ROOT }}/emulator
${{ env.ANDROID_SDK_ROOT }}/platform-tools
${{ env.ANDROID_SDK_ROOT }}/platforms/android-35
${{ env.ANDROID_SDK_ROOT }}/system-images/android-35/google_apis/x86_64
key: ${{ runner.os }}-mobilegl-avd-api35-google_apis-x86_64-pixel_6-v2-${{ hashFiles('android-plugin/run-avd-ci.sh') }}
- name: Download retrace APK
uses: actions/download-artifact@v8
with:
name: MobileGL-plugin-trace-release-${{ github.sha }}.apk
path: android-retrace-apks
- name: Enable KVM
run: |
echo 'KERNEL=="kvm", GROUP="kvm", MODE="0666", OPTIONS+="static_node=kvm"' | sudo tee /etc/udev/rules.d/99-kvm4all.rules
sudo udevadm control --reload-rules
sudo udevadm trigger --name-match=kvm
- name: Create AVD
if: steps.android-avd-cache.outputs.cache-hit != 'true'
run: |
sh android-plugin/run-avd-ci.sh create \
--api-level 35 \
--target google_apis \
--arch x86_64 \
--profile pixel_6 \
--avd-name "${AVD_NAME}"
- name: Launch Emulator
run: |
sh android-plugin/run-avd-ci.sh start \
--avd-name "${AVD_NAME}" \
--gpu "${{ matrix.backend.gpu }}" \
--emulator-log "${EMULATOR_LOG}" \
--pid-file "${EMULATOR_PID_FILE}" \
--boot-timeout 300
- name: Retrace and validate
env:
MOBILEGL_USE_ANGLE: ${{ matrix.backend.name == 'DirectGLES' && '1' || '0' }}
MOBILEGL_TRACE_ANGLE_VARIANT: ${{ matrix.case.name == 'minecraft-1.21.4-fabric-iris-bliss-in-world' && '90a62123d794' || 'ec889e6ea831' }}
MOBILEGL_MAGMA_R11G11B10F_FALLBACK: ${{ matrix.backend.name == 'DirectVulkan' && '1' || '0' }}
run: |
apk_file="android-retrace-apks/MobileGL-plugin-trace-release-${GITHUB_SHA}.apk"
test -f "${apk_file}"
extra_retrace_args=()
# Bliss needs the newer signed ANGLE variant plus sampler mipmap
# min-filter downgrading on ANGLE llvmpipe.
if [ "${{ matrix.backend.name }}" = "DirectGLES" ] && [ "${{ matrix.case.name }}" = "minecraft-1.21.4-fabric-iris-bliss-in-world" ]; then
extra_retrace_args+=(--avoid-angle-llvmpipe-sampler-mipmap-min-filter)
fi
if [ "${{ matrix.case.coherent_as_flush || false }}" = "true" ]; then
extra_retrace_args+=(--coherent-as-flush)
fi
run_retrace() {
timeout "$(( ${{ matrix.case.timeout_seconds }} + 300 ))" sh android-plugin/trace-replay-ci.sh \
--apk-file "${apk_file}" \
--package top.mobilegl.plugin.trace \
--backend "${{ matrix.backend.name }}" \
--result-root android-retrace-result \
--fixture-root android-retrace-fixture \
--case "${{ matrix.case.name }}" \
--trace-archive "${{ matrix.case.trace_archive }}" \
--trace-file "${{ matrix.case.trace_file }}" \
--golden "${{ matrix.case.golden }}" \
--alternate-golden "${{ matrix.case.alternate_golden || '' }}" \
--target-call "${{ matrix.case.target_call }}" \
--width "${{ matrix.case.width }}" \
--height "${{ matrix.case.height }}" \
--ssim-threshold "${{ matrix.case.ssim_threshold || '0.99' }}" \
--crop-x "${{ matrix.case.crop_x }}" \
--crop-y "${{ matrix.case.crop_y }}" \
--crop-width "${{ matrix.case.crop_width }}" \
--crop-height "${{ matrix.case.crop_height }}" \
--timeout-seconds "${{ matrix.case.timeout_seconds }}" \
"${extra_retrace_args[@]}"
}
retrace_status=0
run_retrace || retrace_status=$?
if [ "${retrace_status}" -eq 75 ]; then
echo "::warning::Android emulator infrastructure failed; restarting it and retrying this retrace once."
sh android-plugin/run-avd-ci.sh stop \
--avd-name "${AVD_NAME}" \
--emulator-log "${EMULATOR_LOG}" \
--pid-file "${EMULATOR_PID_FILE}"
adb kill-server || true
sleep 2
sh android-plugin/run-avd-ci.sh start \
--avd-name "${AVD_NAME}" \
--gpu "${{ matrix.backend.gpu }}" \
--emulator-log "${EMULATOR_LOG}" \
--pid-file "${EMULATOR_PID_FILE}" \
--boot-timeout 300
run_retrace
elif [ "${retrace_status}" -ne 0 ]; then
exit "${retrace_status}"
fi
- name: Collect retrace summary inputs
if: always()
run: |
safe_case="$(printf '%s' '${{ matrix.case.name }}' | sed 's/[^A-Za-z0-9._-]/_/g')"
result_dir="android-retrace-result/${safe_case}-${{ matrix.backend.name }}"
mkdir -p "${result_dir}"
if [ -s "${{ matrix.case.golden }}" ]; then
cp "${{ matrix.case.golden }}" "${result_dir}/${safe_case}-${{ matrix.backend.name }}-golden.png"
fi
if [ -n "${{ matrix.case.alternate_golden || '' }}" ] && [ -s "${{ matrix.case.alternate_golden || '' }}" ]; then
cp "${{ matrix.case.alternate_golden || '' }}" "${result_dir}/${safe_case}-${{ matrix.backend.name }}-alternate-golden.png"
fi
- name: Collect emulator diagnostics
if: always()
run: |
mkdir -p android-retrace-result/diagnostics
adb devices -l > android-retrace-result/diagnostics/adb-devices.txt || true
timeout 30 adb logcat -d -t 1000 > android-retrace-result/diagnostics/logcat.txt || true
if [ -f "${EMULATOR_LOG}" ]; then
cp "${EMULATOR_LOG}" android-retrace-result/diagnostics/emulator.log
fi
- name: Stop Emulator
if: always()
run: |
sh android-plugin/run-avd-ci.sh stop \
--avd-name "${AVD_NAME}" \
--emulator-log "${EMULATOR_LOG}" \
--pid-file "${EMULATOR_PID_FILE}"
- name: Upload Android retrace result
if: always()
uses: actions/upload-artifact@v7
with:
name: MobileGL-android-retrace-result-${{ env.date_today }}-${{ github.sha }}-${{ matrix.backend.name }}-${{ matrix.case.name }}
path: android-retrace-result/**
if-no-files-found: warn
retrace-summary:
name: retrace summary
runs-on: ubuntu-latest
needs: retrace
if: always()
steps:
- name: Checkout repo
uses: actions/checkout@v6
- name: Set artifact metadata
run: |
echo "date_today=$(date +'%Y-%m-%d')" >> "$GITHUB_ENV"
- name: Set up Node.js
uses: actions/setup-node@v7
with:
node-version: '22'
- name: Download Android retrace results
uses: actions/download-artifact@v8
with:
pattern: MobileGL-android-retrace-result-*
path: retrace-artifacts
- name: Render retrace summary
run: |
node tools/trace_replay/render_retrace_summary.mjs \
--input retrace-artifacts \
--output-dir android-retrace-summary \
--title "MobileGL Android retrace overview" \
--group-label "Android Emulator" \
--html mobilegl-android-retrace-overview.html
- name: Upload Android retrace summary
uses: actions/upload-artifact@v7
with:
path: android-retrace-summary/mobilegl-android-retrace-overview.html
archive: false
if-no-files-found: error
remove-artifact-clutter:
name: remove artifact clutter
runs-on: ubuntu-latest
needs: retrace-summary
if: always()
permissions:
actions: write
steps:
- name: Delete intermediate Android retrace artifacts
env:
GH_TOKEN: ${{ github.token }}
run: |
declare -A failed_cases=()
while IFS= read -r job_name; do
case_name="${job_name#retrace (*, }"
case_name="${case_name%)}"
failed_cases["${case_name}"]=1
done < <(
gh api --paginate "repos/${GITHUB_REPOSITORY}/actions/runs/${GITHUB_RUN_ID}/jobs?per_page=100" \
--jq '.jobs[] | select(.name | startswith("retrace (")) | select(.conclusion == "failure" or .conclusion == "cancelled" or .conclusion == "timed_out" or .conclusion == "action_required") | .name'
)
if ((${#failed_cases[@]})); then
echo "Retaining fixtures for failed retrace case(s):"
printf ' %s\n' "${!failed_cases[@]}"
else
echo "All retrace jobs succeeded; no fixtures need to be retained."
fi
deleted=0
retained=0
while IFS=$'\t' read -r artifact_id artifact_name; do
if [[ "${artifact_name}" == MobileGL-trace-fixture-* ]]; then
case_name="${artifact_name#MobileGL-trace-fixture-}"
if [[ -v "failed_cases[${case_name}]" ]]; then
echo "Retaining ${artifact_name} (${artifact_id}) for failed retrace."
((retained += 1))
continue
fi
fi
echo "Deleting ${artifact_name} (${artifact_id})"
gh api --method DELETE "repos/${GITHUB_REPOSITORY}/actions/artifacts/${artifact_id}"
((deleted += 1))
done < <(
gh api --paginate "repos/${GITHUB_REPOSITORY}/actions/runs/${GITHUB_RUN_ID}/artifacts?per_page=100" \
--jq '.artifacts[] | select(.name | startswith("MobileGL-trace-fixture-") or startswith("MobileGL-android-retrace-result-") or startswith("trace-fixture-") or startswith("retrace-result-")) | [.id, .name] | @tsv'
)
echo "Deleted ${deleted} intermediate Android artifact(s); retained ${retained} failed-retrace fixture(s)."
+57
View File
@@ -0,0 +1,57 @@
name: Benchmark
on:
push:
branches:
- dev
- Feat/Backend-Direct-GLES
- Feat/Backend-Direct-Vulkan
jobs:
benchmark:
runs-on: ubuntu-latest
env:
# BENCH_ROOT: ${{github.workspace}}/MobileGL/MG_Benchmark
BENCH_ROOT: ${{github.workspace}}
steps:
- name: Set Swap Space
uses: pierotofy/set-swap-space@master
with:
swap-size-gb: 32
- name: Checkout repo
uses: actions/checkout@v4
with:
submodules: true
- name: Get CMake
uses: lukka/get-cmake@latest
- name: Update glslang external sources
working-directory: ${{env.BENCH_ROOT}}/3rdparty/glslang
run: python update_glslang_sources.py
- name: Install clang-20
run: |
sudo apt-get update
sudo apt-get install -y clang-20 clang++-20 lld-20 libc++-20-dev libc++abi-20-dev libvulkan-dev
- name: Show installed toolchain
run: |
clang-20 --version
clang++-20 --version
ld.lld-20 --version || ld.lld --version || true
dpkg -l 'libc++*' || true
- name: Configure CMake
working-directory: ${{env.BENCH_ROOT}}
run: cmake -S . -B build-bench -G Ninja -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=clang-20 -DCMAKE_CXX_COMPILER=clang++-20 -DBENCHMARK_DOWNLOAD_DEPENDENCIES=ON -DBENCHMARK_ENABLE_TESTING=OFF -DMOBILEGL_BUILD_TEST=OFF -DMOBILEGL_BUILD_BENCHMARK=ON -DCMAKE_POLICY_VERSION_MINIMUM=3.5
- name: Build
working-directory: ${{env.BENCH_ROOT}}/build-bench
run: cmake --build .
- name: Benchmark
working-directory: ${{env.BENCH_ROOT}}/build-bench/MobileGL/MG_Benchmark
run: ctest -V -C Release
+24 -526
View File
@@ -6,564 +6,62 @@ on:
- dev
- Feat/Backend-Direct-GLES
- Feat/Backend-Direct-Vulkan
workflow_dispatch:
jobs:
build-linux:
runs-on: ubuntu-latest
env:
BUILD_DIR: build-linux
CCACHE_BASEDIR: ${{ github.workspace }}
CCACHE_COMPRESS: "true"
CCACHE_DIR: ${{ github.workspace }}/.ccache
CCACHE_MAXSIZE: 4G
CCACHE_NOHASHDIR: "true"
steps:
- name: Set Swap Space
uses: pierotofy/set-swap-space@v1.0
with:
swap-size-gb: 32
- name: Checkout repo
uses: actions/checkout@v6
with:
submodules: recursive
- name: Get CMake
uses: lukka/get-cmake@v4.3.3
- name: Restore ccache
uses: actions/cache@v5
with:
path: .ccache
key: ${{ runner.os }}-test-${{ github.job }}-ccache-${{ github.ref_name }}-${{ github.run_id }}
restore-keys: |
${{ runner.os }}-test-${{ github.job }}-ccache-${{ github.ref_name }}-
${{ runner.os }}-test-${{ github.job }}-ccache-
- name: Prepare Vulkan SDK
uses: humbletim/setup-vulkan-sdk@v1.2.1
with:
vulkan-query-version: 1.4.304.1
vulkan-components: Vulkan-Headers, Vulkan-Loader
vulkan-use-cache: true
- name: Update glslang external sources
working-directory: 3rdparty/glslang
run: python update_glslang_sources.py
- name: Install build dependencies
run: |
sudo apt-get update
sudo apt-get install -y ccache clang-20 clang++-20 lld-20 libc++-20-dev libc++abi-20-dev libvulkan-dev libegl1-mesa-dev libgles2-mesa-dev libgl1-mesa-dri mesa-vulkan-drivers ninja-build
- name: Show installed toolchain
run: |
ccache --version
clang-20 --version
clang++-20 --version
ld.lld-20 --version || ld.lld --version || true
dpkg -l 'libc++*' 'libegl*' 'libgles*' 'mesa*' 'vulkan*' || true
- name: Configure CMake
run: |
if [ "${{ secrets.ACTIONS_STEP_DEBUG }}" = "true" ]; then
BUILD_TYPE=Debug
else
BUILD_TYPE=Release
fi
cmake -S . -B "${BUILD_DIR}" -G Ninja \
-DCMAKE_C_COMPILER=clang-20 \
-DCMAKE_CXX_COMPILER=clang++-20 \
-DCMAKE_C_COMPILER_LAUNCHER=ccache \
-DCMAKE_CXX_COMPILER_LAUNCHER=ccache \
-DCMAKE_BUILD_TYPE="${BUILD_TYPE}" \
-DMOBILEGL_LOG_ACTIVE_LEVEL=MOBILEGL_LOG_LEVEL_INFO \
-DMOBILEGL_BUILD_TEST=ON \
-DMOBILEGL_BUILD_BENCHMARK=ON \
-DMOBILEGL_BUILD_TRACE_REPLAY=OFF \
-DBENCHMARK_DOWNLOAD_DEPENDENCIES=ON \
-DBENCHMARK_ENABLE_TESTING=OFF \
-DCMAKE_POLICY_VERSION_MINIMUM=3.5
- name: Build
run: cmake --build "${BUILD_DIR}" --parallel "$(nproc)"
- name: Show ccache stats
if: always()
run: ccache --show-stats
- name: Package Linux runtime
run: |
mkdir -p ci-artifacts
mapfile -t SHARED_LIBS < <(find "${BUILD_DIR}" -type f \( -name '*.so' -o -name '*.so.*' \) -print | sort)
tar \
--exclude='*/CMakeFiles' \
--exclude='*.o' \
--exclude='*.a' \
--exclude='*.ninja*' \
--exclude='build.ninja' \
--exclude='cmake_install.cmake' \
-czf ci-artifacts/mobilegl-linux-runtime.tgz \
"${BUILD_DIR}/CTestTestfile.cmake" \
"${BUILD_DIR}/MobileGL/MG_Test" \
"${BUILD_DIR}/MobileGL/MG_Benchmark" \
"${SHARED_LIBS[@]}"
- name: Upload Linux runtime
uses: actions/upload-artifact@v7
with:
name: mobilegl-linux-runtime
path: ci-artifacts/mobilegl-linux-runtime.tgz
if-no-files-found: error
test:
runs-on: ubuntu-latest
needs: build-linux
steps:
- name: Checkout repo
uses: actions/checkout@v6
- name: Get CMake
uses: lukka/get-cmake@v4.3.3
- name: Install runtime dependencies
run: |
sudo apt-get update
sudo apt-get install -y libvulkan1 libegl1 libgles2 libgl1-mesa-dri mesa-vulkan-drivers
- name: Download Linux runtime
uses: actions/download-artifact@v8
with:
name: mobilegl-linux-runtime
path: .
- name: Unpack Linux runtime
run: tar -xzf mobilegl-linux-runtime.tgz
- name: Normalize CTest command paths
run: |
python - <<'PY'
from pathlib import Path
import re
for path in Path('build-linux').rglob('CTestTestfile.cmake'):
text = path.read_text()
text = re.sub(r'"[^"]*/cmake-[^"]*/bin/cmake"', '"cmake"', text)
path.write_text(text)
PY
- name: Test
working-directory: build-linux
run: |
if [ "${{ secrets.ACTIONS_STEP_DEBUG }}" = "true" ]; then
ctest -V -L unit --no-tests=error
else
ctest --output-on-failure -L unit --no-tests=error
fi
benchmark:
runs-on: ubuntu-latest
needs: build-linux
steps:
- name: Checkout repo
uses: actions/checkout@v6
- name: Get CMake
uses: lukka/get-cmake@v4.3.3
- name: Install runtime dependencies
run: |
sudo apt-get update
sudo apt-get install -y libvulkan1 libegl1 libgles2 libgl1-mesa-dri mesa-vulkan-drivers
- name: Download Linux runtime
uses: actions/download-artifact@v8
with:
name: mobilegl-linux-runtime
path: .
- name: Unpack Linux runtime
run: tar -xzf mobilegl-linux-runtime.tgz
- name: Normalize CTest command paths
run: |
python - <<'PY'
from pathlib import Path
import re
for path in Path('build-linux').rglob('CTestTestfile.cmake'):
text = path.read_text()
text = re.sub(r'"[^"]*/cmake-[^"]*/bin/cmake"', '"cmake"', text)
path.write_text(text)
PY
- name: Benchmark
working-directory: build-linux
run: ctest -V -C Release -L benchmark --no-tests=error
build-retrace:
runs-on: ubuntu-latest
needs:
- build-linux
- test
- benchmark
env:
BUILD_DIR: build-retrace
CCACHE_BASEDIR: ${{ github.workspace }}
CCACHE_COMPRESS: "true"
CCACHE_DIR: ${{ github.workspace }}/.ccache
CCACHE_MAXSIZE: 4G
CCACHE_NOHASHDIR: "true"
MOBILEGL_LIBRARY: ${{ github.workspace }}/build-linux/libMobileGL.so
# TEST_ROOT: ${{github.workspace}}/MobileGL/MG_Test
TEST_ROOT: ${{github.workspace}}
steps:
- name: Set Swap Space
uses: pierotofy/set-swap-space@v1.0
uses: pierotofy/set-swap-space@master
with:
swap-size-gb: 32
- name: Checkout repo
uses: actions/checkout@v6
uses: actions/checkout@v4
with:
submodules: recursive
submodules: true
- name: Get CMake
uses: lukka/get-cmake@v4.3.3
- name: Restore ccache
uses: actions/cache@v5
with:
path: .ccache
key: ${{ runner.os }}-test-${{ github.job }}-ccache-${{ github.ref_name }}-${{ github.run_id }}
restore-keys: |
${{ runner.os }}-test-${{ github.job }}-ccache-${{ github.ref_name }}-
${{ runner.os }}-test-${{ github.job }}-ccache-
- name: Prepare Vulkan SDK
uses: humbletim/setup-vulkan-sdk@v1.2.1
with:
vulkan-query-version: 1.4.304.1
vulkan-components: Vulkan-Headers, Vulkan-Loader
vulkan-use-cache: true
uses: lukka/get-cmake@latest
- name: Update glslang external sources
working-directory: 3rdparty/glslang
working-directory: ${{env.TEST_ROOT}}/3rdparty/glslang
run: python update_glslang_sources.py
- name: Install dependencies
- name: Install clang-20
run: |
sudo apt-get update
sudo apt-get install -y ccache clang-20 clang++-20 lld-20 libc++-20-dev libc++abi-20-dev libvulkan-dev libegl1-mesa-dev libgles2-mesa-dev libgl1-mesa-dri mesa-vulkan-drivers ninja-build
sudo apt-get install -y clang-20 clang++-20 lld-20 libc++-20-dev libc++abi-20-dev libvulkan-dev
- name: Show installed toolchain
run: |
ccache --version
clang-20 --version
clang++-20 --version
ld.lld-20 --version || ld.lld --version || true
dpkg -l 'libc++*' 'libegl*' 'libgles*' 'mesa*' 'vulkan*' || true
- name: Download Linux runtime
uses: actions/download-artifact@v8
with:
name: mobilegl-linux-runtime
path: .
- name: Unpack Linux runtime
run: |
tar -xzf mobilegl-linux-runtime.tgz
test -f "${MOBILEGL_LIBRARY}"
dpkg -l 'libc++*' || true
- name: Configure CMake
working-directory: ${{env.TEST_ROOT}}
run: |
if [ "${{ secrets.ACTIONS_STEP_DEBUG }}" = "true" ]; then
BUILD_TYPE=Debug
if [ "${{ secrets.ACTIONS_STEP_DEBUG }}" == "true" ]; then
cmake -S . -B build-test -G Ninja -DCMAKE_C_COMPILER=clang-20 -DCMAKE_CXX_COMPILER=clang++-20 -DCMAKE_BUILD_TYPE=Debug -DMOBILEGL_BUILD_TEST=ON -DMOBILEGL_BUILD_BENCHMARK=OFF -DCMAKE_POLICY_VERSION_MINIMUM=3.5
else
BUILD_TYPE=Release
cmake -S . -B build-test -G Ninja -DCMAKE_C_COMPILER=clang-20 -DCMAKE_CXX_COMPILER=clang++-20 -DCMAKE_BUILD_TYPE=Release -DMOBILEGL_BUILD_TEST=ON -DMOBILEGL_BUILD_BENCHMARK=OFF -DCMAKE_POLICY_VERSION_MINIMUM=3.5
fi
- name: Build
working-directory: ${{env.TEST_ROOT}}/build-test
run: cmake --build .
cmake -S . -B "${BUILD_DIR}" -G Ninja \
-DCMAKE_C_COMPILER=clang-20 \
-DCMAKE_CXX_COMPILER=clang++-20 \
-DCMAKE_C_COMPILER_LAUNCHER=ccache \
-DCMAKE_CXX_COMPILER_LAUNCHER=ccache \
-DCMAKE_BUILD_TYPE="${BUILD_TYPE}" \
-DMOBILEGL_LOG_ACTIVE_LEVEL=MOBILEGL_LOG_LEVEL_INFO \
-DMOBILEGL_BUILD_TEST=OFF \
-DMOBILEGL_BUILD_BENCHMARK=OFF \
-DMOBILEGL_BUILD_TRACE_REPLAY=ON \
-DMOBILEGL_TRACE_REPLAY_MOBILEGL_LIBRARY="${MOBILEGL_LIBRARY}" \
-DCMAKE_POLICY_VERSION_MINIMUM=3.5
- name: Build trace replay
run: cmake --build "${BUILD_DIR}" --target mobilegl_trace_replay --parallel "$(nproc)"
- name: Show ccache stats
if: always()
run: ccache --show-stats
- name: Normalize CTest command paths
- name: Test
working-directory: ${{env.TEST_ROOT}}/build-test/MobileGL/MG_Test
run: |
python - <<'PY'
from pathlib import Path
import re
for path in Path('build-retrace').rglob('CTestTestfile.cmake'):
text = path.read_text()
text = re.sub(r'"[^"]*/cmake-[^"]*/bin/cmake"', '"cmake"', text)
path.write_text(text)
PY
- name: Package trace replay
run: |
mkdir -p ci-artifacts
tar -czf ci-artifacts/mobilegl-trace-replay.tgz \
build-retrace/tools/trace_replay/mobilegl_trace_replay \
build-retrace/tools/trace_replay/CTestTestfile.cmake
- name: Upload trace replay
uses: actions/upload-artifact@v7
with:
name: mobilegl-trace-replay
path: ci-artifacts/mobilegl-trace-replay.tgz
if-no-files-found: error
trace-cases:
name: trace case matrix
runs-on: ubuntu-latest
needs:
- test
- benchmark
outputs:
names: ${{ steps.trace-cases.outputs.names }}
steps:
- name: Checkout repo
uses: actions/checkout@v6
- name: Load trace cases
id: trace-cases
run: echo "names=$(python3 tools/trace_replay/trace_cases.py --ci --format names)" >> "$GITHUB_OUTPUT"
trace-fixtures:
name: trace fixture (${{ matrix.case }})
runs-on: ubuntu-latest
needs: trace-cases
strategy:
fail-fast: false
max-parallel: 4
matrix:
case: ${{ fromJSON(needs.trace-cases.outputs.names) }}
steps:
- name: Checkout repo
uses: actions/checkout@v6
- name: Fetch trace fixture
run: bash .github/scripts/fetch-trace-fixture-lfs.sh '${{ matrix.case }}'
- name: Stage trace fixture
run: |
safe_case="$(printf '%s' '${{ matrix.case }}' | sed 's/[^A-Za-z0-9._-]/_/g')"
stage_dir="trace-fixtures/${safe_case}"
mkdir -p "${stage_dir}"
python3 tools/trace_replay/trace_cases.py --format fixture-files --case '${{ matrix.case }}' |
while IFS= read -r file; do
cp "${file}" "${stage_dir}/"
done
- name: Upload trace fixture
uses: actions/upload-artifact@v7
with:
name: trace-fixture-${{ matrix.case }}
path: trace-fixtures/**
if-no-files-found: error
retrace:
name: retrace (${{ matrix.backend }}, ${{ matrix.case }})
runs-on: ubuntu-latest
needs:
- build-linux
- build-retrace
- trace-cases
- trace-fixtures
if: ${{ always() && needs.build-linux.result == 'success' && needs.build-retrace.result == 'success' && needs.trace-cases.result == 'success' }}
strategy:
fail-fast: false
max-parallel: 4
matrix:
backend:
- DirectGLES
- DirectVulkan
case: ${{ fromJSON(needs.trace-cases.outputs.names) }}
steps:
- name: Set Swap Space
uses: pierotofy/set-swap-space@v1.0
with:
swap-size-gb: 16
- name: Checkout repo
uses: actions/checkout@v6
- name: Download trace fixture
uses: actions/download-artifact@v8
with:
name: trace-fixture-${{ matrix.case }}
path: trace-fixture-download
- name: Install trace fixture
run: |
mkdir -p tools/trace_replay/fixtures
find trace-fixture-download -type f -exec cp {} tools/trace_replay/fixtures/ \;
- name: Get CMake
uses: lukka/get-cmake@v4.3.3
- name: Install runtime dependencies
run: |
sudo apt-get update
sudo apt-get install -y libvulkan1 libegl1-mesa-dev libgles2-mesa-dev libgl1-mesa-dri mesa-vulkan-drivers
test -e /usr/lib/x86_64-linux-gnu/libEGL.so
test -e /usr/lib/x86_64-linux-gnu/libGLESv2.so
- name: Download Linux runtime
uses: actions/download-artifact@v8
with:
name: mobilegl-linux-runtime
path: .
- name: Download trace replay
uses: actions/download-artifact@v8
with:
name: mobilegl-trace-replay
path: .
- name: Unpack retrace runtime
run: |
tar -xzf mobilegl-linux-runtime.tgz
tar -xzf mobilegl-trace-replay.tgz
test -f build-linux/libMobileGL.so
test -f build-retrace/tools/trace_replay/mobilegl_trace_replay
- name: Retrace and validate
working-directory: build-retrace/tools/trace_replay
run: |
if [ '${{ matrix.backend }}' = 'DirectVulkan' ]; then
export MOBILEGL_MAGMA_R11G11B10F_FALLBACK=1
fi
# The blended depth-write quirk auto-enables only on Qualcomm, which no CI
# runner has, so force it on for the OIT case it exists to fix. ForceOn
# bypasses only the vendor gate, so this exercises the real strip on
# lavapipe. The Android AVD lane deliberately leaves it off, keeping the
# unstripped path covered for the same trace.
if [ '${{ matrix.backend }}' = 'DirectVulkan' ] \
&& [ '${{ matrix.case }}' = 'improved-transparency-minecraft-26.3' ]; then
export MOBILEGL_MAGMA_DISABLE_BLENDED_DEPTH_WRITE=1
fi
ctest -V --no-tests=error -R '^MobileGLTraceReplay\.${{ matrix.case }}\.${{ matrix.backend }}$'
- name: Upload actual image
if: always()
uses: actions/upload-artifact@v7
with:
name: retrace-result-${{ matrix.backend }}-${{ matrix.case }}
path: |
build-retrace/tools/trace_replay/${{ matrix.case }}/actual-images/**
build-retrace/tools/trace_replay/${{ matrix.case }}/${{ matrix.backend }}/output/**
if-no-files-found: warn
retrace-summary:
name: retrace summary
runs-on: ubuntu-latest
needs: retrace
if: ${{ always() && needs.retrace.result != 'skipped' }}
steps:
- name: Checkout repo
uses: actions/checkout@v6
- name: Set artifact metadata
run: |
echo "date_today=$(date +'%Y-%m-%d')" >> "$GITHUB_ENV"
- name: Set up Node.js
uses: actions/setup-node@v7
with:
node-version: '22'
- name: Download retrace results
uses: actions/download-artifact@v8
with:
pattern: retrace-result-*
path: retrace-artifacts
- name: Render retrace summary
run: |
node tools/trace_replay/render_retrace_summary.mjs \
--input retrace-artifacts \
--output-dir retrace-summary \
--title "MobileGL Linux retrace overview" \
--group-label "Linux" \
--html mobilegl-linux-retrace-overview.html
- name: Upload retrace summary
uses: actions/upload-artifact@v7
with:
path: retrace-summary/mobilegl-linux-retrace-overview.html
archive: false
if-no-files-found: error
remove-artifact-clutter:
name: remove artifact clutter
runs-on: ubuntu-latest
needs: retrace-summary
if: always()
permissions:
actions: write
steps:
- name: Delete intermediate Linux retrace artifacts
env:
GH_TOKEN: ${{ github.token }}
run: |
declare -A failed_cases=()
while IFS= read -r job_name; do
case_name="${job_name#retrace (*, }"
case_name="${case_name%)}"
failed_cases["${case_name}"]=1
done < <(
gh api --paginate "repos/${GITHUB_REPOSITORY}/actions/runs/${GITHUB_RUN_ID}/jobs?per_page=100" \
--jq '.jobs[] | select(.name | startswith("retrace (")) | select(.conclusion == "failure" or .conclusion == "cancelled" or .conclusion == "timed_out" or .conclusion == "action_required") | .name'
)
if ((${#failed_cases[@]})); then
echo "Retaining fixtures for failed retrace case(s):"
printf ' %s\n' "${!failed_cases[@]}"
if [ "${{ secrets.ACTIONS_STEP_DEBUG }}" == "true" ]; then
ctest -V
else
echo "All retrace jobs succeeded; no fixtures need to be retained."
ctest
fi
deleted=0
retained=0
while IFS=$'\t' read -r artifact_id artifact_name; do
if [[ "${artifact_name}" == trace-fixture-* ]]; then
case_name="${artifact_name#trace-fixture-}"
if [[ -v "failed_cases[${case_name}]" ]]; then
echo "Retaining ${artifact_name} (${artifact_id}) for failed retrace."
((retained += 1))
continue
fi
fi
echo "Deleting ${artifact_name} (${artifact_id})"
gh api --method DELETE "repos/${GITHUB_REPOSITORY}/actions/artifacts/${artifact_id}"
((deleted += 1))
done < <(
gh api --paginate "repos/${GITHUB_REPOSITORY}/actions/runs/${GITHUB_RUN_ID}/artifacts?per_page=100" \
--jq '.artifacts[] | select(.name | startswith("trace-fixture-") or startswith("retrace-result-")) | [.id, .name] | @tsv'
)
echo "Deleted ${deleted} intermediate Linux artifact(s); retained ${retained} failed-retrace fixture(s)."
+1 -11
View File
@@ -14,14 +14,4 @@ MobileGLCodeManager
.vscode
.clangd
MobileGL/MG_Test/build
/build_*
/cmake-build*
.idea
MobileGL/MG*/build*
MobileGL/MG*/cmake-build*
/android-plugin/.gradle
/android-plugin/build
/android-plugin/app/build
/android-plugin/app/src/trace/jniLibs
/android-plugin/local.properties
tools/trace_replay/work/
/build_*
-15
View File
@@ -16,18 +16,3 @@
[submodule "3rdparty/xxHash"]
path = 3rdparty/xxHash
url = https://github.com/Cyan4973/xxHash.git
[submodule "3rdparty/VulkanMemoryAllocator"]
path = 3rdparty/VulkanMemoryAllocator
url = https://github.com/GPUOpen-LibrariesAndSDKs/VulkanMemoryAllocator.git
[submodule "3rdparty/Vulkan-Utility-Libraries"]
path = 3rdparty/Vulkan-Utility-Libraries
url = https://github.com/KhronosGroup/Vulkan-Utility-Libraries.git
[submodule "3rdparty/Vulkan-Headers"]
path = 3rdparty/Vulkan-Headers
url = https://github.com/KhronosGroup/Vulkan-Headers.git
[submodule "3rdparty/SPIRV-Reflect"]
path = 3rdparty/SPIRV-Reflect
url = https://github.com/KhronosGroup/SPIRV-Reflect.git
[submodule "3rdparty/apitrace"]
path = 3rdparty/apitrace
url = https://github.com/MobileGL-Dev/apitrace.git
Submodule 3rdparty/apitrace deleted from 10935bb5e4
+12 -192
View File
@@ -6,11 +6,6 @@ option(MOBILEGL_BUILD_TEST "Build MobileGL tests"
option(MOBILEGL_BUILD_BENCHMARK "Build MobileGL benchmarks" ON )
option(MOBILEGL_FORCE_RELEASE_OPT "Enable Release optimization flags in Debug build" ON )
option(MOBILEGL_ENABLE_TRACY "Enable tracy for profiling" OFF)
option(MOBILEGL_BUILD_TRACE_REPLAY "Build desktop apitrace replay runner" OFF)
option(MOBILEGL_TRACE_ANGLE_VARIANTS "Enable signed trace-APK ANGLE variant loading" OFF)
option(MOBILEGL_IOS "Build MobileGL for iOS instead of macOS when APPLE is set" OFF)
set(MOBILEGL_LOG_ACTIVE_LEVEL "MOBILEGL_LOG_LEVEL_INFO" CACHE STRING "MobileGL active log level macro")
set(MOBILEGL_VULKAN_LIBRARY "" CACHE FILEPATH "Vulkan loader/MoltenVK library to link for iOS builds")
if (ANDROID)
set(MOBILEGL_BUILD_TEST OFF CACHE BOOL "Build MobileGL tests" FORCE)
@@ -109,20 +104,9 @@ set(SPIRV_CROSS_ENABLE_CPP OFF CACHE BOOL "Disable C++ API target" FORCE)
set(SPIRV_CROSS_CLI OFF CACHE BOOL "Disable CLI binary" FORCE)
set(SPIRV_CROSS_STATIC ON CACHE BOOL "Prefer static libs" FORCE)
set(SPIRV_REFLECT_EXECUTABLE OFF CACHE BOOL "Build spirv-reflect executable" FORCE)
set(SPIRV_REFLECT_STATIC_LIB ON CACHE BOOL "Build a SPIRV-Reflect static library" FORCE)
set(SPIRV_REFLECT_BUILD_TESTS OFF CACHE BOOL "Build the SPIRV-Reflect test suite" FORCE)
set(SPIRV_REFLECT_ENABLE_ASSERTS OFF CACHE BOOL "Enable asserts for debugging" FORCE)
set(SPIRV_REFLECT_ENABLE_ASAN OFF CACHE BOOL "Use address sanitization" FORCE)
set(SPIRV_REFLECT_INSTALL OFF CACHE BOOL "Whether to install" FORCE)
# add_subdirectory(3rdparty/DiligentCore)
add_subdirectory(3rdparty/glslang)
add_subdirectory(3rdparty/SPIRV-Cross)
add_subdirectory(3rdparty/VulkanMemoryAllocator)
add_subdirectory(3rdparty/Vulkan-Headers)
add_subdirectory(3rdparty/Vulkan-Utility-Libraries)
add_subdirectory(3rdparty/SPIRV-Reflect)
set(XXHASH_BUILD_XXHSUM OFF)
option(BUILD_SHARED_LIBS OFF)
@@ -143,7 +127,6 @@ endif ()
set(SOURCE_FILES
MobileGL/Init.cpp
MobileGL/GlobalObjects.cpp
MobileGL/ConfigLoader.cpp
MobileGL/MG_Util/Debug/Log.cpp
@@ -175,8 +158,6 @@ set(SOURCE_FILES
MobileGL/MG_Util/Converters/GLToMG/RenderStateEnumConverter.cpp
MobileGL/MG_Util/Converters/GLToMG/ProgramEnumConverter.cpp
MobileGL/MG_Util/Converters/MGToMG/TextureEnumConverter.cpp
MobileGL/MG_Util/Converters/MGToVk/RenderStateEnumConverter.cpp
MobileGL/MG_Util/Converters/MGToVk/TextureEnumConverter.cpp
MobileGL/MG_Util/Classifiers/TextureEnumClassifier.cpp
@@ -184,25 +165,11 @@ set(SOURCE_FILES
MobileGL/MG_Util/ShaderTranspiler/SpvcSession.cpp
MobileGL/MG_Util/ShaderTranspiler/ShaderSourceProcessor.cpp
MobileGL/MG_Util/ShaderTranspiler/glslang/TMglGlslIoResolver.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/FlattenInterfaceStructPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/EliminateFloatEqualsZeroPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/RenameSamplerFunctionParameterPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/DecomposeWorkgroupVec3Pass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/DecoratePositionInvariantPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/LowerDrawParametersPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/RebaseInstanceIndexPass.cpp
MobileGL/MG_Util/ShaderTranspiler/SpirvPasses/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
MobileGL/MG_Util/SelfTest/DriverPost.cpp
MobileGL/MG_Util/Texture/PixelStoreProcessor.cpp
MobileGL/MG_Util/Texture/TextureFormatProcessor.cpp
@@ -230,7 +197,6 @@ set(SOURCE_FILES
MobileGL/MG_Impl/GLImpl/Exporting/Definitions.cpp
MobileGL/MG_Impl/GLImpl/Getter/GL_Getter.cpp
MobileGL/MG_Impl/GLImpl/Sync/GL_Sync.cpp
MobileGL/MG_Impl/GLImpl/Query/GL_Query.cpp
MobileGL/MG_Impl/Init.cpp
MobileGL/MG_Impl/GetProcAddress.cpp
@@ -243,28 +209,16 @@ set(SOURCE_FILES
MobileGL/MG_Backend/DirectGLES/Utils.cpp
MobileGL/MG_Backend/DirectGLES/Managers.cpp
MobileGL/MG_Backend/DirectVulkan/DirectVulkan.cpp
MobileGL/MG_Backend/DirectVulkan/BackendObject_DirectVulkan.cpp
MobileGL/MG_Backend/DirectVulkan/VmaImpl.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/VulkanRenderer.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/SwapchainObject.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/FrameContext.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/PipelineFactory.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/ProgramFactory.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/UniformManager.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/BufferArena.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/VkBufferManager.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/VertexInputStateBuilder.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/VertexInputStateFactory.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/VkBufferObject.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/VkTextureManager.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/VkTimerQueryManager.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/VkSamplerManager.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/VkClearManager.cpp
MobileGL/MG_Backend/DirectVulkan/Renderer/VkRenderPassManager.cpp
MobileGL/MG_Backend/DirectVulkanTMP/DirectVulkanTMP.cpp
MobileGL/MG_Backend/DirectVulkanTMP/TmpImpl.cpp
MobileGL/MG_Backend/DirectVulkanTMP/BackendObject_DirectVulkanTMP.cpp
MobileGL/MG_Backend/DirectVulkanTMP/Renderer/VulkanContext.cpp
MobileGL/MG_Backend/DirectVulkanTMP/Renderer/SwapchainManager.cpp
MobileGL/MG_Backend/DirectVulkanTMP/Renderer/PipelineManager.cpp
MobileGL/MG_Backend/DirectVulkanTMP/Renderer/FrameContext.cpp
MobileGL/MG_Backend/DirectVulkanTMP/Managers/ProgramManager.cpp
MobileGL/MG_State/GLState/Core.cpp
MobileGL/MG_State/EGLState/Core.cpp
MobileGL/MG_State/GLState/ErrorState/Error.cpp
MobileGL/MG_State/GLState/BufferState/BufferState.cpp
MobileGL/MG_State/GLState/BufferState/BufferObject.cpp
@@ -291,47 +245,14 @@ set(SOURCE_FILES
MobileGL/MG_State/GLState/RenderbufferState/RenderbufferState.cpp
)
if (APPLE AND NOT MOBILEGL_IOS)
list(APPEND SOURCE_FILES
MobileGL/MG_Impl/CGLImpl/CGLImpl.cpp
MobileGL/MG_Impl/CGLImpl/Exporting/Definitions.cpp
MobileGL/MG_Impl/DyldInterpose/DyldInterpose.cpp
MobileGL/MG_Impl/NSOpenGLImpl/NSOpenGLImpl.cpp
)
endif()
if (ANDROID)
list(APPEND SOURCE_FILES
MobileGL/MG_Util/SelfTest/DriverPostJni.cpp
)
endif()
if (WIN32)
list(APPEND SOURCE_FILES
MobileGL/MG_Impl/WGLImpl/WGLImpl.cpp
MobileGL/MG_Impl/WGLImpl/Exporting/Definitions.cpp
)
endif()
set(MOBILEGL_LINK_LIBRARIES
glslang::glslang
spirv-cross-c
SPIRV-Tools-opt
SPIRV-Tools
xxHash::xxhash
GPUOpen::VulkanMemoryAllocator
Vulkan::UtilityHeaders
spirv-reflect-static
)
set(MOBILEGL_COMPILE_DEF
-DVMA_STATIC_VULKAN_FUNCTIONS=0
-DVMA_DYNAMIC_VULKAN_FUNCTIONS=1
-DVMA_VULKAN_VERSION=1001000
)
message(STATUS "MOBILEGL_COMPILE_DEF=${MOBILEGL_COMPILE_DEF}")
set(MOBILEGL_INCLUDE_DIR
${CMAKE_SOURCE_DIR}/include
${CMAKE_SOURCE_DIR}/MobileGL
@@ -341,18 +262,10 @@ set(MOBILEGL_INCLUDE_DIR
${SPIRV-Headers_SOURCE_DIR}/include
)
add_library(${CMAKE_PROJECT_NAME} SHARED
add_library(${CMAKE_PROJECT_NAME} SHARED
${SOURCE_FILES}
)
if (WIN32)
# The wgl* entry points are exported via .def (see the comment in wgl.def);
# only the shared library links it.
target_sources(${CMAKE_PROJECT_NAME} PRIVATE
MobileGL/MG_Impl/WGLImpl/Exporting/wgl.def
)
endif()
if (CMAKE_BUILD_TYPE STREQUAL "Debug")
set_target_properties(${CMAKE_PROJECT_NAME} PROPERTIES
C_VISIBILITY_PRESET default
@@ -372,41 +285,10 @@ target_include_directories(${CMAKE_PROJECT_NAME} PUBLIC
)
target_link_libraries(${CMAKE_PROJECT_NAME}
PUBLIC
PRIVATE
${MOBILEGL_LINK_LIBRARIES}
)
target_compile_definitions(${CMAKE_PROJECT_NAME}
PUBLIC
${MOBILEGL_COMPILE_DEF}
MOBILEGL_LOG_ACTIVE_LEVEL=${MOBILEGL_LOG_ACTIVE_LEVEL}
$<$<BOOL:${MOBILEGL_TRACE_ANGLE_VARIANTS}>:MOBILEGL_TRACE_ANGLE_VARIANTS=1>
)
if(UNIX AND NOT APPLE AND NOT ANDROID)
foreach(MOBILEGL_LOADER_ALIAS
libEGL.so libEGL.so.1)
add_custom_command(TARGET ${CMAKE_PROJECT_NAME} POST_BUILD
COMMAND ${CMAKE_COMMAND} -E create_symlink
"$<TARGET_FILE_NAME:${CMAKE_PROJECT_NAME}>"
"$<TARGET_FILE_DIR:${CMAKE_PROJECT_NAME}>/${MOBILEGL_LOADER_ALIAS}"
COMMENT "Creating ${MOBILEGL_LOADER_ALIAS} alias for Linux GL/EGL loaders"
)
endforeach()
endif()
if(WIN32)
# Drop-in for the classic GL loader path: a copy named opengl32.dll placed
# next to a host executable is what LoadLibrary("opengl32.dll") and gdi32's
# pixel-format forwarding will resolve.
add_custom_command(TARGET ${CMAKE_PROJECT_NAME} POST_BUILD
COMMAND ${CMAKE_COMMAND} -E copy_if_different
"$<TARGET_FILE:${CMAKE_PROJECT_NAME}>"
"$<TARGET_FILE_DIR:${CMAKE_PROJECT_NAME}>/opengl32.dll"
COMMENT "Creating opengl32.dll drop-in copy"
)
endif()
if(NOT ANDROID)
add_library(${CMAKE_PROJECT_NAME}_s STATIC
${SOURCE_FILES}
@@ -431,15 +313,9 @@ if(NOT ANDROID)
)
target_link_libraries(${CMAKE_PROJECT_NAME}_s
PUBLIC
PRIVATE
${MOBILEGL_LINK_LIBRARIES}
)
target_compile_definitions(${CMAKE_PROJECT_NAME}_s
PUBLIC
${MOBILEGL_COMPILE_DEF}
MOBILEGL_LOG_ACTIVE_LEVEL=${MOBILEGL_LOG_ACTIVE_LEVEL}
)
endif()
if (TRACY_ENABLE)
@@ -457,48 +333,7 @@ if (ANDROID)
)
endif()
if (APPLE AND NOT MOBILEGL_IOS)
target_link_libraries(${CMAKE_PROJECT_NAME} PUBLIC
"-framework Cocoa"
"-framework QuartzCore"
"-framework Foundation"
"-framework OpenGL"
objc)
if(TARGET ${CMAKE_PROJECT_NAME}_s)
target_link_libraries(${CMAKE_PROJECT_NAME}_s PUBLIC
"-framework Cocoa"
"-framework QuartzCore"
"-framework Foundation"
"-framework OpenGL"
objc)
endif()
endif()
if (APPLE AND MOBILEGL_IOS)
target_compile_definitions(${CMAKE_PROJECT_NAME} PUBLIC MOBILEGL_IOS=1 _LIBCPP_DISABLE_AVAILABILITY)
target_link_libraries(${CMAKE_PROJECT_NAME} PUBLIC
"-framework CoreGraphics"
"-framework Foundation"
"-framework QuartzCore"
objc)
if (MOBILEGL_VULKAN_LIBRARY)
target_link_libraries(${CMAKE_PROJECT_NAME} PUBLIC "${MOBILEGL_VULKAN_LIBRARY}")
endif()
if(TARGET ${CMAKE_PROJECT_NAME}_s)
target_compile_definitions(${CMAKE_PROJECT_NAME}_s PUBLIC MOBILEGL_IOS=1 _LIBCPP_DISABLE_AVAILABILITY)
target_link_libraries(${CMAKE_PROJECT_NAME}_s PUBLIC
"-framework CoreGraphics"
"-framework Foundation"
"-framework QuartzCore"
objc)
if (MOBILEGL_VULKAN_LIBRARY)
target_link_libraries(${CMAKE_PROJECT_NAME}_s PUBLIC "${MOBILEGL_VULKAN_LIBRARY}")
endif()
endif()
endif()
if (NOT ANDROID AND NOT MOBILEGL_IOS)
if (NOT ANDROID)
find_package(Vulkan)
if (Vulkan_FOUND)
target_link_libraries(${CMAKE_PROJECT_NAME} PUBLIC Vulkan::Vulkan Vulkan::Headers)
@@ -506,17 +341,6 @@ if (NOT ANDROID AND NOT MOBILEGL_IOS)
target_include_directories(${CMAKE_PROJECT_NAME} PUBLIC ${Vulkan_INCLUDE_DIR})
target_include_directories(${CMAKE_PROJECT_NAME}_s PUBLIC ${Vulkan_INCLUDE_DIR})
endif ()
endif ()
if (NOT ANDROID)
# Enable testing in the top-level scope so a CTestTestfile.cmake is emitted
# at the build-tree root. This lets `ctest` be invoked from the top-level
# build directory (IDE "run all tests", CI) and discover every test in the
# subdirectories below, instead of having to descend into each
# MG_Test/MG_Benchmark subdirectory. Tests are tagged with CTest labels
# (unit / benchmark / integration), so e.g. `ctest -L unit` selects just
# the unit suite.
enable_testing()
if (MOBILEGL_BUILD_TEST)
add_subdirectory(MobileGL/MG_Test)
@@ -525,8 +349,4 @@ if (NOT ANDROID)
if (MOBILEGL_BUILD_BENCHMARK)
add_subdirectory(MobileGL/MG_Benchmark)
endif()
if (MOBILEGL_BUILD_TRACE_REPLAY)
add_subdirectory(tools/trace_replay)
endif()
endif()
+1 -81
View File
@@ -14,88 +14,8 @@ 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, 7, 0, "-dev", VersionType::Development};
inline const Version CoreVersion = {26, 2, 0, "-dev", VersionType::Development};
inline const VersionStringFormatAttrib DefaultVersionStringFormatAttrib = {2, 2, 0, true, true};
inline const Uint64 CacheVersion = 0;
extern BackendType ActiveBackendType;
// Tri-state override for device-specific quirks: Auto lets the detected device decide,
// ForceOn/ForceOff bypass the detection in either direction. ForceOn only bypasses the
// device gate - each quirk keeps its structural safety checks.
enum class QuirkOverride : Uint8 {
Auto = 0,
ForceOn,
ForceOff,
};
// Feature toggles parsed once from environment variables in MG_ConfigLoader::Init()
// (ConfigLoader.cpp), before the accepted-env map is destroyed. All Bool fields share
// one truthy rule: the variable is set, non-empty, not "0", and not "false"
// (case-insensitive).
//
// Env variables intentionally NOT mirrored here (kept as live std::getenv at their
// call sites):
// - DISPLAY: X11 session variable, not MobileGL configuration.
// - MOBILEGL_LOG_FILE_PATH: log-file init runs before MG_ConfigLoader::Init
// (see MG_Util/Debug/Log.cpp).
struct FeaturesTable {
// MOBILEGL_DISABLE_TIMERQUERY: do not advertise or use GPU timer queries.
Bool DisableTimerQuery = false;
// MOBILEGL_USE_ANGLE: load ANGLE EGL/GLES libraries.
Bool UseAngle = false;
#if defined(MOBILEGL_TRACE_ANGLE_VARIANTS)
// MOBILEGL_TRACE_ANGLE_VARIANT: signed trace-APK ANGLE build short hash.
String TraceAngleVariant;
#endif
// MOBILEGL_DISABLE_SUBGROUP: force-disable Vulkan shader subgroup support.
Bool DisableSubgroup = false;
// MOBILEGL_MAGMA_R11G11B10F_FALLBACK: use fallback format for R11G11B10F on Vulkan.
Bool MagmaR11G11B10FFallback = false;
// MOBILEGL_MAGMA_FRAMESINFLIGHT: requested Magma frames in flight, defaulting to 3.
Uint32 MagmaFramesInFlight = 3;
// MOBILEGL_AVOID_SAMPLER_MIPMAP_MIN_FILTER: avoid mipmap min filters in samplers,
// resolves certain rendering bugs on ANGLE + llvmpipe.
Bool AvoidSamplerMipmapMinFilter = false;
// MOBILEGL_COHERENT_AS_FLUSH: app-compat for engines (e.g. Flywheel) that write
// GPU-read data through persistent GL_MAP_FLUSH_EXPLICIT_BIT maps they never
// flush. Persistent FLUSH_EXPLICIT map requests are rewritten to coherent
// semantics: writes reach the backend without glFlushMappedBufferRange, and
// flush calls on rewritten maps become error-free no-ops. Non-persistent maps
// keep spec FLUSH_EXPLICIT behavior.
Bool CoherentAsFlush = false;
// MOBILEGL_TRACE_SKIP_AUTODESTROY: skip teardown in the ELF destructor (Init.cpp).
Bool TraceSkipAutodestroy = false;
// MOBILEGL_DISABLE_UBO_RING: force the DirectGLES global-UBO upload back to the
// per-draw glBufferSubData path instead of the persistent-mapped ring allocator
// (negative control / driver-bug escape hatch).
Bool DisableUboRing = false;
// MOBILEGL_RELAXED_SEMANTICS: relax strict core-profile rules (e.g. VAO-0 draws,
// texture-name reuse after delete) even on contexts that explicitly requested a core
// profile. Without it, relaxed semantics still apply to every context that did not
// explicitly request a core profile via EGL_CONTEXT_OPENGL_PROFILE_MASK / a >=3.1
// version request.
Bool RelaxedSemantics = false;
// MOBILEGL_QUIRK_SUBGROUP_PREFIX_SCAN: overrides the shader-source quirk that
// 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
// cross-pipeline vertex position invariance. Sorted-transparency "over" blends,
// gl_FragDepth writers, and fully color-masked attachments are exempt (see
// PipelineFactory::ShouldSuppressDepthWrite). Auto detects Qualcomm.
QuirkOverride MagmaDisableBlendedDepthWriteQuirk = QuirkOverride::Auto;
// MOBILEGL_DISABLE_ROBUST_BUFFER_ACCESS: leave the Vulkan robustBufferAccess device
// feature off. It is enabled by default to match GL's defined out-of-range fetch
// behavior; this escape hatch exists to measure or dodge its GPU cost on a device.
Bool DisableRobustBufferAccess = false;
};
extern FeaturesTable Features;
} // namespace MobileGL::MG_Config
-170
View File
@@ -1,170 +0,0 @@
// MobileGL - MobileGL/ConfigLoader.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 "Config.h"
#include <cerrno>
#include <cstdlib>
#ifndef _WIN32
extern char** environ;
#endif
namespace MobileGL::MG_Config {
// Zero/default-initialized at static-init time (all fields have constexpr-friendly
// defaults), so it is safe to read even if MG_ConfigLoader::Init has not run yet.
FeaturesTable Features;
} // namespace MobileGL::MG_Config
namespace MobileGL::MG_ConfigLoader {
static UniquePtr<UnorderedMap<String, String>> acceptedEnvVariablesMap;
static Bool IsAcceptedPrefix(const String& key) {
return (key.compare(0, 6, "LIBGL_") == 0 || key.compare(0, 9, "MOBILEGL_") == 0);
}
inline void InitializeAcceptedEnvVariables() {
if (!acceptedEnvVariablesMap) {
acceptedEnvVariablesMap = MakeUnique<UnorderedMap<String, String>>();
} else {
acceptedEnvVariablesMap->clear();
}
char** envPtr = nullptr;
#ifdef _WIN32
envPtr = _environ;
#else // POSIX
envPtr = ::environ;
#endif
if (envPtr == nullptr) return;
for (char** env = envPtr; *env != nullptr; ++env) {
String entry(*env);
SizeT pos = entry.find('=');
if (pos != String::npos) {
String key = entry.substr(0, pos);
String value = entry.substr(pos + 1);
if (IsAcceptedPrefix(key)) {
(*acceptedEnvVariablesMap)[key] = value;
MGLOG_D("Config: Accepted env variable: %s=%s", key.c_str(), value.c_str());
}
}
}
}
inline void QueryEnvVariable(const String& key, String& outValue, const String& defaultValue) {
auto it = acceptedEnvVariablesMap->find(key);
if (it != acceptedEnvVariablesMap->end()) {
outValue = it->second;
} else {
outValue = defaultValue;
}
}
// Unified truthy rule for boolean feature env variables: set, non-empty, not "0",
// and not "false" (case-insensitive).
static Bool IsTruthyValue(const String& value) {
if (value.empty() || value == "0") {
return false;
}
String lowered = value;
std::transform(lowered.begin(), lowered.end(), lowered.begin(),
[](unsigned char c) { return static_cast<char>(std::tolower(c)); });
return lowered != "false";
}
inline Bool QueryEnvFlag(const String& key) {
auto it = acceptedEnvVariablesMap->find(key);
return it != acceptedEnvVariablesMap->end() && IsTruthyValue(it->second);
}
// Quirk overrides are tri-state: an unset variable keeps device auto-detection, a truthy
// value forces the quirk on, anything else set ("0", "false", "") forces it off.
inline MG_Config::QuirkOverride QueryEnvQuirkOverride(const String& key) {
auto it = acceptedEnvVariablesMap->find(key);
if (it == acceptedEnvVariablesMap->end()) {
return MG_Config::QuirkOverride::Auto;
}
return IsTruthyValue(it->second) ? MG_Config::QuirkOverride::ForceOn
: MG_Config::QuirkOverride::ForceOff;
}
inline Uint32 QueryEnvUint32(const String& key, Uint32 defaultValue, Uint32 minValue, Uint32 maxValue) {
auto it = acceptedEnvVariablesMap->find(key);
if (it == acceptedEnvVariablesMap->end()) {
return defaultValue;
}
const String& value = it->second;
char* parseEnd = nullptr;
errno = 0;
const unsigned long parsedValue = std::strtoul(value.c_str(), &parseEnd, 10);
if (parseEnd == value.c_str() || *parseEnd != '\0' || errno == ERANGE || parsedValue < minValue ||
parsedValue > maxValue) {
MGLOG_W("Config: Ignoring invalid env variable %s='%s'; expected an integer in range [%u, %u], "
"using default %u",
key.c_str(), value.c_str(), minValue, maxValue, defaultValue);
return defaultValue;
}
return static_cast<Uint32>(parsedValue);
}
inline void InitFeatures() {
auto& features = MG_Config::Features;
features.DisableTimerQuery = QueryEnvFlag("MOBILEGL_DISABLE_TIMERQUERY");
features.UseAngle = QueryEnvFlag("MOBILEGL_USE_ANGLE");
#if defined(MOBILEGL_TRACE_ANGLE_VARIANTS)
QueryEnvVariable("MOBILEGL_TRACE_ANGLE_VARIANT", features.TraceAngleVariant, "");
#endif
features.DisableSubgroup = QueryEnvFlag("MOBILEGL_DISABLE_SUBGROUP");
features.MagmaR11G11B10FFallback = QueryEnvFlag("MOBILEGL_MAGMA_R11G11B10F_FALLBACK");
features.MagmaFramesInFlight = QueryEnvUint32("MOBILEGL_MAGMA_FRAMESINFLIGHT", 3, 1, 64);
features.AvoidSamplerMipmapMinFilter =
QueryEnvFlag("MOBILEGL_AVOID_SAMPLER_MIPMAP_MIN_FILTER");
features.CoherentAsFlush = QueryEnvFlag("MOBILEGL_COHERENT_AS_FLUSH");
features.TraceSkipAutodestroy = QueryEnvFlag("MOBILEGL_TRACE_SKIP_AUTODESTROY");
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");
}
inline void InitBackendType() {
String backendTypeStr;
QueryEnvVariable("MOBILEGL_BACKEND_TYPE", backendTypeStr, "DirectGLES");
#define ENTRY(backendType) \
if (backendTypeStr == #backendType) { \
MG_Config::ActiveBackendType = BackendType::backendType; \
MGLOG_I("Config: Active backend type set to " #backendType); \
return; \
}
ENTRY(DirectGLES)
ENTRY(DirectVulkan)
ENTRY(Unknown)
MG_Config::ActiveBackendType = BackendType::Unknown;
#undef ENTRY
}
void Init() {
MGLOG_D("Loading configuration from environment variables...");
InitializeAcceptedEnvVariables();
InitBackendType();
InitFeatures();
// Destroy the map since we won't need it anymore
acceptedEnvVariablesMap.reset();
}
} // namespace MobileGL::MG_ConfigLoader
+8 -17
View File
@@ -32,14 +32,9 @@
#define MOBILEGL_GLX_API MOBILEGL_API
#define MOBILEGL_GL_API MOBILEGL_API
#define MOBILEGL_EGL_API MOBILEGL_API
#define MOBILEGL_CGL_API MOBILEGL_API
#define MOBILEGL_NSOPENGL_API MOBILEGL_API
#define MOBILEGL_WGL_API MOBILEGL_API
// ====================== MobileGL configurations ======================= //
#ifndef MOBILEGL_LOG_ACTIVE_LEVEL
#define MOBILEGL_LOG_ACTIVE_LEVEL MOBILEGL_LOG_LEVEL_INFO
#endif
#define MOBILEGL_LOG_ENABLE_CONSOLE 0
#define MOBILEGL_LOG_ENABLE_FILE 1
@@ -68,15 +63,11 @@
#endif
// =============================== Utils ================================ //
#if MOBILEGL_LOG_ACTIVE_LEVEL <= MOBILEGL_LOG_LEVEL_DEBUG
#define MOBILEGL_ASSERT(condition, ...) \
do { \
if (!(condition)) { \
MGLOG_F("Assertion failed" __VA_OPT__(": ") __VA_ARGS__); \
MGLOG_F(" at %s:%d (%s)", __FILE__, __LINE__, __func__); \
TRAP; \
} \
} while (0)
#else
#define MOBILEGL_ASSERT(condition, ...)
#endif
#define MOBILEGL_ASSERT(condition, ...) \
do { \
if (!(condition)) { \
MGLOG_F("Assertion failed" __VA_OPT__(": ") __VA_ARGS__); \
MGLOG_F(" at %s:%d (%s)", __FILE__, __LINE__, __func__); \
TRAP; \
} \
} while (0)
+2 -8
View File
@@ -10,17 +10,11 @@
namespace MobileGL {
namespace MG_Config {
BackendType ActiveBackendType;
BackendType ActiveBackendType = BackendType::DirectVulkanTMP;
} // namespace MG_Config
namespace MG_Backend {
// Leak-at-exit storage: the UniquePtr itself lives on the heap and is
// never destroyed by the runtime, so process exit runs no backend
// destructors (static destruction order across TUs is undefined).
// Deterministic teardown happens inside the EGL lifecycle instead:
// the last eglTerminate calls MobileGL::Destroy(), which .reset()s
// these singletons while the process is still healthy.
UniquePtr<BackendObject>& pActiveBackendObject = *new UniquePtr<BackendObject>();
UniquePtr<BackendObject> pActiveBackendObject;
GlobalBackendFunctionsTable gBackendFunctionsTable;
} // namespace MG_Backend
} // namespace MobileGL
+2 -39
View File
@@ -32,14 +32,13 @@
#include <thread>
#include <vector>
#include <cassert>
#include <climits>
#include <cstdlib>
#include <cstdarg>
#include <cstring>
#include <numeric>
#include <expected>
#include <iostream>
#include <optional>
#include <xxhash.h>
#include <algorithm>
#include <stdexcept>
#include <functional>
@@ -52,9 +51,6 @@
// Include FastSTL
#include <FastSTL/UnorderedMap.h>
// Include xxHash
#include <xxhash.h>
// Include spirv_cross
#include <spirv_cross/spirv_cross_c.h>
@@ -64,9 +60,7 @@
#endif
#include <EGL/egl.h>
#define GL_GLEXT_PROTOTYPES
#ifndef NO_GL_H
#include "GL/gl.h"
#endif
#include <GL/glcorearb.h>
#undef GL_GLEXT_PROTOTYPES
@@ -98,44 +92,13 @@
#endif
#ifdef __ANDROID__
#define VK_USE_PLATFORM_ANDROID_KHR
#include <unistd.h>
#include <pthread.h>
#include <android/log.h>
#include <android/native_window.h>
#endif
#ifdef __ANDROID__
#define VK_USE_PLATFORM_ANDROID_KHR
#elif _WIN32
#define VK_USE_PLATFORM_WIN32_KHR
#elif defined(__APPLE__)
#define VK_USE_PLATFORM_METAL_EXT
#elif defined(__linux__)
#define VK_USE_PLATFORM_XLIB_KHR
typedef struct _XDisplay Display;
typedef unsigned long XID;
typedef XID Window;
typedef unsigned long VisualID;
#else
#warning "VK_USE_PLATFORM_*_KHR not defined for this platform!"
#endif
#if defined(VK_USE_PLATFORM_XLIB_KHR)
#pragma push_macro("Bool")
#pragma push_macro("None")
#pragma push_macro("Always")
#pragma push_macro("Status")
#pragma push_macro("LSBFirst")
#pragma push_macro("DestroyAll")
#endif
#include <vulkan/vulkan.h>
#if defined(VK_USE_PLATFORM_XLIB_KHR)
#pragma pop_macro("DestroyAll")
#pragma pop_macro("LSBFirst")
#pragma pop_macro("Status")
#pragma pop_macro("Always")
#pragma pop_macro("None")
#pragma pop_macro("Bool")
#endif
#ifdef TRACY_ENABLE
#include <tracy/Tracy.hpp>
+36 -77
View File
@@ -8,101 +8,60 @@
#include "Init.h"
#include "Config.h"
#include <MG_Impl/Init.h>
#include <MG_Backend/BackendObjects.h>
#include <MG_Backend/DirectVulkan/DirectVulkan.h>
#include <MG_State/GLState/Core.h>
#include <MG_State/EGLState/Core.h>
#include <MG_Impl/GLImpl/Texture/ProxyTexture.h>
#include <MG_Impl/GLImpl/Framebuffer/GL_Framebuffer.h>
#include <atomic>
#include <mutex>
namespace MobileGL {
namespace {
std::atomic<Bool> g_isInitialized = false;
thread_local Bool tl_initializing = false;
std::mutex& InitMutex() {
static std::mutex mutex;
return mutex;
}
void DestroyImpl(Bool logLifecycle) {
if (!g_isInitialized) {
return;
}
if (logLifecycle) {
MGLOG_I("MobileGL closing...");
}
glslang::FinalizeProcess();
MG_Backend::pActiveBackendObject.reset();
MG_State::pGLContext.reset();
MG_State::pEGLContext.reset();
MG_Impl::GLImpl::TextureImpl::pProxyTextureManager.reset();
MG_Impl::GLImpl::FramebufferImpl::pDefaultFramebufferInfo.reset();
MG_Backend::gBackendFunctionsTable = {};
g_isInitialized = false;
if (logLifecycle) {
MG_Util::Debug::Close();
}
// TODO: add and use Destroy functions for other subsystems
}
}
void Initialize() {
if (g_isInitialized) {
MGLOG_D("MobileGL already initialized; skipping duplicate Initialize()");
return;
}
void MG_Initialize() {
MG_Util::Debug::InitFile();
MGLOG_I("Initializing MobileGL...");
MG_ConfigLoader::Init();
MGLOG_I("Config loaded");
MG_State::Init();
MGLOG_D("MG_State initialized");
MGLOG_D("MobileGL State initialized");
MG_Backend::Init();
MGLOG_D("MG_Backend initialized");
MGLOG_D("MobileGL Backend initialized");
MG_Impl::Init();
MGLOG_D("MG_Impl initialized");
MGLOG_D("MobileGL Implementation initialized");
glslang::InitializeProcess();
MGLOG_D("glslang initialized");
g_isInitialized = true;
MGLOG_I("MobileGL initialized");
}
void EnsureInitialized() {
if (g_isInitialized.load(std::memory_order_acquire)) {
return;
}
// Re-entrant call while this thread is already inside Initialize()
// (e.g. an init step routing back through a public entry point).
if (tl_initializing) {
return;
}
const std::lock_guard<std::mutex> lock(InitMutex());
if (g_isInitialized.load(std::memory_order_acquire)) {
return;
}
tl_initializing = true;
Initialize();
tl_initializing = false;
void MG_Destroy() {
MGLOG_I("MobileGL closing...");
glslang::FinalizeProcess();
delete MG_State::pGLContext;
delete MG_Impl::GLImpl::TextureImpl::pProxyTextureManager;
delete MG_Impl::GLImpl::FramebufferImpl::pDefaultFramebufferInfo;
MG_Util::Debug::Close();
// TODO: add and use Destroy functions for other subsystems
}
void Destroy() {
DestroyImpl(true);
#if defined(__linux__) || defined(__APPLE__)
__attribute__((constructor)) static void AutoInit() {
MG_Initialize();
}
// MobileGL's lifecycle is owned entirely by the host-API layers
// (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 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.
__attribute__((destructor)) static void AutoDestroy() {
MG_Destroy();
}
#endif
#ifdef _WIN32
BOOL WINAPI DllMain(HMODULE hModule, DWORD ul_reason_for_call, LPVOID lpReserved) {
switch (ul_reason_for_call) {
case DLL_PROCESS_ATTACH:
MG_Initialize();
break;
case DLL_PROCESS_DETACH:
MG_Destroy();
break;
}
return TRUE;
}
#endif
} // namespace MobileGL
+3 -25
View File
@@ -10,28 +10,6 @@
#include "Includes.h"
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
// 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();
namespace MG_Util::Debug {
void InitFile();
} // namespace MG_Util::Debug
namespace MG_ConfigLoader {
void Init();
} // namespace MG_ConfigLoader
namespace MG_Backend {
void Init();
} // namespace MG_Backend
namespace MG_Impl {
void Init();
} // namespace MG_Impl
} // namespace MobileGL
void MG_Initialize();
void MG_Destroy();
} // namespace MobileGL
-480
View File
@@ -7,489 +7,9 @@
// End of Source File Header
#include "BackendObject.h"
#include "MG_Util/Converters/MGToStr/TextureEnumConverter.h"
#include <algorithm>
#include <cstring>
#include <iomanip>
#include <sstream>
namespace MobileGL::MG_Backend {
namespace {
Bool IsReleaseCurrentRequest(EGLDisplay dpy, EGLSurface draw, EGLSurface read, EGLContext ctx) {
(void)dpy;
return draw == EGL_NO_SURFACE && read == EGL_NO_SURFACE && ctx == EGL_NO_CONTEXT;
}
std::thread::id CurrentThreadKey() {
return std::this_thread::get_id();
}
const char* GetFormatCapabilitySupportString(const FormatCapabilityCache& cache,
SizeT targetIndex,
SizeT formatIndex,
FormatCapability capability) {
if (HasFormatCapability(cache.FullCaps[targetIndex][formatIndex], capability)) return "Full";
if (HasFormatCapability(cache.CaveatCaps[targetIndex][formatIndex], capability)) return "Caveat";
return "None";
}
SizeT GetPrintedFormatNameWidth() {
SizeT width = 0;
for (SizeT formatIndex = 0; formatIndex < kFormatCapabilityFormatCount; ++formatIndex) {
const auto format = static_cast<TextureInternalFormat>(formatIndex);
width = std::max(width, MG_Util::ConvertTextureInternalFormatToString(format).size());
}
return width;
}
SizeT GetCapabilityColumnWidth(FormatCapability capability) {
SizeT width = std::strlen(GetFormatCapabilityName(capability));
width = std::max<SizeT>(width, std::strlen("Caveat"));
return width;
}
String BuildFormatCapabilityHeader(SizeT formatNameWidth) {
std::ostringstream line;
line << std::left << std::setw(static_cast<Int>(formatNameWidth)) << "";
for (FormatCapability capability : kReportedFormatCapabilities) {
line << " | " << std::left << std::setw(static_cast<Int>(GetCapabilityColumnWidth(capability)))
<< GetFormatCapabilityName(capability);
}
return line.str();
}
String BuildFormatCapabilityRow(const FormatCapabilityCache& cache,
SizeT targetIndex,
SizeT formatIndex,
SizeT formatNameWidth) {
const auto format = static_cast<TextureInternalFormat>(formatIndex);
std::ostringstream line;
line << std::left << std::setw(static_cast<Int>(formatNameWidth))
<< MG_Util::ConvertTextureInternalFormatToString(format);
for (FormatCapability capability : kReportedFormatCapabilities) {
line << " | " << std::left << std::setw(static_cast<Int>(GetCapabilityColumnWidth(capability)))
<< GetFormatCapabilitySupportString(cache, targetIndex, formatIndex, capability);
}
return line.str();
}
} // namespace
void FormatCapabilityCache::Clear() {
for (auto& row : FullCaps) {
row.fill(FormatCapabilityFlags{});
}
for (auto& row : CaveatCaps) {
row.fill(FormatCapabilityFlags{});
}
for (auto& row : SampleCounts) {
for (auto& counts : row) {
counts.clear();
}
}
}
Bool HasFormatCapability(FormatCapabilityFlags caps, FormatCapability capability) {
return static_cast<Bool>(caps & capability);
}
SizeT GetFormatCapabilityTargetIndex(TextureTarget target) {
if (target == TextureTarget::Unknown || static_cast<Int>(target) < 0 ||
static_cast<SizeT>(target) >= kFormatCapabilityTextureTargetCount) {
return kFormatCapabilityTargetCount;
}
return static_cast<SizeT>(target);
}
SizeT GetRenderbufferFormatCapabilityTargetIndex() {
return kFormatCapabilityRenderbufferTargetIndex;
}
const char* GetFormatCapabilityName(FormatCapability capability) {
switch (capability) {
case FormatCapability::Creatable:
return "Creatable";
case FormatCapability::Sampled:
return "Sampled";
case FormatCapability::LinearFilter:
return "LinearFilter";
case FormatCapability::GenerateMipmap:
return "GenerateMipmap";
case FormatCapability::TextureGather:
return "TextureGather";
case FormatCapability::TextureShadow:
return "TextureShadow";
case FormatCapability::FramebufferRenderable:
return "FramebufferRenderable";
case FormatCapability::FramebufferLayered:
return "FramebufferLayered";
case FormatCapability::MultisampleTexture:
return "MultisampleTexture";
case FormatCapability::MultisampleRenderbuffer:
return "MultisampleRenderbuffer";
case FormatCapability::ColorAttachment:
return "ColorAttachment";
case FormatCapability::DepthAttachment:
return "DepthAttachment";
case FormatCapability::StencilAttachment:
return "StencilAttachment";
case FormatCapability::TextureBuffer:
return "TextureBuffer";
}
return "Unknown";
}
String GetFormatCapabilityTargetName(SizeT targetIndex) {
if (targetIndex == kFormatCapabilityRenderbufferTargetIndex) {
return "Renderbuffer";
}
if (targetIndex >= kFormatCapabilityTextureTargetCount) {
return "Unknown";
}
return MG_Util::ConvertTextureTargetToString(static_cast<TextureTarget>(targetIndex));
}
void PrintFormatCapabilities(const FormatCapabilityCache& cache) {
const SizeT formatNameWidth = GetPrintedFormatNameWidth();
MGLOG_D("Backend format capabilities:");
for (SizeT targetIndex = 0; targetIndex < kFormatCapabilityTargetCount; ++targetIndex) {
MGLOG_D("");
const String targetName = GetFormatCapabilityTargetName(targetIndex);
MGLOG_D("- %s", targetName.c_str());
const String header = BuildFormatCapabilityHeader(formatNameWidth);
MGLOG_D("%s", header.c_str());
for (SizeT formatIndex = 0; formatIndex < kFormatCapabilityFormatCount; ++formatIndex) {
const String row = BuildFormatCapabilityRow(cache, targetIndex, formatIndex, formatNameWidth);
MGLOG_D("%s", row.c_str());
}
}
}
Bool BackendObject::InitializeEGLDisplay(EGLDisplay dpy, EGLint* major, EGLint* minor) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
if (dpy == EGL_NO_DISPLAY) {
MGLOG_E("InitializeEGLDisplay failed: invalid EGLDisplay");
return false;
}
if (m_eglDisplayInitialized && m_eglDisplay != dpy) {
MGLOG_E("InitializeEGLDisplay failed: backend already bound to a different EGLDisplay");
return false;
}
m_eglDisplay = dpy;
m_eglDisplayInitialized = true;
if (major) {
*major = 1;
}
if (minor) {
*minor = 5;
}
return true;
}
Bool BackendObject::CreateEGLWindowSurface(EGLSurface surface, const WindowHandle& handle) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
return RegisterEGLWindowSurface(surface, handle) && ActivateEGLSurface(surface);
}
Bool BackendObject::ResizeEGLWindowSurface(EGLSurface surface, Uint32 width, Uint32 height) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
auto surfaceIt = m_eglSurfaces.find(surface);
if (surfaceIt == m_eglSurfaces.end() || surfaceIt->second.Kind != SurfaceKind::Window) {
MGLOG_E("ResizeEGLWindowSurface failed: no window surface is initialized");
return false;
}
surfaceIt->second.Window.Width = width;
surfaceIt->second.Window.Height = height;
if (m_eglSurface == surface) {
m_windowHandle.Width = width;
m_windowHandle.Height = height;
}
return true;
}
Bool BackendObject::CreateEGLPbufferSurface(EGLSurface surface, EGLint width, EGLint height) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
return RegisterEGLPbufferSurface(surface, width, height) && ActivateEGLSurface(surface);
}
Bool BackendObject::RegisterEGLWindowSurface(EGLSurface surface, const WindowHandle& handle) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
if (!m_eglDisplayInitialized) {
MGLOG_E("RegisterEGLWindowSurface failed: EGL display is not initialized");
return false;
}
if (surface == EGL_NO_SURFACE) {
MGLOG_E("RegisterEGLWindowSurface failed: invalid EGLSurface");
return false;
}
if (handle.Backend == WindowBackend::Unknown || !handle.Handle) {
MGLOG_E("RegisterEGLWindowSurface failed: invalid native window handle");
return false;
}
auto& state = m_eglSurfaces[surface];
state = EGLSurfaceState{
.Kind = SurfaceKind::Window,
.Window = handle,
.Width = static_cast<EGLint>(std::max<Uint32>(handle.Width, 1)),
.Height = static_cast<EGLint>(std::max<Uint32>(handle.Height, 1)),
};
return true;
}
Bool BackendObject::RegisterEGLPbufferSurface(EGLSurface surface, EGLint width, EGLint height) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
if (!m_eglDisplayInitialized) {
MGLOG_E("RegisterEGLPbufferSurface failed: EGL display is not initialized");
return false;
}
if (surface == EGL_NO_SURFACE) {
MGLOG_E("RegisterEGLPbufferSurface failed: invalid EGLSurface");
return false;
}
if (width <= 0 || height <= 0) {
MGLOG_E("RegisterEGLPbufferSurface failed: invalid size %dx%d", width, height);
return false;
}
m_eglSurfaces[surface] = EGLSurfaceState{
.Kind = SurfaceKind::Pbuffer,
.Width = width,
.Height = height,
};
return true;
}
const BackendObject::EGLSurfaceState* BackendObject::GetRegisteredEGLSurface(EGLSurface surface) const {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
auto surfaceIt = m_eglSurfaces.find(surface);
return surfaceIt == m_eglSurfaces.end() ? nullptr : &surfaceIt->second;
}
Bool BackendObject::ActivateEGLSurface(EGLSurface surface) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
const auto* surfaceState = GetRegisteredEGLSurface(surface);
if (!surfaceState) {
MGLOG_E("ActivateEGLSurface failed: EGL surface is not registered");
return false;
}
if (m_eglSurfaceInitialized && m_eglSurface == surface) {
return true;
}
if (surfaceState->Kind == SurfaceKind::Window) {
SetWindowHandle(surfaceState->Window);
if (!InitWindowSurface()) {
MGLOG_E("ActivateEGLSurface failed: backend InitWindowSurface failed");
return false;
}
} else if (surfaceState->Kind == SurfaceKind::Pbuffer) {
if (!InitPbufferSurface(surfaceState->Width, surfaceState->Height)) {
MGLOG_E("ActivateEGLSurface failed: backend InitPbufferSurface failed");
return false;
}
} else {
MGLOG_E("ActivateEGLSurface failed: unsupported surface kind");
return false;
}
m_eglSurface = surface;
m_eglSurfaceInitialized = true;
m_eglSurfaceKind = surfaceState->Kind;
m_eglCurrentThreads.clear();
m_backendCapabilitiesInitialized = false;
return true;
}
Bool BackendObject::MakeEGLCurrent(EGLDisplay dpy, EGLSurface draw, EGLSurface read, EGLContext ctx) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
const auto threadKey = CurrentThreadKey();
if (IsReleaseCurrentRequest(dpy, draw, read, ctx)) {
ReleaseEGLCurrentThread(threadKey);
return true;
}
if (!m_eglDisplayInitialized || m_eglDisplay != dpy) {
MGLOG_E("MakeEGLCurrent failed: EGL display mismatch or not initialized");
return false;
}
if (!m_eglSurfaceInitialized) {
if (draw != read || !ActivateEGLSurface(draw)) {
MGLOG_E("MakeEGLCurrent failed: EGL surface is not initialized");
return false;
}
}
if (!GetRegisteredEGLSurface(draw) || !GetRegisteredEGLSurface(read)) {
MGLOG_E("MakeEGLCurrent failed: EGL surface is not registered");
return false;
}
if (draw != read) {
MGLOG_E("MakeEGLCurrent failed: separate draw/read surfaces are not supported");
return false;
}
if (draw != m_eglSurface && !ActivateEGLSurface(draw)) {
MGLOG_E("MakeEGLCurrent failed: EGL surface is not backed by this backend");
return false;
}
if (draw == EGL_NO_SURFACE || read == EGL_NO_SURFACE || ctx == EGL_NO_CONTEXT) {
MGLOG_E("MakeEGLCurrent failed: draw/read/context is invalid");
return false;
}
if (!m_backendCapabilitiesInitialized) {
if (!InitCapabilities()) {
MGLOG_E("MakeEGLCurrent failed: InitCapabilities failed");
return false;
}
m_backendCapabilitiesInitialized = true;
}
ReleaseEGLCurrentThread(threadKey);
m_eglCurrentThreads[threadKey] = EGLCurrentState{
.Display = dpy,
.DrawSurface = draw,
.ReadSurface = read,
.Context = ctx,
};
return true;
}
void BackendObject::ResetEGLRuntimeState() {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
m_eglSurfaceInitialized = false;
m_backendCapabilitiesInitialized = false;
m_eglSurfaceKind = SurfaceKind::None;
m_eglSurface = EGL_NO_SURFACE;
m_windowHandle = {};
m_eglCurrentThreads.clear();
}
Bool BackendObject::SwapEGLBuffers(EGLDisplay dpy, EGLSurface draw) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
if (!m_eglDisplayInitialized || m_eglDisplay != dpy) {
MGLOG_E("SwapEGLBuffers failed: EGL display mismatch or not initialized");
return false;
}
const auto currentIt = m_eglCurrentThreads.find(CurrentThreadKey());
if (currentIt == m_eglCurrentThreads.end()) {
MGLOG_E("SwapEGLBuffers failed: no current context attached");
return false;
}
if (currentIt->second.Display != dpy || currentIt->second.DrawSurface != draw ||
currentIt->second.Context == EGL_NO_CONTEXT) {
MGLOG_E("SwapEGLBuffers failed: draw surface is not current on this thread");
return false;
}
if (!m_eglSurfaceInitialized || draw == EGL_NO_SURFACE || draw != m_eglSurface) {
MGLOG_E("SwapEGLBuffers failed: invalid draw surface");
return false;
}
const auto& backendFunctions = GetBackendFunctions();
if (!backendFunctions.Present) {
MGLOG_E("SwapEGLBuffers failed: backend Present function is null");
return false;
}
backendFunctions.Present();
return true;
}
void BackendObject::SetEGLSwapInterval(Int interval) {
const auto& backendFunctions = GetBackendFunctions();
if (backendFunctions.SetSwapInterval) {
backendFunctions.SetSwapInterval(interval);
}
}
Bool BackendObject::IsEGLSurfaceCurrent(EGLSurface surface) const {
if (surface == EGL_NO_SURFACE) {
return false;
}
for (const auto& current : m_eglCurrentThreads) {
if (current.second.DrawSurface == surface || current.second.ReadSurface == surface) {
return true;
}
}
return false;
}
void BackendObject::DestroyPendingEGLSurfaceIfUnused(EGLSurface surface) {
auto surfaceIt = m_eglSurfaces.find(surface);
if (surfaceIt == m_eglSurfaces.end() || !surfaceIt->second.DestroyPending ||
IsEGLSurfaceCurrent(surface)) {
return;
}
m_eglSurfaces.erase(surfaceIt);
if (m_eglSurface == surface) {
OnEGLSurfaceReleased(surface);
ResetEGLRuntimeState();
}
}
void BackendObject::ReleaseEGLCurrentThread(const std::thread::id& threadKey) {
auto currentIt = m_eglCurrentThreads.find(threadKey);
if (currentIt == m_eglCurrentThreads.end()) {
return;
}
const EGLSurface drawSurface = currentIt->second.DrawSurface;
const EGLSurface readSurface = currentIt->second.ReadSurface;
m_eglCurrentThreads.erase(currentIt);
DestroyPendingEGLSurfaceIfUnused(drawSurface);
DestroyPendingEGLSurfaceIfUnused(readSurface);
}
void BackendObject::ReleaseEGLSurface(EGLSurface surface) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
auto surfaceIt = m_eglSurfaces.find(surface);
if (surfaceIt == m_eglSurfaces.end()) {
return;
}
if (IsEGLSurfaceCurrent(surface)) {
surfaceIt->second.DestroyPending = true;
return;
}
m_eglSurfaces.erase(surfaceIt);
if (m_eglSurface == surface) {
OnEGLSurfaceReleased(surface);
ResetEGLRuntimeState();
}
}
void BackendObject::ReleaseEGLResources() {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
ResetEGLRuntimeState();
m_eglSurfaces.clear();
m_eglDisplay = EGL_NO_DISPLAY;
m_eglDisplayInitialized = false;
}
void BackendObject::SetWindowHandle(const WindowHandle& handle) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
m_windowHandle = handle;
}
const FormatCapabilityCache& BackendObject::GetFormatCapabilities() const {
return m_formatCapabilities;
}
FormatCapabilityCache& BackendObject::MutableFormatCapabilities() {
return m_formatCapabilities;
}
Bool BackendObject::InitPbufferSurface(EGLint width, EGLint height) {
(void)width;
(void)height;
return false;
}
void BackendObject::OnEGLSurfaceReleased(EGLSurface surface) {
(void)surface;
}
} // namespace MobileGL::MG_Backend
+4 -302
View File
@@ -8,104 +8,21 @@
#pragma once
#include <Includes.h>
#include "MG_State/GLState/TextureState/TextureEnum.h"
namespace MobileGL {
namespace MG_State::GLState {
class FramebufferObject;
class ITextureObject;
}
enum class BackendType {
DirectGLES,
DirectVulkan,
DirectVulkanTMP,
BackendTypeCount,
Unknown = -1
};
namespace MG_Backend {
enum class FormatCapability : Uint64 {
Creatable = 1ull << 0,
Sampled = 1ull << 1,
LinearFilter = 1ull << 2,
GenerateMipmap = 1ull << 3,
TextureGather = 1ull << 4,
TextureShadow = 1ull << 5,
FramebufferRenderable = 1ull << 6,
FramebufferLayered = 1ull << 7,
MultisampleTexture = 1ull << 8,
MultisampleRenderbuffer = 1ull << 9,
ColorAttachment = 1ull << 10,
DepthAttachment = 1ull << 11,
StencilAttachment = 1ull << 12,
TextureBuffer = 1ull << 13
};
using FormatCapabilityFlags = Flags<FormatCapability>;
inline constexpr Array<FormatCapability, 14> kReportedFormatCapabilities = {
FormatCapability::Creatable,
FormatCapability::Sampled,
FormatCapability::LinearFilter,
FormatCapability::GenerateMipmap,
FormatCapability::TextureGather,
FormatCapability::TextureShadow,
FormatCapability::FramebufferRenderable,
FormatCapability::FramebufferLayered,
FormatCapability::MultisampleTexture,
FormatCapability::MultisampleRenderbuffer,
FormatCapability::ColorAttachment,
FormatCapability::DepthAttachment,
FormatCapability::StencilAttachment,
FormatCapability::TextureBuffer,
};
inline constexpr SizeT kFormatCapabilityTextureTargetCount =
static_cast<SizeT>(TextureTarget::TextureTargetCount);
inline constexpr SizeT kFormatCapabilityRenderbufferTargetIndex = kFormatCapabilityTextureTargetCount;
inline constexpr SizeT kFormatCapabilityTargetCount = kFormatCapabilityTextureTargetCount + 1;
inline constexpr SizeT kFormatCapabilityFormatCount =
static_cast<SizeT>(TextureInternalFormat::TextureInternalFormatCount);
using FormatCapabilityTable =
Array<Array<FormatCapabilityFlags, kFormatCapabilityFormatCount>, kFormatCapabilityTargetCount>;
using FormatSampleCountTable =
Array<Array<Vector<Int>, kFormatCapabilityFormatCount>, kFormatCapabilityTargetCount>;
struct FormatCapabilityCache {
FormatCapabilityTable FullCaps{};
FormatCapabilityTable CaveatCaps{};
FormatSampleCountTable SampleCounts{};
void Clear();
};
Bool HasFormatCapability(FormatCapabilityFlags caps, FormatCapability capability);
SizeT GetFormatCapabilityTargetIndex(TextureTarget target);
SizeT GetRenderbufferFormatCapabilityTargetIndex();
const char* GetFormatCapabilityName(FormatCapability capability);
String GetFormatCapabilityTargetName(SizeT targetIndex);
void PrintFormatCapabilities(const FormatCapabilityCache& cache);
// Opaque backend fence-sync handle, created by GLFunctionsTable::FenceSync
// and released by GLFunctionsTable::DeleteSync.
using BackendSyncHandle = void*;
// Opaque backend timer-query handle, created by
// GLFunctionsTable::BeginTimeElapsedQuery / QueryCounterTimestamp and
// released by GLFunctionsTable::DeleteBackendQuery.
using BackendQueryHandle = void*;
struct GLFunctionsTable {
void (*DrawArrays)(GLenum mode, GLint first, GLsizei count);
void (*DrawElements)(GLenum mode, GLsizei count, GLenum type, const void* indices);
void (*DrawElementsBaseVertex)(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLint basevertex);
void (*MultiDrawArrays)(GLenum mode, const GLint* first, const GLsizei* count, GLsizei drawcount);
void (*MultiDrawElements)(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount);
void (*MultiDrawElementsBaseVertex)(GLenum mode, const GLsizei* count, GLenum type,
@@ -114,10 +31,6 @@ namespace MobileGL {
void (*MultiDrawElementsIndirect)(GLenum mode, GLenum type, const void* indirect, GLsizei drawcount,
GLsizei stride);
void (*MultiDrawArraysIndirect)(GLenum mode, const void* indirect, GLsizei drawcount, GLsizei stride);
void (*MultiDrawElementsIndirectCount)(GLenum mode, GLenum type, const void* indirect,
GLintptr drawcount, GLsizei maxdrawcount, GLsizei stride);
void (*MultiDrawArraysIndirectCount)(GLenum mode, const void* indirect, GLintptr drawcount,
GLsizei maxdrawcount, GLsizei stride);
void (*DrawRangeElementsBaseVertex)(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type,
const void* indices, GLint basevertex);
void (*DrawRangeElements)(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type,
@@ -141,180 +54,29 @@ namespace MobileGL {
void (*ClearBufferfv)(GLenum buffer, GLint drawbuffer, const GLfloat* value);
void (*ClearBufferuiv)(GLenum buffer, GLint drawbuffer, const GLuint* value);
void (*ClearBufferiv)(GLenum buffer, GLint drawbuffer, const GLint* value);
void (*ClearNamedFramebufferfv)(const SharedPtr<MG_State::GLState::FramebufferObject>& framebuffer,
GLenum buffer, GLint drawbuffer, const GLfloat* value);
void (*ClearNamedFramebufferfi)(const SharedPtr<MG_State::GLState::FramebufferObject>& framebuffer,
GLenum buffer, GLint drawbuffer, GLfloat depth, GLint stencil);
void (*BlitFramebuffer)(GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1, GLint dstX0, GLint dstY0,
GLint dstX1, GLint dstY1, GLbitfield mask, GLenum filter);
void (*BlitNamedFramebuffer)(const SharedPtr<MG_State::GLState::FramebufferObject>& readFramebuffer,
const SharedPtr<MG_State::GLState::FramebufferObject>& drawFramebuffer,
GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1,
GLint dstX0, GLint dstY0, GLint dstX1, GLint dstY1,
GLbitfield mask, GLenum filter);
void (*CopyTexImage2D)(GLenum target, GLint level, GLenum internalformat, GLint x, GLint y, GLsizei width,
GLsizei height, GLint border);
void (*CopyTexSubImage2D)(GLenum target, GLint level, GLint xoffset, GLint yoffset, GLint x, GLint y,
GLsizei width, GLsizei height);
void (*CopyImageSubData)(const SharedPtr<MG_State::GLState::ITextureObject>& srcTexture,
GLenum srcTarget, GLint srcLevel, GLint srcX, GLint srcY, GLint srcZ,
const SharedPtr<MG_State::GLState::ITextureObject>& dstTexture,
GLenum dstTarget, GLint dstLevel, GLint dstX, GLint dstY, GLint dstZ,
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);
void (*GetTexImage)(GLenum target, GLint level, GLenum format, GLenum type, GLvoid* pixels);
void (*GetTextureImage)(const SharedPtr<MG_State::GLState::ITextureObject>& texture,
TextureUploadTarget uploadTarget, GLint level, GLenum format, GLenum type,
GLsizei bufSize, GLvoid* pixels);
void (*DispatchCompute)(GLuint numGroupsX, GLuint numGroupsY, GLuint numGroupsZ);
void (*DispatchComputeIndirect)(GLintptr indirect);
void (*MemoryBarrier)(GLbitfield barriers);
void (*MemoryBarrierByRegion)(GLbitfield barriers);
void (*BindImageTexture)(GLuint unit, GLuint texture, GLint level, GLboolean layered, GLint layer,
GLenum access, GLenum format);
void (*GetIntegeri_v)(GLenum target, GLuint index, GLint* data);
void (*GetInteger64i_v)(GLenum target, GLuint index, GLint64* data);
void (*GetProgramiv)(GLuint program, GLenum pname, GLint* params);
void (*GetProgramInterfaceiv)(GLuint program, GLenum programInterface, GLenum pname, GLint* params);
GLuint (*GetProgramResourceIndex)(GLuint program, GLenum programInterface, const GLchar* name);
void (*GetProgramResourceName)(GLuint program, GLenum programInterface, GLuint index, GLsizei bufSize,
GLsizei* length, GLchar* name);
void (*GetProgramResourceiv)(GLuint program, GLenum programInterface, GLuint index, GLsizei propCount,
const GLenum* props, GLsizei bufSize, GLsizei* length, GLint* params);
GLint (*GetProgramResourceLocation)(GLuint program, GLenum programInterface, const GLchar* name);
GLint (*GetProgramResourceLocationIndex)(GLuint program, GLenum programInterface, const GLchar* name);
void (*ShaderStorageBlockBinding)(GLuint program, GLuint storageBlockIndex, GLuint storageBlockBinding);
// GL fence sync objects. All entries are optional (may be null); the
// frontend then falls back to always-signaled sync semantics.
// FenceSync may itself return null when the backend cannot create a
// fence right now (e.g. the calling thread does not own the backend
// context); the frontend treats such a sync as always signaled.
BackendSyncHandle (*FenceSync)();
GLenum (*ClientWaitSync)(BackendSyncHandle sync, GLbitfield flags, GLuint64 timeout);
void (*WaitSync)(BackendSyncHandle sync, GLbitfield flags, GLuint64 timeout);
void (*DeleteSync)(BackendSyncHandle sync);
Bool (*GetSyncStatus)(BackendSyncHandle sync); // true = signaled
// GL timer-query objects (GL_ARB_timer_query). All entries are
// optional (may be null); the frontend then falls back to zero
// results and reports GL_QUERY_COUNTER_BITS == 0.
// BeginTimeElapsedQuery / QueryCounterTimestamp may themselves
// return null when the backend cannot create a query right now;
// the frontend treats such a query as immediately available with
// a zero result.
// Dynamic support check: true only when the live backend can
// actually time at the moment of the call (extension / entry
// points / timestamp valid bits are known then, not at table
// init). Gates the advertised GL_QUERY_COUNTER_BITS.
Bool (*IsTimerQuerySupported)();
BackendQueryHandle (*BeginTimeElapsedQuery)(); // starts a TIME_ELAPSED span
void (*EndTimeElapsedQuery)(BackendQueryHandle query); // ends the span
BackendQueryHandle (*QueryCounterTimestamp)(); // glQueryCounter(GL_TIMESTAMP) one-shot
Bool (*IsQueryResultAvailable)(BackendQueryHandle query); // non-blocking
// Returns true when a final value was produced (*outNanoseconds
// written; the frontend may cache it and release the handle).
// Returns false when the result could not be obtained YET - e.g.
// a Vulkan wait that refuses to block on a not-yet-submitted
// frame serial - in which case the frontend must keep the handle
// and leave the query readable later.
Bool (*GetQueryResult64)(BackendQueryHandle query, Bool wait, Uint64* outNanoseconds);
void (*DeleteBackendQuery)(BackendQueryHandle query);
Int64 (*GetGpuTimestampNs)(); // glGetInteger64v(GL_TIMESTAMP); 0 if unsupported
};
struct GlobalBackendFunctionsTable {
GLFunctionsTable GL;
void (*Present)();
// Optional: applies the app-requested eglSwapInterval to the native
// presentation path (null = backend keeps its own pacing policy).
void (*SetSwapInterval)(Int interval);
};
// Coarse GPU vendor identity for gating device-specific quirks. Detected from the
// Vulkan physical-device vendorID or the GLES GL_VENDOR/GL_RENDERER strings; stays
// Unknown when detection is inconclusive, in which case auto-gated quirks stay off.
enum class GpuVendorKind : Uint8 {
Unknown = 0,
Qualcomm,
Arm,
Nvidia,
Amd,
Intel,
ImgTec,
// Software rasterizers (llvmpipe/lavapipe, SwiftShader).
Software,
};
struct DynamicBackendParameters {
SizeT UniformBufferOffsetAlignment = 256;
// GL_MAX_TEXTURE_MAX_ANISOTROPY_EXT. 1.0 means the backend cannot filter anisotropically,
// which is also why the extension is not advertised in that case.
Float MaxTextureMaxAnisotropy = 1.0f;
Float AliasedLineWidthRangeMin = 1.0f;
Float AliasedLineWidthRangeMax = 1.0f;
Float SmoothLineWidthRangeMin = 1.0f;
Float SmoothLineWidthRangeMax = 1.0f;
Float SmoothLineWidthGranularity = 1.0f;
Float PointSizeRangeMin = 1.0f;
Float PointSizeRangeMax = 1.0f;
Float PointSizeGranularity = 1.0f;
Int Max3DTextureSize = 16384;
Int MaxArrayTextureLayers = 2048;
Int MaxCubeMapTextureSize = 16384;
Int MaxFramebufferWidth = 16384;
Int MaxFramebufferHeight = 16384;
Int MaxFramebufferLayers = 2048;
Int MaxRenderbufferSize = 16384;
Int MaxTextureSize = 16384;
Int MaxColorTextureSamples = 1;
Int MaxDepthTextureSamples = 1;
Int MaxFramebufferSamples = 1;
Int MaxIntegerSamples = 1;
Int MaxSamples = 1;
Int MaxSampleMaskWords = 1;
Int MaxTextureImageUnits = 32;
Int MaxVertexTextureImageUnits = 32;
Int MaxComputeTextureImageUnits = 32;
Int MaxCombinedTextureImageUnits = 192;
Int MaxVertexAttribs = 16;
Int MaxComputeShaderStorageBlocks = 8;
Int MaxCombinedShaderStorageBlocks = 32;
Int MaxComputeUniformBlocks = 12;
Int MaxComputeWorkGroupInvocations = 128;
Int MaxShaderStorageBufferBindings = 8;
Int MaxTextureBufferSize = 65536;
Int MaxUniformBufferBindings = 24;
Int MaxUniformBlockSize = 16384;
Int MaxImageUnits = 8;
Int MaxCombinedImageUniforms = 8;
Int MaxVertexImageUniforms = 0;
Int MaxGeometryImageUniforms = 0;
Int MaxFragmentImageUniforms = 8;
Int MaxComputeImageUniforms = 8;
Int MaxDrawBuffers = 8;
Int MaxColorAttachments = 8;
Int MaxClipDistances = 8;
Int MaxViewports = 16;
Int MaxViewportWidth = 16384;
Int MaxViewportHeight = 16384;
Float ViewportBoundsRangeMin = 0.0f;
Float ViewportBoundsRangeMax = 0.0f;
Int ViewportSubpixelBits = 0;
Bool SupportsWideLines = false;
SizeT MaxShaderStorageBlockSize = 128 * 1024 * 1024;
Uint32 SubgroupSize = 0;
Uint32 SubgroupSupportedStages = 0;
Uint32 SubgroupSupportedFeatures = 0;
Bool SubgroupQuadOperationsInAllStages = false;
GpuVendorKind GpuVendor = GpuVendorKind::Unknown;
};
enum class WindowBackend {
Android,
X11,
MetalLayer,
Win32, // Handle is an HWND
// TODO: Wayland, etc.
// TODO: X11, Wayland, Windows, macOS, etc.
WindowBackendCount,
Unknown = -1
};
@@ -322,8 +84,6 @@ namespace MobileGL {
struct WindowHandle {
WindowBackend Backend = WindowBackend::Unknown;
void* Handle = nullptr;
Uint32 Width = 0;
Uint32 Height = 0;
};
class BackendObject {
@@ -331,20 +91,8 @@ namespace MobileGL {
virtual ~BackendObject() = default;
virtual void Initialize() = 0;
virtual Bool InitCapabilities() = 0;
virtual Bool InitWindowSurface() = 0;
virtual Bool InitializeEGLDisplay(EGLDisplay dpy, EGLint* major, EGLint* minor);
virtual Bool CreateEGLWindowSurface(EGLSurface surface, const WindowHandle& handle);
virtual Bool ResizeEGLWindowSurface(EGLSurface surface, Uint32 width, Uint32 height);
virtual Bool CreateEGLPbufferSurface(EGLSurface surface, EGLint width, EGLint height);
virtual Bool MakeEGLCurrent(EGLDisplay dpy, EGLSurface draw, EGLSurface read, EGLContext ctx);
virtual Bool SwapEGLBuffers(EGLDisplay dpy, EGLSurface draw);
// Forwards the app-requested eglSwapInterval to the backend's native
// presentation path (no-op for backends without a SetSwapInterval hook).
virtual void SetEGLSwapInterval(Int interval);
virtual void ReleaseEGLSurface(EGLSurface surface);
virtual void ReleaseEGLResources();
virtual void InitCapabilities() = 0;
virtual void InitWindowSurface() = 0;
void SetWindowHandle(const WindowHandle& handle);
@@ -352,56 +100,10 @@ namespace MobileGL {
virtual String GetBackendAPIVersionString() const = 0;
virtual const GlobalBackendFunctionsTable& GetBackendFunctions() const = 0;
virtual const DynamicBackendParameters& GetDynamicParameters() const = 0;
const FormatCapabilityCache& GetFormatCapabilities() const;
virtual BackendType GetBackendType() const = 0;
protected:
enum class SurfaceKind {
None,
Window,
Pbuffer
};
struct EGLCurrentState {
EGLDisplay Display = EGL_NO_DISPLAY;
EGLSurface DrawSurface = EGL_NO_SURFACE;
EGLSurface ReadSurface = EGL_NO_SURFACE;
EGLContext Context = EGL_NO_CONTEXT;
};
struct EGLSurfaceState {
SurfaceKind Kind = SurfaceKind::None;
Bool DestroyPending = false;
WindowHandle Window;
EGLint Width = 1;
EGLint Height = 1;
};
void ResetEGLRuntimeState();
Bool RegisterEGLWindowSurface(EGLSurface surface, const WindowHandle& handle);
Bool RegisterEGLPbufferSurface(EGLSurface surface, EGLint width, EGLint height);
const EGLSurfaceState* GetRegisteredEGLSurface(EGLSurface surface) const;
Bool ActivateEGLSurface(EGLSurface surface);
virtual Bool InitPbufferSurface(EGLint width, EGLint height);
virtual void OnEGLSurfaceReleased(EGLSurface surface);
FormatCapabilityCache& MutableFormatCapabilities();
mutable std::recursive_mutex m_eglStateMutex;
FormatCapabilityCache m_formatCapabilities;
WindowHandle m_windowHandle;
EGLDisplay m_eglDisplay = EGL_NO_DISPLAY;
EGLSurface m_eglSurface = EGL_NO_SURFACE;
Bool m_eglDisplayInitialized = false;
Bool m_eglSurfaceInitialized = false;
Bool m_backendCapabilitiesInitialized = false;
SurfaceKind m_eglSurfaceKind = SurfaceKind::None;
UnorderedMap<std::thread::id, EGLCurrentState> m_eglCurrentThreads;
UnorderedMap<EGLSurface, EGLSurfaceState> m_eglSurfaces;
private:
Bool IsEGLSurfaceCurrent(EGLSurface surface) const;
void DestroyPendingEGLSurfaceIfUnused(EGLSurface surface);
void ReleaseEGLCurrentThread(const std::thread::id& threadKey);
};
} // namespace MG_Backend
} // namespace MobileGL
+4 -2
View File
@@ -10,9 +10,11 @@
#include <Includes.h>
#include "BackendObject.h"
#include "DirectGLES/BackendObject_DirectGLES.h"
#include "DirectVulkan/BackendObject_DirectVulkan.h"
#include "DirectVulkanTMP/BackendObject_DirectVulkanTMP.h"
namespace MobileGL::MG_Backend {
extern UniquePtr<BackendObject>& pActiveBackendObject;
extern UniquePtr<BackendObject> pActiveBackendObject;
extern GlobalBackendFunctionsTable gBackendFunctionsTable;
void Init();
} // namespace MobileGL::MG_Backend
File diff suppressed because it is too large Load Diff
@@ -12,26 +12,13 @@
#include <MG_Util/BackendLoaders/OpenGL/Loader.h>
namespace MobileGL::MG_Backend::DirectGLES {
// Populates the same format-capability cache used by backend startup. The caller
// must keep the supplied GLES context current for the duration of this call.
void PopulateFormatCapabilities(const MG_External::GLESFunctionsTable& gl,
const MG_External::GLESCapabilities& capabilities,
FormatCapabilityCache& cache);
class BackendObject_DirectGLES : public BackendObject {
public:
~BackendObject_DirectGLES() override;
void Initialize() override;
Bool InitCapabilities() override;
Bool InitWindowSurface() override;
Bool InitializeEGLDisplay(EGLDisplay dpy, EGLint* major, EGLint* minor) override;
Bool CreateEGLWindowSurface(EGLSurface surface, const WindowHandle& handle) override;
Bool CreateEGLPbufferSurface(EGLSurface surface, EGLint width, EGLint height) override;
Bool MakeEGLCurrent(EGLDisplay dpy, EGLSurface draw, EGLSurface read, EGLContext ctx) override;
Bool SwapEGLBuffers(EGLDisplay dpy, EGLSurface draw) override;
void ReleaseEGLSurface(EGLSurface surface) override;
void ReleaseEGLResources() override;
void InitCapabilities() override;
void InitWindowSurface() override;
const RendererInfo& GetRendererInfo() const override;
String GetBackendAPIVersionString() const override;
@@ -41,12 +28,9 @@ namespace MobileGL::MG_Backend::DirectGLES {
const MG_External::GLESFunctionsTable& GetGLESFunctions() const;
const MG_External::EGLFunctionsTable& GetEGLFunctions() const;
void ApplyGLESCapabilitiesForTesting(const MG_External::GLESCapabilities& capabilities);
private:
void UpdateDynamicBackendParameters();
Bool InitPbufferSurface(EGLint width, EGLint height) override;
void OnEGLSurfaceReleased(EGLSurface surface) override;
Bool m_initialized = false;
MG_External::EGLFunctionsTable m_EGLFunctions;
@@ -54,25 +38,4 @@ namespace MobileGL::MG_Backend::DirectGLES {
MG_External::GLESCapabilities m_GLESCapabilities;
DynamicBackendParameters m_dynamicParameters;
};
// Single-source-of-truth helpers shared with the driver POST
// (MG_Util/SelfTest/DriverPost.cpp), so the identity strings and extension list
// MobileGL reports to applications on this backend cannot drift from what the
// POST screen shows.
// Static identity of the Espryt renderer (renderer/backend names, target GL/GLSL
// versions, ExtraVendor). The Extensions vector inside is live backend state that
// is reconciled after capability init; callers that need the advertised list for
// a known capability set must use BuildAdvertisedExtensions instead.
const RendererInfo& GetRendererIdentity();
// The full OpenGL extension list Espryt advertises (glGetString(GL_EXTENSIONS))
// for a device whose timer queries / anisotropic filtering are (or are not) usable.
// The MOBILEGL_DISABLE_TIMERQUERY escape hatch is applied inside.
Vector<GLExtension> BuildAdvertisedExtensions(Bool timerQueriesSupported, Bool anisotropicFilteringSupported);
// Format: <OpenGL ES Renderer>, OpenGL ES <Major>.<Minor> — the exact string an
// initialized backend returns from GetBackendAPIVersionString (and that ends up
// inside the application-visible GL_RENDERER string).
String FormatBackendAPIVersionString(const String& glesRendererString, Int glesMajor, Int glesMinor);
} // namespace MobileGL::MG_Backend::DirectGLES
File diff suppressed because it is too large Load Diff
@@ -8,8 +8,6 @@
#pragma once
#include <Includes.h>
#include <MG_Backend/BackendObject.h>
#include <MG_State/GLState/FramebufferState/FramebufferObject.h>
#include <MG_State/GLState/TextureState/TextureState.h>
#include <MG_State/GLState/SamplerState/SamplerObject.h>
#include <MG_Util/BackendLoaders/OpenGL/Loader.h>
@@ -19,10 +17,6 @@
operation Utils::CheckGLESError();
namespace MobileGL::MG_Backend::DirectGLES {
// Re-establishes the frontend texture-unit bindings on the native ES context.
// Content uploads use scratch bindings, so draws and dispatches call this after
// texture synchronization.
void BindCurrentTextures();
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);
@@ -31,14 +25,11 @@ namespace MobileGL::MG_Backend::DirectGLES {
void DrawElements(GLenum mode, GLsizei count, GLenum type, const void* indices);
void DrawArrays(GLenum mode, GLint first, GLsizei count);
void DrawElementsBaseVertex(GLenum mode, GLsizei count, GLenum type, const GLvoid* indices, GLint basevertex);
void MultiDrawArrays(GLenum mode, const GLint* first, const GLsizei* count, GLsizei drawcount);
void MultiDrawElements(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount);
void MultiDrawElementsBaseVertex(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount, const GLint* basevertex);
void MultiDrawElementsIndirect(GLenum mode, GLenum type, const void* indirect, GLsizei drawcount, GLsizei stride);
void MultiDrawElementsIndirectCount(GLenum mode, GLenum type, const void* indirect, GLintptr drawcount,
GLsizei maxdrawcount, GLsizei stride);
void MultiDrawArraysIndirect(GLenum mode, const void* indirect, GLsizei drawcount, GLsizei stride);
void DrawRangeElementsBaseVertex(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type,
const void* indices, GLint basevertex);
@@ -55,100 +46,18 @@ namespace MobileGL::MG_Backend::DirectGLES {
GLuint baseinstance);
void DrawArraysInstanced(GLenum mode, GLint first, GLsizei count, GLsizei instancecount);
void DrawArraysIndirect(GLenum mode, const void* indirect);
void ClearNamedFramebufferfv(const SharedPtr<MG_State::GLState::FramebufferObject>& framebuffer,
GLenum buffer, GLint drawbuffer, const GLfloat* value);
void ClearNamedFramebufferfi(const SharedPtr<MG_State::GLState::FramebufferObject>& framebuffer,
GLenum buffer, GLint drawbuffer, GLfloat depth, GLint stencil);
void BlitFramebuffer(GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1, GLint dstX0, GLint dstY0, GLint dstX1,
GLint dstY1, GLbitfield mask, GLenum filter);
void BlitNamedFramebuffer(const SharedPtr<MG_State::GLState::FramebufferObject>& readFramebuffer,
const SharedPtr<MG_State::GLState::FramebufferObject>& drawFramebuffer,
GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1,
GLint dstX0, GLint dstY0, GLint dstX1, GLint dstY1,
GLbitfield mask, GLenum filter);
void CopyTexImage2D(GLenum target, GLint level, GLenum internalformat, GLint x, GLint y, GLsizei width,
GLsizei height, GLint border);
void CopyTexSubImage2D(GLenum target, GLint level, GLint xoffset, GLint yoffset, GLint x, GLint y, GLsizei width,
GLsizei height);
void CopyImageSubData(const SharedPtr<MG_State::GLState::ITextureObject>& srcTexture,
GLenum srcTarget, GLint srcLevel, GLint srcX, GLint srcY, GLint srcZ,
const SharedPtr<MG_State::GLState::ITextureObject>& dstTexture,
GLenum dstTarget, GLint dstLevel, GLint dstX, GLint dstY, GLint dstZ,
GLsizei srcWidth, GLsizei srcHeight, GLsizei srcDepth);
void GenerateMipmap(GLenum target);
const GLubyte* GetString(GLenum name);
void ReadPixels(GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, void* pixels);
void GetTexImage(GLenum target, GLint level, GLenum format, GLenum type, GLvoid* pixels);
void DispatchCompute(GLuint numGroupsX, GLuint numGroupsY, GLuint numGroupsZ);
void DispatchComputeIndirect(GLintptr indirect);
void MemoryBarrier(GLbitfield barriers);
void MemoryBarrierByRegion(GLbitfield barriers);
void BindImageTexture(GLuint unit, GLuint texture, GLint level, GLboolean layered, GLint layer, GLenum access,
GLenum format);
void GetIntegeri_v(GLenum target, GLuint index, GLint* data);
void GetInteger64i_v(GLenum target, GLuint index, GLint64* data);
void GetProgramiv(GLuint program, GLenum pname, GLint* params);
void GetProgramInterfaceiv(GLuint program, GLenum programInterface, GLenum pname, GLint* params);
GLuint GetProgramResourceIndex(GLuint program, GLenum programInterface, const GLchar* name);
void GetProgramResourceName(GLuint program, GLenum programInterface, GLuint index, GLsizei bufSize, GLsizei* length,
GLchar* name);
void GetProgramResourceiv(GLuint program, GLenum programInterface, GLuint index, GLsizei propCount,
const GLenum* props, GLsizei bufSize, GLsizei* length, GLint* params);
GLint GetProgramResourceLocation(GLuint program, GLenum programInterface, const GLchar* name);
GLint GetProgramResourceLocationIndex(GLuint program, GLenum programInterface, const GLchar* name);
void ShaderStorageBlockBinding(GLuint program, GLuint storageBlockIndex, GLuint storageBlockBinding);
Bool InitWindowSurface(NativeWindowType window);
Bool InitPbufferSurface(EGLint width, EGLint height);
Bool MakeCurrent();
Bool ReleaseCurrent();
// True when the backend ES context is current on the calling thread, i.e.
// immediate buffer ops may issue GL calls right now.
Bool IsBackendContextCurrentOnThisThread();
// GL fence sync objects, backed by native ES fences. FenceSync returns null
// (the frontend then falls back to an always-signaled sync) when the calling
// thread does not own the ES context. Waits/queries degrade to "signaled" in
// the same situation, and handles created under a since-destroyed ES context
// are always treated as signaled.
BackendSyncHandle FenceSync();
GLenum ClientWaitSync(BackendSyncHandle sync, GLbitfield flags, GLuint64 timeout);
void WaitSync(BackendSyncHandle sync, GLbitfield flags, GLuint64 timeout);
void DeleteSync(BackendSyncHandle sync);
Bool GetSyncStatus(BackendSyncHandle sync);
// True when GL_EXT_disjoint_timer_query and every entry point the timer
// hooks below need are present. Also gates the E_GL_ARB_timer_query
// advertisement in BackendObject_DirectGLES::InitCapabilities, and is
// registered as the GLFunctionsTable::IsTimerQuerySupported hook: a pure
// capability read needs no current ES context, and it stays false until
// the ES capabilities have been filled in.
Bool AreTimerQueriesSupported();
// GL timer-query objects, backed by GL_EXT_disjoint_timer_query. The
// creators return null (the frontend then falls back to an immediately
// available zero result) when the calling thread does not own the ES
// context or the extension/entry points are missing, and handles created
// under a since-destroyed ES context are always treated as complete with
// a zero result (mirrors the fence-sync handles above).
BackendQueryHandle BeginTimeElapsedQuery();
void EndTimeElapsedQuery(BackendQueryHandle query);
BackendQueryHandle QueryCounterTimestamp();
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
// and release the handle). Returns false only when the calling thread
// does not own the ES context, so the value is genuinely unobtainable
// right now; the handle stays alive and readable later.
Bool GetQueryResult64(BackendQueryHandle query, Bool wait, Uint64* outNanoseconds);
void DeleteBackendQuery(BackendQueryHandle query);
Int64 GetGpuTimestampNs();
void Present();
// Frame-completion watermarks for the buffer-storage pool: CurrentFrameSerial()
// is bumped once per Present(); CompletedFrameSerial() is the newest frame whose
// GPU work has provably finished (advanced by polling a one-fence-per-frame ring).
// A buffer retired during frame N is safe to recycle once CompletedFrameSerial() >= N.
Uint64 CurrentFrameSerial();
Uint64 CompletedFrameSerial();
// Applies (or defers until the window surface exists) the app-requested
// eglSwapInterval on the native EGL surface.
void SetSwapInterval(Int interval);
void SetEGLFuncsTable(const MG_External::EGLFunctionsTable& eglFuncs);
void SetGLESFuncsTable(const MG_External::GLESFunctionsTable& glesFuncs);
void SetGLESCapabilities(const MG_External::GLESCapabilities& capabilities);
File diff suppressed because it is too large Load Diff
+50 -509
View File
@@ -8,324 +8,67 @@
#pragma once
#include <Includes.h>
#include <atomic>
#include <mutex>
#include "DirectGLES.h"
#include "MG_State/GLState/SamplerState/SamplerObject.h"
#include "MG_State/GLState/TextureState/TextureEnum.h"
#include <MG_State/GLState/TextureState/TextureObject.h>
#include <MG_State/GLState/Core.h>
#include <MG_Util/Converters/MGToGL/TextureEnumConverter.h>
namespace MobileGL::MG_Backend::DirectGLES {
String EmulateBaseInstanceInVertexShader(String source, GLenum shaderType);
String PromoteDrawParameterGlobalsToUniforms(String source, GLenum shaderType);
template <typename StateObject, typename BackendObject>
class StateBackendObjectRegistry {
public:
using StatePtr = SharedPtr<StateObject>;
using StateWeakPtr = std::weak_ptr<StateObject>;
using BackendPtr = SharedPtr<BackendObject>;
using BackendMap = UnorderedMap<StateObject*, BackendPtr>;
using StateRefMap = UnorderedMap<StateObject*, StateWeakPtr>;
using iterator = typename BackendMap::iterator;
using const_iterator = typename BackendMap::const_iterator;
BackendPtr& GetOrCreate(const StatePtr& stateObj) {
MOBILEGL_ASSERT(stateObj != nullptr, "State object must not be null");
auto* key = stateObj.get();
auto trackedStateIt = m_stateRefs.find(key);
if (trackedStateIt != m_stateRefs.end() && trackedStateIt->second.expired()) {
EraseByKey(key);
}
m_stateRefs[key] = stateObj;
return m_backendObjects[key];
}
iterator find(StateObject* stateObj) {
if (!IsAlive(stateObj)) {
EraseByKey(stateObj);
return m_backendObjects.end();
}
return m_backendObjects.find(stateObj);
}
const_iterator find(StateObject* stateObj) const {
return const_cast<StateBackendObjectRegistry*>(this)->find(stateObj);
}
iterator begin() { return m_backendObjects.begin(); }
const_iterator begin() const { return m_backendObjects.begin(); }
iterator end() { return m_backendObjects.end(); }
const_iterator end() const { return m_backendObjects.end(); }
void CollectGarbageIfNeeded() {
++m_gcTick;
if (m_gcTick < kGCInterval) {
return;
}
CollectGarbage();
m_gcTick = 0;
}
void CollectGarbageNow() { CollectGarbage(); }
private:
bool IsAlive(StateObject* stateObj) const {
const auto trackedStateIt = m_stateRefs.find(stateObj);
if (trackedStateIt == m_stateRefs.end()) {
return false;
}
return !trackedStateIt->second.expired();
}
void EraseByKey(StateObject* stateObj) {
m_stateRefs.erase(stateObj);
m_backendObjects.erase(stateObj);
}
void CollectGarbage() {
if (m_isCollecting) {
return;
}
m_isCollecting = true;
Vector<StateObject*> staleKeys;
staleKeys.reserve(m_stateRefs.size());
for (const auto& [stateKey, stateWeakRef] : m_stateRefs) {
if (stateWeakRef.expired()) {
staleKeys.push_back(stateKey);
}
}
for (auto* stateKey : staleKeys) {
m_stateRefs.erase(stateKey);
m_backendObjects.erase(stateKey);
}
m_isCollecting = false;
}
private:
static constexpr Uint32 kGCInterval = 1024;
StateRefMap m_stateRefs;
BackendMap m_backendObjects;
Uint32 m_gcTick = 0;
Bool m_isCollecting = false;
};
namespace BufferImpl {
const GLenum TempBufferTarget = GL_ARRAY_BUFFER;
// The DirectGLES storage behind one frontend buffer. Owned (refcounted) by
// the frontend BufferObject; immediate BufferBackendOps keep it current, so
// draw-time "sync" reduces to ensuring the storage exists.
class GLESBufferResource : public MG_State::GLState::BackendBufferResource {
class BackendBufferObject {
public:
~GLESBufferResource() override = default;
BackendBufferObject();
void SyncToBackend(SharedPtr<MG_State::GLState::BufferObject>& stateBufferObject);
Uint GetBackendBufferId() { return m_backendBufferId; }
void Bind(GLenum target = TempBufferTarget);
Uint id = 0;
SizeT storageSize = 0;
Bool storageInitialized = false;
// ES context generation this resource's id belongs to; ids from a
// destroyed context are invalid and must not be deleted or reused.
Uint contextGeneration = 0;
// Frontend change serial the backend storage reflects. When immediate
// ops cannot run (ops unregistered, no current context), this lags and
// EnsureBufferResource falls back to a full re-upload. Atomic: read on
// the context-owning thread while ops on other threads may update it.
std::atomic<Uint64> syncedChangeSerial{0};
// Ops that arrived while no ES context was current on the calling thread
// (or before storage existed); replayed by EnsureBufferResource. The ES
// context migrates between app threads, so deferring ops can race with
// the owning thread replaying them: guard both fields with pendingMutex.
Bool pendingRespecify = false;
VecRange1D pendingRanges;
std::mutex pendingMutex;
// Zero-copy coherent persistent map (EXT_buffer_storage): the GL store is
// immutable, persistently+coherently mapped, and persistentPtr is what the app
// (and the frontend PipeResource) write into directly. While set, draw-time
// sync is a no-op and no per-draw glBufferSubData is issued. Cleared on ES
// context loss.
Bool persistentMapped = false;
void* persistentPtr = nullptr;
private:
void SyncToBackend_glBufferData(SharedPtr<MG_State::GLState::BufferObject>& stateBufferObject);
void SyncToBackend_glBufferSubData(SharedPtr<MG_State::GLState::BufferObject>& stateBufferObject);
void SyncToBackend_glMapBufferRange(SharedPtr<MG_State::GLState::BufferObject>& stateBufferObject,
Bool invalidate = true, Bool unsynchronized = true);
Uint m_backendBufferId = 0;
SizeT m_prevBufferSize = 0;
Bool m_isInitialized = false;
};
// Registered as the frontend's BufferBackendOps at backend init and on
// every MakeCurrent (the ES context can be destroyed and recreated, e.g.
// by the trace replayer's probe context).
void RegisterBufferBackendOps();
void UnregisterBufferBackendOps();
// The ES context died: unregister ops, invalidate all outstanding GL ids
// (they belonged to the dead context) and drop deferred deletes.
void OnBackendContextDestroyed();
// Get-or-create the backend resource and bring its storage up to date
// (creates the GL buffer, replays pending ops, pushes persistent-mapped
// ranges). Requires the ES context to be current. Returns nullptr only
// for null input.
GLESBufferResource* EnsureBufferResource(const SharedPtr<MG_State::GLState::BufferObject>& bufferObject);
// Existing resource or nullptr; performs no GL calls.
GLESBufferResource* GetBufferResource(MG_State::GLState::BufferObject* bufferObject);
// Deletes GL buffers whose owning frontend objects died (possibly on a
// thread without a current ES context). Called from draw-time sync.
void ProcessDeferredBufferReleases();
// glBindBuffer with a redundant-bind cache for GL_ARRAY_BUFFER.
void BindBufferId(GLenum target, Uint id);
void InvalidateArrayBufferBindingCache();
// Redundant-bind caches for the driver-level GL_PIXEL_PACK/UNPACK_BUFFER
// bindings. Every backend readback (glReadPixels / pack-PBO map) and pixel
// upload site routes its binding through these so the shadow always matches
// the driver; the resting state between operations is 0, which keeps any
// path that implicitly assumes "no PBO bound" correct. Scrubbed when a
// buffer id is deleted/pooled (GL resets a deleted buffer's bindings to 0,
// and a recycled name matching the shadow would false-skip the rebind) and
// invalidated on MakeCurrent (context may reset).
void BindPixelPackBufferId(Uint id);
void BindPixelUnpackBufferId(Uint id);
void InvalidatePixelBufferBindingCaches();
// A GL buffer id is being deleted by code outside BufferImpl (e.g. the VAO
// client-attribute staging buffers): scrub every buffer-binding shadow that
// could false-skip when the name is recycled.
void NoteBufferIdDeleted(Uint id);
// Redundant-bind cache for INDEXED buffer bindings (glBindBufferBase/Range on
// GL_UNIFORM_BUFFER / GL_SHADER_STORAGE_BUFFER): skips the GL call when the
// (id, range) already at that index matches, like the array-buffer/texture/
// sampler caches already do. Invalidated on MakeCurrent (context may reset).
void BindBufferBaseCached(GLenum glTarget, Uint index, Uint id);
void BindBufferRangeCached(GLenum glTarget, Uint index, Uint id, GLintptr offset, GLsizeiptr size);
void InvalidateIndexedBufferBindingCache();
// Buffer-storage pool maintenance. TrimBufferPool evicts over-budget entries
// (called once per frame from Present); ClearBufferPool drops all pooled ids
// without glDeleteBuffers (called when the ES context is going away).
void TrimBufferPool();
void ClearBufferPool();
// --- Global-UBO ring ------------------------------------------------------
// One persistently+coherently mapped buffer (EXT_buffer_storage) shared by
// every program's lowered default-uniform block. Each content change is
// bump-allocated into a fresh slot and bound with glBindBufferRange, so the
// CPU never rewrites bytes the GPU may still be reading — the per-draw
// glBufferSubData into one static UBO forced Adreno to resolve that
// write-after-read hazard on every uniform-dirtying draw (MC dirties
// uniforms every draw). Reclamation rides the Present() frame-fence
// watermark; no ring bytes are recycled before their frame's GPU work
// completed.
//
// A program's cached slot, reusable within one frame while the frontend UBO
// content version is unchanged. Cross-frame reuse is intentionally not
// attempted: later same-frame allocations may recycle bytes of completed
// frames, so re-referencing them would need per-bind pinning — rewriting
// GetUBOSize() bytes once per program per frame is far cheaper.
struct UboRingAllocation {
Uint32 contentVersion = ~0u; // frontend UBO content version held at `offset`
Uint32 ringGeneration = 0; // ring identity the slot lives in (0 = never valid)
Uint64 frameSerial = ~Uint64{0}; // frame the slot was written in
SizeT offset = 0;
};
// False when the feature is disabled, EXT_buffer_storage / fences are
// missing, the ES context is not current, or ring creation already failed
// under this context (callers then take the legacy glBufferSubData path).
Bool UboRingAvailable();
// Bump-allocate `size` bytes aligned to GL_UNIFORM_BUFFER_OFFSET_ALIGNMENT.
// Grows the ring (new GL store, generation bump) when the in-flight span
// would be overrun. Returns false when storage (re)creation fails.
Bool UboRingAllocate(SizeT size, SizeT& outOffset);
void* UboRingMappedPtr();
Uint UboRingBufferId();
Uint32 UboRingGeneration();
// Present()-time upkeep: records the frame's high-water mark for reclamation
// and deletes grown-away ring stores once the GPU is done with them.
void UboRingOnPresent();
extern BackendBufferObject* g_boundVertexBufferObject;
extern UnorderedMap<SharedPtr<MG_State::GLState::BufferObject>, SharedPtr<BackendBufferObject>>
g_backendBufferObjects;
} // namespace BufferImpl
namespace VertexArrayImpl {
class BackendVertexArrayObject {
public:
BackendVertexArrayObject();
~BackendVertexArrayObject();
void SyncToBackend(const SharedPtr<MG_State::GLState::VertexArrayObject>& stateVAOObject);
void SyncClientSideAttributesForDrawArrays(
const SharedPtr<MG_State::GLState::VertexArrayObject>& stateVAOObject, GLint first, GLsizei count);
Uint GetBackendVertexArrayId() const { return m_backendVAOId; }
void Bind() const;
void SyncToBackend(SharedPtr<MG_State::GLState::VertexArrayObject>& stateVAOObject);
Uint GetBackendVertexArrayId() { return m_backendVAOId; }
void Bind();
private:
void BindAttributeBuffer(Uint index, const MG_State::GLState::VertexAttribute& attrib);
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>
m_syncedAttributeVersions;
};
extern StateBackendObjectRegistry<MG_State::GLState::VertexArrayObject, BackendVertexArrayObject>
extern UnorderedMap<SharedPtr<MG_State::GLState::VertexArrayObject>, SharedPtr<BackendVertexArrayObject>>
g_backendVertexArrayObjects;
} // namespace VertexArrayImpl
namespace TextureImpl {
inline Bool IsSupportedTextureTarget(TextureTarget target) {
// Rectangle textures need non-normalized sampling ES cannot express; everything else is
// either native or emulated (1D -> 2D with height 1, 1D array -> 2D array, see
// MapToBackendTextureTarget). SPIRV-Cross already emits the matching ESSL samplers and
// coordinate padding for 1D/1D-array shaders.
return target != TextureTarget::TextureRectangle;
}
// ES has no 1D targets: 1D textures are stored as 2D (height 1) and 1D arrays as 2D arrays
// (height 1, layers in depth). Must match SPIRV-Cross's ES 1D-as-2D shader emulation.
inline TextureTarget MapToBackendTextureTarget(TextureTarget target) {
switch (target) {
case TextureTarget::Texture1D:
return TextureTarget::Texture2D;
case TextureTarget::Texture1DArray:
return TextureTarget::Texture2DArray;
default:
return target;
}
}
inline GLenum ConvertTextureTargetToBackendGLEnum(TextureTarget target) {
return MG_Util::ConvertTextureTargetToGLEnum(MapToBackendTextureTarget(target));
}
inline GLenum ConvertTextureUploadTargetToBackendGLEnum(TextureUploadTarget uploadTarget) {
switch (uploadTarget) {
case TextureUploadTarget::Texture1D:
return GL_TEXTURE_2D;
case TextureUploadTarget::Texture1DArray:
return GL_TEXTURE_2D_ARRAY;
default:
return MG_Util::ConvertTextureUploadTargetToGLEnum(uploadTarget);
}
}
// 1D arrays store layers in the state-side height; the ES 2D-array image keeps height 1 and
// moves the layer count into depth.
inline IntVec3 GetBackendUploadSize(TextureTarget stateTarget, const IntVec3& texelSize) {
if (stateTarget == TextureTarget::Texture1DArray) {
return {texelSize.x(), 1, texelSize.y()};
}
return texelSize;
}
inline Bool IsMultisampleTextureTarget(TextureTarget target) {
return target == TextureTarget::Texture2DMultisample ||
target == TextureTarget::Texture2DMultisampleArray;
}
inline Bool SupportsWrapR(TextureTarget target) {
return target == TextureTarget::Texture3D || target == TextureTarget::TextureCubeMap;
if (target == TextureTarget::Texture1D || target == TextureTarget::TextureRectangle ||
target == TextureTarget::Texture2DMultisampleArray || target == TextureTarget::Texture1DArray ||
target == TextureTarget::Texture2DMultisample || target == TextureTarget::Texture2DArray)
return false;
return true;
}
struct StateTextureBasicInfo { // Used for tracking texture state changes
@@ -335,14 +78,11 @@ namespace MobileGL::MG_Backend::DirectGLES {
SizeT depth = 0;
SizeT mipmapLevels = 0;
Uint bufferExternalIndex = 0;
Int samples = 0;
Bool fixedSampleLocations = true;
bool operator==(const StateTextureBasicInfo& other) const {
return internalFormat == other.internalFormat && width == other.width && height == other.height &&
depth == other.depth && mipmapLevels == other.mipmapLevels &&
bufferExternalIndex == other.bufferExternalIndex && samples == other.samples &&
fixedSampleLocations == other.fixedSampleLocations;
bufferExternalIndex == other.bufferExternalIndex;
}
bool operator!=(const StateTextureBasicInfo& other) const { return !(*this == other); }
@@ -352,29 +92,15 @@ namespace MobileGL::MG_Backend::DirectGLES {
class BackendTextureObject {
public:
BackendTextureObject();
// Deletes the GL texture (frontend glDeleteTextures used to leak every
// backend id for the context lifetime) and scrubs the binding/scratch-FBO
// shadows so a recycled name or heap address cannot false-skip a rebind.
~BackendTextureObject();
BackendTextureObject(const BackendTextureObject&) = delete;
BackendTextureObject& operator=(const BackendTextureObject&) = delete;
void SyncMipmapsToBackend(const SharedPtr<MG_State::GLState::ITextureObject>& stateTextureObject);
void SyncBuiltinSamplerToBackend(const SharedPtr<MG_State::GLState::ITextureObject>& stateTextureObject);
void SyncTextureParamsToBackend(const SharedPtr<MG_State::GLState::ITextureObject>& stateTextureObject);
void RequireImageBindableStorage();
void SyncMipmapsToBackend(SharedPtr<MG_State::GLState::ITextureObject>& stateTextureObject);
void SyncBuiltinSamplerToBackend(SharedPtr<MG_State::GLState::ITextureObject>& stateTextureObject);
void SyncTextureParamsToBackend(SharedPtr<MG_State::GLState::ITextureObject>& stateTextureObject);
void Bind(GLenum target, Uint unit = TempTextureUnit);
Uint GetBackendTextureId() const;
Uint GetBackendTextureId();
private:
void RecreateBackendTexture();
Uint m_backendTextureId = 0;
// ES context generation the id was created under; a dtor running after
// that context died must not delete a foreign (recycled) name.
Uint m_contextGeneration = 0;
Bool m_isInitialized = false;
Bool m_imageBindableStorageRequired = false;
Bool m_backendStorageImmutable = false;
StateTextureBasicInfo m_prevTextureInfo;
SamplerParameters m_cacheSamplerParameters;
UintVec2 m_cacheLodRange = {0, 1000};
@@ -387,33 +113,25 @@ namespace MobileGL::MG_Backend::DirectGLES {
void ActivateTextureUnit(Uint unit);
void UnbindTexture(Uint unit, GLenum target);
extern StateBackendObjectRegistry<MG_State::GLState::ITextureObject, BackendTextureObject>
extern UnorderedMap<SharedPtr<MG_State::GLState::ITextureObject>, SharedPtr<BackendTextureObject>>
g_backendTextureObjects;
SharedPtr<BackendTextureObject>& SyncTextureObjectToBackend(
const SharedPtr<MG_State::GLState::ITextureObject>& textureObject,
Bool imageBindableStorageRequired = false);
extern Array<Array<BackendTextureObject*, (SizeT)TextureTarget::TextureTargetCount>,
MG_State::GLState::TextureState::MAX_TEXTURE_IMAGE_UNITS>
g_boundTexturesCache;
extern Uint g_activeTextureUnit;
// Bumped when the backend ES context is destroyed; texture ids stamped with
// an older generation belong to a dead context and must not be deleted.
extern Uint g_textureContextGeneration;
} // namespace TextureImpl
namespace FramebufferImpl {
class BackendFramebufferObject {
public:
BackendFramebufferObject();
void SyncToBackend(const SharedPtr<MG_State::GLState::FramebufferObject>& stateFBOObject,
void SyncToBackend(SharedPtr<MG_State::GLState::FramebufferObject>& stateFBOObject,
FramebufferTarget asTarget);
// Apply only this FBO's read buffer (glReadBuffer) to the backend. Split out so it can
// still run when SyncCurrentFBO skips the READ-target sync because the same GL FBO is
// bound as both draw and read (otherwise glReadBuffer changes would be silently dropped).
void SyncReadBufferToBackend(const SharedPtr<MG_State::GLState::FramebufferObject>& stateFBOObject);
void InvalidateSyncedState();
Uint GetBackendFramebufferId() const { return m_backendFBOId; }
void Bind(FramebufferTarget target) const;
Uint GetBackendFramebufferId() { return m_backendFBOId; }
void Bind(FramebufferTarget target);
bool SyncAttachmentObject(GLenum glFBOTarget,
const MG_State::GLState::FramebufferAttachmentObject& attachmentObject,
GLenum glBackendAttachment);
// FramebufferAttachmentType GetCompactedAttachmentTypeAtDrawBufferIndex(Int index);
GLenum GetBackendAttachmentType(FramebufferAttachmentType frontendAtt) const;
@@ -441,204 +159,28 @@ namespace MobileGL::MG_Backend::DirectGLES {
FramebufferObject::FramebufferAttachmentVersionArray m_syncedFrontendAttachmentVersions = {0};
};
extern StateBackendObjectRegistry<MG_State::GLState::FramebufferObject, BackendFramebufferObject>
extern UnorderedMap<SharedPtr<MG_State::GLState::FramebufferObject>, SharedPtr<BackendFramebufferObject>>
g_backendFramebufferObjects;
extern Array<Uint16, SizeT(FramebufferTarget::FramebufferTargetCount)> g_fboBindVersions;
// Tracks the bound FBO's object version (bumped on any attachment/drawbuffer change)
// per target: re-attaching textures or changing draw buffers on an already-bound FBO
// must re-sync it even when the binding-slot version has not moved.
extern Array<Uint16, SizeT(FramebufferTarget::FramebufferTargetCount)> g_fboSyncedObjectVersions;
extern Array<MG_State::GLState::FramebufferObject*, SizeT(FramebufferTarget::FramebufferTargetCount)>
g_fboSyncedObjects;
// Driver-level READ/DRAW framebuffer-binding shadow. Every backend
// glBindFramebuffer routes through BindFramebufferId so scoped helpers can
// save/restore the current binding without a glGetIntegerv round-trip (that
// query forces a driver pipeline sync) and so redundant rebinds no-op.
// Starts unknown; the first CurrentFramebufferBinding() query pins it from
// the driver once. Invalidated on MakeCurrent (context may reset).
// GL_FRAMEBUFFER binds both targets.
void BindFramebufferId(GLenum fbTarget, Uint id);
Uint CurrentFramebufferBinding(FramebufferTarget target);
void InvalidateFramebufferBindingCache();
} // namespace FramebufferImpl
// Shared scratch framebuffers for the readback/copy/blit emulation paths, with a
// driver-side attachment shadow: repeated uses skip redundant detach/attach GL
// calls, and an attachment left by one use (e.g. a depth copy's DEPTH_STENCIL
// texture) is detached exactly when a later use of another aspect would
// otherwise inherit it (stale cross-aspect attachments made the shared temp FBO
// incomplete and silently degraded later readbacks).
namespace ScratchFBOImpl {
struct ScratchFramebuffer {
Uint id = 0;
// false => attachment state unknown; scrub every point on next use.
// A fresh FBO starts with nothing attached, so creation sets it true.
Bool attachmentsKnown = false;
Uint colorTex = 0;
GLenum colorTarget = 0;
GLint colorLevel = 0;
GLint colorLayer = -1; // >= 0 => attached via glFramebufferTextureLayer
Uint depthTex = 0;
GLenum depthTarget = 0;
GLint depthLevel = 0;
Bool depthHasStencil = false;
// Per-FBO read/draw buffer state (0 = unknown, set on first use).
GLenum readBuffer = 0;
GLenum drawBuffer = 0;
};
ScratchFramebuffer& TempFramebuffer(); // GetTexImage READ / CopyTex*Image2D depth DRAW
ScratchFramebuffer& BlitReadFramebuffer(); // texture-to-texture blit source
ScratchFramebuffer& BlitDrawFramebuffer(); // texture-to-texture blit destination
// Returns the GL id, generating it if needed (requires a current ES context).
Uint EnsureId(ScratchFramebuffer& fb);
// The fb must currently be bound at fbTarget (glReadBuffer/glDrawBuffers
// target the READ/DRAW binding respectively). Each Ensure* performs the
// minimal detach/attach set and keeps the shadow in sync; a failed attach
// records the point as detached so the completeness check fails instead of
// silently reading a stale attachment.
void EnsureColorAttachment2D(ScratchFramebuffer& fb, GLenum fbTarget, Uint tex, GLenum texTarget, GLint level);
void EnsureColorAttachmentLayer(ScratchFramebuffer& fb, GLenum fbTarget, Uint tex, GLint level, GLint layer);
void EnsureDepthAttachment2D(ScratchFramebuffer& fb, GLenum fbTarget, Uint tex, GLenum texTarget, GLint level,
Bool withStencil);
void EnsureNoColorAttachment(ScratchFramebuffer& fb, GLenum fbTarget);
void EnsureNoDepthAttachment(ScratchFramebuffer& fb, GLenum fbTarget);
void EnsureReadBuffer(ScratchFramebuffer& fb, GLenum readBuffer);
void EnsureDrawBuffer(ScratchFramebuffer& fb, GLenum drawBuffer);
// A 1x1 RGBA8-renderbuffer-complete FBO (GenerateMipmap needs a complete
// binding while respecifying texture storage). Attachment is set once at
// creation and never changes.
Uint EnsureCompleteTinyFramebufferId();
// A backend texture id is being deleted or respecified: a scratch FBO still
// referencing it would hold a dangling attachment (ES only auto-detaches
// from the *bound* framebuffer), and a recycled name could false-skip a
// re-attach; force a full scrub on next use.
void NoteTextureIdDeleted(Uint textureId);
// The ES context (and the scratch FBO ids with it) is going away.
void OnBackendContextDestroyed();
} // namespace ScratchFBOImpl
// Driver-level GL_PACK_* pixel-store shadow, the readback-side sibling of the
// upload path's ScopedDefaultUnpackState (Managers.cpp): the backend PACK state
// is written ONLY through ApplyPackState, so scoped helpers can save/restore it
// from the shadow instead of glGetIntegerv (which forces a driver pipeline
// sync), and redundant glPixelStorei calls no-op. The first Apply/Current call
// pins the driver to the shadow by writing all fields once. Invalidated on
// MakeCurrent (context may reset). PACK_IMAGE_HEIGHT/SKIP_IMAGES/SWAP_BYTES/
// LSB_FIRST have no ES equivalents; readbacks honor them on the CPU from the
// frontend context state instead.
namespace PixelStoreImpl {
struct PackState {
GLint Alignment = 4;
GLint RowLength = 0;
GLint SkipRows = 0;
GLint SkipPixels = 0;
Bool operator==(const PackState& o) const {
return Alignment == o.Alignment && RowLength == o.RowLength && SkipRows == o.SkipRows &&
SkipPixels == o.SkipPixels;
}
};
void ApplyPackState(const PackState& desired);
PackState CurrentPackState();
void InvalidatePackStateCache();
} // namespace PixelStoreImpl
// Image uniforms take their unit from the layout(binding=N) qualifier baked into
// the transpiled ESSL; unlike samplers they must not (and in ES cannot) be
// assigned through glUniform1i.
inline Bool IsImageUniformType(GLenum type) {
switch (type) {
case 0x904D: /*GL_IMAGE_2D*/
case 0x904E: /*GL_IMAGE_3D*/
case 0x9050: /*GL_IMAGE_CUBE*/
case 0x9051: /*GL_IMAGE_BUFFER*/
case 0x9053: /*GL_IMAGE_2D_ARRAY*/
case 0x9058: /*GL_INT_IMAGE_2D*/
case 0x9059: /*GL_INT_IMAGE_3D*/
case 0x905B: /*GL_INT_IMAGE_CUBE*/
case 0x905C: /*GL_INT_IMAGE_BUFFER*/
case 0x905E: /*GL_INT_IMAGE_2D_ARRAY*/
case 0x9063: /*GL_UNSIGNED_INT_IMAGE_2D*/
case 0x9064: /*GL_UNSIGNED_INT_IMAGE_3D*/
case 0x9066: /*GL_UNSIGNED_INT_IMAGE_CUBE*/
case 0x9067: /*GL_UNSIGNED_INT_IMAGE_BUFFER*/
case 0x9069: /*GL_UNSIGNED_INT_IMAGE_2D_ARRAY*/
return true;
default:
return false;
}
}
namespace PrgramImpl {
class BackendProgramObjectImpl {
public:
// Per-link cache of a sampler-style uniform's backend location: built once in
// SyncToBackend so draws stop issuing glGetUniformLocation string queries.
// lastAssignedUnit mirrors the program-state value set through glUniform1i
// (program state persists across binds, so caching per program is exact).
struct SamplerUniformBinding {
Uint frontendLocation = 0;
Int backendLocation = -1;
GLenum uniformType = 0;
Int lastAssignedUnit = -1;
};
BackendProgramObjectImpl();
~BackendProgramObjectImpl();
void SyncToBackend(const SharedPtr<MG_State::GLState::ProgramObject>& stateProgramObject);
void Use() const;
void SetBaseInstance(Uint32 baseInstance) const;
void SetBaseInstanceWordIndex(Int32 wordIndex) const;
void SetDrawID(Uint32 drawId) const;
Int GetIndirectParamsBinding() const { return m_indirectParamsBinding; }
void SyncToBackend(SharedPtr<MG_State::GLState::ProgramObject>& stateProgramObject);
void Use();
Uint GetBackendProgramId() const { return m_backendProgramId; }
Uint GetBackendGlobalUBOId() const { return m_backendGlobalUBOId; }
Uint32 GetSnormFallbackClampOutputMask() const { return m_snormFallbackClampOutputMask; }
Uint32 GetUnormFallbackClampOutputMask() const { return m_unormFallbackClampOutputMask; }
Bool HasGlobalUboBlock() const { return m_globalUboBackendBlockIndex >= 0; }
const Vector<Int>& GetUniformBlockBackendIndices() const { return m_uniformBlockBackendIndices; }
Vector<SamplerUniformBinding>& GetSamplerUniformBindings() { return m_samplerUniformBindings; }
Uint32 GetLastUploadedGlobalUboVersion() const { return m_lastUploadedGlobalUboVersion; }
void SetLastUploadedGlobalUboVersion(Uint32 version) { m_lastUploadedGlobalUboVersion = version; }
// Backend-reported GL_UNIFORM_BLOCK_DATA_SIZE of the global block; ring
// bindings must span at least this much (may exceed the frontend's
// reflected size when the transpiled block pads differently).
Int GetGlobalUboBackendBlockSize() const { return m_globalUboBackendBlockSize; }
BufferImpl::UboRingAllocation& GetGlobalUboRingAllocation() { return m_globalUboRingAllocation; }
// Frontend link version this backend program (and its resource caches) was
// built from; a mismatch means every link-derived cache here is stale.
Uint32 GetSyncedLinkVersion() const { return m_syncedLinkVersion; }
private:
void CacheResourceLocations(const SharedPtr<MG_State::GLState::ProgramObject>& stateProgramObject);
Uint m_backendProgramId = 0;
Uint m_backendGlobalUBOId = 0;
Int m_baseInstanceUniformLocation = -1;
Int m_drawIdUniformLocation = -1;
Int m_baseInstanceWordIndexUniformLocation = -1;
Int m_indirectParamsBinding = -1;
Uint32 m_snormFallbackClampOutputMask = 0;
Uint32 m_unormFallbackClampOutputMask = 0;
Bool m_isInitialized = false;
Int m_globalUboBackendBlockIndex = -1;
Int m_globalUboBackendBlockSize = 0;
Vector<Int> m_uniformBlockBackendIndices; // frontend block index -> backend index (-1 = absent)
Vector<SamplerUniformBinding> m_samplerUniformBindings;
Uint32 m_lastUploadedGlobalUboVersion = ~0u;
BufferImpl::UboRingAllocation m_globalUboRingAllocation;
Uint32 m_syncedLinkVersion = ~0u;
};
extern Uint32 g_snormFallbackClampOutputMask;
extern Uint32 g_unormFallbackClampOutputMask;
// Backend id of the last glUseProgram issued through this backend; lets Use()
// skip redundant rebinds. Reset to 0 wherever glUseProgram(0) is issued or the
// ES context is recreated.
extern Uint g_lastUsedBackendProgramId;
extern StateBackendObjectRegistry<MG_State::GLState::ProgramObject, BackendProgramObjectImpl>
extern UnorderedMap<SharedPtr<MG_State::GLState::ProgramObject>, SharedPtr<BackendProgramObjectImpl>>
g_backendProgramObjects;
} // namespace PrgramImpl
@@ -646,9 +188,9 @@ namespace MobileGL::MG_Backend::DirectGLES {
class BackendSamplerObject {
public:
BackendSamplerObject();
void SyncToBackend(const SharedPtr<MG_State::GLState::SamplerObject>& stateSamplerObject);
void SyncToBackend(SharedPtr<MG_State::GLState::SamplerObject>& stateSamplerObject);
void Bind(Uint unit);
Uint GetBackendSamplerId() const;
Uint GetBackendSamplerId();
private:
Uint m_backendSamplerId = 0;
@@ -661,7 +203,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
extern Array<BackendSamplerObject*, MG_State::GLState::TextureState::MAX_TEXTURE_IMAGE_UNITS>
g_boundSamplersCache;
extern StateBackendObjectRegistry<MG_State::GLState::SamplerObject, BackendSamplerObject>
extern UnorderedMap<SharedPtr<MG_State::GLState::SamplerObject>, SharedPtr<BackendSamplerObject>>
g_backendSamplerObjects;
} // namespace SamplerImpl
@@ -670,8 +212,8 @@ namespace MobileGL::MG_Backend::DirectGLES {
public:
BackendRenderbufferObject();
void SyncToBackend(const SharedPtr<MG_State::GLState::RenderbufferObject>& stateRBOObject);
Uint GetBackendRenderbufferId() const { return m_backendRBOId; }
void Bind() const;
Uint GetBackendRenderbufferId() { return m_backendRBOId; }
void Bind();
private:
Uint m_backendRBOId = 0;
@@ -679,10 +221,9 @@ namespace MobileGL::MG_Backend::DirectGLES {
TextureInternalFormat m_cacheInternalFormat = TextureInternalFormat::Unknown;
Int m_cacheWidth = 0;
Int m_cacheHeight = 0;
Int m_cacheSamples = 0;
};
extern StateBackendObjectRegistry<MG_State::GLState::RenderbufferObject, BackendRenderbufferObject>
extern UnorderedMap<SharedPtr<MG_State::GLState::RenderbufferObject>, SharedPtr<BackendRenderbufferObject>>
g_backendRenderbufferObjects;
} // namespace RenderbufferImpl
} // namespace MobileGL::MG_Backend::DirectGLES
+16 -676
View File
@@ -9,7 +9,6 @@
#include "DirectGLES.h"
#include "Utils.h"
#include "Managers.h"
#include "MG_Backend/BackendObjects.h"
#include "MG_Util/Converters/GLToMG/FramebufferEnumConverter.h"
#include "MG_Util/Texture/TextureFormatProcessor.h"
@@ -18,137 +17,28 @@
#include <MG_Util/Converters/GLToStr/GLEnumConverter.h>
#include <MG_Util/Converters/MGToGL/TextureEnumConverter.h>
#include <MG_Util/Converters/MGToGL/FramebufferEnumConverter.h>
#include <MG_Util/Math/HalfFloat.h>
#include <MG_Util/Math/SmallFloat.h>
#include <cmath>
namespace MobileGL::MG_Backend::DirectGLES {
namespace {
Flags<PixelFormatNormalizeOptionBit> GetForcedPixelFormatNormalizeOptions() {
Flags<PixelFormatNormalizeOptionBit> options;
if (g_GLESCapabilities.IsAngleRenderer) {
options |= PixelFormatNormalizeOptionBit::NoRgb16;
options |= PixelFormatNormalizeOptionBit::NoSnorm16;
options |= PixelFormatNormalizeOptionBit::NoSnorm8;
}
return options;
}
namespace BufferImpl {} // namespace BufferImpl
Flags<PixelFormatNormalizeOptionBit> GetDriverPixelFormatNormalizeOptions() {
Flags<PixelFormatNormalizeOptionBit> options = PixelFormatNormalizeOptionBit::NoDepthComponent32;
options |= PixelFormatNormalizeOptionBit::NoRGBA8Snorm;
options |= PixelFormatNormalizeOptionBit::NoRGB16Snorm;
if (!g_GLESCapabilities.SupportsNorm16Texture) {
options |= PixelFormatNormalizeOptionBit::NoNorm16;
}
return options;
}
Flags<PixelFormatNormalizeOptionBit> GetRuntimeFallbackNormalizeOptions(GLenum requestedInternalFormat) {
using namespace MG_Util::TextureFormatProcessor;
const Flags<PixelFormatNormalizeOptionBit> forcedOptions =
GetApplicablePixelFormatNormalizeOptions(requestedInternalFormat, GetForcedPixelFormatNormalizeOptions());
if (forcedOptions) {
return forcedOptions;
}
return GetApplicablePixelFormatNormalizeOptions(requestedInternalFormat,
GetDriverPixelFormatNormalizeOptions());
}
Bool HasCachedFormatCapability(TextureInternalFormat internalFormat,
SizeT targetIndex,
Bool caveat,
FormatCapability capability) {
if (!pActiveBackendObject || targetIndex >= kFormatCapabilityTargetCount) {
return false;
}
const SizeT formatIndex = static_cast<SizeT>(internalFormat);
if (formatIndex >= kFormatCapabilityFormatCount) {
return false;
}
const FormatCapabilityCache& cache = pActiveBackendObject->GetFormatCapabilities();
const FormatCapabilityFlags caps =
caveat ? cache.CaveatCaps[targetIndex][formatIndex] : cache.FullCaps[targetIndex][formatIndex];
return HasFormatCapability(caps, capability);
}
Bool HasAnyCachedFormatCapability(TextureInternalFormat internalFormat,
Bool caveat,
FormatCapability capability) {
for (SizeT targetIndex = 0; targetIndex < kFormatCapabilityTargetCount; ++targetIndex) {
if (HasCachedFormatCapability(internalFormat, targetIndex, caveat, capability)) {
return true;
}
}
return false;
}
Bool ShouldUseCaveatFormat(TextureInternalFormat internalFormat, SizeT targetIndex) {
if (targetIndex < kFormatCapabilityTargetCount) {
const Bool fullCreatable =
HasCachedFormatCapability(internalFormat, targetIndex, false, FormatCapability::Creatable);
const Bool caveatCreatable =
HasCachedFormatCapability(internalFormat, targetIndex, true, FormatCapability::Creatable);
const Bool fullRenderable =
HasCachedFormatCapability(internalFormat, targetIndex, false, FormatCapability::FramebufferRenderable);
const Bool caveatRenderable =
HasCachedFormatCapability(internalFormat, targetIndex, true, FormatCapability::FramebufferRenderable);
return (!fullCreatable && caveatCreatable) || (!fullRenderable && caveatRenderable);
}
if (HasAnyCachedFormatCapability(internalFormat, false, FormatCapability::Creatable)) {
return false;
}
return HasAnyCachedFormatCapability(internalFormat, true, FormatCapability::Creatable);
}
void GenerateFormatInfo(TextureInternalFormat internalFormat,
SizeT targetIndex,
GLenum* outInternalFormat,
GLenum* outFormat,
GLenum* outType) {
using namespace MobileGL::MG_Util::TextureFormatProcessor;
const GLenum requestedInternalFormat = MG_Util::ConvertTextureInternalFormatToGLEnum(internalFormat);
Flags<PixelFormatNormalizeOptionBit> options;
if (!pActiveBackendObject || ShouldUseCaveatFormat(internalFormat, targetIndex)) {
options = GetRuntimeFallbackNormalizeOptions(requestedInternalFormat);
}
NormalizePixelFormat(requestedInternalFormat, options, outInternalFormat, outFormat, outType);
}
} // namespace
namespace VertexArrayImpl {} // namespace VertexArrayImpl
namespace TextureImpl {
void GenerateTextureFormatInfo(TextureInternalFormat internalFormat, GLenum* outInternalFormat,
GLenum* outFormat, GLenum* outType, TextureTarget target) {
GLenum* outFormat, GLenum* outType) {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
#endif
const SizeT targetIndex =
target == TextureTarget::Unknown ? kFormatCapabilityTargetCount : GetFormatCapabilityTargetIndex(target);
GenerateFormatInfo(internalFormat, targetIndex, outInternalFormat, outFormat, outType);
}
void GenerateRenderbufferFormatInfo(TextureInternalFormat internalFormat, GLenum* outInternalFormat,
GLenum* outFormat, GLenum* outType) {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
#endif
GenerateFormatInfo(internalFormat, GetRenderbufferFormatCapabilityTargetIndex(), outInternalFormat,
outFormat, outType);
}
Bool ShouldUseCaveatTextureFormat(TextureInternalFormat internalFormat, TextureTarget target) {
const SizeT targetIndex =
target == TextureTarget::Unknown ? kFormatCapabilityTargetCount : GetFormatCapabilityTargetIndex(target);
return ShouldUseCaveatFormat(internalFormat, targetIndex);
}
Bool ShouldUseCaveatRenderbufferFormat(TextureInternalFormat internalFormat) {
return ShouldUseCaveatFormat(internalFormat, GetRenderbufferFormatCapabilityTargetIndex());
using namespace MobileGL::MG_Util::TextureFormatProcessor;
auto options = (g_GLESCapabilities.SupportsNorm16Texture) ? PixelFormatNormalizeOptionBit::None
: PixelFormatNormalizeOptionBit::NoNorm16;
NormalizePixelFormat(MG_Util::ConvertTextureInternalFormatToGLEnum(internalFormat), options,
outInternalFormat, outFormat, outType);
}
} // namespace TextureImpl
namespace FramebufferImpl {} // namespace FramebufferImpl
namespace PrgramImpl {
String ProcessOutColorLocations(const String& glslCode) {
#ifdef TRACY_ENABLE
@@ -218,163 +108,14 @@ namespace MobileGL::MG_Backend::DirectGLES {
return result;
}
String ClampNormFallbackOutputs(String glslCode, GLenum shaderType, Uint32 snormOutputMask,
Uint32 unormOutputMask) {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
#endif
const Uint32 outputMask = snormOutputMask | unormOutputMask;
if (shaderType != GL_FRAGMENT_SHADER || outputMask == 0) {
return glslCode;
}
const std::regex outputPattern(
R"(layout\s*\(\s*location\s*=\s*([0-9]+)\s*\)\s*out\s+(?:(?:lowp|mediump|highp)\s+)?vec4\s+([A-Za-z_][A-Za-z0-9_]*)\s*;)");
std::sregex_iterator outputIt(glslCode.begin(), glslCode.end(), outputPattern);
std::sregex_iterator outputEnd;
struct OutputClamp {
String Name;
Bool Signed;
};
Vector<OutputClamp> outputClamps;
for (; outputIt != outputEnd; ++outputIt) {
const Uint location = static_cast<Uint>(std::stoul((*outputIt)[1].str()));
if (location < 32 && (outputMask & (1u << location))) {
outputClamps.push_back({(*outputIt)[2].str(), static_cast<Bool>(snormOutputMask & (1u << location))});
}
}
if (outputClamps.empty()) {
return glslCode;
}
const std::regex mainPattern(R"(void\s+main\s*\([^)]*\)\s*\{)");
std::smatch mainMatch;
if (!std::regex_search(glslCode, mainMatch, mainPattern)) {
return glslCode;
}
SizeT bracePos = static_cast<SizeT>(mainMatch.position(0) + mainMatch.length(0) - 1);
Int depth = 0;
for (SizeT pos = bracePos; pos < glslCode.size(); ++pos) {
if (glslCode[pos] == '{') {
++depth;
} else if (glslCode[pos] == '}') {
--depth;
if (depth == 0) {
String clampLine;
for (const OutputClamp& outputClamp : outputClamps) {
const String minValue = outputClamp.Signed ? "-1.0" : "0.0";
clampLine += "\n " + outputClamp.Name + " = clamp(" + outputClamp.Name +
", vec4(" + minValue + "), vec4(1.0));";
}
clampLine += "\n";
glslCode.insert(pos, clampLine);
return glslCode;
}
}
}
return glslCode;
}
String ForceFlatIntegerVaryings(const String& glslCode, GLenum shaderType) {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
#endif
String result = glslCode;
const String integerType = R"((?:(?:lowp|mediump|highp)\s+)?(?:u?int|[iu]vec[234])\b)";
auto addFlatQualifier = [&result, &integerType](const String& qualifier) {
const std::regex pattern("(layout\\s*\\([^)]*\\)\\s*)(?!(?:flat|smooth|noperspective)\\s)(" +
qualifier + "\\s+" + integerType + ")");
result = std::regex_replace(result, pattern, "$1flat $2");
};
switch (shaderType) {
case GL_VERTEX_SHADER:
addFlatQualifier("out");
break;
case GL_GEOMETRY_SHADER:
addFlatQualifier("in");
addFlatQualifier("out");
break;
case GL_FRAGMENT_SHADER:
addFlatQualifier("in");
break;
default:
break;
}
return result;
}
String RemoveLayoutBinding(const String& glslCode) {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
#endif
// Sampler and uniform-block bindings are re-established at draw time through the
// API, so their layout qualifiers are stripped (they may exceed ES limits). SSBO
// blocks and image uniforms are different: ES has no glShaderStorageBlockBinding,
// and image units cannot be set with glUniform1i, so for those declarations the
// binding qualifier is the only binding mechanism and must be preserved.
static std::regex bindingRegex(R"(layout\s*\(\s*binding\s*=\s*\d+\s*\)\s*)");
String result = std::regex_replace(glslCode, bindingRegex, "");
static std::regex bindingRegex2(R"(layout\s*\(\s*binding\s*=\s*\d+\s*,)");
static std::regex keepBindingRegex(R"(\b(buffer|[iu]?image[A-Za-z0-9]*)\b)");
String result;
result.reserve(glslCode.size());
SizeT lineStart = 0;
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_search(line, keepBindingRegex)) {
line = std::regex_replace(line, bindingRegex, "");
line = std::regex_replace(line, bindingRegex2, "layout(");
}
result += line;
if (lastLine) {
break;
}
result += '\n';
lineStart = lineEnd + 1;
}
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;
}
result = std::regex_replace(result, bindingRegex2, "layout(");
return result;
}
} // namespace PrgramImpl
@@ -384,7 +125,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
#ifdef TRACY_ENABLE
ZoneScopedC(TRACY_ZONECOLOR_BACKEND);
#endif
for (GLenum err = g_GLESFuncs.glGetError(); err != GL_NO_ERROR; err = g_GLESFuncs.glGetError()) {
while (GLenum err = g_GLESFuncs.glGetError() != GL_NO_ERROR) {
MGLOG_E("-> GLES Error: %s", MG_Util::ConvertGLEnumToString(err).c_str());
}
}
@@ -425,6 +166,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
return GL_UNIFORM_BUFFER_BINDING;
case GL_FRAMEBUFFER:
return GL_FRAMEBUFFER_BINDING;
case GL_DRAW_FRAMEBUFFER:
return GL_DRAW_FRAMEBUFFER_BINDING;
case GL_READ_FRAMEBUFFER:
@@ -434,6 +176,7 @@ namespace MobileGL::MG_Backend::DirectGLES {
return GL_RENDERBUFFER_BINDING;
case GL_VERTEX_ARRAY:
return GL_VERTEX_ARRAY_BINDING;
case GL_VERTEX_ARRAY_BINDING:
return GL_VERTEX_ARRAY_BINDING;
@@ -487,407 +230,4 @@ namespace MobileGL::MG_Backend::DirectGLES {
}
}
} // namespace Utils
// ---- Client-format readback conversion helpers -------------------------------------------------
// ReadPixels/GetTexImage read a guaranteed wide RGBA(_INTEGER) layout from the ES driver and repack
// it on the CPU into the client's (format, type) layout. Everything here is pure byte shuffling so
// unit tests can assert the exact packed words; field positions follow GL 3.3 table 3.6 and mirror
// the GL CTS packed_pixels oracle (glcPackedPixelsTests.cpp pack_UNSIGNED_* helpers).
namespace ReadbackImpl {
using MG_Util::DecodeHalfBitsToFloat;
using MG_Util::EncodeFloatToHalfBits;
Bool GetReadbackChannelMapping(GLenum format, ReadbackChannelMapping& outMapping) {
switch (format) {
case GL_RED: outMapping = {{0, 0, 0, 0}, 1, false}; return true;
case GL_RED_INTEGER: outMapping = {{0, 0, 0, 0}, 1, true}; return true;
// Desktop-GL single-channel client formats (GL CTS packed_pixels rgba8_format_green/blue):
// the destination holds one component sourced from the named channel of the wide RGBA read.
// GL_ALPHA is mapped here from the raw enum because the state layer folds it into Red for the
// legacy alpha-texture upload hack.
case GL_GREEN: outMapping = {{1, 0, 0, 0}, 1, false}; return true;
case GL_GREEN_INTEGER: outMapping = {{1, 0, 0, 0}, 1, true}; return true;
case GL_BLUE: outMapping = {{2, 0, 0, 0}, 1, false}; return true;
case GL_BLUE_INTEGER: outMapping = {{2, 0, 0, 0}, 1, true}; return true;
case GL_ALPHA: outMapping = {{3, 0, 0, 0}, 1, false}; return true;
case GL_ALPHA_INTEGER: outMapping = {{3, 0, 0, 0}, 1, true}; return true;
case GL_RG: outMapping = {{0, 1, 0, 0}, 2, false}; return true;
case GL_RG_INTEGER: outMapping = {{0, 1, 0, 0}, 2, true}; return true;
case GL_RGB: outMapping = {{0, 1, 2, 0}, 3, false}; return true;
case GL_RGB_INTEGER: outMapping = {{0, 1, 2, 0}, 3, true}; return true;
case GL_BGR: outMapping = {{2, 1, 0, 0}, 3, false}; return true;
case GL_BGR_INTEGER: outMapping = {{2, 1, 0, 0}, 3, true}; return true;
case GL_RGBA: outMapping = {{0, 1, 2, 3}, 4, false}; return true;
case GL_RGBA_INTEGER: outMapping = {{0, 1, 2, 3}, 4, true}; return true;
case GL_BGRA: outMapping = {{2, 1, 0, 3}, 4, false}; return true;
case GL_BGRA_INTEGER: outMapping = {{2, 1, 0, 3}, 4, true}; return true;
default:
return false;
}
}
Bool GetPackedReadbackLayout(GLenum type, PackedReadbackLayout& out) {
switch (type) {
// Non-REV types pack the first format component starting at the most significant bit,
// *_REV types starting at the least significant bit (GL CTS pack_UNSIGNED_SHORT_5_6_5:
// R bits 15-11; pack_UNSIGNED_SHORT_1_5_5_5_REV: R bits 4-0, A bit 15).
case GL_UNSIGNED_BYTE_3_3_2: out = {3, {3, 3, 2, 0}, {5, 2, 0, 0}, 1, false}; return true;
case GL_UNSIGNED_BYTE_2_3_3_REV: out = {3, {3, 3, 2, 0}, {0, 3, 6, 0}, 1, false}; return true;
case GL_UNSIGNED_SHORT_5_6_5: out = {3, {5, 6, 5, 0}, {11, 5, 0, 0}, 2, false}; return true;
case GL_UNSIGNED_SHORT_5_6_5_REV: out = {3, {5, 6, 5, 0}, {0, 5, 11, 0}, 2, false}; return true;
case GL_UNSIGNED_SHORT_4_4_4_4: out = {4, {4, 4, 4, 4}, {12, 8, 4, 0}, 2, false}; return true;
case GL_UNSIGNED_SHORT_4_4_4_4_REV: out = {4, {4, 4, 4, 4}, {0, 4, 8, 12}, 2, false}; return true;
case GL_UNSIGNED_SHORT_5_5_5_1: out = {4, {5, 5, 5, 1}, {11, 6, 1, 0}, 2, false}; return true;
case GL_UNSIGNED_SHORT_1_5_5_5_REV: out = {4, {5, 5, 5, 1}, {0, 5, 10, 15}, 2, false}; return true;
case GL_UNSIGNED_INT_8_8_8_8: out = {4, {8, 8, 8, 8}, {24, 16, 8, 0}, 4, false}; return true;
case GL_UNSIGNED_INT_8_8_8_8_REV: out = {4, {8, 8, 8, 8}, {0, 8, 16, 24}, 4, false}; return true;
case GL_UNSIGNED_INT_10_10_10_2: out = {4, {10, 10, 10, 2}, {22, 12, 2, 0}, 4, false}; return true;
case GL_UNSIGNED_INT_2_10_10_10_REV: out = {4, {10, 10, 10, 2}, {0, 10, 20, 30}, 4, false}; return true;
// Packed-float RGB types: fields hold unsigned small floats; 5_9_9_9_REV's shared 5-bit
// exponent (bits 31-27) is emitted by EncodeSharedExponentRGB9E5, not a component field.
case GL_UNSIGNED_INT_10F_11F_11F_REV: out = {3, {11, 11, 10, 0}, {0, 11, 22, 0}, 4, true}; return true;
case GL_UNSIGNED_INT_5_9_9_9_REV: out = {3, {9, 9, 9, 0}, {0, 9, 18, 0}, 4, true}; return true;
default:
return false;
}
}
SizeT GetReadbackComponentSize(GLenum type) {
PackedReadbackLayout packedLayout{};
if (GetPackedReadbackLayout(type, packedLayout)) {
return packedLayout.byteSize;
}
switch (type) {
case GL_UNSIGNED_BYTE:
case GL_BYTE:
return 1;
case GL_UNSIGNED_SHORT:
case GL_SHORT:
case GL_HALF_FLOAT:
return 2;
case GL_UNSIGNED_INT:
case GL_INT:
case GL_FLOAT:
return 4;
default:
return 0;
}
}
SizeT GetReadbackDstPixelSize(const ReadbackChannelMapping& mapping, GLenum type) {
PackedReadbackLayout packedLayout{};
if (GetPackedReadbackLayout(type, packedLayout)) {
if (packedLayout.fieldCount != mapping.channelCount) {
return 0; // 3-field packed types pair with 3-component formats only, 4 with 4
}
if (mapping.isInteger && packedLayout.isFloatPacked) {
return 0; // packed-float RGB types never pair with integer formats
}
return packedLayout.byteSize;
}
if (mapping.isInteger && (type == GL_FLOAT || type == GL_HALF_FLOAT)) {
return 0;
}
const SizeT componentSize = GetReadbackComponentSize(type);
return componentSize == 0 ? 0 : static_cast<SizeT>(mapping.channelCount) * componentSize;
}
namespace {
void WritePackedReadbackWord(Uint8* dst, Uint32 word, SizeT byteSize) {
switch (byteSize) {
case 1: {
const auto out = static_cast<Uint8>(word);
Memcpy(dst, &out, sizeof(out));
break;
}
case 2: {
const auto out = static_cast<Uint16>(word);
Memcpy(dst, &out, sizeof(out));
break;
}
default:
Memcpy(dst, &word, sizeof(word));
break;
}
}
} // namespace
// Shared encoders live in MG_Util/Math/SmallFloat.h so the upload conversion
// (PixelStoreProcessor) uses byte-identical packing; kept exported here for unit tests.
Uint32 EncodeFloatToUnsignedF11(Float value) { return MG_Util::EncodeFloatToUnsignedF11(value); }
Uint32 EncodeFloatToUnsignedF10(Float value) { return MG_Util::EncodeFloatToUnsignedF10(value); }
Uint32 EncodeSharedExponentRGB9E5(const Float rgb[3]) { return MG_Util::EncodeSharedExponentRGB9E5(rgb); }
void ConvertWideReadbackRow(const Uint8* src, Uint8* dst, SizeT width, GLenum wideType,
const ReadbackChannelMapping& mapping, GLenum type) {
PackedReadbackLayout packedLayout{};
const Bool isPacked = GetPackedReadbackLayout(type, packedLayout);
const SizeT dstComponentSize = GetReadbackComponentSize(type);
const SizeT dstPixelBytes = GetReadbackDstPixelSize(mapping, type);
const SizeT srcPixelBytes = 4 * GetReadbackComponentSize(wideType);
for (SizeT col = 0; col < width; ++col) {
const Uint8* srcPixel = src + col * srcPixelBytes;
Uint8* dstPixel = dst + col * dstPixelBytes;
if (mapping.isInteger) {
Int64 srcValues[4];
for (Int c = 0; c < 4; ++c) {
srcValues[c] = wideType == GL_INT
? static_cast<Int64>(reinterpret_cast<const Int32*>(srcPixel)[c])
: static_cast<Int64>(reinterpret_cast<const Uint32*>(srcPixel)[c]);
}
if (isPacked) {
// Integer sources clamp each component to the unsigned range of its field
// (GL 3.3 section 4.3.1 final conversion).
Uint32 word = 0;
for (Int ch = 0; ch < packedLayout.fieldCount; ++ch) {
const Int64 fieldMax = (Int64{1} << packedLayout.width[ch]) - 1;
const auto v = static_cast<Uint32>(
std::clamp<Int64>(srcValues[mapping.sourceChannel[ch]], 0, fieldMax));
word |= v << packedLayout.shift[ch];
}
WritePackedReadbackWord(dstPixel, word, packedLayout.byteSize);
} else {
for (Int ch = 0; ch < mapping.channelCount; ++ch) {
const Int64 v = srcValues[mapping.sourceChannel[ch]];
Uint8* dstComponent = dstPixel + static_cast<SizeT>(ch) * dstComponentSize;
switch (type) {
case GL_UNSIGNED_BYTE:
*dstComponent = static_cast<Uint8>(std::clamp<Int64>(v, 0, 255));
break;
case GL_BYTE: {
const auto out = static_cast<Int8>(std::clamp<Int64>(v, -128, 127));
Memcpy(dstComponent, &out, sizeof(out));
break;
}
case GL_UNSIGNED_SHORT: {
const auto out = static_cast<Uint16>(std::clamp<Int64>(v, 0, 65535));
Memcpy(dstComponent, &out, sizeof(out));
break;
}
case GL_SHORT: {
const auto out = static_cast<Int16>(std::clamp<Int64>(v, -32768, 32767));
Memcpy(dstComponent, &out, sizeof(out));
break;
}
case GL_UNSIGNED_INT: {
const auto out = static_cast<Uint32>(std::clamp<Int64>(v, 0, 4294967295LL));
Memcpy(dstComponent, &out, sizeof(out));
break;
}
case GL_INT: {
const auto out =
static_cast<Int32>(std::clamp<Int64>(v, -2147483648LL, 2147483647LL));
Memcpy(dstComponent, &out, sizeof(out));
break;
}
default:
break;
}
}
}
} else {
Float srcValues[4];
switch (wideType) {
case GL_UNSIGNED_BYTE:
for (Int c = 0; c < 4; ++c) {
srcValues[c] = static_cast<Float>(srcPixel[c]) / 255.0f;
}
break;
case GL_BYTE:
for (Int c = 0; c < 4; ++c) {
srcValues[c] = std::max(
static_cast<Float>(reinterpret_cast<const Int8*>(srcPixel)[c]) / 127.0f, -1.0f);
}
break;
case GL_UNSIGNED_SHORT:
for (Int c = 0; c < 4; ++c) {
srcValues[c] =
static_cast<Float>(reinterpret_cast<const Uint16*>(srcPixel)[c]) / 65535.0f;
}
break;
case GL_SHORT:
for (Int c = 0; c < 4; ++c) {
srcValues[c] = std::max(
static_cast<Float>(reinterpret_cast<const Int16*>(srcPixel)[c]) / 32767.0f, -1.0f);
}
break;
case GL_HALF_FLOAT:
for (Int c = 0; c < 4; ++c) {
srcValues[c] = DecodeHalfBitsToFloat(reinterpret_cast<const Uint16*>(srcPixel)[c]);
}
break;
default: // GL_FLOAT
for (Int c = 0; c < 4; ++c) {
srcValues[c] = reinterpret_cast<const Float*>(srcPixel)[c];
}
break;
}
if (isPacked) {
Uint32 word = 0;
if (packedLayout.isFloatPacked) {
const Float fields[3] = {srcValues[mapping.sourceChannel[0]],
srcValues[mapping.sourceChannel[1]],
srcValues[mapping.sourceChannel[2]]};
word = type == GL_UNSIGNED_INT_5_9_9_9_REV
? EncodeSharedExponentRGB9E5(fields)
: (EncodeFloatToUnsignedF11(fields[0]) << packedLayout.shift[0]) |
(EncodeFloatToUnsignedF11(fields[1]) << packedLayout.shift[1]) |
(EncodeFloatToUnsignedF10(fields[2]) << packedLayout.shift[2]);
} else {
// Normalized encode: round(clamp(v, 0, 1) * (2^bits - 1)) into each field.
for (Int ch = 0; ch < packedLayout.fieldCount; ++ch) {
const auto fieldMax = static_cast<Float>((1u << packedLayout.width[ch]) - 1u);
const auto v = static_cast<Uint32>(std::llround(
std::clamp(srcValues[mapping.sourceChannel[ch]], 0.0f, 1.0f) * fieldMax));
word |= v << packedLayout.shift[ch];
}
}
WritePackedReadbackWord(dstPixel, word, packedLayout.byteSize);
} else {
for (Int ch = 0; ch < mapping.channelCount; ++ch) {
const Float v = srcValues[mapping.sourceChannel[ch]];
Uint8* dstComponent = dstPixel + static_cast<SizeT>(ch) * dstComponentSize;
switch (type) {
case GL_UNSIGNED_BYTE:
*dstComponent =
static_cast<Uint8>(std::llround(std::clamp(v, 0.0f, 1.0f) * 255.0));
break;
case GL_BYTE: {
const auto out =
static_cast<Int8>(std::llround(std::clamp(v, -1.0f, 1.0f) * 127.0));
Memcpy(dstComponent, &out, sizeof(out));
break;
}
case GL_UNSIGNED_SHORT: {
const auto out =
static_cast<Uint16>(std::llround(std::clamp(v, 0.0f, 1.0f) * 65535.0));
Memcpy(dstComponent, &out, sizeof(out));
break;
}
case GL_SHORT: {
const auto out =
static_cast<Int16>(std::llround(std::clamp(v, -1.0f, 1.0f) * 32767.0));
Memcpy(dstComponent, &out, sizeof(out));
break;
}
case GL_UNSIGNED_INT: {
const auto out = static_cast<Uint32>(
std::llround(static_cast<Double>(std::clamp(v, 0.0f, 1.0f)) * 4294967295.0));
Memcpy(dstComponent, &out, sizeof(out));
break;
}
case GL_INT: {
const auto out = static_cast<Int32>(
std::llround(static_cast<Double>(std::clamp(v, -1.0f, 1.0f)) * 2147483647.0));
Memcpy(dstComponent, &out, sizeof(out));
break;
}
case GL_FLOAT:
Memcpy(dstComponent, &v, sizeof(v));
break;
case GL_HALF_FLOAT: {
const Uint16 out = EncodeFloatToHalfBits(v);
Memcpy(dstComponent, &out, sizeof(out));
break;
}
default:
break;
}
}
}
}
}
}
static SizeT AlignReadbackRow(SizeT rowBytes, Int alignment) {
const SizeT align = alignment > 0 ? static_cast<SizeT>(alignment) : 1;
return (rowBytes + align - 1) / align * align;
}
// Repacks wide RGBA(_INTEGER) rows into the client's (format, type) layout, honoring the
// client-side PACK parameters and the bound pixel-pack buffer. `wide` holds
// `sliceHeight * sliceCount` rows of `width` texels (slice-major, tightly stacked),
// 4 components x GetReadbackComponentSize(wideType) bytes each.
// applyPackImageParams: GL_PACK_IMAGE_HEIGHT / GL_PACK_SKIP_IMAGES apply only to GetTexImage
// of 3D/array images; ReadPixels and 2D GetTexImage ignore them (GL 3.3 sections 4.3.1, 6.1.4).
// Per the GL addressing rules, slice k row j lands at
// SKIP_IMAGES*imageStride + SKIP_ROWS*rowStride + SKIP_PIXELS*pixelBytes
// + k*imageStride + j*rowStride, with imageStride = max(IMAGE_HEIGHT, sliceHeight)*rowStride.
Bool StoreWideRowsToClient(const Uint8* wide, GLenum wideType, GLsizei width, GLsizei sliceHeight,
GLsizei sliceCount, const ReadbackChannelMapping& mapping, GLenum type,
void* pixels, Bool applyPackImageParams) {
const SizeT dstPixelBytes = GetReadbackDstPixelSize(mapping, type);
if (dstPixelBytes == 0) {
return false;
}
PackedReadbackLayout packedLayout{};
const Bool isPackedType = GetPackedReadbackLayout(type, packedLayout);
const SizeT dstComponentSize = GetReadbackComponentSize(type);
const auto& pixelPackBufferObject =
MG_State::pGLContext->GetBufferBindingSlot(BufferTarget::PixelPack).GetBoundObject();
// Destination layout is computed from the client-side PACK parameters; only the actual pixel
// rows are written so skip regions of the destination stay untouched.
const auto packParams = MG_State::pGLContext->GetPixelStoreParameters(false);
const SizeT rowPixels = static_cast<SizeT>(packParams.RowLength > 0 ? packParams.RowLength : width);
const SizeT dstRowStride = AlignReadbackRow(rowPixels * dstPixelBytes, packParams.Alignment);
const SizeT imageRows =
applyPackImageParams && packParams.ImageHeight > 0
? static_cast<SizeT>(packParams.ImageHeight)
: static_cast<SizeT>(sliceHeight);
const SizeT dstImageStride = imageRows * dstRowStride;
const SizeT skipImages =
applyPackImageParams ? static_cast<SizeT>(std::max(packParams.SkipImages, 0)) : SizeT{0};
const SizeT dstSkipOffset = skipImages * dstImageStride +
static_cast<SizeT>(std::max(packParams.SkipRows, 0)) * dstRowStride +
static_cast<SizeT>(std::max(packParams.SkipPixels, 0)) * dstPixelBytes;
const SizeT dstRowBytes = static_cast<SizeT>(width) * dstPixelBytes;
const SizeT pboBaseOffset = reinterpret_cast<SizeT>(pixels); // with a PBO, `pixels` is an offset
if (pixelPackBufferObject) {
const SizeT requiredSize = pboBaseOffset + dstSkipOffset +
static_cast<SizeT>(sliceCount - 1) * dstImageStride +
static_cast<SizeT>(sliceHeight - 1) * dstRowStride + dstRowBytes;
if (requiredSize > pixelPackBufferObject->GetSize()) {
MGLOG_E("Readback conversion: pixel pack buffer is too small");
return true;
}
}
const SizeT srcComponentSize = GetReadbackComponentSize(wideType);
const SizeT srcPixelBytes = 4 * srcComponentSize;
Vector<Uint8> convertedRow(dstRowBytes);
for (GLsizei slice = 0; slice < sliceCount; ++slice) {
for (GLsizei row = 0; row < sliceHeight; ++row) {
const SizeT flatRow = static_cast<SizeT>(slice) * static_cast<SizeT>(sliceHeight) +
static_cast<SizeT>(row);
const Uint8* srcRow = wide + flatRow * static_cast<SizeT>(width) * srcPixelBytes;
ConvertWideReadbackRow(srcRow, convertedRow.data(), static_cast<SizeT>(width), wideType,
mapping, type);
if (packParams.SwapBytes) {
const SizeT groupSize = isPackedType ? packedLayout.byteSize : dstComponentSize;
if (groupSize > 1) {
for (SizeT offset = 0; offset + groupSize <= dstRowBytes; offset += groupSize) {
std::reverse(convertedRow.data() + offset, convertedRow.data() + offset + groupSize);
}
}
}
const SizeT dstOffset = dstSkipOffset + static_cast<SizeT>(slice) * dstImageStride +
static_cast<SizeT>(row) * dstRowStride;
if (pixelPackBufferObject) {
pixelPackBufferObject->WritebackFromBackend({convertedRow.data(), dstRowBytes},
pboBaseOffset + dstOffset);
} else {
Memcpy(static_cast<Uint8*>(pixels) + dstOffset, convertedRow.data(), dstRowBytes);
}
}
}
return true;
}
} // namespace ReadbackImpl
} // namespace MobileGL::MG_Backend::DirectGLES
+4 -66
View File
@@ -14,15 +14,15 @@ namespace MobileGL::MG_Backend::DirectGLES {
namespace DebugImpl {
class ErrorLopper {
public:
static void Loop(const std::function<void(GLenum)>&);
static void Clear();
void Loop(std::function<void(GLenum)>);
void Clear();
ErrorLopper();
~ErrorLopper();
};
class OpenGLScopeMarker {
public:
explicit OpenGLScopeMarker(const String& scopeName);
explicit OpenGLScopeMarker(String scopeName);
~OpenGLScopeMarker();
};
} // namespace DebugImpl
@@ -35,77 +35,15 @@ namespace MobileGL::MG_Backend::DirectGLES {
namespace TextureImpl {
void GenerateTextureFormatInfo(TextureInternalFormat internalFormat, GLenum* outInternalFormat,
GLenum* outFormat, GLenum* outType,
TextureTarget target = TextureTarget::Unknown);
void GenerateRenderbufferFormatInfo(TextureInternalFormat internalFormat, GLenum* outInternalFormat,
GLenum* outFormat, GLenum* outType);
Bool ShouldUseCaveatTextureFormat(TextureInternalFormat internalFormat, TextureTarget target);
Bool ShouldUseCaveatRenderbufferFormat(TextureInternalFormat internalFormat);
GLenum* outFormat, GLenum* outType);
} // namespace TextureImpl
namespace FramebufferImpl {} // namespace FramebufferImpl
// Pure CPU helpers of the client-format readback conversion (ReadPixels/GetTexImage repack a wide
// RGBA(_INTEGER) read into the caller's (format, type) layout). Kept context-free so unit tests can
// exercise the exact packing the GL CTS packed_pixels oracle compares against.
namespace ReadbackImpl {
struct ReadbackChannelMapping {
Int sourceChannel[4]; // RGBA source channel feeding each destination component
Int channelCount; // destination component count
Bool isInteger;
};
Bool GetReadbackChannelMapping(GLenum format, ReadbackChannelMapping& outMapping);
// Byte size of one destination component of `type`; packed types report the packed word size.
// 0 = type not supported by the conversion path.
SizeT GetReadbackComponentSize(GLenum type);
// Bit-field layout of a GL packed pixel type. width/shift are indexed in the client format's
// component order (matching ReadbackChannelMapping); shift is the LSB position of the field in
// the packed word: non-REV types pack the first component from the MSB, *_REV types from the
// LSB (GL 3.3 table 3.6; field positions mirror the GL CTS glcPackedPixelsTests pack_* oracle).
struct PackedReadbackLayout {
Int fieldCount; // format components stored in the packed word
Int width[4]; // bit width of each component's field
Int shift[4]; // LSB bit position of each component's field
SizeT byteSize; // packed word size in bytes (1, 2 or 4)
Bool isFloatPacked; // 10F_11F_11F_REV / 5_9_9_9_REV: fields hold unsigned small floats
};
Bool GetPackedReadbackLayout(GLenum type, PackedReadbackLayout& out);
// Unsigned small-float encoders (EXT_packed_float / EXT_texture_shared_exponent semantics).
Uint32 EncodeFloatToUnsignedF11(Float value);
Uint32 EncodeFloatToUnsignedF10(Float value);
Uint32 EncodeSharedExponentRGB9E5(const Float rgb[3]);
// Destination bytes per pixel for a (format mapping, type) readback pair; 0 when the pair is
// not convertible (unknown type, packed field count != format component count, floating-point
// or packed-float type with an integer format).
SizeT GetReadbackDstPixelSize(const ReadbackChannelMapping& mapping, GLenum type);
// Repacks one row of wide RGBA(_INTEGER) texels (4 components of wideType each) into the
// client's (format, type) layout. src holds width * 4 * GetReadbackComponentSize(wideType)
// bytes, dst receives width * GetReadbackDstPixelSize(mapping, type) bytes.
void ConvertWideReadbackRow(const Uint8* src, Uint8* dst, SizeT width, GLenum wideType,
const ReadbackChannelMapping& mapping, GLenum type);
// Stores wide RGBA(_INTEGER) rows into the client pointer or the bound PACK pixel buffer,
// honoring the client-side PACK pixel-store parameters (row length, alignment, skips,
// swap-bytes, and - when applyPackImageParams - image height/skip images). Shared by the
// DirectGLES and DirectVulkan readback conversion paths.
Bool StoreWideRowsToClient(const Uint8* wide, GLenum wideType, GLsizei width, GLsizei sliceHeight,
GLsizei sliceCount, const ReadbackChannelMapping& mapping, GLenum type,
void* pixels, Bool applyPackImageParams);
} // namespace ReadbackImpl
namespace PrgramImpl {
String ProcessOutColorLocations(const String& glslCode);
String ForceSupporterOutput(const String& glslCode);
String ClampNormFallbackOutputs(String glslCode, GLenum shaderType, Uint32 snormOutputMask,
Uint32 unormOutputMask);
String ForceFlatIntegerVaryings(const String& glslCode, GLenum shaderType);
String RemoveLayoutBinding(const String& glslCode);
String RemoveClipDistanceRedeclaration(const String& glslCode);
} // namespace PrgramImpl
namespace Utils {
@@ -1,840 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/BackendObject_DirectVulkan.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 "BackendObject_DirectVulkan.h"
#include "MG_Backend/BackendObject.h"
#include "DirectVulkan.h"
#include "MG_State/GLState/FramebufferState/FramebufferObject.h"
#include "MG_State/GLState/TextureState/TextureState.h"
#include "MG_Util/Classifiers/TextureEnumClassifier.h"
#include "MG_Util/Converters/MGToGL/TextureEnumConverter.h"
#include "MG_Util/Converters/MGToStr/TextureEnumConverter.h"
#include "MG_Util/Converters/MGToVk/TextureEnumConverter.h"
#include "MG_Util/Texture/TextureFormatProcessor.h"
#include <Config.h>
#include <cstdlib>
#include <cstring>
namespace MobileGL::MG_Backend::DirectVulkan {
namespace {
Bool IsR11G11B10FFallbackEnabled() {
return MG_Config::Features.MagmaR11G11B10FFallback;
}
Bool IsReleaseCurrentRequest(EGLDisplay dpy, EGLSurface draw, EGLSurface read, EGLContext ctx) {
(void)dpy;
return draw == EGL_NO_SURFACE && read == EGL_NO_SURFACE && ctx == EGL_NO_CONTEXT;
}
Bool IsFormatIndexValid(TextureInternalFormat format) {
return format != TextureInternalFormat::Unknown && static_cast<Int>(format) >= 0 &&
static_cast<SizeT>(format) < kFormatCapabilityFormatCount;
}
Bool IsLayeredTarget(TextureTarget target) {
return target == TextureTarget::Texture3D || target == TextureTarget::Texture1DArray ||
target == TextureTarget::Texture2DArray || target == TextureTarget::TextureCubeMap ||
target == TextureTarget::TextureCubeMapArray ||
target == TextureTarget::Texture2DMultisampleArray;
}
Bool IsMultisampleTarget(TextureTarget target) {
return target == TextureTarget::Texture2DMultisample ||
target == TextureTarget::Texture2DMultisampleArray;
}
Bool IsTextureBufferTarget(TextureTarget target) {
return target == TextureTarget::TextureBuffer;
}
Bool IsIntegerInternalFormat(TextureInternalFormat format) {
const GLenum glFormat = MG_Util::ConvertTextureInternalFormatToGLEnum(format);
GLenum normalizedInternalFormat = glFormat;
GLenum imageFormat = GL_RGBA;
GLenum imageType = GL_UNSIGNED_BYTE;
MG_Util::TextureFormatProcessor::NormalizePixelFormat(
glFormat, PixelFormatNormalizeOptionBit::None, &normalizedInternalFormat, &imageFormat, &imageType);
return imageFormat == GL_RED_INTEGER || imageFormat == GL_RG_INTEGER || imageFormat == GL_RGB_INTEGER ||
imageFormat == GL_RGBA_INTEGER;
}
FormatCapabilityFlags GetAttachmentCaps(TextureInternalFormat format) {
FormatCapabilityFlags caps = FormatCapability::FramebufferRenderable;
const Bool isDepth = MG_Util::IsDepthFormatInternalFormat(format);
const Bool isStencil = MG_Util::IsStencilFormatInternalFormat(format);
if (!isDepth && !isStencil) {
caps |= FormatCapability::ColorAttachment;
}
if (isDepth) {
caps |= FormatCapability::DepthAttachment;
}
if (isStencil) {
caps |= FormatCapability::StencilAttachment;
}
return caps;
}
FormatCapabilityFlags BuildVulkanCaps(TextureInternalFormat logicalFormat,
TextureTarget target,
VkFormatFeatureFlags features) {
FormatCapabilityFlags caps;
const Bool isDepth = MG_Util::IsDepthFormatInternalFormat(logicalFormat);
const Bool isStencil = MG_Util::IsStencilFormatInternalFormat(logicalFormat);
const Bool isInteger = IsIntegerInternalFormat(logicalFormat);
if (IsTextureBufferTarget(target)) {
if ((features & VK_FORMAT_FEATURE_UNIFORM_TEXEL_BUFFER_BIT) != 0) {
caps |= FormatCapability::Creatable;
caps |= FormatCapability::Sampled;
caps |= FormatCapability::TextureBuffer;
}
return caps;
}
const Bool sampled = (features & VK_FORMAT_FEATURE_SAMPLED_IMAGE_BIT) != 0;
const Bool linearFilter = (features & VK_FORMAT_FEATURE_SAMPLED_IMAGE_FILTER_LINEAR_BIT) != 0;
const Bool colorRenderable = (features & VK_FORMAT_FEATURE_COLOR_ATTACHMENT_BIT) != 0;
const Bool depthStencilRenderable =
(features & VK_FORMAT_FEATURE_DEPTH_STENCIL_ATTACHMENT_BIT) != 0;
const Bool renderable = (isDepth || isStencil) ? depthStencilRenderable : colorRenderable;
if (sampled || renderable) {
caps |= FormatCapability::Creatable;
}
if (sampled) {
caps |= FormatCapability::Sampled;
if (linearFilter && !isInteger && !isStencil) {
caps |= FormatCapability::LinearFilter;
}
if (!isStencil && (features & VK_FORMAT_FEATURE_BLIT_SRC_BIT) != 0 &&
(features & VK_FORMAT_FEATURE_BLIT_DST_BIT) != 0) {
caps |= FormatCapability::GenerateMipmap;
}
if (!isInteger && !isDepth && !isStencil) {
caps |= FormatCapability::TextureGather;
}
if (isDepth && !isStencil) {
caps |= FormatCapability::TextureShadow;
}
}
if (renderable) {
caps |= GetAttachmentCaps(logicalFormat);
if (IsLayeredTarget(target)) {
caps |= FormatCapability::FramebufferLayered;
}
}
if (IsMultisampleTarget(target)) {
caps |= FormatCapability::MultisampleTexture;
}
return caps;
}
Optional<TextureInternalFormat> ResolveVulkanFallbackLogicalFormat(TextureInternalFormat format) {
switch (format) {
case TextureInternalFormat::RGB:
case TextureInternalFormat::RGB8:
return TextureInternalFormat::RGBA8;
// Legacy low-bit-depth formats with no (or rarely supported) native Vulkan
// encoding; a wider normalized fallback keeps at least the required precision.
case TextureInternalFormat::R3G3B2:
case TextureInternalFormat::RGB4:
case TextureInternalFormat::RGB5:
case TextureInternalFormat::RGBA2:
case TextureInternalFormat::RGBA4:
case TextureInternalFormat::RGB5A1:
return TextureInternalFormat::RGBA8;
case TextureInternalFormat::RGB10:
return TextureInternalFormat::RGB10A2;
case TextureInternalFormat::RGB12:
case TextureInternalFormat::RGBA12:
return TextureInternalFormat::RGBA16;
case TextureInternalFormat::SRGB8:
return TextureInternalFormat::SRGB8Alpha8;
case TextureInternalFormat::RGB8Snorm:
return TextureInternalFormat::RGBA8Snorm;
case TextureInternalFormat::RGB16:
return TextureInternalFormat::RGBA16;
case TextureInternalFormat::RGB16Snorm:
return TextureInternalFormat::RGBA16Snorm;
case TextureInternalFormat::RGB16F:
return TextureInternalFormat::RGBA16F;
case TextureInternalFormat::R11FG11FB10F:
if (IsR11G11B10FFallbackEnabled()) {
return TextureInternalFormat::RGBA16F;
}
return Nullopt;
case TextureInternalFormat::RGB32F:
return TextureInternalFormat::RGBA32F;
case TextureInternalFormat::RGB8I:
return TextureInternalFormat::RGBA8I;
case TextureInternalFormat::RGB8UI:
return TextureInternalFormat::RGBA8UI;
case TextureInternalFormat::RGB16I:
return TextureInternalFormat::RGBA16I;
case TextureInternalFormat::RGB16UI:
return TextureInternalFormat::RGBA16UI;
case TextureInternalFormat::RGB32I:
return TextureInternalFormat::RGBA32I;
case TextureInternalFormat::RGB32UI:
return TextureInternalFormat::RGBA32UI;
default:
return Nullopt;
}
}
Optional<VkFormat> ResolveVulkanFallbackFormat(TextureInternalFormat format) {
const Optional<TextureInternalFormat> fallbackLogicalFormat = ResolveVulkanFallbackLogicalFormat(format);
if (!fallbackLogicalFormat) {
return Nullopt;
}
return MG_Util::ConvertTextureInternalFormatToVkEnum(*fallbackLogicalFormat);
}
Bool HasNewCaveatFormatCaps(FormatCapabilityFlags nativeCaps, FormatCapabilityFlags fallbackCaps) {
for (FormatCapability capability : kReportedFormatCapabilities) {
if (HasFormatCapability(fallbackCaps, capability) &&
!HasFormatCapability(nativeCaps, capability)) {
return true;
}
}
return false;
}
void LogVulkanFormatCaveat(TextureInternalFormat logicalFormat,
SizeT targetIndex,
TextureInternalFormat fallbackFormat) {
MGLOG_D("Caveat: %s %s not fully supported. Reason: native Vulkan format is not fully supported. Fallback: %s",
GetFormatCapabilityTargetName(targetIndex).c_str(),
MG_Util::ConvertTextureInternalFormatToString(logicalFormat).c_str(),
MG_Util::ConvertTextureInternalFormatToString(fallbackFormat).c_str());
}
Vector<Int> BuildSampleCounts(Int maxSamples) {
Vector<Int> counts;
for (Int samples = std::max(maxSamples, 1); samples > 1; samples >>= 1) {
counts.push_back(samples);
}
counts.push_back(1);
return counts;
}
void PopulateFormatCapabilitiesImpl(VkPhysicalDevice physicalDevice,
PFN_vkGetPhysicalDeviceFormatProperties getFormatProperties,
const MG_External::VulkanCapabilities& capabilities,
FormatCapabilityCache& cache) {
cache.Clear();
if (physicalDevice == VK_NULL_HANDLE || getFormatProperties == nullptr) {
return;
}
for (SizeT formatIndex = 0; formatIndex < kFormatCapabilityFormatCount; ++formatIndex) {
const auto logicalFormat = static_cast<TextureInternalFormat>(formatIndex);
if (!IsFormatIndexValid(logicalFormat)) {
continue;
}
VkFormat nativeFormat = MG_Util::ConvertTextureInternalFormatToVkEnum(logicalFormat);
const Optional<TextureInternalFormat> fallbackLogicalFormat =
ResolveVulkanFallbackLogicalFormat(logicalFormat);
VkFormat fallbackFormat = ResolveVulkanFallbackFormat(logicalFormat).value_or(VK_FORMAT_UNDEFINED);
VkFormatProperties nativeProperties{};
if (nativeFormat != VK_FORMAT_UNDEFINED) {
getFormatProperties(physicalDevice, nativeFormat, &nativeProperties);
}
VkFormatProperties fallbackProperties{};
if (fallbackFormat != VK_FORMAT_UNDEFINED && fallbackFormat != nativeFormat) {
getFormatProperties(physicalDevice, fallbackFormat, &fallbackProperties);
}
for (SizeT targetIndex = 0; targetIndex < kFormatCapabilityTextureTargetCount; ++targetIndex) {
const auto target = static_cast<TextureTarget>(targetIndex);
const VkFormatFeatureFlags nativeFeatures =
IsTextureBufferTarget(target) ? nativeProperties.bufferFeatures
: nativeProperties.optimalTilingFeatures;
FormatCapabilityFlags nativeCaps = BuildVulkanCaps(logicalFormat, target, nativeFeatures);
cache.FullCaps[targetIndex][formatIndex] |= nativeCaps;
const VkFormatFeatureFlags fallbackFeatures =
IsTextureBufferTarget(target) ? fallbackProperties.bufferFeatures
: fallbackProperties.optimalTilingFeatures;
FormatCapabilityFlags fallbackCaps = BuildVulkanCaps(logicalFormat, target, fallbackFeatures);
if (fallbackFormat != VK_FORMAT_UNDEFINED && fallbackFormat != nativeFormat) {
cache.CaveatCaps[targetIndex][formatIndex] |= fallbackCaps;
if (fallbackLogicalFormat && HasNewCaveatFormatCaps(nativeCaps, fallbackCaps)) {
LogVulkanFormatCaveat(logicalFormat, targetIndex, *fallbackLogicalFormat);
}
}
if (HasFormatCapability(nativeCaps | fallbackCaps, FormatCapability::MultisampleTexture)) {
const Bool isDepth = MG_Util::IsDepthFormatInternalFormat(logicalFormat);
const Bool isStencil = MG_Util::IsStencilFormatInternalFormat(logicalFormat);
const Bool isInteger = IsIntegerInternalFormat(logicalFormat);
Int maxSamples = capabilities.MaxColorTextureSamples;
if (isDepth || isStencil) {
maxSamples = capabilities.MaxDepthTextureSamples;
} else if (isInteger) {
maxSamples = capabilities.MaxIntegerSamples;
}
cache.SampleCounts[targetIndex][formatIndex] = BuildSampleCounts(maxSamples);
}
}
const SizeT renderbufferTargetIndex = GetRenderbufferFormatCapabilityTargetIndex();
FormatCapabilityFlags renderbufferCaps =
BuildVulkanCaps(logicalFormat, TextureTarget::Texture2D, nativeProperties.optimalTilingFeatures);
renderbufferCaps &= FormatCapability::Creatable;
if ((nativeProperties.optimalTilingFeatures &
(VK_FORMAT_FEATURE_COLOR_ATTACHMENT_BIT | VK_FORMAT_FEATURE_DEPTH_STENCIL_ATTACHMENT_BIT)) != 0) {
renderbufferCaps |= GetAttachmentCaps(logicalFormat);
renderbufferCaps |= FormatCapability::MultisampleRenderbuffer;
}
cache.FullCaps[renderbufferTargetIndex][formatIndex] |= renderbufferCaps;
if (fallbackFormat != VK_FORMAT_UNDEFINED && fallbackFormat != nativeFormat) {
FormatCapabilityFlags fallbackRenderbufferCaps =
BuildVulkanCaps(logicalFormat, TextureTarget::Texture2D,
fallbackProperties.optimalTilingFeatures);
fallbackRenderbufferCaps &= FormatCapability::Creatable;
if ((fallbackProperties.optimalTilingFeatures &
(VK_FORMAT_FEATURE_COLOR_ATTACHMENT_BIT | VK_FORMAT_FEATURE_DEPTH_STENCIL_ATTACHMENT_BIT)) !=
0) {
fallbackRenderbufferCaps |= GetAttachmentCaps(logicalFormat);
fallbackRenderbufferCaps |= FormatCapability::MultisampleRenderbuffer;
}
cache.CaveatCaps[renderbufferTargetIndex][formatIndex] |= fallbackRenderbufferCaps;
if (fallbackLogicalFormat &&
HasNewCaveatFormatCaps(renderbufferCaps, fallbackRenderbufferCaps)) {
LogVulkanFormatCaveat(logicalFormat, renderbufferTargetIndex, *fallbackLogicalFormat);
}
}
const FormatCapabilityFlags rbCaps = cache.FullCaps[renderbufferTargetIndex][formatIndex] |
cache.CaveatCaps[renderbufferTargetIndex][formatIndex];
if (HasFormatCapability(rbCaps, FormatCapability::MultisampleRenderbuffer)) {
cache.SampleCounts[renderbufferTargetIndex][formatIndex] =
BuildSampleCounts(capabilities.MaxFramebufferSamples);
}
}
}
} // namespace
void PopulateFormatCapabilities(VkPhysicalDevice physicalDevice,
PFN_vkGetPhysicalDeviceFormatProperties getFormatProperties,
const MG_External::VulkanCapabilities& capabilities,
FormatCapabilityCache& cache) {
PopulateFormatCapabilitiesImpl(physicalDevice, getFormatProperties, capabilities, cache);
}
BackendObject_DirectVulkan::~BackendObject_DirectVulkan() = default;
BackendObject_DirectVulkan::BackendObject_DirectVulkan(): m_rendererInfo{GetRendererIdentity()} {}
Bool BackendObject_DirectVulkan::InitWindowSurface() {
if (!m_windowHandle.Handle) {
MGLOG_E("Cannot initialize DirectVulkan window surface: native window handle is null");
return false;
}
auto nativeWindow = reinterpret_cast<NativeWindowType>(m_windowHandle.Handle);
// Any renderer instance this assignment replaces is destroyed here;
// fence/timer-query handles stamped with the old generation go stale.
BumpRendererGeneration();
pVulkanRenderer = MakeUnique<MG_Backend::DirectVulkan::VulkanRenderer>(nativeWindow);
MOBILEGL_ASSERT(pVulkanRenderer != nullptr, "InitWindowSurface: VulkanRenderer creation failed");
pVulkanRenderer->Initialize();
return true;
}
Bool BackendObject_DirectVulkan::InitPbufferSurface(EGLint width, EGLint height) {
VulkanRendererConfig config;
config.SurfaceWidth = static_cast<Uint32>(std::max<EGLint>(width, 1));
config.SurfaceHeight = static_cast<Uint32>(std::max<EGLint>(height, 1));
// Any renderer instance this assignment replaces is destroyed here;
// fence/timer-query handles stamped with the old generation go stale.
BumpRendererGeneration();
pVulkanRenderer = MakeUnique<MG_Backend::DirectVulkan::VulkanRenderer>(NativeWindowType{}, config);
MOBILEGL_ASSERT(pVulkanRenderer != nullptr, "InitPbufferSurface: VulkanRenderer creation failed");
pVulkanRenderer->Initialize();
return true;
}
void BackendObject_DirectVulkan::Initialize() {
m_initialized = true;
}
Bool BackendObject_DirectVulkan::InitCapabilities() {
if (!m_initialized) {
MGLOG_E("Cannot initialize capabilities before backend is initialized");
return false;
}
if (!pVulkanRenderer) {
MGLOG_E("Cannot initialize capabilities: Vulkan renderer has not been created");
return false;
}
const auto& physicalDevice = pVulkanRenderer->GetPhysicalDevice();
if (!MG_Util::BackendLoader::QueryVulkanCapabilities(m_vulkanCaps, pVulkanRenderer->GetInstance(),
physicalDevice.handle)) {
MGLOG_W("DirectVulkan: failed to query extended Vulkan capabilities, using basic properties");
MG_Util::BackendLoader::FillInVulkanCapabilities(m_vulkanCaps, physicalDevice.properties);
}
UpdateDynamicBackendParameters();
UpdateAdvertisedExtensions();
PopulateFormatCapabilities(physicalDevice.handle, vkGetPhysicalDeviceFormatProperties, m_vulkanCaps,
MutableFormatCapabilities());
PrintFormatCapabilities(GetFormatCapabilities());
return true;
}
Bool BackendObject_DirectVulkan::InitializeEGLDisplay(EGLDisplay dpy, EGLint* major, EGLint* minor) {
if (!m_initialized) {
MGLOG_E("DirectVulkan backend not initialized");
return false;
}
return BackendObject::InitializeEGLDisplay(dpy, major, minor);
}
Bool BackendObject_DirectVulkan::CreateEGLWindowSurface(EGLSurface surface, const WindowHandle& handle) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
if (!m_initialized) {
MGLOG_E("DirectVulkan backend not initialized");
return false;
}
if (!handle.Handle || (handle.Backend != WindowBackend::Android &&
handle.Backend != WindowBackend::X11 &&
handle.Backend != WindowBackend::MetalLayer &&
handle.Backend != WindowBackend::Win32)) {
MGLOG_E("DirectVulkan backend only supports Android, X11, CAMetalLayer, and Win32 native windows");
return false;
}
return RegisterEGLWindowSurface(surface, handle);
}
Bool BackendObject_DirectVulkan::ResizeEGLWindowSurface(EGLSurface surface, Uint32 width, Uint32 height) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
if (!m_initialized) {
MGLOG_E("DirectVulkan backend not initialized");
return false;
}
if (!BackendObject::ResizeEGLWindowSurface(surface, width, height)) {
return false;
}
if (pVulkanRenderer && m_eglSurface == surface) {
pVulkanRenderer->RequestSwapchainResize(width, height);
}
return true;
}
Bool BackendObject_DirectVulkan::CreateEGLPbufferSurface(EGLSurface surface, EGLint width, EGLint height) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
if (!m_initialized) {
MGLOG_E("DirectVulkan backend not initialized");
return false;
}
return RegisterEGLPbufferSurface(surface, width, height);
}
Bool BackendObject_DirectVulkan::MakeEGLCurrent(EGLDisplay dpy, EGLSurface draw, EGLSurface read, EGLContext ctx) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
return BackendObject::MakeEGLCurrent(dpy, draw, read, ctx);
}
Bool BackendObject_DirectVulkan::SwapEGLBuffers(EGLDisplay dpy, EGLSurface draw) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
if (!pVulkanRenderer) {
MGLOG_E("DirectVulkan renderer is not initialized");
return false;
}
return BackendObject::SwapEGLBuffers(dpy, draw);
}
void BackendObject_DirectVulkan::ReleaseEGLSurface(EGLSurface surface) {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
BackendObject::ReleaseEGLSurface(surface);
}
void BackendObject_DirectVulkan::ReleaseEGLResources() {
const std::lock_guard<std::recursive_mutex> lock(m_eglStateMutex);
// Outstanding fence/timer-query handles now refer to a dead renderer;
// treat them as signaled/available with zero results from here on.
BumpRendererGeneration();
pVulkanRenderer.reset();
BackendObject::ReleaseEGLResources();
}
void BackendObject_DirectVulkan::OnEGLSurfaceReleased(EGLSurface surface) {
(void)surface;
// Outstanding fence/timer-query handles now refer to a dead renderer;
// treat them as signaled/available with zero results from here on.
BumpRendererGeneration();
pVulkanRenderer.reset();
}
const RendererInfo& BackendObject_DirectVulkan::GetRendererInfo() const {
return m_rendererInfo;
}
String BackendObject_DirectVulkan::GetBackendAPIVersionString() const {
if (!m_initialized) {
return "<uninitialized DirectVulkan backend>";
}
return FormatBackendAPIVersionString(m_vulkanCaps.DeviceName, m_vulkanCaps.VulkanAPIVersion.toString(),
m_vulkanCaps.DriverVersionString);
}
const RendererInfo& GetRendererIdentity() {
static const RendererInfo rendererInfo = {
.RendererName = "Magma",
.BackendName = "Direct (Vulkan)",
.ExtraVendor = Nullopt,
.RendererGLInfo =
{
.TargetGLVersion = {3, 3, 0},
.TargetGLSLVersion = {4, 6, 0},
// Baseline advertisement (no shader subgroup, no timer queries); a
// live backend reconciles its copy in UpdateAdvertisedExtensions.
.Extensions = BuildAdvertisedExtensions(false, false, false),
.IsCompatibilityProfile = false
},
.StaticBackendCapability = {.AllowVSOnlyPrograms = false}};
return rendererInfo;
}
Vector<GLExtension> BuildAdvertisedExtensions(Bool shaderSubgroupSupported, Bool timerQueriesSupported,
Bool anisotropicFilteringSupported) {
Vector<GLExtension> extensions = {V_OpenGL30, V_OpenGL31, V_OpenGL32,
V_OpenGL33, E_GL_ARB_draw_buffers_blend, E_GL_ARB_compute_shader,
E_GL_ARB_shader_storage_buffer_object, E_GL_ARB_shader_image_load_store,
E_GL_ARB_program_interface_query, E_GL_ARB_framebuffer_object,
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_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};
if (shaderSubgroupSupported && !MG_Config::Features.DisableSubgroup) {
extensions.push_back(E_GL_KHR_shader_subgroup);
}
// GL_ARB_timer_query gates MC's F3 GPU% (LWJGL checks the extension string);
// only advertised when the device actually supports timestamp queries and the
// MOBILEGL_DISABLE_TIMERQUERY escape hatch is off.
if (timerQueriesSupported && !MG_Config::Features.DisableTimerQuery) {
extensions.push_back(E_GL_ARB_timer_query);
}
// Only advertised when the samplerAnisotropy device feature was granted: without it the
// sampler state is accepted but never applied, and an app trusting the string (LWJGL builds
// GLCapabilities from it) would think it enabled anisotropic filtering.
if (anisotropicFilteringSupported) {
extensions.push_back(E_GL_EXT_texture_filter_anisotropic);
extensions.push_back(E_GL_ARB_texture_filter_anisotropic);
}
return extensions;
}
String FormatBackendAPIVersionString(const String& deviceName, const String& vulkanApiVersionString,
const String& driverVersionString) {
// Format:
// <GPU Name>, Vulkan <Vulkan Version>, Driver <Driver Version>
return deviceName + ", Vulkan " + vulkanApiVersionString + ", Driver " + driverVersionString;
}
BackendType BackendObject_DirectVulkan::GetBackendType() const {
return BackendType::DirectVulkan;
}
const GlobalBackendFunctionsTable& BackendObject_DirectVulkan::GetBackendFunctions() const {
static GlobalBackendFunctionsTable funcsTable;
static Bool funcsTableInitialized = false;
if (!funcsTableInitialized) {
funcsTable.Present = Present;
funcsTable.GL.DrawArrays = DrawArrays;
funcsTable.GL.DrawElements = DrawElements;
funcsTable.GL.DrawElementsBaseVertex = DrawElementsBaseVertex;
funcsTable.GL.MultiDrawArrays = MultiDrawArrays;
funcsTable.GL.MultiDrawElements = MultiDrawElements;
funcsTable.GL.MultiDrawElementsBaseVertex = MultiDrawElementsBaseVertex;
funcsTable.GL.MultiDrawElementsIndirect = MultiDrawElementsIndirect;
funcsTable.GL.MultiDrawArraysIndirect = MultiDrawArraysIndirect;
funcsTable.GL.MultiDrawElementsIndirectCount = MultiDrawElementsIndirectCount;
funcsTable.GL.MultiDrawArraysIndirectCount = MultiDrawArraysIndirectCount;
funcsTable.GL.DrawRangeElementsBaseVertex = DrawRangeElementsBaseVertex;
funcsTable.GL.DrawRangeElements = DrawRangeElements;
funcsTable.GL.DrawElementsInstancedBaseVertexBaseInstance = DrawElementsInstancedBaseVertexBaseInstance;
funcsTable.GL.DrawElementsInstancedBaseVertex = DrawElementsInstancedBaseVertex;
funcsTable.GL.DrawElementsInstancedBaseInstance = DrawElementsInstancedBaseInstance;
funcsTable.GL.DrawElementsInstanced = DrawElementsInstanced;
funcsTable.GL.DrawArraysInstancedBaseInstance = DrawArraysInstancedBaseInstance;
funcsTable.GL.DrawArraysInstanced = DrawArraysInstanced;
funcsTable.GL.DrawElementsIndirect = DrawElementsIndirect;
funcsTable.GL.DrawArraysIndirect = DrawArraysIndirect;
funcsTable.GL.Clear = Clear;
funcsTable.GL.ClearBufferfi = ClearBufferfi;
funcsTable.GL.ClearBufferfv = ClearBufferfv;
funcsTable.GL.ClearBufferuiv = ClearBufferuiv;
funcsTable.GL.ClearBufferiv = ClearBufferiv;
funcsTable.GL.ClearNamedFramebufferfv = ClearNamedFramebufferfv;
funcsTable.GL.ClearNamedFramebufferfi = ClearNamedFramebufferfi;
funcsTable.GL.BlitFramebuffer = BlitFramebuffer;
funcsTable.GL.BlitNamedFramebuffer = BlitNamedFramebuffer;
funcsTable.GL.CopyTexImage2D = CopyTexImage2D;
funcsTable.GL.CopyTexSubImage2D = CopyTexSubImage2D;
funcsTable.GL.CopyImageSubData = CopyImageSubData;
funcsTable.GL.GenerateMipmap = GenerateMipmap;
funcsTable.GL.ReadPixels = ReadPixels;
funcsTable.GL.GetTexImage = GetTexImage;
funcsTable.GL.GetTextureImage = GetTextureImage;
funcsTable.GL.DispatchCompute = DispatchCompute;
funcsTable.GL.DispatchComputeIndirect = DispatchComputeIndirect;
funcsTable.GL.MemoryBarrier = MemoryBarrier;
funcsTable.GL.MemoryBarrierByRegion = MemoryBarrierByRegion;
funcsTable.GL.BindImageTexture = BindImageTexture;
funcsTable.GL.GetIntegeri_v = GetIntegeri_v;
funcsTable.GL.GetInteger64i_v = GetInteger64i_v;
funcsTable.GL.GetProgramiv = GetProgramiv;
funcsTable.GL.GetProgramInterfaceiv = GetProgramInterfaceiv;
funcsTable.GL.GetProgramResourceIndex = GetProgramResourceIndex;
funcsTable.GL.GetProgramResourceName = GetProgramResourceName;
funcsTable.GL.GetProgramResourceiv = GetProgramResourceiv;
funcsTable.GL.GetProgramResourceLocation = GetProgramResourceLocation;
funcsTable.GL.GetProgramResourceLocationIndex = GetProgramResourceLocationIndex;
funcsTable.GL.ShaderStorageBlockBinding = ShaderStorageBlockBinding;
funcsTable.GL.FenceSync = FenceSync;
funcsTable.GL.ClientWaitSync = ClientWaitSync;
funcsTable.GL.WaitSync = WaitSync;
funcsTable.GL.DeleteSync = DeleteSync;
funcsTable.GL.GetSyncStatus = GetSyncStatus;
// Optional timer-query group: left null (the frontend then falls
// back) when disabled via MOBILEGL_DISABLE_TIMERQUERY. The hooks
// themselves additionally degrade to null handles when the device
// lacks timestamp support.
if (!MG_Config::Features.DisableTimerQuery) {
funcsTable.GL.IsTimerQuerySupported = IsTimerQuerySupported;
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;
}
funcsTableInitialized = true;
}
return funcsTable;
}
const DynamicBackendParameters& BackendObject_DirectVulkan::GetDynamicParameters() const {
return m_dynamicParameters;
}
void BackendObject_DirectVulkan::ApplyVulkanCapabilitiesForTesting(
const MG_External::VulkanCapabilities& capabilities) {
m_vulkanCaps = capabilities;
UpdateDynamicBackendParameters();
UpdateAdvertisedExtensions();
MutableFormatCapabilities().Clear();
}
void BackendObject_DirectVulkan::UpdateAdvertisedExtensions() {
// GL_ARB_timer_query gates MC's F3 GPU% (LWJGL checks the extension
// string). InitCapabilities runs after InitWindowSurface has created
// and initialized the renderer, so the advertisement can be gated on
// real device timestamp support. ApplyVulkanCapabilitiesForTesting may
// run without a renderer; no timer query is advertised then. Rebuilding
// the whole list keeps re-runs idempotent.
m_rendererInfo.RendererGLInfo.Extensions = BuildAdvertisedExtensions(
m_vulkanCaps.SupportsShaderSubgroup, pVulkanRenderer && pVulkanRenderer->IsTimerQuerySupported(),
pVulkanRenderer && pVulkanRenderer->IsSamplerAnisotropySupported());
}
void BackendObject_DirectVulkan::UpdateDynamicBackendParameters() {
const auto mapShaderStages = [](Uint32 vkStages) {
Uint32 glStages = 0;
if ((vkStages & VK_SHADER_STAGE_VERTEX_BIT) != 0) glStages |= GL_VERTEX_SHADER_BIT;
if ((vkStages & VK_SHADER_STAGE_TESSELLATION_CONTROL_BIT) != 0) glStages |= GL_TESS_CONTROL_SHADER_BIT;
if ((vkStages & VK_SHADER_STAGE_TESSELLATION_EVALUATION_BIT) != 0) {
glStages |= GL_TESS_EVALUATION_SHADER_BIT;
}
if ((vkStages & VK_SHADER_STAGE_GEOMETRY_BIT) != 0) glStages |= GL_GEOMETRY_SHADER_BIT;
if ((vkStages & VK_SHADER_STAGE_FRAGMENT_BIT) != 0) glStages |= GL_FRAGMENT_SHADER_BIT;
if ((vkStages & VK_SHADER_STAGE_COMPUTE_BIT) != 0) glStages |= GL_COMPUTE_SHADER_BIT;
return glStages;
};
const auto mapSubgroupFeatures = [](Uint32 vkFeatures) {
Uint32 glFeatures = 0;
if ((vkFeatures & VK_SUBGROUP_FEATURE_BASIC_BIT) != 0) {
glFeatures |= GL_SUBGROUP_FEATURE_BASIC_BIT_KHR;
}
if ((vkFeatures & VK_SUBGROUP_FEATURE_VOTE_BIT) != 0) {
glFeatures |= GL_SUBGROUP_FEATURE_VOTE_BIT_KHR;
}
if ((vkFeatures & VK_SUBGROUP_FEATURE_ARITHMETIC_BIT) != 0) {
glFeatures |= GL_SUBGROUP_FEATURE_ARITHMETIC_BIT_KHR;
}
if ((vkFeatures & VK_SUBGROUP_FEATURE_BALLOT_BIT) != 0) {
glFeatures |= GL_SUBGROUP_FEATURE_BALLOT_BIT_KHR;
}
if ((vkFeatures & VK_SUBGROUP_FEATURE_SHUFFLE_BIT) != 0) {
glFeatures |= GL_SUBGROUP_FEATURE_SHUFFLE_BIT_KHR;
}
if ((vkFeatures & VK_SUBGROUP_FEATURE_SHUFFLE_RELATIVE_BIT) != 0) {
glFeatures |= GL_SUBGROUP_FEATURE_SHUFFLE_RELATIVE_BIT_KHR;
}
if ((vkFeatures & VK_SUBGROUP_FEATURE_CLUSTERED_BIT) != 0) {
glFeatures |= GL_SUBGROUP_FEATURE_CLUSTERED_BIT_KHR;
}
if ((vkFeatures & VK_SUBGROUP_FEATURE_QUAD_BIT) != 0) {
glFeatures |= GL_SUBGROUP_FEATURE_QUAD_BIT_KHR;
}
return glFeatures;
};
static constexpr SizeT kMaxAdvertisedShaderStorageBlockSize = 512ull * 1024ull * 1024ull;
m_dynamicParameters.UniformBufferOffsetAlignment = m_vulkanCaps.UniformBufferOffsetAlignment;
m_dynamicParameters.AliasedLineWidthRangeMin = m_vulkanCaps.AliasedLineWidthRangeMin;
m_dynamicParameters.AliasedLineWidthRangeMax = m_vulkanCaps.AliasedLineWidthRangeMax;
// Without the samplerAnisotropy feature the limit is unusable, so report 1.0 (no anisotropy)
// rather than a maximum the sampler manager will never apply.
m_dynamicParameters.MaxTextureMaxAnisotropy =
(pVulkanRenderer && pVulkanRenderer->IsSamplerAnisotropySupported()) ? m_vulkanCaps.MaxSamplerAnisotropy
: 1.0f;
m_dynamicParameters.SmoothLineWidthRangeMin = m_vulkanCaps.SmoothLineWidthRangeMin;
m_dynamicParameters.SmoothLineWidthRangeMax = m_vulkanCaps.SmoothLineWidthRangeMax;
m_dynamicParameters.SmoothLineWidthGranularity = m_vulkanCaps.SmoothLineWidthGranularity;
m_dynamicParameters.PointSizeRangeMin = m_vulkanCaps.PointSizeRangeMin;
m_dynamicParameters.PointSizeRangeMax = m_vulkanCaps.PointSizeRangeMax;
m_dynamicParameters.PointSizeGranularity = m_vulkanCaps.PointSizeGranularity;
m_dynamicParameters.Max3DTextureSize = m_vulkanCaps.Max3DTextureSize;
m_dynamicParameters.MaxArrayTextureLayers = m_vulkanCaps.MaxArrayTextureLayers;
m_dynamicParameters.MaxCubeMapTextureSize = m_vulkanCaps.MaxCubeMapTextureSize;
m_dynamicParameters.MaxFramebufferWidth = m_vulkanCaps.MaxFramebufferWidth;
m_dynamicParameters.MaxFramebufferHeight = m_vulkanCaps.MaxFramebufferHeight;
m_dynamicParameters.MaxFramebufferLayers = m_vulkanCaps.MaxFramebufferLayers;
m_dynamicParameters.MaxRenderbufferSize = m_vulkanCaps.MaxRenderbufferSize;
m_dynamicParameters.MaxTextureSize = m_vulkanCaps.MaxTextureSize;
m_dynamicParameters.MaxColorTextureSamples = m_vulkanCaps.MaxColorTextureSamples;
m_dynamicParameters.MaxDepthTextureSamples = m_vulkanCaps.MaxDepthTextureSamples;
m_dynamicParameters.MaxFramebufferSamples = m_vulkanCaps.MaxFramebufferSamples;
m_dynamicParameters.MaxIntegerSamples = m_vulkanCaps.MaxIntegerSamples;
m_dynamicParameters.MaxSamples = m_vulkanCaps.MaxSamples;
m_dynamicParameters.MaxSampleMaskWords = m_vulkanCaps.MaxSampleMaskWords;
const Int maxSupportedTextureUnits =
static_cast<Int>(MG_State::GLState::TextureState::MAX_TEXTURE_IMAGE_UNITS);
// GL_MAX_TEXTURE_IMAGE_UNITS is a *per-stage* sampler limit. Adreno/Qualcomm report a huge
// maxPerStageDescriptorSampledImages (descriptor-indexing scale), so clamping it only to our
// combined array capacity (192) still advertises 192 per stage. Host code treats this value as
// an array bound: Minecraft's Blaze3D GlStateManager.TEXTURES[] holds 128 entries and Iris
// iterates [0, GL_MAX_TEXTURE_IMAGE_UNITS) over it (CompositeRenderer.renderAll), so any value
// > 128 throws ArrayIndexOutOfBoundsException. Match desktop drivers (32) for the per-stage
// limits while keeping the combined limit at our texture-unit array capacity.
constexpr Int maxPerStageTextureUnits =
static_cast<Int>(MG_State::GLState::TextureState::MAX_PER_STAGE_TEXTURE_IMAGE_UNITS);
m_dynamicParameters.MaxTextureImageUnits =
std::min(m_vulkanCaps.MaxTextureImageUnits, maxPerStageTextureUnits);
m_dynamicParameters.MaxVertexTextureImageUnits =
std::min(m_vulkanCaps.MaxVertexTextureImageUnits, maxPerStageTextureUnits);
m_dynamicParameters.MaxComputeTextureImageUnits =
std::min(m_vulkanCaps.MaxComputeTextureImageUnits, maxPerStageTextureUnits);
m_dynamicParameters.MaxCombinedTextureImageUnits =
std::min(m_vulkanCaps.MaxCombinedTextureImageUnits, maxSupportedTextureUnits);
// Never advertise more attributes than the state layer can store: the current-value array and
// the Uint32 attribute masks the draw path passes around are both bounded by MAX_VERTEX_ATTRIBS.
m_dynamicParameters.MaxVertexAttribs =
std::min(m_vulkanCaps.MaxVertexAttribs,
static_cast<Int>(MG_State::GLState::VertexArrayObject::MAX_VERTEX_ATTRIBS));
m_dynamicParameters.MaxComputeShaderStorageBlocks = m_vulkanCaps.MaxComputeShaderStorageBlocks;
m_dynamicParameters.MaxCombinedShaderStorageBlocks = m_vulkanCaps.MaxCombinedShaderStorageBlocks;
m_dynamicParameters.MaxComputeUniformBlocks = m_vulkanCaps.MaxComputeUniformBlocks;
m_dynamicParameters.MaxComputeWorkGroupInvocations = m_vulkanCaps.MaxComputeWorkGroupInvocations;
m_dynamicParameters.MaxShaderStorageBufferBindings = m_vulkanCaps.MaxShaderStorageBufferBindings;
m_dynamicParameters.MaxTextureBufferSize = m_vulkanCaps.MaxTextureBufferSize;
m_dynamicParameters.MaxUniformBufferBindings = m_vulkanCaps.MaxUniformBufferBindings;
m_dynamicParameters.MaxUniformBlockSize = m_vulkanCaps.MaxUniformBlockSize;
m_dynamicParameters.MaxImageUnits =
std::max(std::min(m_vulkanCaps.MaxImageUnits, maxSupportedTextureUnits), 0);
m_dynamicParameters.MaxCombinedImageUniforms = std::max(m_vulkanCaps.MaxCombinedImageUniforms, 0);
const Int maxPerStageImageUniforms =
std::min(m_dynamicParameters.MaxImageUnits, m_dynamicParameters.MaxCombinedImageUniforms);
// Vulkan uses one descriptor limit for every stage, but non-compute stores/atomics are
// optional device features. VulkanRenderer enables each feature whenever the physical
// device reports it, so these are the exact limits the logical device can compile and run.
m_dynamicParameters.MaxVertexImageUniforms =
m_vulkanCaps.SupportsVertexPipelineStoresAndAtomics ? maxPerStageImageUniforms : 0;
m_dynamicParameters.MaxGeometryImageUniforms =
m_vulkanCaps.SupportsVertexPipelineStoresAndAtomics && m_vulkanCaps.SupportsGeometryShader
? maxPerStageImageUniforms
: 0;
m_dynamicParameters.MaxFragmentImageUniforms =
m_vulkanCaps.SupportsFragmentStoresAndAtomics ? maxPerStageImageUniforms : 0;
m_dynamicParameters.MaxComputeImageUniforms =
std::min(std::max(m_vulkanCaps.MaxComputeImageUniforms, 0), maxPerStageImageUniforms);
const Int maxSupportedDrawBuffers =
static_cast<Int>(MG_State::GLState::FramebufferObject::MAX_DRAW_BUFFERS);
m_dynamicParameters.MaxDrawBuffers = std::min(m_vulkanCaps.MaxDrawBuffers, maxSupportedDrawBuffers);
m_dynamicParameters.MaxColorAttachments = std::min(m_vulkanCaps.MaxColorAttachments, maxSupportedDrawBuffers);
m_dynamicParameters.MaxClipDistances = m_vulkanCaps.MaxClipDistances;
m_dynamicParameters.MaxViewports = m_vulkanCaps.MaxViewports;
m_dynamicParameters.MaxViewportWidth = m_vulkanCaps.MaxViewportWidth;
m_dynamicParameters.MaxViewportHeight = m_vulkanCaps.MaxViewportHeight;
m_dynamicParameters.ViewportBoundsRangeMin = m_vulkanCaps.ViewportBoundsRangeMin;
m_dynamicParameters.ViewportBoundsRangeMax = m_vulkanCaps.ViewportBoundsRangeMax;
m_dynamicParameters.ViewportSubpixelBits = m_vulkanCaps.ViewportSubpixelBits;
m_dynamicParameters.SupportsWideLines = m_vulkanCaps.SupportsWideLines;
m_dynamicParameters.MaxShaderStorageBlockSize =
std::min(m_vulkanCaps.MaxShaderStorageBlockSize, kMaxAdvertisedShaderStorageBlockSize);
if (m_vulkanCaps.SupportsShaderSubgroup) {
m_dynamicParameters.SubgroupSize = m_vulkanCaps.SubgroupSize;
m_dynamicParameters.SubgroupSupportedStages = mapShaderStages(m_vulkanCaps.SubgroupSupportedStages);
m_dynamicParameters.SubgroupSupportedFeatures = mapSubgroupFeatures(m_vulkanCaps.SubgroupSupportedOperations);
m_dynamicParameters.SubgroupQuadOperationsInAllStages = m_vulkanCaps.SubgroupQuadOperationsInAllStages;
} else {
m_dynamicParameters.SubgroupSize = 0;
m_dynamicParameters.SubgroupSupportedStages = 0;
m_dynamicParameters.SubgroupSupportedFeatures = 0;
m_dynamicParameters.SubgroupQuadOperationsInAllStages = false;
}
if (m_dynamicParameters.MaxShaderStorageBlockSize != m_vulkanCaps.MaxShaderStorageBlockSize) {
MGLOG_I("DirectVulkan: clamped GL_MAX_SHADER_STORAGE_BLOCK_SIZE from %zu to %zu",
m_vulkanCaps.MaxShaderStorageBlockSize,
m_dynamicParameters.MaxShaderStorageBlockSize);
}
switch (m_vulkanCaps.VendorId) {
case 0x5143u: // VK_VENDOR_ID: Qualcomm
m_dynamicParameters.GpuVendor = GpuVendorKind::Qualcomm;
break;
case 0x13B5u: // ARM
m_dynamicParameters.GpuVendor = GpuVendorKind::Arm;
break;
case 0x10DEu: // NVIDIA
m_dynamicParameters.GpuVendor = GpuVendorKind::Nvidia;
break;
case 0x1002u: // AMD
m_dynamicParameters.GpuVendor = GpuVendorKind::Amd;
break;
case 0x8086u: // Intel
m_dynamicParameters.GpuVendor = GpuVendorKind::Intel;
break;
case 0x1010u: // Imagination
m_dynamicParameters.GpuVendor = GpuVendorKind::ImgTec;
break;
case 0x10005u: // Mesa software (lavapipe)
case 0x1AE0u: // Google (SwiftShader)
m_dynamicParameters.GpuVendor = GpuVendorKind::Software;
break;
default:
m_dynamicParameters.GpuVendor = GpuVendorKind::Unknown;
break;
}
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,84 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/BackendObject_DirectVulkan.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 <Includes.h>
#include "../BackendObject.h"
#include <MG_Util/BackendLoaders/Vulkan/Loader.h>
namespace MobileGL::MG_Backend::DirectVulkan {
// Populates the same format-capability cache used by backend startup. Passing the
// instance-resolved function keeps standalone callers independent of global loader
// initialization; the physical device must remain valid for the duration of the call.
void PopulateFormatCapabilities(VkPhysicalDevice physicalDevice,
PFN_vkGetPhysicalDeviceFormatProperties getFormatProperties,
const MG_External::VulkanCapabilities& capabilities,
FormatCapabilityCache& cache);
class BackendObject_DirectVulkan : public BackendObject {
public:
BackendObject_DirectVulkan();
~BackendObject_DirectVulkan() override;
void Initialize() override;
Bool InitWindowSurface() override;
Bool InitCapabilities() override;
Bool InitializeEGLDisplay(EGLDisplay dpy, EGLint* major, EGLint* minor) override;
Bool CreateEGLWindowSurface(EGLSurface surface, const WindowHandle& handle) override;
Bool ResizeEGLWindowSurface(EGLSurface surface, Uint32 width, Uint32 height) override;
Bool CreateEGLPbufferSurface(EGLSurface surface, EGLint width, EGLint height) override;
Bool MakeEGLCurrent(EGLDisplay dpy, EGLSurface draw, EGLSurface read, EGLContext ctx) override;
Bool SwapEGLBuffers(EGLDisplay dpy, EGLSurface draw) override;
void ReleaseEGLSurface(EGLSurface surface) override;
void ReleaseEGLResources() override;
const RendererInfo& GetRendererInfo() const override;
String GetBackendAPIVersionString() const override;
const GlobalBackendFunctionsTable& GetBackendFunctions() const override;
const DynamicBackendParameters& GetDynamicParameters() const override;
BackendType GetBackendType() const override;
void ApplyVulkanCapabilitiesForTesting(const MG_External::VulkanCapabilities& capabilities);
private:
Bool InitPbufferSurface(EGLint width, EGLint height) override;
void OnEGLSurfaceReleased(EGLSurface surface) override;
void UpdateAdvertisedExtensions();
void UpdateDynamicBackendParameters();
Bool m_initialized = false;
DynamicBackendParameters m_dynamicParameters;
MG_External::VulkanCapabilities m_vulkanCaps;
RendererInfo m_rendererInfo;
};
// Single-source-of-truth helpers shared with the driver POST
// (MG_Util/SelfTest/DriverPost.cpp), so the identity strings and extension list
// MobileGL reports to applications on this backend cannot drift from what the
// POST screen shows.
// Static identity of the Magma renderer (renderer/backend names, target GL/GLSL
// versions, ExtraVendor) with the baseline extension advertisement (no shader
// subgroup, no timer queries). A live backend copies this in its constructor and
// reconciles the Extensions in UpdateAdvertisedExtensions once real capabilities
// exist; callers that need the advertised list for a known capability set must
// use BuildAdvertisedExtensions instead.
const RendererInfo& GetRendererIdentity();
// The full OpenGL extension list Magma advertises (glGetString(GL_EXTENSIONS)) for
// a device with the given raw capabilities. The MOBILEGL_DISABLE_SUBGROUP and
// MOBILEGL_DISABLE_TIMERQUERY escape hatches are applied inside, so callers pass
// the detected device support (passing an already-gated value is harmless).
Vector<GLExtension> BuildAdvertisedExtensions(Bool shaderSubgroupSupported, Bool timerQueriesSupported,
Bool anisotropicFilteringSupported);
// Format: <GPU Name>, Vulkan <Vulkan Version>, Driver <Driver Version> — the exact
// string an initialized backend returns from GetBackendAPIVersionString (and that
// ends up inside the application-visible GL_RENDERER string).
String FormatBackendAPIVersionString(const String& deviceName, const String& vulkanApiVersionString,
const String& driverVersionString);
} // namespace MobileGL::MG_Backend::DirectVulkan
File diff suppressed because it is too large Load Diff
@@ -1,134 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/DirectVulkan.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 <Includes.h>
#include <MG_Backend/BackendObject.h>
#include "Renderer/VulkanRenderer.h"
namespace MobileGL::MG_Backend::DirectVulkan {
extern UniquePtr<VulkanRenderer>& pVulkanRenderer;
// Generation of the live VulkanRenderer instance, mirroring DirectGLES's
// g_syncContextGeneration. BackendObject_DirectVulkan bumps it wherever
// pVulkanRenderer is reset or recreated; fence and timer-query handles
// stamped with an older generation are stale and resolve as signaled /
// available with zero results instead of dereferencing the destroyed
// renderer's frame serials and query-pool slots.
Uint64 GetRendererGeneration();
void BumpRendererGeneration();
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);
void ClearBufferiv(GLenum buffer, GLint drawbuffer, const GLint* value);
void ClearNamedFramebufferfv(const SharedPtr<MG_State::GLState::FramebufferObject>& framebuffer, GLenum buffer,
GLint drawbuffer, const GLfloat* value);
void ClearNamedFramebufferfi(const SharedPtr<MG_State::GLState::FramebufferObject>& framebuffer, GLenum buffer,
GLint drawbuffer, GLfloat depth, GLint stencil);
void Clear(GLbitfield mask);
void DrawElements(GLenum mode, GLsizei count, GLenum type, const void* indices);
void DrawArrays(GLenum mode, GLint first, GLsizei count);
void DrawElementsBaseVertex(GLenum mode, GLsizei count, GLenum type, const GLvoid* indices, GLint basevertex);
void MultiDrawArrays(GLenum mode, const GLint* first, const GLsizei* count, GLsizei drawcount);
void MultiDrawElements(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount);
void MultiDrawElementsBaseVertex(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount, const GLint* basevertex);
void MultiDrawElementsIndirect(GLenum mode, GLenum type, const void* indirect, GLsizei drawcount, GLsizei stride);
void MultiDrawArraysIndirect(GLenum mode, const void* indirect, GLsizei drawcount, GLsizei stride);
void MultiDrawElementsIndirectCount(GLenum mode, GLenum type, const void* indirect, GLintptr drawcount,
GLsizei maxdrawcount, GLsizei stride);
void MultiDrawArraysIndirectCount(GLenum mode, const void* indirect, GLintptr drawcount,
GLsizei maxdrawcount, GLsizei stride);
void DrawRangeElementsBaseVertex(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type,
const void* indices, GLint basevertex);
void DrawRangeElements(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type, const void* indices);
void DrawElementsInstancedBaseVertexBaseInstance(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLint basevertex, GLuint baseinstance);
void DrawElementsInstancedBaseVertex(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLint basevertex);
void DrawElementsInstancedBaseInstance(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLuint baseinstance);
void DrawElementsInstanced(GLenum mode, GLsizei count, GLenum type, const void* indices, GLsizei instancecount);
void DrawElementsIndirect(GLenum mode, GLenum type, const void* indirect);
void DrawArraysInstancedBaseInstance(GLenum mode, GLint first, GLsizei count, GLsizei instancecount,
GLuint baseinstance);
void DrawArraysInstanced(GLenum mode, GLint first, GLsizei count, GLsizei instancecount);
void DrawArraysIndirect(GLenum mode, const void* indirect);
void BlitFramebuffer(GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1, GLint dstX0, GLint dstY0, GLint dstX1,
GLint dstY1, GLbitfield mask, GLenum filter);
void BlitNamedFramebuffer(const SharedPtr<MG_State::GLState::FramebufferObject>& readFramebuffer,
const SharedPtr<MG_State::GLState::FramebufferObject>& drawFramebuffer,
GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1,
GLint dstX0, GLint dstY0, GLint dstX1, GLint dstY1,
GLbitfield mask, GLenum filter);
void CopyTexImage2D(GLenum target, GLint level, GLenum internalformat, GLint x, GLint y, GLsizei width,
GLsizei height, GLint border);
void CopyTexSubImage2D(GLenum target, GLint level, GLint xoffset, GLint yoffset, GLint x, GLint y, GLsizei width,
GLsizei height);
void CopyImageSubData(const SharedPtr<MG_State::GLState::ITextureObject>& srcTexture,
GLenum srcTarget, GLint srcLevel, GLint srcX, GLint srcY, GLint srcZ,
const SharedPtr<MG_State::GLState::ITextureObject>& dstTexture,
GLenum dstTarget, GLint dstLevel, GLint dstX, GLint dstY, GLint dstZ,
GLsizei srcWidth, GLsizei srcHeight, GLsizei srcDepth);
void GenerateMipmap(GLenum target);
void DispatchCompute(GLuint numGroupsX, GLuint numGroupsY, GLuint numGroupsZ);
void DispatchComputeIndirect(GLintptr indirect);
void MemoryBarrier(GLbitfield barriers);
void MemoryBarrierByRegion(GLbitfield barriers);
void BindImageTexture(GLuint unit, GLuint texture, GLint level, GLboolean layered, GLint layer, GLenum access,
GLenum format);
void GetIntegeri_v(GLenum target, GLuint index, GLint* data);
void GetInteger64i_v(GLenum target, GLuint index, GLint64* data);
void GetProgramiv(GLuint program, GLenum pname, GLint* params);
void GetProgramInterfaceiv(GLuint program, GLenum programInterface, GLenum pname, GLint* params);
GLuint GetProgramResourceIndex(GLuint program, GLenum programInterface, const GLchar* name);
void GetProgramResourceName(GLuint program, GLenum programInterface, GLuint index, GLsizei bufSize,
GLsizei* length, GLchar* name);
void GetProgramResourceiv(GLuint program, GLenum programInterface, GLuint index, GLsizei propCount,
const GLenum* props, GLsizei bufSize, GLsizei* length, GLint* params);
GLint GetProgramResourceLocation(GLuint program, GLenum programInterface, const GLchar* name);
GLint GetProgramResourceLocationIndex(GLuint program, GLenum programInterface, const GLchar* name);
void ShaderStorageBlockBinding(GLuint program, GLuint storageBlockIndex, GLuint storageBlockBinding);
void ReadPixels(GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, void* pixels);
void GetTexImage(GLenum target, GLint level, GLenum format, GLenum type, GLvoid* pixels);
void GetTextureImage(const SharedPtr<MG_State::GLState::ITextureObject>& texture, TextureUploadTarget uploadTarget,
GLint level, GLenum format, GLenum type, GLsizei bufSize, GLvoid* pixels);
// GL fence sync objects, mapped onto the renderer's frame-serial busy
// tracking: a fence captures the frame serial current at creation and is
// signaled once every command recorded under that serial has completed on
// the GPU.
BackendSyncHandle FenceSync();
GLenum ClientWaitSync(BackendSyncHandle sync, GLbitfield flags, GLuint64 timeout);
void WaitSync(BackendSyncHandle sync, GLbitfield flags, GLuint64 timeout);
void DeleteSync(BackendSyncHandle sync);
Bool GetSyncStatus(BackendSyncHandle sync);
// GPU timer queries (GL_TIME_ELAPSED spans and GL_TIMESTAMP one-shots),
// backed by per-frame VkQueryPool timestamp slots. All hooks degrade
// gracefully: null handles when the renderer is absent, the device lacks
// timestamp support, or the frame's pool is exhausted.
// Dynamic support check (GLFunctionsTable::IsTimerQuerySupported): true
// only while a live renderer exists whose device can actually time.
Bool IsTimerQuerySupported();
BackendQueryHandle BeginTimeElapsedQuery();
void EndTimeElapsedQuery(BackendQueryHandle query);
BackendQueryHandle QueryCounterTimestamp();
Bool IsQueryResultAvailable(BackendQueryHandle query);
// Returns true when a final value was produced (outNanoseconds set; the
// frontend may cache it and release the handle), false when the result
// cannot be obtained yet (e.g. a wait refused because the records' frame
// serial is the current unsubmitted frame) - the handle then stays
// readable later.
Bool GetQueryResult64(BackendQueryHandle query, Bool wait, Uint64* outNanoseconds);
void DeleteBackendQuery(BackendQueryHandle query);
// Always 0: Vulkan cannot synchronously sample the GPU clock (timestamps
// only exist as vkCmdWriteTimestamp results); the frontend falls back.
Int64 GetGpuTimestampNs();
void Present();
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,20 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/DirectVulkanResourceState.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 <Includes.h>
namespace MobileGL::MG_State::GLState {
class ProgramObject;
}
namespace MobileGL::MG_Backend::DirectVulkan {
GLuint GetShaderStorageBlockIndex(const MG_State::GLState::ProgramObject& program, const String& name);
GLuint GetShaderStorageBlockBinding(const MG_State::GLState::ProgramObject& program, GLuint blockIndex);
}
@@ -1,138 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/BufferArena.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 "BufferArena.h"
namespace MobileGL::MG_Backend::DirectVulkan {
Bool BufferArena::Initialize(const BufferArenaDesc& desc) {
Shutdown();
MOBILEGL_ASSERT(desc.allocator != nullptr, "BufferArena::Initialize requires valid allocator");
MOBILEGL_ASSERT(desc.frameCount > 0, "BufferArena::Initialize requires non-zero frame count");
MOBILEGL_ASSERT(desc.usage != 0, "BufferArena::Initialize requires non-zero buffer usage");
m_desc = desc;
m_frames.clear();
m_frames.resize(desc.frameCount);
m_deferredReleases.resize(desc.frameCount);
return true;
}
void BufferArena::Shutdown() {
for (auto& frame : m_frames) {
frame.buffer.Destroy();
frame.writeCursor = 0;
}
m_frames.clear();
m_deferredReleases.clear();
m_desc = {};
}
void BufferArena::BeginFrame(Uint32 frameIndex) {
CollectDeferredReleases(frameIndex);
ResetFrame(frameIndex);
}
void BufferArena::ResetFrame(Uint32 frameIndex) {
AssertValidFrameIndex(frameIndex);
m_frames[frameIndex].writeCursor = 0;
}
void BufferArena::CollectDeferredReleases(Uint32 frameIndex) {
AssertValidFrameIndex(frameIndex);
m_deferredReleases[frameIndex].clear();
}
Bool BufferArena::Allocate(Uint32 frameIndex, VkDeviceSize size, VkDeviceSize alignment, BufferSlice& outSlice) {
AssertValidFrameIndex(frameIndex);
MOBILEGL_ASSERT(size > 0, "BufferArena::Allocate requires non-zero size");
auto& frame = m_frames[frameIndex];
const VkDeviceSize resolvedAlignment = alignment > 0 ? alignment : 1;
const VkDeviceSize offset = (frame.writeCursor + resolvedAlignment - 1) & ~(resolvedAlignment - 1);
const VkDeviceSize endOffset = offset + size;
if (!EnsureCapacity(frameIndex, endOffset)) {
return false;
}
frame.writeCursor = endOffset;
outSlice = frame.buffer.GetSlice(offset, size);
return outSlice.IsValid();
}
Bool BufferArena::Upload(Uint32 frameIndex, const void* data, VkDeviceSize size, VkDeviceSize alignment,
BufferSlice& outSlice) {
MOBILEGL_ASSERT(data != nullptr || size == 0, "BufferArena::Upload data pointer is null");
if (!Allocate(frameIndex, size, alignment, outSlice)) {
return false;
}
if (outSlice.mapped != nullptr) {
Memcpy(outSlice.mapped, data, static_cast<SizeT>(size));
return true;
}
return m_frames[frameIndex].buffer.Upload(data, size, outSlice.offset);
}
VkDeviceSize BufferArena::GetWriteCursor(Uint32 frameIndex) const {
AssertValidFrameIndex(frameIndex);
return m_frames[frameIndex].writeCursor;
}
Uint32 BufferArena::GetFrameCount() const {
return static_cast<Uint32>(m_frames.size());
}
Bool BufferArena::EnsureCapacity(Uint32 frameIndex, VkDeviceSize requiredEndOffset) {
AssertValidFrameIndex(frameIndex);
auto& frame = m_frames[frameIndex];
auto& buffer = frame.buffer;
if (buffer.IsValid() && buffer.GetSize() >= requiredEndOffset) {
return true;
}
VkDeviceSize newCapacity = buffer.IsValid() ? buffer.GetSize() : 0;
if (newCapacity < m_desc.minBufferSize) {
newCapacity = m_desc.minBufferSize;
}
if (newCapacity == 0) {
newCapacity = requiredEndOffset;
}
while (newCapacity < requiredEndOffset) {
newCapacity *= 2;
}
if (buffer.IsValid()) {
m_deferredReleases[frameIndex].push_back(std::move(buffer));
}
VkBufferObjectDesc bufferDesc{};
bufferDesc.allocator = m_desc.allocator;
bufferDesc.size = newCapacity;
bufferDesc.usage = m_desc.usage;
bufferDesc.memoryUsage = m_desc.memoryUsage;
bufferDesc.allocationFlags = m_desc.allocationFlags;
if (!buffer.Create(bufferDesc)) {
return false;
}
if (m_desc.persistentlyMapped && buffer.Map() == nullptr) {
buffer.Destroy();
return false;
}
frame.writeCursor = 0;
return true;
}
void BufferArena::AssertValidFrameIndex(Uint32 frameIndex) const {
MOBILEGL_ASSERT(frameIndex < m_frames.size(), "BufferArena frame index out of range");
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,56 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/BufferArena.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 "BufferSlice.h"
#include "VkBufferObject.h"
#include "../VkIncludes.h"
#include <Includes.h>
#include <vk_mem_alloc.h>
namespace MobileGL::MG_Backend::DirectVulkan {
struct BufferArenaDesc {
VmaAllocator allocator = nullptr;
Uint32 frameCount = 0;
VkBufferUsageFlags usage = 0;
VmaMemoryUsage memoryUsage = VMA_MEMORY_USAGE_AUTO;
VmaAllocationCreateFlags allocationFlags = 0;
VkDeviceSize minBufferSize = 0;
Bool persistentlyMapped = false;
};
class BufferArena {
public:
Bool Initialize(const BufferArenaDesc& desc);
void Shutdown();
void BeginFrame(Uint32 frameIndex);
void ResetFrame(Uint32 frameIndex);
void CollectDeferredReleases(Uint32 frameIndex);
Bool Allocate(Uint32 frameIndex, VkDeviceSize size, VkDeviceSize alignment, BufferSlice& outSlice);
Bool Upload(Uint32 frameIndex, const void* data, VkDeviceSize size, VkDeviceSize alignment, BufferSlice& outSlice);
VkDeviceSize GetWriteCursor(Uint32 frameIndex) const;
Uint32 GetFrameCount() const;
private:
struct FrameResources {
VkBufferObject buffer;
VkDeviceSize writeCursor = 0;
};
Bool EnsureCapacity(Uint32 frameIndex, VkDeviceSize requiredEndOffset);
void AssertValidFrameIndex(Uint32 frameIndex) const;
BufferArenaDesc m_desc{};
Vector<FrameResources> m_frames;
Vector<Vector<VkBufferObject>> m_deferredReleases;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,23 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/BufferSlice.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 "../VkIncludes.h"
#include <Includes.h>
namespace MobileGL::MG_Backend::DirectVulkan {
struct BufferSlice {
VkBuffer buffer = VK_NULL_HANDLE;
VkDeviceSize offset = 0;
VkDeviceSize size = 0;
void* mapped = nullptr;
Bool IsValid() const { return buffer != VK_NULL_HANDLE; }
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,334 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/FrameContext.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 "FrameContext.h"
namespace MobileGL::MG_Backend::DirectVulkan {
VkResult FrameContext::Initialize(VkDevice device, VkCommandPool commandPool, Uint32 frameCount) {
Destroy(device, commandPool);
m_frames.assign(frameCount, {});
currentFrameIndex = 0;
m_device = device;
m_commandPool = commandPool;
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;
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];
}
VkSemaphoreCreateInfo semaphoreInfo{VK_STRUCTURE_TYPE_SEMAPHORE_CREATE_INFO};
VkFenceCreateInfo fenceInfo{VK_STRUCTURE_TYPE_FENCE_CREATE_INFO};
fenceInfo.flags = VK_FENCE_CREATE_SIGNALED_BIT;
for (Uint32 i = 0; i < frameCount; ++i) {
result = CreateSyncObjectsForFrame(device, i, semaphoreInfo, fenceInfo);
if (result != VK_SUCCESS) {
Destroy(device, commandPool);
return result;
}
}
return VK_SUCCESS;
}
void FrameContext::Destroy(VkDevice device, VkCommandPool commandPool) {
const Uint32 frameCount = static_cast<Uint32>(m_frames.size());
Vector<VkCommandBuffer> commandBuffers(frameCount, VK_NULL_HANDLE);
for (Uint32 i = 0; i < frameCount; ++i) {
commandBuffers[i] = m_frames[i].commandBuffer;
}
for (Uint32 i = 0; i < frameCount; ++i) {
DestroySyncObjectsForFrame(device, i);
}
DestroySwapchainSemaphores(device);
if (device != VK_NULL_HANDLE && commandPool != VK_NULL_HANDLE && !m_frames.empty()) {
for (auto& frame : m_frames) {
FreeRetiredCommandBuffers(frame);
}
vkFreeCommandBuffers(device, commandPool, frameCount, commandBuffers.data());
}
m_frames.clear();
currentFrameIndex = 0;
m_device = VK_NULL_HANDLE;
m_commandPool = VK_NULL_HANDLE;
}
FrameContext::FrameData& FrameContext::GetCurrent() {
MOBILEGL_ASSERT(!m_frames.empty(), "FrameContext is not initialized");
return m_frames[currentFrameIndex];
}
const FrameContext::FrameData& FrameContext::GetCurrent() const {
MOBILEGL_ASSERT(!m_frames.empty(), "FrameContext is not initialized");
return m_frames[currentFrameIndex];
}
Bool FrameContext::IsCommandRecording() const {
return GetCurrent().isCommandRecording;
}
void FrameContext::AdvanceToNext() {
MOBILEGL_ASSERT(!m_frames.empty(), "FrameContext is not initialized");
currentFrameIndex = (currentFrameIndex + 1) % static_cast<Uint32>(m_frames.size());
GetCurrent().isCommandRecording = false;
GetCurrent().hasCommandBufferRecorded = false;
}
VkCommandBuffer& FrameContext::BeginCommandRecording(VkCommandBufferUsageFlags flags,
const VkCommandBufferInheritanceInfo* pInheritanceInfo) {
auto& frame = GetCurrent();
MOBILEGL_ASSERT(!frame.isCommandRecording, "BeginCommandRecording called while command buffer is already recording");
frame.hasCommandBufferRecorded = false;
VK_VERIFY(vkResetCommandBuffer(frame.commandBuffer, 0), "BeginCommandRecording, vkResetCommandBuffer");
VkCommandBufferBeginInfo beginInfo{};
beginInfo.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_BEGIN_INFO;
beginInfo.flags = flags;
beginInfo.pInheritanceInfo = pInheritanceInfo;
VK_VERIFY(vkBeginCommandBuffer(frame.commandBuffer, &beginInfo), "BeginCommandRecording, vkBeginCommandBuffer");
frame.isCommandRecording = true;
if (m_recordingObserver != nullptr) {
m_recordingObserver->OnFrameCommandRecordingBegan(frame.commandBuffer);
}
return frame.commandBuffer;
}
void FrameContext::EndCommandRecording() {
auto& frame = GetCurrent();
MOBILEGL_ASSERT(frame.isCommandRecording, "EndCommandRecording called without active command buffer recording");
VK_VERIFY(vkEndCommandBuffer(frame.commandBuffer), "EndCommandRecording, vkEndCommandBuffer");
frame.isCommandRecording = false;
frame.hasCommandBufferRecorded = true;
}
VkResult FrameContext::InitializeSwapchainSemaphores(VkDevice device, Uint32 swapchainImageCount) {
DestroySwapchainSemaphores(device);
if (swapchainImageCount == 0) {
return VK_SUCCESS;
}
m_swapchainImageRenderFinishedSemaphores.assign(swapchainImageCount, VK_NULL_HANDLE);
VkSemaphoreCreateInfo semaphoreInfo{VK_STRUCTURE_TYPE_SEMAPHORE_CREATE_INFO};
for (Uint32 imageIndex = 0; imageIndex < swapchainImageCount; ++imageIndex) {
VkResult result =
vkCreateSemaphore(device, &semaphoreInfo, nullptr, &m_swapchainImageRenderFinishedSemaphores[imageIndex]);
if (result != VK_SUCCESS) {
DestroySwapchainSemaphores(device);
return result;
}
}
return VK_SUCCESS;
}
void FrameContext::DestroySwapchainSemaphores(VkDevice device) {
if (device != VK_NULL_HANDLE) {
for (auto semaphore : m_swapchainImageRenderFinishedSemaphores) {
if (semaphore != VK_NULL_HANDLE) {
vkDestroySemaphore(device, semaphore, nullptr);
}
}
}
m_swapchainImageRenderFinishedSemaphores.clear();
}
Bool FrameContext::TransitionToPresent(VkImage image, VkImageLayout oldLayout, VkImageLayout presentLayout) {
auto& frame = GetCurrent();
if (frame.hasCommandBufferRecorded || frame.isCommandRecording || oldLayout == presentLayout ||
oldLayout == VK_IMAGE_LAYOUT_SHARED_PRESENT_KHR) {
return false;
}
auto& commandBuffer = BeginCommandRecording();
VkImageMemoryBarrier presentBarrier{};
presentBarrier.sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER;
presentBarrier.srcAccessMask = 0;
presentBarrier.dstAccessMask = 0;
presentBarrier.oldLayout = oldLayout;
presentBarrier.newLayout = presentLayout;
presentBarrier.srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED;
presentBarrier.dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED;
presentBarrier.image = image;
presentBarrier.subresourceRange.aspectMask = VK_IMAGE_ASPECT_COLOR_BIT;
presentBarrier.subresourceRange.baseMipLevel = 0;
presentBarrier.subresourceRange.levelCount = 1;
presentBarrier.subresourceRange.baseArrayLayer = 0;
presentBarrier.subresourceRange.layerCount = 1;
vkCmdPipelineBarrier(commandBuffer, VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT, VK_PIPELINE_STAGE_BOTTOM_OF_PIPE_BIT, 0, 0,
nullptr, 0, nullptr, 1, &presentBarrier);
EndCommandRecording();
return true;
}
FrameContext::SubmitInfoPacket FrameContext::GetSubmitInfo(Bool shouldSubmitCommandBuffer,
Uint32 swapchainImageIndex) const {
const auto& frame = GetCurrent();
MOBILEGL_ASSERT(!frame.isCommandRecording, "GetSubmitInfo called while command buffer recording is still active");
AssertValidSwapchainImageIndex(swapchainImageIndex);
SubmitInfoPacket packet{};
packet.waitSemaphore = frame.imageAvailableSemaphore;
packet.signalSemaphore = m_swapchainImageRenderFinishedSemaphores[swapchainImageIndex];
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 = shouldSubmitCommandBuffer ? 1U : 0U;
packet.submitInfo.pCommandBuffers = shouldSubmitCommandBuffer ? &packet.commandBuffer : nullptr;
packet.submitInfo.signalSemaphoreCount = 1;
packet.submitInfo.pSignalSemaphores = &packet.signalSemaphore;
return packet;
}
FrameContext::PresentInfoPacket FrameContext::GetPresentInfo(VkSwapchainKHR swapchain, Uint32 imageIndex) const {
AssertValidSwapchainImageIndex(imageIndex);
PresentInfoPacket packet{};
packet.waitSemaphore = m_swapchainImageRenderFinishedSemaphores[imageIndex];
packet.swapchain = swapchain;
packet.imageIndex = imageIndex;
packet.presentInfo.waitSemaphoreCount = 1;
packet.presentInfo.pWaitSemaphores = &packet.waitSemaphore;
packet.presentInfo.swapchainCount = 1;
packet.presentInfo.pSwapchains = &packet.swapchain;
packet.presentInfo.pImageIndices = &packet.imageIndex;
packet.presentInfo.pResults = nullptr;
return packet;
}
VkResult FrameContext::WaitAndAcquireNextImage(VkDevice device, VkSwapchainKHR swapchain, Uint32& outImageIndex,
Uint64 timeout, VkFence acquireFence) {
auto& frame = GetCurrent();
VkResult result = vkWaitForFences(device, 1, &frame.imageInFlightFence, VK_TRUE, timeout);
if (result != VK_SUCCESS) {
return result;
}
// The slot's fence has been waited: every command buffer this slot
// submitted (including mid-frame flushes) has finished executing.
FreeRetiredCommandBuffers(frame);
result = vkAcquireNextImageKHR(device, swapchain, timeout, frame.imageAvailableSemaphore, acquireFence,
&outImageIndex);
if (result != VK_SUCCESS) {
return result;
}
frame.imageAvailableSemaphoreConsumed = false;
return vkResetFences(device, 1, &frame.imageInFlightFence);
}
Uint32 FrameContext::GetCurrentFrameIndex() const {
return currentFrameIndex;
}
Uint32 FrameContext::GetFrameCount() const {
return static_cast<Uint32>(m_frames.size());
}
void FrameContext::SetRecordingObserver(IRecordingObserver* observer) {
m_recordingObserver = observer;
}
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");
VkCommandBufferAllocateInfo allocInfo{};
allocInfo.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_ALLOCATE_INFO;
allocInfo.commandPool = m_commandPool;
allocInfo.level = VK_COMMAND_BUFFER_LEVEL_PRIMARY;
allocInfo.commandBufferCount = 1;
VkCommandBuffer replacement = VK_NULL_HANDLE;
const VkResult result = vkAllocateCommandBuffers(m_device, &allocInfo, &replacement);
if (result != VK_SUCCESS) {
return result;
}
frame.retiredCommandBuffers.push_back(frame.commandBuffer);
frame.commandBuffer = replacement;
return VK_SUCCESS;
}
void FrameContext::FreeRetiredCommandBuffers(FrameData& frame) {
if (frame.retiredCommandBuffers.empty()) {
return;
}
if (m_device != VK_NULL_HANDLE && m_commandPool != VK_NULL_HANDLE) {
vkFreeCommandBuffers(m_device, m_commandPool, static_cast<Uint32>(frame.retiredCommandBuffers.size()),
frame.retiredCommandBuffers.data());
}
frame.retiredCommandBuffers.clear();
}
void FrameContext::AssertValidFrameIndex(Uint32 frameIndex) const {
MOBILEGL_ASSERT(frameIndex < m_frames.size(), "FrameContext index out of range");
}
void FrameContext::AssertValidSwapchainImageIndex(Uint32 imageIndex) const {
MOBILEGL_ASSERT(imageIndex < m_swapchainImageRenderFinishedSemaphores.size(),
"FrameContext swapchain image index out of range");
}
VkResult FrameContext::CreateSyncObjectsForFrame(VkDevice device, Uint32 frameIndex,
const VkSemaphoreCreateInfo& semaphoreInfo,
const VkFenceCreateInfo& fenceInfo) {
AssertValidFrameIndex(frameIndex);
DestroySyncObjectsForFrame(device, frameIndex);
auto& frame = m_frames[frameIndex];
VkResult result =
vkCreateSemaphore(device, &semaphoreInfo, nullptr, &frame.imageAvailableSemaphore);
if (result != VK_SUCCESS) {
return result;
}
result = vkCreateFence(device, &fenceInfo, nullptr, &frame.imageInFlightFence);
if (result != VK_SUCCESS) {
vkDestroySemaphore(device, frame.imageAvailableSemaphore, nullptr);
frame.imageAvailableSemaphore = VK_NULL_HANDLE;
return result;
}
frame.hasCommandBufferRecorded = false;
frame.isCommandRecording = false;
frame.imageAvailableSemaphoreConsumed = false;
return VK_SUCCESS;
}
void FrameContext::DestroySyncObjectsForFrame(VkDevice device, Uint32 frameIndex) {
AssertValidFrameIndex(frameIndex);
auto& frame = m_frames[frameIndex];
if (device != VK_NULL_HANDLE && frame.imageInFlightFence != VK_NULL_HANDLE) {
vkDestroyFence(device, frame.imageInFlightFence, nullptr);
}
frame.imageInFlightFence = VK_NULL_HANDLE;
if (device != VK_NULL_HANDLE && frame.imageAvailableSemaphore != VK_NULL_HANDLE) {
vkDestroySemaphore(device, frame.imageAvailableSemaphore, nullptr);
}
frame.imageAvailableSemaphore = VK_NULL_HANDLE;
frame.isCommandRecording = false;
frame.hasCommandBufferRecorded = false;
frame.imageAvailableSemaphoreConsumed = false;
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,109 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/FrameContext.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 "../VkIncludes.h"
#include <Includes.h>
namespace MobileGL::MG_Backend::DirectVulkan {
class FrameContext {
public:
// Notified immediately after a frame command buffer begins recording
// (before any render pass has been begun); every BeginCommandRecording
// caller funnels through this single seam. Implemented by the renderer
// to prepare per-frame timer-query pools (vkCmdResetQueryPool must be
// recorded outside a render pass).
class IRecordingObserver {
public:
virtual ~IRecordingObserver() = default;
virtual void OnFrameCommandRecordingBegan(VkCommandBuffer commandBuffer) = 0;
};
struct SubmitInfoPacket {
VkPipelineStageFlags waitDstStageMask = VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT;
VkSemaphore waitSemaphore = VK_NULL_HANDLE;
VkSemaphore signalSemaphore = VK_NULL_HANDLE;
VkCommandBuffer commandBuffer = VK_NULL_HANDLE;
VkSubmitInfo submitInfo{VK_STRUCTURE_TYPE_SUBMIT_INFO};
};
struct PresentInfoPacket {
VkSemaphore waitSemaphore = VK_NULL_HANDLE;
VkSwapchainKHR swapchain = VK_NULL_HANDLE;
Uint32 imageIndex = 0;
VkPresentInfoKHR presentInfo{VK_STRUCTURE_TYPE_PRESENT_INFO_KHR};
};
struct FrameData {
VkCommandBuffer commandBuffer = VK_NULL_HANDLE;
VkSemaphore imageAvailableSemaphore = VK_NULL_HANDLE;
VkFence imageInFlightFence = VK_NULL_HANDLE;
Bool isCommandRecording = false;
Bool hasCommandBufferRecorded = false;
Bool imageAvailableSemaphoreConsumed = false;
// 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;
};
VkResult Initialize(VkDevice device, VkCommandPool commandPool, Uint32 frameCount);
void Destroy(VkDevice device, VkCommandPool commandPool);
// Lifecycle functions
FrameData& GetCurrent();
const FrameData& GetCurrent() const;
Bool IsCommandRecording() const;
void AdvanceToNext();
VkCommandBuffer& BeginCommandRecording(VkCommandBufferUsageFlags flags = 0,
const VkCommandBufferInheritanceInfo* pInheritanceInfo = nullptr);
void EndCommandRecording();
VkResult InitializeSwapchainSemaphores(VkDevice device, Uint32 swapchainImageCount);
void DestroySwapchainSemaphores(VkDevice device);
Bool TransitionToPresent(VkImage image, VkImageLayout oldLayout,
VkImageLayout presentLayout = VK_IMAGE_LAYOUT_PRESENT_SRC_KHR);
SubmitInfoPacket GetSubmitInfo(Bool shouldSubmitCommandBuffer, Uint32 swapchainImageIndex) const;
PresentInfoPacket GetPresentInfo(VkSwapchainKHR swapchain, Uint32 imageIndex) const;
VkResult WaitAndAcquireNextImage(VkDevice device, VkSwapchainKHR swapchain, Uint32& outImageIndex,
Uint64 timeout = UINT64_MAX, VkFence acquireFence = VK_NULL_HANDLE);
// 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.
VkResult RetireCurrentCommandBuffer();
Uint32 GetCurrentFrameIndex() const;
Uint32 GetFrameCount() const;
// Observer may be null (no notifications). Not owned.
void SetRecordingObserver(IRecordingObserver* observer);
private:
void AssertValidFrameIndex(Uint32 frameIndex) const;
void AssertValidSwapchainImageIndex(Uint32 imageIndex) const;
VkResult CreateSyncObjectsForFrame(VkDevice device, Uint32 frameIndex,
const VkSemaphoreCreateInfo& semaphoreInfo,
const VkFenceCreateInfo& fenceInfo);
void DestroySyncObjectsForFrame(VkDevice device, Uint32 frameIndex);
void FreeRetiredCommandBuffers(FrameData& frame);
Vector<FrameData> m_frames;
Vector<VkSemaphore> m_swapchainImageRenderFinishedSemaphores;
Uint32 currentFrameIndex = 0;
IRecordingObserver* m_recordingObserver = nullptr;
// Stored at Initialize for retired-command-buffer management.
VkDevice m_device = VK_NULL_HANDLE;
VkCommandPool m_commandPool = VK_NULL_HANDLE;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,420 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/PipelineFactory.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 "PipelineFactory.h"
namespace MobileGL::MG_Backend::DirectVulkan {
static const char* PrimitiveTopologyToString(VkPrimitiveTopology topology) {
switch (topology) {
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_POINT_LIST)
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_LINE_LIST)
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_LINE_STRIP)
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_TRIANGLE_LIST)
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_TRIANGLE_STRIP)
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_TRIANGLE_FAN)
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_LINE_LIST_WITH_ADJACENCY)
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_LINE_STRIP_WITH_ADJACENCY)
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_TRIANGLE_LIST_WITH_ADJACENCY)
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_TRIANGLE_STRIP_WITH_ADJACENCY)
ENUM_STR_CASE(VK_PRIMITIVE_TOPOLOGY_PATCH_LIST)
default:
return "VK_PRIMITIVE_TOPOLOGY_UNKNOWN";
}
}
static const char* SampleCountToString(VkSampleCountFlagBits sampleCount) {
switch (sampleCount) {
ENUM_STR_CASE(VK_SAMPLE_COUNT_1_BIT)
ENUM_STR_CASE(VK_SAMPLE_COUNT_2_BIT)
ENUM_STR_CASE(VK_SAMPLE_COUNT_4_BIT)
ENUM_STR_CASE(VK_SAMPLE_COUNT_8_BIT)
ENUM_STR_CASE(VK_SAMPLE_COUNT_16_BIT)
ENUM_STR_CASE(VK_SAMPLE_COUNT_32_BIT)
ENUM_STR_CASE(VK_SAMPLE_COUNT_64_BIT)
default:
return "VK_SAMPLE_COUNT_UNKNOWN";
}
}
static const char* CullModeToString(VkCullModeFlags cullMode) {
switch (cullMode) {
case VK_CULL_MODE_NONE:
return "VK_CULL_MODE_NONE";
case VK_CULL_MODE_FRONT_BIT:
return "VK_CULL_MODE_FRONT_BIT";
case VK_CULL_MODE_BACK_BIT:
return "VK_CULL_MODE_BACK_BIT";
case VK_CULL_MODE_FRONT_AND_BACK:
return "VK_CULL_MODE_FRONT_AND_BACK";
default:
return "VK_CULL_MODE_UNKNOWN";
}
}
static const char* CompareOpToString(VkCompareOp compareOp) {
switch (compareOp) {
ENUM_STR_CASE(VK_COMPARE_OP_NEVER)
ENUM_STR_CASE(VK_COMPARE_OP_LESS)
ENUM_STR_CASE(VK_COMPARE_OP_EQUAL)
ENUM_STR_CASE(VK_COMPARE_OP_LESS_OR_EQUAL)
ENUM_STR_CASE(VK_COMPARE_OP_GREATER)
ENUM_STR_CASE(VK_COMPARE_OP_NOT_EQUAL)
ENUM_STR_CASE(VK_COMPARE_OP_GREATER_OR_EQUAL)
ENUM_STR_CASE(VK_COMPARE_OP_ALWAYS)
default:
return "VK_COMPARE_OP_UNKNOWN";
}
}
static const char* LogicOpToString(VkLogicOp logicOp) {
switch (logicOp) {
ENUM_STR_CASE(VK_LOGIC_OP_CLEAR)
ENUM_STR_CASE(VK_LOGIC_OP_AND)
ENUM_STR_CASE(VK_LOGIC_OP_AND_REVERSE)
ENUM_STR_CASE(VK_LOGIC_OP_COPY)
ENUM_STR_CASE(VK_LOGIC_OP_AND_INVERTED)
ENUM_STR_CASE(VK_LOGIC_OP_NO_OP)
ENUM_STR_CASE(VK_LOGIC_OP_XOR)
ENUM_STR_CASE(VK_LOGIC_OP_OR)
ENUM_STR_CASE(VK_LOGIC_OP_NOR)
ENUM_STR_CASE(VK_LOGIC_OP_EQUIVALENT)
ENUM_STR_CASE(VK_LOGIC_OP_INVERT)
ENUM_STR_CASE(VK_LOGIC_OP_OR_REVERSE)
ENUM_STR_CASE(VK_LOGIC_OP_COPY_INVERTED)
ENUM_STR_CASE(VK_LOGIC_OP_OR_INVERTED)
ENUM_STR_CASE(VK_LOGIC_OP_NAND)
ENUM_STR_CASE(VK_LOGIC_OP_SET)
default:
return "VK_LOGIC_OP_UNKNOWN";
}
}
PipelineFactory::PipelineFactory(VkDevice device, const VulkanRendererConfig& config):
m_device(device), m_config(config) {
MOBILEGL_ASSERT(m_device != VK_NULL_HANDLE, "PipelineFactory: device is null");
if (m_config.DisablePipelineCache) {
MGLOG_I("DirectVulkan: pipeline cache disabled");
return;
}
VkPipelineCacheCreateInfo pipelineCacheInfo{VK_STRUCTURE_TYPE_PIPELINE_CACHE_CREATE_INFO};
VK_VERIFY(vkCreatePipelineCache(m_device, &pipelineCacheInfo, nullptr, &m_pipelineCache),
"vkCreatePipelineCache");
}
// Must be called once, before any pipeline is created: the flag is not part of the
// pipeline hash, so flipping it mid-life would serve cached pipelines built under the
// old value.
void PipelineFactory::SetSuppressBlendedDepthWrite(Bool enabled) {
s_suppressBlendedDepthWrite = enabled;
}
Bool PipelineFactory::ShouldSuppressBlendedDepthWriteForDevice(MG_Config::QuirkOverride quirkOverride,
Uint32 vendorId) {
static constexpr Uint32 kVendorIdQualcomm = 0x5143;
switch (quirkOverride) {
case MG_Config::QuirkOverride::ForceOn:
return true;
case MG_Config::QuirkOverride::ForceOff:
return false;
case MG_Config::QuirkOverride::Auto:
default:
return vendorId == kVendorIdQualcomm;
}
}
namespace {
// MIN/MAX extremum blending: the signature of a depth-bounds accumulation pass
// (MC 26.3 OIT writes vec4(-linD, linD, deviceZ, 0) under GL_MAX while writing
// depth for its equality chain). MIN/MAX ignore blend factors per the Vulkan spec.
//
// Deliberately the ONLY shape stripped. A quirk should touch as little unrelated
// content as possible, and a trace sweep of every fixture showed the wider
// alternatives all cost more than they fix:
// - additive ONE+ONE with a depth write matched zero draws of the 26.3 chain
// (its transmittance/accumulate passes disable depth writes themselves) - the
// only real content it caught was harmless additive glow effects (Create);
// - sorted-transparency "over" blends (SRC_ALPHA-style) are order-dependent,
// drawn once per surface, and rely on their depth writes for occlusion;
// - separate-alpha accumulation over an over-blending color channel has no
// known pairing with a depth-equality chain (color channel only, see tests).
// If a future workload pairs another blend shape with an equality chain, widen
// this with that evidence in hand rather than pre-emptively.
Bool IsAccumulationBlend(const VkPipelineColorBlendAttachmentState& attachment) {
return attachment.colorBlendOp == VK_BLEND_OP_MIN ||
attachment.colorBlendOp == VK_BLEND_OP_MAX;
}
} // namespace
Bool PipelineFactory::ShouldSuppressDepthWrite(const PipelineCreatePayload& payload) {
if (!payload.depthWriteEnable) {
return false;
}
// A shader that assigns gl_FragDepth supplies depth itself rather than taking the
// pipeline's interpolated Z, so a driver that varies the vertex position math
// between pipelines cannot desynchronize it. (A gl_FragDepth = gl_FragCoord.z
// passthrough is the exception that stays exposed; no known content pairs one with
// an equality chain, and 26.3's composite is a genuine computed-depth writer.)
if (payload.fragmentReplacesDepth) {
return false;
}
for (Uint32 i = 0; i < payload.colorAttachmentCount; ++i) {
const VkPipelineColorBlendAttachmentState& attachment = payload.colorBlendAttachments[i];
if (attachment.blendEnable != VK_TRUE) {
continue;
}
// All color writes masked: blending is moot (depth-prepass pattern that left
// GL_BLEND enabled); stripping the depth write would delete the whole prepass.
if (attachment.colorWriteMask == 0) {
continue;
}
// Any attachment qualifies, not just attachment 0: the 26.3 transmittance pass
// accumulates into a 2-target MRT and must stay stripped.
if (IsAccumulationBlend(attachment)) {
return true;
}
}
return false;
}
PipelineFactory::~PipelineFactory() {
DestroyAll();
if (m_pipelineCache != VK_NULL_HANDLE) {
vkDestroyPipelineCache(m_device, m_pipelineCache, nullptr);
m_pipelineCache = VK_NULL_HANDLE;
}
}
PipelineFactory::HashType PipelineFactory::ComputeHash(const PipelineCreatePayload& payload) const {
XXHASH_VERIFY(XXH64_reset(m_hashState, m_config.CacheVersion));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.programHash, sizeof(payload.programHash)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.vertexInputHash, sizeof(payload.vertexInputHash)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.pipelineLayout, sizeof(payload.pipelineLayout)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.renderPass, sizeof(payload.renderPass)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.colorAttachmentCount, sizeof(payload.colorAttachmentCount)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.rasterizationSamples, sizeof(payload.rasterizationSamples)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.subpass, sizeof(payload.subpass)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.topology, sizeof(payload.topology)));
XXHASH_VERIFY(
XXH64_update(m_hashState, &payload.primitiveRestartEnable, sizeof(payload.primitiveRestartEnable)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.polygonMode, sizeof(payload.polygonMode)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.cullMode, sizeof(payload.cullMode)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.frontFace, sizeof(payload.frontFace)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.depthTestEnable, sizeof(payload.depthTestEnable)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.depthWriteEnable, sizeof(payload.depthWriteEnable)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.depthBiasEnable, sizeof(payload.depthBiasEnable)));
XXHASH_VERIFY(
XXH64_update(m_hashState, &payload.rasterizerDiscardEnable, sizeof(payload.rasterizerDiscardEnable)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.logicOpEnable, sizeof(payload.logicOpEnable)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.stencilTestEnable, sizeof(payload.stencilTestEnable)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.depthCompareOp, sizeof(payload.depthCompareOp)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.logicOp, sizeof(payload.logicOp)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.frontStencilFailOp, sizeof(payload.frontStencilFailOp)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.frontStencilPassOp, sizeof(payload.frontStencilPassOp)));
XXHASH_VERIFY(
XXH64_update(m_hashState, &payload.frontStencilDepthFailOp, sizeof(payload.frontStencilDepthFailOp)));
XXHASH_VERIFY(
XXH64_update(m_hashState, &payload.frontStencilCompareOp, sizeof(payload.frontStencilCompareOp)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.backStencilFailOp, sizeof(payload.backStencilFailOp)));
XXHASH_VERIFY(XXH64_update(m_hashState, &payload.backStencilPassOp, sizeof(payload.backStencilPassOp)));
XXHASH_VERIFY(
XXH64_update(m_hashState, &payload.backStencilDepthFailOp, sizeof(payload.backStencilDepthFailOp)));
XXHASH_VERIFY(
XXH64_update(m_hashState, &payload.backStencilCompareOp, sizeof(payload.backStencilCompareOp)));
XXHASH_VERIFY(
XXH64_update(m_hashState, &payload.fragmentReplacesDepth, sizeof(payload.fragmentReplacesDepth)));
if (payload.colorAttachmentCount > 0) {
XXHASH_VERIFY(XXH64_update(
m_hashState,
payload.colorBlendAttachments.data(),
sizeof(payload.colorBlendAttachments[0]) * payload.colorAttachmentCount));
}
return XXH64_digest(m_hashState);
}
VkPipeline PipelineFactory::GetOrCreatePipeline(const PipelineCreatePayload& payload) {
const HashType hash = ComputeHash(payload);
auto it = m_cache.find(hash);
if (it != m_cache.end()) {
return it->second;
}
VkPipeline pipeline = CreatePipeline(payload);
m_cache.emplace(hash, pipeline);
return pipeline;
}
void PipelineFactory::DestroyAll() {
for (auto& pair : m_cache) {
if (pair.second != VK_NULL_HANDLE) {
vkDestroyPipeline(m_device, pair.second, nullptr);
}
}
m_cache.clear();
}
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");
MOBILEGL_ASSERT(payload.pipelineLayout != VK_NULL_HANDLE, "PipelineFactory: pipelineLayout is null");
MOBILEGL_ASSERT(payload.renderPass != VK_NULL_HANDLE, "PipelineFactory: renderPass is null");
MOBILEGL_ASSERT(payload.colorAttachmentCount <= PipelineCreatePayload::kMaxColorAttachments,
"PipelineFactory: colorAttachmentCount=%u is unexpectedly large",
payload.colorAttachmentCount);
MGLOG_D("PipelineFactory::CreatePipeline: programHash=0x%llx vertexInputHash=0x%llx colorAttachmentCount=%u subpass=%u",
static_cast<unsigned long long>(payload.programHash),
static_cast<unsigned long long>(payload.vertexInputHash),
payload.colorAttachmentCount,
payload.subpass);
static constexpr VkDynamicState kDynamicStates[] = {
VK_DYNAMIC_STATE_VIEWPORT,
VK_DYNAMIC_STATE_SCISSOR,
VK_DYNAMIC_STATE_BLEND_CONSTANTS,
VK_DYNAMIC_STATE_DEPTH_BIAS,
VK_DYNAMIC_STATE_LINE_WIDTH,
VK_DYNAMIC_STATE_STENCIL_COMPARE_MASK,
VK_DYNAMIC_STATE_STENCIL_WRITE_MASK,
VK_DYNAMIC_STATE_STENCIL_REFERENCE
};
VkPipelineDynamicStateCreateInfo dynamicState{};
dynamicState.sType = VK_STRUCTURE_TYPE_PIPELINE_DYNAMIC_STATE_CREATE_INFO;
dynamicState.dynamicStateCount = static_cast<uint32_t>(std::size(kDynamicStates));
dynamicState.pDynamicStates = kDynamicStates;
VkPipelineInputAssemblyStateCreateInfo ia{VK_STRUCTURE_TYPE_PIPELINE_INPUT_ASSEMBLY_STATE_CREATE_INFO};
ia.topology = payload.topology;
ia.primitiveRestartEnable = payload.primitiveRestartEnable ? VK_TRUE : VK_FALSE;
VkPipelineViewportStateCreateInfo vpci{VK_STRUCTURE_TYPE_PIPELINE_VIEWPORT_STATE_CREATE_INFO};
vpci.viewportCount = 1;
vpci.scissorCount = 1;
VkPipelineRasterizationStateCreateInfo raster{VK_STRUCTURE_TYPE_PIPELINE_RASTERIZATION_STATE_CREATE_INFO};
raster.polygonMode = payload.polygonMode;
raster.cullMode = payload.cullMode;
raster.frontFace = payload.frontFace;
raster.depthBiasEnable = payload.depthBiasEnable ? VK_TRUE : VK_FALSE;
raster.rasterizerDiscardEnable = payload.rasterizerDiscardEnable ? VK_TRUE : VK_FALSE;
raster.lineWidth = 1.0f;
VkPipelineMultisampleStateCreateInfo ms{VK_STRUCTURE_TYPE_PIPELINE_MULTISAMPLE_STATE_CREATE_INFO};
ms.rasterizationSamples = payload.rasterizationSamples;
VkPipelineDepthStencilStateCreateInfo depthStencil{VK_STRUCTURE_TYPE_PIPELINE_DEPTH_STENCIL_STATE_CREATE_INFO};
depthStencil.depthTestEnable = payload.depthTestEnable ? VK_TRUE : VK_FALSE;
depthStencil.depthWriteEnable = payload.depthWriteEnable ? VK_TRUE : VK_FALSE;
depthStencil.depthCompareOp = payload.depthCompareOp;
depthStencil.depthBoundsTestEnable = VK_FALSE;
depthStencil.stencilTestEnable = payload.stencilTestEnable ? VK_TRUE : VK_FALSE;
if (payload.stencilTestEnable) {
depthStencil.front.failOp = payload.frontStencilFailOp;
depthStencil.front.passOp = payload.frontStencilPassOp;
depthStencil.front.depthFailOp = payload.frontStencilDepthFailOp;
depthStencil.front.compareOp = payload.frontStencilCompareOp;
depthStencil.front.compareMask = 0xffffffffu;
depthStencil.front.writeMask = 0xffffffffu;
depthStencil.front.reference = 0;
depthStencil.back.failOp = payload.backStencilFailOp;
depthStencil.back.passOp = payload.backStencilPassOp;
depthStencil.back.depthFailOp = payload.backStencilDepthFailOp;
depthStencil.back.compareOp = payload.backStencilCompareOp;
depthStencil.back.compareMask = 0xffffffffu;
depthStencil.back.writeMask = 0xffffffffu;
depthStencil.back.reference = 0;
}
Vector<VkPipelineColorBlendAttachmentState> colorAttachments(payload.colorAttachmentCount);
for (Uint32 i = 0; i < payload.colorAttachmentCount; ++i) {
colorAttachments[i] = payload.colorBlendAttachments[i];
}
// Suppress depth writes on accumulation-blended pipelines when the active driver
// cannot keep vertex positions invariant across the pipelines of a multi-pass
// depth-equality chain (see SetSuppressBlendedDepthWrite). The decision is narrowed
// in ShouldSuppressDepthWrite: sorted-transparency "over" blends (vanilla MC water),
// gl_FragDepth writers, and masked-out attachments keep their depth writes.
// This bakes the decision into the pipeline, which only works because depth write is
// static state here - adding VK_DYNAMIC_STATE_DEPTH_WRITE_ENABLE to kDynamicStates
// would let the record-time value override it and silently disable the quirk.
if (s_suppressBlendedDepthWrite && ShouldSuppressDepthWrite(payload)) {
depthStencil.depthWriteEnable = VK_FALSE;
}
VkPipelineColorBlendStateCreateInfo blend{VK_STRUCTURE_TYPE_PIPELINE_COLOR_BLEND_STATE_CREATE_INFO};
blend.logicOpEnable = payload.logicOpEnable ? VK_TRUE : VK_FALSE;
blend.logicOp = payload.logicOp;
blend.attachmentCount = payload.colorAttachmentCount;
blend.pAttachments = colorAttachments.empty() ? nullptr : colorAttachments.data();
VkGraphicsPipelineCreateInfo gpi{VK_STRUCTURE_TYPE_GRAPHICS_PIPELINE_CREATE_INFO};
gpi.stageCount = static_cast<Uint32>(payload.stages->size());
gpi.pStages = payload.stages->data();
gpi.pVertexInputState = payload.vertexInputState;
gpi.pInputAssemblyState = &ia;
gpi.pViewportState = &vpci;
gpi.pRasterizationState = &raster;
gpi.pMultisampleState = &ms;
gpi.pDepthStencilState = &depthStencil;
gpi.pColorBlendState = &blend;
gpi.pDynamicState = &dynamicState;
gpi.layout = payload.pipelineLayout;
gpi.renderPass = payload.renderPass;
gpi.subpass = payload.subpass;
VkPipeline pipeline = VK_NULL_HANDLE;
const VkResult result = vkCreateGraphicsPipelines(m_device, m_pipelineCache, 1, &gpi, nullptr, &pipeline);
if (result != VK_SUCCESS) {
MGLOG_F("PipelineFactory::CreatePipeline failed: result=%s (%d) programHash=0x%llx vertexInputHash=0x%llx stageCount=%u topology=%s(%d) colorAttachmentCount=%u samples=%s(%d) subpass=%u",
VkResultToString(result),
result,
static_cast<unsigned long long>(payload.programHash),
static_cast<unsigned long long>(payload.vertexInputHash),
gpi.stageCount,
PrimitiveTopologyToString(payload.topology),
payload.topology,
payload.colorAttachmentCount,
SampleCountToString(payload.rasterizationSamples),
payload.rasterizationSamples,
payload.subpass);
MGLOG_F("PipelineFactory::CreatePipeline state: cullMode=%s(0x%x) frontFace=%d depthTest=%d depthWrite=%d depthCompare=%s(%d) depthBias=%d rasterizerDiscard=%d stencilTest=%d logicOpEnable=%d logicOp=%s(%d)",
CullModeToString(payload.cullMode),
static_cast<Uint32>(payload.cullMode),
payload.frontFace,
payload.depthTestEnable ? 1 : 0,
payload.depthWriteEnable ? 1 : 0,
CompareOpToString(payload.depthCompareOp),
payload.depthCompareOp,
payload.depthBiasEnable ? 1 : 0,
payload.rasterizerDiscardEnable ? 1 : 0,
payload.stencilTestEnable ? 1 : 0,
payload.logicOpEnable ? 1 : 0,
LogicOpToString(payload.logicOp),
payload.logicOp);
MGLOG_F("PipelineFactory::CreatePipeline vertex input: bindingCount=%u attributeCount=%u",
payload.vertexInputState->vertexBindingDescriptionCount,
payload.vertexInputState->vertexAttributeDescriptionCount);
for (Uint32 i = 0; i < payload.colorAttachmentCount; ++i) {
const auto& attachment = payload.colorBlendAttachments[i];
MGLOG_F("PipelineFactory::CreatePipeline colorAttachment[%u]: blend=%d colorWriteMask=0x%x srcColor=%d dstColor=%d colorOp=%d srcAlpha=%d dstAlpha=%d alphaOp=%d",
i,
attachment.blendEnable == VK_TRUE ? 1 : 0,
static_cast<Uint32>(attachment.colorWriteMask),
attachment.srcColorBlendFactor,
attachment.dstColorBlendFactor,
attachment.colorBlendOp,
attachment.srcAlphaBlendFactor,
attachment.dstAlphaBlendFactor,
attachment.alphaBlendOp);
}
}
VK_VERIFY(result, "vkCreateGraphicsPipelines");
return pipeline;
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,99 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/PipelineFactory.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 "Config.h"
#include "../VkIncludes.h"
#include "MG_State/GLState/FramebufferState/FramebufferObject.h"
#include <Includes.h>
namespace MobileGL::MG_Backend::DirectVulkan {
class PipelineFactory {
public:
using HashType = Uint64;
struct PipelineCreatePayload {
static constexpr Uint32 kMaxColorAttachments = MG_State::GLState::FramebufferObject::MAX_DRAW_BUFFERS;
HashType programHash = 0;
HashType vertexInputHash = 0;
VkPipelineLayout pipelineLayout = VK_NULL_HANDLE;
VkRenderPass renderPass = VK_NULL_HANDLE;
Uint32 colorAttachmentCount = 1;
VkSampleCountFlagBits rasterizationSamples = VK_SAMPLE_COUNT_1_BIT;
Uint32 subpass = 0;
VkPrimitiveTopology topology = VK_PRIMITIVE_TOPOLOGY_TRIANGLE_LIST;
Bool primitiveRestartEnable = false;
VkPolygonMode polygonMode = VK_POLYGON_MODE_FILL;
VkCullModeFlags cullMode = VK_CULL_MODE_BACK_BIT;
VkFrontFace frontFace = VK_FRONT_FACE_CLOCKWISE;
Bool depthTestEnable = false;
Bool depthWriteEnable = false;
Bool depthBiasEnable = false;
Bool rasterizerDiscardEnable = false;
Bool logicOpEnable = false;
Bool stencilTestEnable = false;
VkCompareOp depthCompareOp = VK_COMPARE_OP_ALWAYS;
VkLogicOp logicOp = VK_LOGIC_OP_COPY;
VkStencilOp frontStencilFailOp = VK_STENCIL_OP_KEEP;
VkStencilOp frontStencilPassOp = VK_STENCIL_OP_KEEP;
VkStencilOp frontStencilDepthFailOp = VK_STENCIL_OP_KEEP;
VkCompareOp frontStencilCompareOp = VK_COMPARE_OP_ALWAYS;
VkStencilOp backStencilFailOp = VK_STENCIL_OP_KEEP;
VkStencilOp backStencilPassOp = VK_STENCIL_OP_KEEP;
VkStencilOp backStencilDepthFailOp = VK_STENCIL_OP_KEEP;
VkCompareOp backStencilCompareOp = VK_COMPARE_OP_ALWAYS;
// The fragment module writes gl_FragDepth (SPIR-V DepthReplacing); exempts the
// pipeline from the blended depth-write quirk (see ShouldSuppressDepthWrite).
Bool fragmentReplacesDepth = false;
Array<VkPipelineColorBlendAttachmentState, kMaxColorAttachments> colorBlendAttachments{};
const Vector<VkPipelineShaderStageCreateInfo>* stages = nullptr;
const VkPipelineVertexInputStateCreateInfo* vertexInputState = nullptr;
};
explicit PipelineFactory(VkDevice device, const VulkanRendererConfig& config);
~PipelineFactory();
PipelineFactory(const PipelineFactory&) = delete;
HashType ComputeHash(const PipelineCreatePayload& payload) const;
VkPipeline GetOrCreatePipeline(const PipelineCreatePayload& payload);
void DestroyAll();
// 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
// cross-pipeline position invariance that some mobile compilers do not provide, even
// with the SPIR-V Invariant decoration; whole primitives then drop out of the later
// passes. Only MIN/MAX extremum blends are stripped - the signature of such a
// chain's depth-bounds pass (MC 26.3 OIT), and per a fixture-wide trace sweep the
// only depth-writing shape the chain actually uses - so every other blend
// (sorted-transparency "over" like vanilla MC water, additive glows, ...) keeps
// its depth writes. Set at renderer initialization based on the active driver.
static void SetSuppressBlendedDepthWrite(Bool enabled);
static Bool IsSuppressBlendedDepthWriteEnabled() { return s_suppressBlendedDepthWrite; }
// Device gate for the quirk: ForceOn/ForceOff bypass detection, Auto enables it on
// the known-affected vendor (Qualcomm).
static Bool ShouldSuppressBlendedDepthWriteForDevice(MG_Config::QuirkOverride quirkOverride,
Uint32 vendorId);
// Pure per-pipeline strip decision (exempts gl_FragDepth writers, masked-out and
// non-accumulation blends); combined with the device flag in CreatePipeline. Static
// and payload-only so tests can pin the contract without a VkDevice.
static Bool ShouldSuppressDepthWrite(const PipelineCreatePayload& payload);
private:
VkPipeline CreatePipeline(const PipelineCreatePayload& payload) const;
VkDevice m_device = VK_NULL_HANDLE;
const VulkanRendererConfig& m_config;
VkPipelineCache m_pipelineCache = VK_NULL_HANDLE;
UnorderedMap<HashType, VkPipeline> m_cache;
static inline XXH64_state_t* m_hashState = XXH64_createState();
static inline Bool s_suppressBlendedDepthWrite = false;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
File diff suppressed because it is too large Load Diff
@@ -1,271 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/ProgramFactory.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 "../VkIncludes.h"
#include "MG_State/GLState/ProgramState/ProgramObject.h"
#include "MG_State/GLState/ProgramState/ShaderObject.h"
#include "MG_State/GLState/TextureState/TextureEnum.h"
#include <Includes.h>
#include <spirv_reflect.h>
namespace MobileGL::MG_Backend::DirectVulkan {
enum class SamplerNumericDomain : Uint8 {
Unknown = 0,
Float,
SignedInteger,
UnsignedInteger,
};
class ProgramFactory {
public:
enum class DescriptorBindingKind : Uint8 {
None = 0,
UniformBufferDynamic,
CombinedImageSampler,
UniformTexelBuffer,
StorageBuffer,
StorageImage
};
enum class CompileOptionBit : Uint {
None = 0,
PositionYFlip = 1 << 0,
PositionZRemap = 1 << 1,
SurfaceRotate90 = 1 << 2,
SurfaceRotate180 = 1 << 3,
SurfaceRotate270 = 1 << 4,
};
using CompileOptionFlags = Flags<CompileOptionBit>;
using HashType = Uint64;
struct VkProgramObject {
static constexpr Uint32 kMaxVertexInputLocations = 32;
HashType hash = 0;
Vector<VkPipelineShaderStageCreateInfo> stages;
Vector<VkShaderModule> modules;
// Layout data (previously in separate VkProgramLayout)
VkDescriptorSetLayout descriptorSetLayout = VK_NULL_HANDLE;
VkPipelineLayout pipelineLayout = VK_NULL_HANDLE;
Vector<DescriptorBindingKind> bindingKinds;
Vector<Uint32> dynamicBindings;
Vector<Int> uniformBlockIndexByBinding;
// Descriptor count per binding (1 except for UBO instance arrays, which occupy one
// binding with descriptorCount = N).
Vector<Uint16> bindingDescriptorCounts;
// Per-element GL uniform block indices for arrayed UBO bindings (count > 1);
// element 0 of a non-arrayed binding stays in uniformBlockIndexByBinding.
UnorderedMap<Uint32, Vector<Int>> arrayedUniformBlockIndicesByBinding;
Vector<String> samplerNameByBinding;
Vector<Int> samplerUniformLocationByBinding;
Vector<TextureTarget> samplerTextureTargetByBinding;
Vector<SamplerNumericDomain> samplerNumericDomainByBinding;
Vector<VkFormat> storageImageFormatByBinding;
Vector<Bool> storageImageUsesBindingFormatByBinding;
Vector<String> storageBlockNameByBinding;
Vector<Int> storageBlockIndexByBinding;
// Set once during ReflectLayout so the per-draw path can skip the whole
// storage-image preparation for the overwhelming majority of programs.
Bool hasStorageImages = false;
Int globalUboBinding = -1;
Uint32 activeVertexInputLocationMask = 0;
Array<GLenum, kMaxVertexInputLocations> vertexInputTypes{};
Uint32 activeFragmentOutputLocationMask = 0;
Array<GLenum, kMaxVertexInputLocations> fragmentOutputTypes{};
ShaderStage rasterizationProducerStage = ShaderStage::Unknown;
Uint32 producerOutputComponentCount = 0;
Uint32 fragmentInputComponentCount = 0;
// The fragment module declares the DepthReplacing execution mode (writes
// gl_FragDepth); shader-computed depth is immune to the cross-pipeline
// position-invariance quirk (see PipelineFactory::ShouldSuppressDepthWrite).
Bool fragmentReplacesDepth = false;
static inline VkDevice s_device = VK_NULL_HANDLE;
VkProgramObject() = default;
VkProgramObject(const VkProgramObject&) = delete;
VkProgramObject& operator=(const VkProgramObject&) = delete;
VkProgramObject(VkProgramObject&& other) noexcept {
hash = other.hash;
stages = std::move(other.stages);
modules = std::move(other.modules);
descriptorSetLayout = other.descriptorSetLayout;
pipelineLayout = other.pipelineLayout;
bindingKinds = std::move(other.bindingKinds);
dynamicBindings = std::move(other.dynamicBindings);
uniformBlockIndexByBinding = std::move(other.uniformBlockIndexByBinding);
bindingDescriptorCounts = std::move(other.bindingDescriptorCounts);
arrayedUniformBlockIndicesByBinding = std::move(other.arrayedUniformBlockIndicesByBinding);
samplerNameByBinding = std::move(other.samplerNameByBinding);
samplerUniformLocationByBinding = std::move(other.samplerUniformLocationByBinding);
samplerTextureTargetByBinding = std::move(other.samplerTextureTargetByBinding);
samplerNumericDomainByBinding = std::move(other.samplerNumericDomainByBinding);
storageImageFormatByBinding = std::move(other.storageImageFormatByBinding);
storageImageUsesBindingFormatByBinding =
std::move(other.storageImageUsesBindingFormatByBinding);
storageBlockNameByBinding = std::move(other.storageBlockNameByBinding);
storageBlockIndexByBinding = std::move(other.storageBlockIndexByBinding);
hasStorageImages = other.hasStorageImages;
globalUboBinding = other.globalUboBinding;
activeVertexInputLocationMask = other.activeVertexInputLocationMask;
vertexInputTypes = other.vertexInputTypes;
activeFragmentOutputLocationMask = other.activeFragmentOutputLocationMask;
fragmentOutputTypes = other.fragmentOutputTypes;
rasterizationProducerStage = other.rasterizationProducerStage;
producerOutputComponentCount = other.producerOutputComponentCount;
fragmentInputComponentCount = other.fragmentInputComponentCount;
fragmentReplacesDepth = other.fragmentReplacesDepth;
other.hash = 0;
other.descriptorSetLayout = VK_NULL_HANDLE;
other.pipelineLayout = VK_NULL_HANDLE;
other.hasStorageImages = false;
other.globalUboBinding = -1;
other.activeVertexInputLocationMask = 0;
other.activeFragmentOutputLocationMask = 0;
other.rasterizationProducerStage = ShaderStage::Unknown;
other.producerOutputComponentCount = 0;
other.fragmentInputComponentCount = 0;
other.fragmentReplacesDepth = false;
}
VkProgramObject& operator=(VkProgramObject&& other) noexcept {
if (this == &other) {
return *this;
}
Destroy();
hash = other.hash;
stages = std::move(other.stages);
modules = std::move(other.modules);
descriptorSetLayout = other.descriptorSetLayout;
pipelineLayout = other.pipelineLayout;
bindingKinds = std::move(other.bindingKinds);
dynamicBindings = std::move(other.dynamicBindings);
uniformBlockIndexByBinding = std::move(other.uniformBlockIndexByBinding);
bindingDescriptorCounts = std::move(other.bindingDescriptorCounts);
arrayedUniformBlockIndicesByBinding = std::move(other.arrayedUniformBlockIndicesByBinding);
samplerNameByBinding = std::move(other.samplerNameByBinding);
samplerUniformLocationByBinding = std::move(other.samplerUniformLocationByBinding);
samplerTextureTargetByBinding = std::move(other.samplerTextureTargetByBinding);
samplerNumericDomainByBinding = std::move(other.samplerNumericDomainByBinding);
storageImageFormatByBinding = std::move(other.storageImageFormatByBinding);
storageImageUsesBindingFormatByBinding =
std::move(other.storageImageUsesBindingFormatByBinding);
storageBlockNameByBinding = std::move(other.storageBlockNameByBinding);
storageBlockIndexByBinding = std::move(other.storageBlockIndexByBinding);
hasStorageImages = other.hasStorageImages;
globalUboBinding = other.globalUboBinding;
activeVertexInputLocationMask = other.activeVertexInputLocationMask;
vertexInputTypes = other.vertexInputTypes;
activeFragmentOutputLocationMask = other.activeFragmentOutputLocationMask;
fragmentOutputTypes = other.fragmentOutputTypes;
rasterizationProducerStage = other.rasterizationProducerStage;
producerOutputComponentCount = other.producerOutputComponentCount;
fragmentInputComponentCount = other.fragmentInputComponentCount;
fragmentReplacesDepth = other.fragmentReplacesDepth;
other.hash = 0;
other.descriptorSetLayout = VK_NULL_HANDLE;
other.pipelineLayout = VK_NULL_HANDLE;
other.hasStorageImages = false;
other.globalUboBinding = -1;
other.activeVertexInputLocationMask = 0;
other.activeFragmentOutputLocationMask = 0;
other.rasterizationProducerStage = ShaderStage::Unknown;
other.producerOutputComponentCount = 0;
other.fragmentInputComponentCount = 0;
other.fragmentReplacesDepth = false;
return *this;
}
~VkProgramObject() {
Destroy();
}
private:
void Destroy() {
if (s_device != VK_NULL_HANDLE) {
if (pipelineLayout != VK_NULL_HANDLE) {
vkDestroyPipelineLayout(s_device, pipelineLayout, nullptr);
pipelineLayout = VK_NULL_HANDLE;
}
if (descriptorSetLayout != VK_NULL_HANDLE) {
vkDestroyDescriptorSetLayout(s_device, descriptorSetLayout, nullptr);
descriptorSetLayout = VK_NULL_HANDLE;
}
for (auto module : modules) {
if (module != VK_NULL_HANDLE) {
vkDestroyShaderModule(s_device, module, nullptr);
}
}
}
modules.clear();
stages.clear();
}
};
explicit ProgramFactory(VkDevice device, const VulkanRendererConfig& config, Uint32 maxBindings = 16,
Bool shaderDrawParametersEnabled = false,
Bool unformattedFloatStorageImagesEnabled = false)
: m_device(device), m_maxBindings(maxBindings), m_config(config),
m_shaderDrawParametersEnabled(shaderDrawParametersEnabled),
m_unformattedFloatStorageImagesEnabled(unformattedFloatStorageImagesEnabled) {
VkProgramObject::s_device = device;
}
~ProgramFactory() = default;
ProgramFactory(const ProgramFactory&) = delete;
HashType ComputeHash(const MG_State::GLState::ProgramObject& program, CompileOptionFlags flags) const;
const VkProgramObject& GetOrCreateProgram(
const MG_State::GLState::ProgramObject& program, CompileOptionFlags flags);
static VkShaderStageFlagBits ToVkStage(ShaderStage stage);
static VkFormat ConvertSpirvImageFormatToVkFormat(SpvImageFormat format);
static SamplerNumericDomain UniformTypeToSamplerNumericDomain(GLenum glType);
// True when any entry point declares the DepthReplacing execution mode, i.e. the
// shader assigns gl_FragDepth. Exposed so the blended depth-write quirk's exemption
// can be pinned by tests. A false negative loses the exemption, so such a shader is
// stripped conservatively and forfeits its depth write.
static Bool ReflectedFragmentReplacesDepth(const SpvReflectShaderModule& reflectModule);
// True when an entry point reads the InstanceIndex builtin. Only gates a diagnostic:
// without shaderDrawParameters such a shader cannot have gl_InstanceID rebased.
static Bool ReflectedReadsInstanceIndexBuiltin(const SpvReflectShaderModule& reflectModule);
private:
struct ProgramLookupCache {
const MG_State::GLState::ProgramObject* program = nullptr;
Uint32 backendStateVersion = 0;
CompileOptionFlags flags{};
HashType hash = 0;
};
static TextureTarget UniformTypeToTextureTarget(GLenum glType);
void ReflectVertexInputs(const Vector<SharedPtr<MG_State::GLState::ShaderObject>>& shaders,
const Vector<Vector<Uint>>& spirv,
VkProgramObject& entry) const;
void ReflectFragmentOutputs(const Vector<SharedPtr<MG_State::GLState::ShaderObject>>& shaders,
const Vector<Vector<Uint>>& spirv,
VkProgramObject& entry) const;
void ReflectLayout(const MG_State::GLState::ProgramObject& program, const Vector<Vector<Uint>>& spirv,
VkProgramObject& entry) const;
VkDevice m_device = VK_NULL_HANDLE;
Uint32 m_maxBindings = 0;
UnorderedMap<HashType, VkProgramObject> m_cache;
const VulkanRendererConfig& m_config;
// True when the device enabled shaderDrawParameters; gates the InstanceIndex rebase pass
// (which needs the DrawParameters capability / gl_BaseInstance builtin).
Bool m_shaderDrawParametersEnabled = false;
// True only when the logical device enabled both
// shaderStorageImageReadWithoutFormat and shaderStorageImageWriteWithoutFormat.
Bool m_unformattedFloatStorageImagesEnabled = false;
mutable ProgramLookupCache m_lastLookup;
static inline XXH64_state_t* m_hashState = XXH64_createState();
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,495 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/SwapchainObject.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 "SwapchainObject.h"
#include "MG_Impl/GLImpl/Framebuffer/GL_Framebuffer.h"
#include "MG_State/GLState/TextureState/TextureObject2D.h"
#if defined(__has_include)
#if __has_include(<vulkan/vk_enum_string_helper.h>)
#include <vulkan/vk_enum_string_helper.h>
#define MOBILEGL_HAS_VK_ENUM_STRING_HELPER 1
#else
#define MOBILEGL_HAS_VK_ENUM_STRING_HELPER 0
#endif
#else
#define MOBILEGL_HAS_VK_ENUM_STRING_HELPER 0
#endif
#if !MOBILEGL_HAS_VK_ENUM_STRING_HELPER
static const char* string_VkFormat(VkFormat) {
return "VkFormat(unknown)";
}
static const char* string_VkColorSpaceKHR(VkColorSpaceKHR) {
return "VkColorSpaceKHR(unknown)";
}
static const char* string_VkPresentModeKHR(VkPresentModeKHR presentMode) {
switch (presentMode) {
case VK_PRESENT_MODE_IMMEDIATE_KHR:
return "VK_PRESENT_MODE_IMMEDIATE_KHR";
case VK_PRESENT_MODE_MAILBOX_KHR:
return "VK_PRESENT_MODE_MAILBOX_KHR";
case VK_PRESENT_MODE_FIFO_KHR:
return "VK_PRESENT_MODE_FIFO_KHR";
case VK_PRESENT_MODE_FIFO_RELAXED_KHR:
return "VK_PRESENT_MODE_FIFO_RELAXED_KHR";
default:
return "VkPresentModeKHR(unknown)";
}
}
static const char* string_VkSurfaceTransformFlagBitsKHR(VkSurfaceTransformFlagBitsKHR) {
return "VkSurfaceTransformFlagBitsKHR(unknown)";
}
#endif
namespace MobileGL::MG_Backend::DirectVulkan {
namespace {
Bool HasStencilComponent(VkFormat format) {
return format == VK_FORMAT_D24_UNORM_S8_UINT || format == VK_FORMAT_D32_SFLOAT_S8_UINT;
}
VkFormat FindSupportedDepthStencilFormat(VkPhysicalDevice physicalDevice) {
const VkFormat candidates[] = {VK_FORMAT_D24_UNORM_S8_UINT, VK_FORMAT_D32_SFLOAT_S8_UINT,
VK_FORMAT_D32_SFLOAT};
for (VkFormat format : candidates) {
VkFormatProperties props{};
vkGetPhysicalDeviceFormatProperties(physicalDevice, format, &props);
if ((props.optimalTilingFeatures & VK_FORMAT_FEATURE_DEPTH_STENCIL_ATTACHMENT_BIT) != 0) {
return format;
}
}
return VK_FORMAT_UNDEFINED;
}
Uint32 FindMemoryType(VkPhysicalDevice physicalDevice, Uint32 typeFilter, VkMemoryPropertyFlags properties) {
VkPhysicalDeviceMemoryProperties memProperties{};
vkGetPhysicalDeviceMemoryProperties(physicalDevice, &memProperties);
for (Uint32 i = 0; i < memProperties.memoryTypeCount; i++) {
if ((typeFilter & (1 << i)) &&
(memProperties.memoryTypes[i].propertyFlags & properties) == properties) {
return i;
}
}
MOBILEGL_ASSERT(false, "Failed to find suitable memory type.");
return 0;
}
} // namespace
SwapchainObject::SwapchainCapabilities SwapchainObject::GetSwapchainCapabilities(VkPhysicalDevice physicalDevice,
VkSurfaceKHR surface) {
SwapchainCapabilities swapchainCapabilities{};
VK_VERIFY(
vkGetPhysicalDeviceSurfaceCapabilitiesKHR(physicalDevice, surface, &swapchainCapabilities.capabilities));
Uint32 formatCount = 0;
VK_VERIFY(vkGetPhysicalDeviceSurfaceFormatsKHR(physicalDevice, surface, &formatCount, nullptr));
if (formatCount != 0) {
swapchainCapabilities.surfaceFormats.resize(formatCount);
VK_VERIFY(vkGetPhysicalDeviceSurfaceFormatsKHR(
physicalDevice, surface, &formatCount, swapchainCapabilities.surfaceFormats.data()));
}
Uint32 presentModeCount = 0;
VK_VERIFY(
vkGetPhysicalDeviceSurfacePresentModesKHR(physicalDevice, surface, &presentModeCount, nullptr));
if (presentModeCount != 0) {
swapchainCapabilities.presentModes.resize(presentModeCount);
VK_VERIFY(vkGetPhysicalDeviceSurfacePresentModesKHR(
physicalDevice, surface, &presentModeCount, swapchainCapabilities.presentModes.data()));
}
return swapchainCapabilities;
}
VkSurfaceFormatKHR SwapchainObject::ChooseSwapchainSurfaceFormat(
const Vector<VkSurfaceFormatKHR>& availableFormats) {
for (const auto& availableFormat : availableFormats) {
if ((availableFormat.format == VK_FORMAT_B8G8R8A8_UNORM ||
availableFormat.format == VK_FORMAT_R8G8B8A8_UNORM) &&
availableFormat.colorSpace == VK_COLOR_SPACE_SRGB_NONLINEAR_KHR) {
return availableFormat;
}
}
for (const auto& availableFormat : availableFormats) {
if ((availableFormat.format == VK_FORMAT_B8G8R8A8_SRGB ||
availableFormat.format == VK_FORMAT_R8G8B8A8_SRGB) &&
availableFormat.colorSpace == VK_COLOR_SPACE_SRGB_NONLINEAR_KHR) {
return availableFormat;
}
}
// TODO: Properly rank other formats
return availableFormats[0];
}
VkPresentModeKHR SwapchainObject::ChooseSwapchainPresentMode(
const Vector<VkPresentModeKHR>& availablePresentModes) {
for (auto desiredPresentMode : s_desiredPresentModes) {
for (const auto& presentMode : availablePresentModes) {
if (presentMode == desiredPresentMode) {
return presentMode;
}
}
}
// TODO: Properly rank other modes
return availablePresentModes[0];
}
void SwapchainObject::Create(VkDevice device, VkPhysicalDevice physicalDevice, VkSurfaceKHR surface,
Uint32 graphicsQueueFamily, Uint32 presentQueueFamily, Uint32 minImageCountHint,
VkExtent2D desiredExtent) {
const auto swapchainCapabilities = GetSwapchainCapabilities(physicalDevice, surface);
MOBILEGL_ASSERT(swapchainCapabilities.IsComplete(),
"SwapchainObject::Create failed: incomplete swapchain capabilities");
MGLOG_I("Got %d surface formats:", swapchainCapabilities.surfaceFormats.size());
for (const auto& sf : swapchainCapabilities.surfaceFormats) {
MGLOG_I(" [%s, %s]", string_VkFormat(sf.format), string_VkColorSpaceKHR(sf.colorSpace));
}
const auto pickedSurfaceFormat = ChooseSwapchainSurfaceFormat(swapchainCapabilities.surfaceFormats);
MGLOG_I("Picked surface format: [%s, %s]", string_VkFormat(pickedSurfaceFormat.format),
string_VkColorSpaceKHR(pickedSurfaceFormat.colorSpace));
MGLOG_I("Got %d present modes:", swapchainCapabilities.presentModes.size());
for (const auto& pm : swapchainCapabilities.presentModes) {
MGLOG_I(" %s", string_VkPresentModeKHR(pm));
}
const auto presentMode = ChooseSwapchainPresentMode(swapchainCapabilities.presentModes);
MGLOG_I("Picked present mode: %s", string_VkPresentModeKHR(presentMode));
const auto& swapchainCaps = swapchainCapabilities.capabilities;
Uint32 targetImageCount = std::max<Uint32>(minImageCountHint, swapchainCaps.minImageCount);
if (swapchainCaps.maxImageCount != 0) {
targetImageCount = std::min(targetImageCount, swapchainCaps.maxImageCount);
}
MGLOG_I("Set minImageCount = %u", targetImageCount);
MGLOG_I("Swapchain currentTransform = %s",
string_VkSurfaceTransformFlagBitsKHR(swapchainCaps.currentTransform));
VkSwapchainCreateInfoKHR createInfo{VK_STRUCTURE_TYPE_SWAPCHAIN_CREATE_INFO_KHR};
createInfo.surface = surface;
createInfo.minImageCount = targetImageCount;
createInfo.imageFormat = pickedSurfaceFormat.format;
createInfo.imageColorSpace = pickedSurfaceFormat.colorSpace;
createInfo.imageExtent = swapchainCaps.currentExtent;
if (createInfo.imageExtent.width == UINT32_MAX || createInfo.imageExtent.height == UINT32_MAX) {
createInfo.imageExtent.width = std::clamp(desiredExtent.width,
swapchainCaps.minImageExtent.width,
swapchainCaps.maxImageExtent.width);
createInfo.imageExtent.height = std::clamp(desiredExtent.height,
swapchainCaps.minImageExtent.height,
swapchainCaps.maxImageExtent.height);
}
const VkExtent2D defaultFramebufferExtent = createInfo.imageExtent;
if (swapchainCaps.currentTransform == VK_SURFACE_TRANSFORM_ROTATE_90_BIT_KHR ||
swapchainCaps.currentTransform == VK_SURFACE_TRANSFORM_ROTATE_270_BIT_KHR) {
std::swap(createInfo.imageExtent.width, createInfo.imageExtent.height);
}
createInfo.imageArrayLayers = 1;
const VkImageUsageFlags requiredImageUsage =
VK_IMAGE_USAGE_COLOR_ATTACHMENT_BIT | VK_IMAGE_USAGE_TRANSFER_DST_BIT;
MOBILEGL_ASSERT((swapchainCaps.supportedUsageFlags & requiredImageUsage) == requiredImageUsage,
"Swapchain does not support required usage flags (COLOR_ATTACHMENT | TRANSFER_DST). "
"supportedUsageFlags=0x%x",
static_cast<Uint32>(swapchainCaps.supportedUsageFlags));
VkImageUsageFlags imageUsage = requiredImageUsage;
if ((swapchainCaps.supportedUsageFlags & VK_IMAGE_USAGE_TRANSFER_SRC_BIT) != 0) {
imageUsage |= VK_IMAGE_USAGE_TRANSFER_SRC_BIT;
}
createInfo.imageUsage = imageUsage;
MGLOG_I("Swapchain imageUsage = 0x%x (supportedUsageFlags = 0x%x)", static_cast<Uint32>(createInfo.imageUsage),
static_cast<Uint32>(swapchainCaps.supportedUsageFlags));
Uint32 queueFamilyIndices[] = {graphicsQueueFamily, presentQueueFamily};
if (graphicsQueueFamily != presentQueueFamily) {
createInfo.imageSharingMode = VK_SHARING_MODE_CONCURRENT;
createInfo.queueFamilyIndexCount = 2;
createInfo.pQueueFamilyIndices = queueFamilyIndices;
} else {
createInfo.imageSharingMode = VK_SHARING_MODE_EXCLUSIVE;
createInfo.queueFamilyIndexCount = 0;
createInfo.pQueueFamilyIndices = nullptr;
}
createInfo.preTransform = swapchainCaps.currentTransform;
MGLOG_I("Set swapchain preTransform = %s", string_VkSurfaceTransformFlagBitsKHR(createInfo.preTransform));
const VkCompositeAlphaFlagBitsKHR compositeAlphaCandidates[] = {
VK_COMPOSITE_ALPHA_OPAQUE_BIT_KHR,
VK_COMPOSITE_ALPHA_PRE_MULTIPLIED_BIT_KHR,
VK_COMPOSITE_ALPHA_POST_MULTIPLIED_BIT_KHR,
VK_COMPOSITE_ALPHA_INHERIT_BIT_KHR
};
createInfo.compositeAlpha = VK_COMPOSITE_ALPHA_OPAQUE_BIT_KHR;
for (auto candidate : compositeAlphaCandidates) {
if ((swapchainCaps.supportedCompositeAlpha & candidate) != 0) {
createInfo.compositeAlpha = candidate;
break;
}
}
createInfo.presentMode = presentMode;
createInfo.clipped = VK_TRUE;
createInfo.oldSwapchain = VK_NULL_HANDLE;
m_surfaceFormat = {createInfo.imageFormat, createInfo.imageColorSpace};
m_extent = createInfo.imageExtent;
m_preTransform = createInfo.preTransform;
VK_VERIFY(vkCreateSwapchainKHR(device, &createInfo, nullptr, &m_swapchain));
Uint32 imageCount = 0;
VK_VERIFY(vkGetSwapchainImagesKHR(device, m_swapchain, &imageCount, nullptr));
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);
CreateImageViews(device);
CreateDepthStencilResources(device, physicalDevice);
MGLOG_I("Swapchain created, extent = %dx%d, swapchain imageCount = %d", m_extent.width, m_extent.height,
imageCount);
// Properly initialize Default FBO here
auto& defaultFBOInfo = MG_Impl::GLImpl::FramebufferImpl::pDefaultFramebufferInfo;
const Int extentWidth = static_cast<Int>(defaultFramebufferExtent.width);
const Int extentHeight = static_cast<Int>(defaultFramebufferExtent.height);
const SizeT defaultAttachmentByteSize =
static_cast<SizeT>(defaultFramebufferExtent.width) *
static_cast<SizeT>(defaultFramebufferExtent.height) * 4;
auto* colorTex = static_cast<MG_State::GLState::TextureObject2D*>(defaultFBOInfo->colorAttachment.get());
colorTex->AllocateStorage(
TextureUploadTarget::Texture2D, 0, {
{extentWidth, extentHeight, 1},
defaultAttachmentByteSize}); // TODO: 4 is format size
TextureInternalFormat depthFormat = TextureInternalFormat::Depth24Stencil8;
switch (m_depthStencilFormat) {
case VK_FORMAT_D24_UNORM_S8_UINT:
depthFormat = TextureInternalFormat::Depth24Stencil8;
break;
case VK_FORMAT_D32_SFLOAT_S8_UINT:
depthFormat = TextureInternalFormat::Depth32FStencil8;
break;
case VK_FORMAT_D32_SFLOAT:
depthFormat = TextureInternalFormat::DepthComponent32F;
break;
default:
depthFormat = TextureInternalFormat::Depth24Stencil8;
break;
}
auto* depthTex = static_cast<MG_State::GLState::TextureObject2D*>(defaultFBOInfo->depthAttachment.get());
depthTex->SetInternalFormat(depthFormat);
depthTex->AllocateStorage(TextureUploadTarget::Texture2D, 0, {
{extentWidth, extentHeight, 1},
defaultAttachmentByteSize}); // TODO: 4 is format size
// The default FBO's stencil attachment must track the swapchain extent:
// FramebufferObject::CheckCompleteness requires every valid attachment
// to share the same dimensions, and Init.cpp leaves a 512x512 placeholder.
// Without this the retrace-layer glReadPixels snapshot fails with
// GL_INVALID_FRAMEBUFFER_OPERATION on DirectVulkan.
TextureInternalFormat stencilFormat = TextureInternalFormat::Depth24Stencil8;
switch (m_depthStencilFormat) {
case VK_FORMAT_D32_SFLOAT_S8_UINT:
stencilFormat = TextureInternalFormat::Depth32FStencil8;
break;
case VK_FORMAT_D24_UNORM_S8_UINT:
stencilFormat = TextureInternalFormat::Depth24Stencil8;
break;
default:
// No stencil plane; mirror the depth format for consistency.
stencilFormat = depthFormat;
break;
}
auto* stencilTex = static_cast<MG_State::GLState::TextureObject2D*>(defaultFBOInfo->stencilAttachment.get());
stencilTex->SetInternalFormat(stencilFormat);
stencilTex->AllocateStorage(TextureUploadTarget::Texture2D, 0, {
{extentWidth, extentHeight, 1},
defaultAttachmentByteSize}); // TODO: 4 is format size
}
void SwapchainObject::CreateDepthStencilResources(VkDevice device, VkPhysicalDevice physicalDevice) {
DestroyDepthStencilResources(device);
const auto imageCount = static_cast<Uint32>(m_images.size());
if (imageCount == 0) {
return;
}
m_depthStencilFormat = FindSupportedDepthStencilFormat(physicalDevice);
MOBILEGL_ASSERT(m_depthStencilFormat != VK_FORMAT_UNDEFINED, "No supported depth/stencil format found.");
m_depthStencilImages.assign(imageCount, VK_NULL_HANDLE);
m_depthStencilImageMemories.assign(imageCount, VK_NULL_HANDLE);
m_depthStencilImageViews.assign(imageCount, VK_NULL_HANDLE);
m_depthStencilImageLayouts.assign(imageCount, VK_IMAGE_LAYOUT_UNDEFINED);
for (Uint32 i = 0; i < imageCount; ++i) {
VkImageCreateInfo imageInfo{};
imageInfo.sType = VK_STRUCTURE_TYPE_IMAGE_CREATE_INFO;
imageInfo.imageType = VK_IMAGE_TYPE_2D;
imageInfo.extent.width = m_extent.width;
imageInfo.extent.height = m_extent.height;
imageInfo.extent.depth = 1;
imageInfo.mipLevels = 1;
imageInfo.arrayLayers = 1;
imageInfo.format = m_depthStencilFormat;
imageInfo.tiling = VK_IMAGE_TILING_OPTIMAL;
imageInfo.initialLayout = VK_IMAGE_LAYOUT_UNDEFINED;
imageInfo.usage = VK_IMAGE_USAGE_DEPTH_STENCIL_ATTACHMENT_BIT | VK_IMAGE_USAGE_TRANSFER_SRC_BIT;
imageInfo.samples = VK_SAMPLE_COUNT_1_BIT;
imageInfo.sharingMode = VK_SHARING_MODE_EXCLUSIVE;
VK_VERIFY(vkCreateImage(device, &imageInfo, nullptr, &m_depthStencilImages[i]), "vkCreateImage(depth)");
VkMemoryRequirements memRequirements{};
vkGetImageMemoryRequirements(device, m_depthStencilImages[i], &memRequirements);
VkMemoryAllocateInfo allocInfo{};
allocInfo.sType = VK_STRUCTURE_TYPE_MEMORY_ALLOCATE_INFO;
allocInfo.allocationSize = memRequirements.size;
allocInfo.memoryTypeIndex =
FindMemoryType(physicalDevice, memRequirements.memoryTypeBits, VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT);
VK_VERIFY(vkAllocateMemory(device, &allocInfo, nullptr, &m_depthStencilImageMemories[i]),
"vkAllocateMemory(depth)");
VK_VERIFY(vkBindImageMemory(device, m_depthStencilImages[i], m_depthStencilImageMemories[i], 0),
"vkBindImageMemory(depth)");
VkImageViewCreateInfo viewInfo{};
viewInfo.sType = VK_STRUCTURE_TYPE_IMAGE_VIEW_CREATE_INFO;
viewInfo.image = m_depthStencilImages[i];
viewInfo.viewType = VK_IMAGE_VIEW_TYPE_2D;
viewInfo.format = m_depthStencilFormat;
viewInfo.subresourceRange.aspectMask = VK_IMAGE_ASPECT_DEPTH_BIT;
if (HasStencilComponent(m_depthStencilFormat)) {
viewInfo.subresourceRange.aspectMask |= VK_IMAGE_ASPECT_STENCIL_BIT;
}
viewInfo.subresourceRange.baseMipLevel = 0;
viewInfo.subresourceRange.levelCount = 1;
viewInfo.subresourceRange.baseArrayLayer = 0;
viewInfo.subresourceRange.layerCount = 1;
VK_VERIFY(vkCreateImageView(device, &viewInfo, nullptr, &m_depthStencilImageViews[i]),
"vkCreateImageView(depth)");
}
}
void SwapchainObject::DestroyDepthStencilResources(VkDevice device) {
for (auto view : m_depthStencilImageViews) {
if (view != VK_NULL_HANDLE) {
vkDestroyImageView(device, view, nullptr);
}
}
m_depthStencilImageViews.clear();
for (auto image : m_depthStencilImages) {
if (image != VK_NULL_HANDLE) {
vkDestroyImage(device, image, nullptr);
}
}
m_depthStencilImages.clear();
for (auto memory : m_depthStencilImageMemories) {
if (memory != VK_NULL_HANDLE) {
vkFreeMemory(device, memory, nullptr);
}
}
m_depthStencilImageMemories.clear();
m_depthStencilImageLayouts.clear();
m_depthStencilFormat = VK_FORMAT_UNDEFINED;
}
void SwapchainObject::Shutdown(VkDevice device) {
DestroyDepthStencilResources(device);
for (auto imageView : m_imageViews) {
vkDestroyImageView(device, imageView, nullptr);
}
m_imageViews.clear();
if (m_swapchain != VK_NULL_HANDLE) {
vkDestroySwapchainKHR(device, m_swapchain, nullptr);
m_swapchain = VK_NULL_HANDLE;
}
m_images.clear();
m_imageLayouts.clear();
m_preTransform = VK_SURFACE_TRANSFORM_IDENTITY_BIT_KHR;
}
VkImage SwapchainObject::GetImage(Uint32 index) const {
MOBILEGL_ASSERT(index < m_images.size(), "Swapchain image index out of range");
return m_images[index];
}
VkImageLayout SwapchainObject::GetImageLayout(Uint32 index) const {
MOBILEGL_ASSERT(index < m_imageLayouts.size(), "Swapchain image layout index out of range");
return m_imageLayouts[index];
}
void SwapchainObject::SetImageLayout(Uint32 index, VkImageLayout layout) {
MOBILEGL_ASSERT(index < m_imageLayouts.size(), "Swapchain image layout index out of range");
m_imageLayouts[index] = layout;
}
VkImage SwapchainObject::GetDepthStencilImage(Uint32 index) const {
MOBILEGL_ASSERT(index < m_depthStencilImages.size(), "Swapchain depth/stencil image index out of range");
return m_depthStencilImages[index];
}
VkImageView SwapchainObject::GetDepthStencilImageView(Uint32 index) const {
MOBILEGL_ASSERT(index < m_depthStencilImageViews.size(),
"Swapchain depth/stencil image view index out of range");
return m_depthStencilImageViews[index];
}
VkImageLayout SwapchainObject::GetDepthStencilImageLayout(Uint32 index) const {
MOBILEGL_ASSERT(index < m_depthStencilImageLayouts.size(),
"Swapchain depth/stencil image layout index out of range");
return m_depthStencilImageLayouts[index];
}
void SwapchainObject::SetDepthStencilImageLayout(Uint32 index, VkImageLayout layout) {
MOBILEGL_ASSERT(index < m_depthStencilImageLayouts.size(),
"Swapchain depth/stencil image layout index out of range");
m_depthStencilImageLayouts[index] = layout;
}
void SwapchainObject::CreateImageViews(VkDevice device) {
m_imageViews.resize(m_images.size(), VK_NULL_HANDLE);
for (SizeT i = 0; i < m_imageViews.size(); i++) {
VkImageViewCreateInfo createInfo{};
createInfo.sType = VK_STRUCTURE_TYPE_IMAGE_VIEW_CREATE_INFO;
createInfo.image = m_images[i];
createInfo.viewType = VK_IMAGE_VIEW_TYPE_2D;
createInfo.format = m_surfaceFormat.format;
createInfo.components.r = VK_COMPONENT_SWIZZLE_IDENTITY;
createInfo.components.g = VK_COMPONENT_SWIZZLE_IDENTITY;
createInfo.components.b = VK_COMPONENT_SWIZZLE_IDENTITY;
createInfo.components.a = VK_COMPONENT_SWIZZLE_IDENTITY;
createInfo.subresourceRange.aspectMask = VK_IMAGE_ASPECT_COLOR_BIT;
createInfo.subresourceRange.baseMipLevel = 0;
createInfo.subresourceRange.levelCount = 1;
createInfo.subresourceRange.baseArrayLayer = 0;
createInfo.subresourceRange.layerCount = 1;
VK_VERIFY(vkCreateImageView(device, &createInfo, nullptr, &m_imageViews[i]));
}
MGLOG_I("Swapchain image views created");
}
#undef VK_VERIFY
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,77 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/SwapchainObject.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 <Includes.h>
#include "../VkIncludes.h"
namespace MobileGL::MG_Backend::DirectVulkan {
class SwapchainObject {
public:
struct SwapchainCapabilities {
VkSurfaceCapabilitiesKHR capabilities;
Vector<VkSurfaceFormatKHR> surfaceFormats;
Vector<VkPresentModeKHR> presentModes;
Bool IsComplete() const {
return !surfaceFormats.empty() && !presentModes.empty();
}
};
static SwapchainCapabilities GetSwapchainCapabilities(VkPhysicalDevice physicalDevice, VkSurfaceKHR surface);
static VkSurfaceFormatKHR ChooseSwapchainSurfaceFormat(const Vector<VkSurfaceFormatKHR>& availableFormats);
static VkPresentModeKHR ChooseSwapchainPresentMode(const Vector<VkPresentModeKHR>& availablePresentModes);
void Create(VkDevice device, VkPhysicalDevice physicalDevice, VkSurfaceKHR surface, Uint32 graphicsQueueFamily,
Uint32 presentQueueFamily, Uint32 minImageCountHint, VkExtent2D desiredExtent);
void Shutdown(VkDevice device);
VkSwapchainKHR GetHandle() const { return m_swapchain; }
const VkSurfaceFormatKHR& GetSurfaceFormat() const { return m_surfaceFormat; }
VkExtent2D GetExtent() const { return m_extent; }
VkSurfaceTransformFlagBitsKHR GetPreTransform() const { return m_preTransform; }
const Vector<VkImage>& GetImages() const { return m_images; }
const Vector<VkImageView>& GetImageViews() const { return m_imageViews; }
VkFormat GetDepthStencilFormat() const { return m_depthStencilFormat; }
const Vector<VkImageView>& GetDepthStencilImageViews() const { return m_depthStencilImageViews; }
VkImage GetDepthStencilImage(Uint32 index) const;
VkImageView GetDepthStencilImageView(Uint32 index) const;
VkImageLayout GetDepthStencilImageLayout(Uint32 index) const;
void SetDepthStencilImageLayout(Uint32 index, VkImageLayout layout);
VkImage GetImage(Uint32 index) const;
VkImageLayout GetImageLayout(Uint32 index) const;
void SetImageLayout(Uint32 index, VkImageLayout layout);
SizeT GetImageCount() const { return m_images.size(); }
private:
void CreateImageViews(VkDevice device);
void CreateDepthStencilResources(VkDevice device, VkPhysicalDevice physicalDevice);
void DestroyDepthStencilResources(VkDevice device);
static constexpr VkPresentModeKHR s_desiredPresentModes[] {
VK_PRESENT_MODE_MAILBOX_KHR,
VK_PRESENT_MODE_IMMEDIATE_KHR,
VK_PRESENT_MODE_FIFO_RELAXED_KHR,
VK_PRESENT_MODE_FIFO_KHR
};
VkSwapchainKHR m_swapchain = VK_NULL_HANDLE;
VkSurfaceFormatKHR m_surfaceFormat{};
VkExtent2D m_extent{};
VkSurfaceTransformFlagBitsKHR m_preTransform = VK_SURFACE_TRANSFORM_IDENTITY_BIT_KHR;
Vector<VkImage> m_images;
Vector<VkImageView> m_imageViews;
Vector<VkImageLayout> m_imageLayouts;
VkFormat m_depthStencilFormat = VK_FORMAT_UNDEFINED;
Vector<VkImage> m_depthStencilImages;
Vector<VkDeviceMemory> m_depthStencilImageMemories;
Vector<VkImageView> m_depthStencilImageViews;
Vector<VkImageLayout> m_depthStencilImageLayouts;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
File diff suppressed because it is too large Load Diff
@@ -1,189 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/UniformDescriptorBinder.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 "ProgramFactory.h"
#include "VkBufferManager.h"
#include "VkSamplerManager.h"
#include "VkTextureManager.h"
#include "../VkIncludes.h"
#include <Includes.h>
namespace MobileGL::MG_State::GLState {
class ITextureObject;
class ProgramObject;
class SamplerObject;
}
namespace MobileGL::MG_Backend::DirectVulkan {
class UniformManager {
public:
struct SamplerBindingOverride {
Uint32 binding = 0;
MG_State::GLState::ITextureObject* texture = nullptr;
const MG_State::GLState::SamplerObject* sampler = nullptr;
VkImageView imageView = VK_NULL_HANDLE;
};
Bool Initialize(VkDevice device, VkBufferManager* bufferManager,
ProgramFactory* programFactory,
VkDeviceSize minUniformBufferOffsetAlignment, Uint32 frameCount,
Uint32 maxBindings = 16, Uint32 setsPerFrame = 64,
VkTextureManager* textureManager = nullptr, VkSamplerManager* samplerManager = nullptr);
void Shutdown();
void BeginFrame(Uint32 frameIndex);
Bool CollectSampledTextures(const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj,
Vector<MG_State::GLState::ITextureObject*>& outTextures);
Bool CollectStorageImageTextures(const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj,
Vector<MG_State::GLState::ITextureObject*>& outTextures) const;
Bool BindProgramUniformBuffers(VkCommandBuffer commandBuffer,
const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj,
Uint32 frameIndex,
VkPipelineBindPoint bindPoint = VK_PIPELINE_BIND_POINT_GRAPHICS,
const SamplerBindingOverride* samplerBindingOverride = nullptr);
// Pure format-policy helper kept public for host regression tests. Formatted storage
// images use their shader qualifier; transformed float images use glBindImageTexture's
// format and never silently fall back to the backing image format.
static VkFormat ResolveStorageImageViewFormat(VkFormat reflectedFormat, GLenum bindingFormat,
VkFormat resourceFormat, Bool useBindingFormat);
private:
struct DescriptorPoolBucket {
VkDescriptorPool handle = VK_NULL_HANDLE;
Uint32 maxSets = 0;
Uint32 allocatedSets = 0;
};
struct DescriptorSetCacheEntry {
Vector<VkDescriptorSet> sets;
Uint32 cursor = 0;
};
struct FrameResources {
Vector<DescriptorPoolBucket> descriptorPools;
UnorderedMap<VkDescriptorSetLayout, DescriptorSetCacheEntry> descriptorSetCacheByLayout;
Vector<VkBufferView> texelBufferViews;
Uint32 activeDescriptorPoolIndex = 0;
Uint32 allocatedSetsThisFrame = 0;
Uint32 peakAllocatedSetsThisFrame = 0;
};
static Bool ResolveSamplerTexture(const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj, Uint32 binding,
SharedPtr<MG_State::GLState::ITextureObject>& outTexture);
// Raw-pointer variant for the per-draw sampled-texture walk (CollectSampledTextures):
// the bound texture stays alive through the draw via GL binding state, so callers that
// only need the pointer skip the SharedPtr copy's atomic refcount churn.
static MG_State::GLState::ITextureObject* ResolveSamplerTextureRaw(
const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj, Uint32 binding);
SharedPtr<MG_State::GLState::ITextureObject> GetFallbackTexture(TextureTarget target) const;
Bool ResolveSamplerDescriptor(VkCommandBuffer commandBuffer, const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj, Uint32 binding,
VkDescriptorImageInfo& outImageInfo) const;
Bool ResolveSamplerDescriptorOverride(const SamplerBindingOverride& samplerBindingOverride,
VkDescriptorImageInfo& outImageInfo) const;
Bool ResolveTexelBufferDescriptor(const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj, Uint32 binding,
Uint32 frameIndex, VkBufferView& outBufferView);
Bool ResolveStorageBufferDescriptor(const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj, Uint32 binding,
VkDescriptorBufferInfo& outBufferInfo) const;
Bool ResolveStorageImageDescriptor(VkCommandBuffer commandBuffer,
const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj, Uint32 binding,
VkDescriptorImageInfo& outImageInfo) const;
// Result of resolving a UBO binding: either a zero-copy direct bind to the app's resident
// VkBuffer (the GLES backend's approach - no per-draw copy) or the CPU payload to upload.
struct UboBindResult {
Bool directBindable = false;
VkBuffer buffer = VK_NULL_HANDLE;
VkDeviceSize range = 0; // reflected block size; constant across draws (hashed)
VkDeviceSize dynamicOffset = 0; // block range start; moves per draw (NOT hashed)
const void* payload = nullptr; // fallback UploadTransient path
VkDeviceSize payloadSize = 0;
};
Bool ResolveUniformBufferPayload(const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj, Uint32 binding,
Uint32 arrayElement, UboBindResult& out) const;
Bool CreateDescriptorPool(Uint32 maxSets, VkDescriptorPool& outPool) const;
Bool GrowFrameDescriptorPool(FrameResources& frame, Uint32 frameIndex);
VkResult AllocateDescriptorSetsFromActivePool(
Uint32 frameIndex, const ProgramFactory::VkProgramObject& programObj, VkDescriptorSet& outDescriptorSet);
VkResult AcquireDescriptorSet(Uint32 frameIndex,
const ProgramFactory::VkProgramObject& programObj,
VkDescriptorSet& outDescriptorSet);
VkDevice m_device = VK_NULL_HANDLE;
VkBufferManager* m_bufferManager = nullptr;
ProgramFactory* m_programFactory = nullptr;
Vector<FrameResources> m_frames;
VkDeviceSize m_minDynamicOffsetAlignment = 1;
Uint32 m_frameCount = 0;
Uint32 m_maxBindings = 0;
Uint32 m_setsPerFrame = 0;
Uint32 m_peakDescriptorSetsObserved = 0;
VkTextureManager* m_textureManager = nullptr;
VkSamplerManager* m_samplerManager = nullptr;
mutable SharedPtr<MG_State::GLState::ITextureObject> m_fallbackTexture2D;
// Per-draw scratch buffers for BindProgramUniformBuffers: reused (clear keeps
// capacity) so the descriptor-write path stops allocating on every draw.
Vector<VkWriteDescriptorSet> m_writesScratch;
Vector<VkDescriptorBufferInfo> m_bufferInfosScratch;
Vector<VkDescriptorImageInfo> m_imageInfosScratch;
Vector<VkBufferView> m_texelBufferViewsScratch;
Vector<Uint32> m_dynamicOffsetsScratch;
// Descriptor-set reuse across consecutive draws (see BindProgramUniformBuffers).
// When a draw's resolved descriptor content is byte-identical to the previous
// draw's, reuse the same VkDescriptorSet and skip AcquireDescriptorSet +
// vkUpdateDescriptorSets - only the bind-time dynamic offsets differ. Reset each
// frame in BeginFrame because the frame's descriptor sets are recycled there.
VkDescriptorSet m_lastBoundDescriptorSet = VK_NULL_HANDLE;
Uint64 m_lastDescriptorSignature = 0;
Bool m_hasLastDescriptor = false;
// 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
// skips recomputing that hash. Across a draw batch the bound sampler set is stable, so a
// binding whose sampler (lifetime id + version, bumped on every setter) and texture
// (lifetime id + params version, bumped on the format/border-color setters that feed the
// key) are unchanged recycles the VkSampler it resolved last draw; a param change bumps
// a version and forces a re-resolve. Both objects are keyed by a never-reused monotonic
// lifetime id, so a freed-and-reallocated sampler or texture at the same heap address
// always gets a fresh id and misses (a raw pointer would false-hit that ABA) - so a
// stale guess can only miss and fall through to the hash, never resolve wrong. Still
// reset each frame alongside the descriptor-set cache. Indexed by binding.
struct SamplerResolveMemo {
Uint64 samplerLifetimeId = 0;
Uint64 textureLifetimeId = 0;
VkSampler sampler = VK_NULL_HANDLE;
Uint16 samplerVersion = 0;
Uint16 textureParamsVersion = 0;
Bool forceNearestFiltering = false;
Bool valid = false;
// ResolveSampledImageViewFormat is pure in (image format, numeric domain), but a
// domain mismatch walks a ~184-entry format table. Memo the resolution per binding
// so a reinterpreted sampler pays that scan once, not once per draw.
VkFormat viewFormatSource = VK_FORMAT_UNDEFINED;
SamplerNumericDomain viewFormatDomain = SamplerNumericDomain::Unknown;
VkFormat viewFormat = VK_FORMAT_UNDEFINED;
Bool viewFormatValid = false;
};
mutable Vector<SamplerResolveMemo> m_samplerResolveMemo;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,64 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VertexInputStateBuilder.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 "VertexInputStateBuilder.h"
namespace MobileGL::MG_Backend::DirectVulkan {
VertexInputStateBuilder::VertexInputStateBuilder() {
Reset();
}
void VertexInputStateBuilder::Reset() {
m_bindings.clear();
m_attributes.clear();
m_state = {};
m_state.sType = VK_STRUCTURE_TYPE_PIPELINE_VERTEX_INPUT_STATE_CREATE_INFO;
m_state.vertexBindingDescriptionCount = 0;
m_state.pVertexBindingDescriptions = nullptr;
m_state.vertexAttributeDescriptionCount = 0;
m_state.pVertexAttributeDescriptions = nullptr;
}
VertexInputStateBuilder& VertexInputStateBuilder::AddBinding(
Uint32 binding, Uint32 stride, VkVertexInputRate inputRate) {
VkVertexInputBindingDescription desc{};
desc.binding = binding;
desc.stride = stride;
desc.inputRate = inputRate;
m_bindings.push_back(desc);
return *this;
}
VertexInputStateBuilder& VertexInputStateBuilder::AddAttribute(
Uint32 location, Uint32 binding, VkFormat format, Uint32 offset) {
VkVertexInputAttributeDescription desc{};
desc.location = location;
desc.binding = binding;
desc.format = format;
desc.offset = offset;
m_attributes.push_back(desc);
return *this;
}
const VkPipelineVertexInputStateCreateInfo& VertexInputStateBuilder::Build() {
m_state.vertexBindingDescriptionCount = static_cast<Uint32>(m_bindings.size());
m_state.pVertexBindingDescriptions = m_bindings.empty() ? nullptr : m_bindings.data();
m_state.vertexAttributeDescriptionCount = static_cast<Uint32>(m_attributes.size());
m_state.pVertexAttributeDescriptions = m_attributes.empty() ? nullptr : m_attributes.data();
return m_state;
}
const Vector<VkVertexInputBindingDescription>& VertexInputStateBuilder::GetBindings() const {
return m_bindings;
}
const Vector<VkVertexInputAttributeDescription>& VertexInputStateBuilder::GetAttributes() const {
return m_attributes;
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,31 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VertexInputStateBuilder.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 <Includes.h>
namespace MobileGL::MG_Backend::DirectVulkan {
class VertexInputStateBuilder {
public:
VertexInputStateBuilder();
void Reset();
VertexInputStateBuilder& AddBinding(
Uint32 binding, Uint32 stride, VkVertexInputRate inputRate = VK_VERTEX_INPUT_RATE_VERTEX);
VertexInputStateBuilder& AddAttribute(Uint32 location, Uint32 binding, VkFormat format, Uint32 offset);
const VkPipelineVertexInputStateCreateInfo& Build();
const Vector<VkVertexInputBindingDescription>& GetBindings() const;
const Vector<VkVertexInputAttributeDescription>& GetAttributes() const;
private:
VkPipelineVertexInputStateCreateInfo m_state{};
Vector<VkVertexInputBindingDescription> m_bindings;
Vector<VkVertexInputAttributeDescription> m_attributes;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,384 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VertexInputStateFactory.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 "VertexInputStateFactory.h"
#include "MG_Util/Converters/MGToStr/DataTypeConverter.h"
#include <utility>
namespace MobileGL::MG_Backend::DirectVulkan {
VertexInputStateFactory::HashType VertexInputStateFactory::ComputeHash(
const MG_State::GLState::VertexArrayObject& vao) const {
XXHASH_VERIFY(XXH64_reset(m_hashState, m_config.CacheVersion));
for (Int i = 0; i < MG_State::GLState::VertexArrayObject::MAX_VERTEX_ATTRIBS; ++i) {
const auto& attr = vao.GetAttribute(i);
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.Enabled, sizeof(attr.Enabled)));
if (!attr.Enabled) {
continue;
}
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.Size, sizeof(attr.Size)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.Type, sizeof(attr.Type)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.Normalized, sizeof(attr.Normalized)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.Stride, sizeof(attr.Stride)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.Offset, sizeof(attr.Offset)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.IsInteger, sizeof(attr.IsInteger)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.IsBgra, sizeof(attr.IsBgra)));
XXHASH_VERIFY(XXH64_update(m_hashState, &attr.Divisor, sizeof(attr.Divisor)));
const SizeT bufferKey = reinterpret_cast<SizeT>(attr.Buffer.get());
XXHASH_VERIFY(XXH64_update(m_hashState, &bufferKey, sizeof(bufferKey)));
}
return XXH64_digest(m_hashState);
}
VertexInputStateFactory::HashType VertexInputStateFactory::GetOrComputeHash(
const MG_State::GLState::VertexArrayObject& vao) const {
HashType hash = 0;
if (!vao.GetBackendHashMemo(hash)) {
hash = ComputeHash(vao);
vao.SetBackendHashMemo(hash);
}
return hash;
}
const VertexInputStateFactory::BackendVertexInputState& VertexInputStateFactory::GetOrCreateVertexInputState(
const MG_State::GLState::VertexArrayObject& vao) {
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()) {
return it->second;
}
VertexInputStateBuilder builder;
Vector<SizeT> bindingBufferKeys;
Vector<SizeT> bindingBaseOffsets;
Vector<Uint32> bindingAttributeLocations;
Vector<Bool> bindingUsesClientMemory;
Vector<VertexStreamConversion> bindingConversions;
Uint32 unsupportedAttribMask = 0;
for (Uint32 location = 0; location < MG_State::GLState::VertexArrayObject::MAX_VERTEX_ATTRIBS; ++location) {
const auto& attr = vao.GetAttribute(location);
if (!attr.Enabled) {
continue;
}
const VkFormat sourceVkFormat =
ToVkVertexFormat(attr.Type, attr.Size, attr.Normalized, attr.IsInteger, attr.IsBgra);
if (sourceVkFormat == VK_FORMAT_UNDEFINED) {
MGLOG_E("Unsupported vertex attribute layout (location=%u, type=%s, size=%d): the array is "
"enabled but cannot be mapped to a VkFormat",
location, MG_Util::ConvertDataTypeToString(attr.Type).c_str(), attr.Size);
unsupportedAttribMask |= (1u << location);
continue;
}
VkFormat vkFormat = sourceVkFormat;
VertexStreamConversion conversion = VertexStreamConversion::None;
if (!SupportsVertexBufferFormat(vkFormat)) {
if (IsScaledIntegerVertexFormat(vkFormat)) {
const VkFormat fallbackFormat = ToFloat32VertexFormat(attr.Size);
if (fallbackFormat != VK_FORMAT_UNDEFINED && SupportsVertexBufferFormat(fallbackFormat)) {
vkFormat = fallbackFormat;
conversion = VertexStreamConversion::ScaledIntegerToFloat32;
MGLOG_W("Vertex attribute location=%u format=%d lacks "
"VK_FORMAT_FEATURE_VERTEX_BUFFER_BIT; using float32 stream format=%d "
"(type=%s size=%d normalized=%s integer=%s)",
location, static_cast<Int>(sourceVkFormat), static_cast<Int>(vkFormat),
MG_Util::ConvertDataTypeToString(attr.Type).c_str(), attr.Size,
attr.Normalized ? "true" : "false", attr.IsInteger ? "true" : "false");
}
}
if (conversion == VertexStreamConversion::None) {
MGLOG_E("Unsupported Vulkan vertex format (location=%u, format=%d, type=%s, size=%d): "
"VK_FORMAT_FEATURE_VERTEX_BUFFER_BIT is unavailable and no semantic fallback exists",
location, static_cast<Int>(sourceVkFormat),
MG_Util::ConvertDataTypeToString(attr.Type).c_str(), attr.Size);
unsupportedAttribMask |= (1u << location);
continue;
}
}
const SizeT attribByteSize = GetAttributeByteSize(attr.Type, attr.Size, attr.IsBgra);
if (attribByteSize == 0) {
MGLOG_E("Vertex attribute with unknown component size (location=%u, type=%s): the array is "
"enabled but cannot be sized",
location, MG_Util::ConvertDataTypeToString(attr.Type).c_str());
unsupportedAttribMask |= (1u << location);
continue;
}
const Uint32 sourceStride =
attr.Stride > 0 ? static_cast<Uint32>(attr.Stride) : static_cast<Uint32>(attribByteSize);
const Bool packedAttribute = attr.Type == DataType::Int2101010Rev ||
attr.Type == DataType::Uint2101010Rev;
const SizeT requiredAlignment = packedAttribute ? attribByteSize : GetComponentSize(attr.Type);
// For a client-memory array attr.Offset holds the raw client pointer, and the
// draw path re-uploads the data to a 16-aligned transient slice with attribute
// offset 0, so only the stride can violate Vulkan's fetch alignment there.
const Bool clientMemoryAttribute = attr.Buffer == nullptr;
if (conversion == VertexStreamConversion::None && requiredAlignment > 1 &&
((sourceStride % requiredAlignment) != 0 ||
(!clientMemoryAttribute && (attr.Offset % requiredAlignment) != 0))) {
// GL accepts arbitrary byte strides and offsets. Core Vulkan vertex fetches do not
// unless VK_EXT_legacy_vertex_attributes is available, so deinterleave this one
// attribute into a tightly packed transient stream without changing its format.
conversion = VertexStreamConversion::Repack;
MGLOG_W("Vertex attribute location=%u uses Vulkan-incompatible alignment "
"(offset=%zu stride=%u required=%zu); using a tightly packed stream",
location, attr.Offset, sourceStride, requiredAlignment);
}
Uint32 stride = sourceStride;
if (conversion == VertexStreamConversion::Repack) {
stride = static_cast<Uint32>(attribByteSize);
} else if (conversion == VertexStreamConversion::ScaledIntegerToFloat32) {
stride = static_cast<Uint32>(attr.Size * static_cast<Int>(sizeof(Float)));
}
const VkVertexInputRate inputRate =
(attr.Divisor == 0) ? VK_VERTEX_INPUT_RATE_VERTEX : VK_VERTEX_INPUT_RATE_INSTANCE;
const SizeT bufferKey = reinterpret_cast<SizeT>(attr.Buffer.get());
const Uint32 binding = static_cast<Uint32>(bindingBufferKeys.size());
bindingBufferKeys.push_back(bufferKey);
bindingBaseOffsets.push_back(attr.Buffer ? attr.Offset : 0);
bindingAttributeLocations.push_back(location);
bindingUsesClientMemory.push_back(attr.Buffer == nullptr);
bindingConversions.push_back(conversion);
builder.AddBinding(binding, stride, inputRate);
builder.AddAttribute(location, binding, vkFormat, 0);
}
const auto& state = builder.Build();
auto& entry = m_cache[hash];
entry.hash = hash;
entry.bindings = builder.GetBindings();
entry.attributes = builder.GetAttributes();
entry.bindingBufferKeys = std::move(bindingBufferKeys);
entry.bindingBaseOffsets = std::move(bindingBaseOffsets);
entry.bindingAttributeLocations = std::move(bindingAttributeLocations);
entry.bindingUsesClientMemory = std::move(bindingUsesClientMemory);
entry.bindingConversions = std::move(bindingConversions);
entry.unsupportedAttribMask = unsupportedAttribMask;
entry.state = state;
entry.state.pVertexBindingDescriptions = entry.bindings.empty() ? nullptr : entry.bindings.data();
entry.state.pVertexAttributeDescriptions = entry.attributes.empty() ? nullptr : entry.attributes.data();
return entry;
}
VkFormat VertexInputStateFactory::ToVkVertexFormat(DataType type, Int size, Bool normalized, Bool isInteger,
Bool isBgra) {
if (isBgra) {
// GL_BGRA: four reversed-order components, always normalized (enforced at validation), only
// legal with GL_UNSIGNED_BYTE or a 2_10_10_10 type. The reversed VkFormats put the
// components back into R,G,B,A order for the shader.
switch (type) {
case DataType::Uint8:
return VK_FORMAT_B8G8R8A8_UNORM;
case DataType::Uint2101010Rev:
return VK_FORMAT_A2R10G10B10_UNORM_PACK32;
case DataType::Int2101010Rev:
return VK_FORMAT_A2R10G10B10_SNORM_PACK32;
default:
return VK_FORMAT_UNDEFINED;
}
}
switch (type) {
case DataType::Uint2101010Rev:
// Packed 2_10_10_10 travels the float-normalizing path only; size is always 4. SNORM/UNORM
// normalize, SSCALED/USCALED cast the packed field to float.
if (isInteger || size != 4) return VK_FORMAT_UNDEFINED;
return normalized ? VK_FORMAT_A2B10G10R10_UNORM_PACK32 : VK_FORMAT_A2B10G10R10_USCALED_PACK32;
case DataType::Int2101010Rev:
if (isInteger || size != 4) return VK_FORMAT_UNDEFINED;
return normalized ? VK_FORMAT_A2B10G10R10_SNORM_PACK32 : VK_FORMAT_A2B10G10R10_SSCALED_PACK32;
case DataType::Float32:
switch (size) {
case 1: return VK_FORMAT_R32_SFLOAT;
case 2: return VK_FORMAT_R32G32_SFLOAT;
case 3: return VK_FORMAT_R32G32B32_SFLOAT;
case 4: return VK_FORMAT_R32G32B32A32_SFLOAT;
default: return VK_FORMAT_UNDEFINED;
}
case DataType::Float16:
// GL_HALF_FLOAT is a floating-point array type: it is never an integer attribute, and
// GL_TRUE for `normalized` is ignored for float types rather than selecting a *NORM format.
if (isInteger) return VK_FORMAT_UNDEFINED;
switch (size) {
case 1: return VK_FORMAT_R16_SFLOAT;
case 2: return VK_FORMAT_R16G16_SFLOAT;
case 3: return VK_FORMAT_R16G16B16_SFLOAT;
case 4: return VK_FORMAT_R16G16B16A16_SFLOAT;
default: return VK_FORMAT_UNDEFINED;
}
case DataType::Int32:
if (!isInteger || normalized) return VK_FORMAT_UNDEFINED;
switch (size) {
case 1: return VK_FORMAT_R32_SINT;
case 2: return VK_FORMAT_R32G32_SINT;
case 3: return VK_FORMAT_R32G32B32_SINT;
case 4: return VK_FORMAT_R32G32B32A32_SINT;
default: return VK_FORMAT_UNDEFINED;
}
case DataType::Uint32:
if (!isInteger || normalized) return VK_FORMAT_UNDEFINED;
switch (size) {
case 1: return VK_FORMAT_R32_UINT;
case 2: return VK_FORMAT_R32G32_UINT;
case 3: return VK_FORMAT_R32G32B32_UINT;
case 4: return VK_FORMAT_R32G32B32A32_UINT;
default: return VK_FORMAT_UNDEFINED;
}
case DataType::Int16:
switch (size) {
case 1:
return isInteger ? VK_FORMAT_R16_SINT : (normalized ? VK_FORMAT_R16_SNORM : VK_FORMAT_R16_SSCALED);
case 2:
return isInteger ? VK_FORMAT_R16G16_SINT
: (normalized ? VK_FORMAT_R16G16_SNORM : VK_FORMAT_R16G16_SSCALED);
case 3:
return isInteger ? VK_FORMAT_R16G16B16_SINT
: (normalized ? VK_FORMAT_R16G16B16_SNORM : VK_FORMAT_R16G16B16_SSCALED);
case 4:
return isInteger ? VK_FORMAT_R16G16B16A16_SINT
: (normalized ? VK_FORMAT_R16G16B16A16_SNORM : VK_FORMAT_R16G16B16A16_SSCALED);
default: return VK_FORMAT_UNDEFINED;
}
case DataType::Uint16:
switch (size) {
case 1:
return isInteger ? VK_FORMAT_R16_UINT : (normalized ? VK_FORMAT_R16_UNORM : VK_FORMAT_R16_USCALED);
case 2:
return isInteger ? VK_FORMAT_R16G16_UINT
: (normalized ? VK_FORMAT_R16G16_UNORM : VK_FORMAT_R16G16_USCALED);
case 3:
return isInteger ? VK_FORMAT_R16G16B16_UINT
: (normalized ? VK_FORMAT_R16G16B16_UNORM : VK_FORMAT_R16G16B16_USCALED);
case 4:
return isInteger ? VK_FORMAT_R16G16B16A16_UINT
: (normalized ? VK_FORMAT_R16G16B16A16_UNORM : VK_FORMAT_R16G16B16A16_USCALED);
default: return VK_FORMAT_UNDEFINED;
}
case DataType::Int8:
switch (size) {
case 1:
return isInteger ? VK_FORMAT_R8_SINT : (normalized ? VK_FORMAT_R8_SNORM : VK_FORMAT_R8_SSCALED);
case 2:
return isInteger ? VK_FORMAT_R8G8_SINT
: (normalized ? VK_FORMAT_R8G8_SNORM : VK_FORMAT_R8G8_SSCALED);
case 3:
return isInteger ? VK_FORMAT_R8G8B8_SINT
: (normalized ? VK_FORMAT_R8G8B8_SNORM : VK_FORMAT_R8G8B8_SSCALED);
case 4:
return isInteger ? VK_FORMAT_R8G8B8A8_SINT
: (normalized ? VK_FORMAT_R8G8B8A8_SNORM : VK_FORMAT_R8G8B8A8_SSCALED);
default: return VK_FORMAT_UNDEFINED;
}
case DataType::Uint8:
switch (size) {
case 1:
return isInteger ? VK_FORMAT_R8_UINT : (normalized ? VK_FORMAT_R8_UNORM : VK_FORMAT_R8_USCALED);
case 2:
return isInteger ? VK_FORMAT_R8G8_UINT
: (normalized ? VK_FORMAT_R8G8_UNORM : VK_FORMAT_R8G8_USCALED);
case 3:
return isInteger ? VK_FORMAT_R8G8B8_UINT
: (normalized ? VK_FORMAT_R8G8B8_UNORM : VK_FORMAT_R8G8B8_USCALED);
case 4:
return isInteger ? VK_FORMAT_R8G8B8A8_UINT
: (normalized ? VK_FORMAT_R8G8B8A8_UNORM : VK_FORMAT_R8G8B8A8_USCALED);
default: return VK_FORMAT_UNDEFINED;
}
default:
return VK_FORMAT_UNDEFINED;
}
}
SizeT VertexInputStateFactory::GetComponentSize(DataType type) {
switch (type) {
case DataType::Int8:
case DataType::Uint8:
return 1;
case DataType::Int16:
case DataType::Uint16:
case DataType::Float16:
return 2;
case DataType::Int32:
case DataType::Uint32:
case DataType::Float32:
case DataType::Fixed32:
return 4;
case DataType::Float64:
return 8;
default:
return 0;
}
}
SizeT VertexInputStateFactory::GetAttributeByteSize(DataType type, Int size, Bool isBgra) {
// The packed 2_10_10_10 types are a single 32-bit word for all 4 components; GL_BGRA is always
// 4 components (GL_UNSIGNED_BYTE x4 = 4 bytes, or a packed word = 4 bytes) -- both are 4 bytes.
if (type == DataType::Int2101010Rev || type == DataType::Uint2101010Rev || isBgra) {
return 4;
}
const SizeT componentSize = GetComponentSize(type);
return componentSize == 0 ? 0 : componentSize * static_cast<SizeT>(size);
}
Bool VertexInputStateFactory::IsScaledIntegerVertexFormat(VkFormat format) {
switch (format) {
case VK_FORMAT_R8_USCALED:
case VK_FORMAT_R8_SSCALED:
case VK_FORMAT_R8G8_USCALED:
case VK_FORMAT_R8G8_SSCALED:
case VK_FORMAT_R8G8B8_USCALED:
case VK_FORMAT_R8G8B8_SSCALED:
case VK_FORMAT_R8G8B8A8_USCALED:
case VK_FORMAT_R8G8B8A8_SSCALED:
case VK_FORMAT_R16_USCALED:
case VK_FORMAT_R16_SSCALED:
case VK_FORMAT_R16G16_USCALED:
case VK_FORMAT_R16G16_SSCALED:
case VK_FORMAT_R16G16B16_USCALED:
case VK_FORMAT_R16G16B16_SSCALED:
case VK_FORMAT_R16G16B16A16_USCALED:
case VK_FORMAT_R16G16B16A16_SSCALED:
return true;
default:
return false;
}
}
VkFormat VertexInputStateFactory::ToFloat32VertexFormat(Int componentCount) {
switch (componentCount) {
case 1: return VK_FORMAT_R32_SFLOAT;
case 2: return VK_FORMAT_R32G32_SFLOAT;
case 3: return VK_FORMAT_R32G32B32_SFLOAT;
case 4: return VK_FORMAT_R32G32B32A32_SFLOAT;
default: return VK_FORMAT_UNDEFINED;
}
}
Bool VertexInputStateFactory::SupportsVertexBufferFormat(VkFormat format) const {
if (m_physicalDevice == VK_NULL_HANDLE || format == VK_FORMAT_UNDEFINED) {
return false;
}
VkFormatProperties properties{};
vkGetPhysicalDeviceFormatProperties(m_physicalDevice, format, &properties);
return (properties.bufferFeatures & VK_FORMAT_FEATURE_VERTEX_BUFFER_BIT) != 0;
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,75 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VertexInputStateFactory.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 "Config.h"
#include "VertexInputStateBuilder.h"
#include "MG_State/GLState/VertexArrayState/VertexArrayObject.h"
#include <Includes.h>
#include "../VkIncludes.h"
namespace MobileGL::MG_Backend::DirectVulkan {
class VertexInputStateFactory {
public:
using HashType = Uint64;
enum class VertexStreamConversion : Uint8 {
None = 0,
Repack,
ScaledIntegerToFloat32,
};
struct BackendVertexInputState {
HashType hash = 0;
Vector<VkVertexInputBindingDescription> bindings;
Vector<VkVertexInputAttributeDescription> attributes;
Vector<SizeT> bindingBufferKeys;
Vector<SizeT> bindingBaseOffsets;
Vector<Uint32> bindingAttributeLocations;
Vector<Bool> bindingUsesClientMemory;
Vector<VertexStreamConversion> bindingConversions;
// Locations whose array is ENABLED but whose GL format has no VkFormat mapping. They are
// 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;
VkPipelineVertexInputStateCreateInfo state{
VK_STRUCTURE_TYPE_PIPELINE_VERTEX_INPUT_STATE_CREATE_INFO
};
};
VertexInputStateFactory(const VulkanRendererConfig& config, VkPhysicalDevice physicalDevice):
m_config(config), m_physicalDevice(physicalDevice) {}
~VertexInputStateFactory() = default;
VertexInputStateFactory(const VertexInputStateFactory&) = delete;
HashType ComputeHash(const MG_State::GLState::VertexArrayObject& vao) const;
// Memoized ComputeHash: reuses the VAO's cached hash while its config version
// is unchanged. Use this on per-draw paths.
HashType GetOrComputeHash(const MG_State::GLState::VertexArrayObject& vao) const;
const BackendVertexInputState& GetOrCreateVertexInputState(
const MG_State::GLState::VertexArrayObject& vao, HashType hash);
const BackendVertexInputState& GetOrCreateVertexInputState(const MG_State::GLState::VertexArrayObject& vao);
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
// an unknown/unsupported type.
static SizeT GetAttributeByteSize(DataType type, Int size, Bool isBgra);
private:
static VkFormat ToVkVertexFormat(DataType type, Int size, Bool normalized, Bool isInteger, Bool isBgra = false);
static Bool IsScaledIntegerVertexFormat(VkFormat format);
static VkFormat ToFloat32VertexFormat(Int componentCount);
Bool SupportsVertexBufferFormat(VkFormat format) const;
const VulkanRendererConfig& m_config;
VkPhysicalDevice m_physicalDevice = VK_NULL_HANDLE;
UnorderedMap<HashType, BackendVertexInputState> m_cache;
static inline XXH64_state_t* m_hashState = XXH64_createState();
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,624 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkBufferManager.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 "VkBufferManager.h"
namespace MobileGL::MG_Backend::DirectVulkan {
namespace {
constexpr VmaAllocationCreateFlags kResidentBufferAllocationFlags =
VMA_ALLOCATION_CREATE_HOST_ACCESS_SEQUENTIAL_WRITE_BIT;
constexpr SizeT kLiveResourcePruneThreshold = 256;
// A zero-copy persistent buffer is created once and never recreated (the app holds
// its mapped pointer), and may be bound to any role, so it carries every usage.
// TRANSFER_DST is added by CreateResidentStorage.
constexpr VkBufferUsageFlags kPersistentBackedUsage =
VK_BUFFER_USAGE_VERTEX_BUFFER_BIT | VK_BUFFER_USAGE_INDEX_BUFFER_BIT |
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;
// 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 =
VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT | VK_MEMORY_PROPERTY_HOST_COHERENT_BIT;
using MG_State::GLState::BackendBufferResource;
using MG_State::GLState::BufferBackendOps;
using MG_State::GLState::BufferObject;
// The manager owned by the active VulkanRenderer; immediate ops route here.
VkBufferManager* g_activeBufferManager = nullptr;
void Ops_Respecify(BufferObject& bufferObject) {
if (g_activeBufferManager) {
g_activeBufferManager->OnRespecify(bufferObject);
}
}
void Ops_SubData(BufferObject& bufferObject, SizeT offset, SizeT size) {
if (g_activeBufferManager) {
g_activeBufferManager->OnSubData(bufferObject, offset, size);
}
}
void Ops_FlushMappedRange(BufferObject& bufferObject, Range1D range,
Flags<BufferMappingAccessBit> appAccess) {
if (g_activeBufferManager) {
g_activeBufferManager->OnFlushMappedRange(bufferObject, range, appAccess);
}
}
void* Ops_AcquirePersistentMap(BufferObject& bufferObject) {
if (g_activeBufferManager) {
return g_activeBufferManager->AcquirePersistentMap(bufferObject);
}
return nullptr;
}
void Ops_OnDestroy(SharedPtr<BackendBufferResource>&& resource) {
if (g_activeBufferManager) {
g_activeBufferManager->OnResourceDestroyed(std::move(resource));
}
// No active manager: the device/allocator is gone or going away and
// Shutdown() already destroyed the storage; dropping the handle here
// must not touch Vulkan. VkBufferResource's dtor destroys via VMA only
// when the allocation is still valid, which Shutdown() cleared.
}
const BufferBackendOps g_vulkanBufferBackendOps = {
.Respecify = Ops_Respecify,
.SubData = Ops_SubData,
.FlushMappedRange = Ops_FlushMappedRange,
.OnDestroy = Ops_OnDestroy,
.AcquirePersistentMap = Ops_AcquirePersistentMap,
};
} // namespace
Bool VkBufferManager::Initialize(const VkBufferManagerInitInfo& initInfo) {
Shutdown();
MOBILEGL_ASSERT(initInfo.allocator != nullptr, "VkBufferManager::Initialize requires valid allocator");
MOBILEGL_ASSERT(initInfo.frameCount > 0, "VkBufferManager::Initialize requires non-zero frame count");
m_initInfo = initInfo;
m_deferredBufferReleases.resize(initInfo.frameCount);
m_deferredResourceReleases.resize(initInfo.frameCount);
m_currentFrameIndex = 0;
m_frameSerial = 1;
m_completedSerialFloor = 0;
if (!InitializeTransientArenas()) {
return false;
}
g_activeBufferManager = this;
MG_State::GLState::SetBufferBackendOps(&g_vulkanBufferBackendOps);
return true;
}
void VkBufferManager::Shutdown() {
if (g_activeBufferManager == this) {
g_activeBufferManager = nullptr;
if (MG_State::GLState::GetBufferBackendOps() == &g_vulkanBufferBackendOps) {
MG_State::GLState::SetBufferBackendOps(nullptr);
}
}
m_transientUploadArena.Shutdown();
DestroyAllDeferredReleases();
ReleaseAllLiveResources();
m_copyProvider = nullptr;
m_initInfo = {};
m_currentFrameIndex = 0;
m_frameSerial = 1;
m_completedSerialFloor = 0;
}
Bool VkBufferManager::RecreateTransientArenas(Uint32 frameCount) {
MOBILEGL_ASSERT(m_initInfo.allocator != nullptr,
"VkBufferManager::RecreateTransientArenas requires initialized manager");
MOBILEGL_ASSERT(frameCount > 0, "VkBufferManager::RecreateTransientArenas requires non-zero frame count");
// Callers guarantee the device is idle around arena recreation.
NotifyDeviceIdle();
m_transientUploadArena.Shutdown();
m_initInfo.frameCount = frameCount;
DestroyAllDeferredReleases();
m_deferredBufferReleases.resize(frameCount);
m_deferredResourceReleases.resize(frameCount);
m_currentFrameIndex = 0;
return InitializeTransientArenas();
}
void VkBufferManager::BeginFrame(Uint32 frameIndex) {
MOBILEGL_ASSERT(frameIndex < m_deferredBufferReleases.size(),
"VkBufferManager::BeginFrame frame index out of range");
m_currentFrameIndex = frameIndex;
++m_frameSerial;
CollectDeferredReleases(frameIndex);
m_transientUploadArena.BeginFrame(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
// remains busy.
if (m_frameSerial > 0) {
m_completedSerialFloor = m_frameSerial - 1;
}
}
void VkBufferManager::NotifyFrameSerialComplete(Uint64 serial) {
// The current serial's work is still being recorded; a completion
// report for it (or beyond) can only come from a stale caller.
if (serial >= m_frameSerial) {
return;
}
m_completedSerialFloor = std::max(m_completedSerialFloor, serial);
}
void VkBufferManager::SetCopyCommandProvider(IBufferCopyCommandProvider* provider) {
m_copyProvider = provider;
}
Uint64 VkBufferManager::GetCompletedSerial() const {
const Uint64 frameCount = m_initInfo.frameCount > 0 ? m_initInfo.frameCount : 1;
const Uint64 completed = m_frameSerial > frameCount ? m_frameSerial - frameCount : 0;
return std::max(completed, m_completedSerialFloor);
}
Bool VkBufferManager::IsResourceBusy(const VkBufferResource& resource) const {
return resource.lastUseSerial > GetCompletedSerial();
}
Bool VkBufferManager::UploadTransient(BufferKind kind, Uint32 frameIndex, const void* data,
VkDeviceSize size, VkDeviceSize alignment, BufferSlice& outSlice) {
(void)kind;
return m_transientUploadArena.Upload(frameIndex, data, size, alignment, outSlice);
}
Bool VkBufferManager::InitializeTransientArenas() {
return m_transientUploadArena.Initialize({
.allocator = m_initInfo.allocator,
.frameCount = m_initInfo.frameCount,
.usage = VK_BUFFER_USAGE_VERTEX_BUFFER_BIT | VK_BUFFER_USAGE_INDEX_BUFFER_BIT |
VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT | VK_BUFFER_USAGE_INDIRECT_BUFFER_BIT |
VK_BUFFER_USAGE_TRANSFER_SRC_BIT,
.memoryUsage = m_initInfo.transientMemoryUsage,
.allocationFlags = m_initInfo.transientAllocationFlags,
.minBufferSize = m_initInfo.minUploadBytes,
.persistentlyMapped = m_initInfo.transientPersistentMapping,
});
}
VkBufferResource* VkBufferManager::ResourceOf(MG_State::GLState::BufferObject& bufferObject) {
return static_cast<VkBufferResource*>(bufferObject.GetBackendResource().get());
}
VkBufferResource* VkBufferManager::GetOrCreateResource(
const SharedPtr<MG_State::GLState::BufferObject>& bufferObject) {
// Return by raw pointer: the resource is owned for its whole lifetime by the BufferObject's
// backend-resource SharedPtr (already set, or set below), so callers that only dereference
// it avoid a static_pointer_cast + SharedPtr refcount inc/dec on every per-draw buffer bind.
const auto& existing = bufferObject->GetBackendResource();
if (existing) {
return static_cast<VkBufferResource*>(existing.get());
}
auto resource = MakeShared<VkBufferResource>();
VkBufferResource* raw = resource.get();
bufferObject->SetBackendResource(resource);
TrackLiveResource(resource);
return raw;
}
void VkBufferManager::TrackLiveResource(const SharedPtr<VkBufferResource>& resource) {
if (m_liveResources.size() >= kLiveResourcePruneThreshold) {
std::erase_if(m_liveResources, [](const WeakPtr<VkBufferResource>& weak) { return weak.expired(); });
}
m_liveResources.push_back(resource);
}
void VkBufferManager::ReleaseAllLiveResources() {
for (auto& weak : m_liveResources) {
if (auto resource = weak.lock()) {
resource->buffer.Destroy();
resource->storageSize = 0;
resource->usageFlags = 0;
resource->lastUseSerial = 0;
resource->pendingFullUpload = true;
resource->transientSlice = {};
resource->transientFrameSerial = 0;
}
}
m_liveResources.clear();
}
Bool VkBufferManager::CreateResidentStorage(VkBufferResource& resource, VkDeviceSize size,
VkBufferUsageFlags usage, VkMemoryPropertyFlags requiredFlags) {
// Staged range copies write resident storage with vkCmdCopyBuffer.
usage |= VK_BUFFER_USAGE_TRANSFER_DST_BIT;
const Bool created = resource.buffer.Create({
.allocator = m_initInfo.allocator,
.size = size,
.usage = usage,
.memoryUsage = VMA_MEMORY_USAGE_AUTO,
.allocationFlags = kResidentBufferAllocationFlags,
.requiredFlags = requiredFlags,
});
if (!created || resource.buffer.Map() == nullptr) {
MGLOG_E("VkBufferManager::CreateResidentStorage failed (size=%llu)",
static_cast<unsigned long long>(size));
resource.buffer.Destroy();
resource.storageSize = 0;
resource.usageFlags = 0;
return false;
}
resource.storageSize = size;
resource.usageFlags = usage;
return true;
}
Bool VkBufferManager::SwapStorageAndUploadAll(VkBufferResource& resource,
MG_State::GLState::BufferObject& bufferObject) {
const VkDeviceSize size = static_cast<VkDeviceSize>(bufferObject.GetSize());
const VkBufferUsageFlags usage = resource.usageFlags;
DeferRelease(std::move(resource.buffer));
if (!CreateResidentStorage(resource, size, usage)) {
resource.pendingFullUpload = true;
return false;
}
if (!resource.buffer.Upload(bufferObject.MappedData(), size, 0)) {
MGLOG_E("VkBufferManager::SwapStorageAndUploadAll: upload failed");
resource.pendingFullUpload = true;
return false;
}
resource.pendingFullUpload = false;
return true;
}
Bool VkBufferManager::StagedRangeCopy(VkBufferResource& resource, MG_State::GLState::BufferObject& bufferObject,
SizeT offset, SizeT size) {
if (!m_copyProvider) {
return false;
}
BufferSlice staging{};
if (!m_transientUploadArena.Upload(m_currentFrameIndex, bufferObject.MappedData() + offset,
static_cast<VkDeviceSize>(size), 16, staging)) {
return false;
}
VkCommandBuffer commandBuffer = m_copyProvider->AcquireBufferCopyCommandBuffer();
if (commandBuffer == VK_NULL_HANDLE) {
return false;
}
// Order the copy after every prior read/write of this buffer, both from
// in-flight frames (submission order) and from commands already recorded
// in this frame's command buffer.
VkMemoryBarrier beforeBarrier{VK_STRUCTURE_TYPE_MEMORY_BARRIER};
beforeBarrier.srcAccessMask = VK_ACCESS_MEMORY_READ_BIT | VK_ACCESS_MEMORY_WRITE_BIT;
beforeBarrier.dstAccessMask = VK_ACCESS_TRANSFER_WRITE_BIT;
vkCmdPipelineBarrier(commandBuffer, VK_PIPELINE_STAGE_ALL_COMMANDS_BIT, VK_PIPELINE_STAGE_TRANSFER_BIT, 0, 1,
&beforeBarrier, 0, nullptr, 0, nullptr);
VkBufferCopy region{};
region.srcOffset = staging.offset;
region.dstOffset = static_cast<VkDeviceSize>(offset);
region.size = static_cast<VkDeviceSize>(size);
vkCmdCopyBuffer(commandBuffer, staging.buffer, resource.buffer.GetHandle(), 1, &region);
VkMemoryBarrier afterBarrier{VK_STRUCTURE_TYPE_MEMORY_BARRIER};
afterBarrier.srcAccessMask = VK_ACCESS_TRANSFER_WRITE_BIT;
afterBarrier.dstAccessMask = VK_ACCESS_MEMORY_READ_BIT | VK_ACCESS_MEMORY_WRITE_BIT;
vkCmdPipelineBarrier(commandBuffer, VK_PIPELINE_STAGE_TRANSFER_BIT, VK_PIPELINE_STAGE_ALL_COMMANDS_BIT, 0, 1,
&afterBarrier, 0, nullptr, 0, nullptr);
resource.lastUseSerial = m_frameSerial;
return true;
}
void VkBufferManager::OnRespecify(MG_State::GLState::BufferObject& bufferObject) {
auto* resource = ResourceOf(bufferObject);
if (!resource) {
return; // lazy: AcquireResidentSlice performs a full upload on creation
}
// Any cached streaming slice refers to the previous contents.
resource->transientFrameSerial = 0;
if (!resource->buffer.IsValid()) {
return; // streaming-only resource: shadow + serial are enough
}
const VkDeviceSize size = static_cast<VkDeviceSize>(bufferObject.GetSize());
if (size == 0) {
DeferRelease(std::move(resource->buffer));
resource->storageSize = 0;
resource->pendingFullUpload = false;
return;
}
if (size != resource->storageSize || IsResourceBusy(*resource)) {
// Conditional orphan: only swap the storage when the old one is
// still referenced by the GPU (or no longer fits).
SwapStorageAndUploadAll(*resource, bufferObject);
return;
}
if (!resource->buffer.Upload(bufferObject.MappedData(), size, 0)) {
MGLOG_E("VkBufferManager::OnRespecify: in-place upload failed");
resource->pendingFullUpload = true;
}
}
void VkBufferManager::OnSubData(MG_State::GLState::BufferObject& bufferObject, SizeT offset, SizeT size) {
auto* resource = ResourceOf(bufferObject);
if (!resource) {
return;
}
resource->transientFrameSerial = 0;
if (!resource->buffer.IsValid() || resource->pendingFullUpload) {
return;
}
if (static_cast<VkDeviceSize>(bufferObject.GetSize()) != resource->storageSize) {
resource->pendingFullUpload = true;
return;
}
if (!IsResourceBusy(*resource)) {
if (!resource->buffer.Upload(bufferObject.MappedData() + offset,
static_cast<VkDeviceSize>(size), static_cast<VkDeviceSize>(offset))) {
MGLOG_E("VkBufferManager::OnSubData: host upload failed");
resource->pendingFullUpload = true;
}
return;
}
// Busy partial write: stage + GPU copy preserves GL ordering within the
// frame and leaves bytes outside the range (possibly GPU-written, e.g.
// SSBO) intact. Fall back to a storage swap if staging is unavailable.
if (!StagedRangeCopy(*resource, bufferObject, offset, size)) {
SwapStorageAndUploadAll(*resource, bufferObject);
}
}
void VkBufferManager::OnFlushMappedRange(MG_State::GLState::BufferObject& bufferObject, Range1D range,
Flags<BufferMappingAccessBit> appAccess) {
auto* resource = ResourceOf(bufferObject);
if (!resource) {
return;
}
resource->transientFrameSerial = 0;
if (!resource->buffer.IsValid() || resource->pendingFullUpload) {
return;
}
if (static_cast<VkDeviceSize>(bufferObject.GetSize()) != resource->storageSize) {
resource->pendingFullUpload = true;
return;
}
const SizeT offset = range.start;
const SizeT size = range.end - range.start;
// GL_MAP_UNSYNCHRONIZED_BIT: the app guarantees it does not overwrite
// data the GPU is still reading; honour it with a direct host write.
if ((appAccess & BufferMappingAccessBit::Unsynchronized) || !IsResourceBusy(*resource)) {
if (!resource->buffer.Upload(bufferObject.MappedData() + offset,
static_cast<VkDeviceSize>(size), static_cast<VkDeviceSize>(offset))) {
MGLOG_E("VkBufferManager::OnFlushMappedRange: host upload failed");
resource->pendingFullUpload = true;
}
return;
}
if (!StagedRangeCopy(*resource, bufferObject, offset, size)) {
SwapStorageAndUploadAll(*resource, bufferObject);
}
}
void VkBufferManager::OnResourceDestroyed(SharedPtr<MG_State::GLState::BackendBufferResource>&& resource) {
if (!resource) {
return;
}
auto vkResource = std::static_pointer_cast<VkBufferResource>(std::move(resource));
if (!vkResource->buffer.IsValid()) {
return;
}
if (m_deferredResourceReleases.empty()) {
vkResource->buffer.Destroy();
return;
}
MOBILEGL_ASSERT(m_currentFrameIndex < m_deferredResourceReleases.size(),
"VkBufferManager::OnResourceDestroyed current frame index out of range");
// Keep the whole resource alive until this frame slot's fence has been
// waited, then the storage is destroyed with it.
m_deferredResourceReleases[m_currentFrameIndex].push_back(std::move(vkResource));
}
void* VkBufferManager::AcquirePersistentMap(MG_State::GLState::BufferObject& bufferObject) {
const VkDeviceSize size = static_cast<VkDeviceSize>(bufferObject.GetSize());
if (size == 0) {
return nullptr;
}
auto resource = std::static_pointer_cast<VkBufferResource>(bufferObject.GetBackendResource());
if (!resource) {
resource = MakeShared<VkBufferResource>();
bufferObject.SetBackendResource(resource);
TrackLiveResource(resource);
}
// Idempotent: an already-backed buffer returns the same mapped base.
if (resource->persistentMapped && resource->buffer.IsValid() && resource->storageSize == size) {
return resource->buffer.GetMappedData();
}
// One-time creation of HOST_VISIBLE + HOST_COHERENT, persistently mapped storage
// carrying every usage (never recreated, so the app's pointer never dangles). Seed
// 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));
if (!CreateResidentStorage(*resource, size, kPersistentBackedUsage, kPersistentBackedRequiredFlags)) {
resource->persistentMapped = false;
resource->storageSize = 0;
resource->usageFlags = 0;
return nullptr;
}
const Uint8* seed = bufferObject.MappedData();
if (seed != nullptr) {
resource->buffer.Upload(seed, size, 0);
}
resource->persistentMapped = true;
resource->pendingFullUpload = false;
resource->storageSize = size;
resource->lastUseSerial = 0;
return resource->buffer.GetMappedData();
}
Bool VkBufferManager::AcquireResidentSlice(BufferKind kind,
const SharedPtr<MG_State::GLState::BufferObject>& bufferObject,
BufferSlice& outSlice) {
const VkBufferUsageFlags requiredUsage = GetVkBufferUsage(kind);
MOBILEGL_ASSERT(requiredUsage != 0, "VkBufferManager::AcquireResidentSlice unsupported buffer kind");
MOBILEGL_ASSERT(bufferObject != nullptr, "VkBufferManager::AcquireResidentSlice requires valid buffer object");
auto resource = GetOrCreateResource(bufferObject);
bufferObject->SyncPersistentMappedRange();
const VkDeviceSize size = static_cast<VkDeviceSize>(bufferObject->GetSize());
if (size == 0) {
MGLOG_E("VkBufferManager::AcquireResidentSlice failed: buffer size is zero");
return false;
}
// Zero-copy persistent buffers already hold the app's live coherent writes in
// host-visible storage carrying every usage; bind directly, no re-upload/staging.
if (resource->persistentMapped && resource->buffer.IsValid() && resource->storageSize == size) {
resource->lastUseSerial = m_frameSerial;
outSlice = resource->buffer.GetSlice(0, size);
return outSlice.IsValid();
}
const Bool needsRecreate = !resource->buffer.IsValid() || resource->storageSize != size ||
((resource->usageFlags & requiredUsage) != requiredUsage) ||
resource->pendingFullUpload;
if (needsRecreate) {
const VkBufferUsageFlags usage = resource->usageFlags | requiredUsage;
DeferRelease(std::move(resource->buffer));
if (!CreateResidentStorage(*resource, size, usage)) {
return false;
}
if (!resource->buffer.Upload(bufferObject->MappedData(), size, 0)) {
MGLOG_E("VkBufferManager::AcquireResidentSlice failed: initial upload failed");
resource->buffer.Destroy();
resource->storageSize = 0;
resource->usageFlags = 0;
return false;
}
resource->pendingFullUpload = false;
}
resource->lastUseSerial = m_frameSerial;
outSlice = resource->buffer.GetSlice(0, size);
return true;
}
Bool VkBufferManager::AcquireStreamedSlice(BufferKind kind,
const SharedPtr<MG_State::GLState::BufferObject>& bufferObject,
BufferSlice& outSlice) {
(void)kind;
MOBILEGL_ASSERT(bufferObject != nullptr, "VkBufferManager::AcquireStreamedSlice requires valid buffer object");
auto resource = GetOrCreateResource(bufferObject);
bufferObject->SyncPersistentMappedRange();
const VkDeviceSize size = static_cast<VkDeviceSize>(bufferObject->GetSize());
if (size == 0) {
MGLOG_E("VkBufferManager::AcquireStreamedSlice failed: buffer size is zero");
return false;
}
const Uint64 changeSerial = bufferObject->GetChangeSerial();
if (resource->transientFrameSerial == m_frameSerial && resource->transientChangeSerial == changeSerial &&
resource->transientSize == size && resource->transientSlice.IsValid()) {
outSlice = resource->transientSlice;
return true;
}
if (!m_transientUploadArena.Upload(m_currentFrameIndex, bufferObject->MappedData(), size, 16,
outSlice)) {
return false;
}
resource->transientSlice = outSlice;
resource->transientFrameSerial = m_frameSerial;
resource->transientChangeSerial = changeSerial;
resource->transientSize = size;
// Streaming path is authoritative now; release resident storage so we do
// not keep a second, stale copy alive (downgrade).
if (resource->buffer.IsValid()) {
DeferRelease(std::move(resource->buffer));
resource->storageSize = 0;
}
return true;
}
void VkBufferManager::DeferRelease(VkBufferObject&& buffer) {
if (!buffer.IsValid()) {
return;
}
if (m_deferredBufferReleases.empty()) {
buffer.Destroy();
return;
}
MOBILEGL_ASSERT(m_currentFrameIndex < m_deferredBufferReleases.size(),
"VkBufferManager::DeferRelease current frame index out of range");
m_deferredBufferReleases[m_currentFrameIndex].push_back(std::move(buffer));
}
void VkBufferManager::CollectDeferredReleases(Uint32 frameIndex) {
MOBILEGL_ASSERT(frameIndex < m_deferredBufferReleases.size(),
"VkBufferManager::CollectDeferredReleases frame index out of range");
m_deferredBufferReleases[frameIndex].clear();
m_deferredResourceReleases[frameIndex].clear();
}
VkBufferUsageFlags VkBufferManager::GetVkBufferUsage(BufferKind kind) {
switch (kind) {
case BufferKind::Vertex:
case BufferKind::Index:
// A GL buffer can be rebound between ARRAY_BUFFER and ELEMENT_ARRAY_BUFFER,
// and may even be used as both within the same draw setup. Keep resident
// vertex/index buffers compatible with both roles from the start so we
// never need to recreate a buffer after it has already been bound.
return VK_BUFFER_USAGE_VERTEX_BUFFER_BIT | VK_BUFFER_USAGE_INDEX_BUFFER_BIT;
case BufferKind::Uniform:
return VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT;
case BufferKind::TextureBuffer:
return VK_BUFFER_USAGE_UNIFORM_TEXEL_BUFFER_BIT;
case BufferKind::ShaderStorage:
return VK_BUFFER_USAGE_STORAGE_BUFFER_BIT | VK_BUFFER_USAGE_INDIRECT_BUFFER_BIT;
case BufferKind::Indirect:
return VK_BUFFER_USAGE_INDIRECT_BUFFER_BIT | VK_BUFFER_USAGE_STORAGE_BUFFER_BIT;
default:
return 0;
}
}
void VkBufferManager::DestroyAllDeferredReleases() {
for (auto& releases : m_deferredBufferReleases) {
for (auto& buffer : releases) {
buffer.Destroy();
}
releases.clear();
}
m_deferredBufferReleases.clear();
for (auto& releases : m_deferredResourceReleases) {
for (auto& resource : releases) {
resource->buffer.Destroy();
}
releases.clear();
}
m_deferredResourceReleases.clear();
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,153 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkBufferManager.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 "BufferArena.h"
#include "MG_State/GLState/BufferState/BufferObject.h"
#include "../VkIncludes.h"
#include <Includes.h>
#include <vk_mem_alloc.h>
namespace MobileGL::MG_Backend::DirectVulkan {
enum class BufferKind : Uint8 {
Vertex,
Index,
Uniform,
TextureBuffer,
ShaderStorage,
Indirect,
};
struct VkBufferManagerInitInfo {
VmaAllocator allocator = nullptr;
Uint32 frameCount = 0;
VkDeviceSize minUploadBytes = 4 * 1024 * 1024;
VmaMemoryUsage transientMemoryUsage = VMA_MEMORY_USAGE_AUTO;
VmaAllocationCreateFlags transientAllocationFlags = VMA_ALLOCATION_CREATE_HOST_ACCESS_SEQUENTIAL_WRITE_BIT;
Bool transientPersistentMapping = false;
};
// The DirectVulkan storage behind one frontend buffer (pipe_resource analogue).
// Owned (refcounted) by the frontend BufferObject; the manager holds only weak
// references (for shutdown) plus strong references on deferred-release lists.
class VkBufferResource : public MG_State::GLState::BackendBufferResource {
public:
~VkBufferResource() override = default;
// Resident storage (may be invalid for streaming-only buffers).
VkBufferObject buffer;
VkDeviceSize storageSize = 0;
VkBufferUsageFlags usageFlags = 0;
// Frame serial of the last GPU reference; drives busy tracking.
Uint64 lastUseSerial = 0;
// Set when an immediate op could not be applied; forces a full re-upload
// on the next AcquireResidentSlice.
Bool pendingFullUpload = false;
// Backs a zero-copy coherent persistent map (PipeResource GPU residency): the
// buffer is HOST_VISIBLE+COHERENT, persistently mapped, carries every usage and is
// never orphaned or recreated. Draw-time acquire binds it directly, no re-upload.
Bool persistentMapped = false;
// Cached transient (streaming) slice for the current frame.
BufferSlice transientSlice{};
Uint64 transientFrameSerial = 0;
Uint64 transientChangeSerial = 0;
VkDeviceSize transientSize = 0;
};
// Supplies a command buffer that is recording and outside any render pass,
// for staged buffer-range copies. Implemented by VulkanRenderer.
class IBufferCopyCommandProvider {
public:
virtual ~IBufferCopyCommandProvider() = default;
virtual VkCommandBuffer AcquireBufferCopyCommandBuffer() = 0;
};
class VkBufferManager {
public:
Bool Initialize(const VkBufferManagerInitInfo& initInfo);
void Shutdown();
// Recreate all per-frame transient arenas
Bool RecreateTransientArenas(Uint32 frameCount);
void BeginFrame(Uint32 frameIndex);
// All previously submitted GPU work has completed (vkDeviceWaitIdle).
void NotifyDeviceIdle();
// A frame slot's submission fence has been waited: every serial up to
// and including `serial` is complete. Raises the completed floor so
// GetCompletedSerial reflects real fence progress instead of only the
// frameSerial-minus-frameCount inference.
void NotifyFrameSerialComplete(Uint64 serial);
void SetCopyCommandProvider(IBufferCopyCommandProvider* provider);
Bool UploadTransient(BufferKind kind, Uint32 frameIndex, const void* data, VkDeviceSize size,
VkDeviceSize alignment, BufferSlice& outSlice);
// Draw-time acquire for resident (device-storage) buffers: ensures the
// resource exists and is fully uploaded, marks it used this frame.
Bool AcquireResidentSlice(BufferKind kind, const SharedPtr<MG_State::GLState::BufferObject>& bufferObject,
BufferSlice& outSlice);
// Draw-time acquire for streamed buffers: uploads the whole shadow into
// the per-frame arena (cached by change serial), releasing any resident
// storage the buffer may still own.
Bool AcquireStreamedSlice(BufferKind kind, const SharedPtr<MG_State::GLState::BufferObject>& bufferObject,
BufferSlice& outSlice);
// Zero-copy persistent map (PipeResource GPU residency): create (once) a
// HOST_VISIBLE+COHERENT, persistently mapped resident buffer carrying every usage,
// seed it from the shadow, and return its mapped base for the app to write into
// directly. Idempotent. Returns nullptr on failure (frontend keeps its shadow).
void* AcquirePersistentMap(MG_State::GLState::BufferObject& bufferObject);
// Immediate ops, dispatched from the frontend BufferBackendOps table.
void OnRespecify(MG_State::GLState::BufferObject& bufferObject);
void OnSubData(MG_State::GLState::BufferObject& bufferObject, SizeT offset, SizeT size);
void OnFlushMappedRange(MG_State::GLState::BufferObject& bufferObject, Range1D range,
Flags<BufferMappingAccessBit> appAccess);
void OnResourceDestroyed(SharedPtr<MG_State::GLState::BackendBufferResource>&& resource);
Uint64 GetFrameSerial() const { return m_frameSerial; }
// Highest frame serial whose GPU work is known complete; serials at or
// below it may be considered signaled. Drives IsResourceBusy and the
// backend GL fence objects.
Uint64 GetCompletedSerial() const;
// Busy = potentially referenced by GPU work that has not been fenced yet
// (including commands recorded for the current, unsubmitted frame).
Bool IsResourceBusy(const VkBufferResource& resource) const;
private:
Bool InitializeTransientArenas();
static VkBufferUsageFlags GetVkBufferUsage(BufferKind kind);
VkBufferResource* GetOrCreateResource(const SharedPtr<MG_State::GLState::BufferObject>& bufferObject);
static VkBufferResource* ResourceOf(MG_State::GLState::BufferObject& bufferObject);
Bool CreateResidentStorage(VkBufferResource& resource, VkDeviceSize size, VkBufferUsageFlags usage,
VkMemoryPropertyFlags requiredFlags = 0);
// Swap storage (conditional orphan) and refill it from the shadow copy.
Bool SwapStorageAndUploadAll(VkBufferResource& resource, MG_State::GLState::BufferObject& bufferObject);
// Record a staging-slice copy into the resident storage, ordered against
// in-flight and already-recorded GPU work.
Bool StagedRangeCopy(VkBufferResource& resource, MG_State::GLState::BufferObject& bufferObject,
SizeT offset, SizeT size);
void DeferRelease(VkBufferObject&& buffer);
void CollectDeferredReleases(Uint32 frameIndex);
void DestroyAllDeferredReleases();
void TrackLiveResource(const SharedPtr<VkBufferResource>& resource);
void ReleaseAllLiveResources();
VkBufferManagerInitInfo m_initInfo{};
BufferArena m_transientUploadArena;
IBufferCopyCommandProvider* m_copyProvider = nullptr;
Vector<Vector<VkBufferObject>> m_deferredBufferReleases;
Vector<Vector<SharedPtr<VkBufferResource>>> m_deferredResourceReleases;
Vector<WeakPtr<VkBufferResource>> m_liveResources;
Uint32 m_currentFrameIndex = 0;
Uint64 m_frameSerial = 1;
Uint64 m_completedSerialFloor = 0;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,191 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkBufferObject.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 "VkBufferObject.h"
namespace MobileGL::MG_Backend::DirectVulkan {
VkBufferObject::VkBufferObject(VkBufferObject&& other) noexcept {
m_allocator = other.m_allocator;
m_buffer = other.m_buffer;
m_allocation = other.m_allocation;
m_mappedData = other.m_mappedData;
m_size = other.m_size;
other.m_allocator = nullptr;
other.m_buffer = VK_NULL_HANDLE;
other.m_allocation = nullptr;
other.m_mappedData = nullptr;
other.m_size = 0;
}
VkBufferObject& VkBufferObject::operator=(VkBufferObject&& other) noexcept {
if (this == &other) {
return *this;
}
Destroy();
m_allocator = other.m_allocator;
m_buffer = other.m_buffer;
m_allocation = other.m_allocation;
m_mappedData = other.m_mappedData;
m_size = other.m_size;
other.m_allocator = nullptr;
other.m_buffer = VK_NULL_HANDLE;
other.m_allocation = nullptr;
other.m_mappedData = nullptr;
other.m_size = 0;
return *this;
}
VkBufferObject::~VkBufferObject() {
Destroy();
}
Bool VkBufferObject::Create(const VkBufferObjectDesc& desc) {
return Create(desc.allocator, desc.size, desc.usage, desc.memoryUsage, desc.allocationFlags,
desc.requiredFlags);
}
Bool VkBufferObject::Create(VmaAllocator allocator, VkDeviceSize size, VkBufferUsageFlags usage,
VmaMemoryUsage memoryUsage, VmaAllocationCreateFlags allocationFlags,
VkMemoryPropertyFlags requiredFlags) {
MOBILEGL_ASSERT(allocator != nullptr, "VkBufferObject::Create requires valid VMA allocator");
MOBILEGL_ASSERT(size > 0, "VkBufferObject::Create requires non-zero size");
Destroy();
m_allocator = allocator;
VkBufferCreateInfo bufferInfo{};
bufferInfo.sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO;
bufferInfo.size = size;
bufferInfo.usage = usage;
bufferInfo.sharingMode = VK_SHARING_MODE_EXCLUSIVE;
VmaAllocationCreateInfo allocationInfo{};
allocationInfo.usage = memoryUsage;
allocationInfo.flags = allocationFlags;
allocationInfo.requiredFlags = requiredFlags;
const VkResult result =
vmaCreateBuffer(m_allocator, &bufferInfo, &allocationInfo, &m_buffer, &m_allocation, nullptr);
if (result != VK_SUCCESS) {
MGLOG_E("VkBufferObject::Create failed: vmaCreateBuffer returned %d", result);
m_allocator = nullptr;
m_buffer = VK_NULL_HANDLE;
m_allocation = nullptr;
m_size = 0;
return false;
}
m_size = size;
return true;
}
void VkBufferObject::Destroy() {
Unmap();
if (m_allocator != nullptr && m_buffer != VK_NULL_HANDLE && m_allocation != nullptr) {
vmaDestroyBuffer(m_allocator, m_buffer, m_allocation);
}
m_buffer = VK_NULL_HANDLE;
m_allocation = nullptr;
m_allocator = nullptr;
m_size = 0;
}
void* VkBufferObject::Map() {
MOBILEGL_ASSERT(IsValid(), "VkBufferObject::Map called on invalid buffer");
if (m_mappedData != nullptr) {
return m_mappedData;
}
const VkResult mapResult = vmaMapMemory(m_allocator, m_allocation, &m_mappedData);
if (mapResult != VK_SUCCESS || m_mappedData == nullptr) {
MGLOG_E("VkBufferObject::Map failed: vmaMapMemory returned %d", mapResult);
m_mappedData = nullptr;
return nullptr;
}
return m_mappedData;
}
void VkBufferObject::Unmap() {
if (!IsValid() || m_mappedData == nullptr) {
m_mappedData = nullptr;
return;
}
vmaUnmapMemory(m_allocator, m_allocation);
m_mappedData = nullptr;
}
Bool VkBufferObject::Upload(const void* data, VkDeviceSize size, VkDeviceSize offset) {
MOBILEGL_ASSERT(IsValid(), "VkBufferObject::Upload called on invalid buffer");
MOBILEGL_ASSERT(data != nullptr || size == 0, "VkBufferObject::Upload data pointer is null");
MOBILEGL_ASSERT(offset + size <= m_size, "VkBufferObject::Upload out of range");
if (size == 0) {
return true;
}
const Bool wasMapped = IsMapped();
void* mapped = wasMapped ? m_mappedData : Map();
if (mapped == nullptr) {
MGLOG_E("VkBufferObject::Upload failed: unable to map buffer");
return false;
}
Memcpy(static_cast<Uint8*>(mapped) + offset, data, static_cast<SizeT>(size));
const VkResult flushResult = vmaFlushAllocation(m_allocator, m_allocation, offset, size);
if (flushResult != VK_SUCCESS) {
MGLOG_E("VkBufferObject::Upload failed: vmaFlushAllocation returned %d", flushResult);
if (!wasMapped) {
Unmap();
}
return false;
}
if (!wasMapped) {
Unmap();
}
return true;
}
Bool VkBufferObject::Invalidate(VkDeviceSize size, VkDeviceSize offset) {
MOBILEGL_ASSERT(IsValid(), "VkBufferObject::Invalidate called on invalid buffer");
MOBILEGL_ASSERT(IsMapped(), "VkBufferObject::Invalidate requires mapped memory");
MOBILEGL_ASSERT(offset <= m_size, "VkBufferObject::Invalidate offset out of range");
const VkDeviceSize resolvedSize = size == VK_WHOLE_SIZE ? m_size - offset : size;
MOBILEGL_ASSERT(offset + resolvedSize <= m_size, "VkBufferObject::Invalidate range out of bounds");
if (resolvedSize == 0) {
return true;
}
const VkResult result = vmaInvalidateAllocation(m_allocator, m_allocation, offset, resolvedSize);
if (result != VK_SUCCESS) {
MGLOG_E("VkBufferObject::Invalidate failed: vmaInvalidateAllocation returned %d", result);
return false;
}
return true;
}
BufferSlice VkBufferObject::GetSlice(VkDeviceSize offset, VkDeviceSize size) const {
MOBILEGL_ASSERT(offset <= m_size, "VkBufferObject::GetSlice offset out of range");
const VkDeviceSize resolvedSize = (size == VK_WHOLE_SIZE) ? (m_size - offset) : size;
MOBILEGL_ASSERT(offset + resolvedSize <= m_size, "VkBufferObject::GetSlice range out of bounds");
BufferSlice slice{};
slice.buffer = m_buffer;
slice.offset = offset;
slice.size = resolvedSize;
slice.mapped = (m_mappedData != nullptr) ? static_cast<Uint8*>(m_mappedData) + offset : nullptr;
return slice;
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,63 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkBufferObject.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 "BufferSlice.h"
#include "../VkIncludes.h"
#include <Includes.h>
#include <vk_mem_alloc.h>
namespace MobileGL::MG_Backend::DirectVulkan {
struct VkBufferObjectDesc {
VmaAllocator allocator = nullptr;
VkDeviceSize size = 0;
VkBufferUsageFlags usage = 0;
VmaMemoryUsage memoryUsage = VMA_MEMORY_USAGE_AUTO;
VmaAllocationCreateFlags allocationFlags = 0;
// Memory property bits the allocation MUST satisfy (e.g. HOST_VISIBLE|HOST_COHERENT
// for a persistently-mapped buffer the app writes into without explicit flushes).
VkMemoryPropertyFlags requiredFlags = 0;
};
class VkBufferObject {
public:
VkBufferObject() = default;
~VkBufferObject();
VkBufferObject(const VkBufferObject&) = delete;
VkBufferObject& operator=(const VkBufferObject&) = delete;
VkBufferObject(VkBufferObject&& other) noexcept;
VkBufferObject& operator=(VkBufferObject&& other) noexcept;
Bool Create(const VkBufferObjectDesc& desc);
Bool Create(VmaAllocator allocator, VkDeviceSize size, VkBufferUsageFlags usage,
VmaMemoryUsage memoryUsage, VmaAllocationCreateFlags allocationFlags = 0,
VkMemoryPropertyFlags requiredFlags = 0);
void Destroy();
void* Map();
void Unmap();
Bool Upload(const void* data, VkDeviceSize size, VkDeviceSize offset = 0);
Bool Invalidate(VkDeviceSize size = VK_WHOLE_SIZE, VkDeviceSize offset = 0);
VkBuffer GetHandle() const { return m_buffer; }
VkDeviceSize GetSize() const { return m_size; }
BufferSlice GetSlice(VkDeviceSize offset = 0, VkDeviceSize size = VK_WHOLE_SIZE) const;
void* GetMappedData() const { return m_mappedData; }
Bool IsMapped() const { return m_mappedData != nullptr; }
Bool IsValid() const { return m_allocator != nullptr && m_buffer != VK_NULL_HANDLE && m_allocation != nullptr; }
private:
VmaAllocator m_allocator = nullptr;
VkBuffer m_buffer = VK_NULL_HANDLE;
VmaAllocation m_allocation = nullptr;
void* m_mappedData = nullptr;
VkDeviceSize m_size = 0;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,402 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkClearManager.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 "VkClearManager.h"
#include "MG_Util/Converters/MGToStr/FramebufferEnumConverter.h"
#include "MG_Util/Converters/MGToStr/TextureEnumConverter.h"
namespace MobileGL::MG_Backend::DirectVulkan {
static Bool IsCubeMapFaceUploadTarget(TextureUploadTarget target) {
return target >= TextureUploadTarget::CubeMapPositiveX &&
target <= TextureUploadTarget::CubeMapNegativeZ;
}
static Bool PendingClearMatchesTextureIdentity(const PendingClearKey& key, const TextureIdentity& identity) {
return key.texture == identity.texture && key.textureLifetimeId == identity.lifetimeId;
}
static Uint32 ResolveAttachmentBaseArrayLayer(TextureUploadTarget target) {
if (!IsCubeMapFaceUploadTarget(target)) {
return 0;
}
return static_cast<Uint32>(target) - static_cast<Uint32>(TextureUploadTarget::CubeMapPositiveX);
}
static Uint32 ResolveAttachmentBaseArrayLayer(
const MG_State::GLState::FramebufferAttachmentObject& attachment) {
if (attachment.IsLayered()) {
return 0;
}
const TextureUploadTarget uploadTarget = attachment.GetTextureUploadTarget();
if (!IsCubeMapFaceUploadTarget(uploadTarget)) {
return static_cast<Uint32>(std::max(attachment.GetTextureLayer(), 0));
}
return ResolveAttachmentBaseArrayLayer(uploadTarget);
}
static Uint32 ResolveAttachmentLayerCount(
const MG_State::GLState::FramebufferAttachmentObject& attachment) {
if (attachment.IsLayered()) {
return static_cast<Uint32>(std::max(attachment.GetSize().z(), 1));
}
return 1u;
}
static const MG_State::GLState::FramebufferAttachmentObject* GetClearableAttachment(
const MG_State::GLState::FramebufferObject& drawFbo, FramebufferAttachmentType attachmentType) {
if (attachmentType == FramebufferAttachmentType::None) {
return nullptr;
}
const auto& attachment = drawFbo.GetAttachment(attachmentType);
if (!attachment.IsTexture() || attachment.IsRenderbuffer()) {
return nullptr;
}
return &attachment;
}
PendingClearKey VkClearManager::MakePendingClearKey(MG_State::GLState::ITextureObject* texture, Uint32 mipLevel,
Uint32 baseArrayLayer, Uint32 layerCount) {
return PendingClearKey {
.texture = texture,
.textureLifetimeId = texture ? texture->GetLifetimeId() : 0,
.mipLevel = mipLevel,
.baseArrayLayer = baseArrayLayer,
.layerCount = layerCount,
};
}
PendingClearKey VkClearManager::MakePendingClearKey(
const MG_State::GLState::FramebufferAttachmentObject& attachment) {
MOBILEGL_ASSERT(attachment.IsTexture() && !attachment.IsRenderbuffer(),
"MakePendingClearKey requires a texture framebuffer attachment");
auto* texture = attachment.GetTexture().get();
MOBILEGL_ASSERT(texture != nullptr, "MakePendingClearKey: texture attachment resolved to null");
const Uint32 mipLevel = static_cast<Uint32>(std::max(attachment.GetTextureLevel(), 0));
const Uint32 baseArrayLayer = ResolveAttachmentBaseArrayLayer(attachment);
const Uint32 layerCount = ResolveAttachmentLayerCount(attachment);
return MakePendingClearKey(texture, mipLevel, baseArrayLayer, layerCount);
}
Bool VkClearManager::Initialize() {
return true;
}
void VkClearManager::Shutdown() {
const std::lock_guard<std::mutex> lock(m_mutex);
m_pendingClears.clear();
m_aliveObjects.clear();
}
TextureIdentity VkClearManager::MakeTextureIdentity(MG_State::GLState::ITextureObject* texture) {
return TextureIdentity {
.texture = texture,
.lifetimeId = texture ? texture->GetLifetimeId() : 0,
};
}
void VkClearManager::MergeClearPayload(ClearAttachmentPayload& dst, const ClearAttachmentPayload& src) {
dst.mask |= src.mask;
if ((src.mask & GL_COLOR_BUFFER_BIT) != 0) {
dst.color = src.color;
}
if ((src.mask & GL_DEPTH_BUFFER_BIT) != 0) {
dst.depth = src.depth;
}
if ((src.mask & GL_STENCIL_BUFFER_BIT) != 0) {
dst.stencil = src.stencil;
}
}
void VkClearManager::ErasePendingClearsForTextureLocked(const TextureIdentity& identity) {
Vector<PendingClearKey> keysToErase;
keysToErase.reserve(m_pendingClears.size());
for (auto it = m_pendingClears.begin(); it != m_pendingClears.end(); ++it) {
if (PendingClearMatchesTextureIdentity(it->first, identity)) {
keysToErase.emplace_back(it->first);
}
}
for (const auto& key : keysToErase) {
m_pendingClears.erase(key);
}
m_aliveObjects.erase(identity);
}
Bool VkClearManager::LockTextureIdentityLocked(const TextureIdentity& identity,
SharedPtr<MG_State::GLState::ITextureObject>& outTexture) {
outTexture.reset();
if (identity.texture == nullptr) {
return false;
}
auto aliveIt = m_aliveObjects.find(identity);
if (aliveIt == m_aliveObjects.end()) {
ErasePendingClearsForTextureLocked(identity);
return false;
}
outTexture = aliveIt->second.lock();
if (!outTexture || outTexture.get() != identity.texture || outTexture->GetLifetimeId() != identity.lifetimeId) {
ErasePendingClearsForTextureLocked(identity);
outTexture.reset();
return false;
}
return true;
}
Bool VkClearManager::LockTextureLocked(const PendingClearKey& key,
SharedPtr<MG_State::GLState::ITextureObject>& outTexture) {
return LockTextureIdentityLocked(TextureIdentity{
.texture = key.texture,
.lifetimeId = key.textureLifetimeId,
}, outTexture);
}
void VkClearManager::QueueClear(GLbitfield mask, const ClearFramebufferPayload& clearPayload,
const MG_State::GLState::FramebufferObject& drawFbo) {
if (mask & GL_COLOR_BUFFER_BIT) {
auto& drawbufs = drawFbo.GetDrawBuffers();
// This should automatically work on default & offscreen FBO
for (auto drawbuf: drawbufs) {
const auto* attachment = GetClearableAttachment(drawFbo, drawbuf);
if (!attachment) {
continue;
}
QueueClear({
.mask = GL_COLOR_BUFFER_BIT,
.color = clearPayload.color
}, *attachment);
MGLOG_D("%s: %s (texture %d) - color = (%.2f, %.2f, %.2f, %.2f)", __func__,
MG_Util::ConvertFramebufferAttachmentTypeToString(drawbuf).c_str(),
attachment->GetTexture()->GetExternalIndex(),
clearPayload.color[0], clearPayload.color[1], clearPayload.color[2], clearPayload.color[3]);
}
}
if (mask & GL_DEPTH_BUFFER_BIT) {
const auto* attachment = GetClearableAttachment(drawFbo, FramebufferAttachmentType::Depth);
if (attachment) {
QueueClear({
.mask = GL_DEPTH_BUFFER_BIT,
.depth = clearPayload.depth,
}, *attachment);
MGLOG_D("%s: Depth (texture %d) - depth = (%.2f)", __func__,
attachment->GetTexture()->GetExternalIndex(), clearPayload.depth);
}
}
if (mask & GL_STENCIL_BUFFER_BIT) {
const auto* attachment = GetClearableAttachment(drawFbo, FramebufferAttachmentType::Stencil);
if (attachment) {
QueueClear({
.mask = GL_STENCIL_BUFFER_BIT,
.stencil = clearPayload.stencil,
}, *attachment);
MGLOG_D("%s: Stencil (texture %d) - stencil = (%u)", __func__,
attachment->GetTexture()->GetExternalIndex(), clearPayload.stencil);
}
}
}
void VkClearManager::QueueClear(const ClearAttachmentPayload& clearPayload,
const SharedPtr<MG_State::GLState::ITextureObject>& texture) {
if (clearPayload.mask == 0 || !texture) {
return;
}
const PendingClearKey key = MakePendingClearKey(texture.get());
const std::lock_guard<std::mutex> lock(m_mutex);
m_aliveObjects[MakeTextureIdentity(texture.get())] = texture;
auto& pending = m_pendingClears[key];
MergeClearPayload(pending, clearPayload);
}
void VkClearManager::QueueClear(const ClearAttachmentPayload& clearPayload,
const MG_State::GLState::FramebufferAttachmentObject& attachment) {
if (clearPayload.mask == 0 || !attachment.IsTexture() || attachment.IsRenderbuffer()) {
return;
}
const auto texture = attachment.GetTexture();
if (!texture) {
return;
}
const PendingClearKey key = MakePendingClearKey(attachment);
const std::lock_guard<std::mutex> lock(m_mutex);
m_aliveObjects[MakeTextureIdentity(texture.get())] = texture;
auto& pending = m_pendingClears[key];
MergeClearPayload(pending, clearPayload);
}
Bool VkClearManager::HasPendingClear(MG_State::GLState::ITextureObject* texture) {
if (texture == nullptr) {
return false;
}
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) {
if (it->first.texture == texture && it->first.textureLifetimeId == lifetimeId) {
SharedPtr<MG_State::GLState::ITextureObject> liveTexture;
return LockTextureLocked(it->first, liveTexture);
}
}
return false;
}
Bool VkClearManager::HasPendingClear(const PendingClearKey& key) {
if (key.texture == nullptr) {
return false;
}
const std::lock_guard<std::mutex> lock(m_mutex);
if (m_pendingClears.find(key) == m_pendingClears.end()) {
return false;
}
SharedPtr<MG_State::GLState::ITextureObject> liveTexture;
return LockTextureLocked(key, liveTexture);
}
Bool VkClearManager::HasPendingClear(const MG_State::GLState::FramebufferAttachmentObject& attachment) {
if (!attachment.IsTexture() || attachment.IsRenderbuffer() || !attachment.GetTexture()) {
return false;
}
return HasPendingClear(MakePendingClearKey(attachment));
}
Bool VkClearManager::GetPendingClear(const PendingClearKey& key, ClearAttachmentPayload& outPayload) {
SharedPtr<MG_State::GLState::ITextureObject> liveTexture;
return GetPendingClear(key, outPayload, liveTexture);
}
Bool VkClearManager::GetPendingClear(const PendingClearKey& key, ClearAttachmentPayload& outPayload,
SharedPtr<MG_State::GLState::ITextureObject>& outTexture) {
if (key.texture == nullptr) {
return false;
}
const std::lock_guard<std::mutex> lock(m_mutex);
if (!LockTextureLocked(key, outTexture)) {
return false;
}
auto it = m_pendingClears.find(key);
if (it == m_pendingClears.end()) {
outTexture.reset();
return false;
}
outPayload = it->second;
MGLOG_D("%s: Got pending clear for texture@%p lifetime=%llu, mip=%u layer=%u count=%u mask=0x%x clear value: color = (%.2f, %.2f, %.2f, %.2f), depth = (%.2f), stencil = (%u)", __func__,
static_cast<void*>(key.texture),
static_cast<unsigned long long>(key.textureLifetimeId),
key.mipLevel, key.baseArrayLayer, key.layerCount,
static_cast<Uint32>(outPayload.mask),
outPayload.color[0], outPayload.color[1], outPayload.color[2], outPayload.color[3],
outPayload.depth,
outPayload.stencil);
return true;
}
Bool VkClearManager::GetPendingClear(const MG_State::GLState::FramebufferAttachmentObject& attachment,
ClearAttachmentPayload& outPayload) {
if (!attachment.IsTexture() || attachment.IsRenderbuffer() || !attachment.GetTexture()) {
MGLOG_D("%s: Failed getting pending clear for non-texture framebuffer attachment", __func__);
return false;
}
return GetPendingClear(MakePendingClearKey(attachment), outPayload);
}
Bool VkClearManager::GetPendingClears(MG_State::GLState::ITextureObject* texture,
Vector<PendingClearEntry>& outEntries) {
outEntries.clear();
if (texture == nullptr) {
return false;
}
const Uint64 lifetimeId = texture->GetLifetimeId();
const std::lock_guard<std::mutex> lock(m_mutex);
SharedPtr<MG_State::GLState::ITextureObject> liveTexture;
if (!LockTextureIdentityLocked(MakeTextureIdentity(texture), liveTexture)) {
return false;
}
for (auto it = m_pendingClears.begin(); it != m_pendingClears.end(); ++it) {
if (it->first.texture == texture && it->first.textureLifetimeId == lifetimeId) {
outEntries.emplace_back(PendingClearEntry{.key = it->first, .payload = it->second});
}
}
return !outEntries.empty();
}
void VkClearManager::PopPendingClear(MG_State::GLState::ITextureObject* texture) {
if (texture == nullptr) {
return;
}
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);
ErasePendingClearsForTextureLocked(identity);
}
void VkClearManager::PopPendingClear(const PendingClearKey& key) {
if (key.texture == nullptr) {
return;
}
{
const std::lock_guard<std::mutex> lock(m_mutex);
auto it = m_pendingClears.find(key);
if (it != m_pendingClears.end()) {
m_pendingClears.erase(it);
}
}
MGLOG_D("%s: Pop pending clear for texture@%p lifetime=%llu mip=%u layer=%u count=%u", __func__,
static_cast<void*>(key.texture), static_cast<unsigned long long>(key.textureLifetimeId),
key.mipLevel, key.baseArrayLayer, key.layerCount);
}
void VkClearManager::PopPendingClear(const MG_State::GLState::FramebufferAttachmentObject& attachment) {
if (!attachment.IsTexture() || attachment.IsRenderbuffer() || !attachment.GetTexture()) {
return;
}
PopPendingClear(MakePendingClearKey(attachment));
}
SizeT VkClearManager::CollectGarbage() {
const std::lock_guard<std::mutex> lock(m_mutex);
m_gcCounter++;
if (m_gcCounter != 0) {
return 0;
}
Vector<TextureIdentity> expiredTextures;
expiredTextures.reserve(m_aliveObjects.size());
for (auto it = m_aliveObjects.begin(); it != m_aliveObjects.end(); ++it) {
if (it->second.expired()) {
expiredTextures.emplace_back(it->first);
}
}
if (expiredTextures.empty()) {
return 0;
}
for (const auto& identity : expiredTextures) {
ErasePendingClearsForTextureLocked(identity);
}
return expiredTextures.size();
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,127 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkClearManager.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 "../VkIncludes.h"
#include "../VulkanRendererConfig.h"
#include "MG_State/GLState/FramebufferState/FramebufferObject.h"
#include "MG_Util/Math/VectorTypes.h"
#include <Includes.h>
#include <unordered_map>
namespace MobileGL::MG_Backend::DirectVulkan {
struct ClearFramebufferPayload {
FloatVec4 color;
Float depth{};
Uint32 stencil{};
};
struct ClearAttachmentPayload {
GLbitfield mask = 0;
FloatVec4 color = FloatVec4(0.0f, 0.0f, 0.0f, 0.0f);
Float depth = 1.0f;
Uint32 stencil = 0;
};
struct PendingClearKey {
MG_State::GLState::ITextureObject* texture = nullptr;
Uint64 textureLifetimeId = 0;
Uint32 mipLevel = 0;
Uint32 baseArrayLayer = 0;
Uint32 layerCount = 1;
Bool operator==(const PendingClearKey& other) const {
return texture == other.texture && textureLifetimeId == other.textureLifetimeId &&
mipLevel == other.mipLevel &&
baseArrayLayer == other.baseArrayLayer && layerCount == other.layerCount;
}
};
struct TextureIdentity {
MG_State::GLState::ITextureObject* texture = nullptr;
Uint64 lifetimeId = 0;
Bool operator==(const TextureIdentity& other) const {
return texture == other.texture && lifetimeId == other.lifetimeId;
}
};
struct PendingClearEntry {
PendingClearKey key{};
ClearAttachmentPayload payload{};
};
struct PendingClearKeyHash {
SizeT operator()(const PendingClearKey& key) const {
const SizeT textureHash = std::hash<MG_State::GLState::ITextureObject*>{}(key.texture);
const SizeT textureLifetimeHash = std::hash<Uint64>{}(key.textureLifetimeId);
const SizeT mipHash = std::hash<Uint32>{}(key.mipLevel);
const SizeT layerHash = std::hash<Uint32>{}(key.baseArrayLayer);
const SizeT layerCountHash = std::hash<Uint32>{}(key.layerCount);
SizeT hash = textureHash;
hash ^= textureLifetimeHash + 0x9e3779b9u + (hash << 6) + (hash >> 2);
hash ^= mipHash + 0x9e3779b9u + (hash << 6) + (hash >> 2);
hash ^= layerHash + 0x9e3779b9u + (hash << 6) + (hash >> 2);
hash ^= layerCountHash + 0x9e3779b9u + (hash << 6) + (hash >> 2);
return hash;
}
};
struct TextureIdentityHash {
SizeT operator()(const TextureIdentity& key) const {
SizeT hash = std::hash<MG_State::GLState::ITextureObject*>{}(key.texture);
hash ^= std::hash<Uint64>{}(key.lifetimeId) + 0x9e3779b9u + (hash << 6) + (hash >> 2);
return hash;
}
};
class VkClearManager {
public:
static PendingClearKey MakePendingClearKey(const MG_State::GLState::FramebufferAttachmentObject& attachment);
static PendingClearKey MakePendingClearKey(MG_State::GLState::ITextureObject* texture, Uint32 mipLevel = 0,
Uint32 baseArrayLayer = 0, Uint32 layerCount = 1);
Bool Initialize();
void Shutdown();
void QueueClear(GLbitfield mask, const ClearFramebufferPayload& clearPayload, const MG_State::GLState::FramebufferObject& drawFbo);
void QueueClear(
const ClearAttachmentPayload& clearPayload,
const SharedPtr<MG_State::GLState::ITextureObject>& texture);
void QueueClear(const ClearAttachmentPayload& clearPayload,
const MG_State::GLState::FramebufferAttachmentObject& attachment);
Bool HasPendingClear(MG_State::GLState::ITextureObject* texture);
Bool HasPendingClear(const PendingClearKey& key);
Bool HasPendingClear(const MG_State::GLState::FramebufferAttachmentObject& attachment);
Bool GetPendingClear(const PendingClearKey& key, ClearAttachmentPayload& outPayload);
Bool GetPendingClear(const PendingClearKey& key, ClearAttachmentPayload& outPayload,
SharedPtr<MG_State::GLState::ITextureObject>& outTexture);
Bool GetPendingClear(const MG_State::GLState::FramebufferAttachmentObject& attachment,
ClearAttachmentPayload& outPayload);
Bool GetPendingClears(MG_State::GLState::ITextureObject* texture, Vector<PendingClearEntry>& outEntries);
void PopPendingClear(MG_State::GLState::ITextureObject* texture);
void PopPendingClear(const PendingClearKey& key);
void PopPendingClear(const MG_State::GLState::FramebufferAttachmentObject& attachment);
SizeT CollectGarbage();
private:
static TextureIdentity MakeTextureIdentity(MG_State::GLState::ITextureObject* texture);
static void MergeClearPayload(ClearAttachmentPayload& dst, const ClearAttachmentPayload& src);
void ErasePendingClearsForTextureLocked(const TextureIdentity& identity);
Bool LockTextureIdentityLocked(const TextureIdentity& identity,
SharedPtr<MG_State::GLState::ITextureObject>& outTexture);
Bool LockTextureLocked(const PendingClearKey& key,
SharedPtr<MG_State::GLState::ITextureObject>& outTexture);
Uint8 m_gcCounter = 0;
mutable std::mutex m_mutex;
std::unordered_map<PendingClearKey, ClearAttachmentPayload, PendingClearKeyHash> m_pendingClears;
std::unordered_map<TextureIdentity, WeakPtr<MG_State::GLState::ITextureObject>, TextureIdentityHash> m_aliveObjects;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
File diff suppressed because it is too large Load Diff
@@ -1,259 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkRenderPassManager.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 "SwapchainObject.h"
#include "VkClearManager.h"
#include "VkTextureManager.h"
#include "../VkIncludes.h"
#include "../VulkanRendererConfig.h"
#include "MG_State/GLState/FramebufferState/FramebufferObject.h"
#include <Includes.h>
#include <vk_mem_alloc.h>
namespace MobileGL::MG_Backend::DirectVulkan {
enum class TrackedAttachmentTarget : Uint8 {
Texture,
Renderbuffer,
SwapchainColor,
SwapchainDepthStencil
};
struct PendingClearAttachmentInfo {
// Index into the render pass attachment descriptions (VkRenderPassBeginInfo::pClearValues space).
Uint32 attachmentIndex = 0;
// Index into the subpass pColorAttachments (VkClearAttachment::colorAttachment space) — the GL
// draw-buffer slot. Differs from attachmentIndex when earlier slots are GL_NONE/incomplete.
// Only meaningful for color clears.
Uint32 colorAttachmentSlot = 0;
PendingClearKey key{};
MG_State::GLState::RenderbufferObject* renderbuffer = nullptr;
Bool hasInlinePayload = false;
ClearAttachmentPayload inlinePayload{};
};
struct TrackedAttachmentLayoutInfo {
TrackedAttachmentTarget target = TrackedAttachmentTarget::Texture;
WeakPtr<MG_State::GLState::ITextureObject> texture;
WeakPtr<MG_State::GLState::RenderbufferObject> renderbuffer;
Uint32 textureMipLevel = 0;
Uint32 swapchainImageIndex = 0;
VkImageLayout finalLayout = VK_IMAGE_LAYOUT_UNDEFINED;
};
struct DepthStencilAttachmentLoadInfo {
VkAttachmentLoadOp depthLoadOp = VK_ATTACHMENT_LOAD_OP_LOAD;
VkAttachmentLoadOp stencilLoadOp = VK_ATTACHMENT_LOAD_OP_LOAD;
VkImageLayout initialLayout = VK_IMAGE_LAYOUT_UNDEFINED;
};
DepthStencilAttachmentLoadInfo ResolveDepthStencilAttachmentLoadInfo(
VkImageLayout trackedLayout, Bool clearDepth, Bool clearStencil);
IntVec2 ResolveRenderPassFramebufferExtent(Bool isDefaultFbo, const TextureSize& attachmentExtent,
VkExtent2D swapchainExtent);
struct RenderPassEntry {
static inline VkDevice s_device;
static inline Vector<VkTextureManager::TextureResource*> s_textureResourcesScratch;
Uint64 hash = 0;
VkRenderPass renderPass = VK_NULL_HANDLE;
VkFramebuffer framebuffer = VK_NULL_HANDLE;
Uint64 compatibilityHash = 0;
Vector<PendingClearAttachmentInfo> pendingClearAttachments;
Vector<TrackedAttachmentLayoutInfo> trackedAttachmentLayouts;
Uint32 attachmentCount = 0;
Uint32 colorAttachmentCount = 0;
Bool hasDepthStencilAttachment = false;
VkSampleCountFlagBits sampleCount = VK_SAMPLE_COUNT_1_BIT;
IntVec2 extent = {0, 0};
// VkFramebufferCreateInfo::layers of the entry's framebuffer (>1 for layered GL attachments).
Uint32 layers = 1;
// Frame counter value of the last GetOrCreateRenderPass hit; drives cache eviction.
Uint64 lastUsedFrame = 0;
RenderPassEntry() = default;
RenderPassEntry(const RenderPassEntry&) = delete;
RenderPassEntry(RenderPassEntry&& that) noexcept {
std::swap(hash, that.hash);
std::swap(renderPass, that.renderPass);
std::swap(framebuffer, that.framebuffer);
std::swap(compatibilityHash, that.compatibilityHash);
std::swap(pendingClearAttachments, that.pendingClearAttachments);
std::swap(trackedAttachmentLayouts, that.trackedAttachmentLayouts);
std::swap(attachmentCount, that.attachmentCount);
std::swap(colorAttachmentCount, that.colorAttachmentCount);
std::swap(hasDepthStencilAttachment, that.hasDepthStencilAttachment);
std::swap(sampleCount, that.sampleCount);
std::swap(extent, that.extent);
std::swap(layers, that.layers);
std::swap(lastUsedFrame, that.lastUsedFrame);
}
RenderPassEntry(
Uint64 hash,
VkRenderPass renderpass,
VkFramebuffer framebuffer,
Uint64 compatibilityHash,
const Vector<PendingClearAttachmentInfo>& pendingClearAttachments,
const Vector<TrackedAttachmentLayoutInfo>& trackedAttachmentLayouts,
Uint32 attachmentCount,
Uint32 colorAttachmentCount,
Bool hasDepthStencilAttachment,
VkSampleCountFlagBits sampleCount,
IntVec2 extent, Uint32 layers):
hash(hash),
renderPass(renderpass),
framebuffer(framebuffer),
compatibilityHash(compatibilityHash),
pendingClearAttachments(Move(pendingClearAttachments)),
trackedAttachmentLayouts(Move(trackedAttachmentLayouts)),
attachmentCount(attachmentCount),
colorAttachmentCount(colorAttachmentCount),
hasDepthStencilAttachment(hasDepthStencilAttachment),
sampleCount(sampleCount),
extent(extent),
layers(layers)
{}
~RenderPassEntry() {
if (renderPass != VK_NULL_HANDLE) {
vkDestroyRenderPass(s_device, renderPass, nullptr);
}
if (framebuffer != VK_NULL_HANDLE) {
vkDestroyFramebuffer(s_device, framebuffer, nullptr);
}
}
Bool CompatibleWith(const RenderPassEntry& that) const {
return this->compatibilityHash == that.compatibilityHash;
}
Bool CompatibleWith(Uint64 compatibilityHash) const {
return this->compatibilityHash == compatibilityHash;
}
};
struct ActiveRenderPassInfo {
Uint64 hash = 0;
Uint64 compatibilityHash = 0;
Vector<TrackedAttachmentLayoutInfo> trackedAttachmentLayouts;
IntVec2 extent = {0, 0};
Bool CompatibleWith(const RenderPassEntry& that) const {
return compatibilityHash == that.compatibilityHash;
}
Bool CompatibleWith(Uint64 thatCompatibilityHash) const {
return compatibilityHash == thatCompatibilityHash;
}
};
class VkRenderPassManager {
public:
using HashType = Uint64;
VkRenderPassManager(VkDevice device,
VkPhysicalDevice physicalDevice, VmaAllocator allocator, const VulkanRendererConfig& config,
VkClearManager& clearManager, VkTextureManager& textureManager, SwapchainObject& swapchainObject);
~VkRenderPassManager();
Bool Initialize();
void Shutdown();
HashType ComputeHash(
const MG_State::GLState::FramebufferObject& fbo,
Uint32 swapchainImageIndex,
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,
const MG_State::GLState::FramebufferAttachmentObject& attachment);
void PopPendingRenderbufferClear(MG_State::GLState::RenderbufferObject* renderbuffer);
// Frame boundary hook: ages the render-pass cache and evicts long-unused
// entries (their command buffers retired many frames ago).
void OnPresent();
static Bool BeginRenderPass(VkCommandBuffer commandBuffer, RenderPassEntry& renderPassEntry);
static Bool EndRenderPass(VkCommandBuffer commandBuffer);
static ActiveRenderPassInfo* GetActiveRenderPass();
private:
VkDevice m_device = VK_NULL_HANDLE;
VkPhysicalDevice m_physicalDevice = VK_NULL_HANDLE;
VmaAllocator m_allocator = nullptr;
const VulkanRendererConfig& m_config;
VkClearManager& m_clearManager;
VkTextureManager& m_textureManager;
SwapchainObject& m_swapchainObject;
UnorderedMap<Uint64, RenderPassEntry> m_renderPasses;
// Monotonic frame counter (bumped in OnPresent) for render-pass cache aging.
Uint64 m_frameCounter = 0;
// 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;
// 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 /
// version change, swapchain rotation, any attachment image recreation (the two epochs),
// or a pending clear. Portable to Vulkan 1.1 (no dynamic_rendering / imageless FB needed).
Bool m_rpFastValid = false;
const MG_State::GLState::FramebufferObject* m_rpFastFbo = nullptr;
Uint16 m_rpFastFboVersion = 0;
Uint32 m_rpFastSwapchainIndex = 0;
Uint64 m_rpFastTexEpoch = 0;
Uint64 m_rpFastRbEpoch = 0;
Uint64 m_rpFastRenderPassHash = 0;
public:
struct RenderbufferResource {
WeakPtr<MG_State::GLState::RenderbufferObject> renderbuffer;
VkImage image = VK_NULL_HANDLE;
VmaAllocation allocation = nullptr;
VkImageView view = VK_NULL_HANDLE;
VkImageLayout layout = VK_IMAGE_LAYOUT_UNDEFINED;
VkFormat format = VK_FORMAT_UNDEFINED;
VkImageAspectFlags aspect = VK_IMAGE_ASPECT_NONE;
VkExtent2D extent = {0, 0};
VkSampleCountFlagBits sampleCount = VK_SAMPLE_COUNT_1_BIT;
TextureInternalFormat internalFormat = TextureInternalFormat::Unknown;
Int samples = 0;
void Destroy(VkDevice device, VmaAllocator allocator);
};
// Public so the renderer's blit/copy/readback bindings can source renderbuffer
// attachments the same way texture attachments go through the texture manager.
RenderbufferResource* GetOrCreateRenderbufferResource(
const SharedPtr<MG_State::GLState::RenderbufferObject>& renderbuffer);
Bool GetPendingRenderbufferClear(MG_State::GLState::RenderbufferObject* renderbuffer,
ClearAttachmentPayload& outPayload) const;
private:
struct PendingRenderbufferClear {
WeakPtr<MG_State::GLState::RenderbufferObject> renderbuffer;
ClearAttachmentPayload payload{};
};
UnorderedMap<MG_State::GLState::RenderbufferObject*, RenderbufferResource> m_renderbufferResources;
UnorderedMap<MG_State::GLState::RenderbufferObject*, PendingRenderbufferClear> m_pendingRenderbufferClears;
Bool HasPendingRenderbufferClear(
const MG_State::GLState::FramebufferAttachmentObject& attachment) const;
void CollectRenderbufferGarbage();
static inline XXH64_state_t* m_hashState = XXH64_createState();
static inline ActiveRenderPassInfo s_activeRenderPass{};
static inline Bool s_hasActiveRenderPass = false;
static inline VkClearManager* s_clearManager = nullptr;
static inline VkTextureManager* s_textureManager = nullptr;
static inline SwapchainObject* s_swapchainObject = nullptr;
static inline VkRenderPassManager* s_renderPassManager = nullptr;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,275 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkSamplerManager.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 "VkSamplerManager.h"
#include "MG_State/GLState/Core.h"
#include <algorithm>
#include <cmath>
namespace MobileGL::MG_Backend::DirectVulkan {
namespace {
Bool UsesBorderColor(const MG_State::GLState::SamplerObject& sampler) {
return sampler.GetWrapS() == SamplerWrapMode::ClampToBorder ||
sampler.GetWrapT() == SamplerWrapMode::ClampToBorder ||
sampler.GetWrapR() == SamplerWrapMode::ClampToBorder;
}
Bool IsDepthTextureFormat(TextureInternalFormat format) {
switch (format) {
case TextureInternalFormat::DepthComponent:
case TextureInternalFormat::DepthComponent16:
case TextureInternalFormat::DepthComponent24:
case TextureInternalFormat::DepthComponent32:
case TextureInternalFormat::DepthComponent32F:
case TextureInternalFormat::Depth24Stencil8:
case TextureInternalFormat::Depth32FStencil8:
case TextureInternalFormat::DepthStencil:
return true;
default:
return false;
}
}
Bool NearlyEqual(Float lhs, Float rhs) {
return std::fabs(lhs - rhs) <= 1e-6f;
}
Float ResolveEffectiveMaxLod(const MG_State::GLState::SamplerObject& sampler) {
if (sampler.GetMipmapMode() == SamplerMipmapMode::None) {
return 0.0f;
}
return sampler.GetMaxLod();
}
Float ResolveEffectiveMinLod(const MG_State::GLState::SamplerObject& sampler, Float effectiveMaxLod) {
return std::min(sampler.GetMinLod(), effectiveMaxLod);
}
} // namespace
Bool VkSamplerManager::Initialize(const InitInfo& initInfo) {
Shutdown();
m_device = initInfo.device;
m_config = initInfo.config;
m_samplerAnisotropySupported = initInfo.samplerAnisotropySupported;
m_maxSamplerAnisotropy = std::max(initInfo.maxSamplerAnisotropy, 1.0f);
MOBILEGL_ASSERT(m_device != VK_NULL_HANDLE && m_config != nullptr,
"VkSamplerManager::Initialize failed: invalid initialization info");
return true;
}
Float VkSamplerManager::ResolveEffectiveMaxAnisotropy(const MG_State::GLState::SamplerObject& sampler,
Bool forceNearestFiltering) const {
if (!m_samplerAnisotropySupported) return 1.0f;
if (forceNearestFiltering) return 1.0f;
// VUID-VkSamplerCreateInfo-anisotropyEnable-01071/01072: anisotropy requires both filters to
// be LINEAR and the value to sit within [1, limits.maxSamplerAnisotropy].
if (sampler.GetMinFilter() != SamplerFilterMode::Linear ||
sampler.GetMagFilter() != SamplerFilterMode::Linear) {
return 1.0f;
}
return std::clamp(sampler.GetMaxAnisotropy(), 1.0f, m_maxSamplerAnisotropy);
}
void VkSamplerManager::Shutdown() {
for (auto& [_, sampler] : m_samplers) {
if (m_device != VK_NULL_HANDLE && sampler.handle != VK_NULL_HANDLE) {
vkDestroySampler(m_device, sampler.handle, nullptr);
}
sampler.handle = VK_NULL_HANDLE;
}
m_samplers.clear();
m_device = VK_NULL_HANDLE;
m_config = nullptr;
}
Uint64 VkSamplerManager::BuildSamplerKey(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture,
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)));
const auto minFilter = sampler.GetMinFilter();
XXHASH_VERIFY(XXH64_update(m_hashState, &minFilter, sizeof(minFilter)));
const auto magFilter = sampler.GetMagFilter();
XXHASH_VERIFY(XXH64_update(m_hashState, &magFilter, sizeof(magFilter)));
const auto mipmapMode = sampler.GetMipmapMode();
XXHASH_VERIFY(XXH64_update(m_hashState, &mipmapMode, sizeof(mipmapMode)));
const auto wrapS = sampler.GetWrapS();
XXHASH_VERIFY(XXH64_update(m_hashState, &wrapS, sizeof(wrapS)));
const auto wrapT = sampler.GetWrapT();
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 = 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)));
const auto lodBias = sampler.GetLodBias();
XXHASH_VERIFY(XXH64_update(m_hashState, &lodBias, sizeof(lodBias)));
// The RESOLVED value, not the GL request: samplers that only differ in an anisotropy Vulkan
// will not apply (NEAREST filtering, or requests past the device limit) must still share one
// VkSampler, while two samplers that really do differ must not collide onto the first one's.
const auto maxAnisotropy = ResolveEffectiveMaxAnisotropy(sampler, forceNearestFiltering);
XXHASH_VERIFY(XXH64_update(m_hashState, &maxAnisotropy, sizeof(maxAnisotropy)));
const auto compareMode = sampler.GetCompareMode();
XXHASH_VERIFY(XXH64_update(m_hashState, &compareMode, sizeof(compareMode)));
const auto compareFunc = ResolveCompareFunc(sampler, texture);
XXHASH_VERIFY(XXH64_update(m_hashState, &compareFunc, sizeof(compareFunc)));
const auto borderColor = ResolveVkBorderColor(sampler, texture);
XXHASH_VERIFY(XXH64_update(m_hashState, &borderColor, sizeof(borderColor)));
return XXH64_digest(m_hashState);
}
VkSampler VkSamplerManager::GetOrCreateSampler(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture,
Bool forceNearestFiltering) {
const Uint64 key = BuildSamplerKey(sampler, texture, forceNearestFiltering);
auto it = m_samplers.find(key);
if (it != m_samplers.end()) {
return it->second.handle;
}
VkSamplerCreateInfo samplerInfo{};
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 ? VK_SAMPLER_MIPMAP_MODE_NEAREST
: ToVkMipmapMode(sampler.GetMipmapMode());
samplerInfo.addressModeU = ToVkAddressMode(sampler.GetWrapS());
samplerInfo.addressModeV = ToVkAddressMode(sampler.GetWrapT());
samplerInfo.addressModeW = ToVkAddressMode(sampler.GetWrapR());
samplerInfo.mipLodBias = sampler.GetLodBias();
// Must use the same resolver as BuildSamplerKey - a divergence would either collide two
// different samplers or silently create duplicates.
const Float maxAnisotropy = ResolveEffectiveMaxAnisotropy(sampler, forceNearestFiltering);
samplerInfo.anisotropyEnable = maxAnisotropy > 1.0f ? VK_TRUE : VK_FALSE;
samplerInfo.maxAnisotropy = maxAnisotropy;
samplerInfo.compareEnable = sampler.GetCompareMode() == SamplerCompareMode::CompareToTexture ? VK_TRUE : VK_FALSE;
samplerInfo.compareOp = ToVkCompareOp(ResolveCompareFunc(sampler, texture));
samplerInfo.maxLod = ResolveEffectiveMaxLod(sampler);
samplerInfo.minLod = ResolveEffectiveMinLod(sampler, samplerInfo.maxLod);
samplerInfo.borderColor = ResolveVkBorderColor(sampler, texture);
samplerInfo.unnormalizedCoordinates = VK_FALSE;
VkSampler vkSampler = VK_NULL_HANDLE;
VK_VERIFY(vkCreateSampler(m_device, &samplerInfo, nullptr, &vkSampler), "vkCreateSampler(texture)");
SamplerCacheEntry entry{};
entry.handle = vkSampler;
entry.externalIndex = sampler.GetExternalIndex();
entry.version = sampler.GetVersion();
m_samplers[key] = entry;
return vkSampler;
}
VkFilter VkSamplerManager::ToVkFilter(SamplerFilterMode mode) {
return mode == SamplerFilterMode::Nearest ? VK_FILTER_NEAREST : VK_FILTER_LINEAR;
}
VkSamplerMipmapMode VkSamplerManager::ToVkMipmapMode(SamplerMipmapMode mode) {
switch (mode) {
case SamplerMipmapMode::Nearest:
return VK_SAMPLER_MIPMAP_MODE_NEAREST;
case SamplerMipmapMode::Linear:
return VK_SAMPLER_MIPMAP_MODE_LINEAR;
case SamplerMipmapMode::None:
default:
return VK_SAMPLER_MIPMAP_MODE_NEAREST;
}
}
VkSamplerAddressMode VkSamplerManager::ToVkAddressMode(SamplerWrapMode mode) {
switch (mode) {
case SamplerWrapMode::ClampToEdge:
return VK_SAMPLER_ADDRESS_MODE_CLAMP_TO_EDGE;
case SamplerWrapMode::MirroredRepeat:
return VK_SAMPLER_ADDRESS_MODE_MIRRORED_REPEAT;
case SamplerWrapMode::Repeat:
return VK_SAMPLER_ADDRESS_MODE_REPEAT;
case SamplerWrapMode::ClampToBorder:
return VK_SAMPLER_ADDRESS_MODE_CLAMP_TO_BORDER;
case SamplerWrapMode::MirrorClampToEdge:
return VK_SAMPLER_ADDRESS_MODE_MIRROR_CLAMP_TO_EDGE;
default:
return VK_SAMPLER_ADDRESS_MODE_REPEAT;
}
}
VkCompareOp VkSamplerManager::ToVkCompareOp(SamplerCompareFunc func) {
switch (func) {
case SamplerCompareFunc::Never:
return VK_COMPARE_OP_NEVER;
case SamplerCompareFunc::Less:
return VK_COMPARE_OP_LESS;
case SamplerCompareFunc::Equal:
return VK_COMPARE_OP_EQUAL;
case SamplerCompareFunc::LessEqual:
return VK_COMPARE_OP_LESS_OR_EQUAL;
case SamplerCompareFunc::Greater:
return VK_COMPARE_OP_GREATER;
case SamplerCompareFunc::NotEqual:
return VK_COMPARE_OP_NOT_EQUAL;
case SamplerCompareFunc::GreaterEqual:
return VK_COMPARE_OP_GREATER_OR_EQUAL;
case SamplerCompareFunc::Always:
default:
return VK_COMPARE_OP_ALWAYS;
}
}
SamplerCompareFunc VkSamplerManager::ResolveCompareFunc(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture) {
const auto compareFunc = sampler.GetSamplerCompareFunc();
if (sampler.GetCompareMode() == SamplerCompareMode::CompareToTexture &&
IsDepthTextureFormat(texture.GetFormat()) && compareFunc == SamplerCompareFunc::Always) {
return SamplerCompareFunc::LessEqual;
}
return compareFunc;
}
VkBorderColor VkSamplerManager::ResolveVkBorderColor(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture) {
if (!UsesBorderColor(sampler)) {
return VK_BORDER_COLOR_FLOAT_TRANSPARENT_BLACK;
}
const auto& borderColor = texture.GetBorderColor();
const Bool isDepthTexture = IsDepthTextureFormat(texture.GetFormat());
if (isDepthTexture) {
if (NearlyEqual(borderColor.x(), 1.0f)) {
return VK_BORDER_COLOR_FLOAT_OPAQUE_WHITE;
}
if (NearlyEqual(borderColor.x(), 0.0f)) {
return VK_BORDER_COLOR_FLOAT_OPAQUE_BLACK;
}
}
const Bool rgbZero = NearlyEqual(borderColor.x(), 0.0f) && NearlyEqual(borderColor.y(), 0.0f) &&
NearlyEqual(borderColor.z(), 0.0f);
if (rgbZero && NearlyEqual(borderColor.w(), 0.0f)) {
return VK_BORDER_COLOR_FLOAT_TRANSPARENT_BLACK;
}
if (rgbZero && NearlyEqual(borderColor.w(), 1.0f)) {
return VK_BORDER_COLOR_FLOAT_OPAQUE_BLACK;
}
if (NearlyEqual(borderColor.x(), 1.0f) && NearlyEqual(borderColor.y(), 1.0f) &&
NearlyEqual(borderColor.z(), 1.0f) && NearlyEqual(borderColor.w(), 1.0f)) {
return VK_BORDER_COLOR_FLOAT_OPAQUE_WHITE;
}
return VK_BORDER_COLOR_FLOAT_TRANSPARENT_BLACK;
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,72 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkSamplerManager.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 "../VkIncludes.h"
#include "../VulkanRendererConfig.h"
#include <Includes.h>
#include <MG_State/GLState/SamplerState/SamplerObject.h>
namespace MobileGL::MG_State::GLState {
class SamplerObject;
class ITextureObject;
}
namespace MobileGL::MG_Backend::DirectVulkan {
class VkSamplerManager {
public:
struct InitInfo {
VkDevice device = VK_NULL_HANDLE;
const VulkanRendererConfig* config = nullptr;
// The samplerAnisotropy device feature was requested and granted at vkCreateDevice.
Bool samplerAnisotropySupported = false;
// VkPhysicalDeviceLimits::maxSamplerAnisotropy.
Float maxSamplerAnisotropy = 1.0f;
};
Bool Initialize(const InitInfo& initInfo);
void Shutdown();
VkSampler GetOrCreateSampler(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture,
Bool forceNearestFiltering = false);
private:
struct SamplerCacheEntry {
VkSampler handle = VK_NULL_HANDLE;
Uint externalIndex = 0;
Uint16 version = 0;
};
Uint64 BuildSamplerKey(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture,
Bool forceNearestFiltering) const;
static VkFilter ToVkFilter(SamplerFilterMode mode);
static VkSamplerMipmapMode ToVkMipmapMode(SamplerMipmapMode mode);
static VkSamplerAddressMode ToVkAddressMode(SamplerWrapMode mode);
static VkCompareOp ToVkCompareOp(SamplerCompareFunc func);
static SamplerCompareFunc ResolveCompareFunc(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture);
static VkBorderColor ResolveVkBorderColor(const MG_State::GLState::SamplerObject& sampler,
const MG_State::GLState::ITextureObject& texture);
// The anisotropy Vulkan will actually apply: 1.0 (i.e. disabled) unless the feature is on and
// the sampler filters linearly both ways, otherwise the GL request clamped to the device limit.
// GL happily carries GL_TEXTURE_MAX_ANISOTROPY on a NEAREST sampler (Blaze3D's blocks do exactly
// that) while Vulkan forbids anisotropyEnable there, so the GL value must never be forwarded raw.
Float ResolveEffectiveMaxAnisotropy(const MG_State::GLState::SamplerObject& sampler,
Bool forceNearestFiltering) const;
VkDevice m_device = VK_NULL_HANDLE;
const VulkanRendererConfig* m_config = nullptr;
Bool m_samplerAnisotropySupported = false;
Float m_maxSamplerAnisotropy = 1.0f;
UnorderedMap<Uint64, SamplerCacheEntry> m_samplers;
static inline XXH64_state_t* m_hashState = XXH64_createState();
};
} // namespace MobileGL::MG_Backend::DirectVulkan
File diff suppressed because it is too large Load Diff
@@ -1,396 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkTextureManager.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 "../VkIncludes.h"
#include <Includes.h>
#include <MG_State/GLState/TextureState/TextureObject.h>
#include <vk_mem_alloc.h>
#include <unordered_map>
#include <unordered_set>
namespace MobileGL::MG_State::GLState {
class ITextureObject;
}
namespace MobileGL::MG_Backend::DirectVulkan {
enum class SamplerNumericDomain : Uint8;
class VkTextureManager {
public:
// Monotonic epoch bumped whenever a texture VkImage is (re)created. The render-pass
// 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; }
struct TextureIdentity {
MG_State::GLState::ITextureObject* texture = nullptr;
Uint64 lifetimeId = 0;
Bool operator==(const TextureIdentity& other) const {
return texture == other.texture && lifetimeId == other.lifetimeId;
}
};
struct TextureIdentityHash {
SizeT operator()(const TextureIdentity& key) const {
SizeT hash = std::hash<MG_State::GLState::ITextureObject*>{}(key.texture);
hash ^= std::hash<Uint64>{}(key.lifetimeId) + 0x9e3779b9u + (hash << 6) + (hash >> 2);
return hash;
}
};
struct InitInfo {
VkDevice device = VK_NULL_HANDLE;
VkPhysicalDevice physicalDevice = VK_NULL_HANDLE;
VmaAllocator allocator = nullptr;
VkCommandPool commandPool = VK_NULL_HANDLE;
VkQueue graphicsQueue = VK_NULL_HANDLE;
Uint32 frameCount = 0;
};
struct TextureResource {
struct AttachmentViewKey {
Uint32 mipLevel = 0;
Uint32 baseArrayLayer = 0;
Uint32 layerCount = 1;
VkImageViewType viewType = VK_IMAGE_VIEW_TYPE_2D;
Bool operator==(const AttachmentViewKey& other) const {
return mipLevel == other.mipLevel &&
baseArrayLayer == other.baseArrayLayer &&
layerCount == other.layerCount &&
viewType == other.viewType;
}
};
struct AttachmentViewKeyHash {
SizeT operator()(const AttachmentViewKey& key) const {
SizeT hash = std::hash<Uint32>{}(key.mipLevel);
hash ^= std::hash<Uint32>{}(key.baseArrayLayer) + 0x9e3779b9u + (hash << 6) + (hash >> 2);
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);
return hash;
}
};
struct StorageImageViewKey {
Uint32 mipLevel = 0;
Uint32 baseArrayLayer = 0;
Uint32 layerCount = 1;
VkImageViewType viewType = VK_IMAGE_VIEW_TYPE_2D;
VkFormat format = VK_FORMAT_UNDEFINED;
Bool operator==(const StorageImageViewKey& other) const {
return mipLevel == other.mipLevel &&
baseArrayLayer == other.baseArrayLayer &&
layerCount == other.layerCount &&
viewType == other.viewType &&
format == other.format;
}
};
struct SampledImageViewKey {
Uint32 baseMipLevel = 0;
Uint32 levelCount = 1;
VkImageViewType viewType = VK_IMAGE_VIEW_TYPE_2D;
VkFormat format = VK_FORMAT_UNDEFINED;
Bool operator==(const SampledImageViewKey& other) const {
return baseMipLevel == other.baseMipLevel &&
levelCount == other.levelCount &&
viewType == other.viewType &&
format == other.format;
}
};
struct SampledImageViewKeyHash {
SizeT operator()(const SampledImageViewKey& key) const {
SizeT hash = std::hash<Uint32>{}(key.baseMipLevel);
hash ^= std::hash<Uint32>{}(key.levelCount) + 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.format)) +
0x9e3779b9u + (hash << 6) + (hash >> 2);
return hash;
}
};
struct StorageImageViewKeyHash {
SizeT operator()(const StorageImageViewKey& key) const {
SizeT hash = std::hash<Uint32>{}(key.mipLevel);
hash ^= std::hash<Uint32>{}(key.baseArrayLayer) + 0x9e3779b9u + (hash << 6) + (hash >> 2);
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.format)) +
0x9e3779b9u + (hash << 6) + (hash >> 2);
return hash;
}
};
VkImage image = VK_NULL_HANDLE;
VmaAllocation allocation = nullptr;
VkImageView fullView = VK_NULL_HANDLE;
VkImageView sampledView = VK_NULL_HANDLE;
Vector<VkImageView> perMipViews;
Vector<VkImageView> perMipSampledViews;
UnorderedMap<AttachmentViewKey, VkImageView, AttachmentViewKeyHash> attachmentViews;
UnorderedMap<SampledImageViewKey, VkImageView, SampledImageViewKeyHash> alternateSampledViews;
UnorderedMap<StorageImageViewKey, VkImageView, StorageImageViewKeyHash> storageImageViews;
VkImageLayout layout = VK_IMAGE_LAYOUT_UNDEFINED;
VkExtent2D extent = {0, 0};
Uint32 depth = 1;
Uint32 arrayLayers = 1;
Uint32 mipLevels = 1;
Uint32 sampledBaseMipLevel = 0;
Uint32 sampledLevelCount = 1;
VkFormat format = VK_FORMAT_UNDEFINED;
VkImageAspectFlags aspect = VK_IMAGE_ASPECT_NONE;
VkImageViewType viewType = VK_IMAGE_VIEW_TYPE_2D;
VkSampleCountFlagBits sampleCount = VK_SAMPLE_COUNT_1_BIT;
VkImageCreateFlags imageCreateFlags = 0;
Uint16 syncedTextureParamsVersion = 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;
// Snapshot of the defined mip-level count at the last sync. Folded into the early-out key
// as defense-in-depth: any path that grows the level set (which resizes the sampled view)
// busts the skip even if it failed to bump the content version.
Uint32 syncedMipLevelCount = 0;
TextureResource() = default;
TextureResource(const TextureResource&) = delete;
TextureResource(TextureResource&& that) noexcept {
std::swap(this->image, that.image);
std::swap(this->allocation, that.allocation);
std::swap(this->fullView, that.fullView);
std::swap(this->sampledView, that.sampledView);
std::swap(this->perMipViews, that.perMipViews);
std::swap(this->perMipSampledViews, that.perMipSampledViews);
std::swap(this->attachmentViews, that.attachmentViews);
std::swap(this->alternateSampledViews, that.alternateSampledViews);
std::swap(this->storageImageViews, that.storageImageViews);
std::swap(this->layout, that.layout);
std::swap(this->extent, that.extent);
std::swap(this->depth, that.depth);
std::swap(this->arrayLayers, that.arrayLayers);
std::swap(this->mipLevels, that.mipLevels);
std::swap(this->sampledBaseMipLevel, that.sampledBaseMipLevel);
std::swap(this->sampledLevelCount, that.sampledLevelCount);
std::swap(this->format, that.format);
std::swap(this->aspect, that.aspect);
std::swap(this->viewType, that.viewType);
std::swap(this->sampleCount, that.sampleCount);
std::swap(this->imageCreateFlags, that.imageCreateFlags);
std::swap(this->syncedTextureParamsVersion, that.syncedTextureParamsVersion);
std::swap(this->syncedContentVersion, that.syncedContentVersion);
std::swap(this->syncedMipLevelCount, that.syncedMipLevelCount);
}
void Reset() {
if (fullView != VK_NULL_HANDLE) {
vkDestroyImageView(s_device, fullView, nullptr);
}
if (sampledView != VK_NULL_HANDLE) {
vkDestroyImageView(s_device, sampledView, nullptr);
}
for (const auto attachmentView : perMipViews) {
if (attachmentView != VK_NULL_HANDLE) {
vkDestroyImageView(s_device, attachmentView, nullptr);
}
}
for (const auto sampledView : perMipSampledViews) {
if (sampledView != VK_NULL_HANDLE) {
vkDestroyImageView(s_device, sampledView, nullptr);
}
}
for (const auto& [_, attachmentView] : attachmentViews) {
if (attachmentView != VK_NULL_HANDLE) {
vkDestroyImageView(s_device, attachmentView, nullptr);
}
}
for (const auto& [_, sampledView] : alternateSampledViews) {
if (sampledView != VK_NULL_HANDLE) {
vkDestroyImageView(s_device, sampledView, nullptr);
}
}
for (const auto& [_, storageImageView] : storageImageViews) {
if (storageImageView != VK_NULL_HANDLE) {
vkDestroyImageView(s_device, storageImageView, nullptr);
}
}
if (image != VK_NULL_HANDLE && allocation != nullptr) {
vmaDestroyImage(s_allocator, image, allocation);
}
fullView = VK_NULL_HANDLE;
sampledView = VK_NULL_HANDLE;
perMipViews.clear();
perMipSampledViews.clear();
attachmentViews.clear();
alternateSampledViews.clear();
storageImageViews.clear();
image = VK_NULL_HANDLE;
allocation = nullptr;
layout = VK_IMAGE_LAYOUT_UNDEFINED;
extent = {0, 0};
depth = 1;
arrayLayers = 1;
mipLevels = 1;
sampledBaseMipLevel = 0;
sampledLevelCount = 1;
format = VK_FORMAT_UNDEFINED;
aspect = VK_IMAGE_ASPECT_NONE;
viewType = VK_IMAGE_VIEW_TYPE_2D;
sampleCount = VK_SAMPLE_COUNT_1_BIT;
imageCreateFlags = 0;
syncedTextureParamsVersion = 0;
syncedContentVersion = 0;
syncedMipLevelCount = 0;
}
~TextureResource() {
Reset();
}
static inline VkDevice s_device = VK_NULL_HANDLE;
static inline VmaAllocator s_allocator = VK_NULL_HANDLE;
};
Bool Initialize(const InitInfo& initInfo);
void Shutdown();
void BeginFrame(Uint32 frameIndex);
TextureResource* SyncTextureAndGetDescriptor(
MG_State::GLState::ITextureObject& texture);
VkImageView GetOrCreateViewAtMipLevel(MG_State::GLState::ITextureObject& texture, Uint32 mipLevel);
VkImageView GetOrCreateAttachmentViewAtMipLevel(MG_State::GLState::ITextureObject& texture, Uint32 mipLevel,
Uint32 baseArrayLayer, Uint32 layerCount,
VkImageViewType viewType);
VkImageView GetOrCreateSampledViewAtMipLevel(MG_State::GLState::ITextureObject& texture, Uint32 mipLevel);
VkImageView GetOrCreateSampledImageView(MG_State::GLState::ITextureObject& texture, VkFormat format);
VkImageView GetOrCreateStorageImageView(MG_State::GLState::ITextureObject& texture, Uint32 mipLevel,
VkFormat format, Bool layered, Int32 layer);
void UpdateTrackedImageLayout(MG_State::GLState::ITextureObject* texture, VkImageLayout newLayout);
void UpdateTrackedImageLayoutAfterAttachmentWrite(VkCommandBuffer commandBuffer,
MG_State::GLState::ITextureObject* texture,
Uint32 writtenMipLevel,
VkImageLayout newLayout);
Bool TransitionTextureForSampling(VkCommandBuffer commandBuffer, MG_State::GLState::ITextureObject& texture);
Bool TransitionTextureForStorageImage(VkCommandBuffer commandBuffer, MG_State::GLState::ITextureObject& texture);
// 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
// true - a false positive merely ends the render pass, a false negative would skip a barrier.
Bool NeedsStorageImagePreparation(MG_State::GLState::ITextureObject& texture) const;
static VkImageAspectFlags ResolveSampledImageViewAspectMask(VkImageAspectFlags imageAspect);
static VkFormat ResolveSampledImageViewFormat(VkFormat imageFormat, SamplerNumericDomain numericDomain);
static Bool AreSampledImageViewFormatsCompatible(VkFormat imageFormat, VkFormat viewFormat);
static Bool AreStorageImageViewFormatsCompatible(VkFormat imageFormat, VkFormat viewFormat);
static Bool TransitionImageLayout(VkCommandBuffer commandBuffer, VkImage image, VkImageLayout& trackedLayout,
VkImageLayout newLayout, VkPipelineStageFlags srcStageMask,
VkPipelineStageFlags dstStageMask, VkAccessFlags srcAccessMask,
VkAccessFlags dstAccessMask, VkImageAspectFlags aspectMask,
Uint32 baseMipLevel = 0, Uint32 levelCount = 1,
Uint32 layerCount = 1);
SizeT CollectGarbage();
// Per-draw sync memo. Within a single SetupDraw the same sampled texture is
// resolved ~3x (SetupDraw's layout-probe loop, its post-transition loop, and
// again inside ResolveSamplerDescriptor). No GL texture mutation can happen
// mid-SetupDraw, and layout is tracked on the TextureResource independently of
// SyncTexture, so after the first successful sync of a texture in a draw the
// heavy SyncTexture work (mip-completeness/resource/view resync + dirty scan)
// is pure redundancy. BeginDrawSyncScope opens a window in which repeat
// SyncTextureAndGetDescriptor calls short-circuit to the already-synced
// resource; EndDrawSyncScope closes it. Use the RAII DrawSyncScope guard.
void BeginDrawSyncScope();
void EndDrawSyncScope();
// RAII guard that opens/closes a per-draw sync memo window (see above).
class DrawSyncScope {
public:
explicit DrawSyncScope(VkTextureManager& manager) : m_manager(manager) { m_manager.BeginDrawSyncScope(); }
~DrawSyncScope() { m_manager.EndDrawSyncScope(); }
DrawSyncScope(const DrawSyncScope&) = delete;
DrawSyncScope& operator=(const DrawSyncScope&) = delete;
private:
VkTextureManager& m_manager;
};
private:
// Bumped in SyncTextureResource right after vmaCreateImage(texture). See GetTextureImageEpoch().
Uint64 m_textureImageEpoch = 1;
Bool SyncTexture(MG_State::GLState::ITextureObject &texture,
TextureResource &outResource);
Bool SyncTextureResource(const MG_State::GLState::ITextureObject &texture,
TextureUploadTarget uploadTarget,
const IntVec3 &texelSize, SizeT byteSize, Uint32 mipLevels,
TextureResource &resource);
Bool SyncTextureViews(const MG_State::GLState::ITextureObject& texture, TextureResource& resource);
VkImageView CreateImageView(VkImage image, VkFormat format, VkImageAspectFlags aspect,
VkImageViewType viewType, Uint32 baseMipLevel, Uint32 levelCount,
Uint32 baseArrayLayer,
Uint32 layerCount,
const VkComponentMapping* components = nullptr,
VkImageUsageFlags viewUsage = 0) const;
Bool UploadDirtyMipLevels(MG_State::GLState::TextureObjectMipmap &mipmapTexture,
TextureUploadTarget uploadTarget,
TextureResource &outResource);
static Bool CheckMipmapCompleteness(const MG_State::GLState::ITextureObject& texture,
TextureUploadTarget& outTarget,
IntVec3& outTexelSize,
SizeT& outByteSize,
Uint32& outMipLevelCount);
static Uint32 GetUploadMipLevelCount(const MG_State::GLState::TextureObjectMipmap& texture, TextureUploadTarget target);
static void ResolveViewMipRange(const MG_State::GLState::ITextureObject& texture, Uint32 mipLevels,
Uint32& outBaseMipLevel, Uint32& outLevelCount);
static VkImageAspectFlags GetAspectMaskForFormat(VkFormat format);
void DeferResourceRelease(TextureResource&& resource);
void DeferViewRelease(VkImageView view);
void CollectDeferredReleases(Uint32 frameIndex);
void DestroyDeferredReleases();
static TextureIdentity MakeTextureIdentity(MG_State::GLState::ITextureObject* texture);
void EraseTrackedTexture(const TextureIdentity& identity);
void PruneStaleTextureAliases(MG_State::GLState::ITextureObject* texture);
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;
Uint32 m_currentFrameIndex = 0;
Uint8 m_gcCounter = 0;
// Active only between BeginDrawSyncScope/EndDrawSyncScope; identities of
// textures already fully synced in the current draw (small N -> flat scan).
Bool m_drawSyncScopeActive = false;
// Per-draw sync memo: the identity plus the resolved resource pointer. The pointer is stable
// across rehash in the node-based m_textureResources and stays valid for the draw (a texture
// synced this draw is alive and is not erased mid-draw), so a repeat sync of the same texture
// returns the resource without re-hashing the identity into m_textureResources.
struct DrawSyncedTexture {
TextureIdentity identity;
TextureResource* resource = nullptr;
};
Vector<DrawSyncedTexture> m_drawSyncedThisDraw;
// 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;
Vector<Vector<TextureResource>> m_deferredReleases;
Vector<Vector<VkImageView>> m_deferredViewReleases;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,179 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkTimerQueryManager.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 "VkTimerQueryManager.h"
namespace MobileGL::MG_Backend::DirectVulkan {
Bool VkTimerQueryManager::Initialize(const InitInfo& initInfo) {
Shutdown();
MOBILEGL_ASSERT(initInfo.device != VK_NULL_HANDLE, "VkTimerQueryManager::Initialize requires valid VkDevice");
MOBILEGL_ASSERT(initInfo.frameCount > 0, "VkTimerQueryManager::Initialize requires non-zero frame count");
if (initInfo.timestampValidBits == 0 || initInfo.timestampPeriodNs <= 0.0f || initInfo.slotsPerPool == 0) {
MGLOG_W("VkTimerQueryManager: timestamps unsupported (validBits=%u, period=%f, slots=%u)",
initInfo.timestampValidBits, initInfo.timestampPeriodNs, initInfo.slotsPerPool);
return false;
}
m_device = initInfo.device;
m_timestampPeriodNs = initInfo.timestampPeriodNs;
m_validBitsMask = initInfo.timestampValidBits >= 64
? ~0ull
: ((1ull << initInfo.timestampValidBits) - 1ull);
m_slotsPerPool = initInfo.slotsPerPool;
m_pools.resize(initInfo.frameCount);
VkQueryPoolCreateInfo poolInfo{};
poolInfo.sType = VK_STRUCTURE_TYPE_QUERY_POOL_CREATE_INFO;
poolInfo.queryType = VK_QUERY_TYPE_TIMESTAMP;
poolInfo.queryCount = m_slotsPerPool;
for (auto& poolState : m_pools) {
const VkResult result = vkCreateQueryPool(m_device, &poolInfo, nullptr, &poolState.pool);
if (result != VK_SUCCESS) {
MGLOG_E("VkTimerQueryManager: vkCreateQueryPool failed with %s", VkResultToString(result));
Shutdown();
return false;
}
}
return true;
}
void VkTimerQueryManager::Shutdown() {
if (m_device != VK_NULL_HANDLE) {
for (auto& poolState : m_pools) {
if (poolState.pool != VK_NULL_HANDLE) {
vkDestroyQueryPool(m_device, poolState.pool, nullptr);
}
}
}
// Records the frontend still holds simply stay unharvested; their
// results read back as 0.
m_pools.clear();
m_device = VK_NULL_HANDLE;
m_timestampPeriodNs = 0.0f;
m_validBitsMask = 0;
m_slotsPerPool = 0;
}
void VkTimerQueryManager::OnFrameCommandRecordingBegan(VkCommandBuffer commandBuffer, Uint32 frameIndex,
Uint64 frameSerial) {
MOBILEGL_ASSERT(frameIndex < m_pools.size(), "VkTimerQueryManager frame index out of range");
auto& poolState = m_pools[frameIndex];
if (poolState.preparedFrameSerial == frameSerial) {
// Recording re-began within the same frame (mid-frame readback
// submit or the Present layout transition); the pool was already
// harvested and reset for this cycle, and resetting again would
// clobber timestamps written earlier in the frame.
return;
}
// Harvest what the pool's previous cycle left behind. The frame slot's
// fence was waited before re-recording, so every executed query is
// already available and the reads return immediately.
DrainPoolPending(poolState);
vkCmdResetQueryPool(commandBuffer, poolState.pool, 0, m_slotsPerPool);
poolState.cursor = 0;
poolState.exhaustionWarned = false;
poolState.preparedFrameSerial = frameSerial;
}
SharedPtr<VkTimerQueryManager::TimestampRecord> VkTimerQueryManager::WriteTimestamp(VkCommandBuffer commandBuffer,
Uint32 frameIndex,
Uint64 frameSerial) {
MOBILEGL_ASSERT(frameIndex < m_pools.size(), "VkTimerQueryManager frame index out of range");
auto& poolState = m_pools[frameIndex];
if (poolState.cursor >= m_slotsPerPool) {
if (!poolState.exhaustionWarned) {
MGLOG_W("VkTimerQueryManager: frame %u timestamp pool exhausted (%u slots); further timer queries "
"this frame fall back to the frontend path",
frameIndex, m_slotsPerPool);
poolState.exhaustionWarned = true;
}
return nullptr;
}
auto record = MakeShared<TimestampRecord>();
record->poolIndex = frameIndex;
record->slot = poolState.cursor++;
record->frameSerial = frameSerial;
vkCmdWriteTimestamp(commandBuffer, VK_PIPELINE_STAGE_BOTTOM_OF_PIPE_BIT, poolState.pool, record->slot);
poolState.pendingRecords.push_back(record);
return record;
}
Bool VkTimerQueryManager::TryHarvest(TimestampRecord& record) {
if (record.harvested) {
return true;
}
if (m_device == VK_NULL_HANDLE || record.poolIndex >= m_pools.size()) {
return false;
}
Uint64 resultWithAvailability[2] = {0, 0};
const VkResult result = vkGetQueryPoolResults(
m_device, m_pools[record.poolIndex].pool, record.slot, 1, sizeof(resultWithAvailability),
resultWithAvailability, sizeof(Uint64), VK_QUERY_RESULT_64_BIT | VK_QUERY_RESULT_WITH_AVAILABILITY_BIT);
if (result != VK_SUCCESS && result != VK_NOT_READY) {
MGLOG_E("VkTimerQueryManager: vkGetQueryPoolResults failed with %s", VkResultToString(result));
return false;
}
if (resultWithAvailability[1] == 0) {
return false;
}
record.rawTicks = resultWithAvailability[0];
record.harvested = true;
return true;
}
void VkTimerQueryManager::InvalidatePendingRecords() {
for (auto& poolState : m_pools) {
DrainPoolPending(poolState);
// Force a harvest-free reset cycle the next time this pool's frame
// begins recording.
poolState.preparedFrameSerial = 0;
}
}
void VkTimerQueryManager::DrainPoolPending(PoolState& poolState) {
for (auto& record : poolState.pendingRecords) {
if (record->harvested) {
continue;
}
if (!TryHarvest(*record)) {
// The commands carrying this timestamp never executed (they
// were dropped, e.g. by a swapchain recreation mid-frame).
// Mark the record resolved-as-invalid so waits on it cannot
// hang; its result reads back as 0.
record->harvested = true;
record->valid = false;
}
}
poolState.pendingRecords.clear();
}
Uint64 VkTimerQueryManager::MaskToValidBits(Uint64 ticks) const {
return ticks & m_validBitsMask;
}
Uint64 VkTimerQueryManager::ElapsedNs(const TimestampRecord& begin, const TimestampRecord& end) const {
if (!begin.valid || !end.valid) {
return 0;
}
const Uint64 deltaTicks = MaskToValidBits(end.rawTicks - begin.rawTicks);
return static_cast<Uint64>(static_cast<double>(deltaTicks) * static_cast<double>(m_timestampPeriodNs));
}
Uint64 VkTimerQueryManager::TimestampNs(const TimestampRecord& record) const {
if (!record.valid) {
return 0;
}
return static_cast<Uint64>(static_cast<double>(MaskToValidBits(record.rawTicks)) *
static_cast<double>(m_timestampPeriodNs));
}
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,111 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VkTimerQueryManager.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 "../VkIncludes.h"
#include <Includes.h>
namespace MobileGL::MG_Backend::DirectVulkan {
// GPU timestamp storage backing the GL timer-query frontend (GL_TIME_ELAPSED
// spans and GL_TIMESTAMP one-shots): one VkQueryPool of timestamp slots per
// frame in flight.
//
// Per-frame lifecycle: right after a frame slot's command buffer begins
// recording (and before any render pass, since vkCmdResetQueryPool must be
// recorded outside one), OnFrameCommandRecordingBegan harvests every
// not-yet-read slot of the pool about to be reused (the slot's frame fence
// was waited before re-recording, so the results are already available),
// records a reset of the whole pool, and rewinds the allocation cursor.
class VkTimerQueryManager {
public:
// One vkCmdWriteTimestamp landing spot. Shared (via SharedPtr) between
// the frontend-held query object and the owning pool's pending list, so
// deleting a query while its result is still in flight never leaves the
// pool with a dangling record.
struct TimestampRecord {
Uint32 poolIndex = 0;
Uint32 slot = 0;
// VkBufferManager frame serial current when the timestamp was
// recorded; result availability is bounded by its completion.
Uint64 frameSerial = 0;
Bool harvested = false;
// Cleared when the recorded commands were dropped before they could
// execute (swapchain recreation abandons the in-progress command
// buffer); the result then reads back as 0.
Bool valid = true;
Uint64 rawTicks = 0;
};
struct InitInfo {
VkDevice device = VK_NULL_HANDLE;
Uint32 frameCount = 0;
Uint32 timestampValidBits = 0;
Float timestampPeriodNs = 0.0f; // nanoseconds per timestamp tick
Uint32 slotsPerPool = 128;
};
Bool Initialize(const InitInfo& initInfo);
// The caller guarantees the device is idle (same contract as the other
// DirectVulkan managers' Shutdown paths).
void Shutdown();
// The per-frame hook described in the class comment. Re-begins within
// the same frame serial (mid-frame readback submits, the Present layout
// transition) are skipped so already-written slots survive.
void OnFrameCommandRecordingBegan(VkCommandBuffer commandBuffer, Uint32 frameIndex, Uint64 frameSerial);
// Allocates a slot from the frame's pool and records a bottom-of-pipe
// vkCmdWriteTimestamp (valid both inside and outside a render pass).
// Returns null on pool exhaustion, with one warning per pool cycle; the
// frontend falls back gracefully on a null handle.
SharedPtr<TimestampRecord> WriteTimestamp(VkCommandBuffer commandBuffer, Uint32 frameIndex,
Uint64 frameSerial);
// Non-blocking single-slot read (WITH_AVAILABILITY, no WAIT). Returns
// true once the record holds its raw ticks. Callers gate this on the
// record's frame serial being complete.
Bool TryHarvest(TimestampRecord& record);
// Reads every pending result that is available (the caller guarantees
// the device is idle) and marks the rest invalid. Called when recorded
// but unsubmitted commands are dropped (swapchain recreation), which
// would otherwise leave slots that never become available. Each pool is
// reset lazily on its next OnFrameCommandRecordingBegan.
void InvalidatePendingRecords();
// end - begin using unsigned wrap arithmetic masked to the queue's
// timestampValidBits, converted to nanoseconds. 0 if either record was
// invalidated.
Uint64 ElapsedNs(const TimestampRecord& begin, const TimestampRecord& end) const;
// Raw GPU timestamp converted to nanoseconds. 0 if invalidated.
Uint64 TimestampNs(const TimestampRecord& record) const;
private:
struct PoolState {
VkQueryPool pool = VK_NULL_HANDLE;
Uint32 cursor = 0;
// Frame serial the pool was last harvested + reset for; guards
// against double resets when recording re-begins mid-frame.
Uint64 preparedFrameSerial = 0;
Bool exhaustionWarned = false;
Vector<SharedPtr<TimestampRecord>> pendingRecords;
};
Uint64 MaskToValidBits(Uint64 ticks) const;
// Harvest (or invalidate, when the result never became available)
// every pending record of a pool and clear its pending list.
void DrainPoolPending(PoolState& pool);
VkDevice m_device = VK_NULL_HANDLE;
Float m_timestampPeriodNs = 0.0f;
Uint64 m_validBitsMask = 0;
Uint32 m_slotsPerPool = 0;
Vector<PoolState> m_pools;
};
} // namespace MobileGL::MG_Backend::DirectVulkan
File diff suppressed because it is too large Load Diff
@@ -1,609 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VulkanRenderer.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 "Config.h"
#include "FrameContext.h"
#include "PipelineFactory.h"
#include "ProgramFactory.h"
#include "SwapchainObject.h"
#include "UniformManager.h"
#include "VertexInputStateFactory.h"
#include "VkBufferObject.h"
#include "VkBufferManager.h"
#include "VkClearManager.h"
#include "VkRenderPassManager.h"
#include "VkSamplerManager.h"
#include "VkTextureManager.h"
#include "VkTimerQueryManager.h"
#include "MG_Util/Math/VectorTypes.h"
#include <Includes.h>
#include <vk_mem_alloc.h>
#include "../VkIncludes.h"
namespace MobileGL::MG_State::GLState {
class FramebufferObject;
class ProgramObject;
class SamplerObject;
class VertexArrayObject;
} // namespace MobileGL::MG_State::GLState
namespace MobileGL::MG_Backend::DirectVulkan {
enum class DrawSetupAspect: Uint8 {
FramebufferObject = 1 << 0,
VertexArrayObject = 1 << 1,
UniformBuffer = 1 << 2,
VertexBuffer = 1 << 3,
IndexBuffer = 1 << 4,
IndirectDrawBuffer = 1 << 5,
Viewport = 1 << 6,
Scissor = 1 << 7,
};
struct DrawCmdParam {
Uint32 vertexCount = 0;
Uint32 instanceCount = 1;
Uint32 firstVertex = 0;
Uint32 firstInstance = 0;
// Indexed-draw metadata for bounding vertex-stream conversion. baseVertex is the
// draw's base-vertex offset; indexRangeIsExactView is true only when the draw
// fetches exactly the indices its IndexBufferView describes (direct DrawElements;
// multi/indirect forms leave it false because the CPU cannot bound their ranges).
Int32 baseVertex = 0;
Bool indexRangeIsExactView = false;
};
struct DrawIndexedCmdParam {
Uint32 indexCount = 0;
Uint32 instanceCount = 1;
Uint32 firstIndex = 0;
Int32 vertexOffset = 0;
Int32 firstInstance = 0;
};
struct DrawCmd {
GLenum mode = GL_TRIANGLES;
DrawCmdParam params;
};
struct IndexBufferView {
GLenum indexType = GL_UNSIGNED_SHORT;
SizeT indexByteOffset = 0;
SizeT indexByteSize = 0;
};
struct DrawIndexedCmd {
GLenum mode = GL_TRIANGLES;
IndexBufferView indexBufferView;
DrawIndexedCmdParam params;
};
struct MultiDrawIndexedCmd {
GLenum mode = GL_TRIANGLES;
IndexBufferView indexBufferView;
Uint32 drawCount = 0;
DrawIndexedCmdParam* pParams = nullptr;
};
struct MultiDrawCmd {
GLenum mode = GL_TRIANGLES;
Uint32 drawCount = 0;
DrawCmdParam* pParams = nullptr;
};
struct QueueFamilyIndices {
Int32 graphicsFamily = -1;
Int32 presentFamily = -1;
};
struct PhysicalDevice {
QueueFamilyIndices queueFamilies;
VkPhysicalDeviceProperties properties;
VkPhysicalDevice handle = VK_NULL_HANDLE;
Bool IsComplete() const {
return handle != VK_NULL_HANDLE && queueFamilies.graphicsFamily != -1 && queueFamilies.presentFamily != -1;
}
};
class VulkanRenderer : public IBufferCopyCommandProvider, public FrameContext::IRecordingObserver {
public:
VulkanRenderer(NativeWindowType window, const VulkanRendererConfig& cfg = {});
~VulkanRenderer();
void Initialize();
void Shutdown();
// IBufferCopyCommandProvider: recording command buffer, outside any
// render pass, for immediate staged buffer copies.
VkCommandBuffer AcquireBufferCopyCommandBuffer() override;
// FrameContext::IRecordingObserver: prepares the frame's timer-query
// pool (harvest + reset) right after the frame command buffer begins
// recording, before any render pass.
void OnFrameCommandRecordingBegan(VkCommandBuffer commandBuffer) override;
Bool SetupDraw(FrameContext::FrameData& frame, GLenum mode, Flags<DrawSetupAspect> aspects,
const DrawCmdParam& drawParams,
const IndexBufferView* pIndexBufferView = nullptr);
void ClearAttachmentsOnActiveRenderPass(VkCommandBuffer commandBuffer,
const RenderPassEntry& compatibleRenderPassEntry);
enum class ScissoredClearPrep {
NotNeeded, // scissor covers the whole target — take the deferred whole-surface path instead
NoOp, // nothing to clear (degenerate target or empty scissor rect)
Ready, // a render pass is active; record vkCmdClearAttachments with the returned rect
};
ScissoredClearPrep PrepareScissoredClear(const MG_State::GLState::FramebufferObject& framebuffer,
VkClearRect& outClearRect);
void Clear(GLbitfield mask);
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);
void ClearBufferiv(GLenum buffer, GLint drawbuffer, const GLint* value);
void ClearNamedFramebufferfv(const SharedPtr<MG_State::GLState::FramebufferObject>& framebuffer,
GLenum buffer, GLint drawbuffer, const GLfloat* value);
void ClearNamedFramebufferfi(const SharedPtr<MG_State::GLState::FramebufferObject>& framebuffer,
GLenum buffer, GLint drawbuffer, GLfloat depth, GLint stencil);
void BlitFramebuffer(GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1,
GLint dstX0, GLint dstY0, GLint dstX1, GLint dstY1,
GLbitfield mask, GLenum filter);
void BlitNamedFramebuffer(const SharedPtr<MG_State::GLState::FramebufferObject>& readFbo,
const SharedPtr<MG_State::GLState::FramebufferObject>& drawFbo,
GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1,
GLint dstX0, GLint dstY0, GLint dstX1, GLint dstY1,
GLbitfield mask, GLenum filter);
void CopyTexSubImage2D(GLenum target, GLint level, GLint xoffset, GLint yoffset,
GLint x, GLint y, GLsizei width, GLsizei height);
void CopyImageSubData(const SharedPtr<MG_State::GLState::ITextureObject>& srcTexture,
GLenum srcTarget, GLint srcLevel, GLint srcX, GLint srcY, GLint srcZ,
const SharedPtr<MG_State::GLState::ITextureObject>& dstTexture,
GLenum dstTarget, GLint dstLevel, GLint dstX, GLint dstY, GLint dstZ,
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);
static SizeT GetReadbackTexelSize(VkFormat sourceFormat);
static Bool ConvertReadbackPixels(const Uint8* sourcePixels, VkFormat sourceFormat,
GLsizei width, GLsizei height, GLenum destinationFormat,
GLenum destinationType, SizeT destinationRowStride,
Uint8* destinationPixels);
void GetTexImage(GLenum target, GLint level, GLenum format, GLenum type, GLvoid* pixels);
void GetTextureImage(const SharedPtr<MG_State::GLState::ITextureObject>& texture,
TextureUploadTarget uploadTarget, GLint level, GLenum format, GLenum type,
GLsizei bufSize, GLvoid* pixels);
void DispatchCompute(GLuint numGroupsX, GLuint numGroupsY, GLuint numGroupsZ);
void DispatchComputeIndirect(GLintptr indirect);
void MemoryBarrier(GLbitfield barriers);
static VkMemoryBarrier BuildMemoryBarrierForGlBarriers(GLbitfield barriers);
void DrawArrays(const DrawCmd& payload);
void DrawElements(const DrawIndexedCmd& payload);
void MultiDrawArrays(const MultiDrawCmd& payload);
void MultiDrawElements(const MultiDrawIndexedCmd& payloads);
void MultiDrawElementsIndirect(GLenum mode, GLenum type, const void* indirect, GLsizei drawcount,
GLsizei stride);
void MultiDrawArraysIndirect(GLenum mode, const void* indirect, GLsizei drawcount, GLsizei stride);
void MultiDrawElementsIndirectCount(GLenum mode, GLenum type, const void* indirect, GLintptr drawcount,
GLsizei maxdrawcount, GLsizei stride);
void Present();
const PhysicalDevice& GetPhysicalDevice() const;
VkInstance GetInstance() const;
Bool IsDrawIndirectCountExtensionEnabled() const;
// GL fence support, expressed in queue-submission indices backed by
// real VkFences. A GL fence captures GetSyncPointSubmitIndex() at
// creation: the index of the submission that will carry the commands
// recorded so far (m_submitCounter + 1 while work is pending, or
// m_submitCounter when nothing has been recorded since the last
// submit). It is signaled once that submission's fence is observed
// signaled - unlike the frame-serial heuristic, this makes fences
// signal as soon as the GPU actually finishes, which MC 1.21.5's
// fence-paced ring buffers rely on to recycle their space.
Uint64 GetSyncPointSubmitIndex() const;
// Non-blocking: polls outstanding submission fences and reports
// whether every submission up to `submitIndex` has completed.
Bool IsSubmitIndexComplete(Uint64 submitIndex);
// Submits the commands recorded so far without waiting (GL flush).
// Recording restarts lazily on a fresh command buffer; the submitted
// one is retired until the frame slot's fence is next waited. Returns
// true when a submission was made.
Bool FlushPendingCommands();
// Flush gated on usefulness: only flushes when `submitIndex` is still
// unsubmitted, so poll loops on already-submitted fences do not split
// the frame's render pass (a full tile load/store on TBDR GPUs).
Bool FlushForSyncPoint(Uint64 submitIndex);
// Blocking wait for a submission index with a nanosecond timeout.
// When the index is still unsubmitted and flushIfPending is set, the
// pending commands are flushed first so the wait can make progress.
Bool WaitForSubmitIndex(Uint64 submitIndex, Uint64 timeoutNs, Bool flushIfPending);
// Frame-serial completion, still used by the timer-query paths (their
// records are bucketed per frame slot).
Bool IsFrameSerialComplete(Uint64 serial) const;
// Blocking wait for a submitted serial. Returns false when the serial
// cannot complete without further submissions (it belongs to the
// current, not-yet-presented frame) or when the wait failed.
Bool WaitForFrameSerial(Uint64 serial, Uint64 timeoutNs);
// GPU timer queries, backing the GL_TIME_ELAPSED / GL_TIMESTAMP
// frontend. Timestamp support (queue timestampValidBits > 0 and a
// non-zero timestampPeriod) is cached at device creation.
Bool IsTimerQuerySupported() const;
// The samplerAnisotropy device feature was granted, so GL_TEXTURE_MAX_ANISOTROPY_EXT is
// honored rather than accepted-and-ignored.
Bool IsSamplerAnisotropySupported() const { return m_samplerAnisotropyFeatureEnabled; }
// Ensures the frame command buffer is recording (same lazy pattern as
// SetupDraw) and writes a bottom-of-pipe timestamp into the current
// frame's pool. Null when unsupported or the pool is exhausted.
SharedPtr<VkTimerQueryManager::TimestampRecord> WriteTimerQueryTimestamp();
// Non-blocking: true once the record's raw ticks are on the CPU
// (harvests the slot once its frame serial has completed).
Bool IsTimerQueryResultReady(VkTimerQueryManager::TimestampRecord& record);
// Blocking wait, mirroring ClientWaitSync's caveat: a record written
// this frame cannot complete until Present submits the commands, so
// this returns false (result reads as 0) instead of deadlocking.
Bool WaitForTimerQueryResult(VkTimerQueryManager::TimestampRecord& record);
Uint64 GetTimerQueryElapsedNs(const VkTimerQueryManager::TimestampRecord& begin,
const VkTimerQueryManager::TimestampRecord& end) const;
Uint64 GetTimerQueryTimestampNs(const VkTimerQueryManager::TimestampRecord& record) const;
void RequestSwapchainResize(Uint32 width, Uint32 height);
// Returns false when the surface is zero-area (minimized/hidden window):
// no new swapchain is installed and presentation must stay suspended.
Bool RecreateSwapchain();
private:
struct BlitUniformData {
float srcRect[4] = {0.f, 0.f, 1.f, 1.f};
float dstRect[4] = {0.f, 0.f, 1.f, 1.f};
Int surfaceTransform = 0;
Int padding[3] = {0, 0, 0};
};
struct BlitResources {
SharedPtr<MG_State::GLState::ProgramObject> program;
SharedPtr<MG_State::GLState::SamplerObject> nearestSampler;
SharedPtr<MG_State::GLState::SamplerObject> linearSampler;
Int srcRectLocation = -1;
Int dstRectLocation = -1;
Int surfaceTransformLocation = -1;
Uint32 samplerBinding = 0;
};
struct DepthMipmapResources {
SharedPtr<MG_State::GLState::ProgramObject> program;
Int srcRectLocation = -1;
Int dstRectLocation = -1;
Int surfaceTransformLocation = -1;
Int srcTexelSizeLocation = -1;
Uint32 samplerBinding = 0;
};
struct DeferredDepthMipmapCleanup {
Vector<VkImageView> imageViews;
Vector<VkFramebuffer> framebuffers;
Vector<VkRenderPass> renderPasses;
Vector<VkPipeline> pipelines;
};
void QueueClearBufferPayload(GLenum buffer, GLint drawbuffer, const ClearAttachmentPayload& clearPayload);
void QueueClearBufferPayloadForFramebuffer(const MG_State::GLState::FramebufferObject& framebuffer,
GLenum buffer, GLint drawbuffer,
const ClearAttachmentPayload& clearPayload);
void RecordScissoredClearBuffer(const MG_State::GLState::FramebufferObject& framebuffer,
GLenum buffer, GLint drawbuffer,
const ClearAttachmentPayload& clearPayload,
const VkClearRect& clearRect);
// ---- Submission fence tracking (GL sync objects) ----
// One record per vkQueueSubmit still in flight, in ascending submit
// order. Present/readback submissions reference the frame slot's
// fence (not pool-owned); mid-frame flushes use pooled fences that are
// recycled once their submission is observed complete.
// Not thread-safe: like the rest of the renderer, the tracker relies
// on GL calls being serialized (launchers migrate the context across
// threads, but calls never run concurrently), so sync-object polls
// may mutate it without locking.
struct SubmitRecord {
Uint64 submitIndex = 0;
// Buffer-manager frame serial the submission was made under; its
// completion raises the completed-serial floor (timer queries and
// buffer busy-tracking live in frame-serial space).
Uint64 frameSerial = 0;
VkFence fence = VK_NULL_HANDLE;
Bool pooledFence = false;
};
// Registers a submission that vkQueueSubmit just made with `fence`.
// Invariant: every graphics-queue submission that outlives its call
// site must be registered so GL fences observe it. Exempt are the
// texture-upload/preserve submits in VkTextureManager, which
// vkWaitForFences inline before returning.
void RegisterSubmit(VkFence fence, Bool pooledFence);
// Builds the submit packet for the frame's pending command buffer
// (consuming the acquire semaphore on the slot's first submission),
// submits it with `fence`, and registers the submission. On failure
// the frame state is left untouched. Shared by the mid-frame flush
// and the readback path so the semaphore-consumption invariant lives
// in one place.
Bool SubmitPendingCommandBuffer(FrameContext::FrameData& frame, VkFence fence, Bool pooledFence);
// Polls in-flight submission fences (prefix order) and advances the
// completed counter past every fence observed signaled.
void RefreshCompletedSubmits();
// All submissions up to `submitIndex` are known complete (their fence
// was waited or the device was idled); drops their records and
// recycles pooled fences.
void OnSubmitsCompletedUpTo(Uint64 submitIndex);
VkFence AcquirePooledSubmitFence();
void DestroySubmitFencePool();
Bool HasPendingRecordedWork() const;
Vector<SubmitRecord> m_inFlightSubmits;
Vector<VkFence> m_freeSubmitFences;
Uint64 m_submitCounter = 0;
Uint64 m_completedSubmitCounter = 0;
NativeWindowType m_window = 0;
void* m_platformDisplay = nullptr;
void* m_platformLibrary = nullptr;
void* m_platformCloseDisplay = nullptr;
VulkanRendererConfig m_config;
Bool m_swapchainResizeRequested = false;
// Presentation is suspended while the window is zero-area (minimized): the
// swapchain is unusable/out of date, so Present drops frames instead of
// submitting on a signaled fence / presenting never-acquired images.
Bool m_presentSuspended = false;
// Vulkan objects
Bool m_validationLayersEnabled = false;
Vector<VkExtensionProperties> m_extensions;
VkInstance m_instance = VK_NULL_HANDLE;
VkDebugUtilsMessengerEXT m_debugMessenger = VK_NULL_HANDLE;
PhysicalDevice m_physicalDevice;
VkDevice m_device = VK_NULL_HANDLE;
VmaAllocator m_allocator = nullptr;
VkSurfaceKHR m_surface = VK_NULL_HANDLE;
SwapchainObject m_swapchainObject;
VkQueue m_graphicsQueue = VK_NULL_HANDLE;
VkQueue m_presentQueue = VK_NULL_HANDLE;
Bool m_drawIndirectCountExtensionEnabled = false;
Bool m_indexTypeUint8ExtensionEnabled = false;
Bool m_logicOpFeatureEnabled = false;
Bool m_multiDrawIndirectFeatureEnabled = false;
Bool m_samplerAnisotropyFeatureEnabled = false;
Bool m_shaderDrawParametersExtensionEnabled = false;
Bool m_shaderDrawParametersFeatureEnabled = false;
Bool m_unformattedFloatStorageImagesEnabled = false;
// fillModeNonSolid gates VK_POLYGON_MODE_LINE/_POINT (glPolygonMode); independentBlend gates
// per-draw-buffer color write masks (glColorMaski). Both are cached at device creation and
// drive a runtime fallback when the device lacks them.
Bool m_fillModeNonSolidFeatureEnabled = false;
Bool m_independentBlendFeatureEnabled = false;
// dualSrcBlend gates GL_SRC1_* blend factors (glBindFragDataLocationIndexed dual-source blend);
// primitiveTopologyListRestart gates primitive restart on *list* topologies (strip/fan restart
// needs no feature). Both cached at device creation and drive a hard-fail-at-draw when absent.
Bool m_dualSrcBlendFeatureEnabled = false;
Bool m_primitiveTopologyListRestartFeatureEnabled = false;
// Cached at device creation from the graphics queue family properties
// and device limits; drives timer-query support.
Uint32 m_timestampValidBits = 0;
Float m_timestampPeriodNs = 0.0f;
Bool m_timerQuerySupported = false;
using PFNDrawIndexedIndirectCountFunc = void(VKAPI_PTR*)(VkCommandBuffer commandBuffer, VkBuffer buffer,
VkDeviceSize offset, VkBuffer countBuffer,
VkDeviceSize countBufferOffset, Uint32 maxDrawCount,
Uint32 stride);
static inline PFNDrawIndexedIndirectCountFunc s_vkCmdDrawIndexedIndirectCount = nullptr;
VkCommandPool m_commandPool = VK_NULL_HANDLE;
VkBufferManager m_bufferManager;
Uint m_imageIndexAcquired = 0;
FrameContext m_frameContext;
UniquePtr<PipelineFactory> m_pipelineFactory;
// Single-slot "last pipeline" memo: skip the per-draw GetOrCreatePipeline work (state
// 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.
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;
UniquePtr<VertexInputStateFactory> m_vertexInputStateFactory;
UniquePtr<VkClearManager> m_clearManager;
UniquePtr<VkRenderPassManager> m_renderPassManager;
UniquePtr<VkTextureManager> m_textureManager;
UniquePtr<VkSamplerManager> m_samplerManager;
UniquePtr<VkTimerQueryManager> m_timerQueryManager;
BlitResources m_blitResources;
DepthMipmapResources m_depthMipmapResources;
Vector<DeferredDepthMipmapCleanup> m_deferredDepthMipmapCleanup;
// Skip the per-draw CollectSampledTextures walk (~5% of the render thread) when the sampled
// texture SET is provably unchanged from the previous draw: same program (lifetime id +
// backend-state version, which covers sampler-uniform reassignment / relink) and transform
// flags, and no texture bind/unbind/delete since (GetTextureBindGeneration). On a hit,
// m_sampledTexturesScratch still holds the previous draw's list and steps 2-4 (feedback /
// layout probe / transition) re-run on it, so layout correctness is unaffected - only the GL
// walk is skipped. The program lifetime id (never reused, unlike the GL name) and the
// monotonic bind generation make the key ABA-proof; the per-command-buffer reset is a cheap
// belt-and-suspenders.
Bool m_lastSampledSetValid = false;
Uint64 m_lastSampledSetProgramLifetimeId = 0;
Uint32 m_lastSampledSetProgramVersion = 0;
ProgramFactory::CompileOptionFlags m_lastSampledSetTransformFlags = {};
Uint64 m_lastSampledSetBindGeneration = 0;
// 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;
Vector<MG_State::GLState::ITextureObject*> m_storageImageTexturesScratch;
Vector<VkBuffer> m_vertexBuffersScratch;
Vector<VkDeviceSize> m_vertexOffsetsScratch;
Vector<VkVertexInputAttributeDescription> m_patchedAttributesScratch;
Vector<Float> m_vertexConversionScratch;
Vector<Uint8> m_vertexRepackScratch;
struct ConvertedVertexStreamKey {
const MG_State::GLState::BufferObject* buffer = nullptr;
Uint64 changeSerial = 0;
SizeT baseOffset = 0;
Uint32 sourceStride = 0;
DataType type = DataType::Float32;
Int size = 0;
Bool normalized = false;
Bool isInteger = false;
VertexInputStateFactory::VertexStreamConversion conversion =
VertexInputStateFactory::VertexStreamConversion::None;
Bool operator==(const ConvertedVertexStreamKey& other) const {
return buffer == other.buffer && changeSerial == other.changeSerial &&
baseOffset == other.baseOffset && sourceStride == other.sourceStride &&
type == other.type && size == other.size && normalized == other.normalized &&
isInteger == other.isInteger && conversion == other.conversion;
}
};
struct ConvertedVertexStreamKeyHash {
SizeT operator()(const ConvertedVertexStreamKey& key) const {
SizeT hash = std::hash<const void*>{}(key.buffer);
auto combine = [&hash](SizeT value) {
hash ^= value + static_cast<SizeT>(0x9e3779b97f4a7c15ull) + (hash << 6) + (hash >> 2);
};
combine(std::hash<Uint64>{}(key.changeSerial));
combine(std::hash<SizeT>{}(key.baseOffset));
combine(std::hash<Uint32>{}(key.sourceStride));
combine(std::hash<Uint32>{}(static_cast<Uint32>(key.type)));
combine(std::hash<Int>{}(key.size));
combine(std::hash<Bool>{}(key.normalized));
combine(std::hash<Bool>{}(key.isInteger));
combine(std::hash<Uint32>{}(static_cast<Uint32>(key.conversion)));
return hash;
}
};
struct ConvertedVertexStream {
BufferSlice slice;
// Number of source elements the cached slice covers. A draw needing a prefix of
// this range reuses the slice (converted streams are tightly packed); a draw
// needing more reconverts and replaces the entry, so per (buffer, layout) a
// frame converts at most the largest range any draw asked for.
SizeT elementCount = 0;
// Pins the source buffer for the frame so its heap address cannot be reused by
// a new BufferObject while this pointer-keyed entry is alive.
SharedPtr<const MG_State::GLState::BufferObject> sourcePin;
};
UnorderedMap<ConvertedVertexStreamKey, ConvertedVertexStream, ConvertedVertexStreamKeyHash>
m_convertedVertexStreams;
void CreateInstance();
VkResult SetupDebugMessenger();
VkResult DestroyDebugMessenger();
VkDebugUtilsMessengerCreateInfoEXT PopulateDebugMessengerCreateInfo();
void CreateSurface();
void PickPhysicalDevice();
void CreateLogicalDeviceAndQueues();
void CreateAllocator();
void DestroyAllocator();
void CreateSwapchain();
void CreateCommandPool();
VkPipeline GetOrCreatePipeline(
GLenum mode,
const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj,
ProgramFactory::CompileOptionFlags transformFlags,
const MG_State::GLState::VertexArrayObject& vao,
const RenderPassEntry& renderPassEntry);
VkPipeline GetOrCreateComputePipeline(const ProgramFactory::VkProgramObject& programObj);
void DestroyComputePipelines();
Bool PrepareStorageImageTextures(
VkCommandBuffer commandBuffer,
const MG_State::GLState::ProgramObject& program,
const ProgramFactory::VkProgramObject& programObj);
Bool UploadAndBindVertexBuffers(VkCommandBuffer commandBuffer, const MG_State::GLState::VertexArrayObject& vao,
const ProgramFactory::VkProgramObject& programObj,
const DrawCmdParam& drawParams,
const IndexBufferView* pIndexBufferView);
Bool UploadAndBindIndexBuffer(FrameContext::FrameData& frame,
const MG_State::GLState::VertexArrayObject& vao,
const IndexBufferView* pIndexBufferView = nullptr);
Bool InitializeBlitResources();
Bool InitializeDepthMipmapResources();
void ShutdownBlitResources();
void ShutdownDepthMipmapResources();
void CollectDeferredDepthMipmapCleanup(Uint32 frameIndex);
void DestroyDeferredDepthMipmapCleanup();
Bool TryBlitToDefaultFramebufferWithShader(FrameContext::FrameData& frame,
MG_State::GLState::FramebufferObject& readFbo,
MG_State::GLState::FramebufferObject& drawFbo,
GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1,
GLint dstX0, GLint dstY0, GLint dstX1, GLint dstY1,
GLenum filter);
Bool MaterializePendingClearForTexture(VkCommandBuffer commandBuffer,
MG_State::GLState::ITextureObject& texture);
Bool MaterializePendingClearForRenderbuffer(
VkCommandBuffer commandBuffer,
const SharedPtr<MG_State::GLState::RenderbufferObject>& renderbuffer);
VkPipeline GetOrCreateBlitPipeline(const RenderPassEntry& renderPassEntry);
Bool GenerateDepthMipmapWithShader(FrameContext::FrameData& frame,
MG_State::GLState::ITextureObject& texture,
VkTextureManager::TextureResource& resource,
Uint32 baseMipLevel,
Uint32 generateMipLevelCount,
const IntVec3& storageBaseTexelSize,
VkImageLayout originalLayout,
VkImageLayout finalLayout);
Bool SubmitReadbackCommandsAndWait(FrameContext::FrameData& frame);
void ShutdownSwapchain();
// Static functions
static Int GetPresentQueueFamilyIndex(const PhysicalDevice& physicalDevice, VkSurfaceKHR surface,
const Vector<VkQueueFamilyProperties>& queueFamilies,
Int preferredFamilyIndex = -1);
static Vector<VkQueueFamilyProperties> GetQueueFamilyFromPhysicalDevice(VkPhysicalDevice device);
static Int GetQueueFamilyIndex(const Vector<VkQueueFamilyProperties>& queueFamilies, VkQueueFlagBits flag);
static Vector<VkExtensionProperties> EnumerateInstanceExtensions();
static Vector<VkExtensionProperties> EnumerateDeviceExtensions(VkPhysicalDevice device);
static Bool IsExtensionSupported(const Vector<VkExtensionProperties>& availableExtensions,
const char* extensionName);
static Bool IsExtensionAlreadyEnabled(const Vector<const char*>& enabledExtensions, const char* extensionName);
static Bool EnableOptionalDeviceExtension(const Vector<VkExtensionProperties>& availableExtensions,
Vector<const char*>& inOutEnabledExtensions,
const char* extensionName);
void ResolveOptionalDeviceExtensions(const Vector<VkExtensionProperties>& availableExtensions,
Vector<const char*>& inOutEnabledExtensions);
static Bool IsNecessaryDeviceExtensionSupported(VkPhysicalDevice device);
static Bool GetMoreCapablePhysicalDevice(VkPhysicalDevice newVkDevice, VkSurfaceKHR surface,
const PhysicalDevice& compareWithDevice,
PhysicalDevice& outBetterDevice);
static constexpr const char* s_validationLayerNames[] = {"VK_LAYER_KHRONOS_validation"};
static constexpr const char* s_deviceExtensionNames[] = {VK_KHR_SWAPCHAIN_EXTENSION_NAME};
static Bool CheckValidationLayerSupport();
static VKAPI_ATTR VkBool32 VKAPI_CALL DebugCallback(VkDebugUtilsMessageSeverityFlagBitsEXT messageSeverity,
VkDebugUtilsMessageTypeFlagsEXT messageType,
const VkDebugUtilsMessengerCallbackDataEXT* pCallbackData,
void* pUserData);
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -1,70 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/VkIncludes.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 "VulkanRendererConfig.h"
#define ENUM_STR_CASE(c) case c: return #c;
namespace MobileGL::MG_Backend::DirectVulkan {
inline const char* VkResultToString(VkResult result) {
switch (result) {
ENUM_STR_CASE(VK_SUCCESS)
ENUM_STR_CASE(VK_NOT_READY)
ENUM_STR_CASE(VK_TIMEOUT)
ENUM_STR_CASE(VK_EVENT_SET)
ENUM_STR_CASE(VK_EVENT_RESET)
ENUM_STR_CASE(VK_INCOMPLETE)
ENUM_STR_CASE(VK_ERROR_OUT_OF_HOST_MEMORY)
ENUM_STR_CASE(VK_ERROR_OUT_OF_DEVICE_MEMORY)
ENUM_STR_CASE(VK_ERROR_INITIALIZATION_FAILED)
ENUM_STR_CASE(VK_ERROR_DEVICE_LOST)
ENUM_STR_CASE(VK_ERROR_MEMORY_MAP_FAILED)
ENUM_STR_CASE(VK_ERROR_LAYER_NOT_PRESENT)
ENUM_STR_CASE(VK_ERROR_EXTENSION_NOT_PRESENT)
ENUM_STR_CASE(VK_ERROR_FEATURE_NOT_PRESENT)
ENUM_STR_CASE(VK_ERROR_INCOMPATIBLE_DRIVER)
ENUM_STR_CASE(VK_ERROR_TOO_MANY_OBJECTS)
ENUM_STR_CASE(VK_ERROR_FORMAT_NOT_SUPPORTED)
ENUM_STR_CASE(VK_ERROR_FRAGMENTED_POOL)
ENUM_STR_CASE(VK_ERROR_UNKNOWN)
ENUM_STR_CASE(VK_ERROR_OUT_OF_POOL_MEMORY)
ENUM_STR_CASE(VK_ERROR_INVALID_EXTERNAL_HANDLE)
ENUM_STR_CASE(VK_ERROR_FRAGMENTATION)
ENUM_STR_CASE(VK_ERROR_INVALID_OPAQUE_CAPTURE_ADDRESS)
ENUM_STR_CASE(VK_PIPELINE_COMPILE_REQUIRED)
ENUM_STR_CASE(VK_ERROR_SURFACE_LOST_KHR)
ENUM_STR_CASE(VK_ERROR_NATIVE_WINDOW_IN_USE_KHR)
ENUM_STR_CASE(VK_SUBOPTIMAL_KHR)
ENUM_STR_CASE(VK_ERROR_OUT_OF_DATE_KHR)
ENUM_STR_CASE(VK_ERROR_INCOMPATIBLE_DISPLAY_KHR)
ENUM_STR_CASE(VK_ERROR_VALIDATION_FAILED_EXT)
ENUM_STR_CASE(VK_ERROR_INVALID_SHADER_NV)
default:
return "VK_RESULT_UNKNOWN";
}
}
} // namespace MobileGL::MG_Backend::DirectVulkan
#define VK_VERIFY(expr, ...) \
do { \
VkResult _vk_verify_result = (expr); \
if (_vk_verify_result != VK_SUCCESS) { \
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" __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); \
MOBILEGL_ASSERT(_xxh_verify_result == XXH_OK, "XXHash error %d at %s:%d" __VA_OPT__(" - ") __VA_ARGS__, _xxh_verify_result, __FILE__, __LINE__); \
} while (0)
@@ -1,32 +0,0 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkan/Renderer/VulkanRendererConfig.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 "Config.h"
namespace MobileGL::MG_Backend::DirectVulkan {
struct VulkanRendererConfig {
// Fallback CPU pipeline depth used when the MOBILEGL_MAGMA_FRAMESINFLIGHT env var is
// unset/invalid. A deeper pipeline lets the CPU run further ahead of the GPU, hiding
// per-frame GPU-completion latency. Whatever value is chosen (env or this fallback) is
// only a request: VulkanRenderer::Initialize clamps it down to the surface's maxImageCount
// (and never below 2), since not every driver allows that many swapchain images.
Uint32 MaxFramesInFlight = 3;
String AppName = "MobileGL-VulkanRenderer";
MobileGL::Version Version = MG_Config::CoreVersion;
Uint64 CacheVersion = MG_Config::CacheVersion;
Uint32 SurfaceWidth = 1;
Uint32 SurfaceHeight = 1;
Bool DisablePipelineCache = false;
#if MOBILEGL_LOG_ACTIVE_LEVEL <= MOBILEGL_LOG_LEVEL_DEBUG
Bool EnableValidationLayers = true;
#else
Bool EnableValidationLayers = false;
#endif
};
} // namespace MobileGL::MG_Backend::DirectVulkan
@@ -0,0 +1,119 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/BackendObject_DirectVulkanTMP.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 "BackendObject_DirectVulkanTMP.h"
#include "MG_Backend/BackendObject.h"
#include "DirectVulkanTMP.h"
#include "TmpImpl.h"
#include <MG_Util/BackendLoaders/OpenGL/Loader.h>
namespace MobileGL::MG_Backend::DirectVulkanTMP {
BackendObject_DirectVulkanTMP::~BackendObject_DirectVulkanTMP() = default;
void BackendObject_DirectVulkanTMP::InitWindowSurface() {
auto nativeWindow = reinterpret_cast<NativeWindowType>(m_windowHandle.Handle);
if (!DirectVulkanTMP::InitWindowSurface(nativeWindow)) {
MGLOG_E("Failed to initialize window surface for DirectVulkanTMP backend");
}
}
void BackendObject_DirectVulkanTMP::Initialize() {
m_initialized = true;
}
void BackendObject_DirectVulkanTMP::InitCapabilities() {
if (!m_initialized) {
MGLOG_E("Cannot initialize capabilities before backend is initialized");
return;
}
MG_Util::BackendLoader::QueryVulkanCapabilities(m_vulkanCaps,
DirectVulkanTMP::GetVulkanState().ctx->GetPhysicalDevice());
UpdateDynamicBackendParameters();
}
const RendererInfo& BackendObject_DirectVulkanTMP::GetRendererInfo() const {
static RendererInfo RendererInfo = {
.RendererName = "Magma-TMP", // Renderer Name
.BackendName = "Direct (Vulkan) TMP", // Backend Name
.ExtraVendor = Nullopt, // Extra vendor
.RendererGLInfo =
{
.TargetGLVersion = {3, 3, 0}, // Target OpenGL Version
.TargetGLSLVersion = {4, 6, 0}, // Target Shading Language Version
.Extensions = {V_OpenGL30, V_OpenGL31, V_OpenGL32, // OpenGL Extensions
V_OpenGL33},
.IsCompatibilityProfile = false // Is Compatibility Profile
},
.StaticBackendCapability = {.AllowVSOnlyPrograms = false} // Backend Capability
};
return RendererInfo;
}
String BackendObject_DirectVulkanTMP::GetBackendAPIVersionString() const {
if (!m_initialized) {
return "<uninitialized DirectVulkanTMP backend>";
}
// Format:
// <GPU Name>, Vulkan <Vulkan Version>, Driver <Driver Version>
// TODO
String str = m_vulkanCaps.DeviceName + ", Vulkan " + m_vulkanCaps.VulkanAPIVersion.toString() + ", Driver " +
m_vulkanCaps.DriverVersionString;
return str;
}
BackendType BackendObject_DirectVulkanTMP::GetBackendType() const {
return BackendType::DirectVulkanTMP;
}
const GlobalBackendFunctionsTable& BackendObject_DirectVulkanTMP::GetBackendFunctions() const {
static GlobalBackendFunctionsTable funcsTable;
static Bool funcsTableInitialized = false;
if (!funcsTableInitialized) {
funcsTable.Present = DirectVulkanTMP::Present;
funcsTable.GL.DrawArrays = DrawArrays;
funcsTable.GL.DrawElements = DrawElements;
funcsTable.GL.DrawElementsBaseVertex = DrawElementsBaseVertex;
funcsTable.GL.MultiDrawElements = MultiDrawElements;
funcsTable.GL.MultiDrawElementsBaseVertex = MultiDrawElementsBaseVertex;
funcsTable.GL.MultiDrawElementsIndirect = MultiDrawElementsIndirect;
funcsTable.GL.MultiDrawArraysIndirect = MultiDrawArraysIndirect;
funcsTable.GL.DrawRangeElementsBaseVertex = DrawRangeElementsBaseVertex;
funcsTable.GL.DrawRangeElements = DrawRangeElements;
funcsTable.GL.DrawElementsInstancedBaseVertexBaseInstance = DrawElementsInstancedBaseVertexBaseInstance;
funcsTable.GL.DrawElementsInstancedBaseVertex = DrawElementsInstancedBaseVertex;
funcsTable.GL.DrawElementsInstancedBaseInstance = DrawElementsInstancedBaseInstance;
funcsTable.GL.DrawElementsInstanced = DrawElementsInstanced;
funcsTable.GL.DrawArraysInstancedBaseInstance = DrawArraysInstancedBaseInstance;
funcsTable.GL.DrawArraysInstanced = DrawArraysInstanced;
funcsTable.GL.DrawElementsIndirect = DrawElementsIndirect;
funcsTable.GL.DrawArraysIndirect = DrawArraysIndirect;
funcsTable.GL.Clear = Clear;
funcsTable.GL.ClearBufferfi = ClearBufferfi;
funcsTable.GL.ClearBufferfv = ClearBufferfv;
funcsTable.GL.ClearBufferuiv = ClearBufferuiv;
funcsTable.GL.ClearBufferiv = ClearBufferiv;
funcsTable.GL.BlitFramebuffer = BlitFramebuffer;
funcsTable.GL.CopyTexImage2D = CopyTexImage2D;
funcsTable.GL.CopyTexSubImage2D = CopyTexSubImage2D;
funcsTable.GL.GenerateMipmap = GenerateMipmap;
funcsTable.GL.ReadPixels = ReadPixels;
funcsTable.GL.GetTexImage = GetTexImage;
funcsTableInitialized = true;
}
return funcsTable;
}
const DynamicBackendParameters& BackendObject_DirectVulkanTMP::GetDynamicParameters() const {
return m_dynamicParameters;
}
void BackendObject_DirectVulkanTMP::UpdateDynamicBackendParameters() {
m_dynamicParameters.UniformBufferOffsetAlignment = m_vulkanCaps.UniformBufferOffsetAlignment;
}
} // namespace MobileGL::MG_Backend::DirectVulkanTMP
@@ -0,0 +1,36 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/BackendObject_DirectVulkanTMP.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 <Includes.h>
#include "../BackendObject.h"
#include <MG_Util/BackendLoaders/Vulkan/Loader.h>
namespace MobileGL::MG_Backend::DirectVulkanTMP {
class BackendObject_DirectVulkanTMP : public BackendObject {
public:
~BackendObject_DirectVulkanTMP() override;
void Initialize() override;
void InitWindowSurface() override;
void InitCapabilities() override;
const RendererInfo& GetRendererInfo() const override;
String GetBackendAPIVersionString() const override;
const GlobalBackendFunctionsTable& GetBackendFunctions() const override;
const DynamicBackendParameters& GetDynamicParameters() const override;
BackendType GetBackendType() const override;
private:
void UpdateDynamicBackendParameters();
Bool m_initialized = false;
DynamicBackendParameters m_dynamicParameters;
MG_External::VulkanCapabilities m_vulkanCaps;
};
} // namespace MobileGL::MG_Backend::DirectVulkanTMP
@@ -0,0 +1,74 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/DirectVulkanTMP.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 "DirectVulkanTMP.h"
#include "TmpImpl.h"
namespace MobileGL::MG_Backend::DirectVulkanTMP {
void Clear(GLbitfield mask) {
MobileGL::MG_Backend::DirectVulkanTMP::TmpImpl::Clear(mask);
}
void DrawElements(GLenum mode, GLsizei count, GLenum type, const void* indices) {
MobileGL::MG_Backend::DirectVulkanTMP::TmpImpl::DrawElements(mode, count, type, indices);
}
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) {}
void ClearBufferiv(GLenum buffer, GLint drawbuffer, const GLint* value) {}
void DrawArrays(GLenum mode, GLint first, GLsizei count) {}
void DrawElementsBaseVertex(GLenum mode, GLsizei count, GLenum type, const GLvoid* indices, GLint basevertex) {}
void MultiDrawElements(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount) {}
void MultiDrawElementsBaseVertex(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount, const GLint* basevertex) {}
void MultiDrawElementsIndirect(GLenum mode, GLenum type, const void* indirect, GLsizei drawcount, GLsizei stride) {}
void MultiDrawArraysIndirect(GLenum mode, const void* indirect, GLsizei drawcount, GLsizei stride) {}
void DrawRangeElementsBaseVertex(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type,
const void* indices, GLint basevertex) {}
void DrawRangeElements(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type, const void* indices) {}
void DrawElementsInstancedBaseVertexBaseInstance(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLint basevertex, GLuint baseinstance) {}
void DrawElementsInstancedBaseVertex(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLint basevertex) {}
void DrawElementsInstancedBaseInstance(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLuint baseinstance) {}
void DrawElementsInstanced(GLenum mode, GLsizei count, GLenum type, const void* indices, GLsizei instancecount) {}
void DrawElementsIndirect(GLenum mode, GLenum type, const void* indirect) {}
void DrawArraysInstancedBaseInstance(GLenum mode, GLint first, GLsizei count, GLsizei instancecount,
GLuint baseinstance) {}
void DrawArraysInstanced(GLenum mode, GLint first, GLsizei count, GLsizei instancecount) {}
void DrawArraysIndirect(GLenum mode, const void* indirect) {}
void BlitFramebuffer(GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1, GLint dstX0, GLint dstY0, GLint dstX1,
GLint dstY1, GLbitfield mask, GLenum filter) {}
void CopyTexImage2D(GLenum target, GLint level, GLenum internalformat, GLint x, GLint y, GLsizei width,
GLsizei height, GLint border) {}
void CopyTexSubImage2D(GLenum target, GLint level, GLint xoffset, GLint yoffset, GLint x, GLint y, GLsizei width,
GLsizei height) {}
void GenerateMipmap(GLenum target) {}
void ReadPixels(GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, void* pixels) {}
void GetTexImage(GLenum target, GLint level, GLenum format, GLenum type, GLvoid* pixels) {}
Bool InitWindowSurface(NativeWindowType window) {
if (!window) {
MGLOG_E("Cannot initialize Vulkan window surface: invalid window handle");
return false;
}
TmpImpl::InitVulkan(window);
return true;
}
void Present() {
TmpImpl::Present();
}
const TmpImpl::VulkanState& GetVulkanState() {
return TmpImpl::GetVulkanState();
}
} // namespace MobileGL::MG_Backend::DirectVulkanTMP
@@ -0,0 +1,58 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/DirectVulkanTMP.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 <Includes.h>
namespace MobileGL::MG_Backend::DirectVulkanTMP {
namespace TmpImpl {
class VulkanState;
} // namespace TmpImpl
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);
void ClearBufferiv(GLenum buffer, GLint drawbuffer, const GLint* value);
void Clear(GLbitfield mask);
void DrawElements(GLenum mode, GLsizei count, GLenum type, const void* indices);
void DrawArrays(GLenum mode, GLint first, GLsizei count);
void DrawElementsBaseVertex(GLenum mode, GLsizei count, GLenum type, const GLvoid* indices, GLint basevertex);
void MultiDrawElements(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount);
void MultiDrawElementsBaseVertex(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount, const GLint* basevertex);
void MultiDrawElementsIndirect(GLenum mode, GLenum type, const void* indirect, GLsizei drawcount, GLsizei stride);
void MultiDrawArraysIndirect(GLenum mode, const void* indirect, GLsizei drawcount, GLsizei stride);
void DrawRangeElementsBaseVertex(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type,
const void* indices, GLint basevertex);
void DrawRangeElements(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type, const void* indices);
void DrawElementsInstancedBaseVertexBaseInstance(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLint basevertex, GLuint baseinstance);
void DrawElementsInstancedBaseVertex(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLint basevertex);
void DrawElementsInstancedBaseInstance(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLuint baseinstance);
void DrawElementsInstanced(GLenum mode, GLsizei count, GLenum type, const void* indices, GLsizei instancecount);
void DrawElementsIndirect(GLenum mode, GLenum type, const void* indirect);
void DrawArraysInstancedBaseInstance(GLenum mode, GLint first, GLsizei count, GLsizei instancecount,
GLuint baseinstance);
void DrawArraysInstanced(GLenum mode, GLint first, GLsizei count, GLsizei instancecount);
void DrawArraysIndirect(GLenum mode, const void* indirect);
void BlitFramebuffer(GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1, GLint dstX0, GLint dstY0, GLint dstX1,
GLint dstY1, GLbitfield mask, GLenum filter);
void CopyTexImage2D(GLenum target, GLint level, GLenum internalformat, GLint x, GLint y, GLsizei width,
GLsizei height, GLint border);
void CopyTexSubImage2D(GLenum target, GLint level, GLint xoffset, GLint yoffset, GLint x, GLint y, GLsizei width,
GLsizei height);
void GenerateMipmap(GLenum target);
void ReadPixels(GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, void* pixels);
void GetTexImage(GLenum target, GLint level, GLenum format, GLenum type, GLvoid* pixels);
Bool InitWindowSurface(NativeWindowType window);
void Present();
const TmpImpl::VulkanState& GetVulkanState();
} // namespace MobileGL::MG_Backend::DirectVulkanTMP
@@ -0,0 +1,447 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/Managers/ProgramManager.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 "ProgramManager.h"
#include "MG_Util/Debug/Log.h"
#include "spirv-tools/libspirv.h"
#include "spirv-tools/optimizer.hpp"
#include "source/opt/constants.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/opt/pass.h"
#include "source/opt/type_manager.h"
#include <bit>
namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager {
namespace {
using ProgramObject = MG_State::GLState::ProgramObject;
using ShaderObject = MG_State::GLState::ShaderObject;
struct PositionTargetInfo {
uint32_t variableId = 0;
uint32_t vectorTypeId = 0;
uint32_t floatTypeId = 0;
uint32_t vectorPtrTypeId = 0;
uint32_t memberIndex = 0;
bool isMember = false;
};
bool IsVec4Float32(spvtools::opt::IRContext* context, uint32_t typeId, uint32_t* outFloatTypeId) {
auto* vecInst = context->get_def_use_mgr()->GetDef(typeId);
if (!vecInst || vecInst->opcode() != spv::Op::OpTypeVector) return false;
if (vecInst->GetSingleWordInOperand(1) != 4) return false;
const uint32_t floatTypeId = vecInst->GetSingleWordInOperand(0);
auto* floatInst = context->get_def_use_mgr()->GetDef(floatTypeId);
if (!floatInst || floatInst->opcode() != spv::Op::OpTypeFloat) return false;
if (floatInst->GetSingleWordInOperand(0) != 32) return false;
if (outFloatTypeId) *outFloatTypeId = floatTypeId;
return true;
}
bool ResolveDirectPositionTarget(spvtools::opt::IRContext* context, uint32_t variableId,
PositionTargetInfo* outTarget) {
auto* varInst = context->get_def_use_mgr()->GetDef(variableId);
if (!varInst || varInst->opcode() != spv::Op::OpVariable) return false;
if (varInst->GetSingleWordInOperand(0) != static_cast<uint32_t>(spv::StorageClass::Output)) return false;
auto* ptrTypeInst = context->get_def_use_mgr()->GetDef(varInst->type_id());
if (!ptrTypeInst || ptrTypeInst->opcode() != spv::Op::OpTypePointer) return false;
if (ptrTypeInst->GetSingleWordInOperand(0) != static_cast<uint32_t>(spv::StorageClass::Output))
return false;
PositionTargetInfo target{};
target.variableId = variableId;
target.vectorTypeId = ptrTypeInst->GetSingleWordInOperand(1);
if (!IsVec4Float32(context, target.vectorTypeId, &target.floatTypeId)) return false;
target.vectorPtrTypeId = varInst->type_id();
target.isMember = false;
*outTarget = target;
return true;
}
uint32_t FindOutputVectorPointerTypeId(spvtools::opt::IRContext* context, uint32_t vectorTypeId) {
auto* vectorType = context->get_type_mgr()->GetType(vectorTypeId);
if (!vectorType) return 0;
spvtools::opt::analysis::Pointer ptrType(vectorType, spv::StorageClass::Output);
return context->get_type_mgr()->GetTypeInstruction(&ptrType);
}
bool ResolveMemberPositionTarget(spvtools::opt::IRContext* context, uint32_t structTypeId, uint32_t memberIndex,
PositionTargetInfo* outTarget) {
auto* structInst = context->get_def_use_mgr()->GetDef(structTypeId);
if (!structInst || structInst->opcode() != spv::Op::OpTypeStruct) return false;
if (memberIndex >= structInst->NumInOperands()) return false;
const uint32_t vectorTypeId = structInst->GetSingleWordInOperand(memberIndex);
uint32_t floatTypeId = 0;
if (!IsVec4Float32(context, vectorTypeId, &floatTypeId)) return false;
const uint32_t vectorPtrTypeId = FindOutputVectorPointerTypeId(context, vectorTypeId);
if (vectorPtrTypeId == 0) return false;
for (auto& inst : context->module()->types_values()) {
if (inst.opcode() != spv::Op::OpVariable) continue;
if (inst.GetSingleWordInOperand(0) != static_cast<uint32_t>(spv::StorageClass::Output)) continue;
auto* ptrTypeInst = context->get_def_use_mgr()->GetDef(inst.type_id());
if (!ptrTypeInst || ptrTypeInst->opcode() != spv::Op::OpTypePointer) continue;
if (ptrTypeInst->GetSingleWordInOperand(0) != static_cast<uint32_t>(spv::StorageClass::Output))
continue;
if (ptrTypeInst->GetSingleWordInOperand(1) != structTypeId) continue;
PositionTargetInfo target{};
target.variableId = inst.result_id();
target.vectorTypeId = vectorTypeId;
target.floatTypeId = floatTypeId;
target.vectorPtrTypeId = vectorPtrTypeId;
target.memberIndex = memberIndex;
target.isMember = true;
*outTarget = target;
return true;
}
return false;
}
bool FindPositionTarget(spvtools::opt::IRContext* context, PositionTargetInfo* outTarget) {
Vector<Pair<uint32_t, uint32_t>> memberCandidates;
constexpr uint32_t kDecorationBuiltIn = static_cast<uint32_t>(spv::Decoration::BuiltIn);
constexpr uint32_t kBuiltInPosition = static_cast<uint32_t>(spv::BuiltIn::Position);
for (auto& inst : context->module()->annotations()) {
if (inst.opcode() == spv::Op::OpDecorate) {
if (inst.NumInOperands() < 3) continue;
if (inst.GetSingleWordInOperand(1) != kDecorationBuiltIn) continue;
if (inst.GetSingleWordInOperand(2) != kBuiltInPosition) continue;
if (ResolveDirectPositionTarget(context, inst.GetSingleWordInOperand(0), outTarget)) return true;
} else if (inst.opcode() == spv::Op::OpMemberDecorate) {
if (inst.NumInOperands() < 4) continue;
if (inst.GetSingleWordInOperand(2) != kDecorationBuiltIn) continue;
if (inst.GetSingleWordInOperand(3) != kBuiltInPosition) continue;
memberCandidates.emplace_back(inst.GetSingleWordInOperand(0), inst.GetSingleWordInOperand(1));
}
}
for (const auto& [structTypeId, memberIndex] : memberCandidates) {
if (ResolveMemberPositionTarget(context, structTypeId, memberIndex, outTarget)) return true;
}
return false;
}
bool InsertPositionFixup(spvtools::opt::IRContext* context, spvtools::opt::Instruction* insertBefore,
const PositionTargetInfo& target, uint32_t halfConstId, bool doYFlip, bool doZRemap) {
using namespace spvtools::opt;
InstructionBuilder builder(context, insertBefore,
IRContext::kAnalysisDefUse | IRContext::kAnalysisInstrToBlockMapping);
uint32_t positionPtrId = target.variableId;
if (target.isMember) {
const uint32_t memberIndexId = builder.GetUintConstantId(target.memberIndex);
if (memberIndexId == 0) return false;
auto* access = builder.AddAccessChain(target.vectorPtrTypeId, target.variableId, {memberIndexId});
if (!access) return false;
positionPtrId = access->result_id();
}
auto* position = builder.AddLoad(target.vectorTypeId, positionPtrId);
if (!position) return false;
auto* x = builder.AddCompositeExtract(target.floatTypeId, position->result_id(), {0});
auto* y = builder.AddCompositeExtract(target.floatTypeId, position->result_id(), {1});
auto* z = builder.AddCompositeExtract(target.floatTypeId, position->result_id(), {2});
auto* w = builder.AddCompositeExtract(target.floatTypeId, position->result_id(), {3});
if (!x || !y || !z || !w) return false;
if (!doYFlip && !doZRemap) return false;
uint32_t yValueId = y->result_id();
if (doYFlip) {
auto* negY = builder.AddUnaryOp(target.floatTypeId, spv::Op::OpFNegate, y->result_id());
if (!negY) return false;
yValueId = negY->result_id();
}
uint32_t zValueId = z->result_id();
if (doZRemap) {
auto* zPlusW = builder.AddBinaryOp(target.floatTypeId, spv::Op::OpFAdd, z->result_id(), w->result_id());
if (!zPlusW) return false;
auto* mappedZ =
builder.AddBinaryOp(target.floatTypeId, spv::Op::OpFMul, zPlusW->result_id(), halfConstId);
if (!mappedZ) return false;
zValueId = mappedZ->result_id();
}
auto* fixedPosition = builder.AddCompositeConstruct(target.vectorTypeId,
{x->result_id(), yValueId, zValueId, w->result_id()});
if (!fixedPosition) return false;
return builder.AddStore(positionPtrId, fixedPosition->result_id()) != nullptr;
}
class GlToVulkanPositionFixPass final : public spvtools::opt::Pass {
public:
const char* name() const override { return "gl-to-vulkan-position-fix"; }
explicit GlToVulkanPositionFixPass(ShaderTransformFlags transformFlags)
: m_transformFlags(transformFlags) {}
Status Process() override {
if (!m_transformFlags) return Status::SuccessWithoutChange;
PositionTargetInfo target{};
if (!FindPositionTarget(context(), &target)) return Status::SuccessWithoutChange;
auto* floatType = context()->get_type_mgr()->GetType(target.floatTypeId);
if (!floatType) return Status::SuccessWithoutChange;
const uint32_t halfBits = std::bit_cast<uint32_t>(0.5f);
const auto* halfConst = context()->get_constant_mgr()->GetConstant(floatType, {halfBits});
auto* halfInst = context()->get_constant_mgr()->GetDefiningInstruction(halfConst);
if (!halfInst) return Status::SuccessWithoutChange;
const uint32_t halfConstId = halfInst->result_id();
const bool doYFlip = (m_transformFlags & ShaderTransformBit::PositionYFlip);
const bool doZRemap = (m_transformFlags & ShaderTransformBit::PositionZRemap);
bool modified = false;
for (auto& entryPoint : get_module()->entry_points()) {
if (entryPoint.opcode() != spv::Op::OpEntryPoint) continue;
if (entryPoint.NumInOperands() < 2) continue;
const auto model = static_cast<spv::ExecutionModel>(entryPoint.GetSingleWordInOperand(0));
if (model != spv::ExecutionModel::Vertex && model != spv::ExecutionModel::TessellationEvaluation &&
model != spv::ExecutionModel::Geometry) {
continue;
}
auto* function = context()->GetFunction(entryPoint.GetSingleWordInOperand(1));
if (!function) continue;
for (auto& bb : *function) {
for (auto instIter = bb.begin(); instIter != bb.end(); ++instIter) {
auto* inst = &*instIter;
const bool needsFixup =
(model == spv::ExecutionModel::Geometry && inst->opcode() == spv::Op::OpEmitVertex) ||
(model != spv::ExecutionModel::Geometry && inst->opcode() == spv::Op::OpReturn);
if (!needsFixup) continue;
modified |= InsertPositionFixup(context(), inst, target, halfConstId, doYFlip, doZRemap);
}
}
}
if (!modified) return Status::SuccessWithoutChange;
context()->InvalidateAnalysesExceptFor(spvtools::opt::IRContext::kAnalysisDefUse |
spvtools::opt::IRContext::kAnalysisInstrToBlockMapping);
return Status::SuccessWithChange;
}
private:
ShaderTransformFlags m_transformFlags;
};
spvtools::Optimizer::PassToken CreateGlToVulkanPositionFixPass(ShaderTransformFlags transformFlags) {
return spvtools::Optimizer::PassToken(MakeUnique<GlToVulkanPositionFixPass>(transformFlags));
}
bool TransformSpirvForVulkanPositionFix(const Vector<Uint32>& input, Vector<Uint32>& output,
ShaderTransformFlags transformFlags) {
if (input.empty()) {
output.clear();
return true;
}
if (!transformFlags) {
output = input;
return true;
}
spvtools::Optimizer optimizer(SPV_ENV_VULKAN_1_3);
spvtools::OptimizerOptions options;
options.set_run_validator(false);
optimizer.RegisterPass(CreateGlToVulkanPositionFixPass(transformFlags));
const bool success = optimizer.Run(input.data(), input.size(), &output, options);
if (!success) {
MGLOG_E("Vulkan: failed to run GL->Vulkan position fix pass");
output = input;
}
return success;
}
ShaderStage PickClipFixupStage(const Vector<SharedPtr<ShaderObject>>& shaders) {
bool hasGeometry = false;
bool hasTessEval = false;
bool hasVertex = false;
for (const auto& shader : shaders) {
if (!shader) continue;
const auto stage = shader->GetShaderStage();
hasGeometry |= (stage == ShaderStage::Geometry);
hasTessEval |= (stage == ShaderStage::TessEval);
hasVertex |= (stage == ShaderStage::Vertex);
}
if (hasGeometry) return ShaderStage::Geometry;
if (hasTessEval) return ShaderStage::TessEval;
if (hasVertex) return ShaderStage::Vertex;
return ShaderStage::Unknown;
}
} // namespace
ProgramManager::~ProgramManager() {
for (auto& [_, stages] : m_cache) {
DestroyStages(stages);
}
m_cache.clear();
}
ProgramManager::HashType ProgramManager::ComputeSourceSpvHash(MG_State::GLState::ProgramObject* program) const {
if (!program) return 0;
XXH64_state_t* state = XXH64_createState();
XXH64_reset(state, 0xC0FFEEu);
auto& spirvs = program->GetGeneratedSpirv();
for (const auto& spv : spirvs) {
if (spv.empty()) continue;
XXH64_update(state, spv.data(), spv.size() * sizeof(Uint));
}
HashType hash = XXH64_digest(state);
XXH64_freeState(state);
return hash;
}
ProgramManager::HashType ProgramManager::ComputeSpvHash(const Vector<Vector<Uint32>>& spirvs) const {
XXH64_state_t* state = XXH64_createState();
XXH64_reset(state, 0xC0FFEEu);
for (const auto& spv : spirvs) {
if (spv.empty()) continue;
XXH64_update(state, spv.data(), spv.size() * sizeof(Uint32));
}
HashType hash = XXH64_digest(state);
XXH64_freeState(state);
return hash;
}
void ProgramManager::BuildPipelineSpirvModules(MG_State::GLState::ProgramObject* program,
Vector<Vector<Uint32>>& outSpirvs,
ShaderTransformFlags transformFlags) const {
outSpirvs.clear();
if (!program) return;
auto& spirvs = program->GetGeneratedSpirv();
auto& shaders = program->GetAttachedShaders();
const ShaderStage fixupStage = PickClipFixupStage(shaders);
outSpirvs.reserve(spirvs.size());
for (SizeT i = 0; i < spirvs.size(); ++i) {
Vector<Uint32> module = spirvs[i];
ShaderStage stage = ShaderStage::Unknown;
if (i < shaders.size() && shaders[i]) stage = shaders[i]->GetShaderStage();
if (!module.empty() && fixupStage != ShaderStage::Unknown && stage == fixupStage) {
Vector<Uint32> transformed;
TransformSpirvForVulkanPositionFix(module, transformed, transformFlags);
module = Move(transformed);
}
outSpirvs.push_back(Move(module));
}
}
ProgramManager::HashType ProgramManager::ComputeSpvHash(MG_State::GLState::ProgramObject* program,
ShaderTransformFlags transformFlags) const {
Vector<Vector<Uint32>> spirvs;
BuildPipelineSpirvModules(program, spirvs, transformFlags);
return ComputeSpvHash(spirvs);
}
ProgramManager::HashType ProgramManager::ComputeProgramHash(MG_State::GLState::ProgramObject* program,
ShaderTransformFlags transformFlags) const {
if (!program) return 0;
const HashType sourceHash = ComputeSourceSpvHash(program);
auto it = m_cache.find(program);
if (it != m_cache.end() && it->second.sourceHash == sourceHash) return it->second.hash;
return ComputeSpvHash(program, transformFlags);
}
VkShaderStageFlagBits ProgramManager::ToVkStage(ShaderStage stage) const {
switch (stage) {
case ShaderStage::Vertex:
return VK_SHADER_STAGE_VERTEX_BIT;
case ShaderStage::Fragment:
return VK_SHADER_STAGE_FRAGMENT_BIT;
case ShaderStage::Geometry:
return VK_SHADER_STAGE_GEOMETRY_BIT;
case ShaderStage::TessControl:
return VK_SHADER_STAGE_TESSELLATION_CONTROL_BIT;
case ShaderStage::TessEval:
return VK_SHADER_STAGE_TESSELLATION_EVALUATION_BIT;
case ShaderStage::Compute:
return VK_SHADER_STAGE_COMPUTE_BIT;
default:
return VK_SHADER_STAGE_ALL_GRAPHICS;
}
}
void ProgramManager::DestroyStages(ProgramStages& stages) {
for (auto module : stages.modules) {
if (module != VK_NULL_HANDLE) {
vkDestroyShaderModule(m_ctx.GetDevice(), module, nullptr);
}
}
stages.modules.clear();
stages.stages.clear();
stages.sourceHash = 0;
stages.hash = 0;
}
Vector<VkPipelineShaderStageCreateInfo>& ProgramManager::CreatePipelineShaderStages(
MG_State::GLState::ProgramObject* program, ShaderTransformFlags transformFlags) {
auto& entry = m_cache[program];
HashType sourceHash = ComputeSourceSpvHash(program);
if (!entry.stages.empty() && entry.sourceHash == sourceHash) return entry.stages;
DestroyStages(entry);
entry.sourceHash = sourceHash;
if (!program) return entry.stages;
Vector<Vector<Uint32>> spirvs;
BuildPipelineSpirvModules(program, spirvs, transformFlags);
entry.hash = ComputeSpvHash(spirvs);
auto& shaders = program->GetAttachedShaders();
for (SizeT i = 0; i < spirvs.size(); ++i) {
auto& spv = spirvs[i];
if (spv.empty()) continue;
VkShaderModuleCreateInfo smci{VK_STRUCTURE_TYPE_SHADER_MODULE_CREATE_INFO};
smci.codeSize = spv.size() * sizeof(Uint);
smci.pCode = spv.data();
VkShaderModule module = VK_NULL_HANDLE;
VK_VERIFY(vkCreateShaderModule(m_ctx.GetDevice(), &smci, nullptr, &module), "vkCreateShaderModule");
VkPipelineShaderStageCreateInfo stage{VK_STRUCTURE_TYPE_PIPELINE_SHADER_STAGE_CREATE_INFO};
ShaderStage shaderStage = ShaderStage::Unknown;
if (i < shaders.size() && shaders[i]) shaderStage = shaders[i]->GetShaderStage();
stage.stage = ToVkStage(shaderStage);
stage.module = module;
stage.pName = "main";
entry.modules.push_back(module);
entry.stages.push_back(stage);
}
return entry.stages;
}
} // namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager
@@ -0,0 +1,58 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/Managers/ProgramManager.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 <Includes.h>
#include "../Renderer/VulkanContext.h"
#include "../Renderer/VkCommon.h"
#include "MG_State/GLState/ProgramState/ProgramObject.h"
namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager {
enum class ShaderTransformBit : Uint {
None = 0,
PositionYFlip = 1 << 0,
PositionZRemap = 1 << 1,
};
using ShaderTransformFlags = Flags<ShaderTransformBit>;
class ProgramManager {
public:
using HashType = Uint64;
explicit ProgramManager(VulkanContext& ctx) : m_ctx(ctx) {}
~ProgramManager();
ProgramManager(const ProgramManager&) = delete;
ProgramManager& operator=(const ProgramManager&) = delete;
Vector<VkPipelineShaderStageCreateInfo>& CreatePipelineShaderStages(
MG_State::GLState::ProgramObject* program,
ShaderTransformFlags transformFlags = ShaderTransformBit::PositionZRemap);
HashType ComputeProgramHash(MG_State::GLState::ProgramObject* program,
ShaderTransformFlags transformFlags = ShaderTransformBit::PositionZRemap) const;
private:
struct ProgramStages {
HashType sourceHash = 0;
HashType hash = 0;
Vector<VkPipelineShaderStageCreateInfo> stages;
Vector<VkShaderModule> modules;
};
void DestroyStages(ProgramStages& stages);
HashType ComputeSourceSpvHash(MG_State::GLState::ProgramObject* program) const;
HashType ComputeSpvHash(MG_State::GLState::ProgramObject* program, ShaderTransformFlags transformFlags) const;
HashType ComputeSpvHash(const Vector<Vector<Uint32>>& spirvs) const;
void BuildPipelineSpirvModules(MG_State::GLState::ProgramObject* program, Vector<Vector<Uint32>>& outSpirvs,
ShaderTransformFlags transformFlags) const;
VkShaderStageFlagBits ToVkStage(ShaderStage stage) const;
VulkanContext& m_ctx;
UnorderedMap<const MG_State::GLState::ProgramObject*, ProgramStages> m_cache;
};
} // namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager
@@ -0,0 +1,49 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/Renderer/FrameContext.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 "FrameContext.h"
namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager {
void FrameContext::Initialize(VulkanContext& ctx, VkCommandPool pool) {
CommandPool = pool;
VkCommandBufferAllocateInfo abci{VK_STRUCTURE_TYPE_COMMAND_BUFFER_ALLOCATE_INFO};
abci.commandPool = pool;
abci.level = VK_COMMAND_BUFFER_LEVEL_PRIMARY;
abci.commandBufferCount = 1;
VK_VERIFY(vkAllocateCommandBuffers(ctx.GetDevice(), &abci, &CommandBuffer), "vkAllocateCommandBuffers");
VkSemaphoreCreateInfo sci{VK_STRUCTURE_TYPE_SEMAPHORE_CREATE_INFO};
VK_VERIFY(vkCreateSemaphore(ctx.GetDevice(), &sci, nullptr, &ImageAvailable), "vkCreateSemaphore");
VK_VERIFY(vkCreateSemaphore(ctx.GetDevice(), &sci, nullptr, &RenderFinished), "vkCreateSemaphore");
VkFenceCreateInfo fci{VK_STRUCTURE_TYPE_FENCE_CREATE_INFO};
fci.flags = VK_FENCE_CREATE_SIGNALED_BIT;
VK_VERIFY(vkCreateFence(ctx.GetDevice(), &fci, nullptr, &InFlightFence), "vkCreateFence");
}
void FrameContext::Cleanup(VulkanContext& ctx) {
auto device = ctx.GetDevice();
if (InFlightFence != VK_NULL_HANDLE) {
vkDestroyFence(device, InFlightFence, nullptr);
InFlightFence = VK_NULL_HANDLE;
}
if (ImageAvailable != VK_NULL_HANDLE) {
vkDestroySemaphore(device, ImageAvailable, nullptr);
ImageAvailable = VK_NULL_HANDLE;
}
if (RenderFinished != VK_NULL_HANDLE) {
vkDestroySemaphore(device, RenderFinished, nullptr);
RenderFinished = VK_NULL_HANDLE;
}
if (CommandBuffer != VK_NULL_HANDLE) {
vkFreeCommandBuffers(device, CommandPool, 1, &CommandBuffer);
CommandBuffer = VK_NULL_HANDLE;
}
CommandPool = VK_NULL_HANDLE;
}
} // namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager
@@ -0,0 +1,42 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/Renderer/FrameContext.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 <Includes.h>
#include "VulkanContext.h"
#include "VkCommon.h"
namespace MobileGL::MG_Backend::DirectVulkanTMP {
struct TrashBuffer {
VkBuffer buffer = VK_NULL_HANDLE;
VkDeviceMemory memory = VK_NULL_HANDLE;
Bool mapped = false;
};
struct TrashImage {
VkImage image = VK_NULL_HANDLE;
VkDeviceMemory memory = VK_NULL_HANDLE;
VkImageView view = VK_NULL_HANDLE;
Vector<VkImageView> mipViews;
};
namespace VkManager {
struct FrameContext {
VkCommandBuffer CommandBuffer = VK_NULL_HANDLE;
VkSemaphore ImageAvailable = VK_NULL_HANDLE;
VkSemaphore RenderFinished = VK_NULL_HANDLE;
VkFence InFlightFence = VK_NULL_HANDLE;
Uint32 CurrentImageIndex = 0;
VkCommandPool CommandPool = VK_NULL_HANDLE;
Vector<TrashBuffer> TrashBuffers;
Vector<TrashImage> TrashImages;
// VkCommandPool CommandPool = VK_NULL_HANDLE;
void Initialize(VulkanContext& ctx, VkCommandPool pool);
void Cleanup(VulkanContext& ctx);
};
} // namespace VkManager
} // namespace MobileGL::MG_Backend::DirectVulkanTMP
@@ -1,4 +1,4 @@
// MobileGL - MobileGL/MG_Impl/NSOpenGLImpl/NSOpenGLImpl.h
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/Renderer/PipelineManager.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
@@ -6,9 +6,6 @@
// SPDX-License-Identifier: LGPL-3.0-only
// End of Source File Header
#pragma once
#include <Includes.h>
namespace MobileGL::MG_Impl::NSOpenGLImpl {
void InstallHooks();
}
namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager {}
@@ -0,0 +1,166 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/Renderer/SwapchainManager.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 "SwapchainManager.h"
namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager {
namespace {
VkSurfaceFormatKHR ChooseSurfaceFormat(const Vector<VkSurfaceFormatKHR>& formats) {
for (const auto& f : formats) {
if ((f.format == VK_FORMAT_R8G8B8A8_UNORM || f.format == VK_FORMAT_B8G8R8A8_UNORM) &&
f.colorSpace == VK_COLOR_SPACE_SRGB_NONLINEAR_KHR) {
return f;
}
}
return formats.empty() ? VkSurfaceFormatKHR{VK_FORMAT_B8G8R8A8_UNORM, VK_COLOR_SPACE_SRGB_NONLINEAR_KHR}
: formats[0];
}
VkPresentModeKHR ChoosePresentMode(const Vector<VkPresentModeKHR>& modes) {
for (const auto& m : modes) {
if (m == VK_PRESENT_MODE_MAILBOX_KHR) return m;
}
return VK_PRESENT_MODE_FIFO_KHR;
}
VkExtent2D ChooseExtent(const VkSurfaceCapabilitiesKHR& caps, ANativeWindow* window) {
if (caps.currentExtent.width != UINT32_MAX) return caps.currentExtent;
// try to get frontend viewport size
VkExtent2D extent{640, 480};
#ifdef VK_USE_PLATFORM_ANDROID_KHR
if (window) {
extent.width = static_cast<Uint32>(ANativeWindow_getWidth(window));
extent.height = static_cast<Uint32>(ANativeWindow_getHeight(window));
}
#else
(void)window;
#endif
extent.width = std::max(caps.minImageExtent.width, std::min(caps.maxImageExtent.width, extent.width));
extent.height = std::max(caps.minImageExtent.height, std::min(caps.maxImageExtent.height, extent.height));
return extent;
}
} // namespace
SwapchainManager::~SwapchainManager() {
DestroySwapchain();
}
void SwapchainManager::Initialize() {
CreateSwapchain(VK_NULL_HANDLE);
}
void SwapchainManager::Recreate() {
DestroySwapchain();
CreateSwapchain(VK_NULL_HANDLE);
}
void SwapchainManager::SetFramebuffers(Vector<VkFramebuffer>&& framebuffers) {
m_framebuffers = Move(framebuffers);
}
void SwapchainManager::DestroySwapchain() {
auto device = m_ctx.GetDevice();
if (device == VK_NULL_HANDLE) return;
for (auto fb : m_framebuffers) {
if (fb != VK_NULL_HANDLE) vkDestroyFramebuffer(device, fb, nullptr);
}
m_framebuffers.clear();
for (auto view : m_imageViews) {
if (view != VK_NULL_HANDLE) vkDestroyImageView(device, view, nullptr);
}
m_imageViews.clear();
m_images.clear();
m_imagesInFlight.clear();
if (m_swapchain != VK_NULL_HANDLE) {
vkDestroySwapchainKHR(device, m_swapchain, nullptr);
m_swapchain = VK_NULL_HANDLE;
}
m_format = VK_FORMAT_UNDEFINED;
m_extent = {0, 0};
}
void SwapchainManager::CreateSwapchain(VkSwapchainKHR oldSwapchain) {
VkSurfaceCapabilitiesKHR caps{};
VK_VERIFY(vkGetPhysicalDeviceSurfaceCapabilitiesKHR(m_ctx.GetPhysicalDevice(), m_ctx.GetSurface(), &caps),
"vkGetPhysicalDeviceSurfaceCapabilitiesKHR");
Uint32 fmtCount = 0;
vkGetPhysicalDeviceSurfaceFormatsKHR(m_ctx.GetPhysicalDevice(), m_ctx.GetSurface(), &fmtCount, nullptr);
Vector<VkSurfaceFormatKHR> formats(fmtCount);
if (fmtCount > 0) {
vkGetPhysicalDeviceSurfaceFormatsKHR(m_ctx.GetPhysicalDevice(), m_ctx.GetSurface(), &fmtCount,
formats.data());
}
Uint32 modeCount = 0;
vkGetPhysicalDeviceSurfacePresentModesKHR(m_ctx.GetPhysicalDevice(), m_ctx.GetSurface(), &modeCount, nullptr);
Vector<VkPresentModeKHR> modes(modeCount);
if (modeCount > 0) {
vkGetPhysicalDeviceSurfacePresentModesKHR(m_ctx.GetPhysicalDevice(), m_ctx.GetSurface(), &modeCount,
modes.data());
}
VkSurfaceFormatKHR surfaceFormat = ChooseSurfaceFormat(formats);
VkPresentModeKHR presentMode = ChoosePresentMode(modes);
VkExtent2D extent;
if (m_viewportSize.x() == 0 || m_viewportSize.y() == 0) {
extent = ChooseExtent(caps, m_ctx.GetWindow());
} else {
extent = {m_viewportSize.x(), m_viewportSize.y()};
}
Uint32 imageCount = caps.minImageCount + 1;
if (caps.maxImageCount > 0 && imageCount > caps.maxImageCount) imageCount = caps.maxImageCount;
VkSwapchainCreateInfoKHR sci{VK_STRUCTURE_TYPE_SWAPCHAIN_CREATE_INFO_KHR};
sci.surface = m_ctx.GetSurface();
sci.minImageCount = imageCount;
sci.imageFormat = surfaceFormat.format;
sci.imageColorSpace = surfaceFormat.colorSpace;
sci.imageExtent = extent;
sci.imageArrayLayers = 1;
sci.imageUsage =
VK_IMAGE_USAGE_COLOR_ATTACHMENT_BIT | VK_IMAGE_USAGE_TRANSFER_DST_BIT | VK_IMAGE_USAGE_TRANSFER_SRC_BIT;
sci.imageSharingMode = VK_SHARING_MODE_EXCLUSIVE;
sci.preTransform = caps.currentTransform;
sci.compositeAlpha = VK_COMPOSITE_ALPHA_OPAQUE_BIT_KHR;
sci.presentMode = presentMode;
sci.clipped = VK_TRUE;
sci.oldSwapchain = oldSwapchain;
VK_VERIFY(vkCreateSwapchainKHR(m_ctx.GetDevice(), &sci, nullptr, &m_swapchain), "vkCreateSwapchainKHR");
Uint32 actualCount = 0;
vkGetSwapchainImagesKHR(m_ctx.GetDevice(), m_swapchain, &actualCount, nullptr);
m_images.resize(actualCount);
vkGetSwapchainImagesKHR(m_ctx.GetDevice(), m_swapchain, &actualCount, m_images.data());
m_imageViews.clear();
m_imageViews.reserve(actualCount);
for (auto image : m_images) {
VkImageViewCreateInfo ivci{VK_STRUCTURE_TYPE_IMAGE_VIEW_CREATE_INFO};
ivci.image = image;
ivci.viewType = VK_IMAGE_VIEW_TYPE_2D;
ivci.format = surfaceFormat.format;
ivci.subresourceRange.aspectMask = VK_IMAGE_ASPECT_COLOR_BIT;
ivci.subresourceRange.baseMipLevel = 0;
ivci.subresourceRange.levelCount = 1;
ivci.subresourceRange.baseArrayLayer = 0;
ivci.subresourceRange.layerCount = 1;
VkImageView view = VK_NULL_HANDLE;
VK_VERIFY(vkCreateImageView(m_ctx.GetDevice(), &ivci, nullptr, &view), "vkCreateImageView swapchain");
m_imageViews.push_back(view);
}
m_imagesInFlight.assign(actualCount, VK_NULL_HANDLE);
m_format = surfaceFormat.format;
m_extent = extent;
}
} // namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager
@@ -0,0 +1,52 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/Renderer/SwapchainManager.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 <Includes.h>
#include "MG_Util/Math/VectorTypes.h"
#include "VulkanContext.h"
#include "VkCommon.h"
namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager {
class SwapchainManager {
public:
explicit SwapchainManager(VulkanContext& ctx) : m_ctx(ctx) {}
~SwapchainManager();
SwapchainManager(const SwapchainManager&) = delete;
SwapchainManager& operator=(const SwapchainManager&) = delete;
void Initialize();
void Recreate();
void SetViewportSize(const UintVec2& size) { m_viewportSize = size; }
VkSwapchainKHR GetSwapchain() const { return m_swapchain; }
VkFormat GetFormat() const { return m_format; }
VkExtent2D GetExtent() const { return m_extent; }
const Vector<VkImage>& GetImages() const { return m_images; }
const Vector<VkImageView>& GetImageViews() const { return m_imageViews; }
const Vector<VkFramebuffer>& GetFramebuffers() const { return m_framebuffers; }
void SetFramebuffers(Vector<VkFramebuffer>&& framebuffers);
Vector<VkFence>& GetImagesInFlight() { return m_imagesInFlight; }
const UintVec2& GetViewportSize() const { return m_viewportSize; }
private:
void CreateSwapchain(VkSwapchainKHR oldSwapchain);
void DestroySwapchain();
VulkanContext& m_ctx;
VkSwapchainKHR m_swapchain = VK_NULL_HANDLE;
VkFormat m_format = VK_FORMAT_UNDEFINED;
VkExtent2D m_extent{0, 0};
Vector<VkImage> m_images;
Vector<VkImageView> m_imageViews;
Vector<VkFramebuffer> m_framebuffers;
Vector<VkFence> m_imagesInFlight;
UintVec2 m_viewportSize{0, 0};
};
} // namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager
@@ -0,0 +1,22 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/Renderer/VkCommon.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 <Includes.h>
namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager {
inline void VkCheck(VkResult result, const char* msg) {
if (result == VK_SUCCESS || result == VK_SUBOPTIMAL_KHR) return;
MGLOG_E("Vulkan error %d at %s", static_cast<Int>(result), msg ? msg : "(unknown)");
throw RuntimeError(msg ? msg : "Vulkan error");
}
} // namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager
#ifndef VK_VERIFY
#define VK_VERIFY(res, msg) ::MobileGL::MG_Backend::DirectVulkanTMP::VkManager::VkCheck((res), (msg))
#endif
@@ -0,0 +1,191 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/Renderer/VulkanContext.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 "VulkanContext.h"
namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager {
namespace {
Bool CheckDeviceExtensionSupport(VkPhysicalDevice device, const Vector<const char*>& required) {
Uint32 count = 0;
vkEnumerateDeviceExtensionProperties(device, nullptr, &count, nullptr);
Vector<VkExtensionProperties> props(count);
if (count > 0) vkEnumerateDeviceExtensionProperties(device, nullptr, &count, props.data());
for (auto* ext : required) {
Bool found = false;
for (const auto& p : props) {
if (std::strcmp(p.extensionName, ext) == 0) {
found = true;
break;
}
}
if (!found) return false;
}
return true;
}
Bool FindGraphicsQueueFamily(VkPhysicalDevice device, VkSurfaceKHR surface, Uint32& outFamily) {
Uint32 count = 0;
vkGetPhysicalDeviceQueueFamilyProperties(device, &count, nullptr);
if (count == 0) return false;
Vector<VkQueueFamilyProperties> props(count);
vkGetPhysicalDeviceQueueFamilyProperties(device, &count, props.data());
for (Uint32 i = 0; i < count; ++i) {
if (!(props[i].queueFlags & VK_QUEUE_GRAPHICS_BIT)) continue;
if (surface != VK_NULL_HANDLE) {
VkBool32 presentSupport = VK_FALSE;
vkGetPhysicalDeviceSurfaceSupportKHR(device, i, surface, &presentSupport);
if (!presentSupport) continue;
}
outFamily = i;
return true;
}
return false;
}
Bool SupportsDynamicRendering(VkPhysicalDevice device) {
VkPhysicalDeviceVulkan13Features vk13{VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_3_FEATURES};
VkPhysicalDeviceFeatures2 features2{VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_FEATURES_2};
features2.pNext = &vk13;
vkGetPhysicalDeviceFeatures2(device, &features2);
return vk13.dynamicRendering == VK_TRUE;
}
Bool SupportsTimelineSemaphore(VkPhysicalDevice device) {
VkPhysicalDeviceVulkan12Features vk12{VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_2_FEATURES};
VkPhysicalDeviceFeatures2 features2{VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_FEATURES_2};
features2.pNext = &vk12;
vkGetPhysicalDeviceFeatures2(device, &features2);
return vk12.timelineSemaphore == VK_TRUE;
}
} // namespace
VulkanContext::~VulkanContext() {
Cleanup();
}
void VulkanContext::Initialize(ANativeWindow* window, const char* appName) {
if (m_instance != VK_NULL_HANDLE) return;
m_window = window;
CreateInstance(appName ? appName : "MobileGL-Vulkan");
CreateSurface(window);
PickPhysicalDevice();
CreateDevice();
}
void VulkanContext::Cleanup() {
if (m_device != VK_NULL_HANDLE) {
vkDeviceWaitIdle(m_device);
vkDestroyDevice(m_device, nullptr);
m_device = VK_NULL_HANDLE;
}
if (m_surface != VK_NULL_HANDLE) {
vkDestroySurfaceKHR(m_instance, m_surface, nullptr);
m_surface = VK_NULL_HANDLE;
}
if (m_instance != VK_NULL_HANDLE) {
vkDestroyInstance(m_instance, nullptr);
m_instance = VK_NULL_HANDLE;
}
m_physicalDevice = VK_NULL_HANDLE;
m_graphicsQueue = VK_NULL_HANDLE;
m_graphicsQueueFamily = ~0u;
m_window = nullptr;
}
void VulkanContext::CreateInstance(const char* appName) {
VkApplicationInfo app{VK_STRUCTURE_TYPE_APPLICATION_INFO};
app.pApplicationName = appName;
app.applicationVersion = VK_MAKE_VERSION(1, 3, 0);
app.pEngineName = "MobileGL";
app.engineVersion = VK_MAKE_VERSION(1, 3, 0);
app.apiVersion = VK_API_VERSION_1_3;
Vector<const char*> extensions;
extensions.push_back(VK_KHR_SURFACE_EXTENSION_NAME);
#ifdef VK_USE_PLATFORM_ANDROID_KHR
extensions.push_back(VK_KHR_ANDROID_SURFACE_EXTENSION_NAME);
#endif
VkInstanceCreateInfo ici{VK_STRUCTURE_TYPE_INSTANCE_CREATE_INFO};
ici.pApplicationInfo = &app;
ici.enabledExtensionCount = static_cast<Uint32>(extensions.size());
ici.ppEnabledExtensionNames = extensions.data();
VK_VERIFY(vkCreateInstance(&ici, nullptr, &m_instance), "vkCreateInstance");
}
void VulkanContext::CreateSurface(ANativeWindow* window) {
if (!window) return;
#ifdef VK_USE_PLATFORM_ANDROID_KHR
VkAndroidSurfaceCreateInfoKHR sci{VK_STRUCTURE_TYPE_ANDROID_SURFACE_CREATE_INFO_KHR};
sci.window = window;
VK_VERIFY(vkCreateAndroidSurfaceKHR(m_instance, &sci, nullptr, &m_surface), "vkCreateAndroidSurfaceKHR");
#else
(void)window;
#endif
}
void VulkanContext::PickPhysicalDevice() {
Uint32 count = 0;
VK_VERIFY(vkEnumeratePhysicalDevices(m_instance, &count, nullptr), "vkEnumeratePhysicalDevices");
if (count == 0) throw RuntimeError("No Vulkan physical devices found");
Vector<VkPhysicalDevice> devices(count);
VK_VERIFY(vkEnumeratePhysicalDevices(m_instance, &count, devices.data()), "vkEnumeratePhysicalDevices list");
Vector<const char*> requiredExts = {VK_KHR_SWAPCHAIN_EXTENSION_NAME};
for (auto device : devices) {
Uint32 family = ~0u;
if (!FindGraphicsQueueFamily(device, m_surface, family)) continue;
if (!CheckDeviceExtensionSupport(device, requiredExts)) continue;
if (!SupportsDynamicRendering(device)) continue;
m_physicalDevice = device;
m_graphicsQueueFamily = family;
break;
}
if (m_physicalDevice == VK_NULL_HANDLE) {
throw RuntimeError("No suitable Vulkan physical device found");
}
}
void VulkanContext::CreateDevice() {
float priority = 1.0f;
VkDeviceQueueCreateInfo qci{VK_STRUCTURE_TYPE_DEVICE_QUEUE_CREATE_INFO};
qci.queueFamilyIndex = m_graphicsQueueFamily;
qci.queueCount = 1;
qci.pQueuePriorities = &priority;
Vector<const char*> deviceExtensions = {VK_KHR_SWAPCHAIN_EXTENSION_NAME};
VkPhysicalDeviceFeatures features{};
VkPhysicalDeviceVulkan13Features vk13{VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_3_FEATURES};
vk13.dynamicRendering = VK_TRUE;
VkPhysicalDeviceVulkan12Features vk12{VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_2_FEATURES};
vk12.timelineSemaphore = SupportsTimelineSemaphore(m_physicalDevice) ? VK_TRUE : VK_FALSE;
VkPhysicalDeviceFeatures2 features2{VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_FEATURES_2};
features2.features = features;
vk13.pNext = &vk12;
features2.pNext = &vk13;
VkDeviceCreateInfo dci{VK_STRUCTURE_TYPE_DEVICE_CREATE_INFO};
dci.pNext = &features2;
dci.queueCreateInfoCount = 1;
dci.pQueueCreateInfos = &qci;
dci.enabledExtensionCount = static_cast<Uint32>(deviceExtensions.size());
dci.ppEnabledExtensionNames = deviceExtensions.data();
dci.pEnabledFeatures = nullptr;
VK_VERIFY(vkCreateDevice(m_physicalDevice, &dci, nullptr, &m_device), "vkCreateDevice");
vkGetDeviceQueue(m_device, m_graphicsQueueFamily, 0, &m_graphicsQueue);
}
} // namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager
@@ -0,0 +1,47 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/Renderer/VulkanContext.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 <Includes.h>
#include "VkCommon.h"
namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager {
class VulkanContext {
public:
VulkanContext() = default;
~VulkanContext();
VulkanContext(const VulkanContext&) = delete;
VulkanContext& operator=(const VulkanContext&) = delete;
void Initialize(ANativeWindow* window, const char* appName);
void Cleanup();
VkInstance GetInstance() const { return m_instance; }
VkPhysicalDevice GetPhysicalDevice() const { return m_physicalDevice; }
VkDevice GetDevice() const { return m_device; }
VkQueue GetGraphicsQueue() const { return m_graphicsQueue; }
Uint32 GetGraphicsQueueFamily() const { return m_graphicsQueueFamily; }
VkSurfaceKHR GetSurface() const { return m_surface; }
ANativeWindow* GetWindow() const { return m_window; }
private:
void CreateInstance(const char* appName);
void CreateSurface(ANativeWindow* window);
void PickPhysicalDevice();
void CreateDevice();
VkInstance m_instance = VK_NULL_HANDLE;
VkPhysicalDevice m_physicalDevice = VK_NULL_HANDLE;
VkDevice m_device = VK_NULL_HANDLE;
VkQueue m_graphicsQueue = VK_NULL_HANDLE;
VkSurfaceKHR m_surface = VK_NULL_HANDLE;
Uint32 m_graphicsQueueFamily = ~0u;
ANativeWindow* m_window = nullptr;
};
} // namespace MobileGL::MG_Backend::DirectVulkanTMP::VkManager
File diff suppressed because it is too large Load Diff
@@ -0,0 +1,308 @@
// MobileGL - MobileGL/MG_Backend/DirectVulkanTMP/TmpImpl.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 <Includes.h>
#include "Renderer/VkCommon.h"
#include "Renderer/VulkanContext.h"
#include "Renderer/SwapchainManager.h"
#include "Renderer/FrameContext.h"
#include "Managers/ProgramManager.h"
#include <MG_State/GLState/Core.h>
namespace MobileGL::MG_Backend::DirectVulkanTMP::TmpImpl {
namespace DV = MobileGL::MG_Backend::DirectVulkanTMP::VkManager;
using BufferObject = MobileGL::MG_State::GLState::BufferObject;
using ProgramObject = MobileGL::MG_State::GLState::ProgramObject;
using VertexArrayObject = MobileGL::MG_State::GLState::VertexArrayObject;
using FramebufferObject = MobileGL::MG_State::GLState::FramebufferObject;
using FramebufferAttachmentObject = MobileGL::MG_State::GLState::FramebufferAttachmentObject;
using ITextureObject = MobileGL::MG_State::GLState::ITextureObject;
using RenderbufferObject = MobileGL::MG_State::GLState::RenderbufferObject;
using SamplerObject = MobileGL::MG_State::GLState::SamplerObject;
using TextureObjectMipmap = MobileGL::MG_State::GLState::TextureObjectMipmap;
using TrashImage = MobileGL::MG_Backend::DirectVulkanTMP::TrashImage;
struct BufferResource {
VkBuffer buffer = VK_NULL_HANDLE;
VkDeviceMemory memory = VK_NULL_HANDLE;
VkDeviceSize size = 0;
VkDeviceSize offset = 0;
VkBufferUsageFlags usage = 0;
VkMemoryPropertyFlags props = 0;
void* mapped = nullptr;
Bool fromRing = false;
Uint32 lastUsedFrame = ~0u;
};
struct BufferResourceSet {
Vector<BufferResource> perFrame;
};
struct PendingBufferCopyBatch {
VkBuffer src = VK_NULL_HANDLE;
VkBuffer dst = VK_NULL_HANDLE;
Vector<VkBufferCopy> regions;
};
struct StagingRing {
BufferResource buffer;
VkDeviceSize capacity = 0;
VkDeviceSize head = 0;
};
struct TextureResource {
VkImage image = VK_NULL_HANDLE;
VkDeviceMemory memory = VK_NULL_HANDLE;
VkImageView view = VK_NULL_HANDLE;
UnorderedMap<Uint32, VkImageView> mipViews;
VkFormat format = VK_FORMAT_UNDEFINED;
VkExtent3D extent{0, 0, 1};
VkImageLayout layout = VK_IMAGE_LAYOUT_UNDEFINED;
Uint32 mipLevels = 1;
Uint16 paramsVersion = 0;
bool valid = false;
};
struct RenderbufferResource {
VkImage image = VK_NULL_HANDLE;
VkDeviceMemory memory = VK_NULL_HANDLE;
VkImageView view = VK_NULL_HANDLE;
VkFormat format = VK_FORMAT_UNDEFINED;
VkExtent2D extent{0, 0};
VkImageLayout layout = VK_IMAGE_LAYOUT_UNDEFINED;
bool valid = false;
};
struct SamplerResource {
VkSampler sampler = VK_NULL_HANDLE;
Uint16 version = 0;
SamplerParameters params{};
};
struct PendingClearInfo {
GLbitfield mask = 0;
FloatVec4 clearColor = {0.0f, 0.0f, 0.0f, 1.0f};
Float clearDepth = 1.0f;
};
struct UniformBindingInfo {
String name;
Uint32 binding = 0;
Uint32 set = 0;
Uint32 blockIndex = 0xFFFFFFFFu;
VkShaderStageFlags stages = 0;
};
struct SamplerBindingInfo {
String name;
Uint32 binding = 0;
Uint32 set = 0;
VkShaderStageFlags stages = 0;
};
struct ProgramResource {
ProgramObject* program = nullptr;
Uint64 spvHash = 0;
DV::ProgramManager::HashType programHash = 0;
Vector<VkPipelineShaderStageCreateInfo>* shaderStages = nullptr;
VkDescriptorSetLayout setLayout = VK_NULL_HANDLE;
VkPipelineLayout pipelineLayout = VK_NULL_HANDLE;
struct UboPool {
Vector<BufferResource> buffers;
Uint32 cursor = 0;
};
Vector<UboPool> uboPools;
SizeT uboSize = 0;
Int32 uboBinding = -1;
Uint32 uniformDescriptorCount = 0;
Uint32 samplerDescriptorCount = 0;
Vector<UniformBindingInfo> uniformBindings;
Vector<SamplerBindingInfo> samplerBindings;
UnorderedMap<String, Uint32> samplerBindingByName;
UnorderedMap<Uint64, VkPipeline> pipelines;
};
struct DescriptorPoolConfig {
Uint32 maxSets = 0;
Uint32 uniformCount = 0;
Uint32 samplerCount = 0;
};
struct FrameDescriptorPools {
Vector<VkDescriptorPool> pools;
DescriptorPoolConfig config;
Uint32 activePool = 0;
};
struct VertexAttribKey {
Uint8 enabled = 0;
Uint8 size = 0;
Uint8 type = 0;
Uint8 normalized = 0;
Uint8 isInteger = 0;
Uint8 divisor = 0;
Uint16 pad = 0;
Uint32 stride = 0;
Uint64 offset = 0;
};
struct AttachmentInfo {
FramebufferAttachmentType type = FramebufferAttachmentType::None;
VkImageView view = VK_NULL_HANDLE;
VkFormat format = VK_FORMAT_UNDEFINED;
VkImageAspectFlags aspect = 0;
VkImageLayout finalLayout = VK_IMAGE_LAYOUT_UNDEFINED;
ITextureObject* texture = nullptr;
RenderbufferObject* renderbuffer = nullptr;
};
struct BackendFramebufferObject {
VkExtent2D extent{0, 0};
Uint32 colorAttachmentCount = 0;
Bool hasDepth = false;
VkFormat depthFormat = VK_FORMAT_UNDEFINED;
VkFormat depthAttachmentFormat = VK_FORMAT_UNDEFINED;
VkFormat stencilAttachmentFormat = VK_FORMAT_UNDEFINED;
Uint64 renderingCompatHash = 0;
Vector<AttachmentInfo> attachments;
Vector<VkFormat> colorAttachmentFormats;
Array<Int32, static_cast<SizeT>(FramebufferAttachmentType::FramebufferAttachmentTypeCount)> attachmentIndex =
{};
FramebufferObject::FramebufferAttachmentVersionArray syncedAttachmentVersions = {0};
FramebufferObject::FramebufferAttachmentArray frontendDrawBuffers = {FramebufferAttachmentType::None};
Array<Int32, FramebufferObject::MAX_DRAW_BUFFERS> drawBufferAttachmentIndices = {};
FramebufferAttachmentType frontendReadBuffer = FramebufferAttachmentType::Color0;
Uint64 configHash = 0;
Uint16 objectVersion = 0;
Bool IsValid() const { return !attachments.empty() && extent.width > 0 && extent.height > 0; }
};
struct VulkanState {
UniquePtr<DV::VulkanContext> ctx;
UniquePtr<DV::SwapchainManager> swapchain;
UniquePtr<DV::ProgramManager> programMgr;
VkCommandPool commandPool = VK_NULL_HANDLE;
Vector<UniquePtr<DV::FrameContext>> frames;
Vector<FrameDescriptorPools> frameDescriptorPools;
Vector<BufferResource> stagingPool;
Vector<Vector<BufferResource>> stagingPending;
Uint32 currentFrame = 0;
Uint32 maxFramesInFlight = 2;
Bool initialized = false;
UintVec2 viewportSize{0, 0};
VkImage depthImage = VK_NULL_HANDLE;
VkDeviceMemory depthMemory = VK_NULL_HANDLE;
VkImageView depthView = VK_NULL_HANDLE;
VkFormat depthFormat = VK_FORMAT_UNDEFINED;
Uint64 defaultRenderingCompatHash = 0;
BufferResource nullUbo;
TextureResource defaultTexture;
SamplerResource defaultSampler;
UnorderedMap<const BufferObject*, BufferResourceSet> buffers;
UnorderedMap<const ITextureObject*, TextureResource> textures;
UnorderedMap<const RenderbufferObject*, RenderbufferResource> renderbuffers;
UnorderedMap<const SamplerObject*, SamplerResource> samplers;
UnorderedMap<const ProgramObject*, ProgramResource> programs;
UnorderedMap<SharedPtr<MG_State::GLState::FramebufferObject>, SharedPtr<BackendFramebufferObject>> framebuffers;
Vector<PendingBufferCopyBatch> pendingBufferCopies;
Vector<StagingRing> stagingRings;
VkExtent2D activeExtent{0, 0};
Vector<VkImageView> activeColorViews;
Vector<VkImageLayout> activeColorLayouts;
Vector<VkFormat> activeColorFormats;
Vector<Int32> activeColorAttachmentIndices;
VkImageView activeDepthView = VK_NULL_HANDLE;
VkImageLayout activeDepthLayout = VK_IMAGE_LAYOUT_UNDEFINED;
VkImageView activeStencilView = VK_NULL_HANDLE;
VkImageLayout activeStencilLayout = VK_IMAGE_LAYOUT_UNDEFINED;
Uint32 activeSwapchainImageIndex = ~0u;
Uint32 activeColorAttachmentCount = 1;
Bool activeHasDepth = false;
VkFormat activeDepthFormat = VK_FORMAT_UNDEFINED;
VkFormat activeDepthAttachmentFormat = VK_FORMAT_UNDEFINED;
VkFormat activeStencilAttachmentFormat = VK_FORMAT_UNDEFINED;
Uint64 activeRenderingCompatHash = 0;
Uint64 activeRenderingConfigHash = 0;
SharedPtr<MG_State::GLState::FramebufferObject> activeStateFBO = nullptr;
SharedPtr<BackendFramebufferObject> activeBackendFBO = nullptr;
Bool activeIsDefault = true;
Bool activeDefaultColorWrites = true;
Uint64 recordingRenderingConfigHash = 0;
Bool cmdBufferBegun = false;
Bool recording = false;
Bool frameSubmitted = false;
FloatVec4 clearColor = {0.0f, 0.0f, 0.0f, 1.0f};
Float clearDepth = 1.0f;
UnorderedMap<SharedPtr<FramebufferObject>, PendingClearInfo> pendingClears;
Vector<VkImageLayout> swapchainImageLayouts;
PFN_vkCmdBeginRendering pfnCmdBeginRendering = nullptr;
PFN_vkCmdEndRendering pfnCmdEndRendering = nullptr;
PFN_vkWaitSemaphores pfnWaitSemaphores = nullptr;
VkSemaphore timelineSemaphore = VK_NULL_HANDLE;
Uint64 timelineValue = 0;
};
void Present();
void FrameBegin();
void InitVulkan(ANativeWindow* window);
const VulkanState& GetVulkanState();
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);
void ClearBufferiv(GLenum buffer, GLint drawbuffer, const GLint* value);
void Clear(GLbitfield mask);
void DrawElements(GLenum mode, GLsizei count, GLenum type, const void* indices);
void DrawArrays(GLenum mode, GLint first, GLsizei count);
void DrawElementsBaseVertex(GLenum mode, GLsizei count, GLenum type, const GLvoid* indices, GLint basevertex);
void MultiDrawElements(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount);
void MultiDrawElementsBaseVertex(GLenum mode, const GLsizei* count, GLenum type, const GLvoid* const* indices,
GLsizei drawcount, const GLint* basevertex);
void MultiDrawElementsIndirect(GLenum mode, GLenum type, const void* indirect, GLsizei drawcount, GLsizei stride);
void MultiDrawArraysIndirect(GLenum mode, const void* indirect, GLsizei drawcount, GLsizei stride);
void DrawRangeElementsBaseVertex(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type,
const void* indices, GLint basevertex);
void DrawRangeElements(GLenum mode, GLuint start, GLuint end, GLsizei count, GLenum type, const void* indices);
void DrawElementsInstancedBaseVertexBaseInstance(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLint basevertex, GLuint baseinstance);
void DrawElementsInstancedBaseVertex(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLint basevertex);
void DrawElementsInstancedBaseInstance(GLenum mode, GLsizei count, GLenum type, const void* indices,
GLsizei instancecount, GLuint baseinstance);
void DrawElementsInstanced(GLenum mode, GLsizei count, GLenum type, const void* indices, GLsizei instancecount);
void DrawElementsIndirect(GLenum mode, GLenum type, const void* indirect);
void DrawArraysInstancedBaseInstance(GLenum mode, GLint first, GLsizei count, GLsizei instancecount,
GLuint baseinstance);
void DrawArraysInstanced(GLenum mode, GLint first, GLsizei count, GLsizei instancecount);
void DrawArraysIndirect(GLenum mode, const void* indirect);
void BlitFramebuffer(GLint srcX0, GLint srcY0, GLint srcX1, GLint srcY1, GLint dstX0, GLint dstY0, GLint dstX1,
GLint dstY1, GLbitfield mask, GLenum filter);
void CopyTexImage2D(GLenum target, GLint level, GLenum internalformat, GLint x, GLint y, GLsizei width,
GLsizei height, GLint border);
void CopyTexSubImage2D(GLenum target, GLint level, GLint xoffset, GLint yoffset, GLint x, GLint y, GLsizei width,
GLsizei height);
void GenerateMipmap(GLenum target);
} // namespace MobileGL::MG_Backend::DirectVulkanTMP::TmpImpl
+2 -2
View File
@@ -52,8 +52,8 @@ namespace MobileGL::MG_Backend {
case BackendType::DirectGLES:
pActiveBackendObject = MakeUnique<DirectGLES::BackendObject_DirectGLES>();
break;
case BackendType::DirectVulkan:
pActiveBackendObject = MakeUnique<DirectVulkan::BackendObject_DirectVulkan>();
case BackendType::DirectVulkanTMP:
pActiveBackendObject = MakeUnique<DirectVulkanTMP::BackendObject_DirectVulkanTMP>();
break;
case BackendType::Unknown:
default:
+2 -2
View File
@@ -61,7 +61,7 @@ static void BM_CreateBufferObjectsAndBindBuffer(benchmark::State& state) {
BENCHMARK(BM_CreateBufferObjectsAndBindBuffer)->Unit(benchmark::kMillisecond)->UseRealTime();
static void BM_DeleteBufferObjects(benchmark::State& state) {
Initialize();
MG_Initialize();
std::vector<GLuint> buffers(BUFFER_COUNT);
for (auto _ : state) {
@@ -179,7 +179,7 @@ static void BM_UpdateDataPartially(benchmark::State& state) {
BENCHMARK(BM_UpdateDataPartially)->Unit(benchmark::kMillisecond)->UseRealTime();
int main(int argc, char** argv) {
Initialize();
MG_Initialize();
benchmark ::MaybeReenterWithoutASLR(argc, argv);
char arg0_default[] = "benchmark";
char* args_default = reinterpret_cast<char*>(arg0_default);
+1 -2
View File
@@ -16,5 +16,4 @@ target_link_libraries(
${LINK_LIBRARIES}
)
add_test(NAME BufferBench COMMAND BufferBench --benchmark_counters_tabular=true)
set_tests_properties(BufferBench PROPERTIES LABELS benchmark)
add_test(NAME BufferBench COMMAND BufferBench --benchmark_counters_tabular=true)
-1
View File
@@ -38,7 +38,6 @@ target_link_libraries(
)
add_test(NAME SanityBench COMMAND SanityBench --benchmark_counters_tabular=true)
set_tests_properties(SanityBench PROPERTIES LABELS benchmark)
add_subdirectory(Program)
add_subdirectory(Buffer)
+1 -2
View File
@@ -16,5 +16,4 @@ target_link_libraries(
${LINK_LIBRARIES}
)
add_test(NAME ProgramBench COMMAND ProgramBench --benchmark_counters_tabular=true)
set_tests_properties(ProgramBench PROPERTIES LABELS benchmark)
add_test(NAME ProgramBench COMMAND ProgramBench --benchmark_counters_tabular=true)
@@ -43,7 +43,7 @@ void main(){
// Use TestMat2 and TestMat3 to prevent optimization
vec2 dummy2 = TestMat2[0];
vec3 dummy3 = TestMat3[0];
oneTexel = (1.0 * (fIn1 * fIn2 * fIn3 * fIn4 * fIn5 * fIn6)) / InSize;
texCoord = Position.xy / OutSize;
@@ -221,7 +221,7 @@ static void BM_CompileAndLink(benchmark::State& state) {
BENCHMARK(BM_CompileAndLink)->Unit(benchmark::kMillisecond);
int main(int argc, char** argv) {
Initialize();
MG_Initialize();
benchmark ::MaybeReenterWithoutASLR(argc, argv);
char arg0_default[] = "benchmark";
char* args_default = reinterpret_cast<char*>(arg0_default);
@@ -234,4 +234,4 @@ int main(int argc, char** argv) {
::benchmark ::RunSpecifiedBenchmarks();
::benchmark ::Shutdown();
return 0;
}
}
-700
View File
@@ -1,700 +0,0 @@
// MobileGL - MobileGL/MG_Impl/CGLImpl/CGLImpl.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 "CGLImpl.h"
#if defined(__APPLE__)
#include "../EGLImpl/EGLImpl.h"
namespace MobileGL::MG_Impl::CGLImpl {
namespace {
struct PixelFormatObject {
Uint32 RetainCount = 1;
Bool DoubleBuffer = true;
GLint ColorSize = 24;
GLint AlphaSize = 8;
GLint DepthSize = 24;
GLint StencilSize = 8;
GLint SampleBuffers = 0;
GLint Samples = 0;
GLint Profile = kCGLOGLPVersion_3_2_Core;
GLint RendererId = 0x4d474c;
};
struct ContextObject {
Uint32 RetainCount = 1;
CGLPixelFormatObj PixelFormat = nullptr;
CGLContextObj Share = nullptr;
EGLDisplay Display = EGL_NO_DISPLAY;
EGLConfig Config = nullptr;
EGLContext Context = EGL_NO_CONTEXT;
EGLSurface Surface = EGL_NO_SURFACE;
void* NSObject = nullptr;
void* View = nullptr;
void* MetalLayer = nullptr;
GLint SwapInterval = 1;
GLint VirtualScreen = 0;
GLint SurfaceBackingSize[2] = {0, 0};
Bool HasDrawable = false;
Bool Locked = false;
};
std::recursive_mutex& RegistryMutex() {
static auto* mutex = new std::recursive_mutex();
return *mutex;
}
Uint64& NextPixelFormatHandle() {
static auto* handle = new Uint64(1);
return *handle;
}
Uint64& NextContextHandle() {
static auto* handle = new Uint64(1);
return *handle;
}
UnorderedMap<CGLPixelFormatObj, PixelFormatObject>& PixelFormats() {
static auto* formats = new UnorderedMap<CGLPixelFormatObj, PixelFormatObject>();
return *formats;
}
UnorderedMap<CGLContextObj, ContextObject>& Contexts() {
static auto* contexts = new UnorderedMap<CGLContextObj, ContextObject>();
return *contexts;
}
UnorderedMap<std::thread::id, CGLContextObj>& CurrentContexts() {
static auto* contexts = new UnorderedMap<std::thread::id, CGLContextObj>();
return *contexts;
}
CGLPixelFormatObj EncodePixelFormat(Uint64 handle) {
return reinterpret_cast<CGLPixelFormatObj>(static_cast<SizeT>(handle));
}
CGLContextObj EncodeContext(Uint64 handle) {
return reinterpret_cast<CGLContextObj>(static_cast<SizeT>(handle));
}
std::thread::id CurrentThreadKey() {
return std::this_thread::get_id();
}
Bool AttributeHasValue(CGLPixelFormatAttribute attrib) {
switch (attrib) {
case kCGLPFAColorSize:
case kCGLPFAAlphaSize:
case kCGLPFADepthSize:
case kCGLPFAStencilSize:
case kCGLPFASampleBuffers:
case kCGLPFASamples:
case kCGLPFARendererID:
case kCGLPFADisplayMask:
case kCGLPFAOpenGLProfile:
return true;
default:
return false;
}
}
void ApplyPixelFormatAttribute(PixelFormatObject& pixelFormat,
CGLPixelFormatAttribute attrib,
GLint value) {
switch (attrib) {
case kCGLPFADoubleBuffer:
pixelFormat.DoubleBuffer = true;
break;
case kCGLPFAColorSize:
pixelFormat.ColorSize = value;
break;
case kCGLPFAAlphaSize:
pixelFormat.AlphaSize = value;
break;
case kCGLPFADepthSize:
pixelFormat.DepthSize = value;
break;
case kCGLPFAStencilSize:
pixelFormat.StencilSize = value;
break;
case kCGLPFASampleBuffers:
pixelFormat.SampleBuffers = value;
break;
case kCGLPFASamples:
pixelFormat.Samples = value;
break;
case kCGLPFAOpenGLProfile:
pixelFormat.Profile = value;
break;
case kCGLPFARendererID:
pixelFormat.RendererId = value;
break;
default:
break;
}
}
Bool InitEGLContext(ContextObject& object, CGLPixelFormatObj pix, CGLContextObj share) {
auto* pixelFormat = [&]() -> PixelFormatObject* {
auto& pixelFormats = PixelFormats();
auto it = pixelFormats.find(pix);
return it == pixelFormats.end() ? nullptr : &it->second;
}();
if (!pixelFormat) {
return false;
}
EGLDisplay display = EGLImpl::GetDisplay(EGL_DEFAULT_DISPLAY);
if (display == EGL_NO_DISPLAY) {
return false;
}
if (!EGLImpl::Initialize(display, nullptr, nullptr)) {
return false;
}
EGLImpl::BindAPI(EGL_OPENGL_API);
const EGLint attribs[] = {
EGL_RED_SIZE, 8,
EGL_GREEN_SIZE, 8,
EGL_BLUE_SIZE, 8,
EGL_ALPHA_SIZE, std::max(pixelFormat->AlphaSize, 0),
EGL_DEPTH_SIZE, std::max(pixelFormat->DepthSize, 0),
EGL_STENCIL_SIZE, std::max(pixelFormat->StencilSize, 0),
EGL_SURFACE_TYPE, EGL_WINDOW_BIT | EGL_PBUFFER_BIT,
EGL_RENDERABLE_TYPE, EGL_OPENGL_BIT,
EGL_NONE,
};
EGLConfig config = nullptr;
EGLint count = 0;
if (!EGLImpl::ChooseConfig(display, attribs, &config, 1, &count) || count <= 0) {
return false;
}
EGLContext shareContext = EGL_NO_CONTEXT;
if (share != nullptr) {
auto& contexts = Contexts();
auto shareIt = contexts.find(share);
if (shareIt == contexts.end()) {
return false;
}
shareContext = shareIt->second.Context;
}
const EGLint contextAttribs[] = {
EGL_CONTEXT_MAJOR_VERSION, 3,
EGL_CONTEXT_MINOR_VERSION, 3,
EGL_NONE,
};
EGLContext eglContext = EGLImpl::CreateContext(display, config, shareContext, contextAttribs);
if (eglContext == EGL_NO_CONTEXT) {
return false;
}
object.Display = display;
object.Config = config;
object.Context = eglContext;
object.PixelFormat = pix;
object.Share = share;
return true;
}
ContextObject* TryGetContext(CGLContextObj ctx) {
auto& contexts = Contexts();
auto it = contexts.find(ctx);
return it == contexts.end() ? nullptr : &it->second;
}
const ContextObject* TryGetContext(CGLContextObj ctx, const std::lock_guard<std::recursive_mutex>&) {
auto& contexts = Contexts();
auto it = contexts.find(ctx);
return it == contexts.end() ? nullptr : &it->second;
}
PixelFormatObject* TryGetPixelFormat(CGLPixelFormatObj pix) {
auto& pixelFormats = PixelFormats();
auto it = pixelFormats.find(pix);
return it == pixelFormats.end() ? nullptr : &it->second;
}
CGLError MakeCurrentLocked(CGLContextObj ctx, ContextObject& object) {
CurrentContexts()[CurrentThreadKey()] = ctx;
if (!object.HasDrawable || object.Surface == EGL_NO_SURFACE) {
return kCGLNoError;
}
if (!EGLImpl::MakeCurrent(object.Display, object.Surface, object.Surface, object.Context)) {
return kCGLBadState;
}
return kCGLNoError;
}
CGLError RecreateSurfaceLocked(CGLContextObj ctx, ContextObject& object) {
if (!object.MetalLayer) {
return kCGLBadDrawable;
}
if (object.Surface != EGL_NO_SURFACE) {
EGLImpl::DestroySurface(object.Display, object.Surface);
object.Surface = EGL_NO_SURFACE;
}
const EGLAttrib attribs[] = {
EGL_WIDTH, std::max<GLint>(object.SurfaceBackingSize[0], 1),
EGL_HEIGHT, std::max<GLint>(object.SurfaceBackingSize[1], 1),
EGL_NONE,
};
EGLSurface surface = EGLImpl::CreatePlatformWindowSurface(object.Display, object.Config,
object.MetalLayer, attribs);
if (surface == EGL_NO_SURFACE) {
object.HasDrawable = false;
return kCGLBadDrawable;
}
object.Surface = surface;
object.HasDrawable = true;
return GetCurrentContext() == ctx ? MakeCurrentLocked(ctx, object) : kCGLNoError;
}
CGLError ResizeSurfaceLocked(ContextObject& object) {
if (object.Surface == EGL_NO_SURFACE) {
return kCGLBadDrawable;
}
return EGLImpl::ResizePlatformWindowSurface(
object.Display, object.Surface,
std::max<GLint>(object.SurfaceBackingSize[0], 1),
std::max<GLint>(object.SurfaceBackingSize[1], 1))
? kCGLNoError
: kCGLBadDrawable;
}
} // namespace
CGLError ChoosePixelFormat(const CGLPixelFormatAttribute* attribs, CGLPixelFormatObj* pix, GLint* npix) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
if (!pix || !npix) {
return kCGLBadAddress;
}
PixelFormatObject object;
if (attribs) {
for (SizeT i = 0; attribs[i] != static_cast<CGLPixelFormatAttribute>(0); ++i) {
const auto attrib = attribs[i];
GLint value = 1;
if (AttributeHasValue(attrib)) {
value = static_cast<GLint>(attribs[++i]);
}
ApplyPixelFormatAttribute(object, attrib, value);
}
}
const auto handle = EncodePixelFormat(NextPixelFormatHandle()++);
PixelFormats()[handle] = object;
*pix = handle;
*npix = 1;
return kCGLNoError;
}
CGLError DestroyPixelFormat(CGLPixelFormatObj pix) {
ReleasePixelFormat(pix);
return kCGLNoError;
}
CGLError DescribePixelFormat(CGLPixelFormatObj pix, GLint pixNum, CGLPixelFormatAttribute attrib, GLint* value) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
if (!value) {
return kCGLBadAddress;
}
if (pixNum != 0 && pixNum != 1) {
return kCGLBadValue;
}
auto* pixelFormat = TryGetPixelFormat(pix);
if (!pixelFormat) {
return kCGLBadPixelFormat;
}
switch (attrib) {
case kCGLPFADoubleBuffer:
*value = pixelFormat->DoubleBuffer ? 1 : 0;
return kCGLNoError;
case kCGLPFAAccelerated:
case kCGLPFAAcceleratedCompute:
case kCGLPFASupportsAutomaticGraphicsSwitching:
*value = 1;
return kCGLNoError;
case kCGLPFAColorSize:
*value = pixelFormat->ColorSize;
return kCGLNoError;
case kCGLPFAAlphaSize:
*value = pixelFormat->AlphaSize;
return kCGLNoError;
case kCGLPFADepthSize:
*value = pixelFormat->DepthSize;
return kCGLNoError;
case kCGLPFAStencilSize:
*value = pixelFormat->StencilSize;
return kCGLNoError;
case kCGLPFASampleBuffers:
*value = pixelFormat->SampleBuffers;
return kCGLNoError;
case kCGLPFASamples:
*value = pixelFormat->Samples;
return kCGLNoError;
case kCGLPFARendererID:
*value = pixelFormat->RendererId;
return kCGLNoError;
case kCGLPFAOpenGLProfile:
*value = pixelFormat->Profile;
return kCGLNoError;
case kCGLPFAVirtualScreenCount:
*value = 1;
return kCGLNoError;
default:
*value = 0;
return kCGLNoError;
}
}
void ReleasePixelFormat(CGLPixelFormatObj pix) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* pixelFormat = TryGetPixelFormat(pix);
if (!pixelFormat) {
return;
}
if (pixelFormat->RetainCount > 1) {
--pixelFormat->RetainCount;
return;
}
PixelFormats().erase(pix);
}
CGLPixelFormatObj RetainPixelFormat(CGLPixelFormatObj pix) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* pixelFormat = TryGetPixelFormat(pix);
if (pixelFormat) {
++pixelFormat->RetainCount;
}
return pix;
}
GLuint GetPixelFormatRetainCount(CGLPixelFormatObj pix) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* pixelFormat = TryGetPixelFormat(pix);
return pixelFormat ? pixelFormat->RetainCount : 0;
}
CGLError CreateContext(CGLPixelFormatObj pix, CGLContextObj share, CGLContextObj* ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
if (!ctx) {
return kCGLBadAddress;
}
if (!TryGetPixelFormat(pix)) {
return kCGLBadPixelFormat;
}
if (share && !TryGetContext(share)) {
return kCGLBadMatch;
}
ContextObject object;
if (!InitEGLContext(object, pix, share)) {
return kCGLBadAlloc;
}
RetainPixelFormat(pix);
const auto handle = EncodeContext(NextContextHandle()++);
Contexts()[handle] = object;
*ctx = handle;
return kCGLNoError;
}
CGLError DestroyContext(CGLContextObj ctx) {
ReleaseContext(ctx);
return kCGLNoError;
}
CGLContextObj RetainContext(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (object) {
++object->RetainCount;
}
return ctx;
}
void ReleaseContext(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (!object) {
return;
}
if (object->RetainCount > 1) {
--object->RetainCount;
return;
}
if (object->Surface != EGL_NO_SURFACE) {
EGLImpl::DestroySurface(object->Display, object->Surface);
}
if (object->Context != EGL_NO_CONTEXT) {
EGLImpl::DestroyContext(object->Display, object->Context);
}
ReleasePixelFormat(object->PixelFormat);
auto& currentContexts = CurrentContexts();
for (auto it = currentContexts.begin(); it != currentContexts.end();) {
if (it->second == ctx) {
it = currentContexts.erase(it);
} else {
++it;
}
}
Contexts().erase(ctx);
}
GLuint GetContextRetainCount(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
return object ? object->RetainCount : 0;
}
CGLPixelFormatObj GetPixelFormat(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
return object ? object->PixelFormat : nullptr;
}
CGLError SetCurrentContext(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
if (!ctx) {
CurrentContexts().erase(CurrentThreadKey());
EGLImpl::MakeCurrent(EGL_NO_DISPLAY, EGL_NO_SURFACE, EGL_NO_SURFACE, EGL_NO_CONTEXT);
return kCGLNoError;
}
auto* object = TryGetContext(ctx);
if (!object) {
return kCGLBadContext;
}
return MakeCurrentLocked(ctx, *object);
}
CGLContextObj GetCurrentContext() {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto& currentContexts = CurrentContexts();
auto it = currentContexts.find(CurrentThreadKey());
return it == currentContexts.end() ? nullptr : it->second;
}
CGLError SetParameter(CGLContextObj ctx, CGLContextParameter pname, const GLint* params) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (!object) {
return kCGLBadContext;
}
if (!params && pname != kCGLCPReclaimResources) {
return kCGLBadAddress;
}
switch (pname) {
case kCGLCPSwapInterval:
object->SwapInterval = params[0];
EGLImpl::SwapInterval(object->Display, object->SwapInterval);
return kCGLNoError;
case kCGLCPSurfaceBackingSize:
{
const GLint width = std::max<GLint>(params[0], 1);
const GLint height = std::max<GLint>(params[1], 1);
if (object->SurfaceBackingSize[0] == width && object->SurfaceBackingSize[1] == height) {
return kCGLNoError;
}
object->SurfaceBackingSize[0] = width;
object->SurfaceBackingSize[1] = height;
if (object->MetalLayer && object->Surface != EGL_NO_SURFACE) {
return ResizeSurfaceLocked(*object);
}
return kCGLNoError;
}
case kCGLCPSurfaceOpacity:
case kCGLCPSurfaceOrder:
case kCGLCPMPSwapsInFlight:
case kCGLCPReclaimResources:
return kCGLNoError;
default:
return kCGLNoError;
}
}
CGLError GetParameter(CGLContextObj ctx, CGLContextParameter pname, GLint* params) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (!object) {
return kCGLBadContext;
}
if (!params) {
return kCGLBadAddress;
}
switch (pname) {
case kCGLCPSwapInterval:
params[0] = object->SwapInterval;
return kCGLNoError;
case kCGLCPSurfaceBackingSize:
params[0] = object->SurfaceBackingSize[0];
params[1] = object->SurfaceBackingSize[1];
return kCGLNoError;
case kCGLCPCurrentRendererID:
params[0] = 0x4d474c;
return kCGLNoError;
case kCGLCPGPUVertexProcessing:
case kCGLCPGPUFragmentProcessing:
case kCGLCPHasDrawable:
params[0] = object->HasDrawable ? 1 : 0;
return kCGLNoError;
case kCGLCPMPSwapsInFlight:
params[0] = 1;
return kCGLNoError;
default:
params[0] = 0;
return kCGLNoError;
}
}
CGLError UpdateContext(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
return TryGetContext(ctx) ? kCGLNoError : kCGLBadContext;
}
CGLError ClearDrawable(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (!object) {
return kCGLBadContext;
}
if (object->Surface != EGL_NO_SURFACE) {
EGLImpl::DestroySurface(object->Display, object->Surface);
}
object->Surface = EGL_NO_SURFACE;
object->View = nullptr;
object->MetalLayer = nullptr;
object->HasDrawable = false;
return kCGLNoError;
}
CGLError FlushDrawable(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (!object) {
return kCGLBadContext;
}
if (!object->HasDrawable || object->Surface == EGL_NO_SURFACE) {
return kCGLBadDrawable;
}
const auto currentError = MakeCurrentLocked(ctx, *object);
if (currentError != kCGLNoError) {
return currentError;
}
return EGLImpl::SwapBuffers(object->Display, object->Surface) ? kCGLNoError : kCGLBadDrawable;
}
CGLError LockContext(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (!object) {
return kCGLBadContext;
}
object->Locked = true;
return kCGLNoError;
}
CGLError UnlockContext(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (!object) {
return kCGLBadContext;
}
object->Locked = false;
return kCGLNoError;
}
void GetVersion(GLint* majorvers, GLint* minorvers) {
if (majorvers) {
*majorvers = 1;
}
if (minorvers) {
*minorvers = 0;
}
}
const char* ErrorString(CGLError error) {
switch (error) {
case kCGLNoError:
return "no error";
case kCGLBadAttribute:
return "invalid pixel format attribute";
case kCGLBadPixelFormat:
return "invalid pixel format";
case kCGLBadContext:
return "invalid context";
case kCGLBadDrawable:
return "invalid drawable";
case kCGLBadState:
return "invalid context state";
case kCGLBadValue:
return "invalid numerical value";
case kCGLBadMatch:
return "invalid share context";
case kCGLBadAddress:
return "invalid pointer";
case kCGLBadAlloc:
return "invalid memory allocation";
default:
return "unknown CGL error";
}
}
CGLError AttachDrawable(CGLContextObj ctx, void* nsView, void* metalLayer, GLint width, GLint height) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (!object) {
return kCGLBadContext;
}
if (!metalLayer) {
return kCGLBadDrawable;
}
width = std::max<GLint>(width, 1);
height = std::max<GLint>(height, 1);
const Bool sameSize = object->SurfaceBackingSize[0] == width && object->SurfaceBackingSize[1] == height;
object->SurfaceBackingSize[0] = width;
object->SurfaceBackingSize[1] = height;
if (object->Surface != EGL_NO_SURFACE && object->MetalLayer == metalLayer && sameSize) {
object->View = nsView;
object->HasDrawable = true;
return kCGLNoError;
}
if (object->Surface != EGL_NO_SURFACE && object->MetalLayer == metalLayer) {
object->View = nsView;
object->HasDrawable = true;
return ResizeSurfaceLocked(*object);
}
if (object->Surface != EGL_NO_SURFACE) {
EGLImpl::DestroySurface(object->Display, object->Surface);
object->Surface = EGL_NO_SURFACE;
}
object->View = nsView;
object->MetalLayer = metalLayer;
const auto recreateError = RecreateSurfaceLocked(ctx, *object);
if (recreateError != kCGLNoError) {
return recreateError;
}
return kCGLNoError;
}
void* GetContextNSObject(CGLContextObj ctx) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
return object ? object->NSObject : nullptr;
}
void SetContextNSObject(CGLContextObj ctx, void* nsObject) {
const std::lock_guard<std::recursive_mutex> lock(RegistryMutex());
auto* object = TryGetContext(ctx);
if (object) {
object->NSObject = nsObject;
}
}
} // namespace MobileGL::MG_Impl::CGLImpl
#endif
-49
View File
@@ -1,49 +0,0 @@
// MobileGL - MobileGL/MG_Impl/CGLImpl/CGLImpl.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 <Includes.h>
#if defined(__APPLE__)
#ifndef GL_SILENCE_DEPRECATION
#define GL_SILENCE_DEPRECATION
#endif
#include <OpenGL/OpenGL.h>
namespace MobileGL::MG_Impl::CGLImpl {
CGLError ChoosePixelFormat(const CGLPixelFormatAttribute* attribs, CGLPixelFormatObj* pix, GLint* npix);
CGLError DestroyPixelFormat(CGLPixelFormatObj pix);
CGLError DescribePixelFormat(CGLPixelFormatObj pix, GLint pixNum, CGLPixelFormatAttribute attrib, GLint* value);
void ReleasePixelFormat(CGLPixelFormatObj pix);
CGLPixelFormatObj RetainPixelFormat(CGLPixelFormatObj pix);
GLuint GetPixelFormatRetainCount(CGLPixelFormatObj pix);
CGLError CreateContext(CGLPixelFormatObj pix, CGLContextObj share, CGLContextObj* ctx);
CGLError DestroyContext(CGLContextObj ctx);
CGLContextObj RetainContext(CGLContextObj ctx);
void ReleaseContext(CGLContextObj ctx);
GLuint GetContextRetainCount(CGLContextObj ctx);
CGLPixelFormatObj GetPixelFormat(CGLContextObj ctx);
CGLError SetCurrentContext(CGLContextObj ctx);
CGLContextObj GetCurrentContext();
CGLError SetParameter(CGLContextObj ctx, CGLContextParameter pname, const GLint* params);
CGLError GetParameter(CGLContextObj ctx, CGLContextParameter pname, GLint* params);
CGLError UpdateContext(CGLContextObj ctx);
CGLError ClearDrawable(CGLContextObj ctx);
CGLError FlushDrawable(CGLContextObj ctx);
CGLError LockContext(CGLContextObj ctx);
CGLError UnlockContext(CGLContextObj ctx);
void GetVersion(GLint* majorvers, GLint* minorvers);
const char* ErrorString(CGLError error);
CGLError AttachDrawable(CGLContextObj ctx, void* nsView, void* metalLayer, GLint width, GLint height);
void* GetContextNSObject(CGLContextObj ctx);
void SetContextNSObject(CGLContextObj ctx, void* nsObject);
}
#endif
@@ -1,110 +0,0 @@
// MobileGL - MobileGL/MG_Impl/CGLImpl/Exporting/Definitions.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 "../CGLImpl.h"
#if defined(__APPLE__)
MOBILEGL_CGL_API CGLError CGLChoosePixelFormat(const CGLPixelFormatAttribute* attribs,
CGLPixelFormatObj* pix,
GLint* npix) {
return MobileGL::MG_Impl::CGLImpl::ChoosePixelFormat(attribs, pix, npix);
}
MOBILEGL_CGL_API CGLError CGLDestroyPixelFormat(CGLPixelFormatObj pix) {
return MobileGL::MG_Impl::CGLImpl::DestroyPixelFormat(pix);
}
MOBILEGL_CGL_API CGLError CGLDescribePixelFormat(CGLPixelFormatObj pix,
GLint pix_num,
CGLPixelFormatAttribute attrib,
GLint* value) {
return MobileGL::MG_Impl::CGLImpl::DescribePixelFormat(pix, pix_num, attrib, value);
}
MOBILEGL_CGL_API void CGLReleasePixelFormat(CGLPixelFormatObj pix) {
MobileGL::MG_Impl::CGLImpl::ReleasePixelFormat(pix);
}
MOBILEGL_CGL_API CGLPixelFormatObj CGLRetainPixelFormat(CGLPixelFormatObj pix) {
return MobileGL::MG_Impl::CGLImpl::RetainPixelFormat(pix);
}
MOBILEGL_CGL_API GLuint CGLGetPixelFormatRetainCount(CGLPixelFormatObj pix) {
return MobileGL::MG_Impl::CGLImpl::GetPixelFormatRetainCount(pix);
}
MOBILEGL_CGL_API CGLError CGLCreateContext(CGLPixelFormatObj pix, CGLContextObj share, CGLContextObj* ctx) {
return MobileGL::MG_Impl::CGLImpl::CreateContext(pix, share, ctx);
}
MOBILEGL_CGL_API CGLError CGLDestroyContext(CGLContextObj ctx) {
return MobileGL::MG_Impl::CGLImpl::DestroyContext(ctx);
}
MOBILEGL_CGL_API CGLContextObj CGLRetainContext(CGLContextObj ctx) {
return MobileGL::MG_Impl::CGLImpl::RetainContext(ctx);
}
MOBILEGL_CGL_API void CGLReleaseContext(CGLContextObj ctx) {
MobileGL::MG_Impl::CGLImpl::ReleaseContext(ctx);
}
MOBILEGL_CGL_API GLuint CGLGetContextRetainCount(CGLContextObj ctx) {
return MobileGL::MG_Impl::CGLImpl::GetContextRetainCount(ctx);
}
MOBILEGL_CGL_API CGLPixelFormatObj CGLGetPixelFormat(CGLContextObj ctx) {
return MobileGL::MG_Impl::CGLImpl::GetPixelFormat(ctx);
}
MOBILEGL_CGL_API CGLError CGLSetCurrentContext(CGLContextObj ctx) {
return MobileGL::MG_Impl::CGLImpl::SetCurrentContext(ctx);
}
MOBILEGL_CGL_API CGLContextObj CGLGetCurrentContext(void) {
return MobileGL::MG_Impl::CGLImpl::GetCurrentContext();
}
MOBILEGL_CGL_API CGLError CGLSetParameter(CGLContextObj ctx, CGLContextParameter pname, const GLint* params) {
return MobileGL::MG_Impl::CGLImpl::SetParameter(ctx, pname, params);
}
MOBILEGL_CGL_API CGLError CGLGetParameter(CGLContextObj ctx, CGLContextParameter pname, GLint* params) {
return MobileGL::MG_Impl::CGLImpl::GetParameter(ctx, pname, params);
}
MOBILEGL_CGL_API CGLError CGLUpdateContext(CGLContextObj ctx) {
return MobileGL::MG_Impl::CGLImpl::UpdateContext(ctx);
}
MOBILEGL_CGL_API CGLError CGLClearDrawable(CGLContextObj ctx) {
return MobileGL::MG_Impl::CGLImpl::ClearDrawable(ctx);
}
MOBILEGL_CGL_API CGLError CGLFlushDrawable(CGLContextObj ctx) {
return MobileGL::MG_Impl::CGLImpl::FlushDrawable(ctx);
}
MOBILEGL_CGL_API CGLError CGLLockContext(CGLContextObj ctx) {
return MobileGL::MG_Impl::CGLImpl::LockContext(ctx);
}
MOBILEGL_CGL_API CGLError CGLUnlockContext(CGLContextObj ctx) {
return MobileGL::MG_Impl::CGLImpl::UnlockContext(ctx);
}
MOBILEGL_CGL_API void CGLGetVersion(GLint* majorvers, GLint* minorvers) {
MobileGL::MG_Impl::CGLImpl::GetVersion(majorvers, minorvers);
}
MOBILEGL_CGL_API const char* CGLErrorString(CGLError error) {
return MobileGL::MG_Impl::CGLImpl::ErrorString(error);
}
#endif
@@ -1,56 +0,0 @@
// MobileGL - MobileGL/MG_Impl/DyldInterpose/DyldInterpose.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 <Includes.h>
#if defined(__APPLE__)
#include "MG_Impl/GetProcAddress.h"
#include <dlfcn.h>
namespace {
struct DyldInterposeEntry {
const void* Replacement;
const void* Replacee;
};
bool IsGLProcName(const char* name) {
if (name == nullptr) {
return false;
}
if (strncmp(name, "CGL", 3) == 0) {
return true;
}
if (strncmp(name, "gl", 2) != 0) {
return false;
}
// Avoid stealing glfw*/glib*/glX*/global application symbols.
return name[2] >= 'A' && name[2] <= 'Z' && name[2] != 'X';
}
void* MobileGLDlsym(void* handle, const char* symbol) {
if (IsGLProcName(symbol)) {
if (void* proc = MobileGL::MG_Impl::GetProcAddress(symbol)) {
return proc;
}
}
return dlsym(handle, symbol);
}
__attribute__((used)) static const DyldInterposeEntry kMobileGLDyldInterpose[]
__attribute__((section("__DATA,__interpose"))) = {
{reinterpret_cast<const void*>(MobileGLDlsym), reinterpret_cast<const void*>(dlsym)},
};
} // namespace
#endif

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