Skip to content

Workaround GCC AVX512 movemask truncation issue #574

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
stefanatwork merged 2 commits into from
Dec 4, 2025
+19 −4
Merged

Workaround GCC AVX512 movemask truncation issue #574

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
6e85acc
Workaround GCC AVX512 movemask truncation issue
stefanatwork Dec 2, 2025
ebc6ad0
ci: add Ubuntu 24.04 GCC AVX512 test job
kraszkow Dec 3, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 14 additions & 4 deletions .github/workflows/continuous.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,11 +13,11 @@ jobs:
# GPU tests
################################################################################
###############################################################################

########################################
# Linux DG2
########################################

linux-GPU-JIT-build:
secrets: inherit
uses: intel-innersource/libraries.devops.renderkit.workflows/.github/workflows/docker.yml@main
Expand Down Expand Up @@ -153,7 +153,7 @@ jobs:
cmd: |
cmake --preset windows-DG2-JIT-INTERNAL-L0RTAS -DCMAKE_BUILD_TYPE=Release -DEMBREE_TESTING_INTENSITY=3
cmake --build build --config Release --target test_package

windows-DG2-JIT-INTERNAL-test:
secrets: inherit
uses: intel-innersource/libraries.devops.renderkit.workflows/.github/workflows/windows.yml@main
Expand Down Expand Up @@ -275,6 +275,16 @@ jobs:
cmake --build build --config RelWithDebInfo --target build
cmake --build build --config RelWithDebInfo --target test_package

linux-ubuntu24_04-GCC:
secrets: inherit
uses: intel-innersource/libraries.devops.renderkit.workflows/.github/workflows/docker.yml@main
with:
image: embree/ubuntu:24.04
runs-on: '[ "Linux", "docker", "avx512" ]'
cmd: |
cmake -B build -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTING=ON -DEMBREE_TESTING_INSTALL_TESTS=ON -DCMAKE_INSTALL_TESTDIR="testing" -DEMBREE_TESTING_INTENSITY=3
cmake --build build --config Release --target build
cmake --build build --config Release --target test_package

########################################
# Windows
Expand Down Expand Up @@ -337,7 +347,7 @@ jobs:
cmake --preset macos-x64-CLANG -DCMAKE_BUILD_TYPE=RelWithDebInfo -DEMBREE_TESTING_INTENSITY=2
cmake --build build --config RelWithDebInfo --target build
cmake --build build --config RelWithDebInfo --target test_package

macosx-ARM-CLANG-NEON-AVX2-universal-binaries:
secrets: inherit
uses: intel-innersource/libraries.devops.renderkit.workflows/.github/workflows/macos.yml@main
Expand Down
5 changes: 5 additions & 0 deletions kernels/bvh/bvh_intersector_hybrid.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -796,7 +796,12 @@ namespace embree
#endif

/* return if there are no valid rays */
#if defined(__GNUC__) && (__GNUC__ < 14)
/* Before GCC 14, AVX512 masks could end up broken by optimization */
unsigned short valid_bits = movemask(valid) & 0xffff;
#else
size_t valid_bits = movemask(valid);
#endif
if (unlikely(valid_bits == 0)) return;

/* verify correct input */
Expand Down