Always ensure gpu_threads count >= warp size of 32 #8656
+37
−13
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
8 tasks
alexreinking
approved these changes
Jun 24, 2025
ac74c4b to
53b1567
Compare
Tile the image with `TailStrategy::GuardWithIf` whenever `can_parallelize(c, 3)` is encountered and when the split factor is smaller than GPU's warp size (=32). For 2D/3D tile schedules having mixed tail strategies, pick the most conservative one, i.e. `TailStrategy::GuardWithIf`. Enable `iir_blur` tests on the Github Actions/Buildbot.
53b1567 to
1452961
Compare
mcourteaux
approved these changes
Jun 24, 2025
|
The Testbench passes for targets Error message: |
antonysigma
commented
Jun 26, 2025
Comment on lines
-35
to
+48
| configure_file(${IMAGE} rgba.png COPYONLY) | ||
| add_test(NAME iir_blur_filter | ||
| COMMAND iir_blur_filter rgba.png out.png) | ||
| set_tests_properties(iir_blur_filter PROPERTIES | ||
| LABELS iir_blur | ||
| PASS_REGULAR_EXPRESSION "Success!" | ||
| SKIP_REGULAR_EXPRESSION "\\[SKIP\\]") | ||
| if (Halide_TARGET MATCHES "opencl") | ||
| # Error message: | ||
| # | ||
| # Error: OpenCL error: CL_INVALID_COMMAND_QUEUE clFinish failed | ||
| message(WARNING "Skipping Mullapudi2016's GPU auto-schedules for OpenCL target.") | ||
| else () | ||
| configure_file(${IMAGE} rgba.png COPYONLY) | ||
| add_test(NAME iir_blur_filter | ||
| COMMAND iir_blur_filter rgba.png out.png) | ||
| set_tests_properties(iir_blur_filter PROPERTIES | ||
| LABELS iir_blur | ||
| PASS_REGULAR_EXPRESSION "Success!" | ||
| SKIP_REGULAR_EXPRESSION "\\[SKIP\\]") | ||
| endif () |
There was a problem hiding this comment.
@alexreinking Help wanted to troubleshoot the CMake (re-)configure error:
[2433/4259] Building CXX object test/correctness/CMakeFiles/correctness_fuzz_simplify.dir/fuzz_simplify.cpp.o
FAILED: test/correctness/CMakeFiles/correctness_fuzz_simplify.dir/fuzz_simplify.cpp.o
/Users/halidenightly/build_bot/venv/lib/python3.9/site-packages/cmake/data/bin/cmake -E env CCACHE_CPP2=yes CCACHE_HASHDIR=yes CCACHE_SLOPPINESS=pch_defines,time_macros,include_file_mtime,include_file_ctime /opt/homebrew/bin/ccache /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/c++ -DHALIDE_ENABLE_RTTI -DHALIDE_KEEP_MACROS -DHALIDE_VERSION_MAJOR=20 -DHALIDE_VERSION_MINOR=0 -DHALIDE_VERSION_PATCH=0 -DHALIDE_WITH_EXCEPTIONS -I/Users/halidenightly/build_bot/worker/halide-testbranch-main-llvm20-arm-64-osx-cmake/halide-source/test/common -I/Users/halidenightly/build_bot/worker/halide-testbranch-main-llvm20-arm-64-osx-cmake/halide-source/tools -I/Users/halidenightly/build_bot/worker/halide-testbranch-main-llvm20-arm-64-osx-cmake/halide-build/include -I/Users/halidenightly/build_bot/worker/halide-testbranch-main-llvm20-arm-64-osx-cmake/halide-source/src/runtime -Xclang -fno-pch-timestamp -std=c++17 -arch arm64 -isysroot /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX13.1.sdk -fvisibility=hidden -fvisibility-inlines-hidden -Wall -Wvla-extension -Wcast-qual -Wignored-qualifiers -Wimplicit-fallthrough -Woverloaded-virtual -Wno-old-style-cast -Winconsistent-missing-destructor-override -Winconsistent-missing-override -Wdeprecated-declarations -Wno-double-promotion -Wno-float-conversion -Wno-float-equal -Wno-missing-field-initializers -Wno-shadow -Wno-sign-conversion -Wno-switch-enum -Wno-undef -Wno-unused-function -Wno-unused-macros -Wno-unused-parameter -Wno-c++98-compat-pedantic -Wno-c++98-compat -Wno-cast-align -Wno-comma -Wno-covered-switch-default -Wno-documentation-unknown-command -Wno-documentation -Wno-exit-time-destructors -Wno-global-constructors -Wno-implicit-float-conversion -Wno-implicit-int-conversion -Wno-implicit-int-float-conversion -Wno-missing-prototypes -Wno-nonportable-system-include-path -Wno-reserved-id-macro -Wno-shadow-field-in-constructor -Wno-shadow-field -Wno-shorten-64-to-32 -Wno-undefined-func-template -Wno-unused-member-function -Wno-unused-template -Winvalid-pch -Xclang -include-pch -Xclang /Users/halidenightly/build_bot/worker/halide-testbranch-main-llvm20-arm-64-osx-cmake/halide-build/test/CMakeFiles/_test_internal.dir/cmake_pch.hxx.pch -Xclang -include -Xclang /Users/halidenightly/build_bot/worker/halide-testbranch-main-llvm20-arm-64-osx-cmake/halide-build/test/CMakeFiles/_test_internal.dir/cmake_pch.hxx -MD -MT test/correctness/CMakeFiles/correctness_fuzz_simplify.dir/fuzz_simplify.cpp.o -MF test/correctness/CMakeFiles/correctness_fuzz_simplify.dir/fuzz_simplify.cpp.o.d -o test/correctness/CMakeFiles/correctness_fuzz_simplify.dir/fuzz_simplify.cpp.o -c /Users/halidenightly/build_bot/worker/halide-testbranch-main-llvm20-arm-64-osx-cmake/halide-source/test/correctness/fuzz_simplify.cpp
fatal error: file '/Users/halidenightly/build_bot/worker/halide-testbranch-main-llvm20-arm-64-osx-cmake/halide-build/include/Halide.h' has been modified since the precompiled header '/Users/halidenightly/build_bot/worker/halide-testbranch-main-llvm20-arm-64-osx-cmake/halide-build/test/CMakeFiles/_test_internal.dir/cmake_pch.hxx.pch' was built: size changed (was 1338765, now 1338767)
note: please rebuild precompiled header '/Users/halidenightly/build_bot/worker/halide-testbranch-main-llvm20-arm-64-osx-cmake/halide-build/test/CMakeFiles/_test_internal.dir/cmake_pch.hxx.pch'
1 error generated.
How do I skip iir_blur tests for GPU target host-opencl ?
There was a problem hiding this comment.
Or, is it simply a regression of the Buildbot server?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Tile the image with
TailStrategy::GuardWithIfwhenevercan_parallelize(c, 3)is encountered and when the split factor (3) is smaller than GPU's warp size (=32).For 2D/3D tile schedules having mixed tail strategies, pick the most conservative one, i.e.
TailStrategy::GuardWithIf.Enable
iir_blurtests on the Github Actions/Buildbot.Before:
After:
Note: We are not trying to compete with manual GPU schedules. We just want it to pass correctness tests.
cc'ed @alexreinking and @mcourteaux .