workaround for hanging issue on A770 action #1726
Closed
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
Otherwise, we get -999 Unknown PI error after second device ref: intel/llvm#10982
ginkgo-bot
added
the
reg:ci-cd
yhmtsai
force-pushed
the
a770_action_workaround
branch
2 times, most recently
from
f76e723 to
e2a9110
Compare
yhmtsai
force-pushed
the
a770_action_workaround
branch
from
e2a9110 to
229df3c
Compare
upsj
approved these changes
Nov 16, 2024
| cmake .. -DCMAKE_INSTALL_PREFIX=install_ginkgo -DCMAKE_CXX_FLAGS="-Wpedantic -ffp-model=precise" -DCMAKE_CXX_COMPILER=${{ matrix.config.compiler }} -DCMAKE_BUILD_TYPE=${{ matrix.config.build_type }} -DGINKGO_MIXED_PRECISION=${{ matrix.config.mixed }} -DGINKGO_BUILD_CUDA=OFF -DGINKGO_BUILD_HIP=OFF -DGINKGO_BUILD_MPI=OFF -DGINKGO_DPCPP_SINGLE_MODE=ON | ||
| make -j8 | ||
| ONEAPI_DEVICE_SELECTOR=level_zero:gpu ctest -j10 --output-on-failure | ||
| cmake .. -GNinja -DCMAKE_INSTALL_PREFIX=install_ginkgo -DCMAKE_CXX_FLAGS="-Wpedantic -ffp-model=precise -fsycl-default-sub-group-size=16 -Wno-unused-command-line-argument -Wno-deprecated" -DCMAKE_CXX_COMPILER=${{ matrix.config.compiler }} -DCMAKE_BUILD_TYPE=${{ matrix.config.build_type }} -DGINKGO_MIXED_PRECISION=${{ matrix.config.mixed }} -DGINKGO_BUILD_CUDA=OFF -DGINKGO_BUILD_HIP=OFF -DGINKGO_BUILD_MPI=OFF -DGINKGO_DPCPP_SINGLE_MODE=ON |
There was a problem hiding this comment.
I would keep -Wdeprecated around, especially now that we are starting to fix those deprecated usages.
|
the workaround is not needed anymore as the machine runs normally now |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
The default sub-group size 8 for most kernels (I assume) is available on A770, but it can not run or terminiate successfully with this kind of kernels. It might be from some mismatch between driver/kernel/compiler.
We pass the
-fsycl-default-sub-group-size=16such that it does not hang for now.Interestingly, the cooperative group with sub-group size 8 works for now.
I was inspired by that because job with 8 hangs but job with 16 works.