Last active
March 17, 2025 01:01
-
-
Save AmosLewis/d123bfbad471b2f353310431cabc8265 to your computer and use it in GitHub Desktop.
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
# downdload the issue1.sh by the following wget | |
# wget https://gist.githubusercontent.com/AmosLewis/00fdb4e9a96f29c188828e3ff4ea29ef/raw/8377ffd2f53e58e12a9adae7b92b3d5a7f35d98b/bisect-issue1.sh | |
(bisect.venv) ➜ bisect git:(main) ✗ python ./bisect_packages.py \ | |
--good-ref=6eadf3df6f57a20283da8318f9178b0d92ae50eb \ | |
--bad-ref=00e88733e6b8c8cdb351d4516509f56daebdf604 \ | |
--test-script=/sharedfile/attn/bisect/issue1.sh | |
Welcome to bisect_packages.py! | |
------------------------------------------------------------------ | |
--------- Configuration ------------------------------------------ | |
------------------------------------------------------------------ | |
Searching range : '6eadf3df6f57a20283da8318f9178b0d92ae50eb' - '00e88733e6b8c8cdb351d4516509f56daebdf604' | |
Using working directory : '/home/chi/.iree/bisect' | |
Using test script : '/sharedfile/attn/bisect/issue1.sh' | |
Current platform is 'Linux-6.8.0-52-generic-x86_64-with-glibc2.35', platform.system is 'Linux'. | |
Current Python version is '3.11.1 (main, Oct 7 2024, 06:16:08) [GCC 11.4.0]'. This script requires 3.11. | |
Found gh at '/usr/bin/gh'. | |
------------------------------------------------------------------ | |
------------------------------------------------------------------ | |
--------- Running git bisect ------------------------------------- | |
------------------------------------------------------------------ | |
Bisecting: 3 revisions left to test after this (roughly 2 steps) | |
[87483506196da31623fc3c3316dcec8805bb61b0] Integrates/llvm 20250310: Bump to llvm/llvm-project@967ab7e (#20207) | |
running '/home/chi/.iree/bisect/bisect_run_script.sh' | |
++ git rev-parse BISECT_HEAD | |
+ REF_HASH=87483506196da31623fc3c3316dcec8805bb61b0 | |
+ python /home/chi/src/iree/build_tools/pkgci/bisect/../setup_venv.py /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0/.venv --artifact-path=/home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0 --fetch-git-ref=87483506196da31623fc3c3316dcec8805bb61b0 | |
Finding workflow run for ref: 87483506196da31623fc3c3316dcec8805bb61b0 | |
Using normalized ref: 87483506196da31623fc3c3316dcec8805bb61b0 | |
Running command to list workflow runs: | |
gh api -H Accept: application/vnd.github+json -H X-GitHub-Api-Version: 2022-11-28 /repos/iree-org/iree/actions/workflows/pkgci.yml/runs?head_sha=87483506196da31623fc3c3316dcec8805bb61b0 | |
Found workflow run: https://github.com/iree-org/iree/actions/runs/13796360116 | |
Installing wheels: [(PosixPath('/home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0'), 'iree-base-compiler'), (PosixPath('/home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0'), 'iree-base-runtime')] | |
Running command: /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0/.venv/bin/python -m pip install --no-deps --no-index -f /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0 --force-reinstall iree-base-compiler | |
Looking in links: /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0 | |
Processing /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0/iree_base_compiler-3.3.0.dev0+87483506196da31623fc3c3316dcec8805bb61b0-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl | |
Installing collected packages: iree-base-compiler | |
Attempting uninstall: iree-base-compiler | |
Found existing installation: iree-base-compiler 3.3.0.dev0+87483506196da31623fc3c3316dcec8805bb61b0 | |
Uninstalling iree-base-compiler-3.3.0.dev0+87483506196da31623fc3c3316dcec8805bb61b0: | |
Successfully uninstalled iree-base-compiler-3.3.0.dev0+87483506196da31623fc3c3316dcec8805bb61b0 | |
Successfully installed iree-base-compiler-3.3.0.dev0+87483506196da31623fc3c3316dcec8805bb61b0 | |
Running command: /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0/.venv/bin/python -m pip install --no-deps --no-index -f /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0 --force-reinstall iree-base-runtime | |
Looking in links: /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0 | |
Processing /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0/iree_base_runtime-3.3.0.dev0+87483506196da31623fc3c3316dcec8805bb61b0-cp311-cp311-manylinux_2_28_x86_64.whl | |
Installing collected packages: iree-base-runtime | |
Attempting uninstall: iree-base-runtime | |
Found existing installation: iree-base-runtime 3.3.0.dev0+87483506196da31623fc3c3316dcec8805bb61b0 | |
Uninstalling iree-base-runtime-3.3.0.dev0+87483506196da31623fc3c3316dcec8805bb61b0: | |
Successfully uninstalled iree-base-runtime-3.3.0.dev0+87483506196da31623fc3c3316dcec8805bb61b0 | |
Successfully installed iree-base-runtime-3.3.0.dev0+87483506196da31623fc3c3316dcec8805bb61b0 | |
venv setup complete at '/home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0/.venv'. Activate it with | |
source /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0/.venv/bin/activate | |
+ PATH=/home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0/.venv/bin:/usr/lib/git-core:/usr/lib/git-core:/sharedfile/bisect.venv/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin | |
+ set +e | |
+ iree-compile /sharedfile/attn/128/fp8_attn.mlir --iree-hip-target=gfx942 -o=/sharedfile/attn/128/fp8_attn.vmfb --iree-hal-target-device=hip --iree-dispatch-creation-enable-aggressive-fusion=true --iree-global-opt-propagate-transposes=true --iree-opt-aggressively-propagate-transposes=true --iree-opt-data-tiling=false '--iree-preprocessing-pass-pipeline=builtin.module(util.func(iree-preprocessing-generalize-linalg-matmul-experimental))' --iree-hal-indirect-command-buffers=true --iree-stream-resource-memory-model=discrete --iree-hal-memoization=true --iree-opt-strip-assertions | |
+ ROCR_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 | |
+ iree-benchmark-module --hip_use_streams=true --module=/sharedfile/attn/128/fp8_attn.vmfb --parameters=model=/sharedfile/attn/fp8_attn.irpa --device=hip://4 --function=prefill_bs4 --input=4x128xi64=@/sharedfile/128/prefill/prefill_token_ids_4x128xi64.bin --input=4xi64=@/sharedfile/128/prefill/prefill_seq_lens_4xi64.bin --input=4x4xi64=@/sharedfile/128/prefill/prefill_seq_block_ids_4x4xi64.bin --input=261x2097152xf8E4M3FNUZ=@/sharedfile/128/prefill/prefill_cache_state_261x2097152xf8E4M3FNUZ.bin --benchmark_repetitions=3 | |
2025-03-16T17:52:04-07:00 | |
Running /home/chi/.iree/bisect/87483506196da31623fc3c3316dcec8805bb61b0/.venv/lib/python3.11/site-packages/iree/_runtime_libs/iree-benchmark-module | |
Run on (96 X 3810.79 MHz CPU s) | |
CPU Caches: | |
L1 Data 32 KiB (x96) | |
L1 Instruction 32 KiB (x96) | |
L2 Unified 1024 KiB (x96) | |
L3 Unified 32768 KiB (x16) | |
Load Average: 2.80, 2.53, 3.70 | |
***WARNING*** CPU scaling is enabled, the benchmark real time measurements may be noisy and will incur extra overhead. | |
------------------------------------------------------------------------------------------------------- | |
Benchmark Time CPU Iterations UserCounters... | |
------------------------------------------------------------------------------------------------------- | |
BM_prefill_bs4/process_time/real_time 27.5 ms 28.0 ms 26 items_per_second=36.3144/s | |
BM_prefill_bs4/process_time/real_time 27.5 ms 27.9 ms 26 items_per_second=36.3179/s | |
BM_prefill_bs4/process_time/real_time 27.6 ms 27.9 ms 26 items_per_second=36.2956/s | |
BM_prefill_bs4/process_time/real_time_mean 27.5 ms 27.9 ms 3 items_per_second=36.3093/s | |
BM_prefill_bs4/process_time/real_time_median 27.5 ms 27.9 ms 3 items_per_second=36.3144/s | |
BM_prefill_bs4/process_time/real_time_stddev 0.009 ms 0.025 ms 3 items_per_second=0.0119748/s | |
BM_prefill_bs4/process_time/real_time_cv 0.03 % 0.09 % 3 items_per_second=0.03% | |
+ RET_VALUE=0 | |
+ '[' 0 -ne 0 ']' | |
Bisecting: 1 revision left to test after this (roughly 1 step) | |
[3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b] [Codegen] Set full workgroup tile sizes to 0 (#20159) | |
running '/home/chi/.iree/bisect/bisect_run_script.sh' | |
++ git rev-parse BISECT_HEAD | |
+ REF_HASH=3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
+ python /home/chi/src/iree/build_tools/pkgci/bisect/../setup_venv.py /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b/.venv --artifact-path=/home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b --fetch-git-ref=3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Finding workflow run for ref: 3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Using normalized ref: 3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Running command to list workflow runs: | |
gh api -H Accept: application/vnd.github+json -H X-GitHub-Api-Version: 2022-11-28 /repos/iree-org/iree/actions/workflows/pkgci.yml/runs?head_sha=3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Found workflow run: https://github.com/iree-org/iree/actions/runs/13797422275 | |
Installing wheels: [(PosixPath('/home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b'), 'iree-base-compiler'), (PosixPath('/home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b'), 'iree-base-runtime')] | |
Running command: /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b/.venv/bin/python -m pip install --no-deps --no-index -f /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b --force-reinstall iree-base-compiler | |
Looking in links: /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Processing /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b/iree_base_compiler-3.3.0.dev0+3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl | |
Installing collected packages: iree-base-compiler | |
Attempting uninstall: iree-base-compiler | |
Found existing installation: iree-base-compiler 3.3.0.dev0+3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Uninstalling iree-base-compiler-3.3.0.dev0+3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b: | |
Successfully uninstalled iree-base-compiler-3.3.0.dev0+3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Successfully installed iree-base-compiler-3.3.0.dev0+3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Running command: /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b/.venv/bin/python -m pip install --no-deps --no-index -f /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b --force-reinstall iree-base-runtime | |
Looking in links: /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Processing /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b/iree_base_runtime-3.3.0.dev0+3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b-cp311-cp311-manylinux_2_28_x86_64.whl | |
Installing collected packages: iree-base-runtime | |
Attempting uninstall: iree-base-runtime | |
Found existing installation: iree-base-runtime 3.3.0.dev0+3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Uninstalling iree-base-runtime-3.3.0.dev0+3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b: | |
Successfully uninstalled iree-base-runtime-3.3.0.dev0+3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
Successfully installed iree-base-runtime-3.3.0.dev0+3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b | |
venv setup complete at '/home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b/.venv'. Activate it with | |
source /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b/.venv/bin/activate | |
+ PATH=/home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b/.venv/bin:/usr/lib/git-core:/usr/lib/git-core:/sharedfile/bisect.venv/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin | |
+ set +e | |
+ iree-compile /sharedfile/attn/128/fp8_attn.mlir --iree-hip-target=gfx942 -o=/sharedfile/attn/128/fp8_attn.vmfb --iree-hal-target-device=hip --iree-dispatch-creation-enable-aggressive-fusion=true --iree-global-opt-propagate-transposes=true --iree-opt-aggressively-propagate-transposes=true --iree-opt-data-tiling=false '--iree-preprocessing-pass-pipeline=builtin.module(util.func(iree-preprocessing-generalize-linalg-matmul-experimental))' --iree-hal-indirect-command-buffers=true --iree-stream-resource-memory-model=discrete --iree-hal-memoization=true --iree-opt-strip-assertions | |
+ ROCR_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 | |
+ iree-benchmark-module --hip_use_streams=true --module=/sharedfile/attn/128/fp8_attn.vmfb --parameters=model=/sharedfile/attn/fp8_attn.irpa --device=hip://4 --function=prefill_bs4 --input=4x128xi64=@/sharedfile/128/prefill/prefill_token_ids_4x128xi64.bin --input=4xi64=@/sharedfile/128/prefill/prefill_seq_lens_4xi64.bin --input=4x4xi64=@/sharedfile/128/prefill/prefill_seq_block_ids_4x4xi64.bin --input=261x2097152xf8E4M3FNUZ=@/sharedfile/128/prefill/prefill_cache_state_261x2097152xf8E4M3FNUZ.bin --benchmark_repetitions=3 | |
2025-03-16T17:52:25-07:00 | |
Running /home/chi/.iree/bisect/3ddcef5ec1a31824353f9c1b0f94ef01ed0b784b/.venv/lib/python3.11/site-packages/iree/_runtime_libs/iree-benchmark-module | |
Run on (96 X 3810.79 MHz CPU s) | |
CPU Caches: | |
L1 Data 32 KiB (x96) | |
L1 Instruction 32 KiB (x96) | |
L2 Unified 1024 KiB (x96) | |
L3 Unified 32768 KiB (x16) | |
Load Average: 2.91, 2.58, 3.69 | |
***WARNING*** CPU scaling is enabled, the benchmark real time measurements may be noisy and will incur extra overhead. | |
------------------------------------------------------------------------------------------------------- | |
Benchmark Time CPU Iterations UserCounters... | |
------------------------------------------------------------------------------------------------------- | |
BM_prefill_bs4/process_time/real_time 29.7 ms 30.3 ms 24 items_per_second=33.6416/s | |
BM_prefill_bs4/process_time/real_time 29.7 ms 30.4 ms 24 items_per_second=33.6332/s | |
BM_prefill_bs4/process_time/real_time 29.7 ms 30.3 ms 24 items_per_second=33.6153/s | |
BM_prefill_bs4/process_time/real_time_mean 29.7 ms 30.3 ms 3 items_per_second=33.63/s | |
BM_prefill_bs4/process_time/real_time_median 29.7 ms 30.3 ms 3 items_per_second=33.6332/s | |
BM_prefill_bs4/process_time/real_time_stddev 0.012 ms 0.025 ms 3 items_per_second=0.0134274/s | |
BM_prefill_bs4/process_time/real_time_cv 0.04 % 0.08 % 3 items_per_second=0.04% | |
+ RET_VALUE=0 | |
+ '[' 0 -ne 0 ']' | |
Bisecting: 0 revisions left to test after this (roughly 0 steps) | |
[779989e89fa6e800005e6fcb0fb59f2c6d1a8839] Add `iree-opt-level` and per-pipeline optimization flags (#20047) | |
running '/home/chi/.iree/bisect/bisect_run_script.sh' | |
++ git rev-parse BISECT_HEAD | |
+ REF_HASH=779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
+ python /home/chi/src/iree/build_tools/pkgci/bisect/../setup_venv.py /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839/.venv --artifact-path=/home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839 --fetch-git-ref=779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Finding workflow run for ref: 779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Using normalized ref: 779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Running command to list workflow runs: | |
gh api -H Accept: application/vnd.github+json -H X-GitHub-Api-Version: 2022-11-28 /repos/iree-org/iree/actions/workflows/pkgci.yml/runs?head_sha=779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Found workflow run: https://github.com/iree-org/iree/actions/runs/13799689055 | |
Installing wheels: [(PosixPath('/home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839'), 'iree-base-compiler'), (PosixPath('/home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839'), 'iree-base-runtime')] | |
Running command: /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839/.venv/bin/python -m pip install --no-deps --no-index -f /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839 --force-reinstall iree-base-compiler | |
Looking in links: /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Processing /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839/iree_base_compiler-3.3.0.dev0+779989e89fa6e800005e6fcb0fb59f2c6d1a8839-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl | |
Installing collected packages: iree-base-compiler | |
Attempting uninstall: iree-base-compiler | |
Found existing installation: iree-base-compiler 3.3.0.dev0+779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Uninstalling iree-base-compiler-3.3.0.dev0+779989e89fa6e800005e6fcb0fb59f2c6d1a8839: | |
Successfully uninstalled iree-base-compiler-3.3.0.dev0+779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Successfully installed iree-base-compiler-3.3.0.dev0+779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Running command: /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839/.venv/bin/python -m pip install --no-deps --no-index -f /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839 --force-reinstall iree-base-runtime | |
Looking in links: /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Processing /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839/iree_base_runtime-3.3.0.dev0+779989e89fa6e800005e6fcb0fb59f2c6d1a8839-cp311-cp311-manylinux_2_28_x86_64.whl | |
Installing collected packages: iree-base-runtime | |
Attempting uninstall: iree-base-runtime | |
Found existing installation: iree-base-runtime 3.3.0.dev0+779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Uninstalling iree-base-runtime-3.3.0.dev0+779989e89fa6e800005e6fcb0fb59f2c6d1a8839: | |
Successfully uninstalled iree-base-runtime-3.3.0.dev0+779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
Successfully installed iree-base-runtime-3.3.0.dev0+779989e89fa6e800005e6fcb0fb59f2c6d1a8839 | |
venv setup complete at '/home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839/.venv'. Activate it with | |
source /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839/.venv/bin/activate | |
+ PATH=/home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839/.venv/bin:/usr/lib/git-core:/usr/lib/git-core:/sharedfile/bisect.venv/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin | |
+ set +e | |
+ iree-compile /sharedfile/attn/128/fp8_attn.mlir --iree-hip-target=gfx942 -o=/sharedfile/attn/128/fp8_attn.vmfb --iree-hal-target-device=hip --iree-dispatch-creation-enable-aggressive-fusion=true --iree-global-opt-propagate-transposes=true --iree-opt-aggressively-propagate-transposes=true --iree-opt-data-tiling=false '--iree-preprocessing-pass-pipeline=builtin.module(util.func(iree-preprocessing-generalize-linalg-matmul-experimental))' --iree-hal-indirect-command-buffers=true --iree-stream-resource-memory-model=discrete --iree-hal-memoization=true --iree-opt-strip-assertions | |
+ ROCR_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 | |
+ iree-benchmark-module --hip_use_streams=true --module=/sharedfile/attn/128/fp8_attn.vmfb --parameters=model=/sharedfile/attn/fp8_attn.irpa --device=hip://4 --function=prefill_bs4 --input=4x128xi64=@/sharedfile/128/prefill/prefill_token_ids_4x128xi64.bin --input=4xi64=@/sharedfile/128/prefill/prefill_seq_lens_4xi64.bin --input=4x4xi64=@/sharedfile/128/prefill/prefill_seq_block_ids_4x4xi64.bin --input=261x2097152xf8E4M3FNUZ=@/sharedfile/128/prefill/prefill_cache_state_261x2097152xf8E4M3FNUZ.bin --benchmark_repetitions=3 | |
2025-03-16T17:52:46-07:00 | |
Running /home/chi/.iree/bisect/779989e89fa6e800005e6fcb0fb59f2c6d1a8839/.venv/lib/python3.11/site-packages/iree/_runtime_libs/iree-benchmark-module | |
Run on (96 X 3810.79 MHz CPU s) | |
CPU Caches: | |
L1 Data 32 KiB (x96) | |
L1 Instruction 32 KiB (x96) | |
L2 Unified 1024 KiB (x96) | |
L3 Unified 32768 KiB (x16) | |
Load Average: 3.24, 2.67, 3.70 | |
***WARNING*** CPU scaling is enabled, the benchmark real time measurements may be noisy and will incur extra overhead. | |
------------------------------------------------------------------------------------------------------- | |
Benchmark Time CPU Iterations UserCounters... | |
------------------------------------------------------------------------------------------------------- | |
BM_prefill_bs4/process_time/real_time 29.8 ms 30.6 ms 24 items_per_second=33.5489/s | |
BM_prefill_bs4/process_time/real_time 29.8 ms 30.7 ms 24 items_per_second=33.5225/s | |
BM_prefill_bs4/process_time/real_time 29.8 ms 30.7 ms 24 items_per_second=33.5171/s | |
BM_prefill_bs4/process_time/real_time_mean 29.8 ms 30.7 ms 3 items_per_second=33.5295/s | |
BM_prefill_bs4/process_time/real_time_median 29.8 ms 30.7 ms 3 items_per_second=33.5225/s | |
BM_prefill_bs4/process_time/real_time_stddev 0.015 ms 0.056 ms 3 items_per_second=0.0170314/s | |
BM_prefill_bs4/process_time/real_time_cv 0.05 % 0.18 % 3 items_per_second=0.05% | |
+ RET_VALUE=0 | |
+ '[' 0 -ne 0 ']' | |
00e88733e6b8c8cdb351d4516509f56daebdf604 is the first bad commit | |
commit 00e88733e6b8c8cdb351d4516509f56daebdf604 | |
Author: Quinn Dawkins <[email protected]> | |
Date: Tue Mar 11 19:17:21 2025 -0400 | |
Revert "[AMDGPU] Do not rewrite or approximate math functions on ROCm" (#20215) | |
Reverts iree-org/iree#19970 due to failures on rdna3: | |
https://github.com/iree-org/iree/actions/runs/13796654893/job/38590777349 | |
``` | |
test_pow_types_int64_int64::model.mlir::model.mlir::gpu_rocm_rdna3 _ | |
EXEC @test_pow_types_int64_int64 | |
[FAILED] result[0]: element at index 1 (31) does not match the expected (32) | |
expected: | |
3xi64=1 32 729 | |
actual: | |
3xi64=1 31 729 | |
``` | |
.../compiler/Codegen/Common/MathTransformPass.cpp | 12 +- | |
.../Codegen/Common/test/math_transform.mlir | 39 ++--- | |
tests/e2e/math/math_ops_llvm-cpu.json | 96 ++++++------ | |
tests/e2e/math/math_ops_rocm.json | 172 +++++++++++---------- | |
4 files changed, 151 insertions(+), 168 deletions(-) | |
bisect found first bad commit | |
(bisect.venv) ➜ bisect git:(main) ✗ python ./bisect_packages.py \ | |
--good-ref=6eadf3df6f57a20283da8318f9178b0d92ae50eb \ | |
--bad-ref=87483506196da31623fc3c3316dcec8805bb61b0 \ | |
--test-script=/sharedfile/attn/bisect/issue1.sh | |
Welcome to bisect_packages.py! | |
------------------------------------------------------------------ | |
--------- Configuration ------------------------------------------ | |
------------------------------------------------------------------ | |
Searching range : '6eadf3df6f57a20283da8318f9178b0d92ae50eb' - '87483506196da31623fc3c3316dcec8805bb61b0' | |
Using working directory : '/home/chi/.iree/bisect' | |
Using test script : '/sharedfile/attn/bisect/issue1.sh' | |
Current platform is 'Linux-6.8.0-52-generic-x86_64-with-glibc2.35', platform.system is 'Linux'. | |
Current Python version is '3.11.1 (main, Oct 7 2024, 06:16:08) [GCC 11.4.0]'. This script requires 3.11. | |
Found gh at '/usr/bin/gh'. | |
------------------------------------------------------------------ | |
------------------------------------------------------------------ | |
--------- Running git bisect ------------------------------------- | |
------------------------------------------------------------------ | |
Bisecting: 0 revisions left to test after this (roughly 1 step) | |
[63ed28bca54c61d85180a6be1821f356477c01ab] [GPU] Add support for conv padding when decomposing im2col (#20203) | |
running '/home/chi/.iree/bisect/bisect_run_script.sh' | |
++ git rev-parse BISECT_HEAD | |
+ REF_HASH=63ed28bca54c61d85180a6be1821f356477c01ab | |
+ python /home/chi/src/iree/build_tools/pkgci/bisect/../setup_venv.py /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/.venv --artifact-path=/home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab --fetch-git-ref=63ed28bca54c61d85180a6be1821f356477c01ab | |
Finding workflow run for ref: 63ed28bca54c61d85180a6be1821f356477c01ab | |
Using normalized ref: 63ed28bca54c61d85180a6be1821f356477c01ab | |
Running command to list workflow runs: | |
gh api -H Accept: application/vnd.github+json -H X-GitHub-Api-Version: 2022-11-28 /repos/iree-org/iree/actions/workflows/pkgci.yml/runs?head_sha=63ed28bca54c61d85180a6be1821f356477c01ab | |
Found workflow run: https://github.com/iree-org/iree/actions/runs/13794855550 | |
Package iree-base-compiler not found in cache. Fetching from linux_x86_64_release_packages... | |
Fetching artifacts for workflow run: 13794855550 | |
Found artifacts: | |
linux_x86_64_release_packages: /repos/iree-org/iree/actions/artifacts/2732137639/zip | |
Downloading artifact /repos/iree-org/iree/actions/artifacts/2732137639/zip | |
Extracting /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/linux_x86_64_release_packages.zip | |
Installing wheels: [(PosixPath('/home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab'), 'iree-base-compiler'), (PosixPath('/home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab'), 'iree-base-runtime')] | |
Creating venv at /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/.venv | |
Running command: /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/.venv/bin/python -m pip install --no-deps --no-index -f /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab --force-reinstall iree-base-compiler | |
Looking in links: /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab | |
Processing /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/iree_base_compiler-3.3.0.dev0+63ed28bca54c61d85180a6be1821f356477c01ab-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl | |
Installing collected packages: iree-base-compiler | |
Successfully installed iree-base-compiler-3.3.0.dev0+63ed28bca54c61d85180a6be1821f356477c01ab | |
Running command: /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/.venv/bin/python -m pip install --no-deps --no-index -f /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab --force-reinstall iree-base-runtime | |
Looking in links: /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab | |
Processing /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/iree_base_runtime-3.3.0.dev0+63ed28bca54c61d85180a6be1821f356477c01ab-cp311-cp311-manylinux_2_28_x86_64.whl | |
Installing collected packages: iree-base-runtime | |
Successfully installed iree-base-runtime-3.3.0.dev0+63ed28bca54c61d85180a6be1821f356477c01ab | |
venv setup complete at '/home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/.venv'. Activate it with | |
source /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/.venv/bin/activate | |
+ PATH=/home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/.venv/bin:/usr/lib/git-core:/usr/lib/git-core:/sharedfile/bisect.venv/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin | |
+ set +e | |
+ iree-compile /sharedfile/attn/128/fp8_attn.mlir --iree-hip-target=gfx942 -o=/sharedfile/attn/128/fp8_attn.vmfb --iree-hal-target-device=hip --iree-dispatch-creation-enable-aggressive-fusion=true --iree-global-opt-propagate-transposes=true --iree-opt-aggressively-propagate-transposes=true --iree-opt-data-tiling=false '--iree-preprocessing-pass-pipeline=builtin.module(util.func(iree-preprocessing-generalize-linalg-matmul-experimental))' --iree-hal-indirect-command-buffers=true --iree-stream-resource-memory-model=discrete --iree-hal-memoization=true --iree-opt-strip-assertions | |
+ ROCR_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 | |
+ iree-benchmark-module --hip_use_streams=true --module=/sharedfile/attn/128/fp8_attn.vmfb --parameters=model=/sharedfile/attn/fp8_attn.irpa --device=hip://4 --function=prefill_bs4 --input=4x128xi64=@/sharedfile/128/prefill/prefill_token_ids_4x128xi64.bin --input=4xi64=@/sharedfile/128/prefill/prefill_seq_lens_4xi64.bin --input=4x4xi64=@/sharedfile/128/prefill/prefill_seq_block_ids_4x4xi64.bin --input=261x2097152xf8E4M3FNUZ=@/sharedfile/128/prefill/prefill_cache_state_261x2097152xf8E4M3FNUZ.bin --benchmark_repetitions=3 | |
2025-03-16T17:56:28-07:00 | |
Running /home/chi/.iree/bisect/63ed28bca54c61d85180a6be1821f356477c01ab/.venv/lib/python3.11/site-packages/iree/_runtime_libs/iree-benchmark-module | |
Run on (96 X 3810.79 MHz CPU s) | |
CPU Caches: | |
L1 Data 32 KiB (x96) | |
L1 Instruction 32 KiB (x96) | |
L2 Unified 1024 KiB (x96) | |
L3 Unified 32768 KiB (x16) | |
Load Average: 2.70, 2.60, 3.44 | |
***WARNING*** CPU scaling is enabled, the benchmark real time measurements may be noisy and will incur extra overhead. | |
------------------------------------------------------------------------------------------------------- | |
Benchmark Time CPU Iterations UserCounters... | |
------------------------------------------------------------------------------------------------------- | |
BM_prefill_bs4/process_time/real_time 26.4 ms 27.2 ms 27 items_per_second=37.9342/s | |
BM_prefill_bs4/process_time/real_time 26.4 ms 27.3 ms 27 items_per_second=37.893/s | |
BM_prefill_bs4/process_time/real_time 26.2 ms 27.1 ms 27 items_per_second=38.0967/s | |
BM_prefill_bs4/process_time/real_time_mean 26.3 ms 27.2 ms 3 items_per_second=37.9746/s | |
BM_prefill_bs4/process_time/real_time_median 26.4 ms 27.2 ms 3 items_per_second=37.9342/s | |
BM_prefill_bs4/process_time/real_time_stddev 0.075 ms 0.101 ms 3 items_per_second=0.107699/s | |
BM_prefill_bs4/process_time/real_time_cv 0.28 % 0.37 % 3 items_per_second=0.28% | |
+ RET_VALUE=0 | |
+ '[' 0 -ne 0 ']' | |
87483506196da31623fc3c3316dcec8805bb61b0 is the first bad commit | |
commit 87483506196da31623fc3c3316dcec8805bb61b0 | |
Author: Bangtian Liu <[email protected]> | |
Date: Tue Mar 11 15:18:06 2025 -0400 | |
Integrates/llvm 20250310: Bump to llvm/llvm-project@967ab7e (#20207) | |
bump to | |
llvm@[967ab7e08e62a35cc65f34e21fbeb00abf3eb83f](https://github.com/iree-org/llvm-project/commit/967ab7e08e62a35cc65f34e21fbeb00abf3eb83f) | |
create one revert | |
commit@[c190c8d](https://github.com/iree-org/llvm-project/commit/c190c8df3a501b21f7f018ec3dd4293651736560) | |
to this PR https://github.com/llvm/llvm-project/pull/129850. | |
add the commit | |
[@d80a859](https://github.com/iree-org/llvm-project/commit/d80a85942af26ebfa883054136eb664d13f2714b) | |
to fix bazel test error. | |
PR #129850 attempted to use TableGen to define | |
OneShotBufferizePassOption for controlling the bufferization of all | |
operations. However, it appears that the TableGen implementation does | |
not yet fully support all the options available in the manually defined | |
OneShotBufferizationOptions struct. | |
For the usage of all the vector transformation options, follow Jakub's | |
suggestion below to pick up the default values instead of setting them | |
manually. | |
```c++ | |
VectorTransformsOptions defaultOptions; | |
... | |
vector::populateVectorTransposeLoweringPatterns( | |
patterns, defaultOptions.vectorTransposeLowering); | |
``` | |
--------- | |
Signed-off-by: Bangtian Liu <[email protected]> | |
.../InputConversion/test/apply_pdl_patterns_tosa.mlir | 10 ++++++---- | |
.../TOSA/InputConversion/test/convert_i48_to_i64.mlir | 4 ++-- | |
.../input/TOSA/InputConversion/test/tosa.pdl.mlir | 6 +++++- | |
compiler/plugins/target/CUDA/CUDATarget.cpp | 2 +- | |
compiler/plugins/target/LLVMCPU/LLVMCPUTarget.cpp | 2 +- | |
.../src/iree/compiler/Codegen/LLVMCPU/ConvertToLLVM.cpp | 5 +++-- | |
.../Codegen/LLVMCPU/LLVMCPUVectorTransposeLowering.cpp | 4 ++-- | |
.../Codegen/LLVMCPU/LLVMCPUVirtualVectorLowering.cpp | 2 +- | |
.../LLVMCPU/test/select_x86_64_lowering_strategy.mlir | 2 +- | |
.../src/iree/compiler/Codegen/LLVMGPU/ConvertToNVVM.cpp | 9 +++++---- | |
.../src/iree/compiler/Codegen/LLVMGPU/ConvertToROCDL.cpp | 9 +++++---- | |
.../compiler/Codegen/LLVMGPU/LLVMGPUVectorLowering.cpp | 7 ++++--- | |
.../compiler/Codegen/SPIRV/SPIRVFinalVectorLowering.cpp | 6 ++++-- | |
.../Codegen/SPIRV/SPIRVInitialVectorLowering.cpp | 6 ++++-- | |
.../iree/compiler/Dialect/HAL/Utils/LLVMLinkerUtils.cpp | 2 +- | |
samples/custom_dispatch/cpu/mlp_plugin/mlp_tosa.mlir | 10 ++++++---- | |
.../cpu/mlp_plugin/mlp_tosa_spec.pdl.mlir | 6 +++++- | |
tests/e2e/tosa_ops/const.mlir | 2 +- | |
tests/e2e/tosa_ops/gather.mlir | 4 ++-- | |
tests/e2e/tosa_ops/matmul.mlir | 8 ++++++-- | |
tests/e2e/tosa_ops/mul.mlir | 4 ++-- | |
tests/e2e/tosa_ops/mul_shift.mlir | 2 +- | |
tests/e2e/tosa_ops/pad.mlir | 15 +++++++++------ | |
tests/e2e/tosa_ops/reshape.mlir | 6 +++--- | |
tests/e2e/tosa_ops/while.mlir | 16 ++++++++-------- | |
third_party/llvm-project | 2 +- | |
26 files changed, 89 insertions(+), 62 deletions(-) | |
bisect found first bad commit |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment