Skip to content

[AIROCMLIR-43] tuningRunner improvements - Add state file for crash recovery and improve logging #2208

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

Open
mirza-halilcevic wants to merge 34 commits into
base: develop
Choose a base branch
from
+1,093 −434
Open

[AIROCMLIR-43] tuningRunner improvements - Add state file for crash recovery and improve logging #2208

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
34 commits
Select commit Hold shift + click to select a range
9c0838e
Improve time and metadata tracking, output file format has changed.
mirza-halilcevic Jan 10, 2026
9f6c671
Add support for stdin.
mirza-halilcevic Jan 10, 2026
7eb9851
Revert to old output format.
mirza-halilcevic Jan 10, 2026
44ead00
Add state file for crash and interrupt recovery.
mirza-halilcevic Jan 10, 2026
8a93456
Add --retry-failed option.
mirza-halilcevic Jan 10, 2026
aeedb2a
Use proper python logger.
mirza-halilcevic Jan 11, 2026
d54ab9e
Improve readability of output.
mirza-halilcevic Jan 12, 2026
a2c338e
Log warnings from tuning driver.
mirza-halilcevic Jan 12, 2026
58bada8
Reintroduce --quiet flag.
mirza-halilcevic Jan 17, 2026
1cf2e36
Let important exceptions propagate and clean up code.
mirza-halilcevic Jan 18, 2026
110dccb
Simplify state file and support multiple contexts.
mirza-halilcevic Jan 18, 2026
90132b2
Address copilot comments.
mirza-halilcevic Jan 18, 2026
b781b40
Show tuning-driver output during failures.
mirza-halilcevic Jan 19, 2026
056b12c
Add --status option.
mirza-halilcevic Jan 19, 2026
baa2279
Merge remote-tracking branch 'origin/develop' into tuning-logging
mirza-halilcevic Jan 21, 2026
7041c64
Improve order of logs for easier tracking.
mirza-halilcevic Jan 21, 2026
8f4259e
Update github ci python version.
mirza-halilcevic Jan 21, 2026
daacd66
Merge branch 'develop' into tuning-logging
mirza-halilcevic Jan 23, 2026
58f4f2f
Merge remote-tracking branch 'origin/develop' into tuning-logging
mirza-halilcevic Jan 27, 2026
e6cb102
Merge remote-tracking branch 'origin/develop' into tuning-logging
mirza-halilcevic Feb 2, 2026
aeb430c
Fix state transitions.
mirza-halilcevic Feb 2, 2026
c39a3d2
Add timeout option.
mirza-halilcevic Feb 2, 2026
2b6e394
Improve output file format.
mirza-halilcevic Feb 3, 2026
f1a885f
Simplify output file writing.
mirza-halilcevic Feb 3, 2026
03ffa08
Merge branch 'develop' into tuning-logging
mirza-halilcevic Feb 4, 2026
aa6a586
Merge branch 'develop' into tuning-logging
mirza-halilcevic Feb 4, 2026
b2dc435
Address code review comments.
mirza-halilcevic Feb 5, 2026
2b562ae
Merge remote-tracking branch 'origin/tuning-logging' into tuning-logging
mirza-halilcevic Feb 5, 2026
d964457
Use llvm dbgs instead of errs where appropriate.
mirza-halilcevic Feb 5, 2026
408100a
Warn if env vars are set.
mirza-halilcevic Feb 5, 2026
cd459b2
Merge remote-tracking branch 'origin/develop' into tuning-logging
mirza-halilcevic Feb 6, 2026
54eec0c
Merge branch 'develop' into tuning-logging
mirza-halilcevic Feb 6, 2026
7ef8977
Merge branch 'develop' into tuning-logging
mirza-halilcevic Feb 10, 2026
73f5725
Merge branch 'develop' into tuning-logging
mirza-halilcevic Feb 13, 2026
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
2 changes: 1 addition & 1 deletion .github/workflows/ci.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,7 @@ jobs:
py-checks:
runs-on: ubuntu-latest
container:
image: python:3.8
image: python:3.10
options: --user root
steps:
- uses: actions/checkout@v4
Expand Down
1 change: 0 additions & 1 deletion mlir/lib/Dialect/Rock/Tuning/RockTuningImpl.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -974,7 +974,6 @@ createTunableParamSpace(ModuleOp mod, TuningParamSetKind kind,
// greedy is not implemented for non-accel
if (!archInfo.isAccel(op) && kind == TuningParamSetKind::Greedy) {
kind = TuningParamSetKind::Exhaustive;
// TODO: tuningRunner hides this warning
llvm::errs() << "Greedy tuning not implemented for non-accel, using "
"Exhaustive instead\n";
}
Expand Down
4 changes: 2 additions & 2 deletions mlir/utils/jenkins/Jenkinsfile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1179,10 +1179,10 @@ PY
stage("Tune Fusion") {
dir('build') {
// Tune resnet50
sh """python3 ./bin/tuningRunner.py --quiet --abort-on-error --op fusion --test-dir ../mlir/test/fusion/resnet50-e2e/ -o tuning_fusion_${CHIP}.tsv"""
sh """python3 ./bin/tuningRunner.py --abort-on-error --op fusion --test-dir ../mlir/test/fusion/resnet50-e2e/ -o tuning_fusion_${CHIP}.tsv"""

// Tune bert
sh """python3 ./bin/tuningRunner.py --quiet --abort-on-error --op fusion --test-dir ../mlir/test/xmir/bert-torch-tosa-e2e/ -o tuning_fusion_${CHIP}.tsv"""
sh """python3 ./bin/tuningRunner.py --abort-on-error --op fusion --test-dir ../mlir/test/xmir/bert-torch-tosa-e2e/ -o tuning_fusion_${CHIP}.tsv"""
}
sh 'rm -f build/CMakeCache.txt'
}
Expand Down
4 changes: 2 additions & 2 deletions mlir/utils/jenkins/Jenkinsfile.downstream
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -150,12 +150,12 @@ pipeline {
dir('build') {
timeout(time: 60, activity: true, unit: 'MINUTES') {
// Tune gemms, fail if the DB is not created
sh """python3 ./bin/tuningRunner.py --quiet --abort-on-error \
sh """python3 ./bin/tuningRunner.py --abort-on-error \
--operation gemm \
--configs-file=../mlir/utils/jenkins/ci-configs/selected-gemm-configs \
--output=tuning_gemm.tsv
[ -f tuning_gemm.tsv ]"""
sh """python3 ./bin/tuningRunner.py --quiet --abort-on-error \
sh """python3 ./bin/tuningRunner.py --abort-on-error \
--operation conv \
--configs-file=../mlir/utils/jenkins/ci-configs/selected-conv-configs \
--output=tuning_conv.tsv
Expand Down
Loading