-
Notifications
You must be signed in to change notification settings - Fork 7
[LLVM Integrate] Bump llvm, torch-mlir #105
Conversation
c844b26
to
4237f03
Compare
tools/aot/execute_test.template.cpp
Outdated
// ##READ_REFERENCE_TENSORS_INTO_NPY_ARRAY##// | ||
// cnpy::NpyArray refInput0 = reference_tensors["Input0"]; | ||
// cnpy::NpyArray refInput1 = reference_tensors["Input1"]; | ||
// cnpy::NpyArray refInput2 = reference_tensors["Input2"]; | ||
// cnpy::NpyArray refOutput0 = reference_tensors["Output0"]; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
(pre-emptive, I know this is very much WIP but wanted to throw it out there)
I'd avoid clang-format changes on this file, since it might break some of the match-replace patterns when we instantiate this template in aot_compile
. You might want to ignore this file from clang-format, or make sure the patterns are correctly updated.
find . -type f -name "*.cpp" -a ! -name "*execute_test.template.cpp" -o -name "*.h" | xargs clang-format -i | ||
if [ -n "$(git status --porcelain)" ]; then | ||
echo "Please run 'find . -type f -name "*.cpp" -o -name "*.h" | xargs clang-format -i' and commit changes." | ||
echo "Please run 'find . -type f -name "*.cpp" -a ! -name "*execute_test.template.cpp" -o -name "*.h" | xargs clang-format -i' and commit changes." |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Prefer using a .clang-format-ignore
file at the project root to track ignored files, to avoid this command from blowing up.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, thanks for working on this.
LLVM (Nov 13)
Torch-MLIR (Jan 21)
StableHlo (Oct 25 - No Change)
Also fix a lit test