Fix stack buffer overflow when XNNPACK tensor has too many dims #3233
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/3233
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 5c94e49 with merge base 3b0f271 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D56450593 |
swolchok
requested review from
mcr229 and
cccclai
and removed request for
mcr229
There was a problem hiding this comment.
Looks good to me! Maybe I'll let @mcr229 and @digantdesai for more detailed review...
| @@ -38,9 +38,11 @@ def define_common_targets(): | |||
| preprocessor_flags = [ | |||
| # "-DENABLE_XNNPACK_PROFILING", | |||
| ], | |||
| exported_deps = [ | |||
| "//executorch/runtime/backend:interface", | |||
There was a problem hiding this comment.
Any specific reason we make it an exported_deps?
There was a problem hiding this comment.
there's an include in a header (I think XNNExecutor.h), so dependent libraries that include that header need this dep. This is what exported_deps is for.
| * LICENSE file in the root directory of this source tree. | ||
| */ | ||
|
|
||
| #include <executorch/backends/xnnpack/runtime/XNNExecutor.h> |
swolchok
force-pushed
the
export-D56450593
branch
from
c9ce128 to
c9bf908
Compare
…rch#3233) Summary: Noticed this overflow when I was looking through the XNNPACK backend. I am not very familiar with executorch or XNNPACK, so please be critical in review! In particular, my test is crashing in xnn_delete_runtime and I don't know why; it was previously getting a UBSAN failure for the stack buffer overflow before I patched XNNExecutor.cpp Differential Revision: D56450593
|
This pull request was exported from Phabricator. Differential Revision: D56450593 |
…rch#3233) Summary: Noticed this overflow when I was looking through the XNNPACK backend. I am not very familiar with executorch or XNNPACK, so please be critical in review! In particular, my test is crashing in xnn_delete_runtime and I don't know why; it was previously getting a UBSAN failure for the stack buffer overflow before I patched XNNExecutor.cpp Differential Revision: D56450593
swolchok
force-pushed
the
export-D56450593
branch
from
c9bf908 to
bbcab0a
Compare
|
This pull request was exported from Phabricator. Differential Revision: D56450593 |
…rch#3233) Summary: Noticed this overflow when I was looking through the XNNPACK backend. I am not very familiar with executorch or XNNPACK, so please be critical in review! In particular, my test is crashing in xnn_delete_runtime and I don't know why; it was previously getting a UBSAN failure for the stack buffer overflow before I patched XNNExecutor.cpp Differential Revision: D56450593
swolchok
force-pushed
the
export-D56450593
branch
from
bbcab0a to
9c7a733
Compare
|
This pull request was exported from Phabricator. Differential Revision: D56450593 |
…rch#3233) Summary: Noticed this overflow when I was looking through the XNNPACK backend. I am not very familiar with executorch or XNNPACK, so please be critical in review! In particular, my test is crashing in xnn_delete_runtime and I don't know why; it was previously getting a UBSAN failure for the stack buffer overflow before I patched XNNExecutor.cpp Differential Revision: D56450593
swolchok
force-pushed
the
export-D56450593
branch
from
9c7a733 to
5c94e49
Compare
|
This pull request was exported from Phabricator. Differential Revision: D56450593 |
|
This pull request has been merged in dc726f9. |
Summary:
Noticed this overflow when I was looking through the XNNPACK backend.
I am not very familiar with executorch or XNNPACK, so please be critical in review! In particular, my test is crashing in xnn_delete_runtime and I don't know why; it was previously getting a UBSAN failure for the stack buffer overflow before I patched XNNExecutor.cpp
Differential Revision: D56450593