Conversation
…ytorch#6682) move fd loader Co-authored-by: lind <lind@devvm053.nha0.facebook.com>
Differential Revision: D65481575 Pull Request resolved: pytorch#6637
Differential Revision: D65067790 Pull Request resolved: pytorch#6494
Differential Revision: D65504497 Pull Request resolved: pytorch#6677
…#6660) Change-Id: I80953a699e4861b901af4b2fb17d47d3d7efcedd Signed-off-by: Per Åstrand <per.astrand@arm.com>
Differential Revision: D65552835 Pull Request resolved: pytorch#6697
Differential Revision: D65454239 Pull Request resolved: pytorch#6651
Differential Revision: D65057149 Pull Request resolved: pytorch#6580
Differential Revision: D65306801 Pull Request resolved: pytorch#6648
Pull Request resolved: pytorch#6675 ## Context Copy-pasted from the newly added `maybe_fake_u16vec3` function in the codegen script: > There is a latency benefit to using u16vecn variables to store texture position variables instead of ivecn, likely due to reduced register pressure. However, SwiftShader does not support 16 bit integer types in shaders, so this is a crude way to fallback to using ivecn to store texture positions so that testing with SwiftShader is still possible. ghstack-source-id: 252234981 @exported-using-ghexport Differential Revision: [D65501674](https://our.internmc.facebook.com/intern/diff/D65501674/) Co-authored-by: Stephen Jia <ssjia@meta.com>
Pull Request resolved: pytorch#6524 Earlier only android platofrms had support for sleef ghstack-source-id: 252186435 @exported-using-ghexport //oss lint broken on unrelated issue @bypass-github-export-checks @exported-using-ghexport Differential Revision: [D64571782](https://our.internmc.facebook.com/intern/diff/D64571782/) Co-authored-by: Kimish Patel <kimishpatel@fb.com>
* Add max_pool2d op to Arm backend. - Adds node visitor and unittests - Adds remove_getitem_op pass to convert (maxpool_get inidices + getitem) -> maxpool2d op * Expected failures only for FVP
This change makes it uneccessary to dump intermediates by default for running the reference_model
Adding mean and where ops optimized on HiFi Co-authored-by: dijopaul <dijopaul@cadence.com>
fix seg fault Co-authored-by: lind <lind@devvm053.nha0.facebook.com>
Differential Revision: D65188297 Pull Request resolved: pytorch#6686
Differential Revision: D65608097 Pull Request resolved: pytorch#6683
Pull Request resolved: pytorch#6712 As title. Currently, the compile option to skip memory metadata tagging is not being passed correctly to `vulkan_preprocess`. ghstack-source-id: 252359943 @exported-using-ghexport Differential Revision: [D65600049](https://our.internmc.facebook.com/intern/diff/D65600049/) Co-authored-by: Stephen Jia <ssjia@meta.com>
Revert "Run tosa_reference_model using python binding (pytorch#6658)" This reverts commit 4bbe994.
Differential Revision: D65370235 Pull Request resolved: pytorch#6724
Differential Revision: D65447532 Pull Request resolved: pytorch#6647
Differential Revision: D65104400 Pull Request resolved: pytorch#6563
Add TOSA specification details to the Arm Backend. * Mandate the need for a TOSA version in the compile spec list passed to the Arm backend and propagate the information to node visitors for serialization handling. * Add TOSA version string to all TOSA tests * Adds handling of TOSA 0.80 BI and MI profile as separate serialization handlers for ADD as an example. Signed-off-by: Per Åstrand <per.astrand@arm.com>
Differential Revision: D65684570 Pull Request resolved: pytorch#6740
…6744) * Update torchao pinned commit to latest * Correctly set _GLIBCXX_USE_CXX11_ABI pybind compile options * Remove unwanted change * Update CMakeLists.txt Co-authored-by: Nikita Shulga <2453524+malfet@users.noreply.github.com> * Update CMakeLists.txt to remove the empty else branch --------- Co-authored-by: Nikita Shulga <2453524+malfet@users.noreply.github.com>
Differential Revision: D65268257 Pull Request resolved: pytorch#6590
Differential Revision: D65382714 Pull Request resolved: pytorch#6699
ynimmagahad a problem deploying
to
update-commit-hash
December 7, 2024 02:55 — with 
GitHub Actions
Failure
ynimmagarequested a deployment
to
upload-benchmark-results
December 8, 2024 00:50 — with GitHub Actions
Queued
ynimmagarequested a deployment
to
upload-benchmark-results
December 8, 2024 02:41 — with GitHub Actions
Queued
ynimmagahad a problem deploying
to
update-commit-hash
December 8, 2024 03:03 — with GitHub Actions
Failure
ynimmagahad a problem deploying
to
update-commit-hash
December 9, 2024 03:01 — with GitHub Actions
Failure
ynimmagarequested a deployment
to
upload-benchmark-results
December 10, 2024 00:46 — with GitHub Actions
Queued
ynimmagarequested a deployment
to
upload-benchmark-results
December 10, 2024 02:37 — with GitHub Actions
Queued
ynimmagahad a problem deploying
to
update-commit-hash
December 10, 2024 02:59 — with GitHub Actions
Failure
ynimmagahad a problem deploying
to
update-commit-hash
December 11, 2024 02:57 — with GitHub Actions
Failure
ynimmagarequested a deployment
to
upload-benchmark-results
December 12, 2024 00:44 — with GitHub Actions
Queued
ynimmagarequested a deployment
to
upload-benchmark-results
December 12, 2024 02:35 — with GitHub Actions
Queued
ynimmagahad a problem deploying
to
update-commit-hash
December 12, 2024 02:57 — with GitHub Actions
Failure
ynimmagahad a problem deploying
to
update-commit-hash
December 13, 2024 02:59 — with GitHub Actions
Failure
ynimmagarequested a deployment
to
upload-benchmark-results
December 14, 2024 00:43 — with GitHub Actions
Queued
ynimmagarequested a deployment
to
upload-benchmark-results
December 14, 2024 02:24 — with GitHub Actions
Queued
ynimmagahad a problem deploying
to
update-commit-hash
December 14, 2024 02:52 — with GitHub Actions
Failure
ynimmagahad a problem deploying
to
update-commit-hash
December 15, 2024 03:02 — with GitHub Actions
Failure
ynimmagarequested a deployment
to
upload-benchmark-results
December 16, 2024 00:47 — with GitHub Actions
Queued
ynimmagarequested a deployment
to
upload-benchmark-results
December 16, 2024 02:39 — with GitHub Actions
Queued
ynimmagahad a problem deploying
to
update-commit-hash
December 16, 2024 03:01 — with GitHub Actions
Failure
ynimmagahad a problem deploying
to
update-commit-hash
December 17, 2024 02:56 — with GitHub Actions
Failure
ynimmagarequested a deployment
to
upload-benchmark-results
December 18, 2024 00:43 — with GitHub Actions
Queued
ynimmagarequested a deployment
to
upload-benchmark-results
December 18, 2024 02:23 — with GitHub Actions
Queued
ynimmagahad a problem deploying
to
update-commit-hash
December 18, 2024 02:52 — with GitHub Actions
Failure
ynimmagahad a problem deploying
to
update-commit-hash
December 19, 2024 02:52 — with GitHub Actions
Failure
ynimmagahad a problem deploying
to
upload-benchmark-results
December 20, 2024 13:35 — with GitHub Actions
Failure
ynimmagahad a problem deploying
to
upload-benchmark-results
December 20, 2024 13:35 — with GitHub Actions
Failure
ynimmaga pushed a commit
that referenced
this pull request
Mar 21, 2025
) In preparation for replacing the checked-in files in shim/ with https://github.com/facebook/buck2-shims-meta as a submodule. The portions of this diff that are not a pure move are: creation of shim_et/BUCK symlink shim/tools to shim_et/tools so that fbsource//tools movement of Android toolchains from shim/BUCK to shim_et/BUCK because they are not present in buck2-shims-meta edits to .buckconfig edit to Utils.cmake to point at shim_et// instead of shim// edit to shim_et/README.md (note that shim/README.md will come back when shim/ is a submodule pointing to buck2-shims-meta)
suryasidd pushed a commit
that referenced
this pull request
Aug 14, 2025
Differential Revision: D76157744 Pull Request resolved: pytorch#11501
suryasidd pushed a commit
that referenced
this pull request
Aug 14, 2025
BNNS copy crashes the process when the dtypes differ (pytorch#11714). With the example in this PR (pytorch#11714), we crash the process on main. Here is the stack trace from LLDB: ``` Process 19234 stopped * thread #1, queue = 'com.apple.main-thread', stop reason = signal SIGABRT frame #0: 0x0000000190ac9388 libsystem_kernel.dylib`__pthread_kill + 8 libsystem_kernel.dylib`__pthread_kill: -> 0x190ac9388 <+8>: b.lo 0x190ac93a8 ; <+40> 0x190ac938c <+12>: pacibsp 0x190ac9390 <+16>: stp x29, x30, [sp, #-0x10]! 0x190ac9394 <+20>: mov x29, sp (lldb) bt * thread #1, queue = 'com.apple.main-thread', stop reason = signal SIGABRT * frame #0: 0x0000000190ac9388 libsystem_kernel.dylib`__pthread_kill + 8 frame #1: 0x0000000190b0288c libsystem_pthread.dylib`pthread_kill + 296 frame #2: 0x0000000190a0bc60 libsystem_c.dylib`abort + 124 frame #3: 0x0000000190910174 libsystem_malloc.dylib`malloc_vreport + 892 frame #4: 0x0000000190913c90 libsystem_malloc.dylib`malloc_report + 64 frame #5: 0x000000019091821c libsystem_malloc.dylib`___BUG_IN_CLIENT_OF_LIBMALLOC_POINTER_BEING_FREED_WAS_NOT_ALLOCATED + 32 frame #6: 0x000000019d2f4084 libBNNS.dylib`___lldb_unnamed_symbol1620 + 564 frame #7: 0x000000019d2f5bac libBNNS.dylib`___lldb_unnamed_symbol1628 + 680 frame #8: 0x000000019d69ce48 libBNNS.dylib`BNNSCopy + 616 frame #9: 0x000000030c74d950 _portable_lib.cpython-310-darwin.so`(anonymous namespace)::copy_using_bnns(executorchcoreml::MultiArray const&, executorchcoreml::MultiArray&) + 188 frame #10: 0x000000030c74cfdc _portable_lib.cpython-310-darwin.so`(anonymous namespace)::copy(executorchcoreml::MultiArray const&, executorchcoreml::MultiArray&, executorchcoreml::MultiArray::CopyOptions) + 72 frame #11: 0x000000030c74ceec _portable_lib.cpython-310-darwin.so`executorchcoreml::MultiArray::copy(executorchcoreml::MultiArray&, executorchcoreml::MultiArray::CopyOptions) const + 148 frame #12: 0x000000030c7488d4 _portable_lib.cpython-310-darwin.so`invocation function for block in (anonymous namespace)::copy(MLMultiArray*, executorchcoreml::MultiArray&) + 376 frame #13: 0x000000030c748ac8 _portable_lib.cpython-310-darwin.so`invocation function for block in (anonymous namespace)::copy(MLMultiArray*, executorchcoreml::MultiArray&) + 52 frame #14: 0x000000019ad33f4c CoreML`CoreML::MultiArrayBuffer::getBytesWithHandler(void (void const*, unsigned long) block_pointer) const + 340 frame #15: 0x000000019ad34138 CoreML`-[MLMultiArray(ScopedBufferAccess) getBytesWithHandler:] + 152 frame #16: 0x000000030c7485ec _portable_lib.cpython-310-darwin.so`(anonymous namespace)::copy(MLMultiArray*, executorchcoreml::MultiArray&) + 296 frame #17: 0x000000030c744f68 _portable_lib.cpython-310-darwin.so`(anonymous namespace)::set_outputs(std::__1::vector<executorchcoreml::MultiArray, std::__1::allocator<executorchcoreml::MultiArray>>&, NSArray<MLMultiArray*>*) + 180 ``` With this PR, the process succeeds.
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
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
20 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.
Update executorch to latest