Add cast for output from FP16 -> FP32; remove remaining casts#9
Closed
DenisVieriu97 wants to merge 3 commits intodev/georgep/f32tof16from
Closed
Add cast for output from FP16 -> FP32; remove remaining casts#9DenisVieriu97 wants to merge 3 commits intodev/georgep/f32tof16from
DenisVieriu97 wants to merge 3 commits intodev/georgep/f32tof16from
Conversation
georgepaw
suggested changes
Dec 13, 2023
| for j in range(len(node.args[i])): | ||
| # Call get_node explicitly to preserve the output signature. | ||
| output_nodes.append(graphNodes.get_node(node.args[i][j].name)) | ||
| out_node = graphNodes.get_node(node.args[i][j].name) |
There was a problem hiding this comment.
Is there a reason for this change? The casts will be elided by MPSGraph. If you are concerned about these casts, we should instead just change the get_node method to default to not doing the cast.
Owner
Author
There was a problem hiding this comment.
I was seeing additional casts in the graph (from FP16->FP32) which were causing lots of overhead
There was a problem hiding this comment.
Good point - I wasn't expecting this to be a big overhead. In my PR I've removed these and also ensured to only add output casts when the outputs in EdgeIR were FP32
georgepaw
force-pushed
the
dev/georgep/f32tof16
branch
5 times, most recently
from
35a3b96 to
7825bcd
Compare
DenisVieriu97
force-pushed
the
dev/georgep/f32tof16
branch
from
7825bcd to
a2661be
Compare
georgepaw
force-pushed
the
dev/georgep/f32tof16
branch
from
a2661be to
a480991
Compare
manuelcandales
pushed a commit
that referenced
this pull request
Aug 27, 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 pytorch#15: 0x000000019ad34138 CoreML`-[MLMultiArray(ScopedBufferAccess) getBytesWithHandler:] + 152 frame pytorch#16: 0x000000030c7485ec _portable_lib.cpython-310-darwin.so`(anonymous namespace)::copy(MLMultiArray*, executorchcoreml::MultiArray&) + 296 frame pytorch#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 free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
No description provided.