MSVC fixes #1
Draft
MSVC fixes #1
Conversation
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
wkpark
pushed a commit
that referenced
this pull request
Dec 29, 2023
Turned command sequence mentioned in https://dev-discuss.pytorch.org/t/how-to-get-a-fast-debug-build/1597 and in various discussions into a tool that I use almost daily to debug crashes or correctness issues in the codebase Essentially it allows one to turn this: ``` Process 87729 stopped * thread #1, queue = 'com.apple.main-thread', stop reason = breakpoint 1.1 frame #0: 0x00000001023d55a8 libtorch_python.dylib`at::indexing::impl::applySelect(at::Tensor const&, long long, c10::SymInt, long long, c10::Device const&, std::__1::optional<c10::ArrayRef<c10::SymInt>> const&) libtorch_python.dylib`at::indexing::impl::applySelect: -> 0x1023d55a8 <+0>: sub sp, sp, #0xd0 0x1023d55ac <+4>: stp x24, x23, [sp, #0x90] 0x1023d55b0 <+8>: stp x22, x21, [sp, #0xa0] 0x1023d55b4 <+12>: stp x20, x19, [sp, #0xb0] ``` into this ``` Process 87741 stopped * thread #1, queue = 'com.apple.main-thread', stop reason = breakpoint 1.1 frame #0: 0x00000001024e2628 libtorch_python.dylib`at::indexing::impl::applySelect(self=0x00000001004ee8a8, dim=0, index=(data_ = 3), real_dim=0, (null)=0x000000016fdfe535, self_sizes= Has Value=true ) at TensorIndexing.h:239:7 236 const at::Device& /*self_device*/, 237 const c10::optional<SymIntArrayRef>& self_sizes) { 238 // See NOTE [nested tensor size for indexing] -> 239 if (self_sizes.has_value()) { 240 auto maybe_index = index.maybe_as_int(); 241 if (maybe_index.has_value()) { 242 TORCH_CHECK_INDEX( ``` while retaining good performance for the rest of the codebase Pull Request resolved: pytorch#116521 Approved by: https://github.com/atalman
wkpark
pushed a commit
that referenced
this pull request
Jan 10, 2024
…6938) As [`newFunctionWithName:`](https://developer.apple.com/documentation/metal/mtllibrary/1515524-newfunctionwithname) does not accept error argument, do not attempt to print it as it'll be guaranteed `nil` at that point, that results in a classic null pointer dereference, when `TORCH_CHECK` will attempt to construct `std::string` from it. See below backtrace for example: ``` thread #1, queue = 'metal gpu stream', stop reason = EXC_BAD_ACCESS (code=1, address=0x0) frame #0: 0x000000018a316dc4 libsystem_platform.dylib`_platform_strlen + 4 frame #1: 0x00000001471011bc libtorch_cpu.dylib`std::__1::__constexpr_strlen[abi:v160006](__str=0x0000000000000000) at cstring:114:10 frame pytorch#2: 0x0000000147100c24 libtorch_cpu.dylib`std::__1::char_traits<char>::length(__s=0x0000000000000000) at char_traits.h:220:12 * frame pytorch#3: 0x0000000147100bf0 libtorch_cpu.dylib`std::__1::basic_ostream<char, std::__1::char_traits<char>>& std::__1::operator<<[abi:v160006]<std::__1::char_traits<char>>(__os=0x000000016fdfb3a0, __str=0x0000000000000000) at ostream:901:57 frame pytorch#4: 0x0000000147100bb4 libtorch_cpu.dylib`std::__1::basic_ostream<char, std::__1::char_traits<char>>& c10::detail::_str<char const*>(ss=0x000000016fdfb3a0, t=0x000000016fdfb5d0) at StringUtil.h:55:6 frame pytorch#5: 0x00000001471007ac libtorch_cpu.dylib`std::__1::basic_ostream<char, std::__1::char_traits<char>>& c10::detail::_str<char const*, char const*>(ss=0x000000016fdfb3a0, t=0x000000016fdfb4f8, args=0x000000016fdfb5d0) at StringUtil.h:68:10 frame pytorch#6: 0x0000000147101444 libtorch_cpu.dylib`std::__1::basic_ostream<char, std::__1::char_traits<char>>& c10::detail::_str<std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char>>, char const*, char const*>(ss=0x000000016fdfb3a0, t="index_select_32bit_idx32", args=0x000000016fdfb4f8, args=0x000000016fdfb5d0) at StringUtil.h:68:10 frame pytorch#7: 0x0000000147101404 libtorch_cpu.dylib`std::__1::basic_ostream<char, std::__1::char_traits<char>>& c10::detail::_str<char const*, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char>>, char const*, char const*>(ss=0x000000016fdfb3a0, t=0x000000016fdfb500, args="index_select_32bit_idx32", args=0x000000016fdfb4f8, args=0x000000016fdfb5d0) at StringUtil.h:68:10 frame pytorch#8: 0x000000014710137c libtorch_cpu.dylib`c10::detail::_str_wrapper<char const*, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char>> const&, char const*, char const* const&>::call(args=0x000000016fdfb500, args="index_select_32bit_idx32", args=0x000000016fdfb4f8, args=0x000000016fdfb5d0) at StringUtil.h:75:5 frame pytorch#9: 0x0000000147101310 libtorch_cpu.dylib`decltype(auto) c10::str<char [53], std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char>>, char [10], char const*>(args={a\xcb\xa7H\x01\0\0\0}, args="index_select_32bit_idx32", args={\x96\xcb\xa7H\x01\0\0\0}, args=0x000000016fdfb5d0) at StringUtil.h:111:10 frame pytorch#10: 0x0000000147100210 libtorch_cpu.dylib`decltype(auto) c10::detail::torchCheckMsgImpl<char [53], std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char>>, char [10], char const*>((null)="Expected indexFunction to be true, but got false. (Could this error message be improved? If so, please report an enhancement request to PyTorch.)", args={a\xcb\xa7H\x01\0\0\0}, args="index_select_32bit_idx32", args={\x96\xcb\xa7H\x01\0\0\0}, args=0x000000016fdfb5d0) at Exception.h:453:10 frame pytorch#11: 0x00000001470fffe8 libtorch_cpu.dylib`at::mps::MPSDevice::metalIndexingPSO(this=0x0000600000381670, kernel="index_select_32bit_idx32") at MPSDevice.mm:62:3 ``` This was introduced by pytorch#99855 that replaced `newFunctionWithName:constantValues:error:` with `newFunctionWithName:` Pull Request resolved: pytorch#116938 Approved by: https://github.com/Skylion007
|
@wkpark Is there any update on this PR yet? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
WIP