[pull] main from llvm:main #699
Merged
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
Tmp.moveFrom(*this); about 10 lines above leaves *this in a zombie state with a bucket array still allocated. This patch fixes the memory leak by calling deallocateBuckets().
Passing BlockOrder by value results in deep copies of DenseMap. This patch replaces them with const l-value references.
ISD.Symbol is already of type MCSymbolMachO *. Identified with readability-redundant-casting.
These functions are declared in Hexagon.h. Identified with readability-redundant-declaration.
defineMacro takes Twine, which can be constructed from const std::string &. Identified with readability-redundant-string-cstr.
VPPartialReductionRecipe doesn't yet support an EVL variant, and we guard against this by not calling convertToAbstractRecipes when we're tail folding with EVL. However recently some things got shuffled around which means we may detect some scaled reductions in collectScaledReductions and store them in ScaledReductionMap, where outside of convertToAbstractRecipes we may look them up and start e.g. adding a scale factor to an otherwise regular VPReductionPHI. This fixes it by skipping collectScaledReductions, and fixes #167861
…#168022) AluOpcode includes fixed point as well, which isn't possible here.
This patch support PADD_W, PSUB_W, PSADD_W, PSADDU_W, PSSUB_W, PSSUBU_W, PAADD_W and PAADDU_W
Since 1939eb3, std::filesystem is enabled by default in MSVC builds too.
…7845) The llvm-mingw toolchains defaults to `_WIN32_WINNT=0x601`, so this configuration is covered by our CI build matrix.
…#163724) -- This commit includes the basic infra/utilities to add matchers for linalg.*conv*/*pool* ops - such that given a `linalg.generic` op it identifies which linalg.*conv*/*pool* op it is. -- It adds a few representative linalg.*conv*/*pool* ops to demo the matchers' capability and does so as part of `linalg-specialize-generic-ops` pass. -- The goal is directed towards addressing the aim of [[RFC] Op explosion in Linalg](https://discourse.llvm.org/t/rfc-op-explosion-in-linalg/82863) iteratively for `*conv*/*pooling*` ops. -- This is part-1 of a series of PRs aimed to add matchers for Convolution ops. -- For further details, refer to #163374 (review) Signed-off-by: Abhishek Varma <abhvarma@amd.com>
Instead of having one large Dockerfile building multiple images with relatively confusing inheritance, explicitly have three standalone Dockerfiles each building one image. Then, tie the three images together using the docker-compose file which explicitly versions the base image used by the Android and the Github Actions images.
…163541) Currently, in-loop reductions for AnyOf and FindIV are not supported. They were implicitly blocked. This happened because RecurrenceDescriptor::getReductionOpChain could not detect their recurrence chain. The reason is that RecurrenceDescriptor::getOpcode was set to Instruction::Or, but the recurrence chains of AnyOf and FindIV do not actually contain an Instruction::Or. This patch explicitly disables in-loop reductions for AnyOf and FindIV instead of relying on getReductionOpChain to implicitly prevent them.
Without this, any RAII objects held in the task's captures aren't destroyed in a similar fashion to the task being run. If those objects in turn interact with the thread pool itself, chaos ensues. This comes up quite naturally with RAII-objects used for synchronization such as RAII-powered latches or releasing a mutex, etc. A unit test is crafted that tries to very directly test that the logic of the thread pool continues to hold even with an RAII object. This isn't the only type of failure mode (a deadlock due to mutexes in the captures can also occur), but seemed the easiest to test.
…rand (#161501) InlineAsmLowering rejected inline assembly with memory reference inputs if the values passed to the inline asm weren't pointers. The DAG lowering however handled them just fine. This patch updates InlineAsmLowering to store such values on the stack, and then use the stack pointer as the "indirect" version of the operand.
This commit ensures that gather and scatter operations with int64 index tensors can be created. This aligns with the EXT_INT64 extension.
…S1POE2) (#164912) Add assembly/disassembly support for AArch64 `FEAT_S1POE2` (Stage 1 Permission Overlay Extension 2), as blogged about here: * https://developer.arm.com/community/arm-community-blogs/b/architectures-and-processors-blog/posts/future-architecture-technologies-poe2-and-vmte and as documented here: * https://developer.arm.com/documentation/109697/2025_09/Future-Architecture-Technologies Co-authored-by: Rodolfo Wottrich <rodolfo.wottrich@arm.com>
This patch ensures we switch out of streaming mode before TLS-descriptor calls. ZA state will also be preserved when using the new SME ABI lowering (`-aarch64-new-sme-abi`). Fixes #152165
…163528) The IR verifier will carsh if there is any instructions located before phi-node. The `infer-address-spaces` pass would like to insert `addrspacecast` before phi-node in some corner cases. Indeed, since the operand pointer(phi-node's incoming value) has been determined to `NewAS` by the pass, it is safe to `addrspacecast` it immediately after the position where defined it. Co-authored-by: Kerang Mao <krmao@birentech.com>
Precommit test fixups for #167113
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
17 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.
See Commits and Changes for more details.
Created by
pull[bot] (v2.0.0-alpha.4)
Can you help keep this open source service alive? 💖 Please sponsor : )