AMDGPU: Make fmul_legacy intrinsic propagate poison
| Delta | File | |
|---|---|---|
| +16 | -0 | llvm/test/Transforms/InstCombine/AMDGPU/fmul_legacy.ll |
| +5 | -0 | llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp |
| +21 | -0 | 2 files |
AMDGPU: Make fmed3 intrinsic propagate poison
| Delta | File | |
|---|---|---|
| +24 | -0 | llvm/test/Transforms/InstCombine/AMDGPU/amdgcn-intrinsics.ll |
| +5 | -0 | llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp |
| +29 | -0 | 2 files |
Only propagate if both inputs poison
| Delta | File | |
|---|---|---|
| +16 | -8 | llvm/test/Transforms/InstCombine/AMDGPU/amdgcn-intrinsics.ll |
| +2 | -1 | llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp |
| +18 | -9 | 2 files |
AMDGPU: Make ballot intrinsic propagate poison
| Delta | File | |
|---|---|---|
| +16 | -0 | llvm/test/Transforms/InstCombine/AMDGPU/amdgcn-intrinsics.ll |
| +5 | -1 | llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp |
| +21 | -1 | 2 files |
AMDGPU: Make cvt_pknorm and cvt_pk intrinsics propagate poison
| Delta | File | |
|---|---|---|
| +96 | -0 | llvm/test/Transforms/InstCombine/AMDGPU/amdgcn-intrinsics.ll |
| +3 | -0 | llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp |
| +99 | -0 | 2 files |
[VPlan] Fix crash with inloop fmuladd reductions with blend (#131154) When visiting in-loop reduction links, we previously crashed if we had an fmuladd with a blend after it in the chain. This fixes it by lifting the existing blend folding to also handle fmuladd. This also simplifies the code structure slightly for an upcoming patch I want to post to handle in-loop AnyOf reductions. I removed the PhiR->isInLoop() check since it's already guarded at the top of the parent Header->Phis() loop.
| Delta | File | |
|---|---|---|
| +89 | -6 | llvm/test/Transforms/LoopVectorize/reduction-inloop.ll |
| +13 | -14 | llvm/lib/Transforms/Vectorize/LoopVectorize.cpp |
| +102 | -20 | 2 files |
[X86] Avoid repeated hash lookups (NFC) (#131069)
| Delta | File | |
|---|---|---|
| +1 | -1 | llvm/lib/Target/X86/X86PreTileConfig.cpp |
| +1 | -1 | 1 files |
[IPO] Avoid repeated hash lookups (NFC) (#131068)
| Delta | File | |
|---|---|---|
| +10 | -7 | llvm/lib/Transforms/IPO/IROutliner.cpp |
| +10 | -7 | 1 files |
[CodeGen] Avoid repeated hash lookups (NFC) (#131067)
| Delta | File | |
|---|---|---|
| +2 | -4 | llvm/lib/CodeGen/RDFLiveness.cpp |
| +2 | -4 | 1 files |
[Analysis] Avoid repeated hash lookups (NFC) (#131066)
| Delta | File | |
|---|---|---|
| +4 | -2 | llvm/lib/Analysis/MemoryDependenceAnalysis.cpp |
| +4 | -2 | 1 files |
[Sema] Avoid repeated hash lookups (NFC) (#131065)
| Delta | File | |
|---|---|---|
| +3 | -3 | clang/lib/Sema/SemaOpenACCClause.cpp |
| +3 | -3 | 1 files |
[AST] Avoid repeated hash lookups (NFC) (#131064)
| Delta | File | |
|---|---|---|
| +4 | -5 | clang/lib/AST/ASTImporter.cpp |
| +4 | -5 | 1 files |
[libc] Make RPC server handling header only (#131205) Summary: This patch moves the RPC server handling to be a header only utility stored in the `shared/` directory. This is intended to be shared within LLVM for the loaders and `offload/` handling. Generally, this makes it easier to share code without weird cross-project binaries being plucked out of the build system. It also allows us to soon move the loader interface out of the `libc` project so that we don't need to bootstrap those and can build them in LLVM.
[SandboxVec][PassManager][NFC] Fix PM printing
It currently prints '(' instead of '<'. This patch fixes it.
| Delta | File | |
|---|---|---|
| +9 | -9 | llvm/include/llvm/SandboxIR/PassManager.h |
| +2 | -2 | llvm/unittests/SandboxIR/PassTest.cpp |
| +11 | -11 | 2 files |
[LLVM] Update llvm-mca / MCA maintainers (#131187) Adding myself as a co-maintainer alongside @adibiagio , who invited me to take on this role.
| Delta | File | |
|---|---|---|
| +3 | -1 | llvm/Maintainers.md |
| +3 | -1 | 1 files |
Lex: add support for `i128` and `ui128` suffixes (#130993) Microsoft's compiler supports an extension for 128-bit literals. This is referenced in `intsafe.h` which is included transitievly. When building with modules, the literal parsing causes a failure due to the missing support for the extension. To alleviate this issue, support parsing this literal, especially now that there is the BitInt extension. Take the opportunity to tighten up the code slightly by ensuring that we do not access out-of-bounds characters when lexing the token.
[ORC] Make runAllocActions and runDeallocActions asynchorous. Part of ongoing work to make core ORC operations asynchronous. This only affects the runner utilities, not the AllocationActions themselves. The AllocationActions will be made asynchronous in a future patch.
Reapply "Use global TimerGroups for both new pass manager and old pass manager timers" (#131173) (#131217) This reverts commit 31ebe6647b7f1fc7f6778a5438175b12f82357ae. The reason for the test failure is likely due to `Name2PairMap::getTimerGroup(...)` not holding a lock.
[Support] Prevent leaking unique lock files (#130984) Prior to this PR, failing to get the host ID would leave the unique lock file on the file system. This is now fixed by constructing `RemoveUniqueLockFileOnSignal` earlier. This PR also removes one call to `sys::fs::remove()` that is now redundant and another that was redundant even before this patch.
| Delta | File | |
|---|---|---|
| +4 | -9 | llvm/lib/Support/LockFileManager.cpp |
| +4 | -9 | 1 files |
[CodeGen] Use early return to simplify SplitEditor::defFromParent [NFC]
| Delta | File | |
|---|---|---|
| +26 | -28 | llvm/lib/CodeGen/SplitKit.cpp |
| +26 | -28 | 1 files |
[mlir][affine] Guard invalid dim attribute in the test-reify-bound pass (#129013) Computing the bound of affine op (ValueBoundsConstraintSet::computeBound) crashes due to the invalid dim value given to the op. It is necessary for the pass to check the dim attribute not to be greater than the rank of the input type. Fixes https://github.com/llvm/llvm-project/issues/128807
| Delta | File | |
|---|---|---|
| +47 | -0 | mlir/test/Dialect/Affine/invalid-reify-bound-dim.mlir |
| +21 | -0 | mlir/test/lib/Dialect/Affine/TestReifyValueBounds.cpp |
| +68 | -0 | 2 files |
[SLP] Precommit test (#131236) Precommit test for https://github.com/llvm/llvm-project/pull/131229
| Delta | File | |
|---|---|---|
| +150 | -0 | llvm/test/Transforms/SLPVectorizer/AMDGPU/extract-ordering.ll |
| +150 | -0 | 1 files |
XFAIL malloc_zone.cpp for darwin/lsan (#131234) Silence darwin bot while we investigate the problem rdar://145873843 Co-authored-by: Mariusz Borsa <m_borsa at apple.com>
| Delta | File | |
|---|---|---|
| +2 | -0 | compiler-rt/test/sanitizer_common/TestCases/Darwin/malloc_zone.cpp |
| +2 | -0 | 1 files |
[RISCV] Remove some unnecessary hasSideEffects = 0 and sink some to their base class. NFC (#131044) Some of these were already present in their base class and therefore redundant. Others were missing from their base classes. Maybe leftover from when VXRM was modeled with side effects?
| Delta | File | |
|---|---|---|
| +11 | -23 | llvm/lib/Target/RISCV/RISCVInstrInfoVPseudos.td |
| +11 | -23 | 1 files |
[lldb] Split some lldb-server tests to avoid timeout (#129614) Split test cases out of TestLldbGdbServer.py and TestGdbRemoteFork.py into separate files to avoid hitting the 600s timeout limit. The inferior used by these tests (main.cpp) takes approximately 20s to compile with a Debug build of clang, causing timeouts when a single test file contains many tests. By grouping similar tests into separate files, we can prevent timeouts and improve overall test efficiency.
[lldb-dap] Support vscode launch URLs (#125843)
This commit adds support for starting debug sessions through special
`vscode://llvm-vs-code-extensions.lldb-dap/start?config={launch-config}`
URIs. This allows tighter integration with custom scripts. One potential
use case is providing similar functionality to `xcdebug`, see #125777
for some discussion on that use case.
The functionality was inspired by @vadimcn's CodeLLDB extension, which
[provides similar
functionality](https://github.com/vadimcn/codelldb/blob/master/MANUAL.md#debugging-externally-launched-code).| Delta | File | |
|---|---|---|
| +78 | -0 | lldb/tools/lldb-dap/src-ts/uri-launch-handler.ts |
| +20 | -1 | lldb/tools/lldb-dap/README.md |
| +5 | -0 | lldb/tools/lldb-dap/src-ts/extension.ts |
| +2 | -1 | lldb/tools/lldb-dap/package.json |
| +105 | -2 | 4 files |
[LLDB] Fix handling of auto_continue for stop hooks
| Delta | File | |
|---|---|---|
| +7 | -10 | lldb/source/Target/Target.cpp |
| +7 | -10 | 1 files |
Simplify Target::RunStopHooks() (#129578) Introduce `StopHookResult::NoPreference` and simplify control flow in `Target::RunStopHooks()`. The algorithm is (in order): 1. "Auto continue" set on any hook -> continue 2. "Stop demanded" by any hook -> stop 3. "Continue requested" by any hook -> continue 4. No hooks, or "no preference" only (default stance) -> stop The new `NoPreference` lets us keep the default stance, distinguishing case 3. and 4.
| Delta | File | |
|---|---|---|
| +22 | -57 | lldb/source/Target/Target.cpp |
| +1 | -0 | lldb/include/lldb/Target/Target.h |
| +23 | -57 | 2 files |
[stdio][baremetal] Fix templating for print functions (#131232) This addresses build errors from https://github.com/llvm/llvm-project/pull/111559.
| Delta | File | |
|---|---|---|
| +2 | -2 | libc/src/stdio/baremetal/vprintf.cpp |
| +2 | -2 | libc/src/stdio/baremetal/printf.cpp |
| +4 | -4 | 2 files |
[Clang driver] Diagnose -maix-shared-lib-tls-model-opt on wrong targets
| Delta | File | |
|---|---|---|
| +3 | -4 | clang/test/Driver/aix-shared-lib-tls-model-opt.c |
| +5 | -0 | clang/lib/Driver/ToolChains/Arch/PPC.cpp |
| +8 | -4 | 2 files |