Upstream merge 55 #1899
Merged
Upstream merge 55 #1899
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
…65d2
e8215edf65d2 merge main into amd-staging (#2797)
12503e782eac merge main into amd-staging
e8976e92f655 [clang][Preprocessor] Add peekNextPPToken, makes look ahead next token without side-effects (#143898)
4d2b79b04a9e [lldb] Fix build for #145017
bfde1477617f [NFC][AMDGPU] Update and.ll test and automate check line generation (#145371)
24438aa4886d [lldb] Use Socket::CreatePair for launching debugserver (#145017)
371f12f96dc0 Revert "[lldb] Add count for number of DWO files loaded in statistics" (#145494)
666cad99f87e merge main into amd-staging (#2795)
61b99ca51282 [AArch64] Consider StreamingSVE in shouldExpandGetActiveLaneMask (#144722)
049d61ad65b0 [flang][AArch64] Always link compiler-rt to flang after libgcc (#144710)
594ebe63406b [X86] combineSelect - move vselect(cond, pshufb(x), pshufb(y)) -> or(pshufb(x), pshufb(y)) fold (#145475)
54b522f6fdff [AMDGPU] Fixup a201f8872a63 (#145486)
e4b9aa61923f [bazel] Port d31ba5256327d30f264c2f671bf197877b242cde
45c5eb168fff [bazel] mlir_copts doesn't exist
cf9546b82661 [lldb] Remove GDBRemoteCommunication::ConnectLocally (#145293)
46e1e9f104c8 Reapply "[lldb/cmake] Plugin layering enforcement mechanism (#144543)" (#145305)
a201f8872a63 [AMDGPU] Replace dynamic VGPR feature with attribute (#133444)
6cfa03f1f1d9 [ORC] Drop unused LinkGraphLinkingLayer::Plugin::notifyLoaded method. (#145457)
5fa55b2dfc2c Revert "[flang][OpenMP] Skip runtime mapping with no offload targets (#144534)" (#145478)
73e4f8a71f2b ARM: Use member initializer list (#145459)
8d9911e4a06c [Option] Use a range-based for loop (NFC) (#145446)
d5c8024daef6 [mlir][bazel]: Add FuncUtil rule in bazel files (#145463)
0112f12eb61b [EarlyCSE] Remove void return restriction for call CSE (#145320)
8ec0552a7f1f Reapply "[CUDA][HIP] Add a __device__ version of std::__glibcxx_assert_fail() (#144886)
610b881c620f merge main into amd-staging
f70473878166 [verify-uselistorder] Use llvm::is_sorted (NFC) (#145444)
1247fddf3676 [SimplifyCFG] Relax `cttz` cost check in `simplifySwitchOfPowersOfTwo`
c5972da34a08 [mlir][Transforms] Dialect Conversion: Simplify block-inline handling (#145308)
3e1e368824e7 [AMDGPU][SDAG] Add tests for ISD::PTRADD DAG combines (#142738)
3a3ede40243d merge main into amd-staging (#2794)
b1dcf78378aa [X86][APX] Fix issue of push2/pop2 instr with stack clash protection (#145303)
ef048471f701 [NVPTX][NFC] Rearrange the TMA-S2G intrinsics (#144903)
b9d1642423f8 [NFC] Remove UnwindTable dependency on CIE, and FDE (#142520)
290fc1ea114a MC,AsmPrinter: Report redefinition error instead of crashing in more cases
d31ba5256327 [mlir][Interface] Factor out common IndexingMapOpInterface behavior in a new generic interface (#145313)
0c33799e374a [JITLink] Include target addend in out-of-range error (#145423)
e435558ff9b4 [clang-tidy] add 'IgnoreMarcos' option to 'special-member-functions' check (#143550)
5ce5ed4b8565 [mlir] Allow using non-attribute properties in declarative rewrite patterns (#143071)
f4df9f1c6e3a [mlir][func]: Fixed linkage problem in func dialect (#145456)
809e2900fff9 X86: Use reportFatalUsageError for tiny code model error (#145279)
ffc7d5ae2d79 Add `REQUIRES: asserts` to test added in #145149 because it uses the `-debug-only=` flag.
0e461d1781f1 [clang][bytecode] Only use int128 if it is available
3ba7a872bf5f [mlir][func]: Introduce ReplaceFuncSignature tranform operation (#143381)
f4068cbfdc05 merge main into amd-staging (#2792)
37eb465710b4 Reland "[Modules] Record whether VarDecl initializers contain side effects" (#145447)
83c32fd01be1 merge main into amd-staging
0fcced7d791f [RISCV][NFC] Zce always implies/requires Zca (#145442)
f6ab1f02ec0e [RISCV] Support LLVM IR intrinsics for XAndesVBFHCvt (#145321)
f0d898f36bce DAG: Move get_dynamic_area_offset type check to IR verifier (#145268)
90a6819cfe08 [MemProf] Update the DISubprogram linkageName for clones (#145385)
ca04d74564d7 [ORC] DLLImportDefinitionGenerator should use LookupKind::Static.
88b98d3367f3 [RISCV] Add ISel pattern for generating QC_BREV32 (#145288)
bd6ee6ac2133 [C23][Parser] Accept single variadic parameter function declarator in type name (#145362)
975d4df147ad [lldb] Remove an unused local variable (NFC) (#145212)
d4d37d84303f [BOLT] Remove a redundant call to std::unique_ptr<T>::get (NFC) (#145211)
2f9c97c030c3 [RISCV] Add Andes AX45MPV processor definition (#145267)
9d570d568b37 [ValueTracking] Return true for AddrSpaceCast in canCreateUndefOrPoison (#144686)
a314ac4d2299 [Reland][InstCombine] Iterative replacement in PtrReplacer (#145410)
310a62f88aa9 [HLSL][RootSignature] Plug-in serialization and add full sample testcase (#144769)
a2aa812a3173 [mlir][python] bind block predecessors and successors (#145116)
087444c17767 [Comgr][NFC] Nest spirv lit tests in spirv directory
0d24ae9c26c4 merge main into amd-staging
bc5e5c0114fb [libc] wcpncpy implementation (#145430)
10d46cf0d57a [libc] mbtowc implementation (#145405)
23b66a68f1f1 [ExtractAPI] Include `virtual` keyword for methods (#145412)
4b9f7cd856f1 [flang] flang manpage overhaul (#144948)
98e8ef2273b3 [libc] Fix broken links in libc (#145199)
d37325ea9536 Revert "[MLGO][IR2Vec] Integrating IR2Vec with MLInliner (#143479)" (#145418)
92a7f6fbbe48 [lldb][rpc] Fix bug in convert script for RPC (#145419)
d031383235ee merge main into amd-staging (#2774)
3782eb60f849 [mlir][TilingInterface] NFC Improve comment for `tiledAndFusedOps` member of `SCFTileAndFuseResult` (#145397)
509fb931b441 Fix min_vec_size.ll test for changes in vector-combine (#145392)
a45fda6aeba3 switch type and value ordering for arith `Constant[XX]Op` (#144636)
97ad0f4b3d3c [DAGCombiner][RISCV] Don't propagate the exact flag from udiv/sdiv to urem/srem. (#145387)
06d78ba953dc [lldb] Fix warnings
61a969b86757 Revert "[MSAN] handle assorted AVX permutations" (#145404)
af2c06ecd610 [MLGO][IR2Vec] Integrating IR2Vec with MLInliner (#143479)
329ae868cbc0 Revert "[Modules] Record whether VarDecl initializers contain side effects" (#145407)
d715ecba797c Revert "[flang][fir] Add fir.if -> scf.if and add filecheck test … (#142965)" (#145345)
179d724867fe [CIR] Clean up enum attributes (#144999)
6cf656eca717 [NFC][Clang][AST] Drop `llvm::` in front of `ArrayRef`/`MutableArrayRef` (#145207)
66214410c405 [CIR] Add support for DumpRecordLayouts (#145058)
5f74d9bb6280 [mlir][linalg] Add support for inlined const to isaFillOpInterface (#144870)
653d0d007375 [mlir][python] add `MLIR_BINDINGS_PYTHON_INSTALL_PREFIX` to make bindings install dir configurable (#124878)
7bc956d3d643 [mlir][PartialReductionTilingInterface] Add support for `ReductionTilingStrategy::PartialReductionOuterParallel` in `tileUsingSCF`. (#143988)
6c232f440ff3 [CUDA] Add missing intrinsics to cuda headers, take #2 (#144851)
5d01697ec6cb [LAA] Be more careful when evaluating AddRecs at symbolic max BTC. (#128061)
bf4afb08fe1c [CostModel] improveShuffleKindFromMask - recognise a SK_PermuteSingleSrc incorrectly tagged as SK_PermuteTwoSrc (#145352)
f4d31cdee3d7 [CIR] Add bitfield offset calculation for big-endian targets (#145067)
e6f98ff4a81d Fix variable naming style in PPCBoolRetToInt.cpp (#144533)
3095d3a47d62 [lldb] Add count for number of DWO files loaded in statistics (#144424)
97e826617253 [CIR] Remove redundant operation trait and use AllTypesMatch instead (#144950)
b215c8e18ff9 [libc] wcpcpy implementation (#144802)
71817856f7f4 [mlir][PartialReductionTilingInterface] Generalize implementation of `tileUsingSCF` for `ReductionTilingStrategy::PartialOuterReduction`. (#143467)
e80acd4fae60 [clang][nvlink-wrapper] Add support for opt-remarks command line options (#145365)
0f173a0f9af5 [DirectX] make firstbitlow intrinsic use first argument instead of return for overload type (#145350)
ff865b639af0 [lldb] Fix SBMemoryRegionInfoListExtensions iter to yield unique refe… (#144815)
7ce76e1ad124 [NVPTX] Rename register classes after float register removal (NFC) (#145255)
43ae009a9b31 [RISCV] Make All VType Parts Optional (#144971)
a6eb5eee38ee [RISCV][NFC] Remove hasStdExtCOrZca (#145139)
f1c1456b9137 [gn build] Port c594f6e69743
e391301e0e4d [lldb] Use `proc` instead of `pro` to avoid command ambiguity
ab17ff05621f [RISCV] Add Zvfh tests for vp.splice. NFC
53edba8091df [RISCV] Add vp.reverse tests for Zvfh and fractional lmuls. NFC
735d721de451 [PowerPC] Fix handling of undefs in the PPC::isSplatShuffleMask query (#145149)
319a51a5ffb8 [Modules] Record whether VarDecl initializers contain side effects (#143739)
836ff367d0bc [clang][bytecode] Fix IntegralAP::{isMin,isMax} (#145339)
c445ca5412a7 [clang][bytecode] Remove incorrect assertion (#145341)
ed75e55a70aa [NFC][Clang][AST] Adopt `llvm::copy` in Clang AST (#145192)
f40909f605fd [RISCV] Add SiFive X390 scheduling model (#143938)
8d83d046376e [lldb] add plugin names to process save-core error output. (#143126)
1bc63265afbe [lldb][DWARFASTParserClang] Support constant index encoding of DW_AT_object_pointer (#144998)
081adc15e898 [Triple][CodeGen] Fix `Triple::isTargetEHABICompatible()` for NetBSD (#143549)
7a3356951053 [RISCV] Factor out common SiFive7 scheduling model into an abstraction layer (#144442)
13bb7948c914 [mlir][spirv]: Add Image to Vulkan Storage Class Map (#144899)
a0329eaa0cdc [lldb] upgrade HandleFrameFormatVariable callees to llvm::Expected (#144731)
c594f6e69743 Revert "[clang-tidy] Add new check `readability-use-numeric-limits`" (#145355)
909cbcf988fe [lldb-dap] Fix URL label and error code in DAPError (#145010)
c0ce9adf603c [llvm] remove incorrect `LLVM_ABI` annotation usage (#144606)
10b61fea8a1a [llvm] include GenericLoopInfoImpl for full implementation (#144621)
5a16645a3da2 Reland "[lldb][DWARF] Remove object_pointer from ParsedDWARFAttributes (#145065)" (#145126)
634fe0de5041 [lldb][DWARF] Support retrieving DW_FORM_implicit_const value with DWARFDebugInfoEntry::GetAttributeValue (#145328)
8d7a8fcc3ab9 [SHT_LLVM_BB_ADDR_MAP] Encode and decode callsite offsets in a newly-introduced SHT_LLVM_BB_ADDR_MAP version. (#144426)
698e9f56558e [lldb] Add support for NoneType to decorator skipIfBuildType (#145342)
3dc9f2da29f7 Silence "non-void function does not return a value in all control paths" in Release builds. NFC
576085c94855 [SelectionDAG][RISCV] Add support for splitting vp.splice (#145184)
61113acc3b88 merge main into amd-staging
1128a4fd2c3a [HLSL] Don't use CreateRuntimeFunction for intrinsics (#145334)
9f7567d33a6c [PreISelIntrinsicLowering] Reuse previously generated GlobalVariable for memset_pattern16 when possible (#144677)
ff8049a23e46 [InlineCost] Allow simplifying to non-Constant values (NFCI) (#145083)
b05436364512 [LV] Add tests showing incorrect vector interleaving with early exits.
068af5bfb47a [lld][BP] Print total size of startup symbols (#145106)
daa2a587cc01 [TRE] Adjust function entry count when using instrumented profiles (#143987)
44936c8d13f9 [CUDA][HIP] add options `--[no-]offload-inc` (#140106)
b0366eeb7eff [MLIR] Add support for int8/uint8 properties (#145019)
a50cb6ca3e12 [AMDGPU][True16][CodeGen] fix a predicate bug in VGPRImm with f16/bf16 (#144942)
bb8c42e85987 [LV] Extend FindLastIV to unsigned case (#141752)
4313f9b82168 merge main into amd-staging (#2768)
6d17eb5126b3 [NFC][Clang][AST] Use `llvm::copy` instead of `memcpy` in StringLiteral (#145187)
f4ca2231969a [LAA] Update early-exit test to cover last valid & first invalid access.
d0e5d6fd6180 [CodeGen][CodeLayout] Fix segfault on access to deleted block in MBP. (#142357)
cd91d0fff929 [mlir][py] Don't use a CMake iteration to find nanobind target to suppress warnings (NFC) (#143863)
cccb82e55291 [HLSL][SPIRV] Allow large z value in numthreads (#144934)
879a55793a2d [ExpandVariadics] Clean up intrinsic declaration lookup (NFC)
05491e0359ed [clang-tidy] add 'IgnoreMarcos' option to 'avoid-goto' check (#143554)
117e51de8a83 [flang][cmake] Don't pass -fno-strict-aliasing for GCC either
a65e0edd6ac5 PowerPC: Stop reporting memcpy as an alias of memmove on AIX (#143836)
58987d2e34e6 RuntimeLibcalls: Pass in ABI name from MCOptions (#144894)
c7d9b6ed5d6d [MLIR] Fix incorrect slice contiguity inference in `vector::isContiguousSlice` (#142422)
fb209929e1b1 [flang][OpenMP] Set isNewBlock directly on OpenMP constructs (#144593)
9a6a87da6e61 [AutoUpgrade] Remove unnecessary name check (NFCI)
9e704a0aa158 [MLIR][MemRef] Add `alloca` support for `erase_dead_alloc_and_stores` (#142131)
23ba0fdbfc2a [clang][bytecode] Fix assignInteger() with allocated primtypes (#145302)
9881a507a2a0 [bazel] Port 8584b216b87085a913fe39be15bfa4ab4754aeb9
c5b256a0e480 [MLIR][MPI] adding MLIRDLTIDialect when linking MLIRMPIDialect (#145316)
8584b216b870 Lower allreduce (#144716)
7e77aaebfbfc [bazel] Add missing dependency for 4af96a9d83335b3b59f3441af47c879c7a9eb183
bd809ffb4b5f [bazel] Make lld:ELF glob wider so it doesn't break as easily
680bce007273 [bazel] Add missing dependency for c7165587e49605452f96249412f123b47b78bb81
1e95349dbe32 Revert "ELF: Add branch-to-branch optimization."
6d8d4cf9a46b [lldb] Disable flaky TestDetachResumes.py on Windows x86_64 (#145301)
1d907c28b64b [VectorCombine][X86] fmaddsub.ll - add test variants without any undef elements
ab4b19fbc3b1 merge main into amd-staging
cfcb7888c71c [EarlyCSE] Fix dead store elimination for unwinding readnone calls (#145287)
e56384ff540e [IRTranslator] Remove unnecessary isIntrinsic() check (NFC)
7360ed0159a4 [mlir][transform] Drop redundant padding_dimensions spec from pad_tiling_interface (#145257)
d9a99afbfc32 [mlir][transform] Plumb a simplified form of AffineMin folding into t… (#145170)
02d2a1646a4a [Offload] Fix entry_points.td test (#145292)
18f667d80414 Revert "[lldb/cmake] Plugin layering enforcement mechanism (#144543)"
b1b8f67eab08 [mlir][Transforms] Add 1:N support to `replaceUsesOfBlockArgument` (#145171)
2545d6f7237b [ARM] Add MVE test coverage for LD2/ST2 shuffle costs. NFC
69549e0c2a54 merge main into amd-staging (#2767)
0662045bdf6c [TLI] Add support for pvalloc() (#144949)
714b2fdf3a38 [lldb] Add BRIEF_DOCS for cmake properties defined in #144543
5c22793eadd8 [lldb-dap][test] Refactor runInTerminal Tests. (#144954)
e7c1da7c8ef3 [lldb/cmake] Plugin layering enforcement mechanism (#144543)
613c38a9923a [Offload] Fix type mismatch warning in test (#143700)
b9c979d3698e [mlir][Transforms] Dialect conversion: Simplify `replaceOp` implementation (#145155)
092ef1da4594 AMDGPU: Use reportFatalUsageError for unsupported disassembly error (#145264)
c5629f2b6004 [lldb] Add Socket::CreatePair (#145015)
bc6faf9a0208 [X86] X86LegalizerInfo - use LegalFor instead if LegalIf for simple ISA/test pairs (#144675)
be000986322e [libc++] Remove a bunch of now unnecessary indirections in __tree (#142397)
43260b01dd6e [libc++] Add test to ensure that the mangling of types stays the same (#143556)
98a6fed09657 [mlir][tosa] Allow zero-points to be unranked (#143770)
b31413a96603 [MLIR][AArch64] Simplify LowerContractionToSVEI8MMPattern.cpp:getExtOperand (NFC) (#144909)
4af96a9d8333 [MLIR] Determine contiguousness of memrefs with dynamic dimensions (#142421)
1c78d8d9d7bc [clang][bytecode] Fix shifts with an allocated RHS (#145280)
529662a6b5f1 [mlir] Allow accessing DialectResourceBlobManager::blobMap (#142352)
2dcf43634084 AMDGPU: Remove legacy pass manager version of AMDGPUAttributor (#145262)
48155f93dd84 CodeGen: Emit error if getRegisterByName fails (#145194)
6b129d6bbf39 AsmPrinter: Do not use report_fatal_error for unhandled ConstantExpr (#145275)
db051e8800d3 AsmPrinter: Do not use report_fatal_error for unknown appending linkage (#145269)
16607f643749 AMDGPU: Fix typo in argument allocation error message (#145265)
338ee673bde3 AsmPrinter: Do not use report_fatal_error for AIX XXStructor error (#145273)
74aab3045d18 [TSan, NFC] Eliminate useless calculations in TSan (#145283)
20c04a646bcf [NFC][Sanitizer] Fix incorrect desc of [beg, end] to [beg, end)
4795b2b5a3d7 [RISCV] Fix comment typo and indentation for class SchedNary. NFC.
ddfc7cb61f02 [mlir][tosa] Check negative output size in resize shape inference (#143382)
86beba930111 [PredicateInfo] Cache ssa.copy declarations (NFC) (#145020)
ae8c85c9ce89 [Passes] Remove LoopInterchange from O1 pipeline (#145071)
1e58e9c4b273 [PredicateInfo] Don't store Def in ValueDFS (NFC) (#145022)
c6be4ff0c896 [PredicateInfo] Don't use depth first walk (NFCI) (#145016)
fccc6ee70218 [C++20] [Modules] Don't make enum constant members always visible
11bdc5ccdee3 merge main into amd-staging
4be4b82e74f1 AMDGPU: Use reportFatalUsageError for unhandled calling conventions (#145261)
38920964babf ARM: Move ABI enum from TargetMachine to TargetParser (#144725)
b79bf5ad771f merge main into amd-staging (#2766)
26f3f24a4f0a [MLIR][NFC] Declare RuntimeVerifiableOpInterface for memref ops that have an implementation (#145230)
6a0593b0a383 [AMDGPU] Extend wave reduce intrinsics for i32 type (#126469)
da0c21bd4b62 [mlir][gpu] Fix bug with GPU hardware intrinsic global location (#144923)
ed155ff9f25c AMDGPU: Avoid report_fatal_error on ds ordered intrinsics (#145202)
c7165587e496 [mlir][affine|ValueBounds] Add transform to simplify affine min max ops with ValueBoundsOpInterface (#145068)
89c61449e607 [AMDGPU] Baseline gfx1250 speed model. (#145217)
7620509c947a merge main into amd-staging
32f911f3e83b [InstCombine] Fold `ceil(X / (2 ^ C)) == 0` -> `X == 0` (#143683)
76ae9aa4d2fb [CodeGen] Use range-based for loops (NFC) (#145251)
6023ba2bf789 [mlir] Migrate away from TypeRange(std::nullopt) (NFC) (#145246)
a17b5bce8c9b [clang-reorder-fields] Prevent rewriting unsupported cases (#142149)
b00ddce731c6 [mlir][affine] Fix a crash when cast incompatible type (#145162)
f9fce4975bba [ORC] Fix potential stack corruption in Platform::lookupInitSymbols.
9844085ea813 [lldb] Migrate away from ValueRange(std::nullopt) (NFC) (#145245)
7a4360426c34 merge main into amd-staging (#2764)
4154ada1d485 [Xtensa] Implement Xtensa Floating Point Option. (#136086)
8f01edfa112f [C API] Add getter/setter for samesign flag on icmp (#145247)
40eee8ec7fb6 [AMDGPU] Add s_setprio_inc_wg gfx1250 instruction (#145152)
f9c9a32e9cbe [LV] Add tests with fmax reductions without fast-math flags.
0cb9c8f0006f merge main into amd-staging
99cdc26c94ee [CI] Cleanup buildkite test report script
214ca3161bf5 [CI] Test all projects when CI scripts change
80b9fcf8fdf2 Revert "[lldb] Fix qEcho message handling (#145072)" (#145241)
ee414e35047c [CI] Refactor out some early exits in compute_projects
f2eb5d416ed1 [SelectionDAG] Handle `fneg`/`fabs`/`fcopysign` in `SimplifyDemandedBits` (#139239)
078475d6c153 [ARM] Add test coverage for #144845 and regenerate tests. NFC
cfd6ab5a299e merge main into amd-staging (#2762)
584a2c2e7c54 AMDGPU: Avoid report_fatal_error for reporting libcalls (#145134)
ee5dcdc275b1 [mlir] fix assertion failure in remove-dead-values (#144849)
96493c514efa AMDGPU: Use reportFatalUsageError for regalloc flag error (#145198)
dec93ae45492 [mlir] Migrate away from ValueRange(std::nullopt) (NFC) (#145210)
2ac293f5ac4c [lld] Use llvm::partition_point (NFC) (#145209)
5d7d8d627a41 [Instrumentation] Drop "const" from a return type (NFC) (#145208)
0dc0aeb14f1e [LV] Add additional tests for replicating calls returning structs.
26390f22b8aa [VectorCombine] foldShuffleOfShuffles - fold shuffle(shuffle(x,y),poison) length changing masks (#144690)
1c20ae381202 merge main into amd-staging
d2c0451d05d9 [lldb][DWAFASTParserClang][NFC] Rename GetCXXObjectParameter to GetObjectParameter
c7d9eabf4a9c [AVR] Don't apply post-indexing on mismatched pointers (#145224)
f78819aeef32 Revert "Revert "[RISCV] Remove B and Zbc extension from Andes series cpus." (#144402)"
8583882bdcab [AArch64] Remove unnecessary DL variable. NFC
075cb691a5e8 [MLIR] Add logging/tracing to DataFlow analysis and RemoveDeadValues (NFC) (#144695)
6ae5b89553a9 Make getStridesAndOffset const (#145148)
58b939abe508 [VPlan] Support matching constants in narrowInterleaveGroups.
cbfec48697ad Revert "[X86][NFC] Use std::move to avoid copy" (#145215)
448c18b7cf75 merge main into amd-staging (#2759)
f51d8730b309 [InstSimplify] Simplify 'x u>= 1' to true when x is known non-zero (#145204)
fa0b84f23c08 [AMDGPU] Rename call instructions from b64 to i64 (#145103)
dfa676c23357 merge main into amd-staging
b7d0c9b9d8e2 [SelectionDAG][RISCV] Treat zext nneg as sext in PromoteIntOp_ZERO_EXTEND if the promoted input is sign extended. (#145120)
f280d3b705de AMDGPU: Avoid report_fatal_error for getRegisterByName subtarget case (#145173)
570885128351 Revert "[ValueTracking] Improve `Bitcast` handling to match SDAG" (#145191)
24b502018bef merge main into amd-staging (#2758)
757c80d88a6a Add `REQUIRES: x86` to test added in 141197 to skip when x86 target is not present.
c455f4a32d91 [LV] Add more tests for narrowing interleave groups with live-ins.
e066f35c6981 [lldb] Fix qEcho message handling (#145072)
bce0c5b0a4ca merge main into amd-staging
056b52df344f [clang-doc] Precommit test for global variables (#145069)
550ed9ef198e [gn build] Port e7dd223ec451
e7dd223ec451 [clang-tidy] Add new check `readability-use-numeric-limits` (#127430)
2ed089fb18b9 [clang/docs] add a missing brace in ClangFormatStyleOptions.rst (#145145)
6c8c816b175b [libc] Fix feature check for riscv (#145169)
0c47628515dc Re-commit "[RISCV] Properly support RISCVISD::LLA in getTargetConstantFromLoad. (#145112)"
fc36e47a4959 Revert "[RISCV] Properly support RISCVISD::LLA in getTargetConstantFromLoad. (#145112)"
1da864b574f6 [RISCV] Properly support RISCVISD::LLA in getTargetConstantFromLoad. (#145112)
9f7a15539441 [VPlan] Update packScalarIntoVector to take and return wide value (NFC)
4c1a1009ad8a [X86][NFC] Use std::move to avoid copy (#141455)
463ce0131047 [CodeGen] Remove extraneous casts to VariableID (NFC) (#145144)
ae372bfca890 [CodeGen] Use range-based for loops (NFC) (#145142)
e6ebf8f99ba2 [mlir] Migrate away from ArrayRef(std::nullopt) (NFC) (#145140)
99af99c6657d [TSan] Fix p == end == ShadowMem::end in ShadowSet (#144994)
a961ba88e1d4 AMDGPU: Use reportFatalUsageError for LDS mixed absolute addresses (#145135)
60d1276b0e0d [VPlan] Pass operand index to canNarrowLoad. (NFC)
281329d842de merge main into amd-staging (#2756)
831fcb5e91a6 [libc++] constexpr flat_map (#137453)
77941eba7f01 [CostModel] Add a DstTy to getShuffleCost (#141634)
7f46e193edba merge main into amd-staging
ea321392ebc4 [C++][Modules] A module directive may only appear as the first preprocessing tokens in a file (#144233)
1b5d6ec68553 [clang-tidy] count class member initializers as statements in 'readability-function-size' (#131669)
369e8403b616 [gn build] Port 1bb2328fd3ad
1bb2328fd3ad [libc++] Implement `views::join_with` (#65536)
d6a486c221c1 [ORC] Apply MachO::CPU_SUBTYPE_MASK to comparison in getDylibInterfac… (#145154)
2f5d965bb598 [VPlan] Use EMIT-SCALAR when printing casts.
bb372963dfce [MLIR] Add optional cached symbol tables to LLVM conversion patterns (#144032)
0921bfd81d57 [mlir][Transforms] Dialect conversion: Add missing erasure notifications (#145030)
4a4582dd788b [CIR] Upstream BinAssign for ComplexType (#144868)
5f2135df171e AMDGPU: Really delete AMDGPUAnnotateKernelFeatures (#145136)
c69f97cb95a2 [Clang] Add WG21 papers from the Sofia meeting to the status page
685af55fe004 [libc++] Simplify <limits> a bit (#140021)
2050d2e1815b [libc++] Simplify std::function further after removing allocator support (#144443)
32dbaf12be54 [mlir][Transforms][NFC] Dialect conversion: Rename internal functions (#145018)
437346378fd4 [GlobalISel] Widen vector loads from aligned ptrs (#144309)
bc31fe0062db merge main into amd-staging (#2754)
1fec092fd74a [AArch64][GlobalISel] Allow selecting FPR index loads. (#143835)
cb4f329004b8 [RISCV] Fix HasStdExtCOrZcfOrZce Syntax (#145141)
9cc9efc48333 [lld][COFF] Remove duplicate strtab entries (#141197)
0fa0c3c2333c AMDGPU: Use reportFatalUsageError in AMDGPULowerBufferFatPointers (#145132)
981f8e1380b6 [TableGen] Remove redundant control flow statements (NFC) (#145143)
4c2b9317f8fc [mlir][ods] Document InferTypeOpInterface behavior. (#145060)
70312802184f AMDGPU: Use reportFatalUsageError for unsupported code object version (#145133)
fa117715ca36 [RISCV] Implement Feature Bit for Q (#145001)
f4661310550d AMDGPU: Use reportFatalUsageError in AMDGPULowerModuleLDS (#145130)
9aa18946e798 merge main into amd-staging
17e8465a3eb0 AArch64: Replace AArch64MCExpr with MCSpecifierExpr
72979093e79c Revert "[Reland][InstCombine] Iterative replacement in PtrReplacer" (#145137)
8050a6e0732c [clang-doc] add support for concepts (#144430)
6d18060e4223 Manual update of LLVM_MAIN_REVISION to 541644 (#2752)
2dfcc4375faa [clang-doc] Precommit concept tests (#144160)
c6f9ce297faa merge main into amd-staging (#2750)
6b9fe9e0bc74 [HLSL] Emit a version in the dx.rootsignatures metadata (#145113)
be7e4113c843 [NFC] Add comment to describe the intention use of newly added `avail-extern-gv-in-addrspace-to-local` (#144911)
ad36800d6525 [bazel][lld] Follow up 491b82a5ec1a (llvmorg-21-init-16192-g491b82a5ec1a)
353fbecaadeb Manual update of LLVM_MAIN_REVISION to 541644
1753aba0342b [X86] combineINSERT_SUBVECTOR - directly fold to X86ISD::SUBV_BROADCAST_LOAD to prevent vector split infinite loop (#145077)
1db9afb102ac [lldb/crashlog] Make registers always available & fix x29/x30 parsing (#145104)
b7be8786af42 Reapply "[CI] Migrate to runtimes build" (#143612)
94865edfa85a [Reland][InstCombine] Iterative replacement in PtrReplacer (#144626)
e6ee2c7c7b36 [HLSL][RootSignature] Implement validation of resource ranges for `RootDescriptors` (#140962)
a91154343780 [libc] Implemented wcrtomb internal function and public libc function (#144596)
486a24ec34ac [openmp] - Ensure openmp-config.cmake honors proper install suffix (#2745)
f0d32575a119 [BOLT][NFCI] Use FileSymbols for local symbol disambiguation (#89088)
5886f0a18342 [RISCV] Allow larger offset when matching build_vector as vid sequence (#144756)
0c2191b3a79d [AMDGPU] Omit image waits in function prologue on gfx1250 (#145097)
4c97a91dc085 [libc] Added closing quote (#145101)
9c1ce31f5463 [mlir][vector] Add unroll patterns for vector.load and vector.store (#143420)
b6445ac0c549 [GlobalISel] Create a common register_vector_matchinfo (#144306)
58f48011b322 [lldb] Add support for x86_64h to scripted process (#145099)
6110dead894b [clang][scan-deps] Add option to disable caching stat failures (#144000)
491b82a5ec1a ELF: Add branch-to-branch optimization.
3b9795b3d3b2 [AArch64] Add CodeGen support for scalar FEAT_CPA (#105669)
b506ab2f0c8c merge main into amd-staging
f8ffb4e7cd94 [VPlan] Simplify ExtractLastElement(Broadcast(A)) -> A.
d078ce7c98a3 [libc] mbrtowc implementation (#144760)
3a66e2065296 [AMDGPU] Add gfx1250 runlines to vop3 dpp tests. NFC. (#145089)
d8924d4da78f [LLDB] Explicitly use python for version fixup (#144217)
4959e8a1dadd [BOLT][NFCI] Use heuristic for matching split global functions (#90429)
6d8c6ef90c1a [BOLT][NFC] Simplify doTrace in BAT mode (#143233)
227f759644bb [mlir][python] expose operation.block (#145088)
affcc5e728c8 [AMDGPU] Add s_wait_xcnt gfx1250 instruction (#145086)
2a4207e7322c [DirectX] Don't limit visitGetElementPtrInst to global ptrs (#144959)
958dc8602651 [AMDGPU] Don't insert wait instructions that are not supported by gfx1250 (#145084)
b5d5708128e9 [HLSL] Add descriptor table metadata parsing (#142492)
8d2eea96b391 [AMDGPU] gfx1250 SOPP MC tests. NFC. (#145082)
c103bbc83669 [LV] Consider whether vscale is a known power of two for iteration check (#144963)
f159774352c3 [mlir][core|ptr] Add `PtrLikeTypeInterface` and casting ops to the `ptr` dialect (#137469)
925dbc798828 [flang][OpenMP] Fix namespace nesting after PR144960
3f42c6bddd24 [DirectX] Scalarize `extractelement` and `insertelement` with dynamic indices (#141676)
521adc9fa270 [VPlan] Use createScalarZExtOrTrunc when expanding expandVPWidenIntOrFpInduction
ff6367b47071 [[mlir][Vector] Add simple folders for `vector.from_element`/`vector.to_elements` (#144444)
bae48ac3c0e6 [mlir][bazel] add missing deps for XeGPUTransforms
7f74a377d094 [LV] Regenerate uniform_across_vf* check lines.
ab8b8c1e138a [TargetParser][cmake] Be Smarter about TableGen Deps (#144848)
e40b9c22d636 suppress intermittent mlir dbg test for autopromo to pass (#2744)
04e2e581ac00 [RISCV] Treat bf16->f32 as separate ExtKind in combineOp_VLToVWOp_VL. (#144653)
adc6228ea07e [mlir][xegpu] Refine layout assignment in XeGPU SIMT distribution. (#142687)
016868be0d8c merge main into amd-staging (#2739)
0d21c956a5c1 [BPF] Handle nested wrapper structs in BPF map definition traversal (#144097)
33a92af1b226 [msan] Add off-by-default flag to fix false negatives from partially undefined constant fixed-length vectors (#143837)
f8ee5774b612 [X86] combineConcatVectorOps - only concat AVX1 v4i64 shift-by-32 to a shuffle if the concat is free (#145043)
65cb3bcf327d [Clang][PowerPC] Add __dmr1024 type and DMF integer calculation builtins (#142480)
8d6e29d0d3e3 [libc] Reworked CharacterConverter isComplete into isFull and isEmpty (#144799)
7157f33c6c88 [libc++] Disable a std::unexpected test in modules build (#144466)
6ddb3a69c159 [AMDGPU] Add another test showing unwanted VALU codegen (#145062)
945ce1aa3d29 [libc++] Update the value of __cpp_lib_constrained_equality after P3379R0 (#144553)
edbaf19c46c6 [AMDGPU] Fix a potential integer overflow in GCNRegPressure when true16 is enabled (#144968)
379a609dadc1 [mlir][arith][transforms] Adds f4E2M1FN support to truncf and extf (#144157)
940ff110d78b [InstCombine] fix hwasan mistake in "remove dead loads" (#145057)
877511920dcf Revert "[lldb][DWARF] Remove object_pointer from ParsedDWARFAttributes" (#145065)
bfef8732be1b msan: Support free_sized and free_aligned_sized from C23 (#144529)
6ba1955ba22c [flang][OpenMP] Fix ignore-target-data.f90 test
72de0e45846b [TableGen][Docs] Fix empty list syntax in TableGen doc. (#145041)
770b16cd4939 [BOLT][test] Update X86/perf2bolt-spe.test (#145061)
32fc625a3fa2 Reapply "Reapply "[clang][bytecode] Allocate IntegralAP and Floating … (#145014)
151ee0faad42 [X86] SimplifyDemandedVectorEltsForTargetNode - ensure X86ISD::VPERMILPV node use v2f64/v4f32 types
749e4a53d252 [lldb] Fix ASCII art in CommandObjectProtocolServer (NFC)
6e86b7e34b94 [AMDGPU] Do not replace SALU floating point multiply with VALU-only ldexp (#145048)
78971916da04 [gn build] Port 9524bfb27020
9524bfb27020 [lldb] Add Model Context Protocol (MCP) support to LLDB (#143628)
887222e3526f [mlir] Migrate away from ArrayRef(std::nullopt) (NFC) (#144989)
3f1de197b1c3 [Offload] Rework compiling device code for unit test suites (#144776)
c734377544fc [DAG] foldMaskedMerge - fix Wparentheses operator precedence warning. NFC.
b533b0ec34ac Define a DataFlowSolver helper that loads sensible default analyses (#143415)
71e20c6c86e0 Fix references to required libraries when building LLVM with ASAN and MultiThreaded[Debug] on Windows (#139657)
05b4bfe19eab [ExtractAPI] Include +/- symbols for ObjC methods (#145035)
9dc59cc95b17 [MLIR] Incorrect track of usedKey in setPropertiesFromParsedAttr (#144789)
b017b4ce9a45 [lldb][DWARF] Remove object_pointer from ParsedDWARFAttributes (#144880)
95c6c11c747d [X86] combineConcatVectorOps - only always concat logic ops on AVX512 targets (#145036)
20d57e77f670 [lldb][AIX] Added base file for AIX Register Context (#144645)
a2b8a93ff9cf [VPlan] Pass NumUnrolledElems as operand to VPWidenPointerInductionRecipe. NFC (#119859)
96ab74bf175f [InstCombine] remove undef loads, such as memcpy from undef (#143958)
f242360e156b [Offload] Add type information to device info nodes (#144535)
e0633d59b9d0 [Offload] Check for initialization (#144370)
bd36f7331a9f [CIR] Add initial support for bitfields in structs (#142041)
1c4bc3ae0472 suppress intermittent mlir dbg test for autopromo to pass
225768d1f9f2 [X86] combineConcatVectorOps - add tests showing v4i64 shift-by-32 with unnecessary concatenation
3ff69c80786e [LV] Add early-exit-with-store tests (#140899)
269cb22ae82f [mlir][transform] extract a minimal DomainAndOperandsAffineMapT… (#145034)
376b71442d03 [NFC][TableGen][DecoderEmitter] Use structured binding in range for loop (#144890)
c3929fbf0ef2 [NFC][LLVM] Reorder SVE_SME codegen predicate classes.
8db272ffcf9a [mlir][SparseTensor] Re-enable tests on AArch64 (#143387)
152d4b8a01e8 [AArch64] Use indexed dup for 128b segmented splat (#144688)
349f8d67d4ee [flang][OpenMP] Skip runtime mapping with no offload targets (#144534)
c8c4bd1ebc6e [LV] Stengthen loop-invariance checks in isPredicatedInst (#140744)
b334ffd4f4e7 [VPlan] Refine return types in VPBuilder (NFC) (#108858)
6edf2eb36470 [MLIR] Print more user-friendly error message when generating local reproducer and threading is enabled (#144905)
6c0ac888c59c [mlir][arith][NFC] Remove redundant trait declaration (#145007)
fd97dfbb78e3 [LV] Don't mark ptrs as safe to speculate if fed by UB/poison op. (#143204)
5835f1e0a33a [AsmPrinter] Fix crash when remarks section is unsupported (#144724)
a5b1093f7827 [MLIR] Add ReturnLike trait to memref.atomic_yield (#144932)
e970f59e6b20 [flang][OpenMP] Reintroduce TODO for FIR lowering of linear clause (#144883)
8650c801381d [flang][OpenMP] Do not skip privatization of linear variable if it is OmpPreDetermined (#144315)
e8be733a3c33 [VPlan] Remove redundant ExtractLastElement from vector-to-scalar VPI.
1b83f10072b3 [AMDGPU] Fix to prevent sinking of PERMLANE_SWAP instruction (#144423)
e5559ca45f21 [flang][OpenMP] Move lowering of ATOMIC to separate file, NFC (#144960)
cbd496581fb6 [NFC][AMDGPU] Automate any_extend_vector_inreg.ll check line generation (#145013)
a700dbafaf94 merge main into amd-staging
3e99aa6c0a36 [PredicateInfo] Clean up DFS sorting (NFC) (#144943)
4ec6d127c185 [X86] movmsk-cmp.ll - regenerate VPTERNLOG asm comments
68732ce8e019 [LLVM][CodeGen][SVE] Add isel for bfloat unordered reductions. (#143540)
227cd56e1330 [NFC] Update transform-op-pad-tiling-interface.mlir
7af545237f85 [NFC] Update transform-op-pad-tiling-interface-multiple-of.mlir
00c18d04ab63 [mlir][Transforms] Add a PadTilingInterface transformation and hook i… (#144991)
7085065c02da [BOLT] Support pre-aggregated returns (#143296)
a5fa5bd2a890 [X86] ptest.ll - add test coverage for #144861 load chains
95bd05d7cae3 [mlir][Func][NFC] Simplify implementation after #144706 (#145006)
238abf8ba823 [gn build] Port f75973949b0e
f75973949b0e [BOLT][AArch64] Add support for SPE brstack format (#129231)
dd4776d429bd AMDGPU: Remove AMDGPUInstrInfo class (#144984)
f577516d91dc [mlir][arith] Add back ElementwiseMappable to `arith.trunci` (#145000)
61972054f3fc [CodeGen] Limit number of analyzed predecessors
eb0f1dc00e5d [LLD][ELF] Include offset when adding Thunk symbols (#144995)
b85387dfe8e6 [SCCP] Check instruction type before querying PredicateInfo (NFC)
f704782c63ac [AArch64][SelectionDAG] Fix UDOT regression (#144907)
14e89b061fde [C++20] [Modules] Add exported modules as transitive imported modules
874773635d31 [SLP] NFC: Simplify CandidateVFs initialization (#144882)
b6b8fa3b15d3 [llvm-cov][gcov] Support multi-files coverage in one basic block (#144504)
bc14e5e5e9c7 [AMDGPU] Add trivial gfx1250 runlines to MC tests. NFC. (#144988)
8973be462c49 [PredicateInfo] Avoid duplicate stack in scope check (NFC)
c361bffa50f1 AMDGPU: Remove legacy pass manager version of AMDGPUUnifyMetadata (#144985)
1cae21da47b1 AMDGPU: Remove legacy PM version of AMDGPUPromoteAllocaToVector (#144986)
cbb5e244f756 [PredicateInfo] Remove redundant EdgeOnly member (NFC)
d196124dd223 [PredicateInfo] Remove unnecessary EdgeUsesOnly set (NFC) (#144912)
090f409538d2 [gn build] Port 0f302f38b001
06e08f38e1ec AArch64: Use reportFatalUsageError for unsupported calling conv (#144840)
0f302f38b001 [clang] Add managarm support (#144791)
b8e723cca5b7 merge main into amd-staging (#2735)
7cbb1411550e [clang] Migrate away from ArrayRef(std::nullopt) (NFC) (#144982)
b8d3efa18962 [BOLT][Linux] Fix linux_banner lookup (#144962)
69974658f079 [AMDGPU] Initial support for gfx1250 target. (#144965)
64fe3236476a [llvm] Migrate away from ArrayRef(std::nullopt) (NFC) (#144967)
618b3f0d8923 merge main into amd-staging
bb51c5d4b85a [SDPatternMatch] Add m_Poison matcher (#144860)
8c1a970f7ead merge main into amd-staging (#2734)
50c5ecd35402 [NFC] [Serialization] Remove a useless comment
4f991cc99523 [lldb-dap] Make connection URLs match lldb (#144770)
513bcf6d012f [yaml2obj] Report error when the input filename does not exist
a05393a879b2 [RISCV] Add symbol parsing support for XAndesPerf branch instructions (#137748)
5cbed34404a3 [X86] Remove CLDEMOTE from Arrowlake and later hybrid processors (#144833)
fa7646008ef3 [DirectX] Add Root Signature Version Support and Update Test IR Format (#144957)
efd42b9b1d65 WebAssembly: Stop directly using RuntimeLibcalls.def (#143054)
91439817e8d1 ARM: Avoid using isTarget wrappers around Triple predicates (#144705)
6001a8bb9457 [WholeProgramDevirt] Add check for AvailableExternal and give up icall.branch.funnel (#143468)
6581268fd31e merge main into amd-staging
dad64877c811 [llvm] Remove an extraneous cast (NFC) (#144955)
d8e6d74c6905 [LV] Consider EVL legality for TTI tail folding preference (#144790)
8b92248febc3 merge main into amd-staging (#2730)
b96370131d15 [TTI] Plumb CostKind through getPartialReductionCost (#144953)
dfb5cadf5e81 [SPARC][IAS] Properly set implied feature sets for ISA levels/extensions (#143232)
a8edda195c9f [llvm] Remove unused includes (NFC) (#144941)
a9d175f1735a [CodeGen] Use range-based for loops (NFC) (#144939)
d3a2931d8af8 [libc] Use `raise_except_if_required` for log2f. (#144961)
53ea522d1b87 [LV] Introduce and use VPBuilder::createScalarZExtOrTrunc [nfc] (#144946)
2f3a8fd0b332 [libc] Skip x87 floating point register and only update mxcsr for x86_64 targets when raising exceptions inside math functions. (#144951)
53336ad488c9 [Offload] Move (most) global state to an `OffloadContext` struct (#144494)
9fd22cb56d4c [flang][NFC] Move new code to right place (#144551)
03692aa40487 [memprof] Use a lambda instead of std::bind (NFC) (#144940)
3b672e1d7b73 [llvm] Use "= delete" to delete constructors (NFC) (#144938)
7349864d2c7c [ADT] Remove an extraneous variable (NFC) (#144937)
dc058a3d84ed [TableGen] Use ListSeparator (NFC) (#144936)
28808dda2c53 [flang] Fixed test added in #144830.
b7891e5798e4 Xfail ../flang/test/Lower/OpenMP/target-enter-data-default-openmp52.f90
6d8edc0288e7 merge main into amd-staging
8631b4f1b4f3 [flang] Set low probability for array repacking code. (#144830)
baf35d7a829e [libc] Fix bug in LIBC_CONF_ERRNO_MODE being undefined (#144896)
633e740e3453 [Clang][AMDGPU][Driver] Add `avail-extern-gv-in-addrspace-to-local` option when ThinTLO is enabled (#144914)
3bee9ba0156e AMDGPU/GFX12: Fix s_barrier_signal_isfirst for single-wave workgroups (#143634)
403f62cd116d Reland "[flang] Restructure runtime to avoid recursion (relanding) (#143993)" (#2725)
74ec1c287a88 [LV] Add tests interleaving extended and multiply/accumulate reductions.
836201f1177c Allow bf16 operands on new MFMAs (#144925)
f780955e1df9 lsan: fix macos build after #144604 (#144818)
0b8179b2adbc [ValueTracking] Improve `Bitcast` handling to match SDAG (#125935)
3246e45cf074 [Comgr] Handle amdgcnspirv when used as an ISA name
73694ef880ca [hipcc] fix cmake warnings
c0c71463f6bc [InstCombine] Optimize sub(sext(add(x,y)),sext(add(x,z))). (#144174)
6ce86538c11b [mlir][cf] Preserve branch weights during cf.cond_br canonicalization. (#144822)
0816bb32ac37 [Matrix] Fix heap-use-after-free after 0fa373c77ded203eddb.
3fe62682ef9c [mlir][vector] Use `result` consistently as the result argument name (#144739)
c0cc81cdc03c [NFC][Clang] Adopt simplified `getTrailingObjects` in ASTReader (#144438)
5cf7d871b030 [NFC][Clang][AST] Adopt simplified `getTrailingObjects` in AST (#144432)
b8337349d9b6 [libc][math] Skip setting errno and floating point exception for math functions when LIBC_MATH flag has LIBC_MATH_NO_ERRNO and LIBC_MATH_NO_EXCEPT. (#144920)
c1ac87b32786 [libc] Fix setjmp build order. (#144917)
e16dfd05fd8e [Comgr] Add test for Comgr Profiler (#2676)
01d648a42939 [HLSL][SPIRV] Reapply "[HLSL][SPIRV] Add vk::constant_id attribute." (#144902)
f4db14229cd9 [SCCP] Move logic for removing ssa.copy into Solver (NFC)
36af7345dfb8 Reapply "[Clang] Enable -fextend-variable-liveness at -Og (#118026)"
3de01d07c33c Fix bazel build after #144594, mark variable as potentially unused (#144910)
c96bb6f129d8 [RemoveDIs][NFC] Remove dbg intrinsic handling code from SelectionDAG… (#2723)
5eb24fde11cd [SelectionDAG][RISCV] Preserve nneg flag when folding (trunc (zext X))->(zext X). (#144807)
fdb572681de7 Fix bazel build after #144594 (#144904)
3516ad05dfd6 [RISCV] Update SpacemiT X60 scheduling latencies based on hardware measurements (#144730)
eb694b28461f [mlir][arith] Delete mul ext canonicalizations (#144844)
89efae916a5d [Flang][OpenMP] Update default MapType for Map Clauses and OpenMP 5.2 (#144715)
19360e62d0d1 [gn build] port bf79d4819ede (ppc -gen-target-features)
4c6f398b8660 [Clang] Add standalone AMDGPU SPIR-V toolchain (#144576)
7e8f1f5f7275 [HIP] Remove dots in HIP runtime path (#143792)
802fa92aee35 [PredicateInfo] Avoid duplicate hash lookup (NFC)
de1e8f73b3b5 merge main into amd-staging (#2722)
fe0ce9beeca4 Reland "[flang] Restructure runtime to avoid recursion (relanding) (#143993)"
a4e4527c4b44 [mlir][Transforms] Fix replaceUsesOfBlockArgument API (#144706)
278ece7c80d3 [InstCombine][NFC] Pre-commit tests for #125935 (#144111)
e75e2485f2e5 [AMDGPU][True16][Codegen] keep srcmod/clamp/omod from v_s_xxx_f16 when moved to VALU (#144781)
f87b6625d64c [MLIR][NFC] Fixed some Typos (#144263)
bf79d4819ede [Reland] [PowerPC] frontend get target feature from backend with cpu name (#144594)
5645d6710904 Implement `sigsetjmp` and `siglongjmp` for darwin/aarch64 (#139555)
83381ba832a5 [LLDB] Add negative number parsing to DIL (#144557)
493a359237e8 [lldb][AArch64] Fix live process test for Linux's mte_ctrl register
e873fd157eda [BOLT] Gadget scanner: do not crash on debug-printing CFI instructions (#136151)
09e794c4bb13 [HIP] Emit the CUID value in the module with the new driver (#144570)
c079040eea5c [lldb] add has methods to all DemangledNameInfo attributes (#144549)
dae5104eed45 [lldb][DWARFASTParserClang] Make GetCXXObjectParameter public and call it from unit-tests (#144879)
5148e085386f Fix build issue caused by commit #0fe78c4 (#144888)
c4c2d777f4ae [VPlan] Fix handling of ReductionStartVector for rdxs when unrolling.
e478a22d540d [LLVM][IRBuilder] Use NUW arithmetic for Create{ElementCount,TypeSize}. (#143532)
936c5566db01 [flang][OpenMP] Handle REQUIRES ADMO in lowering (#144362)
2b4d75729022 [BOLT] Gadget scanner: detect authentication oracles (#135663)
eeb5ac0f5373 [RemoveDIs][NFC] Remove dbg intrinsic handling code from SelectionDAG ISel (#144702)
b73720cf6c53 [analyzer] Conversion to CheckerFamily: DynamicTypePropagation (#144735)
e33f13ba4824 [mlir][arith] Add overflow flags to `arith.trunci` (#144863)
046e2f545ef5 [LV] Add interleaving test with partial reductions and non-const start.
30824c449a89 [lldb][DWARFASTParserClang] GetCXXObjectParameter to take DeclContext DIE parameter (#144876)
d94d5596c3b4 merge main into amd-staging
74054cab7a3e [HashRecognize] Make it a non-PM analysis (#144742)
6cb1133528f3 Remove old dbg intrinsics and reland 9eb0020555fc (#2712)
656f54b55aec merge main into amd-staging (#2717)
1c35fe4e6b25 RuntimeLibcalls: Pass in exception handling type (#144696)
305953a32ded MC: Move ExceptionHandling enum to Support (#144692)
5bee2c34bde1 RuntimeLibcalls: Pass in FloatABI and EABI type (#144691)
0fe78c4a2905 [NFC] [Serialization] Some Code Cleanups for Name lookup table things
af51c9d9df9d [LV][NFC] Add branch weight test showing incorrect behaviour (#144682)
97c1a2444574 [mlir][linalg] Add option to pad dynamic dims to `linalg::rewriteAsPaddedOp` (#144354)
c0a9c908a697 [MLIR][NVVM-Docs] Fix rendering of a few tables in NVVM Docs (#144764)
6273c5d4d354 [lldb][AArch64] Correctly invalidate svg when vg is written (#140875)
20245bbf6697 [libc++][NFC] Format abs.pass.cpp test
584cc3768705 [libc++] Move std::abs into __math/abs.h (#139586)
5f94bc37f414 merge main into amd-staging
681db064d221 [TSan] Make Shadow/Meta region inclusive-exclusive (#144647)
8f82c027c896 [AMDGPU] New RegBankSelect: Add rules for `G_PTRTOINT` and `G_INTTOPTR` (#142604)
62fe5e428acc [NFC][AMDGPU] print more info when debugging SIInsertWaitcnts pass (#144629)
0e1aab1ec833 [AMDGPU] Improve test coverage for G_INTTOPTR and G_PTRTOINT (#142603)
1ab0e7dd60e2 [LICM] Hoisting writeonly calls (#143799)
650b451d0065 [libc++] Simplify the implementation of pointer_traits a bit (#142260)
52ff58c33003 Revert "[CUDA][HIP] Add a __device__ version of std::__glibcxx_assert_fail()" (#144850)
7ceea22a7ada [AMDGPU] New RegBankSelect: Add Ptr32/Ptr64/Ptr128 (#142602)
26d4b3cb4ca2 [RISCV] Don't prefer QC_EXT for SEXT_INREG patterns when Zbb is enabled (#144837)
db8e6fc64534 [AMDGPU] New RegBanKSelect: Add S128 types (#142601)
e73bff89ef8e [AMDGPU] New RegBankSelect: Handle all 32/64 bit pointer types for B32/B64 rule (#142560)
50a7511138a4 [BOLT][AArch64] Fix PREL Relocs on RHEL8 (#144505)
a9a71b6d3118 [libc++][test] Don't pass ill-formed UTF-8 to MAKE_STRING_VIEW (#136403)
3e795c60c73e [lldb] Disable TestTargetWatchAddress on Windows x86_64 (#144779)
408e55098d7d [RISCV] Add support for handling one tied operand in the source instruction for compress patterns (#143660)
9ee55e717308 [CIR] Implement folder for VecSplatOp (#143771)
2c2ad9a096e7 Reapply "[Clang] Profile singly-resolved UnresolvedLookupExpr with the declaration" (#140680)
03461c9c6e21 [mlir][gpu][spirv] Remove rotation semantics of gpu.shuffle up/down (#139105)
590066bee70d [NVPTX] Add family-specific architectures support (#141899)
ac2b7c192a4e merge main into amd-staging (#2715)
7b989ade35a4 [lldb/crashlog] Make interactive mode the new default (#144839)
b2060516acf8 [Comgr][NFC] Create level 3 heading for env vars
1175145be827 merge main into amd-staging (#2710)
bb6ea768bfc7 merge main into amd-staging
351303c28e8f [mlir][docs] Fix broken links to Traits documentation. (#144820)
5875fafdc547 [X86] Remove CLDEMOTE from Alderlake and later hybrid processors (#144662)
bfee625821c0 [NVPTX] Attach Range attr to setmaxnreg and fence intrinsics (#144120)
faf9295f4e3a [RISCV] Fix a bug where AVL is the last MI in MBB. (#144668)
2bcdfa198aa5 [CIR] Add side effect attribute to call operations (#144201)
6e5ee4aa98f1 [RISCV] Save vector registers in interrupt handler. (#143808)
874a02f05b6e ARM: Move ABI helpers from Subtarget to TargetMachine (#144680)
bc8908a4e93b ARM: Move declaration of supportSplitCSR to be public (#144679)
d265105b8f50 Augment `uncaught-exception.test` fuzzer test to be msvc-compatible (#125924)
bf32c1aa75fa merge main into amd-staging
b60b3d9bb4ac merge main into amd-staging (#2708)
5f69d680e2cc Revert "[HLSL][SPIRV] Add vk::constant_id attribute." (#144812)
61c8b31cb57d Reapply "[DebugInfo][RemoveDIs] Remove a swathe of debug-intrinsic code (#144389)"
691448b8a8e2 [HeterogeneousDwarf] Remove remants of old approach
bb1f5c3189c4 [AArch64] Lower jump table cases threshold to 10 (#143632)
780c0ef7fb97 [llvm] explicitly link llvm-jitlink-executor with ExecutionEngine (#144778)
e0933ab5ae48 Revert "[lldb][target] Add progress report for wait-attaching to process" (#144810)
00ae89a1cbec Revert "[IPO] Added attributor for identifying invariant loads" (#144808)
7b6963ea672f [compiler-rt] [Fuzzer] Fix tests linking buildbot failure (#144495)
118bfcda46c1 [MLIR][XEGPU] Add blocking support for scatter ops (#144766)
51aa6a4993ea [lldb-dap] Use protocol types for ReadMemory request (#144552)
c4d7ea804968 [llvm-debuginfo-analyzer] Apply various memory savings in Core/LVxxx base classes (#144399)
d10079e305ac [RISCV] Reduce the VL of both operands in VMERGE_VVM (#144759)
ac37a0df949a [mlir] Fix integer comparison warning (#144794)
4dca4459a328 [CIR] Upstream ComplexType builtin_complex (#144225)
03bdc0a1f68a [lldb][target] Add progress report for wait-attaching to process (#144768)
86d1d6b2c0c1 [clang] Use TargetInfo to determine device kernel calling convention (#144728)
7aecd7ecacb4 [mlir][Vector] Add `vector.to_elements` op (#141457)
b85e92990fde Hexagon v87 v89 elf flags (#144584)
96bbe472ef01 Revert "[mlir][spirv] Fix int type declaration duplication when serializing" and follow up commits (#144773)
d67553495c62 Revert "[llvm-cov] Export decision coverage to output json" (#144783)
b9a9a1b9bbf0 Revert "[RemoveDIs][NFC] Remove dbg intrinsic handling code from SelectionDAG ISel (#144702)"
a88e65580965 [llvm] build Blake3 source with LLVM_EXPORTS defined (#144753)
a630ca6f6c47 [lldb][breakpoint] Grey out disabled breakpoints (#91404)
23b8f11b27f1 [VPlan] Remove redundant VPWidenRecipe constructors (NFC)
22a69a266d82 lsan: Support free_sized and free_aligned_sized from C23 (#144604)
6a5452238e5e merge main into amd-staging
d4b7c0d8b437 [Remarks] Auto-detect remark parser format (#144554)
67c52aacae2a [CIR] Upstream support for IncompleteArrayType (#144138)
d0262124db4e merge main into amd-staging (#2702)
c04fc5596ec8 [MemCpyOpt] allow some undef contents overread in processMemCpyMemCpyDependence (#143745)
fb0651959b1b [AArch64] fix trampoline implementation: actually use X15 (#143892)
88d250729eb0 Revert "[llvm-cov] Export decision coverage to output json" (#144783)
156a64c585fa [HashRecognize] Tighten pre-conditions for analysis (#144757)
f13b9e364366 [HashRecognize] Don't const-qualify Values in result (#144752)
a94eb27a29ef [HashRecognize] Fix big-endian CRC tables (#144754)
ca9a09dbe679 [libc++] Fix a typo in documentation (#144763)
ab6beeca9ccc [llvm-cov] Export decision coverage to output json (#144335)
8c3fbaf0ee73 [Clang][OpenMP][LoopTransformations] Fix incorrect number of generated loops for Tile and Reverse directives (#140532)
17f5b8b52a35 [flang][driver] add ability to look up feature flags without setting them (#144559)
3f3526f36d23 [NFC][PowerPC] pre-commit running the update_llc_test_checks.py for all-atomics.ll,loop-comment.ll etc (#144411)
071a6feabd7a [TTI] Remove PPC hasActiveVectorLength impl, simplify interface (NFC). (#142310)
dfe4d44d8de6 Revert "[VPlan] Remove unnecessary DomTreeUpdater flush (NFC)." (#144758)
6f4e4ea17745 [libc] Internal getrandom implementation (#144427)
835d3034fe96 [AArch64] improve zero-cycle regmov test (#143680)
82acd8c377e9 [PowerPC] Add code to spill and restore DMRp registers (#142443)
d9f7979a63ce sanitizer_common: add unsupported test for free_sized and free_aligned_sized from C23 (#144727)
298f1c276f4f Revert "Add missing intrinsics to cuda headers" (#144755)
77bc25485135 [AArch64] Fix build failure with -Werror (#144749)
2a8c65e983b3 [CodeGen][NFC] Fix quadratic c-t for large jump tables
d008d5c9c0c6 Regen lits
4084ffcf1e69 [flang] Show types in DumpEvExpr (#143743)
fe3933da15b5 [mlir][vector] Support complete folding in single pass for vector.insert/vector.extract (#142124)
00189211486d [DAG] add (~a | x) & (a | y) -> (a & (x ^ y)) ^y for foldMaskedMerge (#144342)
9827440f1e72 [libc++] Optimize ranges::{for_each, for_each_n} for segmented iterators (#132896)
dd40c460c42d [libc++] Clean up casts in std::forward_list (#130310)
2a41350aabd8 Fix bazel build issue caused by #142986 second attempt (#144721 didnt… (#144743)
6d785ca4218b [Clang] Fix the clang/test/PCH/ignored-pch.c test. (#144737)
13510c07364d [libc++] Make list constexpr as part of P3372R3 (#129799)
a2cee0544963 [scudo] Make report pointers const. (#144624)
0fa373c77ded [Matrix] Propagate shape information through PHI insts (#141681)
b5aaf9d988ff [InstCombine] Implement vp.reverse reordering/elimination through binop/unop (#143963)
5d502aeddf2a [flang][OpenMP] Clarify confusing error message (#144707)
9da9d32670dd [AMDGPU][True16][CodeGen] sext i16 inreg in true16 mode (#144024)
8b8a3699dbdb [AArch64] Use dupq (SVE2.1) for segmented lane splats (#144482)
3af4d4e8100f [HLSL][SPIR-V] Fix LinkageAttribute emission for BuiltIn (#144701)
b53c1e4ee810 [AArch64] Add ISel for postindex ld1/st1 in big-endian (#144387)
e4c3b037bc7f [InstCombine] Fold `tan(x) * cos(x) => sin(x)` (#136319)
8fc20bffabe7 Fix bazel build issue caused by 142986 (#144721)
36038a1048b2 [RemoveDIs][NFC] Remove dbg intrinsic handling code from SelectionDAG ISel (#144702)
516ebdc8da78 [openmp] - add support for openmp-config.cmake (#2299)
a1ff70eb52a5 merge main into amd-staging
6f4add34801e [compiler-rt] [Fuzzer] Fix ARMv7 test link failure by linking unwinder (#144495)
ffd7229d228a merge main into amd-staging (#2699)
ee070d08163a [mlir][bufferization] Support custom types (1/N) (#142986)
40d2f392106f [Sema][ObjC] Loosen restrictions on reinterpret_cast involving indirect ARC-managed pointers (#144458)
9db7502d229b [libc++] Move __has_iterator_typedefs to the up-to-C++17 implementation of iterator_traits (#144265)
1d6f1029f7e8 [mlir] [python] Fixed the return type of `MemRefType.get_strides_and_offset` (#144523)
c5613dc86350 [MLIR] Mark LLVM::FMAOp as legal (#144671)
bdac9580f3bc [nfc][jt] Drop `std::optional` pointers (#144548)
fda6b751f1b1 Fix libc++ restarter job.
671caef379c6 [Flang][OpenMP] Update relevant warnings to emit when OMP >= v5.2 (#144492)
1f34d68c4f08 [Remarks] Remove yaml-strtab format (#144527)
c4d99704e220 Revert "Reland [Driver] Add support for GCC installation detection in… (#144684)
6729da647afa [mlir][amdgpu][nfc] Add PatternBenefit to populate methods (#144663)
68471d29eed2 Revert "Reapply "[clang][bytecode] Allocate IntegralAP and Floating types usi… (#144676)"
8584abb05a84 [mlir] mlir/test/lit.local.cfg -> mlir/test/Target/SPIRV/lit.local.cfg (#144685)
a83d3362f686 [flang][OpenMP] Don't allow DO CONCURRENT inside of a loop nest (#144506)
4b2ab1494bc0 [flang][OpenMP] Don't crash on iterator modifier in declare mapper (#144359)
66580f77b826 [mlir][Transforms][NFC] Dialect Conversion: Keep `unresolvedMaterializations` up to date (#144254)
a1c2a7129398 [mlir][bufferization] Use Type instead of Value in unknown conversion (#144658)
6265ca686dfe [AArch64] Add Cortex-A320 scheduling model (#144385)
7c15edb30693 Reapply "[clang][bytecode] Allocate IntegralAP and Floating types usi… (#144676)
34a48941498d [X86] detectZextAbsDiff - use SDPatternMatch::m_Abs() matcher. NFC.
d8e8ab79773f [AArch64][SME] Fix restoring callee-saves from FP with hazard padding (#143371)
8a469da8b234 [mlir] remove unnecessary atomic_rmw expansions (#144515)
66d6964a5501 Fix tests failing on fuchsia clang x86_64 builders (#144655)
c3efe7d64ceb [libcxx] [test] Fix odr_signature tests with optimizations enabled (#144317)
c21440fb3dbe merge main into amd-staging
4aca3dc48b09 Reland: [GlobalISel] prevent G_UNMERGE_VALUES for vectors with different elements (#144661)
b5967264b0fb CodeGen: Move ABI option enums to support (#142912)
b7ef5dbac91f [LLVM][ComplexDeinterleaving] Update splat identification to include vector ConstantInt/FP. (#144516)
d3441f734820 [LV] Change getSmallBestKnownTC to return an ElementCount (NFC) (#141793)
acde20b5605f [HLSL][SPIRV] Add vk::constant_id attribute. (#143544)
561eca44e763 [PowerPC] Split tests into asm and mir parts (NFC)
ee4c2bb68752 [lldb][test] explicit-member-function-quals.cpp: add -glldb
a13b7cc00c5f [LICM] Support hoisting of non-argmemonly readonly calls (#144497)
5a9cc93a2058 Fix for bazel build #142079 (#144665)
dac0820b2778 [Thumb2] Regenerate some test checks. NFC
cd8248f3e856 Fixed a bug in `-fsanitize-kcfi-arity` (#142867)
bb00fd087a3c [gn build] Port 669627d0c77e
757a0e6d3b61 [SystemZ] Treat FAKE_USE instructions as instructions without a size (#144390)
6fcdde2a4eb9 [runtimes] Allow use of external llvm-lit on standalone builds (#144347)
58c4fa96cb11 Fix bazel build for #142771 (#144659)
7f08d969ab05 merge main into amd-staging (#2697)
fe42d34274ca [clang][headers]Remove unnecessary guard of !defined(__SCE__). (#144522)
8e157fdbb7b4 [CIR] Add support for __builtin_assume (#144376)
355725a25e6b [TSan] Fix missing inst cleanup (#144067)
43e1a5a411d9 [mlir][mesh] adding option for traversal order in sharding propagation (#144079)
669627d0c77e Add check 'cppcoreguidelines-use-enum-class' (#138282)
c16dc63b44ae [OMPIRBuilder] Fix gcc -Wparentheses warning [NFC]
7ea7ccd24d60 [PowerPC][AIX] Specify pointer info and alignment for stack store (#144526)
255b55c602f7 [GlobalOpt] Use cast instead of dyn_cast. NFC (#144634)
59d6fbb8ffe0 [flang][fir] Provide allocation block for `fir.local` when required (#144521)
ca29c632f06f [RISCV] Support non-power-of-2 types when expanding memcmp
ba40a7bc2e65 [LoopVectorize] Vectorize fixed-order recurrence with vscale x 1. (#142772)
49df87e71b73 [libc][printf] Fix out-of-range shift in float320 printf (#144542)
a38932ac3c0a Revert "[GlobalISel] prevent G_UNMERGE_VALUES for vectors with different elements" (#144650)
e07b1b26c38b [DAG] Implement SDPatternMatch `m_Abs()` matcher (#144512)
45ea46c44636 Reland [Driver] Add support for GCC installation detection in Baremetal toolchain (#144640)
44b715293fca [PhaseOrdering][X86] Copy FMUL+ADDSUB/FMADDSUB build vector patterns from codegen tests
0875bee2b101 [X86] combineAndNotIntoANDNP - pull out repeated SDLoc(). NFC.
dac94f28e696 [X86] combineAndNotOrIntoAndNotAnd - pull out repeated SDLoc(). NFC.
896e187a6e92 [X86] combineAndMaskToShift - pull out repeated SDLoc(). NFC.
4d71f20b287e [GlobalISel] prevent G_UNMERGE_VALUES for vectors with different elements (#133335)
10f29a607205 [MSan] Fix wrong unpoison size in SignalAction (#144071)
74206d4cb6d9 merge main into amd-staging
74687180dde0 [mlir][emitc] Make CExpression trait into interface (#142771)
323d03ddebc5 merge main into amd-staging (#2694)
ad9e591fd53f [SelectionDAG][RISCV] Fold (add (vscale * C0), (vscale * C1)) to (vscale * (C0 + C1)) in getNode. (#144565)
7b9d10d2e641 PowerPC: Fix using long double libm functions for f128 intrinsics (#144382)
af49a650e172 PowerPC: Add baseline tests for more f128 libcall handling (#144381)
71909c76afcb merge main into amd-staging (#2692)
e14f327d8094 [RISCV] Pre-test for #144461
a2ad65661ad5 [RISCV] Add patterns for generating QC_CTO and QC_CLO (#144532)
0cf587cf1cef merge main into amd-staging
a96a3f1b26ba [lldb][Minidump Parser] Implement a range data vector for minidump memory ranges (#136040)
8ddada41df04 [RISCV] Add Andes XAndesVBFHCvt (Andes Vector BFLOAT16 Conversion) extension (#144320)
9265b1f0cff7 LowerTypeTests: Use jump table entry type as value type of jump table alias.
9767ba624c65 Revert "[DebugInfo][RemoveDIs] Remove a swathe of debug-intrinsic code (#144389)"
719fd290a5c8 Merge commit '9eb0020555fc' into amd/merge/upstream_merge_20250617200249
0defde8e0633 [AMDGPU] Support D16 folding for image.sample with multiple extractelement and fptrunc users (#141758)
86a09f36154f [MLIR][XeGPU] Clean up xegpu op tests (#144592)
4e090b6e84e3 [lldb] Re-insert code to search for a binary by filepath if provided
133b818543da Revert "[CUDA][HIP] Add a __device__ version of std::__glibcxx_assert_fail() (#136133)"
99e263228f45 github: Add mips backend to PR autolabeler (#140909)
abbdd1670d8b [llvm] minor fixes for clang-cl Windows DLL build (#144386)
64155a32297f Added clarifying comment to 'LLVMLinkInMCJIT' and 'LLVMLinkInInterpreter' (#92467)
15482c83aa2b [ElimAvailExtern] Add an option to allow to convert global variables in a specified address space to local (#144287)
c21a4c6c43bb [Xtensa] Implement Xtensa Interrupt/Exception/Debug Options. (#143820)
f2d2c99866df [clang] Remove separate evaluation step for static class member init. (#142713)
b164d3613ad9 [gn build] Port 628274dadf92
6652961ae5fe [gn build] Manually port 556e69b7
535291409cc7 [gn build] Port 9ec75a50bc48
a871b919ed13 [gn build] Port 9e0186d925f0
628274dadf92 [NFC] Extract Printing portions of DWARFCFIProgram to new files (#143762)
a9811340b75b AMDGPU: Report special input intrinsics as free (#141948)
f3af1cd08cd4 [RISCV] Set the exact flag on the SRL created for converting vscale to a read of vlenb. (#144571)
f08474ab1fa9 AMDGPU: Add baseline cost model tests for special argument intrinsics (#141947)
54015f36c682 AMDGPU: Cost model for minimumnum/maximumnum (#141946)
70343c8d4427 [mlir][flang] Added Weighted[Region]BranchOpInterface's. (#142079)
af65cb68f553 AMDGPU: Move fpenvIEEEMode into TTI (#141945)
bec9ac2dafe1 [llvm] Lower latency bonus threshold in function specialization. (#143954)
3800a83160a4 AMDGPU: Reduce cost of f64 copysign (#141944)
c9b281638885 AMDGPU: Fix cost model for 16-bit operations on gfx8 (#141943)
cb63b75e32a4 Revert "[lldb-dap] Refactoring DebugCommunication to improve test consistency. (#143818)
87b13ada1096 [HLSL][RootSignature] Implement serialization of remaining Root Elements (#143198)
1cd18bc894b9 AMDGPU: Add cost model tests for minimumnum/maximumnum (#141904)
fd7e46b86422 Revert "[libc++] Remove trailing newline from _LIBCPP_ASSERTION_HANDLER calls" (#144615)
f25f2f7de4f8 [MLIR][XeGPU] Extend unrolling support for scatter ops with chunk_size (#144447)
3f33c8482fc0 [clang] Add release note for int->enum conversion change. (#144407)
f1b8b0709428 merge main into amd-staging
362b9d78b4ee [lldb-dap] Refactoring DebugCommunication to improve test consistency. (#143818)
f2bef5679c48 Revert "[flang] Restructure runtime to avoid recursion (relanding) (#143993)"
2e1e941da692 Merge commit '2bf3ccabfa37' into amd/merge/upstream_merge_20250617150616
6fb36db4818a [LinkerWrapper] Fix 'save-temps' when targeting SPIR-V (#144605)
844e41c2aced [libc] Moved shared constexpr to the top (#144569)
a5a0d880736f [libc++] Remove trailing newline from _LIBCPP_ASSERTION_HANDLER calls (#143573)
ecfb8fe5c187 Revert stack "[Driver] Add support for GCC installation detection in … (#144603)
3c7df98c7b2a [clang-tidy] Add missing colon in the docs of performance-enum-size (#144525)
8d1610afd0db [libc++] Mark two assertion tests as unsupported in C++03 mode
ef0ce24a920f [OMPIRBuilder] - Make offloading input data persist for deferred target tasks (#133499)
49bf8d38d80c [gn build] Manually port b4e39e4f
908f74a25e01 [llvm] re-order LLVM_ABI and extern on NoKernelInfoEndLTO decl (#144601)
c677a11c8d32 [lldb] Add support to list/enable/disable remaining plugin types. (#143970)
9ae4d2e01331 [LLVM] [Support] Disable `ioctl()` terminal size check on Solaris (#144600)
7c4b2be983e9 [libc++][NFC] Refactor basic_streambuf to use public API functions when possible (#144547)
a79186c1ea62 [Driver] Fix Arm/AArch64 Link Argument tests (#144582)
73f307a5ca30 [HLSL] Use ExtVector for firstbit intrinsics (#142679)
8dcf4ba63595 [AMDGPU] Fix getAsmVOP3Base call agruments. (#144572)
7b7b5a397da1 [AMDGPU] Remove AsmVOP3OpSel field completely. NFCI. (#144574)
19658d147498 [llvm] annotate interfaces in llvm/Target for DLL export (#143615)
b1aa845595c4 [libc++][NFC] Consistently qualify calls to C functions in <fstream> (#144539)
667c7860ef5c [CIR] Handle global string literals as char array initializer (#144384)
57828fec760f Revert "[clang][bytecode] Allocate IntegralAP and Floating types using an allocator (#144246)"
7046f5546ca9 merge main into amd-staging (#2663)
9cb754509608 [Driver] Add option to force undefined symbols during linking in BareMetal toolchain object. (#132807)
dd65e6e0608c [Offload][libc] Add cmake cache AMDGPU buildbot (#144500)
b59d4cf05447 [Reland] Adjust bit cast instruction filter for DXIL Prepare pass (#143783)
1f10c6a277fb [Matrix] Hoist more IRBuilder<>'s. NFC
391dafd8af9c [RISCV] Consolidate both copies of getLMUL1VT [nfc] (#144568)
80f3a28bbe7c Revert "lsan: Support free_sized and free_aligned_sized from C23" (#144575)
8cd05b88ec62 [NFC][HLSL] Move Sema work from `ParseMicrosoftRootSignatureAttributeArgs` (#143184)
bb288de4e0e7 [LoopPeel] Support last iteration peeling of min/max intrinsics (#143598)
2ab9c35ea93f [DXContainer] Update DXContainer to match D3D12 spec (#143201)
31523de4b000 [Driver] Fix link order of BareMetal toolchain object (#132806)
e29bb9a03824 [IR2Vec] Consider only reachable BBs and non-debug instructions (#143476)
0cfc59ff5172 [mlir][bazel] remove extra empty space for XeGPUTransforms
b876b3fa98cf Add missing intrinsics to cuda headers (#143664)
030b5519ec13 [mlir][bazel] add missing deps for XeGPUTransforms
4943e746909d fixup! [Remarks] Elaborate on called intrinsics (#143985)
7ea710fafa57 Fix/reapply "[libc] Migrate stdio tests to ErrnoCheckingTest. (#144134)
3a06e9a710b7 Conditionalise the addition of Aarch64 function Multi versioning support on aarch64 target (#143749)
ed07b54b38c6 [CIR][NFCI] Represent Complex RValues As Single Value (#144519)
9e0186d925f0 [HLSL][RootSignature] Implement `ResourceRange` as an `IntervalMap` (#140957)
9dd1c66e8ffb [VPlan] Expand VPWidenIntOrFpInductionRecipe into separate recipes (#118638)
8513066f2c49 [clangd] Implement LSP 3.17 positionEncoding (#142903)
ec230aa7a7d1 [Driver] Add support for crtbegin.o, crtend.o and libgloss lib to BareMetal toolchain object (#121830)
526310e916af [Remarks] Elaborate on called intrinsics (#143985)
01a7a21a4b80 [CMake] Add BINARY_DIR argument for add_lit_testsuites (#144431)
8063bd153c6a [MLIR][XeGPU] Add support for elementwise ops in Wg to Sg distribute pass [1/N] (#142797)
556e69b7f432 [lldb] make lit use the same Python executable for building and testing (#143756)
5baf351ba819 [BPF] Do not allow gotol in the middle of asm insn (#144545)
b14e03d85550 [LLDB] Consolidate C++ string buffer summaries (#144258)
0c608175c11c [X86] matchLogicBlend - convert to SDPatternMatch matching. NFC. (#144546)
4e884dd993e0 SPARC: Remove SparcMCExpr.h
382e3fdbb476 […
umangyadav
requested review from
dhernandez0,
stefankoncarevic,
djramic,
Copilot and
mirza-halilcevic
There was a problem hiding this comment.
Pull Request Overview
This PR merges an upstream ROCm/LLVM sync and incorporates required adaptations in clang and CIR, including:
- Introduce the Managarm toolchain (new files and integrates in Driver and CMake).
- Rename kernel calling convention APIs from
getOpenCLKernelCallingConvtogetDeviceKernelCallingConv. - Enhance CIRRecordLayoutBuilder to implement detailed bitfield layout and add vector splat simplification in CIRSimplify.
Reviewed Changes
Copilot reviewed 294 out of 3143 changed files in this pull request and generated no comments.
Show a summary per file
| File | Description |
|---|---|
| clang/lib/Driver/ToolChains/Managarm.{h,cpp} | Add Managarm target support |
| clang/lib/Driver/ToolChains/Gnu.cpp | Handle Managarm triple in LDM option |
| clang/lib/CodeGen/TargetInfo.{h,cpp} | Rename OpenCL kernel conv to device kernel conv |
| clang/lib/CIR/CodeGen/CIRGenRecordLayoutBuilder.cpp | Full rewrite of bitfield layout algorithm |
| clang/lib/CIR/Dialect/Transforms/CIRSimplify.cpp | Add SimplifyVecSplat pattern |
Comments suppressed due to low confidence (3)
external/llvm-project/clang/lib/Driver/ToolChains/Gnu.cpp:230
- The linker target string "aarch64managarm" lacks a hyphen; change it to "aarch64-managarm" to be consistent with other multiarch triples.
if (T.isOSManagarm())
external/llvm-project/clang/lib/Driver/ToolChains/Managarm.cpp:1
- Add integration tests for the new Managarm toolchain to verify invocation, sysroot detection, and include path resolution.
//===----------------------------------------------------------------------===//
external/llvm-project/clang/lib/CIR/Dialect/Transforms/CIRSimplify.cpp:263
- Introduce unit tests for the
SimplifyVecSplatpattern to ensure constant splats are correctly folded intoConstantOp.
struct SimplifyVecSplat : public OpRewritePattern<VecSplatOp> {
dhernandez0
reviewed
Jun 25, 2025
dhernandez0
reviewed
Jun 25, 2025
dhernandez0
reviewed
Jun 25, 2025
dhernandez0
reviewed
Jun 25, 2025
|
|
||
| options = Options(debug=args.debug, quiet=args.quiet, | ||
| arch=arch, flags=rocmlir_gen_flags, concurrent_tests=args.jobs) | ||
| arch=arch, flags=rocmlir_gen_flags, concurrent_tests=args.jobs, numCU=getNumCU(arch)) |
There was a problem hiding this comment.
same here, this seems unrelated to the upstream merge
dhernandez0
reviewed
Jun 25, 2025
|
|
||
| options = Options(debug=args.debug, quiet=args.quiet, | ||
| arch=arch, flags=rocmlir_gen_flags, concurrent_tests=args.jobs, numCU=getNumCU(arch)) | ||
| arch=arch, flags=rocmlir_gen_flags, concurrent_tests=args.jobs, numCU=getNumCU(getChip())) |
There was a problem hiding this comment.
is this related to the upstream merge?
There was a problem hiding this comment.
no,
ParameterSweep script is failing, cherry picked changes from this PR to make it run
#1898
There was a problem hiding this comment.
sounds good, can we revert the changes once CI passes? and have separate PRs if needed. So, it's easier to keep track of changes (if they are part of a big upstream merge, it's not so easy).
There was a problem hiding this comment.
I'll revert this once parameterSweeps passes
dorde-antic
requested changes
Jun 25, 2025
|
@dorde-antic I enabled Navi4x in PR CI so that i do not have to manual testing for it for upstream merge. But now that i think about it i do not need to keep it since i am running nightly as well which will run on Navi4x. |
dhernandez0
approved these changes
Jun 25, 2025
dorde-antic
approved these changes
Jun 25, 2025
|
CI Problems :
|
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.
Moves copy of LLVM to this commit :
ROCm/llvm-project@e8215ed
Required some changes in rocMLIR due to following commits :
llvm/llvm-project#144636
llvm/llvm-project#144658
llvm/llvm-project#142986
Files that are different between
rocm/llvm-projectandexternal/llvm-projecthere is the changelist (Recommend viewing in notepad++ with "diff" as language)
llvm_diffs.txt
mlir_diffs.txt
Testing :
MI300
ninja check-mlir check-rocmlirpasses.MI355 tests pending