-
Notifications
You must be signed in to change notification settings - Fork 791
LLVM and SPIRV-LLVM-Translator pulldown (WW04 2025) #16781
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Merged
Merged
Changes from 1 commit
Commits
Show all changes
2204 commits
Select commit
Hold shift + click to select a range
d7fb4a2
[LLVM][Clang][AArch64] Implement AArch64 build attributes (#118771)
sivan-shani e30b703
[mlir:python] Construct PyOperation objects in-place on the Python he…
hawkinsp d80eb92
[llvm] Remove `undef` from `llvm/test/Transforms` tests (#123889)
Yeaseen ccd7795
[SLP][NFC]Add a test with potential alternate node, marked for minbit…
alexey-bataev 4f40b07
[X86][AVX10.2-SATCVT][NFC] Remove NE from intrinsic and instruction n…
phoebewang 1df50e6
[PS5][Driver] Don't link with --build-id when -static (#123930)
playstation-edd a2063ba
[clangd][NFC] Delete a pessimizing move
kadircet a7a8694
Remove references to mips within Android (#123856)
hiraditya 70a16b9
[HIP] Support managed variables using the new driver (#123437)
jhuber6 4170d61
[gn] port 6aeffcdb9130
nico d0a89e7
[gn] fix mistake in d7fb4a275c98f
nico c938436
Revert "[GISel] Add more FP opcodes to CSE (#123624)" (#123954)
dklimkin aa580c2
[clang-tidy] Add `EnableQtSupport` option to modernize-use-integer-si…
qt-tatiana 93d35ad
AMDGPU: Delete FillMFMAShadowMutation (#123861)
arsenm 6b486f4
[Clang][Arch] Disable mve.fp when explicit -mfpu option (#123028)
flopop01 b0d35cf
[SSAUpdater] Avoid scanning basic blocks to find instruction order. (…
jcranmer-intel 58c6d44
[CMake] Remove HAVE_DLFCN_H and HAVE_DLADDR (#123879)
MaskRay 5a9b74d
[mlir] Fix a warning
kazutakahirata 75ce2dc
[llvm][AArch64] apple-m4 does not have FEAT_{SPEv1p2,SEL2,MPAM} (#123…
jroelofs fa6f88a
[MLIR][XeGPU] Allow some nd ops to have argument shapes mismatch for …
kurapov-peter 13d09df
[X86] Simplify ArrayRef construction. NFC (#123899)
topperc 9e6494c
[CodeGen] Rename RegisterMaskPair to VRegMaskOrUnit. NFC (#123799)
topperc f63e8ed
Revert "[Modules] Delay deserialization of preferred_name attribute a…
ilya-biryukov 038544b
Merge from 'main' to 'sycl-web' (180 commits)
frasercrmck 16298e4
[X86] var-permute-256.ll - regenerate VPTERNLOG comments
RKSimon 603529b
[X86] add/sub signed sat vec tests - regenerate VPTERNLOG comments
RKSimon e6c7d6a
[X86] avx512-broadcast-unfold.ll - regenerate VPTERNLOG comments
RKSimon bb754f2
[X86] avx512 intrinsics tests - regenerate VPTERNLOG comments
RKSimon a25f2cb
[X86] vector rotate tests - regenerate VPTERNLOG comments
RKSimon 44f3168
[X86] vector reduction tests - regenerate VPTERNLOG comments
RKSimon 2476417
Reapply "[sanitizer][NFCI] Add Options parameter to LowerAllowCheckPa…
thurstond f3c3a9b
[libc][cmake] error if user disables sanitizers but wants scudo (#123…
nickdesaulniers ddb8607
[libc][docs] disable pthreads docs (#123824)
nickdesaulniers 8e79ade
[libc][LIBC_ADD_NULL_CHECKS] replace volatile deref with __builtin_tr…
nickdesaulniers f78359c
[mlir][spirv] Add definition for OpEmitVertex and OpEndPrimitive (#12…
IgWod-IMG 511dc26
[lldb][DWARFASTParserClang][NFCI] Factor out CV-qualifier/is_static p…
Michael137 b1943f4
[BranchFolding] Remove getBranchDebugLoc() (#114613)
ellishg a56ba1f
[ValueTracking] Handle recursive select/PHI in ComputeKnownBits (#114…
goldsteinn afcbcae
[mlir][OpenMP] inscan reduction modifier and scan op mlir support (#…
anchuraj 9f83c4e
[flang][cuda] Allocate descriptor in managed memory on rebox block ar…
clementval 053795b
Merge from 'sycl' to 'sycl-web' (23 commits)
mdfazlay 7bf188f
[NFC] Minor fix to tryEmitAbstract type in EmitCXXNewAllocSize (#123433)
andykaylor b40739a
Revert "[LLVM][Clang][AArch64] Implement AArch64 build attributes (#1…
kazutakahirata 195a1fc
Reapply "[Clang][Sema] Use the correct lookup context when building o…
sdkrystian 137d706
[SCEV] Do not attempt to collect loop guards for loops without predec…
juliannagele 146ee98
[RISCV] Remove duplicate WriteRes<WriteJalr for MIPSP8700. (#123865)
topperc ec15b24
[llvm][Support] Only enable backtrace test when it's enabled (#123852)
jroelofs 5895932
[RISCV][VLOpt] Reorganize visit order and worklist management (#123973)
preames 340706f
[PowerPC] Fix saving of Link Register when using ROP Protect (#123101)
stefanp-synopsys 1687aa2
[RISCV][VLOPT] Don't reduce the VL is the same as CommonVL (#123878)
michaelmaitland 5ede7b6
Revert "Reapply "[Clang][Sema] Use the correct lookup context when bu…
sdkrystian 7ad8a3d
[MemProf] Simplify edge iterations (NFC) (#123469)
teresajohnson df5f3c3
Merge from 'main' to 'sycl-web' (135 commits)
mdfazlay 3057d0f
Android defaults to pic (#123955)
hiraditya 22a4f6e
Merge from 'sycl' to 'sycl-web' (1 commits)
7e622b6
[TOSA] Change PadOp padding to tosa.shape (#123133)
Jerry-Ge 719f0d9
[HLSL] Fix global resource initialization (#123394)
hekota 1cf0af3
[AMDGPU][True16][MC] true16 for v_cmpx_class_f16 (#123251)
broxigarchen 18e9d3d
[AMDGPU][True16][MC] true16 for v_cmpx_xx_u/i16 (#123424)
broxigarchen 213e03c
[Clang] Fix handling of immediate escalation for inherited constructo…
cor3ntin c6e7b4a
[flang][cuda][NFC] Add kernel name in translation error (#123987)
clementval 2656928
Reland "[HLSL] Implement the `reflect` HLSL function" (#123853)
Icohedron 8fb4230
[SYCL] AST support for SYCL kernel entry point functions. (#122379)
tahonermann 517334b
[TableGen] Add maps from Write/ReadType to the parent WriteRes/ReadAd…
topperc dee80e9
Merge from 'main' to 'sycl-web' (176 commits)
mdfazlay ac94fad
[SPIR-V] Rename internal command line flags for optimization level an…
VyacheslavLevytskyy 68c6b2e
[ASTMatchers][NFC] use `Matcher<QualType>` instead of `DynTypedMatche…
HerrCai0907 a2c683b
[LLD][COFF] Use EC symbol table for exports defined in module definit…
cjacek 4e9d5a3
[LLD][COFF] Add support for the -defArm64Native argument (#123850)
cjacek a77250f
[mlir] Add C and Python interface for file range (#123276)
jpienaar 223bd0c
[libc++] Avoid unnecessary instantiations for __copy_cvref_t (#123718)
philnik777 27ccc99
[RISCV][VLOpt] Minor worklist invariant cleanup [NFC] (#123989)
preames e0ae889
[SandboxIR][Doc] Add Quick start notes (#123992)
630177c
[compiler-rt][rtsan] Fix madvise/posix_madvise for macOs. (#124020)
devnexen 939f290
[compiler-rt][rtsan] getsockopt/setsockopt interception. (#124004)
devnexen fd08713
[SandboxVec][Legality] Diamond reuse multi input (#123426)
6436089
[BoundsSafety][Doc] Add BoundsSafetyAdoptionGuide.rst (#120674)
rapidsna a939a9f
[LLDB-DAP] Send Progress update message over DAP (#123837)
Jlalond 6e498bc
[flang][cuda] Handle simple device pointer allocation (#123996)
clementval 98de5df
[mlir] Add NamedAttribute ctor taking StringRef. NFC. (#123974)
kuhar 2dc1c95
[SandboxVec][VecUtils] Implement VecUtils::getLowest() (#124024)
96dbd00
[RISCV] Re-generate test checks so we pick up implicit on whole regi…
topperc 9fbf5cf
[clang][modules] Partially revert 48d0eb518 to fix -gmodules output (…
benlangmuir 8f45452
workflows/release-binaries: Restrict jobs based on owner instead of r…
tstellar cb35278
Merge from 'sycl' to 'sycl-web' (1 commits)
8110af7
[SandboxVec][BottomUpVec] Fix codegen when packing constants. (#124033)
d5457e4
[NFC][Index] Disable LSAN on crash recovery tests (#124035)
vitalybuka eaaac05
[Sema] Fix a warning
kazutakahirata a9d2834
[llvm][CodeGen] Fix the issue caused by live interval checking in win…
huaatian 847acbb
[ExceptionDemo] Transition example from MCJIT to ORC and fix errors (…
epitavy 23d2a18
PeepholeOpt: Remove unnecessary check for null TargetInstrInfo (#123929)
arsenm d3aea77
[SLP] Move transformMaskAfterShuffle into BaseShuffleAnalysis and use…
HanKuanChen 1042ddc
[gn] port ec15b242505
nico 6f69ade
PeepholeOpt: Remove null TargetRegisterInfo check (#123933)
arsenm 2646e2d
PeepholeOpt: Stop allocating tiny helper classes (NFC) (#123936)
arsenm ba70368
[Clang][Driver] Support linker relaxation options for LoongArch (#123…
zhaoqi5 15c2d4b
PeepholeOpt: Remove check for subreg index on a def operand (#123943)
arsenm 9cefa3e
[msan] Generalize handleIntrinsicByApplyingToShadow by adding bitcast…
thurstond ba3e6f0
[RISCV][VLOPT] Remove dead passthru check in getOperandLog2EEW. NFC (…
lukel97 0fe8e70
Revert "Reland "[HLSL] Implement the `reflect` HLSL function"" (#124046)
inbelic 3ef90f8
[emitc] Fix the translation switchop with argument of expressionop (#…
jacquesguan 5e174f8
Merge from 'main' to 'sycl-web' (36 commits)
Chenyang-L b46fcb9
[Clang] Implement CWG 2628 "Implicit deduction guides should propagat…
zyn0217 652ff20
[flang][cuda] Adding atomicadd as a cudadevice intrinsic and converti…
Renaud-K 892a804
[NVPTX] Stop using 16-bit CAS instructions from PTX (#120220)
akshayrdeodhar 1c5d971
[Signals] Exclude dladdr for AIX after #123879
MaskRay 7575072
[llvm-exegesis] Disable exhaustive tests on Windows
boomanaiden154 19834b4
[LoongArch] Support sc.q instruction for 128bit cmpxchg operation (#1…
tangaac 0bcf34e
[Clang] [NFC] Mark `UnresolvedSetImpl`'s move operations as defaulted…
MagentaTreehouse d80b814
[LoongArch] Summary llvm20 release notes
SixWeining 3c7a878
[LoongArch] Summary clang20 release notes
SixWeining aa273fd
[LoongArch] Update lld20 release notes
SixWeining 163935a
[RISCV] Add Qualcomm uC Xqcilo (Large Offset Load Store) extension (#…
hchandel de209fa
[CodeGen] Introduce Static Data Splitter pass (#122183)
mingmingl-llvm d15f3e8
[LoopInterchange] Constrain LI within supported loop nest depth (#118…
madhur13490 646f034
[gn build] Port de209fa11b54
llvmgnsyncbot daa1820
[Flang][OpenMP] Fix copyin allocatable lowering to MLIR (#122097)
kaviya2510 ea49d47
Specify triple for llc test
mingmingl-llvm 5d8390d
Temporarily disable test on Fuchsia
mingmingl-llvm c3dfd34
[WebAssembly] Add unreachable before catch destinations (#123915)
aheejin ba17485
[clang][CodeComplete] Use HeuristicResolver to resolve DependentNameT…
HighCommander4 220004d
[GISel] Add more FP opcodes to CSE (#123949)
lialan 3fb8c5b
[X86] Fix invalid instructions on x32 with large stack frames (#124041)
mconst 8eb99bb
Reland [LLD] [COFF] Fix linking MSVC generated implib header objects …
mstorsjo cd5694e
[StrTable] Switch the option parser to `llvm::StringTable` (#123308)
chandlerc 2b67ece
Android no longer supports arm < 7 (#123952)
hiraditya 2a51a0d
Remove reference to android-mips (#124021)
hiraditya 091741a
[libfuzzer] Clarify -max_len behavior on bigger files (#123095)
chfast 70d7c84
[mlir][docs] Add usage/example of OpAsmOpInterface (#123610)
ZenithalHourlyRate 4b0df28
[clang][Tooling] Prefer <atomic> for atomic_* family in C++
kadircet 7781381
[SDAG] Use BatchAAResults for querying alias analysis (AA) results (#…
MacDue d7c14c8
[IR] Replace of PointerType::getUnqual(Type) with opaque version (NFC…
junlarsen 9fd9263
[libc++abi] Remove support for Android 4 and older (#124054)
brad0 6bc68d0
[libc++] Remove support for Android 4 and older (#124062)
brad0 0429bfe
[libc++] Remove a few unused includes (#124025)
philnik777 ee99c4d
[LLVM][Clang][AArch64] Implement AArch64 build attributes (#123990)
sivan-shani 7fb97be
[ConstraintElimination] Add eq/ne facts to signed constraint system (…
zsrkmyn 08195f3
[libc++] Inline basic_streambuf functions (#123379)
philnik777 8388040
[mlir][tosa] Add NaN Propagation Mode Support (#121951)
FranklandJack 1930635
[clang][Modules] Raise empty.modulemap expected size to <70KB to fix …
asb cad6bba
[C++20][Modules] Fix crash/compiler error due broken AST links (#123648)
dmpolukhin 2e6cc79
[MLIR][NVVM] Migrate CpAsyncOp to intrinsics (#123789)
durga4github cb714e7
[DebugInfo][InstrRef] Avoid producing broken DW_OP_deref_sizes (#123967)
jmorse ad6d808
[lldb][DWARFASTParserClang] Make C++ method parsing aware of explicit…
Michael137 fa7f0e5
[NVPTX] Add Bulk Copy Prefetch Intrinsics (#123226)
abhilash1910 17756aa
[Clang] [Release Notes] Implicit lifetimes are a C++23 feature
cor3ntin a802093
Revert "[lldb][DWARFASTParserClang] Make C++ method parsing aware of …
Michael137 05fbc38
[VPlan] Move VPBlockUtils to VPlanUtils.h (NFC)
fhahn 4bcdb26
Revert "[lldb][test] Remove compiler version check and use regex" (#1…
Michael137 4f26edd
[NFC][YAML] Add `IO::error()` (#123475)
vitalybuka 1311b36
[llvm][Support] Put back filename into FileToRemoveList (#124065)
vitalybuka 636bc72
Reland "[lldb][DWARFASTParserClang] Make C++ method parsing aware of …
Michael137 3ea2b54
[lldb/windows] Make "anonymous" pipe names more unique (#123905)
labath 0236cb6
[lldb] Enable "frame diagnose" on linux (#123217)
labath 6f68481
[libc++] Use [[clang::no_specializations]] to diagnose invalid user s…
philnik777 9705500
[libclc] Move nextafter to the CLC library (#124097)
frasercrmck e069518
SCEV: cover a codepath in isImpliedCondBalancedTypes (#123070)
artagnon 0e944a3
[SCFToGPU] Convert scf.parallel+scf.reduce to gpu.all_reduce (#122782)
tkarna 90e9895
[X86] Handle BSF/BSR "zero-input pass through" behaviour (#123623)
RKSimon 0c66644
[bazel]Fix bazel build past 2e6cc79f816d942ab09d6a310cd925c1da148aa9 …
dklimkin 590e5e2
[M68k] Fix llc pass test after 3630d9ef65b30af7e4ca78e668649bbc48b5be66
darkbuck d3d605b
[FileCheck] Use move semantics instead of std::swap. NFC. (#123304)
jayfoad fb3fa41
MachineRegisterInfo: Use variable for TRI
arsenm 6fdaaaf
[AMDGPU] SIPeepholeSDWA: Disable on existing SDWA instructions (#123942)
frederik-h d8eb4ac
[Support] Remove ciso646 include (#123578)
nikic ff55c9b
[llvm][amdgpu] Handle indirect refs to LDS GVs during LDS lowering (#…
ergawy 92b839e
[gn] fix mistake in ee99c4d4845db
nico e28e935
AMDGPU: Make vector_shuffle legal for v2i32 with v_pk_mov_b32 (#123684)
arsenm 0c71fdd
[NVPTX] Fix ctor / dtor lowering when NVPTX target is not enabled (#1…
jhuber6 99d450e
Revert "[AMDGPU] SIPeepholeSDWA: Disable on existing SDWA instruction…
nico 4d3a530
[bazel]Fix(2) bazel build past 2e6cc79f816d942ab09d6a310cd925c1da148a…
dklimkin 25653e5
[AVX10.2] Update convert chapter intrinsic and mnemonics names (#123656)
mikolaj-pirog 1f0964f
[llvm][Docs] Clarify finding maintainers
DavidSpickett 26b61e1
[LoopVectorizer] Propagate underlying instruction to the cloned insta…
NickGuy-Arm 6206f54
[AMDGPU] Occupancy w.r.t. workgroup size range is also a range (#123748)
lucas-rami c3b40c7
[X86] Regenerate test checks (NFC)
nikic f61d93f
[Flang] Generate math.acos op for non-precise acos intrinsic calls (#…
jsjodin 6fe0fc6
[CallingConv] Return ArrayRef from AllocateRegBlock() (NFC) (#124120)
nikic 9498fc2
Merge from 'sycl' to 'sycl-web' (6 commits)
e1aa1e4
[WPD]Provide branch weight for checking mode. (#124084)
mingmingl-llvm 96410ed
mark test as unsupported as I investigate test failure on certain env…
mingmingl-llvm cb426b1
[NFC][DirectX] Clean-up of `DXIL.td` (#124005)
inbelic 4bd0440
[RISCV] Add RVVConstraint to SiFive custom matrix multiply instructio…
topperc 1937a36
[RISCV] Add @earlyclobber to SiFive custom matrix multiply instructio…
topperc df29995
[mlir] Fix warnings
kazutakahirata bca6dbd
[X86] Add additional i128 abi test (NFC)
nikic 7db4ba3
[GlobalMerge][NFC] Fix inaccurate comments (#124136)
michaelmaitland fa29929
[SLP][NFC]Modernize code base in several places
alexey-bataev e622468
[AST] Migrate away from PointerUnion::dyn_cast (NFC) (#124074)
kazutakahirata 113e1fd
[CodeGen] Migrate away from PointerUnion::dyn_cast (NFC) (#124076)
kazutakahirata d050083
[lldb] Avoid repeated map lookups (NFC) (#124077)
kazutakahirata bb019dd
[CodeGen] Avoid repeated hash lookups (NFC) (#124078)
kazutakahirata bda39a6
[ExecutionEngine] Include <map> (#124083)
kazutakahirata 7ddeea3
[LLVM][AMDGPU] MC support for ds_bpermute_fi_b32 (#124108)
Acim-Maravic d8cd8d5
[SLP] getSpillCost - fully populate IntrinsicCostAttributes to improv…
RKSimon 2f76e2b
[Driver] -fno-plt: warn for unsupported targets
MaskRay 082b148
[TableGen] Pass CodeGenProcModel reference instead of index to addWri…
topperc 1a8f49f
[mlir][python][cmake] Allow skipping nanobind compile options changes…
ScottTodd 8c138be
[flang][cuda] Handle pointer allocation with source (#124070)
clementval caf0540
[LoopVectorizer] Add support for chaining partial reductions (#120272)
NickGuy-Arm a245309
[llvm-profdata] Add block percent to detailed summary (#105915)
ellishg e00f1f8
[ELF] Error for executable .note.GNU-stack unless -z execstack or -r
MaskRay 6d4e72a
[GVN] Add extra vscale tests with different types. NFC
davemgreen 3d72619
[InstCombine] Add a test for splitting scalable structs. NFC
davemgreen e062224
[test] Remove misleading ''
MaskRay ed51271
[Offload] Make MemoryManager threshold ENV var size_t type. (#124063)
hidekisaito 6a97897
[NFC][modules] Create objects on the stack (#124034)
vitalybuka c7e6ca7
[SLP][NFC]Add dump() method for ScheduleData struct type for better d…
alexey-bataev 66e49e3
[YAML] Don't validate `Fill::Size` after error (#123280)
vitalybuka ff17a41
[lldb] Remove support and workarounds for Android 4 and older (#124047)
brad0 5a7d92f
[NFC] Remove invalid features from test and autogenerate checks. (#12…
labrinea e0cd57d
[compiler-rt] Remove support and workarounds for Android 4 and older …
brad0 6045146
[libc] change return type of pthread_setspecific to int in generated …
alexprabhat99 0290693
[llvm-profdata] XFAIL broken test on windows (#124165)
ellishg db6b7a8
[libc][NFC] Strip all training whitespace and missing newlines (#124163)
jhuber6 2f6b0b4
[RISCV] Add SiFive sf.vqmacc tests to vmv-copy.mir. NFC (#124075)
topperc bec4c7f
[InstCombine] Unpack scalable struct loads/stores. (#123986)
davemgreen 3dec24d
Stats are sorted before they are printed. Try fixing test failure by …
mingmingl-llvm 6d5697f
[SystemZ] Fix ICE with i128->i64 uaddo carry chain
uweigand 0d01908
[TailDup] Allow large number of predecessors/successors without phis.…
fhahn 1c28b92
[Clang] __has_builtin should return false for aux triple builtins (#1…
sarnex a6211a6
[gn build] Port ff17a4136ded
llvmgnsyncbot 4cf1fe2
[lldb] Add missing operations to GetOpcodeDataSize (#120163)
JDevlieghere 775d0f3
[GVN] Handle scalable vectors with the same size in VNCoercion (#123984)
davemgreen c9b7303
Add [[maybe_unused]] to a variable used only in assert in VPlan.h (#1…
basioli-k c3ecbe6
Disable the test again.
mingmingl-llvm 4018317
[Clang] restrict use of attribute names reserved by the C++ standard …
a-tarasyuk 4600de9
Merge from 'main' to 'sycl-web' (90 commits)
frasercrmck 76271f1
Merge from 'sycl' to 'sycl-web' (10 commits)
915f3ed
Merge from 'main' to 'sycl-web' (70 commits)
mdfazlay f260c2e
adds support for SPV_EXT_arithmetic_fence (#2954)
bashbaug cbc1b1b
Update for deprecation of llvm::PointerUnion::get (#2961)
svenvh 789f0e6
Update more llvm::PointerUnion::get calls (#2962)
svenvh 45da762
Align translation of `OpCooperativeMatrixLengthKHR` to match the spec…
vmaksimo aaf026b
Adapt to change introduced by #122928 in upstream. (#2965)
AlexVlx 41f6405
Update dwarf-public-names.ll after LLVM change (#2966)
svenvh 33696aa
Use `poison` instead of `undef` values wherever possible (#2960)
vmaksimo 70c569c
[SLP]Adjust NumberOfParts value for adjusted number of buildvector sc…
alexey-bataev 9f10ac0
[libclc] Fix build after bad merge
jsji 3665b1a
[HIP] Remove XFail from accessor.cpp
jsji 69b0c17
Update spirv-headers-tag.conf despite revert of dd33e595
jsji 6f7939d
[Clang][Driver] Add --no-offload-new-driver to sycl cuda test after 9…
jsji 66cfb5c
[Clang][Driver] Update test after 0d352b2ea767 (#26825)
jsji 4b22e2c
[SYCL][E2E] XFAIL multisource.cpp for now
jsji 248b2d2
[SYCL] XFail sycl-kernel-entry-point*pch/module first
jsji 87c57b3
[SYCL][CUDA] include PTX in new driver
jsji 64daf45
[SYCL][NVPTX] Use ptx_kernel cc instead of metadata
jsji 16914cb
Merge remote-tracking branch 'origin/sycl' into llvmspirv_pulldown
jsji 594e3ce
[SYCL] Remove unecessary cc change
jsji 5a6655a
[NFC] Add comments about early exiting in populateKernels
jsji File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
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
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.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I can't really answer to what's going on here, sorry. I suspect that this indicates we're not passing the right flag to control the new offload driver? The
false
should essentially be equivalent to us explicitly passing-fno-offload-new-driver
to the driver.Perhaps this is okay for now but we need to investigate this properly.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The new offload driver is currently not enabled by default for intel/llvm. The plan is to move to the new model this year.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Makes sense, thanks. But might it be easier to explicitly disable the new offload driver by passing the option, rather than have to change the default values of various
hasFlag
checks?There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, this is the easiest workaround to let cuda sycl use new offload driver for now. Once we switch the default to new offload driver, we should remove this workaround.