[PATCH] D158984: [Clang] Allow __declspec(noalias) to access inaccessible memory

This revision was automatically updated to reflect the committed changes. Closed by commit rG14cc7a077275: [Clang] Allow __declspec(noalias) to access inaccessible memory (authored by nikic). Herald added a project: clang. Herald added a subscriber: cfe-commits. Repository: rG LLVM Github Monor

[PATCH] D74183: [IRGen] Add an alignment attribute to underaligned sret parameters

nikic added a comment. For the record, this change caused an 1.5% compile-time regression on tramp3d-v4 (http://llvm-compile-time-tracker.com/compare.php?from=43a6d285bfead762ac472a6e62beedc9f88bce89&to=de98cf92e301ab559a7417f1eca5cfa53624c9e1&stat=instructions). As there was also a 0.9% increa

[PATCH] D74183: [IRGen] Add an alignment attribute to underaligned sret parameters

nikic added a comment. In D74183#1941741 , @efriedma wrote: > If it's just tramp3d-v4, I'm not that concerned... but that's a weird result. > On x86 in particular, alignment markings have almost no effect on > optimization, generally. I've just looked

[PATCH] D74787: [IRBuilder] Always respect inserter/folder

This revision was automatically updated to reflect the committed changes. Closed by commit rGf12fb2d99b8d: [IRBuilder] Always respect inserter/folder (authored by nikic). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D74787/new/ https://reviews.llvm.

[PATCH] D73835: [IRBuilder] Virtualize IRBuilder

This revision was automatically updated to reflect the committed changes. Closed by commit rG0765d3824d06: [IRBuilder] Virtualize IRBuilder (authored by nikic). Herald added a project: clang. Herald added a subscriber: cfe-commits. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION

[PATCH] D73835: [IRBuilder] Virtualize IRBuilder

nikic reopened this revision. nikic added a comment. This revision is now accepted and ready to land. Reverted because of failures on http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/22164. Not sure why this is the only builder that's failing, but I believe the pr

[PATCH] D73835: [IRBuilder] Virtualize IRBuilder

nikic added a comment. Tried what happens if the copy constructor is deleted: https://gist.github.com/nikic/1e6ab5bbf42cfe48e7b848e60a2df180 It looks like the referenced occurrence in DIBuilder was the only place that was using the copy constructor "for good reason", most of the other uses eith

[PATCH] D73835: [IRBuilder] Virtualize IRBuilder

This revision was automatically updated to reflect the committed changes. Closed by commit rG3eaa53e80543: Reapply "[IRBuilder] Virtualize IRBuilder" (authored by nikic). Changed prior to commit: https://reviews.llvm.org/D73835?vs=244871&id=245004#toc Repository: rG LLVM Github Monorepo CHA

[PATCH] D74787: [IRBuilder] Always respect inserter/folder

nikic created this revision. nikic added reviewers: nhaehnle, Meinersbur, spatel, lebedev.ri. Herald added subscribers: llvm-commits, cfe-commits, hiraditya. Herald added a reviewer: jdoerfert. Herald added projects: clang, LLVM. Some IRBuilder methods that were originally defined on IRBuilderBase

[PATCH] D74787: [IRBuilder] Always respect inserter/folder

nikic marked an inline comment as done. nikic added inline comments. Comment at: llvm/test/Transforms/InstCombine/saturating-add-sub.ll:2 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py -; RUN: opt < %s -instcombine -S | FileCheck %s +; RUN: opt < %s -i

[PATCH] D146003: [StandardInstrumentations] Verify function doesn't change if analyses are preserved

nikic requested changes to this revision. nikic added a comment. This revision now requires changes to proceed. nontrivial-unswitch-markloopasdeleted.ll failure looks legit. Comment at: llvm/lib/Passes/StandardInstrumentations.cpp:1127 +if (!PassPA.allAnalysesInSetPreserved

[PATCH] D146104: Use *{Map,Set}::contains (NFC)

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D146104/new/ https://reviews.llvm.org/D146104 ___ c

[PATCH] D146069: [StandardInstrumentations] Rename -verify-cfg-preserved -> -verify-preserved-analyses

nikic added a comment. To me `verify-preserved-analyses` sounds like "if DominatorTree is preserved, calls `DominatorTree::verify()`, etc". That is, call verify on the preserved analyses. Maybe `verify-change-reporting` or so? Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION

[PATCH] D146003: [StandardInstrumentations] Verify function doesn't change if analyses are preserved

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LG Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D146003/new/ https://reviews.llvm.org/D146003 ___ cfe

[PATCH] D146069: [StandardInstrumentations] Rename -verify-cfg-preserved -> -verify-analysis-invalidation

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LG Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D146069/new/ https://reviews.llvm.org/D146069 ___ cfe

[PATCH] D146200: [PassBuilder] Support O0 in default pipelines

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGa8f6b5763e89: [PassBuilder] Support O0 in default pipelines (authored by nikic). Herald added projects: clang, OpenMP, Flang. Herald added subscriber

[PATCH] D146466: [clang] diagnose function fallthrough

nikic added a comment. @nickdesaulniers I don't think we want to handle that on the LLVM side. That will be fundamentally unreliable. If Clang wishes to make "recoverable" ubsan have arbitrary but well-defined behavior in the case of UB, it needs to generate appropriate IR to model that. For ex

[PATCH] D145265: [Pipeline] Remove GlobalCleanupPM

nikic added a comment. I think this (removing GlobalCleanupPM) makes sense at a high level, but this is a pretty substantial pipeline change and will need more thorough evaluation (e.g. looking at IR diffs on test suite). For example, it is not at all obvious to me that moving InstCombine befor

[PATCH] D143630: [clang] update test case

nikic requested changes to this revision. nikic added a comment. This revision now requires changes to proceed. (Marking this as changes requested per previous comment.) Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D143630/new/ https://reviews.llvm

[PATCH] D144352: Do not emit exception diagnostics from coroutines and coroutine lambdas

This revision was automatically updated to reflect the committed changes. Closed by commit rGf332498f9880: [Clang] Do not emit exception diagnostics from coroutines and coroutine lambdas (authored by denizevrenci, committed by nikic). Changed prior to commit: https://reviews.llvm.org/D144352?vs

[PATCH] D144970: [llvm-c] Remove bindings for creating legacy passes

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGf7ca01333214: [llvm-c] Remove bindings for creating legacy passes (authored by nikic). Herald added a project: clang. Herald added a subscriber: cfe-

[PATCH] D144963: [UTC] Include return type/attributes under --version 2

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGa26d3031cf89: [UTC] Include return type/attributes under --version 2 (authored by nikic). Herald added a project: clang. Herald added a subscriber: c

[PATCH] D145150: clang: Emit nofpclass(nan inf) for -ffinite-math-only

nikic resigned from this revision. nikic added a comment. (Sounds reasonable conceptually, but I'm not familiar enough with clang codegen for the details.) CHANGES SINCE LAST ACTION https://reviews.llvm.org/D145150/new/ https://reviews.llvm.org/D145150 __

[PATCH] D142872: Honor the fwrapv option when generating the inbounds GEP .

nikic requested changes to this revision. nikic added a comment. This revision now requires changes to proceed. The original issue has been fixed by the mentioned LLVM changes. As such, I don't think we need to actually do anything on the Clang side. Please let me know if you're still seeing iss

[PATCH] D145149: [UTC] Enable --function-signature by default

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGfb309041f0c3: [UTC] Enable --function-signature by default (authored by nikic). Herald added a project: clang. Herald added a subscriber: cfe-commits

[PATCH] D127812: [AArch64] FMV support and necessary target features dependencies.

nikic added inline comments. Comment at: compiler-rt/lib/builtins/cpu_model.c:1338 + hwcap = getauxval(AT_HWCAP); + hwcap2 = getauxval(AT_HWCAP2); +#endif // defined(__FreeBSD__) This breaks the build with glibc 2.17. Repository: rG LLVM Github Monorepo CH

[PATCH] D53891: [LTO] Add option to enable LTOUnit splitting, and disable unless needed

nikic added a comment. Herald added a project: All. In D53891#2116541 , @hans wrote: > Would be possible to add some documentation for this flag? From a quick > search it's not clear to me what it does and when one is supposed to use it. Ping on adding d

[PATCH] D146003: [StandardInstrumentations] Check that the number of instructions doesn't change if analyses are preserved

nikic requested changes to this revision. nikic added a comment. This revision now requires changes to proceed. Please split out the fixes to individual passes. Comment at: llvm/lib/Passes/StandardInstrumentations.cpp:48 + cl::Hidden,

[PATCH] D145951: [docs] Add more complete documentation for -f[no]split-lto-unit

nikic added a comment. Thanks! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D145951/new/ https://reviews.llvm.org/D145951 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cg

[PATCH] D150997: [llvm] Split out DenseMapInfo specialization

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM. The diff looks weird for some reason, but the downloadable `.diff` file looks fine. CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150997/new/ https://reviews.llvm.org/D150997 __

[PATCH] D148216: Add support for annotations in UpdateTestChecks (NFC)

nikic added inline comments. Comment at: llvm/utils/UpdateTestChecks/common.py:1286 + if value == default_value: +continue if action.dest == 'filters': hnrklssn wrote: > nikic wrote: > > We should also not print the `all` argument for `--check-g

[PATCH] D148216: Add support for annotations in UpdateTestChecks (NFC)

nikic added inline comments. Comment at: llvm/utils/UpdateTestChecks/common.py:1286 + if value == default_value: +continue if action.dest == 'filters': nikic wrote: > hnrklssn wrote: > > nikic wrote: > > > We should also not print the `all` argu

[PATCH] D148216: Add support for annotations in UpdateTestChecks (NFC)

nikic added inline comments. Comment at: llvm/utils/UpdateTestChecks/common.py:1286 + if value == default_value: +continue if action.dest == 'filters': nikic wrote: > nikic wrote: > > hnrklssn wrote: > > > nikic wrote: > > > > We should also not

[PATCH] D153314: [clang] Replace uses of CGBuilderTy::CreateElementBitCast (NFC)

nikic added inline comments. Comment at: clang/lib/CodeGen/CGNonTrivialStruct.cpp:370 llvm::Value *DstArrayEnd = CGF.Builder.CreateInBoundsGEP(CGF.Int8Ty, BC.getPointer(), SizeInBytes); DstArrayEnd = CGF.Builder.CreateBitCast( Only `BC.getPoin

[PATCH] D148216: Add support for annotations in UpdateTestChecks (NFC)

nikic added a comment. There are some test failures. I believe there is one bug with the handling of unnamed globals. Previously we produced this: ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals --version 2 ; RUN: opt -S < %s | FileCheck

[PATCH] D144970: [llvm-c] Remove bindings for creating legacy passes

nikic added a comment. In D144970#4434305 , @yamt wrote: > at least some of these stuff is used by wasm-micro-runtime. do you have a > suggestion about how to adapt it? > https://github.com/bytecodealliance/wasm-micro-runtime/blob/72fc872afe9a51228b2a32b

[PATCH] D153694: [clang][CodeGen] Remove no-op EmitCastToVoidPtr (NFC)

nikic added inline comments. Comment at: clang/lib/CodeGen/ItaniumCXXABI.cpp:1473 QualType SrcRecordTy, QualType DestTy) { auto *ClassDecl = barannikov88 wrot

[PATCH] D153728: [llvm] Move AttributeMask to a separate header

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D153728/new/ https://reviews.llvm.org/D153728 ___ c

[PATCH] D153314: [clang] Replace uses of CGBuilderTy::CreateElementBitCast (NFC)

nikic added inline comments. Comment at: clang/lib/CodeGen/Targets/AArch64.cpp:522-523 auto *Load = CGF.Builder.CreateLoad(VAListAddr); Address Addr = Address(Load, CGF.Int8Ty, SlotSize); -return CGF.Builder.CreateElementBitCast(Addr, CGF.ConvertTypeForMem(Ty)); +

[PATCH] D153314: [clang] Replace uses of CGBuilderTy::CreateElementBitCast (NFC)

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D153314/new/ https://reviews.llvm.org/D153314 ___ c

[PATCH] D148240: [Coroutines] Directly remove unnecessary lifetime intrinsics

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG243e62b9d876: [Coroutines] Directly remove unnecessary lifetime intrinsics (authored by nikic). Herald added a project: clang. Herald added a subscri

[PATCH] D145043: [llvm-c] Remove PassRegistry and initialization APIs

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG62ef97e0631f: [llvm-c] Remove PassRegistry and initialization APIs (authored by nikic). Herald added a project: clang. Herald added a subscriber: cfe

[PATCH] D145403: [Pipeline] Don't run EarlyFPM in LTO post link

nikic added a comment. Any reason why this hasn't been landed yet? Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D145403/new/ https://reviews.llvm.org/D145403 ___ cfe-commits mailing list cfe-commits@list

[PATCH] D150997: [llvm] Split out DenseMapInfo specialization

nikic added a comment. The pre-merge checks have a build failure in clang-tools-extra. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150997/new/ https://reviews.llvm.org/D150997 ___ cfe-commits mailing l

[PATCH] D148216: Add support for annotations in UpdateTestChecks (NFC)

nikic added inline comments. Comment at: clang/test/utils/update_cc_test_checks/Inputs/annotations.c.expected:12 +// CHECK-NEXT:[[TMP1:%.*]] = load i32, ptr [[X]], align 4 +// CHECK-NEXT:ret i32 [[TMP1]] +// hnrklssn wrote: > hnrklssn wrote: > > nikic wr

[PATCH] D151195: [Driver] Fix test for use of ld from devtoolset

This revision was automatically updated to reflect the committed changes. Closed by commit rG75cce50fd2d3: [Driver] Fix test for use of ld from devtoolset (NFC) (authored by nikic). Herald added a project: clang. Herald added a subscriber: cfe-commits. Repository: rG LLVM Github Monorepo CHANG

[PATCH] D150997: [llvm] Split out DenseMapInfo specialization

nikic requested changes to this revision. nikic added a comment. This revision now requires changes to proceed. (Moving this off the review queue per above comments. There is a missing include in a unit test.) CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150997/new/ https://reviews.ll

[PATCH] D148216: Add support for annotations in UpdateTestChecks (NFC)

nikic added inline comments. Comment at: clang/test/utils/update_cc_test_checks/Inputs/annotations.c.expected:12 +// CHECK-NEXT:[[TMP1:%.*]] = load i32, ptr [[X]], align 4 +// CHECK-NEXT:ret i32 [[TMP1]] +// nikic wrote: > hnrklssn wrote: > > hnrklssn wr

[PATCH] D150670: [WebAssembly] Disable generation of fshl/fshr for rotates

nikic added a comment. In D150670#4368241 , @pmatos wrote: > In D150670#4368238 , @pmatos wrote: > >> In D150670#4352163 , @nikic wrote: >> >>> 1. Say that we prefer prese

[PATCH] D148216: Add support for annotations in UpdateTestChecks (NFC)

nikic added inline comments. Comment at: clang/test/utils/update_cc_test_checks/Inputs/annotations.c.expected:12 +// CHECK-NEXT:[[TMP1:%.*]] = load i32, ptr [[X]], align 4 +// CHECK-NEXT:ret i32 [[TMP1]] +// hnrklssn wrote: > nikic wrote: > > nikic wrote

[PATCH] D145265: [Pipeline] Remove GlobalCleanupPM

nikic added a comment. Unfortunately I'm seeing a number of Rust optimization regressions with this change. I'll try to reduce those to some PhaseOrdering tests. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D145265/new/ https://reviews.llvm.org/D1

[PATCH] D145265: [Pipeline] Remove GlobalCleanupPM

nikic added a comment. Here are two inputs that currently fold down to constants and no longer do so with this patch: https://gist.github.com/nikic/4a714ea550bf2252543570585f642af2 These need further reduction for PhaseOrdering tests, but should be a good starting point for analysis... Reposi

[PATCH] D148216: Add support for annotations in UpdateTestChecks (NFC)

nikic added inline comments. Comment at: clang/test/utils/update_cc_test_checks/Inputs/annotations.c.expected:12 +// CHECK-NEXT:[[TMP1:%.*]] = load i32, ptr [[X]], align 4 +// CHECK-NEXT:ret i32 [[TMP1]] +// hnrklssn wrote: > nikic wrote: > > hnrklssn wr

[PATCH] D150887: [clang] Convert a few tests to opaque pointers

nikic added a comment. I wanted to check whether you plan to do more opaque pointer test conversions in clang. Just to make sure we don't duplicate work... Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150887/new/ https://reviews.llvm.org/D150887

[PATCH] D150670: [WebAssembly] Disable generation of fshl/fshr for rotates

nikic added inline comments. Comment at: llvm/test/Transforms/InstCombine/fsh.ll:664 +; CHECK-NEXT:[[T1:%.*]] = and i32 [[A:%.*]], -65536 +; CHECK-NEXT:[[T2:%.*]] = call i32 @llvm.fshl.i32(i32 [[T1]], i32 [[T1]], i32 16) ; CHECK-NEXT:ret i32 [[T2]]

[PATCH] D150670: [WebAssembly] Disable generation of fshl/fshr for rotates

nikic added inline comments. Comment at: llvm/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp:927-933 +if (I->getOperand(0) != I->getOperand(1)) { + APInt DemandedMaskLHS(DemandedMask.lshr(ShiftAmt)); + APInt DemandedMaskRHS(DemandedMask.shl(

[PATCH] D151701: [HIP] Add missing __hip_atomic_fetch_sub support

nikic resigned from this revision. nikic added a comment. (Looks reasonable, but is pretty far outside my area of expertise...) Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D151701/new/ https://reviews.llvm.org/D151701

[PATCH] D150670: [InstCombine] Disable generation of fshl/fshr for rotates

nikic added a comment. Can you please drop all wasm related tests and instead add an InstCombine test for the fsh+and pattern? It would also be good to have a test where we can fold one side to a constant, but that constant is not zero. We should then consider whether that is profitable or not

[PATCH] D150670: [InstCombine] Disable generation of fshl/fshr for rotates

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM, let's give it a try. The patch description needs an update though. Comment at: llvm/lib/Transforms/InstCombine/InstCombineSimplifyDemanded.cpp:920 +return I;

[PATCH] D152023: [UBSan] Consider zero input to __builtin_clz/ctz to be undefined independent of the target.

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D152023/new/ https://reviews.llvm.org/D152023 ___ c

[PATCH] D142473: [UTC] Add --version argument

nikic added a comment. Sorry, I did not interpret those comments as blocking concerns. Happy to revert this if they were intended as such -- this is effectively an NFC patch until DEFAULT_VERSION gets bumped, so we can still backtrack to a different approach (though I'm not aware of any that wo

[PATCH] D137526: [C++20] [Modules] [NFC] Add Preprocessor methods for named modules - for ClangScanDeps (1/4)

nikic added a comment. It looks like this causes a minor compile-time regression (about 0.2% on `O0` builds): http://llvm-compile-time-tracker.com/compare.php?from=ad81d019a62d33e5d1f232844b83d3e164e4b160&to=6470706bc0ed2425704bc8c9e121a72e2d272616&stat=instructions:u I wanted to check whether

[PATCH] D143630: [clang] update test case

nikic added a comment. Hm, I don't think the clang/test/CodeGen/X86/x86-builtins-intrinsic.cl file exists on current main? Is this based on top of some other patch? Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D143630/new/ https://reviews.llvm.org

[PATCH] D144195: [XCore] Adapt Clang tests to opaque pointers.

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D144195/new/ https://reviews.llvm.org/D144195 ___ c

[PATCH] D144334: [Clang] Add C++2b attribute [[assume(expression)]]

nikic resigned from this revision. nikic added a comment. (not a clang reviewer) Comment at: clang/test/CodeGenCXX/cxx2b-assume.cpp:55 +// CHECK-OPT-NEXT: tail call void @llvm.assume(i1 %[[CMP]]) +// CHECK-OPT-NEXT: ret i32 43 +// Isn't the assume expressio

[PATCH] D145788: [CodeGen] Only consider innermost cast for !heapallocsite

This revision was automatically updated to reflect the committed changes. Closed by commit rGcac4d7ff4652: [CodeGen] Only consider innermost cast for !heapallocsite (authored by nikic). Herald added a project: clang. Herald added a subscriber: cfe-commits. Repository: rG LLVM Github Monorepo C

[PATCH] D150520: [clang] Convert a few tests to opaque pointers

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150520/new/ https://reviews.llvm.org/D150520 ___ c

[PATCH] D150520: [clang] Convert a few tests to opaque pointers

nikic accepted this revision. nikic added a comment. Still LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150520/new/ https://reviews.llvm.org/D150520 ___ cfe-commits mailing list cfe-commits@lists.l

[PATCH] D150530: [clang] Convert a few OpenMP tests to use opaque pointers

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Comment at: clang/test/OpenMP/atomic_capture_codegen.cpp:868 -// CHECK: [[NEW_BF_VALUE:%.+]] = load i8, i8* [[BITCAST1]] -// CHECK: [[RES:%.+]] = cmpxchg i8* getelementptr

[PATCH] D150652: [clang] Convert a few OpenMP tests to opaque pointers

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM, thanks! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150652/new/ https://reviews.llvm.org/D150652

[PATCH] D150680: [clang] Convert a few OpenMP tests to opaque pointers

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150680/new/ https://reviews.llvm.org/D150680 ___ c

[PATCH] D150682: [clang] Convert a few OpenMP tests to opaque pointers

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150682/new/ https://reviews.llvm.org/D150682 ___ c

[PATCH] D150704: [clang] Convert several smaller OpenMP tests to opaque pointers

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150704/new/ https://reviews.llvm.org/D150704 ___ c

[PATCH] D150694: [clang] Convert NVPTX OpenMP tests to opaque pointers

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150694/new/ https://reviews.llvm.org/D150694 ___ c

[PATCH] D150733: [clang] Convert remaining OpenMP tests to opaque pointers

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D150733/new/ https://reviews.llvm.org/D150733 ___ c

[PATCH] D150670: [WebAssembly] Disable generation of fshl/fshr for rotates

nikic requested changes to this revision. nikic added a comment. This revision now requires changes to proceed. This doesn't looks like a wasm specific problem. You get essentially the same issue on any target that has a rotate instruction but no funnel shift instruction. Here are just a couple

[PATCH] D145265: [Pipeline] Remove GlobalCleanupPM

nikic added inline comments. Comment at: llvm/lib/Passes/PassBuilderPipelines.cpp:476 + FPM.addPass(EarlyCSEPass()); + Why the extra EarlyCSE pass in the `O1` pipeline? Comment at: llvm/lib/Passes/PassBuilderPipelines.cpp:547 + Simplify

[PATCH] D150670: [WebAssembly] Disable generation of fshl/fshr for rotates

nikic added a comment. In D150670#4352055 , @pmatos wrote: > In D150670#4351147 , @nikic wrote: > >> This doesn't looks like a wasm specific problem. You get essentially the >> same issue on any target that has a

[PATCH] D150829: [clang] Convert several tests to opaque pointers

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Comment at: clang/test/CodeGenCXX/2011-12-19-init-list-ctor.cpp:22 +// CHECK: store i32 0, ptr @arr +// CHECK: call void @_ZN1AC1EPKc(ptr {{[^,]*}} getelementptr inbounds

[PATCH] D150887: [clang] Convert a few tests to opaque pointers

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Comment at: clang/test/CodeGenCXX/const-init-cxx11.cpp:353 }; - // CHECK: @_ZN14VirtualMembersL13sGlobalMemoryE = internal global { i8** } { i8** getelementptr inboun

[PATCH] D72538: [ThinLTO] Add additional ThinLTO pipeline testing with new PM

nikic added a comment. Herald added subscribers: wlei, pcwang-thead, ormris, wenlei. Herald added a project: All. Would it be possible to cut down the Clang side tests to only check parts that Clang controls in some way, e.g. that SLPVectorizer is enabled? This test is something of a PITA becaus

[PATCH] D148904: [Pipelines] Don't run ForceFunctionAttrs post-link

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGe7e4c7632075: [Pipelines] Don't run ForceFunctionAttrs post-link (authored by nikic). Herald added a project: clang. Herald added a subscriber: cfe-c

[PATCH] D149001: [InstSimplify] sdiv a (1 srem b) --> a

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG946b32680311: [InstSimplify] sdiv a (1 srem b) --> a (authored by floatshadow, committed by nikic). Herald added a project: clang. Herald added a sub

[PATCH] D72538: [ThinLTO] Add additional ThinLTO pipeline testing with new PM

nikic added a comment. In D72538#4296119 , @tejohnson wrote: > In D72538#4291552 , @nikic wrote: > >> Would it be possible to cut down the Clang side tests to only check parts >> that Clang controls in some way, e.

[PATCH] D72538: [ThinLTO] Add additional ThinLTO pipeline testing with new PM

nikic added a comment. Thanks! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D72538/new/ https://reviews.llvm.org/D72538 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-

[PATCH] D149210: [IR] Change shufflevector undef mask to poison

nikic requested changes to this revision. nikic added a comment. This revision now requires changes to proceed. Could you please split the change to printing + the test updates from all the other changes? The code changes get lost in the large diff. Comment at: llvm/include/ll

[PATCH] D148216: Add support for annotations in UpdateTestChecks (NFC)

nikic added inline comments. Comment at: clang/test/utils/update_cc_test_checks/Inputs/annotations.c.expected:12 +// CHECK-NEXT:[[TMP1:%.*]] = load i32, ptr [[X]], align 4 +// CHECK-NEXT:ret i32 [[TMP1]] +// hnrklssn wrote: > delcypher wrote: > > @hnrkls

[PATCH] D149210: [IR] Change shufflevector undef mask to poison

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM assuming tests pass. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D149210/new/ https://reviews.llvm.org/D149210

[PATCH] D143624: Inlining: Run the legacy AlwaysInliner before the regular inliner.

nikic added a comment. In D143624#4315468 , @dmgreen wrote: > It looks like there is quite a lot more optimization that happens to the > function being always-inlined (__SSAT) before this change. Through multiple > rounds of instcombine, almost to the e

[PATCH] D143624: Inlining: Run the legacy AlwaysInliner before the regular inliner.

nikic added a comment. In D143624#4316357 , @aeubanks wrote: > In D143624#4315508 , @nikic wrote: > >> In D143624#4315468 , @dmgreen >> wrote: >> >>> It looks like there

[PATCH] D140538: [Clang][CodeGen] Use poison instead of undef for dummy values in CGExpr [NFC]

nikic added a comment. The demangling test shouldn't be modified at all. CHANGES SINCE LAST ACTION https://reviews.llvm.org/D140538/new/ https://reviews.llvm.org/D140538 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org

[PATCH] D149641: [docs] Hide collaboration and include graphs in doxygen docs

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. Herald added a subscriber: StephenFan. LGTM, only the inheritance graph is useful, which this preserves if I understand correctly (CLASS_GRAPH is still YES). Repository: rG LLVM Github Monore

[PATCH] D152321: [clang] Replace use of Type::getPointerTo() (NFC)

nikic added inline comments. Comment at: clang/lib/CodeGen/Address.h:135 llvm::Constant *BitCast = llvm::ConstantExpr::getBitCast( -getPointer(), ElemTy->getPointerTo(getAddressSpace())); +getPointer(), llvm::PointerType::get(ElemTy, getAddressSpace()));

[PATCH] D152447: [Clang] Remove -no-opaque-pointers cc1 flag

This revision was automatically updated to reflect the committed changes. Closed by commit rG066fb7a58c5a: [Clang] Remove -no-opaque-pointers cc1 flag (authored by nikic). Herald added a project: clang. Herald added a subscriber: cfe-commits. Changed prior to commit: https://reviews.llvm.org/D1

[PATCH] D152321: [clang] Replace use of Type::getPointerTo() (NFC)

nikic added a comment. In D152321#4403719 , @JOE1994 wrote: > Some tests that run with `-no-opaque-pointers` began failing after I updated > the revision to get rid of bitcasts. > The bitcasts are still needed if running Clang with `-no-opaque-pointers`.

[PATCH] D152321: [clang] Replace use of Type::getPointerTo() (NFC)

nikic added inline comments. Comment at: clang/lib/CodeGen/CGBuilder.h:172 +auto *PtrTy = llvm::PointerType::get(Ty, Addr.getAddressSpace()); return Address(CreateBitCast(Addr.getPointer(), PtrTy, Name), Ty, Addr.getAlignment(), Addr.isKnownNonNull())

[PATCH] D152522: [NFC][SetVector] Update some usages of SetVector to SmallSetVector

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D152522/new/ https://reviews.llvm.org/D152522 ___ c

[PATCH] D152226: [FunctionAttrs] Propagate some func/arg/ret attributes from caller to callsite (WIP)

nikic added inline comments. Comment at: llvm/lib/Transforms/Utils/InferCallsiteAttrs.cpp:597 + // callsite violating the constraint. + if (checkCallerDoesNotAccessMemory() && !CB->doesNotAccessMemory()) { +// Wait until we know we actually need it to do potentially expensi

[PATCH] D152752: [MS] Fix passing aligned records by value in some cases

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D152752/new/ https://reviews.llvm.org/D152752 ___ c

[PATCH] D142823: Intrinsics: Allow tablegen to mark parameters with dereferenceable

nikic accepted this revision. nikic added a comment. This revision is now accepted and ready to land. LGTM CHANGES SINCE LAST ACTION https://reviews.llvm.org/D142823/new/ https://reviews.llvm.org/D142823 ___ cfe-commits mailing list cfe-commits@lis

<    1   2   3   4   5   >