[PATCH] D89158: [NewPM] Provide method to run all pipeline callbacks, used for -O0

Meinersbur added subscribers: serge-sans-paille, Meinersbur. Meinersbur added a comment. Did you consider just removing the `-NEXT` suffix? Polly might not be the only case, other statically linked pass-plugins may also insert themselves at extension points. @serge-sans-paille might have some i

[PATCH] D91054: [Clang][OpenMP] Frontend work for sections - D89671

Meinersbur added inline comments. Comment at: clang/lib/CodeGen/CGStmtOpenMP.cpp:3552-3554 +using BodyGenCallbackTy = +llvm::function_ref; Instead of duplicating it here, why not using `OpenMPIRBuilder::BodyGenCallbackTy`? Comment

[PATCH] D89974: [driver][CUDA] Use CMake's FindCUDA as default --cuda-path.

Meinersbur added a comment. The host's build environment already influences the compiler's defaults, e.g. D88929 . Especially libomptarget heavily uses it, and I would argue that these should be consistent, unless explicitly overwritten. However, I see that Ope

[PATCH] D89974: [driver][CUDA] Use CMake's FindCUDA as default --cuda-path.

Meinersbur updated this revision to Diff 300046. Meinersbur edited the summary of this revision. Meinersbur added a comment. Re-add accidentally removed space Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D89974/new/ https://reviews.llvm.org/D89974

[PATCH] D89974: [driver][CUDA] Use CMake's FindCUDA as default --cuda-path.

Meinersbur created this revision. Meinersbur added reviewers: jdoerfert, RaviNarayanaswamy, ye-luo, jhuber6, phosek, grokos, tambre, rnk, sylvestre.ledru, gtbercea, tra, yaxunl, Hahnfeld, jdenny. Meinersbur added a project: clang. Herald added subscribers: openmp-commits, dang, ormris, mgorny. He

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur added a comment. Thank you Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76342/new/ https://reviews.llvm.org/D76342 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.o

[PATCH] D83261: [OPENMP]Redesign of OMPExecutableDirective representation.

Meinersbur added a comment. I think the ActOnOpenMPDeclareMapperDirective changes should be separated into its own patch. Otherwise, LGTM. Comment at: clang/lib/Sema/SemaTemplateInstantiateDecl.cpp:3329 } - OMPDeclareMapperDecl *NewDMD = SemaRef.ActOnOpenMPDeclareMapperDi

[PATCH] D84780: Setting the /bigobj option globally for Windows debug build. https://bugs.llvm.org/show_bug.cgi?id=46733

Meinersbur added a comment. Hasn't this already been done in 80bd6ae13ea23d453a1f45d6ccdbfc7d0c877bb0 ? Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D84780/new/ https://reviews.l

[PATCH] D84038: On Windows build, making the /bigobj flag global , instead of passing it per file.

This revision was automatically updated to reflect the committed changes. Closed by commit rG80bd6ae13ea2: On Windows build, making the /bigobj flag global , instead of passing it per… (authored by zahiraam, committed by Meinersbur). Herald added projects: clang, LLDB. Herald added subscribers: l

[PATCH] D83261: [OPENMP]Redesign of OMPExecutableDirective representation.

Meinersbur added a comment. In D83261#2162561 , @ABataev wrote: > 1. OMPChildren class uses standard TrailingObjects harness instead of manual > calculation. Note that that having a separate object defeats the purpose of `TrailingObjects` of having jus

[PATCH] D83261: [OPENMP]Redesign of OMPExecutableDirective representation.

Meinersbur added a comment. In D83261#2162561 , @ABataev wrote: > 1. OMPChildren class uses standard TrailingObjects harness instead of manual > calculation. I was going to argue that OMPExecutableDirective could derive from TrailingObjects as well, bu

[PATCH] D67833: [OpenMP 5.0] Codegen support to pass user-defined mapper functions to runtime

Meinersbur added a comment. Seems that it already has been applied ;-) Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D67833/new/ https://reviews.llvm.org/D67833 ___ cfe-commits mailing list cfe-commits@

[PATCH] D83261: [OPENMP]Redesign of OMPExecutableDirective representation.

Meinersbur added a comment. AFAICS this extract out the handling of subnodes of OMPExecutableDirectives into the OMPChildren class which is made optional since `OMPChildren *OMPExecutableDirectives::Data` can be nullptr. However, since it also stores clauses, it seems that about every executabl

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur added a comment. ping @ABataev Have you checked the alternative scheme? Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76342/new/ https://reviews.llvm.org/D76342 ___ cfe-commits mailing list

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur marked an inline comment as done. Meinersbur added inline comments. Comment at: clang/include/clang/AST/StmtOpenMP.h:4781-4784 +/// This represents the '#pragma omp tile' loop transformation directive. +class OMPTileDirective final +: public OMPLoopDirective, +

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur marked an inline comment as done. Meinersbur added inline comments. Comment at: clang/include/clang/AST/StmtOpenMP.h:4781-4784 +/// This represents the '#pragma omp tile' loop transformation directive. +class OMPTileDirective final +: public OMPLoopDirective, +

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76342/new/ https://reviews.llvm.org/D76342 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. Herald added a subscriber: sstefan1. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69088/new/ https://reviews.llvm.org/D69088 ___ cfe-commits mailing list cfe-commits@l

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur added a comment. Herald added a subscriber: sstefan1. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76342/new/ https://reviews.llvm.org/D76342 ___ cfe-commits mailing list cfe-commits@l

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#2023114 , @kbarton wrote: > @Meinersbur I missed the RFC and discussion on the cfe-dev mailing list. > Could you post a link here so that it's included in the history? See the collection of links in a previous comme

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur added a comment. Herald added a reviewer: DavidTruby. Herald added a project: LLVM. Herald added a subscriber: llvm-commits. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76342/new/ https://reviews.llvm.org/D76342 ___

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69088/new/ https://reviews.llvm.org/D69088 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D78192: Support compiler extensions as a normal component

Meinersbur accepted this revision. Meinersbur added a comment. This revision is now accepted and ready to land. LGTM. As with every change to the build system, please look out broken configurations after committing. You will have noticed already that there is a lot of variety in how others buil

[PATCH] D78192: Support compiler extensions as a normal component

Meinersbur added a comment. I looks to work fine under my Windows and Linux configuration. One Point though: llvm-config --libs Extensions returns [...] -lPollyPPCG -lPollyISL -lPolly I think `-lPolly` must be first to work with bfd.ld order semantics. See https://stackoverflow.com/questi

[PATCH] D78192: Support compiler extensions as a normal component

Meinersbur added a comment. Is this ready to try out? In D78192#1996859 , @serge-sans-paille wrote: > Note that this should solve in an elegant (?) way the multiple link errors > found when linking clang with LLVM_DYLIB. Aren't these fixed already?

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur marked an inline comment as done. Meinersbur added inline comments. Comment at: clang/include/clang/AST/StmtOpenMP.h:4781-4784 +/// This represents the '#pragma omp tile' loop transformation directive. +class OMPTileDirective final +: public OMPLoopDirective, +

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69088/new/ https://reviews.llvm.org/D69088 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur added a comment. pong Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76342/new/ https://reviews.llvm.org/D76342 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69088/new/ https://reviews.llvm.org/D69088 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur added a comment. Herald added a subscriber: yaxunl. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76342/new/ https://reviews.llvm.org/D76342 ___ cfe-commits mailing list cfe-commits@lis

[PATCH] D77058: [Clang] Add llvm.loop.unroll.disable to loops with -fno-unroll-loops.

Meinersbur added a comment. LGTM, since it continues current practice. `optnone` will always be the more annoying. In D77058#1964714 , @fhahn wrote: > IIUC we don't need a patch similar like this one for optnone, as it already > gets added to the functi

[PATCH] D77058: [Clang] Add llvm.loop.unroll.disable to loops with -fno-unroll-loops.

Meinersbur added a comment. Note that loop-metadata is best-effort only and may be forgotten in the optimization pipeline. Do we also need an equivalent to `-Xclang -disable-O0-optnone`? Personally, I don't like to the `optnone` approach: There have been many post on llvm-dev using `clang -emi

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69088/new/ https://reviews.llvm.org/D69088 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76342/new/ https://reviews.llvm.org/D76342 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur added a comment. In D76342#1941591 , @JonChesterfield wrote: > edit: actually you've already done the clang-format on trunk as I hoped, > phabricator mislead me. Apologies for the noise The previous diff was against a branch with only the wh

[PATCH] D76342: [OpenMP] Implement '#pragma omp tile'

Meinersbur added a comment. ping Repository: rC Clang CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76342/new/ https://reviews.llvm.org/D76342 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman

[PATCH] D74801: [ADT][NFC] SCCIterator: Change hasLoop() to hasCycle()

Meinersbur closed this revision. Meinersbur added a comment. In D74801#189 , @baziotis wrote: > I committed that > (https://reviews.llvm.org/rG21390eab4c05e0ed7e7d13ada9e85f62b87ea484) but as > it seems, I should have added the differential division

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

Meinersbur added a comment. Nice work 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 -instcombine -instcombine-infin

[PATCH] D74757: Fix compiler extension in standalone mode

Meinersbur added a comment. The Windows pre-merge check seem to have failed? Comment at: llvm/cmake/modules/AddLLVM.cmake:886 # process_llvm_pass_plugins([NO_GEN]) # Change to `[GEN_CONFIG]` Comment at: llvm/cmake/modules/LLVMConfig.cmake

[PATCH] D70572: [Serialization] #pragma clang transform

Meinersbur updated this revision to Diff 234663. Meinersbur added a comment. - Adapt for refactored ASTRecordReader/Writer Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D70572/new/ https://reviews.llvm.org/D70572 Files: clang/include/clang/Serial

[PATCH] D69091: [Sema] #pragma clang transform

Meinersbur updated this revision to Diff 234659. Meinersbur added a comment. - Rework TransformedTree a bit Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69091/new/ https://reviews.llvm.org/D69091 Files: clang/include/clang/AST/ASTNodeTraverser.

[PATCH] D61446: Generalize the pass registration mechanism used by Polly to any third-party tool

Meinersbur added a comment. Btw, this patch did not apply cleanly on trunk. I instead test the version from your repository . Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.ll

[PATCH] D61446: Generalize the pass registration mechanism used by Polly to any third-party tool

Meinersbur accepted this revision. Meinersbur added a comment. This revision is now accepted and ready to land. I tested the configurations mentioned at https://groups.google.com/d/topic/polly-dev/vxumPMhrSEs/discussion again and did not find any issues that were not there before. Therefore IMHO

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#1772141 , @rjmccall wrote: > This is a major new language feature, and code review is probably not the > right venue for reviewing it; there should be a thread on cfe-dev. My > apologies if that's already been discu

[PATCH] D71447: [CIndex] #pragma clang transform

Meinersbur created this revision. Meinersbur added reviewers: abramobagnara, ddunbar, aganea, hfinkel, rjmccall, kbarton, SjoerdMeijer, aaron.ballman, ABataev, fhahn, hsaito, hans, greened, dmgreen, Ayal, asavonic, rtrieu, dorit, rsmith, tyler.nowicki, jdoerfert. Herald added a subscriber: arpham

[PATCH] D70032: [docs] #pragma clang transform

Meinersbur updated this revision to Diff 233717. Meinersbur added a comment. - Add release notes Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D70032/new/ https://reviews.llvm.org/D70032 Files: clang/docs/LanguageExtensions.rst clang/docs/Pragm

[PATCH] D69092: [CodeGen] #pragma clang transform

Meinersbur updated this revision to Diff 233713. Meinersbur added a comment. - Remove handling of OpenMP simd and LoopHints by CGTransform Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69092/new/ https://reviews.llvm.org/D69092 Files: clang/lib/

[PATCH] D70572: [Serialization] #pragma clang transform

Meinersbur updated this revision to Diff 233714. Meinersbur added a comment. - Rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D70572/new/ https://reviews.llvm.org/D70572 Files: clang/include/clang/Serialization/ASTBitCodes.h clang/include

[PATCH] D69091: [Sema] #pragma clang transform

Meinersbur updated this revision to Diff 233707. Meinersbur added a comment. - Remove handling of LoopHints (#pragma clang loop) and OpenMP (#pragma omp simd) to reducethe size of the patches. Compatibility with OpenMP is still a goal (to implement the upcoming OpenMP loop transformations su

[PATCH] D69089: [Parser] #pragma clang transform

Meinersbur updated this revision to Diff 233704. Meinersbur added a comment. - Simplify transformation classes Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69089/new/ https://reviews.llvm.org/D69089 Files: clang/include/clang/AST/StmtTransform.

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69088/new/ https://reviews.llvm.org/D69088 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D61446: Generalize the pass registration mechanism used by Polly to any third-party tool

Meinersbur added a comment. In D61446#1763278 , @serge-sans-paille wrote: > @Meinersbur I have a strange symbol issue when activating the GPu part: > https://github.com/serge-sans-paille/llvm-project/pull/2/checks?check_run_id=324901896#step:5:4888 > An

[PATCH] D69628: [Clang] Pragma vectorize_width() implies vectorize(enable), take 3

Meinersbur accepted this revision. Meinersbur added a comment. This revision is now accepted and ready to land. LGTM CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69628/new/ https://reviews.llvm.org/D69628 ___ cfe-commits mailing list cfe-c

[PATCH] D61446: Generalize the pass registration mechanism used by Polly to any third-party tool

Meinersbur added a comment. I still have to try out the patch. In D61446#1756211 , @serge-sans-paille wrote: > @Meinersbur to make your reviewer job easier, I've setup validation for that > patch, see https://github.com/serge-sans-paille/llvm-project/pu

[PATCH] D69092: [CodeGen] #pragma clang transform

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69092/new/ https://reviews.llvm.org/D69092 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D70572: [Serialization] #pragma clang transform

Meinersbur updated this revision to Diff 230670. Meinersbur added a comment. - Remove some leftovers Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D70572/new/ https://reviews.llvm.org/D70572 Files: clang/include/clang/Serialization/ASTBitCodes.h

[PATCH] D70572: [Serialization] #pragma clang transform

Meinersbur created this revision. Meinersbur added reviewers: hfinkel, kbarton, SjoerdMeijer, aaron.ballman, ABataev, fhahn, hsaito, hans, greened, dmgreen, Ayal, asavonic, rtrieu, dorit, rsmith, tyler.nowicki, jdoerfert. Herald added subscribers: cfe-commits, zzheng. Herald added a project: clan

[PATCH] D69091: [Sema] #pragma clang transform

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69091/new/ https://reviews.llvm.org/D69091 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D69089: [Parser] #pragma clang transform

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69089/new/ https://reviews.llvm.org/D69089 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur updated this revision to Diff 230094. Meinersbur added a comment. - Address @ABataev's review Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69088/new/ https://reviews.llvm.org/D69088 Files: clang/include/clang/Basic/LangOptions.def

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69088/new/ https://reviews.llvm.org/D69088 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/

[PATCH] D69628: [Clang] Pragma vectorize_width() implies vectorize(enable), take 3

Meinersbur added inline comments. Comment at: clang/lib/CodeGen/CGLoopInfo.cpp:302-306 +// Imply vectorize.enable when it is not already disabled/enabled. +Args.push_back( +MDNode::get(Ctx, {MDString::get(Ctx, "llvm.loop.vectorize.enable"), +

[PATCH] D69897: Add #pragma clang loop vectorize_assume_alignment(n)

Meinersbur added a comment. Could you elaborate why this is better than `__builtin_assume_aligned`? Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69897/new/ https://reviews.llvm.org/D69897 ___ cfe-comm

[PATCH] D69628: [Clang] Pragma vectorize_width() implies vectorize(enable), take 3

Meinersbur added a comment. Sorry for the delay, for some reason I did not see a notification for the update. Comment at: clang/lib/CodeGen/CGLoopInfo.cpp:302-306 +// Imply vectorize.enable when it is not already disabled/enabled. +Args.push_back( +MDNode::get(

[PATCH] D69825: [Clang][Driver] Bypass cc1 process and re-enter driver main

Meinersbur added inline comments. Comment at: clang/lib/Driver/Job.cpp:347 +StringRef DriverExe = llvm::sys::path::stem(D.ClangExecutable); +if (CommandExe.equals_lower(DriverExe)) + ClangDriverMain = Driver::Main; Why is this check necessary? Why no

[PATCH] D69785: [OpenMP] Introduce the OpenMP-IR-Builder

Meinersbur added inline comments. Comment at: llvm/include/llvm/IR/OpenMPKinds.def:186 +///{ + +#ifndef OMP_IDENT_FLAG JonChesterfield wrote: > jdoerfert wrote: > > Meinersbur wrote: > > > jdoerfert wrote: > > > > JonChesterfield wrote: > > > > > Sharing constant

[PATCH] D70032: [docs] #pragma clang transform

Meinersbur created this revision. Meinersbur added reviewers: hfinkel, kbarton, SjoerdMeijer, aaron.ballman, ABataev, fhahn, hsaito, hans, greened, dmgreen, Ayal, asavonic, rtrieu, dorit, rsmith, tyler.nowicki, jdoerfert. Herald added a subscriber: arphaman. Herald added a project: clang. Meiners

[PATCH] D69089: [Parser] #pragma clang transform

Meinersbur updated this revision to Diff 228297. Meinersbur added a comment. - Switch to monorepo - Move Transform.h/cpp to clangBasic - Separate Transform and TransformClause Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69089/new/ https://reviews

[PATCH] D69897: Add #pragma clang loop aligned

Meinersbur added a comment. In D69897#1735983 , @lebedev.ri wrote: > It would be really best if this could just make use of > `CodeGenFunction::EmitAlignmentAssumption()`, > else this is a heavy hammer with loose handle. Since it is a property of the d

[PATCH] D69897: Add #pragma clang loop aligned

Meinersbur added a comment. Please see my comments in D69900 as well. For the font-end side, I suggest the syntax `vectorize_assume_alignment(32)` instead. We also need to define what this assumes to be aligned. The first element of the array? Every array eleme

[PATCH] D69853: [OpenMP][NFCI] Introduce llvm/IR/OpenMPConstants.h

Meinersbur added inline comments. Comment at: clang/lib/Sema/SemaOpenMP.cpp:4011 unsigned NamedModifiersNumber = 0; - SmallVector FoundNameModifiers( - OMPD_unknown + 1); + SmallVector + FoundNameModifiers(unsigned(OMPD_unknown) + 1); ABataev wrote: >

[PATCH] D69785: [OpenMP] Introduce the OpenMP-IR-Builder

Meinersbur added inline comments. Comment at: clang/include/clang/Basic/LangOptions.def:215 LANGOPT(OpenMPCUDAMode, 1, 0, "Generate code for OpenMP pragmas in SIMT/SPMD mode") +LANGOPT(OpenMPNewCodegen , 1, 0, "Use the experimental OpenMP-IR-Builder codegen path.") LANGO

[PATCH] D69825: [Clang][Driver] Bypass cc1 process and re-enter driver main

Meinersbur added inline comments. Comment at: clang/lib/Driver/Job.cpp:340-355 + typedef int (*ClangDriverMainFunc)(SmallVectorImpl &); + ClangDriverMainFunc ClangDriverMain = nullptr; + + if (ReenterTool) { +StringRef F = llvm::sys::path::filename(Executable); +if (F.

[PATCH] D69825: [Clang][Driver] Bypass cc1 process and re-enter driver main

Meinersbur added a comment. This patch reduced the build time from 12 to 7 minutes? Sounds too good to be true. Thanks for the work! I would primarily interested in this to avoid determining the `-cc1` command line when debugging. However, the implementation of calling it's own main function gi

[PATCH] D69764: [clang-format] Add Left/Right Const (East/West , Before/After) fixer capability

Meinersbur added a comment. In D69764#1732232 , @MyDeveloperDay wrote: > @pollydev too (a previously clang-format clean directory) will generate > changes running `clang-format --const-style=west -i -n *.cpp` > > ScopBuilder.cpp:74:9: warning: code sho

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur updated this revision to Diff 227561. Meinersbur added a comment. - Implement -f(no-)experimental-transform-pragma Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69088/new/ https://reviews.llvm.org/D69088 Files: clang/include/clang/Bas

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur updated this revision to Diff 227550. Meinersbur added a comment. - Use PRAGMA_ANNOTATION - Monorepo layout Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D69088/new/ https://reviews.llvm.org/D69088 Files: clang/include/clang/Basic/Toke

[PATCH] D69628: [Clang] Pragma vectorize_width() implies vectorize(enable), take 3

Meinersbur requested changes to this revision. Meinersbur added inline comments. This revision now requires changes to proceed. Comment at: clang/lib/CodeGen/CGLoopInfo.cpp:272-277 + if (Attrs.VectorizeWidth > 1 && + Attrs.VectorizeEnable == LoopAttributes::Unspecified) +

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#1715038 , @hsaito wrote: > If there is a precedence, just follow that. Else, how to spell an > experimental clang pragma would be a good discussion topic by itself. If I > need to provide a discussion starter, I'd sa

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#1713623 , @hsaito wrote: > @Meinersbur, if I remember correctly, there was an RFC discussion on this > topic, right? If yes, would you post the pointer to that? I need a refresher > on what has been discussed/settled

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#1714020 , @hsaito wrote: > Personally, I like the intent. I don't foresee a clear (enough) path to get > there. This leads to hesitation of adding a new non-experimental pragma and > present it to programmers. If you

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#1713933 , @hsaito wrote: > Have we established general consensus for the desire to have the flexible > enough loop optimization pass ordering to accomplish the outcome of the new > directive, and shared vision for th

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#1713915 , @ABataev wrote: > Just do not allow this form with respect_order clause. What exactly would be the rules what is allowed and what isn't? We can just not allow not mixing `#pragma clang loop` and `#pragma c

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#1713831 , @tyler.nowicki wrote: > That approach would avoid the inevitable conflicts of having both loop and > transform pragmas on the same loop. I fear it will give us far worse ambiguities. Consider: #pragma

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#1713831 , @tyler.nowicki wrote: > In D69088#1713648 , @Meinersbur > wrote: > > > In D69088#1713623 , @hsaito wrote: > > > > > @Meiners

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#1713623 , @hsaito wrote: > @Meinersbur, if I remember correctly, there was an RFC discussion on this > topic, right? If yes, would you post the pointer to that? I need a refresher > on what has been discussed/settled

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur added a comment. In D69088#1713147 , @ABataev wrote: > Why not try to improve the existing #pragma clang loop rather than add a new > pragma with almost the same behavior? The behavior and syntax is different. #pragma clang loop ignores the o

[PATCH] D69089: [Parser] #pragma clang transform

Meinersbur created this revision. Meinersbur added reviewers: hfinkel, kbarton, SjoerdMeijer, aaron.ballman, ABataev, fhahn, hsaito, hans, greened, dmgreen, reames, Ayal, asavonic, rtrieu, dorit, rsmith, tyler.nowicki. Herald added a subscriber: mgorny. Herald added a reviewer: jdoerfert. Parse

[PATCH] D69088: [Lex] #pragma clang transform

Meinersbur created this revision. Meinersbur added reviewers: hfinkel, kbarton, SjoerdMeijer, aaron.ballman, ABataev, fhahn, hsaito, hans, greened, dmgreen, reames, Ayal, asavonic, rtrieu, dorit, rsmith, tyler.nowicki. Herald added a reviewer: bollu. Herald added a reviewer: jdoerfert. Herald add

[PATCH] D68551: [clang-format] [NFC] Ensure clang-format is itself clang-formatted.

Meinersbur added a comment. We should seek having only one formatting policy for the entire project, it's too confusing otherwise (e.g. LLVMSupport/ADT is also used by clang-format, are they force-formatted as well? What about the tools/utils subdirectories?). If you would like to gradually mov

[PATCH] D68551: [clang-format] [NFC] Ensure clang-format is itself clang-formatted.

Meinersbur added a comment. In D68551#1697842 , @mitchell-stellar wrote: > I agree that a system in place that either enforces clang-formatting on > commit or after the fact would be ideal. Otherwise, I don't see a need to > have to approve these NFC co

[PATCH] D61446: Generalize the pass registration mechanism used by Polly to any third-party tool

Meinersbur added inline comments. Comment at: polly/lib/Support/RegisterPasses.cpp:726 + +#ifndef LLVM_POLLY_LINK_INTO_TOOLS +extern "C" LLVM_ATTRIBUTE_WEAK ::llvm::PassPluginLibraryInfo Meinersbur wrote: > [serious] `LLVM_POLLY_LINK_INTO_TOOLS` is a cmake config

[PATCH] D67978: [OpenMP 5.0] Fix user-defined mapper lookup in sema

This revision was automatically updated to reflect the committed changes. Closed by commit rL373023: [OpenMP 5.0] Fix user-defined mapper lookup in sema (authored by Meinersbur, committed by ). Herald added a project: LLVM. Herald added a subscriber: llvm-commits. Changed prior to commit: https

[PATCH] D61446: Generalize the pass registration mechanism used by Polly to any third-party tool

Meinersbur added a comment. Keep in mind that for static linking you will need something that pulls-in a symbol from the pass static library. In this patch, `NewPMDriver.cpp` does it for `opt` by calling `get##Ext##PluginInfo()`. In clang, it is `BackendUtil.cpp`. But nothing for `bugpoint` hen

[PATCH] D61446: Generalize the pass registration mechanism used by Polly to any third-party tool

Meinersbur requested changes to this revision. Meinersbur added inline comments. This revision now requires changes to proceed. Comment at: llvm/test/Feature/load_extension.ll:1 +; RUN: not test -f %llvmshlibdir/libBye%shlibext || opt %s -load=%llvmshlibdir/libBye%shlibext -good

[PATCH] D61446: Generalize the pass registration mechanism used by Polly to any third-party tool

Meinersbur added a comment. Sorry for the break. Unfortunately the patch does not apply cleanly anymore. Can you rebase to latest trunk? I surprisingly did not know about `__declspec(selectany)`. I still think having the symbol only defined in that loadable module, but not in the statically li

[PATCH] D66796: [clang] Loop pragma vectorize(disable)

Meinersbur added a comment. Therefore, `vectorize(disable)` would also disable interleaving? CHANGES SINCE LAST ACTION https://reviews.llvm.org/D66796/new/ https://reviews.llvm.org/D66796 ___ cfe-commits mailing list cfe-commits@lists.llvm.org ht

[PATCH] D64564: Loop pragma parsing. NFC.

Meinersbur accepted this revision. Meinersbur added inline comments. This revision is now accepted and ready to land. Comment at: clang/lib/Parse/ParsePragma.cpp:1010 + StringRef Str = PragmaName.getIdentifierInfo()->getName(); + std::string ClangLoopStr = (Twine("clang loop ")

[PATCH] D66199: [docs] loop pragmas

Meinersbur added inline comments. Comment at: docs/LanguageExtensions.rst:3068-3069 +There are loop hints that control transformations (e.g. vectorization, loop +unrolling) and there loop hints that set transformation options (e.g. +``vectorize_width``, ``unroll_count``). Prag

[PATCH] D66199: [docs] loop pragmas

Meinersbur added a comment. Since this is user documentation, we should only add it here once it is true. Comment at: docs/LanguageExtensions.rst:3068-3069 +There are loop hints that control transformations (e.g. vectorization, loop +unrolling) and there loop hints that set t

<    1   2   3   4   5   >