[PATCH] D66673: [OPENMP][NVPTX]Fix critical region codegen.

ABataev created this revision. ABataev added a reviewer: ABataev. Herald added subscribers: cfe-commits, guansong, jholewinski. Herald added a reviewer: jdoerfert. Herald added a project: clang. Previously critical regions were emitted with the barrier making it a worksharing construct though it i

[PATCH] D66673: [OPENMP][NVPTX]Fix critical region codegen.

ABataev added a comment. In D66673#1643544 , @jdoerfert wrote: > I guess IR test should be affected already and it would be good to have the > run time test that breaks with barriers. Runtime test is libomptarget/deviceRTLs/nvptx/test/parallel/spmd_par

[PATCH] D66673: [OPENMP][NVPTX]Fix critical region codegen.

ABataev updated this revision to Diff 216934. ABataev added a comment. Fix the test Repository: rC Clang CHANGES SINCE LAST ACTION https://reviews.llvm.org/D66673/new/ https://reviews.llvm.org/D66673 Files: lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp test/OpenMP/nvptx_parallel_codegen.cpp

[PATCH] D66485: [Clang][Bundler] Use llvm-objcopy for creating fat object files

ABataev added inline comments. Comment at: clang/tools/clang-offload-bundler/ClangOffloadBundler.cpp:485 +"=" + InputFileNames[I])); +ObjcopyArgs.push_back(InputFileNames[HostInputIndex]); +ObjcopyArgs.push_back(OutputFileNames.front())

[PATCH] D66485: [Clang][Bundler] Use llvm-objcopy for creating fat object files

ABataev added inline comments. Comment at: clang/test/Driver/clang-offload-bundler.c:233 +// RUN: clang-offload-bundler -type=o -targets=host-%itanium_abi_triple,openmp-powerpc64le-ibm-linux-gnu,openmp-x86_64-pc-linux-gnu -inputs=%t.o,%t.tgt1,%t.tgt2 -outputs=%t.bundle3.o -###

[PATCH] D66485: [Clang][Bundler] Use llvm-objcopy for creating fat object files

ABataev accepted this revision. ABataev added a comment. This revision is now accepted and ready to land. LG Repository: rC Clang CHANGES SINCE LAST ACTION https://reviews.llvm.org/D66485/new/ https://reviews.llvm.org/D66485 ___ cfe-commits ma

[PATCH] D66673: [OPENMP][NVPTX]Fix critical region codegen.

This revision was automatically updated to reflect the committed changes. Closed by commit rL369946: [OPENMP][NVPTX]Fix critical region codegen. (authored by ABataev, committed by ). Herald added a project: LLVM. Herald added a subscriber: llvm-commits. Repository: rL LLVM CHANGES SINCE LAST A

[PATCH] D66601: [Clang][Bundler] Do not require host triple for extracting device bundles

ABataev added inline comments. Comment at: clang/tools/clang-offload-bundler/ClangOffloadBundler.cpp:888 + // treat missing host triple as error if we do unbundling. + if ((Unbundle && HostTargetNum > 1) || (!Unbundle && HostTargetNum != 1)) { Error = true; ---

[PATCH] D66601: [Clang][Bundler] Do not require host triple for extracting device bundles

ABataev added inline comments. Comment at: clang/tools/clang-offload-bundler/ClangOffloadBundler.cpp:888 + // treat missing host triple as error if we do unbundling. + if ((Unbundle && HostTargetNum > 1) || (!Unbundle && HostTargetNum != 1)) { Error = true; ---

[PATCH] D66601: [Clang][Bundler] Do not require host triple for extracting device bundles

ABataev added inline comments. Comment at: clang/test/Driver/clang-offload-bundler.c:228-231 +// Check that we can extract target parts without providing host triple. +// RUN: clang-offload-bundler -type=ast -targets=openmp-powerpc64le-ibm-linux-gnu -outputs=%t.res.tgt1 -inputs

[PATCH] D66598: [Clang][Bundler] Fix for a hang when unbundling fat binary

ABataev added a comment. Could you provide some more details about the problem? CHANGES SINCE LAST ACTION https://reviews.llvm.org/D66598/new/ https://reviews.llvm.org/D66598 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.l

[PATCH] D66598: [Clang][Bundler] Fix for a hang when unbundling fat binary

ABataev added inline comments. Comment at: clang/tools/clang-offload-bundler/ClangOffloadBundler.cpp:295 return StringRef(); - +CurBundleInfo = NextBundleInfo++; return CurBundleInfo->first(); Maybe just: ``` const BundleInfo &Bundle = *CurBundleI

[PATCH] D66598: [Clang][Bundler] Fix for a hang when unbundling fat binary

ABataev accepted this revision. ABataev added a comment. This revision is now accepted and ready to land. LG CHANGES SINCE LAST ACTION https://reviews.llvm.org/D66598/new/ https://reviews.llvm.org/D66598 ___ cfe-commits mailing list cfe-commits@l

[PATCH] D66601: [Clang][Bundler] Do not require host triple for extracting device bundles

ABataev added inline comments. Comment at: clang/tools/clang-offload-bundler/ClangOffloadBundler.cpp:888 + // treat missing host triple as error if we do unbundling. + if ((Unbundle && HostTargetNum > 1) || (!Unbundle && HostTargetNum != 1)) { Error = true; ---

[PATCH] D66601: [Clang][Bundler] Do not require host triple for extracting device bundles

ABataev accepted this revision. ABataev added a comment. This revision is now accepted and ready to land. LG CHANGES SINCE LAST ACTION https://reviews.llvm.org/D66601/new/ https://reviews.llvm.org/D66601 ___ cfe-commits mailing list cfe-commits@l

[PATCH] D64375: [OpenMP][Docs] Provide implementation status details

ABataev added inline comments. Comment at: clang/docs/OpenMPSupport.rst:204 ++--+--+--++ +| device extension | c

[PATCH] D66668: [OPENMP][Analysis] Add analysis of the map clauses.

This revision was automatically updated to reflect the committed changes. Closed by commit rL370214: [OPENMP][Analysis] Add analysis of the map clauses. (authored by ABataev, committed by ). Herald added a project: LLVM. Herald added a subscriber: llvm-commits. Repository: rL LLVM CHANGES SINC

[PATCH] D66559: [OPENMP] Update the diagnosis message for canonical loop form

ABataev added a comment. What about the check for `!=`? it would be good to allow it only for OpenMP 5.0. Also, add/update the tests. Comment at: clang/include/clang/Basic/DiagnosticSemaKinds.td:9027-9032 +def err_omp_before_50_loop_not_canonical_cond : Error< "condition of

[PATCH] D66559: [OPENMP] Update the diagnosis message for canonical loop form

ABataev added a comment. In D66559#1649775 , @cchen wrote: > Haven't updated the lit test for OpenMP 5.0 yet since I'm not able to check > OPENMP version by using _OPENMP with preprocessor condition checking. You don't need `_OPENMP` macro for this. The

[PATCH] D66559: [OPENMP] Update the diagnosis message for canonical loop form

ABataev added a comment. In D66559#1651756 , @cchen wrote: > "parallel_for_codegen" cannot pass now since it has two test cases for "!=" > and it's hard to just add "fopenmp-version=50" since it will break lots of > other test cases that have different c

[PATCH] D66559: [OPENMP] Update the diagnosis message for canonical loop form

ABataev accepted this revision. ABataev added a comment. This revision is now accepted and ready to land. LG, thanks for the fix! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D66559/new/ https://reviews.llvm.org/D66559 __

[PATCH] D66559: [OPENMP] Update the diagnosis message for canonical loop form

ABataev added inline comments. Comment at: clang/lib/Sema/SemaOpenMP.cpp:5423 + SemaRef.Diag(DefaultLoc, diag::err_omp_loop_not_canonical_cond) + << IneqCondIsCanonical << LCDecl; return true; I would not suggest to rely o

[PATCH] D66559: [OPENMP] Update the diagnosis message for canonical loop form

ABataev accepted this revision. ABataev added a comment. LG Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D66559/new/ https://reviews.llvm.org/D66559 ___ cfe-commits mailing list cfe-commits@lists.llvm.

[PATCH] D64375: [OpenMP][Docs] Provide implementation status details

ABataev accepted this revision. ABataev added a comment. This revision is now accepted and ready to land. LG Comment at: clang/docs/OpenMPSupport.rst:204 ++--+--+--+-

[PATCH] D67294: Register and parse a simplified version of '#pragma omp declare variant'

ABataev added inline comments. Comment at: include/clang/Basic/Attr.td:3202 +def OMPDeclareVariantDecl : Attr { + let Spellings = [Pragma<"omp", "declare variant">]; For the first patch we don't need the attribute, just parsing/sema analysis

[PATCH] D66559: [OPENMP] Update the diagnosis message for canonical loop form

ABataev added a comment. In D66559#1663887 , @cchen wrote: > @ABataev, could you mark it as ready to land, please. Thanks. I accepted it already, don't know what else I can do. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://

[PATCH] D71969: [OpenMP] diagnose zero-length array section in the depend clause

ABataev added inline comments. Comment at: clang/lib/Sema/SemaOpenMP.cpp:15031-15033 +Length->EvaluateAsInt(Result, Context)) { + llvm::APSInt ConstLength = Result.Val.getInt(); + if (ConstLength.getSExtValue() == 0) { Just `.. Lengt

[PATCH] D71884: [OpenMP] Fix formatting of OpenMP error message.

This revision was automatically updated to reflect the committed changes. Closed by commit rG87a004d0f8c2: [OpenMP] Fix formatting of OpenMP error message, by Wang Tianqing. (authored by ABataev). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D71884/n

[PATCH] D71969: [OpenMP] diagnose zero-length array section in the depend clause

ABataev accepted this revision. ABataev added a comment. This revision is now accepted and ready to land. LG CHANGES SINCE LAST ACTION https://reviews.llvm.org/D71969/new/ https://reviews.llvm.org/D71969 ___ cfe-commits mailing list cfe-commits@l

[PATCH] D69585: PerformPendingInstatiations() already in the PCH

ABataev added a comment. I don't see any crashes in OpenMP tests, other tests just must be updated by the author, if this patch will ever going to be landed. I don't think it is a good idea to have a not fully functional implementation, even guarded by the option. CHANGES SINCE LAST ACTION

[PATCH] D72811: [WIP][OPENMP5.0] allow lvalue for motion clause

ABataev added inline comments. Comment at: clang/lib/CodeGen/CGOpenMPRuntime.cpp:7709-7712 isa(Next->getAssociatedExpression()) || isa(Next->getAssociatedExpression()) || -isa(Next->getAssociatedExpression())) && +

[PATCH] D86562: [OPENMP][NFC]Release notes for OpenMP in clang (11.x).

ABataev added a comment. In D86562#2274036 , @hans wrote: > Please go ahead and commit. Hi Hans, can you commit it yourself? I have some technical issues and not able to commit it. And I don't know when it is going to be resolved. :( Repository: rG L

[PATCH] D86097: [OpenMP][AMDGCN] Generate global variables and attributes for AMDGCN

ABataev added inline comments. Comment at: clang/lib/CodeGen/CGOpenMPRuntime.h:684 const OffloadDeviceGlobalVarEntryInfoActTy &Action); - private: Restore original formatting Comment at: clang/lib/CodeGen/CGOpenMPRuntime.h:2482-2487

[PATCH] D86562: [OPENMP][NFC]Release notes for OpenMP in clang (11.x).

ABataev added a comment. In D86562#2274595 , @hans wrote: > In D86562#2274039 , @ABataev wrote: > >> In D86562#2274036 , @hans wrote: >> >>> Please go ahead and commit. >> >>

[PATCH] D86558: [OPENMP]Add support for allocate vars in untied tasks.

This revision was automatically updated to reflect the committed changes. Closed by commit rG738bab743b5c: [OPENMP]Add support for allocate vars in untied tasks. (authored by ABataev). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D86558/new/ https:/

[PATCH] D84887: [OPENMP]Fix codegen for is_device_ptr component, captured by reference.

This revision was automatically updated to reflect the committed changes. Closed by commit rG9e3842d60351: [OPENMP]Fix codegen for is_device_ptr component, captured by reference. (authored by ABataev). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D84

[PATCH] D85762: [OPENMP]Do not allow threadprivates as base for array-like reduction.

This revision was automatically updated to reflect the committed changes. Closed by commit rG4341c6618dec: [OPENMP]Do not allow threadprivates as base for array-like reduction. (authored by ABataev). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D8576

[PATCH] D84470: [OpenMP 5.0] Fix user-defined mapper privatization in tasks

ABataev updated this revision to Diff 292330. ABataev added a comment. Herald added a project: clang. Herald added a subscriber: cfe-commits. Updated + added clang test. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D84470/new/ https://reviews.llvm.

[PATCH] D84470: [OpenMP 5.0] Fix user-defined mapper privatization in tasks

ABataev added a comment. In D84470#2277991 , @ronlieb wrote: > The latest patch applied cleanly to our downstream port. > builds fine, tests very nicely as well. All the failing SOLLVE task wait > depend tests now pass. Could you accept it then? Reposi

[PATCH] D84470: [OpenMP 5.0] Fix user-defined mapper privatization in tasks

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGd5ce8233bfcf: [OpenMP 5.0] Fix user-defined mapper privatization in tasks (authored by ABataev). Repository: rG LLVM Github Monorepo CHANGES SINC

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

ABataev added a comment. In D76342#2262414 , @Meinersbur wrote: > Rebase after D83261 I will try to adapt it to the new interface (it won't work for `simd`, still need to capture the variables) and will send the patch ba

[PATCH] D88179: [OPENMP]PR47606: Do not update the lastprivate item if it was captured by reference as firstprivate data member.

ABataev created this revision. ABataev added reviewers: jdoerfert, mikerice. Herald added subscribers: guansong, yaxunl. Herald added a project: clang. ABataev requested review of this revision. No need to make final copy from the firsptrivate/lastprivate copy to the original item if the item is

[PATCH] D88240: [OPENMP]Fix PR47621: Variable used by task inside a template function is not made firstprivate by default

ABataev created this revision. ABataev added a reviewer: jdoerfert. Herald added subscribers: guansong, yaxunl. Herald added a project: clang. ABataev requested review of this revision. Herald added a subscriber: sstefan1. Need to fix a check for the variable if it is declared in the inner OpenMP

[PATCH] D88179: [OPENMP]PR47606: Do not update the lastprivate item if it was captured by reference as firstprivate data member.

This revision was automatically updated to reflect the committed changes. Closed by commit rGa9fca98ee4f6: [OPENMP]PR47606: Do not update the lastprivate item if it was captured by… (authored by ABataev). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/

[PATCH] D88240: [OPENMP]Fix PR47621: Variable used by task inside a template function is not made firstprivate by default

This revision was automatically updated to reflect the committed changes. Closed by commit rGd1419c9fdab1: [OPENMP]Fix PR47621: Variable used by task inside a template function is not… (authored by ABataev). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.o

[PATCH] D86239: [OPENMP]Fix PR47158, case 3: allow devic_typein nested declare target region.

ABataev created this revision. ABataev added reviewers: jdoerfert, cchen. Herald added subscribers: guansong, yaxunl. Herald added a reviewer: aaron.ballman. Herald added a project: clang. ABataev requested review of this revision. Herald added a subscriber: sstefan1. OpenMP 5.0 supports nested de

[PATCH] D86164: [OPENMP]Fix PR47158, case 2: do not report host-only functions in unused function in device mode.

This revision was automatically updated to reflect the committed changes. Closed by commit rGfb4acd37fe60: [OPENMP]Fix PR47158, case 2: do not report host-only functions in unused… (authored by ABataev). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D

[PATCH] D84887: [OPENMP]Fix codegen for is_device_ptr component, captured by reference.

ABataev updated this revision to Diff 286651. ABataev added a comment. Rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D84887/new/ https://reviews.llvm.org/D84887 Files: clang/lib/CodeGen/CGOpenMPRuntime.cpp clang/test/OpenMP/target_is_dev

[PATCH] D85762: [OPENMP]Do not allow threadprivates as base for array-like reduction.

ABataev updated this revision to Diff 286658. ABataev added a comment. Rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D85762/new/ https://reviews.llvm.org/D85762 Files: clang/lib/Sema/SemaOpenMP.cpp clang/test/OpenMP/parallel_reduction_me

[PATCH] D84887: [OPENMP]Fix codegen for is_device_ptr component, captured by reference.

ABataev updated this revision to Diff 287053. ABataev added a comment. Rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D84887/new/ https://reviews.llvm.org/D84887 Files: clang/lib/CodeGen/CGOpenMPRuntime.cpp clang/test/OpenMP/target_is_dev

[PATCH] D86239: [OPENMP]Fix PR47158, case 3: allow devic_typein nested declare target region.

This revision was automatically updated to reflect the committed changes. Closed by commit rGbedc841a5098: [OPENMP]Fix PR47158, case 3: allow devic_typein nested declare target region. (authored by ABataev). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.o

[PATCH] D84887: [OPENMP]Fix codegen for is_device_ptr component, captured by reference.

ABataev updated this revision to Diff 287396. ABataev added a comment. Rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D84887/new/ https://reviews.llvm.org/D84887 Files: clang/lib/CodeGen/CGOpenMPRuntime.cpp clang/test/OpenMP/target_is_dev

[PATCH] D85762: [OPENMP]Do not allow threadprivates as base for array-like reduction.

ABataev updated this revision to Diff 287406. ABataev added a comment. Rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D85762/new/ https://reviews.llvm.org/D85762 Files: clang/lib/Sema/SemaOpenMP.cpp clang/test/OpenMP/parallel_reduction_me

[PATCH] D86097: [OpenMP][AMDGCN] Generate global variables and attributes for AMDGCN

ABataev added a comment. Reformat the code Comment at: clang/lib/CodeGen/CGOpenMPRuntime.h:498 /// far. + class OffloadEntriesInfoManagerTy { Remove unnecessary formatting changes. Comment at: clang/lib/CodeGen/CGOpenMPRuntime.h:2479-2

[PATCH] D86558: [OPENMP]Add support for allocate vars in untied tasks.

ABataev created this revision. ABataev added a reviewer: jdoerfert. Herald added subscribers: guansong, yaxunl. Herald added a project: clang. ABataev requested review of this revision. Herald added a subscriber: sstefan1. Local vars, marked with pragma allocate, mustbe allocate by the call of the

[PATCH] D86562: [OPENMP][NFC]Release notes for OpenMP in clang (11.x).

ABataev created this revision. ABataev added reviewers: jdoerfert, kkwli0, RaviNarayanaswamy, hfinkel. Herald added subscribers: jfb, guansong, yaxunl. Herald added a project: clang. ABataev requested review of this revision. Herald added a subscriber: sstefan1. Repository: rG LLVM Github Monore

[PATCH] D86562: [OPENMP][NFC]Release notes for OpenMP in clang (11.x).

ABataev updated this revision to Diff 287776. ABataev added a comment. + complex types for NVPTX Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D86562/new/ https://reviews.llvm.org/D86562 Files: clang/docs/ReleaseNotes.rst Index: clang/docs/Rele

[PATCH] D86562: [OPENMP][NFC]Release notes for OpenMP in clang (11.x).

ABataev added a comment. In D86562#2237215 , @jdoerfert wrote: > Nit: Maybe we should mention `complex` support on nvptx devices. Ok, will add. > The rest looks good to me. Others might want to add stuff That was the intention. > but we can merge this

[PATCH] D86562: [OPENMP][NFC]Release notes for OpenMP in clang (11.x).

ABataev updated this revision to Diff 28. ABataev added a comment. Misprint fix Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D86562/new/ https://reviews.llvm.org/D86562 Files: clang/docs/ReleaseNotes.rst Index: clang/docs/ReleaseNotes.rst

[PATCH] D86562: [OPENMP][NFC]Release notes for OpenMP in clang (11.x).

ABataev added a comment. In D86562#2237388 , @kkwli0 wrote: > `present` map type and motion modifier > `default(firstprivate)` Present was added after 11.0 branching. Not aure about default(firstprivate) Repository: rG LLVM Github Monorepo CHANGES SI

[PATCH] D86562: [OPENMP][NFC]Release notes for OpenMP in clang (11.x).

ABataev added a comment. In D86562#2237399 , @kkwli0 wrote: > `default(firstprivate)` was added in > https://reviews.llvm.org/rG78443666bc18a6957d279a0f58319c8a3e57771a Yest, it was before clang 11 was branched, so need to add to the list of the changes

[PATCH] D85735: [OpenMP] Context selector extensions for template functions

ABataev added inline comments. Comment at: clang/include/clang/Sema/Sema.h:10042 void ActOnFinishedFunctionDefinitionInOpenMPDeclareVariantScope( - FunctionDecl *FD, FunctionDecl *BaseFD); + Decl *D, SmallVectorImpl &Bases); JonChesterfield wrote:

[PATCH] D86562: [OPENMP][NFC]Release notes for OpenMP in clang (11.x).

ABataev updated this revision to Diff 287906. ABataev added a comment. Added default(firstprivate) Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D86562/new/ https://reviews.llvm.org/D86562 Files: clang/docs/ReleaseNotes.rst Index: clang/docs/Re

[PATCH] D86562: [OPENMP][NFC]Release notes for OpenMP in clang (11.x).

ABataev added a comment. In D86562#2239136 , @hans wrote: >>> but we can merge this right away (IMHO) >> >> It is up to Hans if he has some time. But I think I covered almost all new >> features (+ complex for NVPTX), so after the update it can be merged

[PATCH] D85150: [OpenMP] Ensure testing for versions 4.5 and default - Part 2

ABataev accepted this revision. ABataev 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/D85150/new/ https://reviews.llvm.org/D85150 ___ c

[PATCH] D85214: [OpenMP] Ensure testing for versions 4.5 and default - Part 3

ABataev accepted this revision. ABataev 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/D85214/new/ https://reviews.llvm.org/D85214 ___ c

[PATCH] D84887: [OPENMP]Fix codegen for is_device_ptr component, captured by reference.

ABataev updated this revision to Diff 288700. ABataev added a comment. Herald added a subscriber: danielkiss. Rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D84887/new/ https://reviews.llvm.org/D84887 Files: clang/lib/CodeGen/CGOpenMPRuntim

[PATCH] D85762: [OPENMP]Do not allow threadprivates as base for array-like reduction.

ABataev updated this revision to Diff 288701. ABataev added a comment. Herald added a subscriber: danielkiss. Rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D85762/new/ https://reviews.llvm.org/D85762 Files: clang/lib/Sema/SemaOpenMP.cpp

[PATCH] D86558: [OPENMP]Add support for allocate vars in untied tasks.

ABataev updated this revision to Diff 288711. ABataev added a comment. Herald added a subscriber: danielkiss. Rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D86558/new/ https://reviews.llvm.org/D86558 Files: clang/lib/CodeGen/CGOpenMPRuntim

[PATCH] D86858: [OpenMP] Fix infinite loop in Sema::isOpenMPGlobalCapturedDecl()

ABataev accepted this revision. ABataev 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/D86858/new/ https://reviews.llvm.org/D86858 ___ c

[PATCH] D86858: [OpenMP] Fix infinite loop in Sema::isOpenMPGlobalCapturedDecl()

This revision was automatically updated to reflect the committed changes. Closed by commit rG2114f71aaa8d: [OpenMP] Fix infinite loop in Sema::isOpenMPGlobalCapturedDecl() (authored by nullptr.cpp, committed by ABataev). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://

[PATCH] D92427: [OPENMP51] Add present modifier in defaultmap clause

ABataev added inline comments. Comment at: clang/test/OpenMP/target_defaultmap_codegen_01.cpp:1523 -// CK26-DAG: [[SIZES:@.+]] = {{.+}}constant [3 x i64] [i64 4096, i64 4, i64 {{.+}}] +// CK26-DAG: [[SIZES:@.+]] = {{.+}}constant [3 x i64] [i64 4, i64 4096, i64 {{.+}}] // Map

[PATCH] D91373: [OpenMP5.0] Support more kinds of lvalues in map clauses

ABataev added a comment. Restore original formatting in test cases that were not directly affected by the patch. Also, I would start with a single kind of expression rather than trying to cover as many kinds of expressions as possible. It makes it easier to understand and to review it. =

[PATCH] D92427: [OPENMP51] Add present modifier in defaultmap clause

ABataev accepted this revision. ABataev 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/D92427/new/ https://reviews.llvm.org/D92427 ___ c

[PATCH] D93068: [clang-offload-bundler] Add option -allow-missing-bundles

ABataev added a comment. In D93068#2457946 , @yaxunl wrote: > @ABataev Is this patch OK for OpenMP? It is NFC for OpenMP toolchain but > affects using clang-offload-bundler as a standalone tool. Thanks. Yes, I think it is ok. CHANGES SINCE LAST ACTION

[PATCH] D93525: [OpenMP] Add unbundling of archives containing bundled object files into device specific archives

ABataev added inline comments. Comment at: clang/tools/clang-offload-bundler/ClangOffloadBundler.cpp:142-144 + } else { +return Triple; + } No need `else` here Comment at: clang/tools/clang-offload-bundler/ClangOffloadBundler.cpp:953-957

[PATCH] D91944: OpenMP 5.0 metadirective

ABataev added inline comments. Comment at: clang/lib/Sema/TreeTransform.h:8383-8388 + DeclarationNameInfo DirName; + getDerived().getSema().StartOpenMPDSABlock(OMPD_metadirective, DirName, + nullptr, D->getBeginLoc()); + StmtResult R

[PATCH] D90704: [OpenMP] target nested `use_device_ptr() if()` and is_device_ptr trigger asserts

ABataev added a comment. It would be good if you could identify the object which leads to a crash, I mean a target region, variable, etc. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D90704/new/ https://reviews.llvm.org/D90704 ___

[PATCH] D90704: [OpenMP] target nested `use_device_ptr() if()` and is_device_ptr trigger asserts

ABataev added inline comments. Comment at: clang/lib/CodeGen/CGOpenMPRuntime.cpp:2958-2970 +auto PerDevice = OffloadEntriesTargetRegion.find(DeviceID); +if (PerDevice != OffloadEntriesTargetRegion.end()) { + auto PerFile = PerDevice->second.find(FileID); + if (P

[PATCH] D90704: [OpenMP] target nested `use_device_ptr() if()` and is_device_ptr trigger asserts

ABataev added inline comments. Comment at: clang/lib/CodeGen/CGOpenMPRuntime.cpp:2958-2970 +auto PerDevice = OffloadEntriesTargetRegion.find(DeviceID); +if (PerDevice != OffloadEntriesTargetRegion.end()) { + auto PerFile = PerDevice->second.find(FileID); + if (P

[PATCH] D90704: [OpenMP] target nested `use_device_ptr() if()` and is_device_ptr trigger asserts

ABataev added inline comments. Comment at: clang/lib/CodeGen/CGOpenMPRuntime.cpp:2958-2970 +auto PerDevice = OffloadEntriesTargetRegion.find(DeviceID); +if (PerDevice != OffloadEntriesTargetRegion.end()) { + auto PerFile = PerDevice->second.find(FileID); + if (P

[PATCH] D90704: [OpenMP] target nested `use_device_ptr() if()` and is_device_ptr trigger asserts

ABataev added inline comments. Comment at: clang/lib/CodeGen/CGOpenMPRuntime.h:617 + StringRef ParentName, unsigned LineNum, + bool Registered = true) const; /// brief Applies action \a Action on all registere

[PATCH] D90704: [OpenMP] target nested `use_device_ptr() if()` and is_device_ptr trigger asserts

ABataev accepted this revision. ABataev 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/D90704/new/ https://reviews.llvm.org/D90704 ___ c

[PATCH] D91261: [OPENMP]Do not use OMP_MAP_TARGET_PARAM for data movement directives.

ABataev created this revision. ABataev added a reviewer: jdoerfert. Herald added subscribers: guansong, yaxunl. Herald added a project: clang. ABataev requested review of this revision. OMP_MAP_TARGET_PARAM flag is used to mark the data that shoud be passed as arguments to the target kernels, noth

[PATCH] D91370: [OPENMP]Fix PR48076: Check map types array before accessing its front.

ABataev created this revision. ABataev added a reviewer: jdoerfert. Herald added subscribers: guansong, yaxunl. Herald added a project: clang. ABataev requested review of this revision. Herald added a subscriber: sstefan1. Need to check if there are map types for the components before trying to ac

[PATCH] D91370: [OPENMP]Fix PR48076: Check map types array before accessing its front.

ABataev added a comment. In D91370#2392107 , @jhuber6 wrote: > This stops it from crashing Clang, but I'm not sure if it's fixing the > underlying problem. When I compile and run this program it crashes in > libomptarget. If you get rid of the `float f[5

[PATCH] D91370: [OPENMP]Fix PR48076: Check map types array before accessing its front.

ABataev added a comment. In D91370#2392171 , @jhuber6 wrote: > LGTM then. > > Any clue what the cause of the other problem is? Will investigate it and prepare a new patch Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews

[PATCH] D91376: [OPENMP]Fix PR47790: segfault in frontend while parsing Objective-C with OpenMP.

ABataev created this revision. ABataev added a reviewer: jdoerfert. Herald added subscribers: guansong, yaxunl. Herald added a project: clang. ABataev requested review of this revision. Herald added a subscriber: sstefan1. Need to check if the sema is actually finishing a function decl. Reposito

[PATCH] D91370: [OPENMP]Fix PR48076: Check map types array before accessing its front.

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG3c6b457bee0c: [OPENMP]Fix PR48076: Check map types array before accessing its front. (authored by ABataev). Repository: rG LLVM Github Monorepo C

[PATCH] D91370: [OPENMP]Fix PR48076: Check map types array before accessing its front.

ABataev added a comment. In D91370#2392171 , @jhuber6 wrote: > LGTM then. > > Any clue what the cause of the other problem is? Could you try your test with the patch from https://reviews.llvm.org/D86119? Repository: rG LLVM Github Monorepo CHANGES SI

[PATCH] D91370: [OPENMP]Fix PR48076: Check map types array before accessing its front.

ABataev added a comment. In D91370#2392310 , @jhuber6 wrote: > In D91370#2392267 , @ABataev wrote: > >> In D91370#2392171 , @jhuber6 wrote: >> >>> LGTM then. >>> >>> Any clue

[PATCH] D91370: [OPENMP]Fix PR48076: Check map types array before accessing its front.

ABataev added a comment. In D91370#2392317 , @jhuber6 wrote: > In D91370#2392313 , @ABataev wrote: > >> In D91370#2392310 , @jhuber6 wrote: >> >>> In D91370#2392267

[PATCH] D91376: [OPENMP]Fix PR47790: segfault in frontend while parsing Objective-C with OpenMP.

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG07b568a9c8db: [OPENMP]Fix PR47790: segfault in frontend while parsing Objective-C with OpenMP. (authored by ABataev). Repository: rG LLVM Github M

[PATCH] D91370: [OPENMP]Fix PR48076: Check map types array before accessing its front.

ABataev added a comment. In D91370#2392342 , @jhuber6 wrote: > In D91370#2392319 , @ABataev wrote: > >> From libomptarget. It must be different. Could you copy it here? > > Yeah, I think there's some extra informati

[PATCH] D91370: [OPENMP]Fix PR48076: Check map types array before accessing its front.

ABataev added a comment. In D91370#2392641 , @jhuber6 wrote: > In D91370#2392616 , @ABataev wrote: > >> Could you try this patch: >> >> diff --git a/clang/lib/CodeGen/CGOpenMPRuntime.cpp >> b/clang/lib/CodeGen/CG

[PATCH] D91627: [OPENMP] Fix PR47999: correctly map implicit firstprivates in outer tasks.

ABataev created this revision. ABataev added a reviewer: jdoerfert. Herald added subscribers: guansong, yaxunl. Herald added a project: clang. ABataev requested review of this revision. Herald added a subscriber: sstefan1. If the variable is implicitly firstprivatized in the inner task-based regio

[PATCH] D91644: [OPENMP]Honor constantness of captured variables.

ABataev created this revision. ABataev added a reviewer: jdoerfert. Herald added subscribers: guansong, yaxunl. Herald added a project: clang. ABataev requested review of this revision. Herald added a subscriber: sstefan1. Fixes bug reported via Stackoverflow: https://stackoverflow.com/questions/6

[PATCH] D91627: [OPENMP] Fix PR47999: correctly map implicit firstprivates in outer tasks.

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG0333567c47c0: [OPENMP] Fix PR47999: correctly map implicit firstprivates in outer tasks. (authored by ABataev). Repository: rG LLVM Github Monorep

[PATCH] D86119: [OPENMP50]Allow overlapping mapping in target constrcuts.

ABataev added inline comments. Comment at: openmp/libomptarget/src/omptarget.cpp:233 MapperComponents -.Components[target_data_function == targetDataEnd ? I : E - I - 1]; +.Components[target_data_function == targetDataEnd ? E - I - 1 : I]; Ma

[PATCH] D91660: [OPENMP]Fix PR48174: compile-time crash with target enter data on a global struct.

ABataev created this revision. ABataev added a reviewer: jdoerfert. Herald added subscribers: guansong, yaxunl. Herald added a project: clang. ABataev requested review of this revision. Herald added a subscriber: sstefan1. The compiler should treat array subscript with base pointer as a first poin

<    6   7   8   9   10   11   12   13   14   15   >