[flang] [llvm] [clang] [mlir] [libc] [flang] Fold MATMUL() (PR #72176)

https://github.com/clementval approved this pull request. LGTM https://github.com/llvm/llvm-project/pull/72176 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [llvm] [clang-tools-extra] [flang] [flang][openacc/mp][NFC] Remove unused baseAddr argument (PR #75537)

https://github.com/clementval updated https://github.com/llvm/llvm-project/pull/75537 >From ae0e461d8c9c013d7d3c2d16d811df6b30e66abd Mon Sep 17 00:00:00 2001 From: Valentin Clement Date: Thu, 14 Dec 2023 13:58:33 -0800 Subject: [PATCH] [flang][openacc/mp][NFC] Remove unused baseAddr argument -

[clang] [flang] [clang-tools-extra] [llvm] [flang][openacc/mp][NFC] Remove unused baseAddr argument (PR #75537)

https://github.com/clementval closed https://github.com/llvm/llvm-project/pull/75537 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[lldb] [llvm] [libcxx] [mlir] [flang] [libc] [libcxxabi] [clang-tools-extra] [clang] [compiler-rt] [mlir][flang][openacc] Support device_type on loop construct (PR #76892)

https://github.com/clementval updated https://github.com/llvm/llvm-project/pull/76892 >From ee784de88ed77d406eefe9f4ea65e823bde53e2e Mon Sep 17 00:00:00 2001 From: Valentin Clement Date: Tue, 2 Jan 2024 16:08:02 -0800 Subject: [PATCH 1/2] [mlir][flang][openacc] Support device_type on loop cons

[clang] [flang] [mlir] [libc] [compiler-rt] [lldb] [clang-tools-extra] [libcxxabi] [libcxx] [llvm] [mlir][flang][openacc] Support device_type on loop construct (PR #76892)

https://github.com/clementval closed https://github.com/llvm/llvm-project/pull/76892 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [flang][cuda] Enable cuda with -x cuda option (PR #84944)

https://github.com/clementval edited https://github.com/llvm/llvm-project/pull/84944 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [flang][cuda] Enable cuda with -x cuda option (PR #84944)

@@ -0,0 +1,13 @@ +! Test -fcuda option +! RUN: %flang -fc1 -cpp -fcuda -fdebug-unparse %s -o - | FileCheck %s clementval wrote: Added https://github.com/llvm/llvm-project/pull/84944 ___ cfe-commits mailing list cfe-com

[clang] [flang] [flang][cuda] Enable cuda with -x cuda option (PR #84944)

https://github.com/clementval closed https://github.com/llvm/llvm-project/pull/84944 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [flang][cuda] Add -fcuda option (PR #84944)

https://github.com/clementval created https://github.com/llvm/llvm-project/pull/84944 Flang driver was already able to enable the CUDA language feature base on the file extension but there was no command line option. This PR adds one. >From e42e8fe7f1dfe503a6735ef76e9d6483f5c9b5ec Mon Sep 17

[clang] [flang] [flang][cuda] Add -fcuda option (PR #84944)

@@ -0,0 +1,13 @@ +! Test -fcuda option +! RUN: %flang -fc1 -cpp -fcuda -fdebug-unparse %s -o - | FileCheck %s clementval wrote: Without it, it would just fail during parsing. Do you want a test that check the failure? https://github.com/llvm/llvm-project/pull/8

[clang] [flang] [flang][cuda] Add -fcuda option (PR #84944)

https://github.com/clementval edited https://github.com/llvm/llvm-project/pull/84944 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [flang][cuda] Add -fcuda option (PR #84944)

@@ -6488,6 +6488,9 @@ defm stack_arrays : BoolOptionWithoutMarshalling<"f", "stack-arrays", defm loop_versioning : BoolOptionWithoutMarshalling<"f", "version-loops-for-stride", PosFlag, NegFlag>; + +def fcuda : Flag<["-"], "fcuda">, Group, clementval wro

[clang] [flang] [flang][cuda] Add -fcuda option (PR #84944)

https://github.com/clementval updated https://github.com/llvm/llvm-project/pull/84944 >From e42e8fe7f1dfe503a6735ef76e9d6483f5c9b5ec Mon Sep 17 00:00:00 2001 From: Valentin Clement Date: Tue, 12 Mar 2024 09:33:17 -0700 Subject: [PATCH 1/2] [flang][cuda] Add -fcuda option --- clang/include/cla

[clang] [flang] [flang][cuda] Add -fcuda option (PR #84944)

https://github.com/clementval deleted https://github.com/llvm/llvm-project/pull/84944 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [flang][cuda] Add -fcuda option (PR #84944)

@@ -6488,6 +6488,9 @@ defm stack_arrays : BoolOptionWithoutMarshalling<"f", "stack-arrays", defm loop_versioning : BoolOptionWithoutMarshalling<"f", "version-loops-for-stride", PosFlag, NegFlag>; + +def fcuda : Flag<["-"], "fcuda">, Group, clementval wro

[clang] [flang] [flang] Enable polymorphic lowering by default (PR #83285)

https://github.com/clementval approved this pull request. Looks great! Thanks Jean to work on this! https://github.com/llvm/llvm-project/pull/83285 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/

[clang] [llvm] [Frontend] Add leaf constructs and association to OpenMP/ACC directives (PR #83625)

https://github.com/clementval approved this pull request. Looks good to me from the OpenACC point of view. Someone from OpenMP should approve as well. https://github.com/llvm/llvm-project/pull/83625 ___ cfe-commits mailing list cfe-commits@lists.llvm.

[flang] [clang] [flang]Add support for -moutline-atomics and -mno-outline-atomics (PR #78755)

clementval wrote: This is breaking couple of arm buildbots. https://github.com/llvm/llvm-project/pull/78755 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[flang] [clang] [flang]Add support for -moutline-atomics and -mno-outline-atomics (PR #78755)

clementval wrote: If you think this is the right fix then you should push it to solve the buildbots or revert the initial commit. https://github.com/llvm/llvm-project/pull/78755 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llv

[clang] [flang][openacc] Support assumed shape array in firstprivate recipe (PR #68640)

https://github.com/clementval updated https://github.com/llvm/llvm-project/pull/68640 >From 8459381299602a80b26ea44b5d099f59b0169ba3 Mon Sep 17 00:00:00 2001 From: Valentin Clement Date: Mon, 9 Oct 2023 15:00:12 -0700 Subject: [PATCH] [flang][openacc] Support assumed shape array in firstprivate

[clang] [flang] [flang][cuda][driver] Make sure flang does not switch to cc1 (PR #104613)

https://github.com/clementval created https://github.com/llvm/llvm-project/pull/104613 Flang is switch to cc1 when we use `-x cuda`. Make sure we can use fc1 with cuda fortran input. The current pipeline will fail at MLIR level for the moment. >From 36e18ae7bfa6fd2647da0b78bf2c332969b1091d M

[clang] [flang] [flang][cuda][driver] Make sure flang does not switch to cc1 (PR #104613)

https://github.com/clementval edited https://github.com/llvm/llvm-project/pull/104613 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [flang][cuda][driver] Make sure flang does not switch to cc1 (PR #104613)

https://github.com/clementval updated https://github.com/llvm/llvm-project/pull/104613 >From 36e18ae7bfa6fd2647da0b78bf2c332969b1091d Mon Sep 17 00:00:00 2001 From: Valentin Clement Date: Fri, 16 Aug 2024 09:23:29 -0700 Subject: [PATCH 1/2] [flang][cuda][driver] Make sure flang does not switch

[clang] [flang] [flang][Frontend] Implement printing defined macros via -dM (PR #87627)

clementval wrote: > How is this code used outside of LLVM? Why do people want to use _without_ > LLVM? Just curious. I can see language features in IDE like vscode being a good candidate. https://github.com/llvm/llvm-project/pull/87627 ___ cfe-commi

[clang] [Driver] Restrict Ofast deprecation help message to Clang (PR #101682)

clementval wrote: > Thanks for the clarification! Cc some flang developers to confirm this and > #101701 are the direction as > https://discourse.llvm.org/t/rfc-deprecate-ofast-in-flang/80243 doesn't have > lots of recorded discussion: @banach-space @clementval @klausler 👍 This was discussed

[clang] [llvm] [OpenMP] OpenMP 5.1 "assume" directive parsing support (PR #92731)

clementval wrote: This is still breaking a buildbot. https://lab.llvm.org/buildbot/#/builders/157/builds/4246 Are you working on a fix? https://github.com/llvm/llvm-project/pull/92731 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://li

[clang] [llvm] [OpenMP] OpenMP 5.1 "assume" directive parsing support (PR #92731)

clementval wrote: Yeah it was probably the case. It's green now. https://github.com/llvm/llvm-project/pull/92731 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [flang][cuda][driver] Make sure flang does not switch to cc1 (PR #104613)

https://github.com/clementval closed https://github.com/llvm/llvm-project/pull/104613 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] Revert "[flang][cuda][driver] Make sure flang does not switch to cc1" (PR #104632)

https://github.com/clementval created https://github.com/llvm/llvm-project/pull/104632 Reverts llvm/llvm-project#104613 >From cf8d13e27b647afaca66457d93b45a7da0e9cdfb Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Valentin=20Clement=20=28=E3=83=90=E3=83=AC=E3=83=B3?= =?UTF-8?q?=E3=82=BF=E3=82=A4=E3=

[clang] [flang] Revert "[flang][cuda][driver] Make sure flang does not switch to cc1" (PR #104632)

https://github.com/clementval closed https://github.com/llvm/llvm-project/pull/104632 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [llvm] [openmp] [flang][driver] rename flang-new to flang (PR #110023)

clementval wrote: > > Could you add an experimental message ("The openmp support in Flang is > > experimental") when compiling with OpenMP? The warning can be in the Driver > > code that forwards the `-fopenmp` flag to the driver. > > Hm. I'm not yet fully convinced that we should be doing thi

[clang] [flang] [lld] [llvm] [Flang][NFC] Move runtime library files to flang-rt. (PR #110298)

clementval wrote: > @clementval You agreed to a new folder structure in > https://discourse.llvm.org/t/rfc-names-for-flang-rt-libraries/84321/4. Could > you please review this PR again? > > The "Check code formatting" stage does not pass by design. For moved/renamed > files it requires a comp

[clang] [llvm] [clang][OpenMP] New OpenMP 6.0 assumption clause, 'no_openmp_constructs' (PR #125933)

clementval wrote: This breaks falling build bots. Can you fix or revert please? https://github.com/llvm/llvm-project/pull/125933 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [flang] [flang][OpenMP] Upstream first part of `do concurrent` mapping (PR #126026)

@@ -6910,6 +6910,10 @@ defm loop_versioning : BoolOptionWithoutMarshalling<"f", "version-loops-for-stri def fhermetic_module_files : Flag<["-"], "fhermetic-module-files">, Group, HelpText<"Emit hermetic module files (no nested USE association)">; + +def do_concurrent_paral

[clang] [flang] [flang][OpenMP] Upstream first part of `do concurrent` mapping (PR #126026)

@@ -142,6 +142,12 @@ static llvm::cl::opt llvm::cl::desc("enable openmp device compilation"), llvm::cl::init(false)); +static llvm::cl::opt enableDoConcurrentToOpenMPConversion( +"fdo-concurrent-parallel", clem

[clang] [flang] [flang][OpenMP] Upstream first part of `do concurrent` mapping (PR #126026)

@@ -6910,6 +6910,10 @@ defm loop_versioning : BoolOptionWithoutMarshalling<"f", "version-loops-for-stri def fhermetic_module_files : Flag<["-"], "fhermetic-module-files">, Group, HelpText<"Emit hermetic module files (no nested USE association)">; + +def do_concurrent_paral

[clang] [flang] [lld] [llvm] [Flang][NFC] Move runtime library files to flang-rt. (PR #110298)

clementval wrote: Ok. Thanks for the explanation. This look ok for me. I'll let @jeanPerier give the final approval since he has followed this work more closely. https://github.com/llvm/llvm-project/pull/110298 ___ cfe-commits mailing list cfe-commits

[clang] [flang] [lld] [llvm] [Flang][NFC] Move runtime library files to flang-rt. (PR #110298)

clementval wrote: There are some missing files in in flang-rt/CUDA both headers and cpp files. https://github.com/llvm/llvm-project/pull/110298 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-

[clang] [flang] [flang][OpenMP] Upstream `do concurrent` loop-nest detection. (PR #127595)

clementval wrote: > Thanks @skatrak and @bhandarkar-pranav for the approval. > > @kiranchandramohan @clementval I think there is a pretty simple solution that > enables us to mark multi-range loop nests. I think we can add an optional > attribute to the `fir::DoLoopOp` to store the loop nest d

[clang] [CIR] Initial implementation of lowering CIR to MLIR (PR #127835)

@@ -2958,6 +2958,8 @@ defm clangir : BoolFOption<"clangir", BothFlags<[], [ClangOption, CC1Option], "">>; def emit_cir : Flag<["-"], "emit-cir">, Visibility<[ClangOption, CC1Option]>, Group, HelpText<"Build ASTs and then lower to ClangIR">; +def emit_cir_mlir : Flag<["-"],

[clang] [CIR] Initial implementation of lowering CIR to MLIR (PR #127835)

@@ -2958,6 +2958,8 @@ defm clangir : BoolFOption<"clangir", BothFlags<[], [ClangOption, CC1Option], "">>; def emit_cir : Flag<["-"], "emit-cir">, Visibility<[ClangOption, CC1Option]>, Group, HelpText<"Build ASTs and then lower to ClangIR">; +def emit_cir_mlir : Flag<["-"],

[clang] [flang] [flang][OpenMP] Upstream `do concurrent` loop-nest detection. (PR #127595)

clementval wrote: > @clementval @jeanPerier can you please take a look at the PR and > @kiranchandramohan's comment above? 🙏 Yes, I think we should add a proper operations to represent the do concurrent. It would also allow use to carry the locality mapping into the IR representation and appl

[clang] [CIR] Initial implementation of lowering CIR to MLIR (PR #127835)

@@ -2958,6 +2958,8 @@ defm clangir : BoolFOption<"clangir", BothFlags<[], [ClangOption, CC1Option], "">>; def emit_cir : Flag<["-"], "emit-cir">, Visibility<[ClangOption, CC1Option]>, Group, HelpText<"Build ASTs and then lower to ClangIR">; +def emit_cir_mlir : Flag<["-"],

[clang] [CIR] Initial implementation of lowering CIR to MLIR (PR #127835)

@@ -2958,6 +2958,8 @@ defm clangir : BoolFOption<"clangir", BothFlags<[], [ClangOption, CC1Option], "">>; def emit_cir : Flag<["-"], "emit-cir">, Visibility<[ClangOption, CC1Option]>, Group, HelpText<"Build ASTs and then lower to ClangIR">; +def emit_cir_mlir : Flag<["-"],

[clang] [flang] [flang][OpenMP] Upstream `do concurrent` loop-nest detection. (PR #127595)

clementval wrote: I would prefer to have a new op. Other dialect also represent this with multiple operations (sci.for vs. sci.parallel). But I will let @jeanPerier weight in on this. I guess the op was not very necessary for your first few patches but representing the do concurrent accuratel

[clang] [OpenACC][CIR] Implement 'data' construct lowering (PR #135038)

https://github.com/clementval approved this pull request. LGTM from the OpenACC dialect point of view. https://github.com/llvm/llvm-project/pull/135038 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/list

[clang] [OpenACC][CIR] Implement 'device_type' clause lowering for 'init'/'sh… (PR #135102)

@@ -4,4 +4,17 @@ void acc_init(void) { // CHECK: cir.func @acc_init() { #pragma acc init // CHECK-NEXT: acc.init loc(#{{[a-zA-Z0-9]+}}){{$}} + +#pragma acc init device_type(*) clementval wrote: What happens if you have this? ``` #pragma acc init device_type

[clang] [OpenACC][CIR] Implement 'device_type' clause lowering for 'init'/'sh… (PR #135102)

https://github.com/clementval edited https://github.com/llvm/llvm-project/pull/135102 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [OpenACC][CIR] Implement 'device_type' clause lowering for 'init'/'sh… (PR #135102)

https://github.com/clementval approved this pull request. LGTM. https://github.com/llvm/llvm-project/pull/135102 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [OpenACC][CIR] Implement 'data' construct lowering, lower OACC->LLVMIR (PR #135038)

clementval wrote: > 2- This patch adds the infrastructure/calls to do the OpenACCDialect->LLVM-IR > lowering. Unfortunately only a handful of constructs are actually functional > in the OpenACC dialect, of which data is one (hence the choice to do it here, > and why I chose to do it as one pat

[clang] [OpenACC][CIR] Implement 'data' construct lowering (PR #135038)

clementval wrote: > > > 2- This patch adds the infrastructure/calls to do the > > > OpenACCDialect->LLVM-IR lowering. Unfortunately only a handful of > > > constructs are actually functional in the OpenACC dialect, of which data > > > is one (hence the choice to do it here, and why I chose to

[clang] [OpenACC] Switch Clang to use the Flang 'appertainment' rules for cla… (PR #135372)

clementval wrote: > Clang/Flang disagree whether 'seq' and 'self' can appear on the same > serial-loop construct. I see no prose to support this? `seq` and `self` are not in the same set for `serial loop` and seems to be accepted. https://godbolt.org/z/dzdoE3xKr https://github.com/llvm/llvm

[clang] [llvm] [OpenACC] Switch Clang to use the Flang 'appertainment' rules for cla… (PR #135372)

clementval wrote: `if_present` #135422 merged `finalize` #135415 merged https://github.com/llvm/llvm-project/pull/135372 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits