[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

jhuber6 updated this revision to Diff 424449. jhuber6 added a comment. Moving some stuff. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D124220/new/ https://reviews.llvm.org/D124220 Files: clang/include/clang/Driver/Options.td clang/lib/Driver/

[PATCH] D124253: [Clang][OpenMP] Fix the issue that one temp cubin file is not removed after compilation

jhuber6 added a comment. Is this really the cause? `nvptx::assemble` should call `createOutputFile` which makes a temp file to output to that is added to `TempFiles`. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D124253/new/ https://reviews.llvm.o

[PATCH] D120272: [CUDA] Add driver support for compiling CUDA with the new driver

jhuber6 added a comment. Ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D120272/new/ https://reviews.llvm.org/D120272 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi

[PATCH] D120272: [CUDA] Add driver support for compiling CUDA with the new driver

jhuber6 added inline comments. Comment at: clang/lib/Driver/ToolChains/Clang.cpp:6223-6224 if (IsCuda || IsHIP) { -if (Args.hasFlag(options::OPT_fgpu_rdc, options::OPT_fno_gpu_rdc, false)) +if (Args.hasFlag(options::OPT_fgpu_rdc, options::OPT_fno_gpu_rdc, false) || +

[PATCH] D123812: [CUDA] Add wrapper code generation for registering CUDA images

jhuber6 updated this revision to Diff 424534. jhuber6 added a comment. Herald added a subscriber: MaskRay. Adding warning for using both `-fno-gpu-rdc` and `-foffload-new-driver`. I think this is a good warning to have for now while this is being worked in as opt-in. Once this has matured I plan

[PATCH] D123812: [CUDA] Add wrapper code generation for registering CUDA images

jhuber6 updated this revision to Diff 424536. jhuber6 added a comment. Applied changes to wrong commit, whoops. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123812/new/ https://reviews.llvm.org/D123812 Files: clang/test/Driver/linker-wrapper-im

[PATCH] D120272: [CUDA] Add driver support for compiling CUDA with the new driver

jhuber6 updated this revision to Diff 424537. jhuber6 added a comment. Adding warning for using both -fno-gpu-rdc and -foffload-new-driver. I think this is a good warning to have for now while this is being worked in as opt-in. Once this has matured I plan on adding the necessary logic to handle

[PATCH] D120272: [CUDA] Add driver support for compiling CUDA with the new driver

jhuber6 added inline comments. Comment at: clang/lib/Driver/ToolChains/Clang.cpp:6223-6224 if (IsCuda || IsHIP) { -if (Args.hasFlag(options::OPT_fgpu_rdc, options::OPT_fno_gpu_rdc, false)) +if (Args.hasFlag(options::OPT_fgpu_rdc, options::OPT_fno_gpu_rdc, false) || +

[PATCH] D120272: [CUDA] Add driver support for compiling CUDA with the new driver

jhuber6 added inline comments. Comment at: clang/lib/Driver/ToolChains/Clang.cpp:6225 if (IsCuda || IsHIP) { -if (Args.hasFlag(options::OPT_fgpu_rdc, options::OPT_fno_gpu_rdc, false)) +if (Args.hasArg(options::OPT_fno_gpu_rdc) && IsCudaDevice && +Args.hasArg(op

[PATCH] D120272: [CUDA] Add driver support for compiling CUDA with the new driver

jhuber6 updated this revision to Diff 424543. jhuber6 added a comment. Making suggested changes. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D120272/new/ https://reviews.llvm.org/D120272 Files: clang/include/clang/Basic/Cuda.h clang/include/c

[PATCH] D124292: [OpenMP] Use CUDA's non-RDC mode when LTO has whole program visibility

jhuber6 created this revision. jhuber6 added reviewers: jdoerfert, tianshilei1992, tra, JonChesterfield. Herald added subscribers: mattd, guansong, inglorion, yaxunl. Herald added a project: All. jhuber6 requested review of this revision. Herald added subscribers: cfe-commits, sstefan1. Herald adde

[PATCH] D124292: [OpenMP] Use CUDA's non-RDC mode when LTO has whole program visibility

jhuber6 updated this revision to Diff 424587. jhuber6 added a comment. Add test line Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D124292/new/ https://reviews.llvm.org/D124292 Files: clang/test/Driver/linker-wrapper.c clang/tools/clang-linker-

[PATCH] D124253: [Clang][OpenMP] Fix the issue that temp cubin files are not removed after compilation when using new OpenMP driver

jhuber6 accepted this revision. jhuber6 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/D124253/new/ https://reviews.llvm.org/D124253 ___

[PATCH] D124292: [OpenMP] Use CUDA's non-RDC mode when LTO has whole program visibility

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG3530c35c6609: [OpenMP] Use CUDA's non-RDC mode when LTO has whole program visibility (authored by jhuber6). Repository: rG LLVM Github Monorepo C

[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

jhuber6 updated this revision to Diff 425214. jhuber6 added a comment. Small change, returning an empty action type should let us stop without an extra check. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D124220/new/ https://reviews.llvm.org/D1242

[PATCH] D120272: [CUDA] Add driver support for compiling CUDA with the new driver

jhuber6 updated this revision to Diff 425261. jhuber6 added a comment. Changing this to simply require that the user manually passes `-fgpu-rdc` in order to use the new driver. I think this makes more sense in the short term and we can move to make the new driver the default rdc approach later.

[PATCH] D120272: [CUDA] Add driver support for compiling CUDA with the new driver

jhuber6 added inline comments. Comment at: clang/lib/Driver/ToolChains/Clang.cpp:6223-6224 if (IsCuda || IsHIP) { -if (Args.hasFlag(options::OPT_fgpu_rdc, options::OPT_fno_gpu_rdc, false)) +if (Args.hasFlag(options::OPT_fgpu_rdc, options::OPT_fno_gpu_rdc, false) || +

[PATCH] D123313: [OpenMP] Make clang argument handling for the new driver more generic

jhuber6 added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123313/new/ https://reviews.llvm.org/D123313 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi

[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

jhuber6 added a comment. ping Comment at: clang/include/clang/Driver/Options.td:2535-2538 +def foffload_device_only : Flag<["-"], "foffload-device-only">, + HelpText<"Only compile for the offloading device.">; +def foffload_host_only : Flag<["-"], "foffload-host-only">, + Hel

[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

jhuber6 added inline comments. Comment at: clang/include/clang/Driver/Options.td:2535-2538 +def foffload_device_only : Flag<["-"], "foffload-device-only">, + HelpText<"Only compile for the offloading device.">; +def foffload_host_only : Flag<["-"], "foffload-host-only">, + Help

[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

jhuber6 updated this revision to Diff 425301. jhuber6 added a comment. Adding `-fno` variants and replacing the uses of `--cuda-host-only` with and alias and this flag. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D124220/new/ https://reviews.llvm

[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

jhuber6 updated this revision to Diff 425336. jhuber6 added a comment. Updating to not use `-f/-fno` versions. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D124220/new/ https://reviews.llvm.org/D124220 Files: clang/include/clang/Driver/Options.t

[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

jhuber6 added inline comments. Comment at: clang/include/clang/Driver/Options.td:2527 HelpText<"Use the static host OpenMP runtime while linking.">; +def offload_device_only : Flag<["-"], "offload-device-only">, + HelpText<"Only compile for the offloading device.">; -

[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

jhuber6 updated this revision to Diff 425344. jhuber6 added a comment. Making suggested changes. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D124220/new/ https://reviews.llvm.org/D124220 Files: clang/include/clang/Driver/Options.td clang/lib/

[PATCH] D124525: [OpenMP][ClangLinkerWrapper] Extending linker wrapper to embed metadata for multi-arch fat binaries

jhuber6 added inline comments. Comment at: clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp:1161 -LinkedImages.push_back(*ImageOrErr); +LinkedImages.emplace_back(TheArch, *ImageOrErr); } I'm doing something similar in D123810, I just used the

[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

jhuber6 added inline comments. Comment at: clang/lib/Driver/Driver.cpp:4215 Action *HostAction) const { - if (!isa(HostAction)) + if (Args.hasArg(options::OPT_offload_host_only)) return HostAction; tra wrote: > jhube

[PATCH] D124525: [OpenMP][ClangLinkerWrapper] Extending linker wrapper to embed metadata for multi-arch fat binaries

jhuber6 added inline comments. Comment at: clang/tools/clang-linker-wrapper/OffloadWrapper.cpp:246 IRBuilder<> Builder(BasicBlock::Create(C, "entry", Func)); + // Create calls to __tgt_register_image_info for each image + auto *NullPtr = llvm::ConstantPointerNull::get(Builde

[PATCH] D123325: [Clang] Make enabling the new driver more generic

jhuber6 updated this revision to Diff 425785. jhuber6 added a comment. Ping and update. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123325/new/ https://reviews.llvm.org/D123325 Files: clang/include/clang/Driver/Options.td clang/lib/Driver/Dr

[PATCH] D120273: [OpenMP] Allow CUDA to be linked with OpenMP using the new driver

jhuber6 updated this revision to Diff 425807. jhuber6 added a comment. Ping and rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D120273/new/ https://reviews.llvm.org/D120273 Files: clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp In

[PATCH] D124624: [OpenMP] Add variant extension that applies to declarations

jhuber6 created this revision. jhuber6 added reviewers: jdoerfert, tianshilei1992. Herald added subscribers: guansong, yaxunl. Herald added a reviewer: aaron.ballman. Herald added a project: All. jhuber6 requested review of this revision. Herald added subscribers: llvm-commits, cfe-commits, sstefan

[PATCH] D123325: [Clang] Make enabling the new driver more generic

jhuber6 added inline comments. Comment at: clang/include/clang/Driver/Options.td:2535-2538 +def foffload_new_driver : Flag<["-"], "foffload-new-driver">, Flags<[CC1Option]>, Group, + HelpText<"Use the new driver for offloading compilation.">; +def fno_offload_new_driver : Flag<

[PATCH] D123325: [Clang] Make enabling the new driver more generic

jhuber6 added inline comments. Comment at: clang/include/clang/Driver/Options.td:2535-2538 +def foffload_new_driver : Flag<["-"], "foffload-new-driver">, Flags<[CC1Option]>, Group, + HelpText<"Use the new driver for offloading compilation.">; +def fno_offload_new_driver : Flag<

[PATCH] D123325: [Clang] Make enabling the new driver more generic

jhuber6 updated this revision to Diff 425879. jhuber6 added a comment. Changin `-f[no-]offload-new-driver` into `--[no-]offload-new-driver` Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123325/new/ https://reviews.llvm.org/D123325 Files: clang/i

[PATCH] D123460: [OpenMP] Make generating offloading entries more generic

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG643c9b22ef52: [OpenMP] Make generating offloading entries more generic (authored by jhuber6). Changed prior to commit: https://reviews.llvm.org/D1

[PATCH] D123313: [OpenMP] Make clang argument handling for the new driver more generic

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGca6bbe008512: [OpenMP] Make clang argument handling for the new driver more generic (authored by jhuber6). Repository: rG LLVM Github Monorepo CH

[PATCH] D123325: [Clang] Make enabling the new driver more generic

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG4e2b5a6693e2: [Clang] Make enabling the new driver more generic (authored by jhuber6). Changed prior to commit: https://reviews.llvm.org/D123325?v

[PATCH] D120272: [CUDA] Add driver support for compiling CUDA with the new driver

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGc5e5b54350fe: [CUDA] Add driver support for compiling CUDA with the new driver (authored by jhuber6). Changed prior to commit: https://reviews.llv

[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

jhuber6 updated this revision to Diff 426049. jhuber6 added a comment. Rebase after landing the cuda support for the new driver. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D124220/new/ https://reviews.llvm.org/D124220 Files: clang/include/clan

[PATCH] D120273: [OpenMP] Allow CUDA to be linked with OpenMP using the new driver

jhuber6 updated this revision to Diff 426050. jhuber6 added a comment. Adding test Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D120273/new/ https://reviews.llvm.org/D120273 Files: clang/test/Driver/linker-wrapper.c clang/tools/clang-linker-wr

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 updated this revision to Diff 426051. jhuber6 added a comment. Herald added a subscriber: mattd. Rebase Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123471/new/ https://reviews.llvm.org/D123471 Files: clang/include/clang/Basic/LangOptio

[PATCH] D124220: [OpenMP] Add options to only compile the host or device when offloading

This revision was automatically updated to reflect the committed changes. Closed by commit rG47d66255701a: [OpenMP] Add options to only compile the host or device when offloading (authored by jhuber6). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D12

[PATCH] D120273: [OpenMP] Allow CUDA to be linked with OpenMP using the new driver

This revision was automatically updated to reflect the committed changes. Closed by commit rGd9c64d33b98b: [OpenMP] Allow CUDA to be linked with OpenMP using the new driver (authored by jhuber6). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D120273/n

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 updated this revision to Diff 426150. jhuber6 added a comment. Fixed missing info flag for `--offload-new-driver`. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123471/new/ https://reviews.llvm.org/D123471 Files: clang/include/clang/Basi

[PATCH] D124721: [OpenMP] Allow compiling multiple target architectures with OpenMP

jhuber6 created this revision. jhuber6 added reviewers: jdoerfert, JonChesterfield, yaxunl, saiislam, tianshilei1992, tra. Herald added subscribers: kerbowa, guansong, jvesely. Herald added a project: All. jhuber6 requested review of this revision. Herald added subscribers: cfe-commits, sstefan1,

[PATCH] D124721: [OpenMP] Allow compiling multiple target architectures with OpenMP

jhuber6 added inline comments. Comment at: clang/test/Driver/amdgpu-openmp-toolchain-new.c:6 // RUN: | FileCheck %s +// RUN: %clang -### --target=x86_64-unknown-linux-gnu -fopenmp -fopenmp-targets=amdgcn-amd-amdhsa \ +// RUN: --offload-arch=gfx906 --libomptarget-a

[PATCH] D124721: [OpenMP] Allow compiling multiple target architectures with OpenMP

jhuber6 added inline comments. Comment at: clang/test/Driver/amdgpu-openmp-toolchain-new.c:6 // RUN: | FileCheck %s +// RUN: %clang -### --target=x86_64-unknown-linux-gnu -fopenmp -fopenmp-targets=amdgcn-amd-amdhsa \ +// RUN: --offload-arch=gfx906 --libomptarget-a

[PATCH] D124721: [OpenMP] Allow compiling multiple target architectures with OpenMP

jhuber6 added inline comments. Comment at: clang/test/Driver/amdgpu-openmp-toolchain-new.c:6 // RUN: | FileCheck %s +// RUN: %clang -### --target=x86_64-unknown-linux-gnu -fopenmp -fopenmp-targets=amdgcn-amd-amdhsa \ +// RUN: --offload-arch=gfx906 --libomptarget-a

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 updated this revision to Diff 426404. jhuber6 added a comment. Fix test. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123471/new/ https://reviews.llvm.org/D123471 Files: clang/include/clang/Basic/LangOptions.def clang/include/clang/Dr

[PATCH] D123810: [Cuda] Add initial support for wrapping CUDA images in the new driver.

jhuber6 updated this revision to Diff 426420. jhuber6 added a comment. Updating to use the `OffloadKind` enum rather than the string. I will also probably simplify some of the logic for handling multiple files here in a later patch. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACT

[PATCH] D123810: [Cuda] Add initial support for wrapping CUDA images in the new driver.

jhuber6 updated this revision to Diff 426437. jhuber6 added a comment. Fix test Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123810/new/ https://reviews.llvm.org/D123810 Files: clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp clang/too

[PATCH] D123812: [CUDA] Add wrapper code generation for registering CUDA images

jhuber6 updated this revision to Diff 426467. jhuber6 added a comment. Updating code generation. Previously we would seg-fault in the case that no offloading entries were created. To solve this we simply check that Begin != End before trying to register anything. Repository: rG LLVM Github M

[PATCH] D124721: [OpenMP] Allow compiling multiple target architectures with OpenMP

jhuber6 updated this revision to Diff 426521. jhuber6 added a comment. Changing slightly, I'm using the `getArgsForToolchain` to only get the `--offload-arch` options for that toolchain. This lets us quality it with options like `-Xopenmp-target=` so we can now specify architectures per-toolcha

[PATCH] D124721: [OpenMP] Allow compiling multiple target architectures with OpenMP

jhuber6 added inline comments. Comment at: clang/test/Driver/openmp-offload-gpu-new.c:56 + +// RUN: %clang -### --target=x86_64-unknown-linux-gnu -ccc-print-bindings -fopenmp \ +// RUN: -fopenmp-targets=nvptx64-nvidia-cuda,amdgcn-amd-amdhsa -Xopenmp-target=nvptx64-nvidia-

[PATCH] D124525: [OpenMP][ClangLinkerWrapper] Extending linker wrapper to embed metadata for multi-arch fat binaries

jhuber6 added a comment. I'm suggesting instead we define a new `__tgt_device_image` and a new `__tgt_register_lib` function to support this. This new `__tgt_device_image` will simply contain a pointer to an optional information struct. struct __tgt_device_image_v2 { void* ImageStar

[PATCH] D125050: [OpenMP] Try to Infer target triples using the offloading architecture

jhuber6 created this revision. jhuber6 added reviewers: jdoerfert, yaxunl, JonChesterfield, saiislam, ronlieb. Herald added a subscriber: guansong. Herald added a project: All. jhuber6 requested review of this revision. Herald added subscribers: cfe-commits, sstefan1, MaskRay. Herald added a projec

[PATCH] D124721: [OpenMP] Allow compiling multiple target architectures with OpenMP

jhuber6 added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D124721/new/ https://reviews.llvm.org/D124721 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi

[PATCH] D124721: [OpenMP] Allow compiling multiple target architectures with OpenMP

jhuber6 added inline comments. Comment at: clang/lib/Driver/ToolChains/AMDGPUOpenMP.cpp:313 + if (BoundArch.empty()) +checkSystemForAMDGPU(Args, *this, Arch); DAL->AddJoinedArg(nullptr, Opts.getOption(options::OPT_march_EQ), Arch); tra wrote:

[PATCH] D125050: [OpenMP] Try to Infer target triples using the offloading architecture

jhuber6 updated this revision to Diff 427496. jhuber6 added a comment. Address comments. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D125050/new/ https://reviews.llvm.org/D125050 Files: clang/include/clang/Basic/DiagnosticDriverKinds.td clang

[PATCH] D125050: [OpenMP] Try to Infer target triples using the offloading architecture

jhuber6 marked 3 inline comments as done. jhuber6 added inline comments. Comment at: clang/lib/Driver/Driver.cpp:788-789 + options::OPT_fno_openmp, false) && + (C.getInputArgs().hasArg(options::OPT_fopenmp_targets_EQ) || + C.getInputArgs()

[PATCH] D125050: [OpenMP] Try to Infer target triples using the offloading architecture

jhuber6 marked an inline comment as done. jhuber6 added a comment. In D125050#3496289 , @saiislam wrote: > Looks good to me. > > Will it work with `-fno-openmp`? Sometimes `-fno-openmp` is used by the > end-user to override system provided `-fopenmp` fla

[PATCH] D125050: [OpenMP] Try to Infer target triples using the offloading architecture

jhuber6 updated this revision to Diff 427599. jhuber6 added a comment. Add test for disabling openmp with `-fno-openmp` Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D125050/new/ https://reviews.llvm.org/D125050 Files: clang/include/clang/Basic/D

[PATCH] D123812: [CUDA] Add wrapper code generation for registering CUDA images

jhuber6 added a comment. ping Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123812/new/ https://reviews.llvm.org/D123812 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi

[PATCH] D125092: [OpenMP] Add basic support for properly handling static libraries

jhuber6 created this revision. jhuber6 added reviewers: jdoerfert, JonChesterfield, tianshilei1992. Herald added subscribers: guansong, yaxunl. Herald added a project: All. jhuber6 requested review of this revision. Herald added subscribers: cfe-commits, sstefan1. Herald added a project: clang. Cu

[PATCH] D125092: [OpenMP] Add basic support for properly handling static libraries

This revision was automatically updated to reflect the committed changes. Closed by commit rGe12905b4d5f9: [OpenMP] Add basic support for properly handling static libraries (authored by jhuber6). Changed prior to commit: https://reviews.llvm.org/D125092?vs=427618&id=427646#toc Repository: rG

[PATCH] D123812: [CUDA] Add wrapper code generation for registering CUDA images

jhuber6 added a comment. In D123812#3496914 , @yaxunl wrote: > LGTM. Did you forget to accept the revision? D123810 and D123471 still need to be looked at, but these are mostly non-

[PATCH] D125092: [OpenMP] Add basic support for properly handling static libraries

jhuber6 added a comment. In D125092#3497033 , @saiislam wrote: >> Ideally we could just put this on the linker itself, but nvlink doesn't seem >> to support .a files. > > Yes, nvlink does not support archives. So we used a wrapper to extract cubin > fil

[PATCH] D125050: [OpenMP] Try to Infer target triples using the offloading architecture

jhuber6 added inline comments. Comment at: clang/lib/Driver/Driver.cpp:789 + (C.getInputArgs().hasArg(options::OPT_fopenmp_targets_EQ) || + C.getInputArgs().hasArg(options::OPT_offload_arch_EQ)); + if (IsOpenMPOffloading) { yaxunl wrote: > probably sh

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 added a comment. In D123471#3497169 , @tra wrote: >> Type struct __tgt_offload_entry { >> >> void*addr; // Pointer to the offload entry info. >> // (function or global) >> char*name; // Name of the funct

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 added inline comments. Comment at: clang/lib/CodeGen/CGCUDARuntime.h:58-70 + OffloadRegionKernelEntry = 0x0, +}; + +/// The kind flag of the global variable entry. +enum OffloadVarEntryKindFlag : uint32_t { + /// Mark the entry as a global variable.

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 updated this revision to Diff 427701. jhuber6 added a comment. Changing enum values from a bitfield to simple enumeration. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123471/new/ https://reviews.llvm.org/D123471 Files: clang/include/cl

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 added inline comments. Comment at: clang/lib/CodeGen/CGCUDARuntime.h:56-70 +enum OffloadRegionEntryKindFlag : uint32_t { + /// Mark the region entry as a kernel. + OffloadRegionKernelEntry = 0x0, +}; + +/// The kind flag of the global variable entry.

[PATCH] D124721: [OpenMP] Allow compiling multiple target architectures with OpenMP

This revision was automatically updated to reflect the committed changes. Closed by commit rG8477a0d769a0: [OpenMP] Allow compiling multiple target architectures with OpenMP (authored by jhuber6). Changed prior to commit: https://reviews.llvm.org/D124721?vs=426521&id=427734#toc Repository: r

[PATCH] D125050: [OpenMP] Try to Infer target triples using the offloading architecture

This revision was automatically updated to reflect the committed changes. Closed by commit rG509b631f84e9: [OpenMP] Try to Infer target triples using the offloading architecture (authored by jhuber6). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D125

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 added inline comments. Comment at: clang/lib/CodeGen/CGCUDARuntime.h:56-70 +enum OffloadRegionEntryKindFlag : uint32_t { + /// Mark the region entry as a kernel. + OffloadRegionKernelEntry = 0x0, +}; + +/// The kind flag of the global variable entry.

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 updated this revision to Diff 427766. jhuber6 added a comment. removing enum Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D123471/new/ https://reviews.llvm.org/D123471 Files: clang/include/clang/Basic/LangOptions.def clang/include/clan

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 added inline comments. Comment at: clang/lib/CodeGen/CGCUDARuntime.h:58 + /// Mark the entry as a global variable. + OffloadGlobalEntry = 0x0, + /// Mark the entry as a managed global variable. tra wrote: > We're still using the same numeri

[PATCH] D123471: [CUDA] Create offloading entries when using the new driver

jhuber6 added inline comments. Comment at: clang/lib/CodeGen/CGCUDARuntime.h:58 + /// Mark the entry as a global variable. + OffloadGlobalEntry = 0x0, + /// Mark the entry as a managed global variable. tra wrote: > jhuber6 wrote: > > tra wrote: > >

[PATCH] D125165: [Clang] Introduce clang-offload-binary tool to bundle device files

jhuber6 created this revision. jhuber6 added reviewers: jdoerfert, JonChesterfield, saiislam, yaxunl, tra. Herald added subscribers: ormris, kerbowa, mgorny, jvesely. Herald added a project: All. jhuber6 requested review of this revision. Herald added subscribers: cfe-commits, sstefan1, MaskRay. He

[PATCH] D125165: [Clang] Introduce clang-offload-binary tool to bundle device files

jhuber6 updated this revision to Diff 427849. jhuber6 added a comment. Fix test. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D125165/new/ https://reviews.llvm.org/D125165 Files: clang/docs/ClangOffloadBinary.rst clang/include/clang/Basic/Code

[PATCH] D125165: [Clang] Introduce clang-offload-binary tool to bundle device files

jhuber6 updated this revision to Diff 427858. jhuber6 added a comment. Fix missing file in test. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D125165/new/ https://reviews.llvm.org/D125165 Files: clang/docs/ClangOffloadBinary.rst clang/include/

[PATCH] D129435: [Clang] Parse toolchain-specific offloading arguments directly

jhuber6 added a comment. In D129435#3642450 , @yaxunl wrote: > need a test There's a test `clang/test/Driver/openmp-offload-gpu-new.c` already making sure that the following works, so this patch should maintain the same functionality we had before.

[PATCH] D128914: [HIP] Add support for handling HIP in the linker wrapper

jhuber6 added a comment. In D128914#3642558 , @JonChesterfield wrote: > Code looks good to me. It's hard to be sure whether it works without running > a bunch of hip test cases through it, have you already done so? If it doesn't > work out of the box i

[PATCH] D128914: [HIP] Add support for handling HIP in the linker wrapper

jhuber6 added a comment. In D128914#3642869 , @yaxunl wrote: > In D128914#3642567 , @jhuber6 wrote: > >> In D128914#3642558 , >> @JonChesterfield wrote: >> >>> Code looks

[PATCH] D129435: [Clang] Parse toolchain-specific offloading arguments directly

This revision was automatically updated to reflect the committed changes. Closed by commit rG359e4a824731: [Clang] Parse toolchain-specific offloading arguments directly (authored by jhuber6). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D129435/new/

[PATCH] D128850: [HIP] Generate offloading entries for HIP with the new driver.

This revision was automatically updated to reflect the committed changes. Closed by commit rGe88d53d25f3b: [HIP] Generate offloading entries for HIP with the new driver. (authored by jhuber6). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D128850/new/

[PATCH] D128914: [HIP] Add support for handling HIP in the linker wrapper

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGce091eb3b91f: [HIP] Add support for handling HIP in the linker wrapper (authored by jhuber6). Changed prior to commit: https://reviews.llvm.org/D1

[PATCH] D128914: [HIP] Add support for handling HIP in the linker wrapper

jhuber6 added a comment. In D128914#3643451 , @yaxunl wrote: > If you only unregister fatbin once for the whole program, then it should be > safe -fgpu-rdc. I am not sure if that is the case. it should be here, the generated handle is private to the reg

[PATCH] D128914: [HIP] Add support for handling HIP in the linker wrapper

jhuber6 added a comment. In D128914#3643802 , @tra wrote: > For what it's worth, NCCL is the only > nontrivial library that needs RDC compilation that I'm aware of. > It's also self-contained for RDC purposes we only n

[PATCH] D127304: [LinkerWrapper] Embed OffloadBinaries for OpenMP offloading images

jhuber6 updated this revision to Diff 443810. jhuber6 added a comment. Removing `.image` from section name. This was originally done to avoid it getting the `SHF_EXCLUDE` flag when it should not be removed by the linker. That behaviour was moved to the `!exclude` metadata in a previous patch.

[PATCH] D128914: [HIP] Add support for handling HIP in the linker wrapper

jhuber6 added a comment. In D128914#3644022 , @thakis wrote: > This breaks check-clang on mac: http://45.33.8.238/macm1/39907/step_7.txt > > Please take a look and revert for now if it takes a while to fix. I changed some of the argument formats in a pre

[PATCH] D128914: [HIP] Add support for handling HIP in the linker wrapper

jhuber6 added a comment. In D128914#3644022 , @thakis wrote: > This breaks check-clang on mac: http://45.33.8.238/macm1/39907/step_7.txt > > Please take a look and revert for now if it takes a while to fix. Let me know if rGfe6a391357fc

[PATCH] D129534: [OpenMP] Do not link static library with `-nogpulib`

jhuber6 created this revision. jhuber6 added reviewers: jdoerfert, JonChesterfield, tianshilei1992, ABataev. Herald added subscribers: guansong, yaxunl. Herald added a project: All. jhuber6 requested review of this revision. Herald added subscribers: cfe-commits, sstefan1, MaskRay. Herald added a p

[PATCH] D129534: [OpenMP] Do not link static library with `-nogpulib`

This revision was automatically updated to reflect the committed changes. Closed by commit rGd214bfe78d8d: [OpenMP] Do not link static library with `-nogpulib` (authored by jhuber6). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D129534/new/ https://

[PATCH] D129581: [Clang] Rework LTO argument handling in the linker wrapper

jhuber6 created this revision. jhuber6 added reviewers: jdoerfert, JonChesterfield, MaskRay, tra, ABataev, yaxunl, tianshilei1992. Herald added subscribers: StephenFan, inglorion. Herald added a project: All. jhuber6 requested review of this revision. Herald added subscribers: cfe-commits, sstefan

[PATCH] D129586: [LinkerWrapper] Support remarks files for device LTO

jhuber6 created this revision. jhuber6 added reviewers: jdoerfert, tianshilei1992, JonChesterfield. Herald added subscribers: wenlei, inglorion. Herald added a project: All. jhuber6 requested review of this revision. Herald added a project: clang. Herald added a subscriber: cfe-commits. This patch

[PATCH] D128090: [Clang][OpenMP] Process multi-arch compilation options given via -march

jhuber6 added a comment. Sorry never noticed this revision. The purpose of this patch seems to be supporting something like this clang input.c -fopenmp -fopenmp-targets=nvptx64 -Xopenmp-target=nvptx64 -march=sm_70 -Xopenmp-target=nvptx64 -march=sm_80 Right now the above works if you replace

[PATCH] D129655: [CUDA] Allow the new driver to compile CUDA in non-RDC mode

jhuber6 created this revision. jhuber6 added reviewers: jdoerfert, tra, JonChesterfield, yaxunl. Herald added subscribers: mattd, carlosgalvezp. Herald added a project: All. jhuber6 requested review of this revision. Herald added subscribers: cfe-commits, MaskRay. Herald added a project: clang. Th

[PATCH] D128090: [Clang][OpenMP] Process multi-arch compilation options given via -march

jhuber6 added a subscriber: tra. jhuber6 added a comment. In D128090#3648879 , @saiislam wrote: > `-Xopenmp-target -march ` used to be the only option to target a specific sub > arch before `--offload-arch`. But, it doesn't support multiple archs. > This

[PATCH] D128090: [Clang][OpenMP] Process multi-arch compilation options given via -march

jhuber6 added a comment. In D128090#3648984 , @tra wrote: > At some point we should start consolidating the ways we can specify an > offload target and try to avoid adding new ones until then. Agreed, that was my intention with making `--offload-arch` w

[PATCH] D128090: [Clang][OpenMP] Process multi-arch compilation options given via -march

jhuber6 added a comment. In D128090#3649059 , @tra wrote: > In D128090#3648999 , @jhuber6 wrote: > >> Right now there's `CLANG_OPENMP_NVPTX_DEFAULT_ARCH`, which is defined by >> CMake to be the architecture of th

<    1   2   3   4   5   6   7   8   9   10   >