[PATCH] D122897: [clang][CUDA][Windows] Fix compilation error on Windows with `uint32_t __nvvm_get_smem_pointer`

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGc23147106f7e: [clang][CUDA][Windows] Fix compilation error on Windows with `uint32_t… (authored by emankov). Herald added a subscriber: cfe-commits.

[PATCH] D107492: [clang] Replace asm with __asm__ in cuda header

emankov added inline comments. Comment at: clang/lib/Headers/__clang_cuda_device_functions.h:1043 } -#else // CUDA_VERSION >= 9020 +#else // CUDA_VERSION >= 9020 // CUDA no longer provides inline assembly (or bitcode) implementation of these JonChesterfield wr

[PATCH] D107492: [clang] Replace asm with __asm__ in cuda header

emankov requested changes to this revision. emankov added inline comments. This revision now requires changes to proceed. Comment at: clang/lib/Headers/__clang_cuda_device_functions.h:1043 } -#else // CUDA_VERSION >= 9020 +#else // CUDA_VERSION >= 9020 // CUDA no longer provid

[PATCH] D89832: [CUDA] Extract CUDA version from cuda.h if version.txt is not found

emankov added a comment. D89832 eliminates 47332 on Windows as well. Tested against the same CUDA versions as for D89752 . Repository: rG LLVM Github Monorepo CHANGES SINCE LAST

[PATCH] D89832: [CUDA] Extract CUDA version from cuda.h if version.txt is not found

emankov added a comment. I confirm that D89752 eliminates 47332 on Windows. Tested against the following CUDA versions: 7.0, 7,5, 9,2, 10.0, 10,1, 10.2, 11.0 Update 1, 11.1. Repository: rG LLVM Github Monorepo C

[PATCH] D89832: [CUDA] Extract CUDA version from cuda.h if version.txt is not found

emankov accepted this revision. emankov added inline comments. This revision is now accepted and ready to land. Comment at: clang/lib/Driver/ToolChains/Cuda.cpp:76-77 +return CudaVersion::CUDA_102; + if (raw_version < 11010) +return CudaVersion::CUDA_110; + return CudaV

[PATCH] D89832: [CUDA] Extract CUDA version from cuda.h if version.txt is not found

emankov requested changes to this revision. emankov added inline comments. This revision now requires changes to proceed. Comment at: clang/lib/Driver/ToolChains/Cuda.cpp:76-77 +return CudaVersion::CUDA_102; + if (raw_version < 11010) +return CudaVersion::CUDA_110; + re

[PATCH] D89752: [CUDA] Improve clang's ability to detect recent CUDA versions.

emankov added inline comments. Comment at: clang/lib/Driver/ToolChains/Cuda.cpp:161 + if (FS.exists(LibDevicePath + "/libdevice.10.bc")) { +Version = CudaVersion::LATEST; +DetectedVersionIsNotSupported = Version > CudaVersion::LATEST_SUPPORTED; -

[PATCH] D89752: [CUDA] Improve clang's ability to detect recent CUDA versions.

emankov added inline comments. Comment at: clang/lib/Driver/ToolChains/Cuda.cpp:161 + if (FS.exists(LibDevicePath + "/libdevice.10.bc")) { +Version = CudaVersion::LATEST; +DetectedVersionIsNotSupported = Version > CudaVersion::LATEST_SUPPORTED; -

[PATCH] D89752: [CUDA] Improve clang's ability to detect recent CUDA versions.

emankov accepted this revision. emankov added a comment. This revision is now accepted and ready to land. Thank you! I've successfully applied your change in `Cuda.cpp` to the following clang versions: `10.0.0`, `10.0.1`, `11.0.0`, and `12.0.0git`; and have created the working patches (#206

[PATCH] D89752: [CUDA] Improve clang's ability to detect recent CUDA versions.

emankov added inline comments. Comment at: clang/lib/Driver/ToolChains/Cuda.cpp:158 if (!VersionFile) { - // CUDA 7.0 doesn't have a version.txt, so guess that's our version if - // version.txt isn't present. - Version = CudaVersion::CUDA_70; + // CUDA 7.

[PATCH] D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3)

emankov added a comment. In D60220#1725633 , @dim wrote: > $ cat check-isnan.cpp > #include > > int check_isnan(double d) > { > return ::__isnan(d); > } > $ clang -c check-isnan.cpp > Why can't the regular `isnan` be used instead? Or

[PATCH] D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3)

emankov added a comment. >> It looks like CUDA doesn't support `double` argument for device function >> __isnan on FreeBSD. > > It's actually the opposite -- FreeBSD does not provide *host*-side > `__isnan(double)` -- the error complains that it's the host code that tried > to use `__isnan` an

[PATCH] D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3)

emankov added a comment. In D60220#1723350 , @6yearold wrote: > I'm seeing quite similar errors on FreeBSD with Clang 8 and 9: > Any idea how to fix this? It looks like CUDA doesn't support `double` argument for device function __isnan on FreeBSD. 1.

[PATCH] D60818: [CUDA][Windows] restrict long double device functions declarations to Windows

This revision was automatically updated to reflect the committed changes. Closed by commit rC358654: [CUDA][Windows] Restrict long double device functions declarations to Windows (authored by emankov, committed by ). Changed prior to commit: https://reviews.llvm.org/D60818?vs=195546&id=195704#t

[PATCH] D60818: [CUDA][Windows] restrict long double functions declarations to Windows

emankov created this revision. emankov added a reviewer: tra. emankov added a project: clang. Herald added a subscriber: cfe-commits. As agreed in D60220 , make `long double` declarations unobservable on non-windows platforms. Repository: rC Clang https://reviews.l

[PATCH] D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3)

emankov added a comment. In D60220#1456449 , @tra wrote: > It's not a big deal at the moment -- there are no `long double` users in CUDA > on linux yet. You can clean up in another commit. > BTW, you may want to make commit description somewhat more conc

[PATCH] D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3)

emankov added a comment. Oooh, sorry, but I've just pushed the fix. But with the following words: "Add missing long double device functions' declarations. Provide only declarations to prevent any use of long double on the device side, because CUDA does not support long double on the device side

[PATCH] D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3)

This revision was automatically updated to reflect the committed changes. Closed by commit rC357779: [CUDA][Windows] Last fix for the clang Bug 38811 "Clang fails to compile with… (authored by emankov, committed by ). Changed prior to commit: https://reviews.llvm.org/D60220?vs=193874&id=193904#

[PATCH] D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3)

emankov updated this revision to Diff 193874. emankov added a comment. Provide only declarations for missing long double device functions to prevent any use of `long double` on the device side, because CUDA does not support `long double` on the device side. [Testing] {Windows 10, Ubuntu 16.04.5

[PATCH] D60220: [CUDA][Windows] Final fix for bug 38811 (Step 3 of 3)

emankov created this revision. emankov added a reviewer: tra. emankov added a project: clang. Herald added a subscriber: cfe-commits. Last fix for the clang Bug 38811 "Clang fails to compile with CUDA-9.x on Windows". **[IMPORTANT]** With that last f

[PATCH] D59423: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3)

This revision was automatically updated to reflect the committed changes. Closed by commit rC356291: [CUDA][Windows] Partial fix for bug 38811 (Step 2 of 3) (authored by emankov, committed by ). Herald added a subscriber: cfe-commits. Changed prior to commit: https://reviews.llvm.org/D59423?vs=

[PATCH] D59361: [CUDA][Windows] Partial fix for bug 38811 (Step 1 of 3)

This revision was automatically updated to reflect the committed changes. Closed by commit rC356255: [CUDA][Windows] Partial fix for bug #38811 (Step 1 of 3) (authored by emankov, committed by ). Changed prior to commit: https://reviews.llvm.org/D59361?vs=190607&id=190809#toc Repository: rC

[PATCH] D59361: [CUDA][Windows] Partial fix for bug 38811 (Step 1 of 3)

emankov created this revision. emankov added a reviewer: tra. Herald added subscribers: cfe-commits, jdoerfert. Herald added a project: clang. Partial fix for the clang Bug 38811 "Clang fails to compile with CUDA-9.x on Windows". Adding `defined(_W

[PATCH] D37386: [AMDGPU] Implement infrastructure to set options in AMDGPUToolChain

emankov accepted this revision. emankov added a comment. LGTM https://reviews.llvm.org/D37386 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D37386: [AMDGPU] Implement infrastructure to set options in AMDGPUToolChain

emankov added inline comments. Comment at: lib/Driver/ToolChains/AMDGPU.cpp:60-62 + for (auto *A : Args) { +DAL->append(A); + } redundant braces Comment at: lib/Driver/ToolChains/AMDGPU.cpp:74-75 +// as they defined that way in Option