r319909 - [NVPTX, CUDA] Added llvm.nvvm.fns intrinsic and matching __nvvm_fns builtin in clang.

Author: tra Date: Wed Dec 6 09:50:05 2017 New Revision: 319909 URL: http://llvm.org/viewvc/llvm-project?rev=319909=rev Log: [NVPTX,CUDA] Added llvm.nvvm.fns intrinsic and matching __nvvm_fns builtin in clang. Differential Revision: https://reviews.llvm.org/D40872 Modified:

r319908 - [CUDA] Added overloads for '[unsigned] long' variants of shfl builtins.

Author: tra Date: Wed Dec 6 09:40:35 2017 New Revision: 319908 URL: http://llvm.org/viewvc/llvm-project?rev=319908=rev Log: [CUDA] Added overloads for '[unsigned] long' variants of shfl builtins. Differential Revision: https://reviews.llvm.org/D40871 Modified:

r319485 - [CUDA] Tweak CUDA wrappers to make cuda-9 work with libc++

Author: tra Date: Thu Nov 30 14:22:21 2017 New Revision: 319485 URL: http://llvm.org/viewvc/llvm-project?rev=319485=rev Log: [CUDA] Tweak CUDA wrappers to make cuda-9 work with libc++ CUDA-9 headers check for specific libc++ version and ifdef out some of the definitions we need if LIBCPP_VERSION

r319201 - [CUDA] Report "unsupported VLA" errors only on device side.

Author: tra Date: Tue Nov 28 10:51:42 2017 New Revision: 319201 URL: http://llvm.org/viewvc/llvm-project?rev=319201=rev Log: [CUDA] Report "unsupported VLA" errors only on device side. This fixes erroneously reported CUDA compilation errors in host-side code during device-side compilation. I've

Re: r318601 - [OpenMP] Show error if VLAs are not supported

Proposed fix: https://reviews.llvm.org/D40275 On Mon, Nov 20, 2017 at 4:13 PM, Artem Belevich wrote: > This change breaks CUDA as clang now reports an error during device-side > compilation when VLA is used in the *host-side* code. >

Re: r318601 - [OpenMP] Show error if VLAs are not supported

This change breaks CUDA as clang now reports an error during device-side compilation when VLA is used in the *host-side* code. http://lab.llvm.org:8011/builders/clang-cuda-build/builds/15591/steps/ninja%20build%20simple%20CUDA%20tests/logs/stdio E.g. I would expect this code to compile

r315624 - [CUDA] Added __hmma_m16n16k16_* builtins to support mma instructions on sm_70

Author: tra Date: Thu Oct 12 14:32:19 2017 New Revision: 315624 URL: http://llvm.org/viewvc/llvm-project?rev=315624=rev Log: [CUDA] Added __hmma_m16n16k16_* builtins to support mma instructions on sm_70 Differential Revision: https://reviews.llvm.org/D38742 Added:

r314334 - [CUDA] Work around conflicting function definitions in CUDA-9 headers.

Author: tra Date: Wed Sep 27 12:07:15 2017 New Revision: 314334 URL: http://llvm.org/viewvc/llvm-project?rev=314334=rev Log: [CUDA] Work around conflicting function definitions in CUDA-9 headers. Differential Revision: https://reviews.llvm.org/D38326 Modified:

r314135 - [NVPTX] added match.{any, all}.sync instructions, intrinsics & builtins.

Author: tra Date: Mon Sep 25 11:53:57 2017 New Revision: 314135 URL: http://llvm.org/viewvc/llvm-project?rev=314135=rev Log: [NVPTX] added match.{any,all}.sync instructions, intrinsics & builtins. Differential Revision: https://reviews.llvm.org/D38191 Modified:

r314129 - [CUDA] Fix names of __nvvm_vote* intrinsics.

Author: tra Date: Mon Sep 25 10:55:26 2017 New Revision: 314129 URL: http://llvm.org/viewvc/llvm-project?rev=314129=rev Log: [CUDA] Fix names of __nvvm_vote* intrinsics. Also fixed a syntax error in activemask(). Differential Revision: https://reviews.llvm.org/D38188 Modified:

r313899 - [CUDA] Fixed order of words in the names of shfl builtins.

Author: tra Date: Thu Sep 21 11:46:39 2017 New Revision: 313899 URL: http://llvm.org/viewvc/llvm-project?rev=313899=rev Log: [CUDA] Fixed order of words in the names of shfl builtins. Differential Revision: https://reviews.llvm.org/D38147 Modified:

r313898 - [NVPTX] Implemented bar.warp.sync, barrier.sync, and vote{.sync} instructions/intrinsics/builtins.

Author: tra Date: Thu Sep 21 11:44:49 2017 New Revision: 313898 URL: http://llvm.org/viewvc/llvm-project?rev=313898=rev Log: [NVPTX] Implemented bar.warp.sync, barrier.sync, and vote{.sync} instructions/intrinsics/builtins. Differential Revision: https://reviews.llvm.org/D38148 Modified:

r313820 - [NVPTX] Implemented shfl.sync instruction and supporting intrinsics/builtins.

Author: tra Date: Wed Sep 20 14:23:07 2017 New Revision: 313820 URL: http://llvm.org/viewvc/llvm-project?rev=313820=rev Log: [NVPTX] Implemented shfl.sync instruction and supporting intrinsics/builtins. Differential Revision: https://reviews.llvm.org/D38090 Added:

r313369 - [CUDA] Work around a new quirk in CUDA9 headers.

Author: tra Date: Fri Sep 15 10:30:53 2017 New Revision: 313369 URL: http://llvm.org/viewvc/llvm-project?rev=313369=rev Log: [CUDA] Work around a new quirk in CUDA9 headers. In CUDA-9 some of device-side math functions that we need are conditionally defined within '#if _GLIBCXX_MATH_H'. We need

r312734 - [CUDA] Added rudimentary support for CUDA-9 and sm_70.

Author: tra Date: Thu Sep 7 11:14:32 2017 New Revision: 312734 URL: http://llvm.org/viewvc/llvm-project?rev=312734=rev Log: [CUDA] Added rudimentary support for CUDA-9 and sm_70. For now CUDA-9 is not included in the list of CUDA versions clang searches for, so the path to CUDA-9 must be

r308675 - [NVPTX] Add lowering of i128 params.

Author: tra Date: Thu Jul 20 14:16:03 2017 New Revision: 308675 URL: http://llvm.org/viewvc/llvm-project?rev=308675=rev Log: [NVPTX] Add lowering of i128 params. The patch adds support of i128 params lowering. The changes are quite trivial to support i128 as a "special case" of integer type.

r290982 - [CUDA] Pre-include sm_60 and sm_61 headers.

Author: tra Date: Wed Jan 4 12:39:29 2017 New Revision: 290982 URL: http://llvm.org/viewvc/llvm-project?rev=290982=rev Log: [CUDA] Pre-include sm_60 and sm_61 headers. CUDA-8.0 comes with new headers which nvcc pre-includes via cuda_runtime.h Clang now makes them available as well.

r289287 - [CUDA,Driver] Added --no-cuda-gpu-arch= option.

Author: tra Date: Fri Dec 9 16:59:17 2016 New Revision: 289287 URL: http://llvm.org/viewvc/llvm-project?rev=289287=rev Log: [CUDA,Driver] Added --no-cuda-gpu-arch= option. This allows us to negate preceding --cuda-gpu-arch=X. This comes handy when user needs to override default flags set for

r288962 - [CUDA] Improve target attribute checking for function templates.

Author: tra Date: Wed Dec 7 13:27:16 2016 New Revision: 288962 URL: http://llvm.org/viewvc/llvm-project?rev=288962=rev Log: [CUDA] Improve target attribute checking for function templates. * __host__ __device__ functions are no longer considered to be redeclarations of __host__ or __device__

r288406 - Send compiler output to /dev/null in defsym.s test.

Author: tra Date: Thu Dec 1 13:34:35 2016 New Revision: 288406 URL: http://llvm.org/viewvc/llvm-project?rev=288406=rev Log: Send compiler output to /dev/null in defsym.s test. Fixes test failures if tests are run in a read-only source tree. Modified: cfe/trunk/test/Driver/defsym.s

[PATCH] D26774: [CUDA] Driver changes to support CUDA compilation on MacOS.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM, with couple of minor nits. Comment at: clang/lib/Driver/Driver.cpp:3650-3654 + + // Intentionally omitted from the switch above: llvm::Triple::CUDA. CUDA + // compiles

[PATCH] D26268: [CUDA] Use only the GVALinkage on function definitions.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. Comment at: clang/test/SemaCUDA/add-inline-in-definition.cu:13-14 +// +// The trickiness here comes from the fact that the FunctionDecl bar() sees for +// foo() does not

[PATCH] D25640: [CUDA] [AST] Allow isInlineDefinitionExternallyVisible to be called on functions without bodies.

tra added a comment. I'm OK with the change, but the comments suggest that things may be more complicated. How about disabling assert for CUDA only? https://reviews.llvm.org/D25640 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

[PATCH] D25809: [CUDA] Improved target attribute-based overloading.

tra updated this revision to Diff 75816. tra added a comment. - handle using declarations found in the overload set we check. https://reviews.llvm.org/D25809 Files: include/clang/Basic/DiagnosticSemaKinds.td include/clang/Sema/Sema.h lib/Sema/SemaCUDA.cpp lib/Sema/SemaDecl.cpp

[PATCH] D25845: [CUDA] Ignore implicit target attributes during function template instantiation.

tra updated this revision to Diff 75812. tra added a comment. - Instead of relying on the first attribute we find, check all matching ones. - Specializations inherit their target attributes from their base template only. Their effective target always matches that of the template and is no

[PATCH] D25888: Add support for __builtin_os_log_format[_buffer_size]

tra added a comment. CodeGen's CMakeLists.txt needs to be updated to link with clangAnalysis now, otherwise it breaks builds that use shared libraries. Fixed in r285037. diff --git a/lib/CodeGen/CMakeLists.txt b/lib/CodeGen/CMakeLists.txt index f5d5d69..9cf34f6 100644 ---

r285037 - Link clangCodeGen with clangAnalysis required after r284990.

Author: tra Date: Mon Oct 24 17:52:39 2016 New Revision: 285037 URL: http://llvm.org/viewvc/llvm-project?rev=285037=rev Log: Link clangCodeGen with clangAnalysis required after r284990. Fixes build break for configurations that use shared libraries. Modified:

[PATCH] D25809: [CUDA] Improved target attribute-based overloading.

tra added inline comments. Comment at: lib/Sema/SemaCUDA.cpp:791 + CUDAFunctionTarget NewTarget = IdentifyCUDATarget(NewFD); + for (auto OldND : Previous) { +FunctionDecl *OldFD = OldND->getAsFunction(); jlebar wrote: > tra wrote: > > jlebar wrote: > > >

[PATCH] D25809: [CUDA] Improved target attribute-based overloading.

tra updated this revision to Diff 75652. tra marked an inline comment as done. tra added a comment. Addressed remaining nits. https://reviews.llvm.org/D25809 Files: include/clang/Basic/DiagnosticSemaKinds.td include/clang/Sema/Sema.h lib/Sema/SemaCUDA.cpp lib/Sema/SemaDecl.cpp

[PATCH] D25809: [CUDA] Improved target attribute-based overloading.

tra updated this revision to Diff 75516. tra added a comment. removed pragma check from IdentifyCUDATarget for real. https://reviews.llvm.org/D25809 Files: include/clang/Basic/DiagnosticSemaKinds.td include/clang/Sema/Sema.h lib/Sema/SemaCUDA.cpp lib/Sema/SemaDecl.cpp

[PATCH] D25809: [CUDA] Improved target attribute-based overloading.

tra added inline comments. Comment at: lib/Sema/SemaCUDA.cpp:87 + + if ((HasHostAttr && HasDeviceAttr) || ForceCUDAHostDeviceDepth > 0) +return CFT_HostDevice; jlebar wrote: > Checking ForceCUDAHostDeviceDepth here is...yeah. Especially because the >

[PATCH] D25809: [CUDA] Improved target attribute-based overloading.

tra updated this revision to Diff 75515. tra marked 5 inline comments as done. tra added a comment. addressed jlebar's comments. https://reviews.llvm.org/D25809 Files: include/clang/Basic/DiagnosticSemaKinds.td include/clang/Sema/Sema.h lib/Sema/SemaCUDA.cpp lib/Sema/SemaDecl.cpp

[PATCH] D25845: [CUDA] Ignore implicit target attributes during function template instantiation.

tra added inline comments. Comment at: include/clang/Sema/Sema.h:9396 + CUDAFunctionTarget IdentifyCUDATarget(const FunctionDecl *D, +bool IgnoreImplicitHDAttr = false); CUDAFunctionTarget IdentifyCUDATarget(const AttributeList *Attr);

[PATCH] D25845: [CUDA] Ignore implicit target attributes during function template instantiation.

tra updated this revision to Diff 75511. tra marked 5 inline comments as done. tra added a comment. addressed jlebar's comments. https://reviews.llvm.org/D25845 Files: include/clang/Sema/Sema.h lib/Sema/SemaCUDA.cpp lib/Sema/SemaDeclAttr.cpp lib/Sema/SemaTemplate.cpp

[PATCH] D25845: [CUDA] Ignore implicit target attributes during function template instantiation.

tra added a comment. In https://reviews.llvm.org/D25845#576819, @jlebar wrote: > > This patch changes the way we match target attributes of base template vs > > attributes used in explicit instantiation or specialization so that only > > explicitly specified attributes are considered. > >

[PATCH] D25845: [CUDA] Ignore implicit target attributes during function template instantiation.

tra updated this revision to Diff 75482. tra added a comment. Added a comment explaining expected constexpr function template matching behavior. https://reviews.llvm.org/D25845 Files: include/clang/Sema/Sema.h lib/Sema/SemaCUDA.cpp lib/Sema/SemaDeclAttr.cpp lib/Sema/SemaTemplate.cpp

r284879 - Declare H and H new/delete.

Author: tra Date: Fri Oct 21 15:34:05 2016 New Revision: 284879 URL: http://llvm.org/viewvc/llvm-project?rev=284879=rev Log: Declare H and H new/delete. Modified: cfe/trunk/lib/Sema/SemaExprCXX.cpp cfe/trunk/test/SemaCUDA/overloaded-delete.cu Modified: cfe/trunk/lib/Sema/SemaExprCXX.cpp

[PATCH] D25839: Removed unused function argument. NFC.

This revision was automatically updated to reflect the committed changes. Closed by commit rL284843: Removed unused function argument. NFC. (authored by tra). Changed prior to commit: https://reviews.llvm.org/D25839?vs=75339=75447#toc Repository: rL LLVM https://reviews.llvm.org/D25839

r284843 - Removed unused function argument. NFC.

Author: tra Date: Fri Oct 21 12:15:46 2016 New Revision: 284843 URL: http://llvm.org/viewvc/llvm-project?rev=284843=rev Log: Removed unused function argument. NFC. Differential Revision: https://reviews.llvm.org/D25839 Modified: cfe/trunk/include/clang/Sema/Sema.h

[PATCH] D25845: [CUDA] Ignore implicit target attributes during function template instantiation.

tra created this revision. tra added reviewers: jlebar, rsmith. tra added a subscriber: cfe-commits. Some functions and templates are treated as `__host__` `__device__` even when they don't have explicitly specified target attributes. What's worse, this treatment may change depending on command

[PATCH] D25839: Removed unused function argument. NFC.

tra created this revision. tra added a reviewer: jlebar. tra added a subscriber: cfe-commits. https://reviews.llvm.org/D25839 Files: include/clang/Sema/Sema.h lib/Sema/SemaCUDA.cpp lib/Sema/SemaDecl.cpp Index: lib/Sema/SemaDecl.cpp

[PATCH] D25809: [CUDA] Improved target attribute-based overloading.

tra created this revision. tra added reviewers: jlebar, rsmith. tra added a subscriber: cfe-commits. Current behavior: - __host__ __device__ (HD) functions are considered to be redeclarations of `__host__` (H) of `__device__` (D) functions with same signature. - Target attributes are not taken

[PATCH] D25796: [CUDA] Create __host__ and device variants of standard allocator declarations.

tra created this revision. tra added a reviewer: jlebar. tra added a subscriber: cfe-commits. Implicit functions are treated as if they were __host__ __device__ and clang does not allow overloading those with __host__ or __device__ variants. In order for users to provide their own standard

[PATCH] D25755: [CUDA] Rework tests now that we emit deferred diagnostics during sema. Test-only change.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. https://reviews.llvm.org/D25755 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

r283964 - Added REQUIRED triples to the test that fails on some ARM buildbots.

Author: tra Date: Tue Oct 11 21:08:08 2016 New Revision: 283964 URL: http://llvm.org/viewvc/llvm-project?rev=283964=rev Log: Added REQUIRED triples to the test that fails on some ARM buildbots. Modified: cfe/trunk/test/SemaCUDA/function-overload-hd.cu Modified:

[PATCH] D25129: [CUDA] Disallow __constant__ local variables.

tra added inline comments. > jlebar wrote in DiagnosticSemaKinds.td:6727 > That's still a "global variable"? Or do you think calling it such will be > confusing? It's not clear whether you mean global storage class or global namespace. The code checks for global storage, but error message is

[PATCH] D25129: [CUDA] Disallow __constant__ local variables.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. > DiagnosticSemaKinds.td:6727 > def err_cuda_extern_shared : Error<"__shared__ variable %0 cannot be > 'extern'">; > +def err_cuda_nonglobal_constant : Error<"__constant__ variables must

[PATCH] D25125: [CUDA] Disallow 'extern __shared__' variables.

tra accepted this revision. tra added a reviewer: tra. tra added a comment. This revision is now accepted and ready to land. LGTM. https://reviews.llvm.org/D25125 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

Re: [PATCH] D25036: [CUDA] Disallow exceptions in device code.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. One question, LGTM otherwise. Comment at: clang/lib/Sema/SemaExprCXX.cpp:688 @@ +687,3 @@ + if (getLangOpts().CUDA) +CheckCUDAExceptionExpr(OpLoc, "throw"); +

Re: [PATCH] D24946: [CUDA] Added support for CUDA-8

This revision was automatically updated to reflect the committed changes. Closed by commit rL282610: [CUDA] Added support for CUDA-8 (authored by tra). Changed prior to commit: https://reviews.llvm.org/D24946?vs=72707=72863#toc Repository: rL LLVM https://reviews.llvm.org/D24946 Files:

Re: [PATCH] D24944: [CUDA] Added __nvvm_atom_{sys|cta}_* builtins for sm_60 GPUs.

This revision was automatically updated to reflect the committed changes. Closed by commit rL282609: [CUDA] added __nvvm_atom_{sys|cta}_* builtins. (authored by tra). Changed prior to commit: https://reviews.llvm.org/D24944?vs=72584=72862#toc Repository: rL LLVM

r282610 - [CUDA] Added support for CUDA-8

Author: tra Date: Wed Sep 28 12:47:40 2016 New Revision: 282610 URL: http://llvm.org/viewvc/llvm-project?rev=282610=rev Log: [CUDA] Added support for CUDA-8 Differential Revision: https://reviews.llvm.org/D24946 Added:

r282609 - [CUDA] added __nvvm_atom_{sys|cta}_* builtins.

Author: tra Date: Wed Sep 28 12:47:35 2016 New Revision: 282609 URL: http://llvm.org/viewvc/llvm-project?rev=282609=rev Log: [CUDA] added __nvvm_atom_{sys|cta}_* builtins. These builtins are available on sm_60+ GPU only. Differential Revision: https://reviews.llvm.org/D24944 Modified:

Re: [PATCH] D24946: [CUDA] Added support for CUDA-8

tra updated this revision to Diff 72707. tra added a comment. addressed Justin's comments. https://reviews.llvm.org/D24946 Files: lib/Driver/ToolChains.cpp lib/Headers/__clang_cuda_runtime_wrapper.h test/Driver/Inputs/CUDA/usr/local/cuda/nvvm/libdevice/libdevice.compute_20.10.bc

Re: [PATCH] D24946: [CUDA] Added support for CUDA-8

tra marked an inline comment as done. Comment at: lib/Headers/__clang_cuda_runtime_wrapper.h:156 @@ +155,3 @@ +#endif +#endif + jlebar wrote: > I don't understand what we are doing here... > > We're saying, if __USE_FAST_MATH__ is defined, and if it's not equal

Re: [PATCH] D24977: [CUDA] Declare our __device__ math functions in the same inline namespace as our standard library.

tra added a comment. OK. https://reviews.llvm.org/D24977 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Re: [PATCH] D24979: [CUDA] Support and std::min/max on the device.

tra added a subscriber: echristo. tra added a comment. This looks like fix-includes and it may be somewhat shaky if users start messing with include paths. You may want to get @echristo's input on that. I' personally would prefer to force-include these files. I suspect it will not change

Re: [PATCH] D24977: [CUDA] Declare our __device__ math functions in the same inline namespace as our standard library.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. That is way too much knowledge about details of standard library implementation. If it changes, I suspect users will end up with a rather uninformative error. Is there a way to produce somewhat more

Re: [PATCH] D24975: [CUDA] Add #pragma clang force_cuda_host_device_{begin, end} pragmas.

tra added a comment. LGTM. Should we add new pragma description to docs/LanguageExtensions.rst ? https://reviews.llvm.org/D24975 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D24946: [CUDA] Added support for CUDA-8

tra created this revision. tra added a reviewer: jlebar. tra added a subscriber: cfe-commits. Herald added a subscriber: jlebar. https://reviews.llvm.org/D24946 Files: lib/Driver/ToolChains.cpp lib/Headers/__clang_cuda_runtime_wrapper.h

[PATCH] D24944: [CUDA] Added __nvvm_atom_{sys|cta}_* builtins for sm_60 GPUs.

tra created this revision. tra added a reviewer: jlebar. tra added a subscriber: cfe-commits. Herald added subscribers: jlebar, jholewinski. https://reviews.llvm.org/D24944 Files: include/clang/Basic/BuiltinsNVPTX.def lib/Basic/Targets.cpp lib/CodeGen/CGBuiltin.cpp

r281557 - Revert r281457 "Supports adding insertion around non-insertion replacements."

Author: tra Date: Wed Sep 14 18:03:06 2016 New Revision: 281557 URL: http://llvm.org/viewvc/llvm-project?rev=281557=rev Log: Revert r281457 "Supports adding insertion around non-insertion replacements." Commit was breaking our internal tests. Modified:

Re: [PATCH] D24590: [test-suite] [CUDA] Update README.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM Comment at: External/CUDA/README:58 @@ +57,3 @@ + There's a cuda-tests-simple target that excludes tests that take a long time + to build (thrust). + It may

Re: [PATCH] D24589: [test-suite] [CUDA] Add and tests.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. https://reviews.llvm.org/D24589 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

Re: [PATCH] D24588: [CUDA] Make __clang_cuda_cmath.h compatible with libc++.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. https://reviews.llvm.org/D24588 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

Re: [PATCH] D24581: [CUDA] Add test checking our ability to take a function pointer to a __global__ function on the host side.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. https://reviews.llvm.org/D24581 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

Re: r281351 - Add a class ObjCProtocolQualifiers to wrap APIs for ObjC protocol list.

Thanks for the quick fix. ASAN is happy now. --Artem On Tue, Sep 13, 2016 at 3:09 PM, Manman <m...@apple.com> wrote: > I checked in r281404. Hopefully it will fix the issue. > > Let me know if it does not. > > Thanks, > Manman > > On Sep 13, 2016, at 3:03 PM,

Re: [PATCH] D24522: [CUDA] Do not merge CUDA target attributes.

This revision was automatically updated to reflect the committed changes. Closed by commit rL281406: [CUDA] Do not merge CUDA target attributes. (authored by tra). Changed prior to commit: https://reviews.llvm.org/D24522?vs=71244=71249#toc Repository: rL LLVM

r281406 - [CUDA] Do not merge CUDA target attributes.

Author: tra Date: Tue Sep 13 17:16:30 2016 New Revision: 281406 URL: http://llvm.org/viewvc/llvm-project?rev=281406=rev Log: [CUDA] Do not merge CUDA target attributes. CUDA target attributes are used for function overloading and must not be merged. This fixes a bug where attributes were

Re: [PATCH] D24522: [CUDA] Do not merge CUDA target attributes.

tra updated this revision to Diff 71244. tra marked an inline comment as done. tra added a comment. Removed REQUIRED lines. https://reviews.llvm.org/D24522 Files: lib/Sema/SemaDecl.cpp test/SemaCUDA/function-overload.cu test/SemaCUDA/target_attr_inheritance.cu Index:

Re: r281351 - Add a class ObjCProtocolQualifiers to wrap APIs for ObjC protocol list.

Manman, FYI, It appears that some of your ObjC commits today trigger asan error. Sanitizer bots are broken by PR30341, so they don't report the issue yet. --Artem $ llvm/tools/clang/clang -cc1 -internal-isystem llvm/tools/clang/staging/include -nostdsysteminc -fblocks -fsyntax-only

[PATCH] D24522: [CUDA] Do not merge CUDA target attributes.

tra created this revision. tra added a reviewer: jlebar. tra added a subscriber: cfe-commits. Herald added a subscriber: jlebar. CUDA target attributes are used for function overloading and must not be merged. This fixes a bug where attributes were inherited during function template

Re: [PATCH] D24407: [CUDA] Make __GCC_ATOMIC_XXX_LOCK_FREE macros the same on host/device.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM https://reviews.llvm.org/D24407 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

Re: [PATCH] D23526: [CUDA] Collapsed offload actions should not be top-level jobs.

This revision was automatically updated to reflect the committed changes. Closed by commit rL279455: [CUDA] Collapsed offload actions should not be top-level jobs. (authored by tra). Changed prior to commit: https://reviews.llvm.org/D23526?vs=68100=68896#toc Repository: rL LLVM

r279455 - [CUDA] Collapsed offload actions should not be top-level jobs.

Author: tra Date: Mon Aug 22 13:50:34 2016 New Revision: 279455 URL: http://llvm.org/viewvc/llvm-project?rev=279455=rev Log: [CUDA] Collapsed offload actions should not be top-level jobs. If they are, we end up with the last intermediary output preserved in the current directory after

Re: [PATCH] D23627: [CUDA] Improve handling of math functions.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM, but we may want someone familiar with math library to take a look. Comment at: clang/lib/Headers/__clang_cuda_cmath.h:125-133 @@ -122,8 +124,11 @@ __DEVICE__ float

Re: [PATCH] D23627: [CUDA] Improve handling of math functions.

tra added inline comments. Comment at: clang/lib/Headers/__clang_cuda_cmath.h:125-133 @@ -122,8 +124,11 @@ __DEVICE__ float modf(float __x, float *__iptr) { return ::modff(__x, __iptr); } -__DEVICE__ float nexttoward(float __from, float __to) { +__DEVICE__ float

Re: [PATCH] D23526: [CUDA] Collapsed offload actions should not be top-level jobs.

tra marked 2 inline comments as done. tra added a comment. https://reviews.llvm.org/D23526 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Re: [PATCH] D23526: [CUDA] Collapsed offload actions should not be top-level jobs.

tra updated this revision to Diff 68100. tra added a comment. Addressed comments. https://reviews.llvm.org/D23526 Files: lib/Driver/Driver.cpp test/Driver/cuda-bindings.cu Index: test/Driver/cuda-bindings.cu === --- /dev/null

[PATCH] D23526: [CUDA] Collapsed offload actions should not be top-level jobs.

tra created this revision. tra added reviewers: jlebar, sfantao. tra added a subscriber: cfe-commits. If they are, we end up with the last intermediary output preserved in the current directory after compilation. Added a test case to verify that we're using appropriate filenames for outputs of

Re: [PATCH] D23429: [CUDA] Place GPU binary into .nv_fatbin section and align it by 8.

This revision was automatically updated to reflect the committed changes. tra marked an inline comment as done. Closed by commit rL278549: [CUDA] Place GPU binary into .nv_fatbin section and align it by 8. (authored by tra). Changed prior to commit:

r278549 - [CUDA] Place GPU binary into .nv_fatbin section and align it by 8.

Author: tra Date: Fri Aug 12 13:44:01 2016 New Revision: 278549 URL: http://llvm.org/viewvc/llvm-project?rev=278549=rev Log: [CUDA] Place GPU binary into .nv_fatbin section and align it by 8. This matches the way nvcc encapsulates GPU binaries into host object file. Now cuobjdump can deal with

Re: [PATCH] D23429: [CUDA] Place GPU binary into .nv_fatbin section and align it by 8.

tra marked an inline comment as done. Comment at: lib/CodeGen/CGCUDANV.cpp:62-69 @@ -60,3 +61,10 @@ llvm::ConstantInt::get(SizeTy, 0)}; auto ConstStr = CGM.GetAddrOfConstantCString(Str, Name.c_str()); +llvm::GlobalVariable *GV = +

Re: [PATCH] D23429: [CUDA] Place GPU binary into .nv_fatbin section and align it by 8.

tra updated this revision to Diff 67851. tra added a comment. Reverted argument type to std::string https://reviews.llvm.org/D23429 Files: lib/CodeGen/CGCUDANV.cpp test/CodeGenCUDA/device-stub.cu Index: test/CodeGenCUDA/device-stub.cu

Re: [PATCH] D23341: [CUDA] Include CUDA headers before anything else.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. OK. This should help with PR26966. https://reviews.llvm.org/D23341 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

[PATCH] D23429: [CUDA] Place GPU binary into .nv_fatbin section and align it by 8.

tra created this revision. tra added a reviewer: jlebar. tra added a subscriber: cfe-commits. This matches the way nvcc encapsulates GPU binaries into host object file. Now cuobjdump can deal with clang-compiled object files. https://reviews.llvm.org/D23429 Files: lib/CodeGen/CGCUDANV.cpp

Re: [PATCH] D23335: [CUDA] Reject calls to __device__ functions from host variable global initializers.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. Comment at: clang/test/SemaCUDA/global-initializers-host.cu:2 @@ +1,3 @@ +// RUN: %clang_cc1 %s --std=c++11 -triple x86_64-linux-unknown -emit-llvm -o - -verify + +#include

Re: [PATCH] D22946: [CUDA] Regression test to make sure C++ include path are forwarded to host and device frontends.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. https://reviews.llvm.org/D22946 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

Re: [PATCH] D22946: [CUDA] Regression test to make sure C++ include path are forwarded to host and device frontends.

tra added inline comments. Comment at: test/Driver/cuda-detect.cu:75-77 @@ -74,1 +74,5 @@ +// Verify that compiler accepts CUDA syntax with "-x cuda-cpp-output". +// RUN: %clang -Werror -fsyntax-only -x cuda-cpp-output -c %s +// +// Verify that C++ include paths are passed for

Re: [PATCH] D23239: [CUDA] Add __device__ overloads for placement new and delete.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. https://reviews.llvm.org/D23239 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

Re: [PATCH] D23242: [CUDA] Raise an error if a wrong-side call is codegen'ed.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. Few nits, but looks good otherwise. Should be add few tests for calling device functions from host-side global initializers? Perhaps for device->host, too, as there may be unexpected interplay with

Re: [PATCH] D22946: [CUDA] Regression test to make sure C++ include path are forwarded to host and device frontends.

tra added a comment. cuda-detect.cu may be a better place for this test. Comment at: test/Driver/cuda-simple.cu:27 @@ -16,1 +26,3 @@ +// CHECK-CXXINCLUDE: clang{{.*}} "-cc1" "-triple" "x86_64--linux-gnu" {{.*}}"-internal-isystem" "{{.+}}/include/c++/4.8" +// CHECK-CXXINCLUDE:

Re: [PATCH] D23241: Add the notion of deferred diagnostics.

tra added inline comments. Comment at: clang/lib/CodeGen/CodeGenModule.cpp:2886 @@ +2885,3 @@ + // Check if this function has diagnostics that should be emitted when we + // codegen it. If so, don't eit this function definition, but don't emit the + // diags just yet.

Re: [PATCH] D23240: [CUDA] Print a "previous-decl" note when calling an illegal member fn.

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. https://reviews.llvm.org/D23240 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

Re: [PATCH] D23239: [CUDA] Add __device__ overloads for placement new and delete.

tra added a comment. I think we need to add `noexcept` for these in c++11. https://reviews.llvm.org/D23239 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Re: [PATCH] D23238: [CUDA] Rename CheckCUDATarget to IsAllowedCUDACall. NFC

tra accepted this revision. tra added a comment. This revision is now accepted and ready to land. LGTM. https://reviews.llvm.org/D23238 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

Re: r277542 - [CUDA] Fix libdevice selection.

i Artem, >> >> This broke >> http://lab.llvm.org:8080/green/job/clang-stage1-cmake-RA-incremental_check/26856 >> , >> can you take a look? >> >> Thanks, >> >> On Tue, Aug 2, 2016 at 4:12 PM, Artem Belevich via cfe-commits >> <cfe-commi

r277552 - [CUDA] Updated CUDA tests that must run w/o CUDA installation.

Author: tra Date: Tue Aug 2 18:43:04 2016 New Revision: 277552 URL: http://llvm.org/viewvc/llvm-project?rev=277552=rev Log: [CUDA] Updated CUDA tests that must run w/o CUDA installation. Fixes test failures after r277542 on systems that don't have CUDA installed. Modified:

Re: r277542 - [CUDA] Fix libdevice selection.

? > > Thanks, > > On Tue, Aug 2, 2016 at 4:12 PM, Artem Belevich via cfe-commits > <cfe-commits@lists.llvm.org> wrote: > > Author: tra > > Date: Tue Aug 2 18:12:51 2016 > > New Revision: 277542 > > > > URL: http://llvm.org/viewvc/llvm-project?rev

Re: [PATCH] D23037: [CUDA] Fix libdevice selection.

This revision was automatically updated to reflect the committed changes. Closed by commit rL277542: [CUDA] Fix libdevice selection. (authored by tra). Changed prior to commit: https://reviews.llvm.org/D23037?vs=66579=66589#toc Repository: rL LLVM https://reviews.llvm.org/D23037 Files:

<    1   2   3   4   5   6   7   8   9   >