r366683 - Updated the signature for some stack related intrinsics (CLANG)

2019-07-22 Thread Christudasan Devadasan via cfe-commits
Author: cdevadas Date: Mon Jul 22 05:50:30 2019 New Revision: 366683 URL: http://llvm.org/viewvc/llvm-project?rev=366683&view=rev Log: Updated the signature for some stack related intrinsics (CLANG) Modified the intrinsics int_addressofreturnaddress, int_frameaddress & int_sponentry. This commit

r365643 - [AMDGPU] Increased the number of implicit argument bytes for both OpenCL and HIP (CLANG).

2019-07-10 Thread Christudasan Devadasan via cfe-commits
Author: cdevadas Date: Wed Jul 10 08:10:08 2019 New Revision: 365643 URL: http://llvm.org/viewvc/llvm-project?rev=365643&view=rev Log: [AMDGPU] Increased the number of implicit argument bytes for both OpenCL and HIP (CLANG). To enable a new implicit kernel argument, increased the number of argum

[clang] [AMDGPU] Check wavefrontsize for GFX11 WMMA builtins (PR #79980)

2024-01-30 Thread Christudasan Devadasan via cfe-commits
@@ -21,14 +21,14 @@ void test_amdgcn_wmma_f32_16x16x16_bf16_w64(global v4f* out4f, v16h a16h, v16h b global v8s* out8s, v4i a4i, v4i b4i, v8s c8s, global v4i* out4i, v2i a2i, v2i b2i, v4i

[clang] [AMDGPU] Check wavefrontsize for GFX11 WMMA builtins (PR #79980)

2024-01-30 Thread Christudasan Devadasan via cfe-commits
https://github.com/cdevadas approved this pull request. https://github.com/llvm/llvm-project/pull/79980 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [HIP] Always add -fnative-half-arguments-and-returns cmdline option. (PR #113335)

2024-10-22 Thread Christudasan Devadasan via cfe-commits
https://github.com/cdevadas created https://github.com/llvm/llvm-project/pull/113335 This command-line option is now required while building the HIP applications (mainly for the host side) after we enabled __fp16 args and return values with patches D133885 & D145345. >From e843f0f792b83aa7b545e

[clang] [HIP] Always add -fnative-half-arguments-and-returns cmdline option. (PR #113335)

2024-10-22 Thread Christudasan Devadasan via cfe-commits
cdevadas wrote: * **#113335** https://app.graphite.dev/github/pr/llvm/llvm-project/113335?utm_source=stack-comment-icon"; target="_blank">https://static.graphite.dev/graphite-32x32-black.png"; alt="Graphite" width="10px" height="10px"/> 👈 * `main` This stack of pull requests is managed by Grap

[clang] [HIP] Always add -fnative-half-arguments-and-returns cmdline option. (PR #113335)

2024-10-22 Thread Christudasan Devadasan via cfe-commits
cdevadas wrote: ### Merge activity * **Oct 22, 12:36 PM EDT**: A user started a stack merge that includes this pull request via [Graphite](https://app.graphite.dev/github/pr/llvm/llvm-project/113335). https://github.com/llvm/llvm-project/pull/113335 __

[clang] [HIP] Always add -fnative-half-arguments-and-returns cmdline option. (PR #113335)

2024-10-22 Thread Christudasan Devadasan via cfe-commits
https://github.com/cdevadas closed https://github.com/llvm/llvm-project/pull/113335 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [HIP] Always add -fnative-half-arguments-and-returns cmdline option. (PR #113335)

2024-10-22 Thread Christudasan Devadasan via cfe-commits
https://github.com/cdevadas ready_for_review https://github.com/llvm/llvm-project/pull/113335 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [Clang][OpenCL][AMDGPU] Allow a kernel to call another kernel (PR #115821)

2024-11-27 Thread Christudasan Devadasan via cfe-commits
@@ -0,0 +1,43 @@ +// RUN: %clang_cc1 -triple amdgcn-unknown-unknown -emit-llvm -o - %s | FileCheck %s cdevadas wrote: I believe this test makes more sense to have the handpicked patterns to specifically look for the kernel &stub pair generated for the definitio

[clang] [Clang][OpenCL][AMDGPU] OpenCL Kernel stubs should be assigned alwaysinline attribute (PR #137769)

2025-05-05 Thread Christudasan Devadasan via cfe-commits
cdevadas wrote: > Apparently the inliner has an arbitrary size threshold even for single uses. > I don't see the point of that, we should probably remove that I had a similar concern. For a single use and in this case especially when the call `Inst` is the only instruction in the Caller, it sh

[clang] [Clang][OpenCL][AMDGPU] OpenCL Kernel stubs should be assigned alwaysinline attribute (PR #137769)

2025-04-29 Thread Christudasan Devadasan via cfe-commits
cdevadas wrote: > In practice this should be a single use of an internal function and should > not require this hint. Is this papering over a different issue? Why do you think it is an internal function? I thought it's an externally callable stub. https://github.com/llvm/llvm-project/pull/137