derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
68 lines
2.1 KiB
LLVM
68 lines
2.1 KiB
LLVM
; RUN: llc < %s -mcpu=sm_35 | FileCheck %s
|
|
|
|
target datalayout = "e-i64:64-v16:16-v32:32-n16:32:64"
|
|
target triple = "nvptx64-nvidia-cuda"
|
|
|
|
; CUDA code
|
|
; __global__ void foo(int b, int s) {
|
|
; #pragma unroll
|
|
; for (int i = 0; i < 4; ++i) {
|
|
; if (cond(i))
|
|
; use((b + i) * s);
|
|
; }
|
|
; }
|
|
define ptx_kernel void @foo(i32 %b, i32 %s) {
|
|
; CHECK-LABEL: .visible .entry foo(
|
|
entry:
|
|
; CHECK: ld.param.b32 [[s:%r[0-9]+]], [foo_param_1];
|
|
; CHECK: ld.param.b32 [[b:%r[0-9]+]], [foo_param_0];
|
|
%call = tail call zeroext i1 @cond(i32 0)
|
|
br i1 %call, label %if.then, label %for.inc
|
|
|
|
if.then: ; preds = %entry
|
|
%mul = mul nsw i32 %b, %s
|
|
; CHECK: mul.lo.s32 [[a0:%r[0-9]+]], [[b]], [[s]]
|
|
tail call void @use(i32 %mul)
|
|
br label %for.inc
|
|
|
|
for.inc: ; preds = %entry, %if.then
|
|
%call.1 = tail call zeroext i1 @cond(i32 1)
|
|
br i1 %call.1, label %if.then.1, label %for.inc.1
|
|
|
|
if.then.1: ; preds = %for.inc
|
|
%add.1 = add nsw i32 %b, 1
|
|
%mul.1 = mul nsw i32 %add.1, %s
|
|
; CHECK: add.s32 [[a1:%r[0-9]+]], [[a0]], [[s]]
|
|
tail call void @use(i32 %mul.1)
|
|
br label %for.inc.1
|
|
|
|
for.inc.1: ; preds = %if.then.1, %for.inc
|
|
%call.2 = tail call zeroext i1 @cond(i32 2)
|
|
br i1 %call.2, label %if.then.2, label %for.inc.2
|
|
|
|
if.then.2: ; preds = %for.inc.1
|
|
%add.2 = add nsw i32 %b, 2
|
|
%mul.2 = mul nsw i32 %add.2, %s
|
|
; CHECK: add.s32 [[a2:%r[0-9]+]], [[a1]], [[s]]
|
|
tail call void @use(i32 %mul.2)
|
|
br label %for.inc.2
|
|
|
|
for.inc.2: ; preds = %if.then.2, %for.inc.1
|
|
%call.3 = tail call zeroext i1 @cond(i32 3)
|
|
br i1 %call.3, label %if.then.3, label %for.inc.3
|
|
|
|
if.then.3: ; preds = %for.inc.2
|
|
%add.3 = add nsw i32 %b, 3
|
|
%mul.3 = mul nsw i32 %add.3, %s
|
|
; CHECK: add.s32 [[a3:%r[0-9]+]], [[a2]], [[s]]
|
|
tail call void @use(i32 %mul.3)
|
|
br label %for.inc.3
|
|
|
|
for.inc.3: ; preds = %if.then.3, %for.inc.2
|
|
ret void
|
|
}
|
|
|
|
declare zeroext i1 @cond(i32)
|
|
|
|
declare void @use(i32)
|