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
71 lines
2.9 KiB
LLVM
71 lines
2.9 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5
|
|
; RUN: llc -mattr=+sve2 -verify-machineinstrs < %s -o - | FileCheck %s
|
|
|
|
target triple = "aarch64-unknown-linux-gnu"
|
|
|
|
; This test is reduced from a real world example that would cause the DAGCombiner to hang.
|
|
|
|
define void @histcnt_loop(ptr %0, i64 %1, ptr %2, i64 %3, i64 %4) {
|
|
; CHECK-LABEL: histcnt_loop:
|
|
; CHECK: // %bb.0: // %entry
|
|
; CHECK-NEXT: mov z0.d, #1 // =0x1
|
|
; CHECK-NEXT: ptrue p0.d
|
|
; CHECK-NEXT: mov x8, xzr
|
|
; CHECK-NEXT: add x9, x0, x1
|
|
; CHECK-NEXT: .LBB0_1: // %loop
|
|
; CHECK-NEXT: // =>This Inner Loop Header: Depth=1
|
|
; CHECK-NEXT: ld1h { z1.d }, p0/z, [x0, x8, lsl #1]
|
|
; CHECK-NEXT: lsl x10, x8, #1
|
|
; CHECK-NEXT: add x11, x0, x10
|
|
; CHECK-NEXT: add x10, x9, x10
|
|
; CHECK-NEXT: lsl z1.d, z1.d, #1
|
|
; CHECK-NEXT: ld1h { z4.d }, p0/z, [x11, #1, mul vl]
|
|
; CHECK-NEXT: ld1h { z5.d }, p0/z, [x10, #1, mul vl]
|
|
; CHECK-NEXT: histcnt z2.d, p0/z, z1.d, z1.d
|
|
; CHECK-NEXT: ld1h { z3.d }, p0/z, [x2, z1.d]
|
|
; CHECK-NEXT: mad z2.d, p0/m, z0.d, z3.d
|
|
; CHECK-NEXT: ld1h { z3.d }, p0/z, [x9, x8, lsl #1]
|
|
; CHECK-NEXT: add x8, x8, x3
|
|
; CHECK-NEXT: cmp x4, x8
|
|
; CHECK-NEXT: st1h { z2.d }, p0, [x2, z1.d]
|
|
; CHECK-NEXT: lsl z1.d, z4.d, #1
|
|
; CHECK-NEXT: histcnt z2.d, p0/z, z1.d, z1.d
|
|
; CHECK-NEXT: ld1h { z4.d }, p0/z, [x2, z1.d]
|
|
; CHECK-NEXT: mad z2.d, p0/m, z0.d, z4.d
|
|
; CHECK-NEXT: st1h { z2.d }, p0, [x2, z1.d]
|
|
; CHECK-NEXT: lsl z1.d, z3.d, #1
|
|
; CHECK-NEXT: histcnt z2.d, p0/z, z1.d, z1.d
|
|
; CHECK-NEXT: ld1h { z3.d }, p0/z, [x2, z1.d]
|
|
; CHECK-NEXT: mad z2.d, p0/m, z0.d, z3.d
|
|
; CHECK-NEXT: st1h { z2.d }, p0, [x2, z1.d]
|
|
; CHECK-NEXT: lsl z1.d, z5.d, #1
|
|
; CHECK-NEXT: histcnt z2.d, p0/z, z1.d, z1.d
|
|
; CHECK-NEXT: ld1h { z3.d }, p0/z, [x2, z1.d]
|
|
; CHECK-NEXT: mad z2.d, p0/m, z0.d, z3.d
|
|
; CHECK-NEXT: st1h { z2.d }, p0, [x2, z1.d]
|
|
; CHECK-NEXT: b.ne .LBB0_1
|
|
; CHECK-NEXT: // %bb.2: // %exit
|
|
; CHECK-NEXT: ret
|
|
entry:
|
|
br label %loop
|
|
|
|
loop:
|
|
%6 = phi i64 [ 0, %entry ], [ %15, %loop ]
|
|
%7 = getelementptr inbounds nuw i16, ptr %0, i64 %6
|
|
%8 = getelementptr inbounds nuw i8, ptr %7, i64 %1
|
|
%9 = load <vscale x 4 x i16>, ptr %7, align 2
|
|
%10 = load <vscale x 4 x i16>, ptr %8, align 2
|
|
%11 = zext <vscale x 4 x i16> %9 to <vscale x 4 x i64>
|
|
%12 = zext <vscale x 4 x i16> %10 to <vscale x 4 x i64>
|
|
%13 = getelementptr inbounds nuw [16 x i16], ptr %2, i64 0, <vscale x 4 x i64> %11
|
|
%14 = getelementptr inbounds nuw [16 x i16], ptr %2, i64 0, <vscale x 4 x i64> %12
|
|
call void @llvm.experimental.vector.histogram.add.nxv4p0.i16(<vscale x 4 x ptr> %13, i16 1, <vscale x 4 x i1> splat (i1 true))
|
|
call void @llvm.experimental.vector.histogram.add.nxv4p0.i16(<vscale x 4 x ptr> %14, i16 1, <vscale x 4 x i1> splat (i1 true))
|
|
%15 = add nuw i64 %6, %3
|
|
%16 = icmp eq i64 %15, %4
|
|
br i1 %16, label %exit, label %loop
|
|
|
|
exit:
|
|
ret void
|
|
}
|