File: llvm.r600.dot4.ll

package info (click to toggle)
llvm-toolchain-18 1%3A18.1.8-18
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid, trixie
  • size: 1,908,340 kB
  • sloc: cpp: 6,667,937; ansic: 1,440,452; asm: 883,619; python: 230,549; objc: 76,880; f90: 74,238; lisp: 35,989; pascal: 16,571; sh: 10,229; perl: 7,459; ml: 5,047; awk: 3,523; makefile: 2,987; javascript: 2,149; xml: 892; fortran: 649; cs: 573
file content (11 lines) | stat: -rw-r--r-- 526 bytes parent folder | download | duplicates (9)
1
2
3
4
5
6
7
8
9
10
11
; RUN: llc -mtriple=r600 -mcpu=redwood -verify-machineinstrs < %s

declare float @llvm.r600.dot4(<4 x float>, <4 x float>) nounwind readnone

define amdgpu_kernel void @test_dp4(ptr addrspace(1) %out, ptr addrspace(1) %a, ptr addrspace(1) %b) nounwind {
  %src0 = load <4 x float>, ptr addrspace(1) %a, align 16
  %src1 = load <4 x float>, ptr addrspace(1) %b, align 16
  %dp4 = call float @llvm.r600.dot4(<4 x float> %src0, <4 x float> %src1) nounwind readnone
  store float %dp4, ptr addrspace(1) %out, align 4
  ret void
}