diff options
author | Dimitry Andric <dim@FreeBSD.org> | 2015-01-18 16:17:27 +0000 |
---|---|---|
committer | Dimitry Andric <dim@FreeBSD.org> | 2015-01-18 16:17:27 +0000 |
commit | 67c32a98315f785a9ec9d531c1f571a0196c7463 (patch) | |
tree | 4abb9cbeecc7901726dd0b4a37369596c852e9ef /test/CodeGen/R600/fsub.ll | |
parent | 9f61947910e6ab40de38e6b4034751ef1513200f (diff) | |
download | src-67c32a98315f785a9ec9d531c1f571a0196c7463.tar.gz src-67c32a98315f785a9ec9d531c1f571a0196c7463.zip |
Vendor import of llvm RELEASE_360/rc1 tag r226102 (effectively, 3.6.0 RC1):vendor/llvm/llvm-release_360-r226102
Notes
Notes:
svn path=/vendor/llvm/dist/; revision=277323
svn path=/vendor/llvm/llvm-release_360-r226102/; revision=277324; tag=vendor/llvm/llvm-release_360-r226102
Diffstat (limited to 'test/CodeGen/R600/fsub.ll')
-rw-r--r-- | test/CodeGen/R600/fsub.ll | 93 |
1 files changed, 59 insertions, 34 deletions
diff --git a/test/CodeGen/R600/fsub.ll b/test/CodeGen/R600/fsub.ll index 4f74efba4d8b..4fe47e7badf3 100644 --- a/test/CodeGen/R600/fsub.ll +++ b/test/CodeGen/R600/fsub.ll @@ -1,14 +1,25 @@ -; RUN: llc < %s -march=r600 -mcpu=redwood | FileCheck %s --check-prefix=R600-CHECK -; RUN: llc < %s -march=r600 -mcpu=SI -verify-machineinstrs | FileCheck %s --check-prefix=SI-CHECK - -; R600-CHECK: @fsub_f32 -; R600-CHECK: ADD {{\** *}}T{{[0-9]+\.[XYZW]}}, KC0[2].Z, -KC0[2].W -; SI-CHECK: @fsub_f32 -; SI-CHECK: V_SUB_F32 -define void @fsub_f32(float addrspace(1)* %out, float %a, float %b) { -entry: - %0 = fsub float %a, %b - store float %0, float addrspace(1)* %out +; RUN: llc -march=r600 -mcpu=redwood < %s | FileCheck -check-prefix=R600 -check-prefix=FUNC %s +; RUN: llc -march=amdgcn -mcpu=SI -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s + + +; FUNC-LABEL: {{^}}v_fsub_f32: +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} +define void @v_fsub_f32(float addrspace(1)* %out, float addrspace(1)* %in) { + %b_ptr = getelementptr float addrspace(1)* %in, i32 1 + %a = load float addrspace(1)* %in, align 4 + %b = load float addrspace(1)* %b_ptr, align 4 + %result = fsub float %a, %b + store float %result, float addrspace(1)* %out, align 4 + ret void +} + +; FUNC-LABEL: {{^}}s_fsub_f32: +; R600: ADD {{\** *}}T{{[0-9]+\.[XYZW]}}, KC0[2].Z, -KC0[2].W + +; SI: v_sub_f32_e32 {{v[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}} +define void @s_fsub_f32(float addrspace(1)* %out, float %a, float %b) { + %sub = fsub float %a, %b + store float %sub, float addrspace(1)* %out, align 4 ret void } @@ -16,34 +27,48 @@ declare float @llvm.R600.load.input(i32) readnone declare void @llvm.AMDGPU.store.output(float, i32) -; R600-CHECK: @fsub_v2f32 -; R600-CHECK-DAG: ADD {{\** *}}T{{[0-9]+\.[XYZW]}}, KC0[3].X, -KC0[3].Z -; R600-CHECK-DAG: ADD {{\** *}}T{{[0-9]+\.[XYZW]}}, KC0[2].W, -KC0[3].Y -; SI-CHECK: @fsub_v2f32 -; SI-CHECK: V_SUB_F32 -; SI-CHECK: V_SUB_F32 +; FUNC-LABEL: {{^}}fsub_v2f32: +; R600-DAG: ADD {{\** *}}T{{[0-9]+\.[XYZW]}}, KC0[3].X, -KC0[3].Z +; R600-DAG: ADD {{\** *}}T{{[0-9]+\.[XYZW]}}, KC0[2].W, -KC0[3].Y + +; FIXME: Should be using SGPR directly for first operand +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} define void @fsub_v2f32(<2 x float> addrspace(1)* %out, <2 x float> %a, <2 x float> %b) { -entry: - %0 = fsub <2 x float> %a, %b - store <2 x float> %0, <2 x float> addrspace(1)* %out + %sub = fsub <2 x float> %a, %b + store <2 x float> %sub, <2 x float> addrspace(1)* %out, align 8 ret void } -; R600-CHECK: @fsub_v4f32 -; R600-CHECK: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW], -T[0-9]+\.[XYZW]}} -; R600-CHECK: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW], -T[0-9]+\.[XYZW]}} -; R600-CHECK: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW], -T[0-9]+\.[XYZW]}} -; R600-CHECK: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW], -T[0-9]+\.[XYZW]}} -; SI-CHECK: @fsub_v4f32 -; SI-CHECK: V_SUB_F32 -; SI-CHECK: V_SUB_F32 -; SI-CHECK: V_SUB_F32 -; SI-CHECK: V_SUB_F32 -define void @fsub_v4f32(<4 x float> addrspace(1)* %out, <4 x float> addrspace(1)* %in) { +; FUNC-LABEL: {{^}}v_fsub_v4f32: +; R600: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW], -T[0-9]+\.[XYZW]}} +; R600: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW], -T[0-9]+\.[XYZW]}} +; R600: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW], -T[0-9]+\.[XYZW]}} +; R600: ADD {{\** *}}T{{[0-9]+\.[XYZW], T[0-9]+\.[XYZW], -T[0-9]+\.[XYZW]}} + +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} +define void @v_fsub_v4f32(<4 x float> addrspace(1)* %out, <4 x float> addrspace(1)* %in) { %b_ptr = getelementptr <4 x float> addrspace(1)* %in, i32 1 - %a = load <4 x float> addrspace(1) * %in - %b = load <4 x float> addrspace(1) * %b_ptr + %a = load <4 x float> addrspace(1)* %in, align 16 + %b = load <4 x float> addrspace(1)* %b_ptr, align 16 + %result = fsub <4 x float> %a, %b + store <4 x float> %result, <4 x float> addrspace(1)* %out, align 16 + ret void +} + +; FIXME: Should be using SGPR directly for first operand + +; FUNC-LABEL: {{^}}s_fsub_v4f32: +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} +; SI: v_subrev_f32_e32 {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}} +; SI: s_endpgm +define void @s_fsub_v4f32(<4 x float> addrspace(1)* %out, <4 x float> %a, <4 x float> %b) { %result = fsub <4 x float> %a, %b - store <4 x float> %result, <4 x float> addrspace(1)* %out + store <4 x float> %result, <4 x float> addrspace(1)* %out, align 16 ret void } |