Skip to content

Commit c7f25b6

Browse files
committed
[AMDGPU] Updated some tests to run on gfx940. NFC.
1 parent 895653d commit c7f25b6

File tree

5 files changed

+18
-9
lines changed

5 files changed

+18
-9
lines changed

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.fdot2.ll

Lines changed: 6 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,11 +1,12 @@
1-
; RUN: llc -march=amdgcn -mcpu=gfx906 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX906
1+
; RUN: llc -march=amdgcn -mcpu=gfx906 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX9,GFX906
2+
; RUN: llc -march=amdgcn -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX9,GFX940
23
; RUN: llc -march=amdgcn -mcpu=gfx1011 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX10
34
; RUN: llc -march=amdgcn -mcpu=gfx1012 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX10
45

56
declare float @llvm.amdgcn.fdot2(<2 x half> %a, <2 x half> %b, float %c, i1 %clamp)
67

78
; GCN-LABEL: {{^}}test_llvm_amdgcn_fdot2_clamp
8-
; GFX906: v_dot2_f32_f16 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}} clamp{{$}}
9+
; GFX9: v_dot2_f32_f16 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}} clamp{{$}}
910
; GFX10: v_dot2_f32_f16 v{{[0-9]+}}, s{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}} clamp{{$}}
1011
define amdgpu_kernel void @test_llvm_amdgcn_fdot2_clamp(
1112
float addrspace(1)* %r,
@@ -23,6 +24,7 @@ entry:
2324

2425
; GCN-LABEL: {{^}}test_llvm_amdgcn_fdot2_no_clamp
2526
; GFX906: v_dot2_f32_f16 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}{{$}}
27+
; GFX940: v_dot2c_f32_f16_e32 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}{{$}}
2628
; GFX10: v_dot2c_f32_f16_e32 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}{{$}}
2729
define amdgpu_kernel void @test_llvm_amdgcn_fdot2_no_clamp(
2830
float addrspace(1)* %r,
@@ -38,8 +40,9 @@ entry:
3840
ret void
3941
}
4042

41-
; GFX906-LABEL: {{^}}fdot2_inline_literal
43+
; GFX9-LABEL: {{^}}fdot2_inline_literal
4244
; GFX906: v_dot2_f32_f16 v{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}, 1.0
45+
; GFX940: v_dot2c_f32_f16_e32 v{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}{{$}}
4346
define float @fdot2_inline_literal(<2 x half> %a, <2 x half> %b) {
4447
%ret = tail call float @llvm.amdgcn.fdot2(<2 x half> %a, <2 x half> %b, float 1.0, i1 false)
4548
ret float %ret

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.udot4.ll

Lines changed: 4 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,11 +1,12 @@
1-
; RUN: llc -march=amdgcn -mcpu=gfx906 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX906
1+
; RUN: llc -march=amdgcn -mcpu=gfx906 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX9
2+
; RUN: llc -march=amdgcn -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX9
23
; RUN: llc -march=amdgcn -mcpu=gfx1011 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX10
34
; RUN: llc -march=amdgcn -mcpu=gfx1012 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX10
45

56
declare i32 @llvm.amdgcn.udot4(i32 %a, i32 %b, i32 %c, i1 %clamp)
67

78
; GCN-LABEL: {{^}}test_llvm_amdgcn_udot4_clamp
8-
; GFX906: v_dot4_u32_u8 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}} clamp{{$}}
9+
; GFX9: v_dot4_u32_u8 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}} clamp{{$}}
910
; GFX10: v_dot4_u32_u8 v{{[0-9]+}}, s{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}} clamp{{$}}
1011
define amdgpu_kernel void @test_llvm_amdgcn_udot4_clamp(
1112
i32 addrspace(1)* %r,
@@ -24,7 +25,7 @@ entry:
2425
}
2526

2627
; GCN-LABEL: {{^}}test_llvm_amdgcn_udot4_no_clamp
27-
; GFX906: v_dot4_u32_u8 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}{{$}}
28+
; GFX9: v_dot4_u32_u8 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}{{$}}
2829
; GFX10: v_dot4_u32_u8 v{{[0-9]+}}, s{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}{{$}}
2930
define amdgpu_kernel void @test_llvm_amdgcn_udot4_no_clamp(
3031
i32 addrspace(1)* %r,

llvm/test/CodeGen/AMDGPU/llvm.amdgcn.udot8.ll

Lines changed: 4 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,11 +1,12 @@
1-
; RUN: llc -march=amdgcn -mcpu=gfx906 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX906
1+
; RUN: llc -march=amdgcn -mcpu=gfx906 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX9
2+
; RUN: llc -march=amdgcn -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX9
23
; RUN: llc -march=amdgcn -mcpu=gfx1011 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX10
34
; RUN: llc -march=amdgcn -mcpu=gfx1012 -verify-machineinstrs < %s | FileCheck %s --check-prefixes=GCN,GFX10
45

56
declare i32 @llvm.amdgcn.udot8(i32 %a, i32 %b, i32 %c, i1 %clamp)
67

78
; GCN-LABEL: {{^}}test_llvm_amdgcn_udot8_clamp
8-
; GFX906: v_dot8_u32_u4 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}} clamp{{$}}
9+
; GFX9: v_dot8_u32_u4 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}} clamp{{$}}
910
; GFX10: v_dot8_u32_u4 v{{[0-9]+}}, s{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}} clamp{{$}}
1011
define amdgpu_kernel void @test_llvm_amdgcn_udot8_clamp(
1112
i32 addrspace(1)* %r,
@@ -24,7 +25,7 @@ entry:
2425
}
2526

2627
; GCN-LABEL: {{^}}test_llvm_amdgcn_udot8_no_clamp
27-
; GFX906: v_dot8_u32_u4 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}{{$}}
28+
; GFX9: v_dot8_u32_u4 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}{{$}}
2829
; GFX10: v_dot8_u32_u4 v{{[0-9]+}}, s{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}{{$}}
2930
define amdgpu_kernel void @test_llvm_amdgcn_udot8_no_clamp(
3031
i32 addrspace(1)* %r,

llvm/test/CodeGen/AMDGPU/mfma-cd-select.ll

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,8 @@
11
; RUN: llc -march=amdgcn -mcpu=gfx908 -verify-machineinstrs < %s | FileCheck --enable-var-scope --check-prefixes=GCN,GFX908 %s
22
; RUN: llc -march=amdgcn -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck --enable-var-scope --check-prefixes=GCN,GFX90A %s
33
; RUN: llc -global-isel -march=amdgcn -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck --enable-var-scope --check-prefixes=GCN,GFX90A %s
4+
; RUN: llc -march=amdgcn -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck --enable-var-scope --check-prefixes=GCN,GFX90A %s
5+
; RUN: llc -global-isel -march=amdgcn -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck --enable-var-scope --check-prefixes=GCN,GFX90A %s
46

57
declare <32 x float> @llvm.amdgcn.mfma.f32.32x32x1f32(float, float, <32 x float>, i32, i32, i32)
68

llvm/test/CodeGen/AMDGPU/mfma-vgpr-cd-select.ll

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,7 @@
11
; RUN: llc -march=amdgcn -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck --enable-var-scope --check-prefixes=GCN %s
22
; RUN: llc -global-isel -march=amdgcn -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck --enable-var-scope --check-prefixes=GCN %s
3+
; RUN: llc -march=amdgcn -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck --enable-var-scope --check-prefixes=GCN %s
4+
; RUN: llc -global-isel -march=amdgcn -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck --enable-var-scope --check-prefixes=GCN %s
35

46
declare <32 x float> @llvm.amdgcn.mfma.f32.32x32x1f32(float, float, <32 x float>, i32, i32, i32)
57
declare <16 x float> @llvm.amdgcn.mfma.f32.16x16x1f32(float, float, <16 x float>, i32, i32, i32)

0 commit comments

Comments
 (0)