[llvm] [AMDGPU] Add GFX12 test coverage for export instructions (PR #100415)
via llvm-commits
llvm-commits at lists.llvm.org
Wed Jul 24 09:29:01 PDT 2024
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-backend-amdgpu
Author: Jay Foad (jayfoad)
<details>
<summary>Changes</summary>
---
Patch is 38.15 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/100415.diff
3 Files Affected:
- (modified) llvm/test/CodeGen/AMDGPU/llvm.amdgcn.exp.ll (+89-88)
- (modified) llvm/test/CodeGen/AMDGPU/llvm.amdgcn.exp.prim.ll (+2-1)
- (modified) llvm/test/CodeGen/AMDGPU/llvm.amdgcn.exp.row.ll (+67)
``````````diff
diff --git a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.exp.ll b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.exp.ll
index 67e4feed21fac..c120c58ce6f33 100644
--- a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.exp.ll
+++ b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.exp.ll
@@ -1,14 +1,15 @@
; RUN: llc -mtriple=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -strict-whitespace -check-prefixes=GCN,GFX8,PREGFX11 %s
; RUN: llc -mtriple=amdgcn -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -strict-whitespace -check-prefixes=GCN,GFX10,PREGFX11 %s
; RUN: llc -mtriple=amdgcn -mcpu=gfx1100 -amdgpu-enable-vopd=0 -verify-machineinstrs < %s | FileCheck -strict-whitespace -check-prefixes=GCN,GFX11 %s
+; RUN: llc -mtriple=amdgcn -mcpu=gfx1200 -amdgpu-enable-vopd=0 -verify-machineinstrs < %s | FileCheck -strict-whitespace -check-prefixes=GCN,GFX11 %s
declare void @llvm.amdgcn.exp.f32(i32, i32, float, float, float, float, i1, i1) #1
declare void @llvm.amdgcn.exp.i32(i32, i32, i32, i32, i32, i32, i1, i1) #1
declare float @llvm.amdgcn.raw.ptr.buffer.load.f32(ptr addrspace(8), i32, i32, i32) #2
; GCN-LABEL: {{^}}test_export_zeroes_f32:
-; GCN: exp mrt0 off, off, off, off{{$}}
-; GCN: exp mrt0 off, off, off, off done{{$}}
+; GCN: {{exp|export}} mrt0 off, off, off, off{{$}}
+; GCN: {{exp|export}} mrt0 off, off, off, off done{{$}}
define amdgpu_kernel void @test_export_zeroes_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 0, i32 0, float 0.0, float 0.0, float 0.0, float 0.0, i1 false, i1 false)
@@ -23,7 +24,7 @@ define amdgpu_kernel void @test_export_zeroes_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp mrt0 [[SRC0]], off, off, off done{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], off, off, off done{{$}}
define amdgpu_kernel void @test_export_en_src0_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 0, i32 1, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
ret void
@@ -34,7 +35,7 @@ define amdgpu_kernel void @test_export_en_src0_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp mrt0 off, [[SRC1]], off, off done{{$}}
+; GCN: {{exp|export}} mrt0 off, [[SRC1]], off, off done{{$}}
define amdgpu_kernel void @test_export_en_src1_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 0, i32 2, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
ret void
@@ -45,7 +46,7 @@ define amdgpu_kernel void @test_export_en_src1_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp mrt0 off, off, [[SRC2]], off done{{$}}
+; GCN: {{exp|export}} mrt0 off, off, [[SRC2]], off done{{$}}
define amdgpu_kernel void @test_export_en_src2_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 0, i32 4, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
ret void
@@ -56,7 +57,7 @@ define amdgpu_kernel void @test_export_en_src2_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp mrt0 off, off, off, [[SRC3]] done{{$}}
+; GCN: {{exp|export}} mrt0 off, off, off, [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_en_src3_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 0, i32 8, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
ret void
@@ -67,7 +68,7 @@ define amdgpu_kernel void @test_export_en_src3_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp mrt0 [[SRC0]], [[SRC1]], off, off done{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], [[SRC1]], off, off done{{$}}
define amdgpu_kernel void @test_export_en_src0_src1_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 0, i32 3, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
ret void
@@ -78,7 +79,7 @@ define amdgpu_kernel void @test_export_en_src0_src1_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp mrt0 [[SRC0]], off, [[SRC2]], off done{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], off, [[SRC2]], off done{{$}}
define amdgpu_kernel void @test_export_en_src0_src2_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 0, i32 5, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
ret void
@@ -89,8 +90,8 @@ define amdgpu_kernel void @test_export_en_src0_src2_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp mrt0 [[SRC0]], off, off, [[SRC3]]{{$}}
-; GCN: exp mrt0 [[SRC0]], off, off, [[SRC3]] done{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], off, off, [[SRC3]]{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], off, off, [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_en_src0_src3_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 0, i32 9, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 0, i32 9, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
@@ -102,8 +103,8 @@ define amdgpu_kernel void @test_export_en_src0_src3_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; GCN: exp mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_en_src0_src1_src2_src3_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
@@ -112,8 +113,8 @@ define amdgpu_kernel void @test_export_en_src0_src1_src2_src3_f32() #0 {
; GCN-LABEL: {{^}}test_export_mrt7_f32:
; GCN-DAG: v_mov_b32_e32 [[VHALF:v[0-9]+]], 0.5
-; GCN: exp mrt7 [[VHALF]], [[VHALF]], [[VHALF]], [[VHALF]]{{$}}
-; GCN: exp mrt7 [[VHALF]], [[VHALF]], [[VHALF]], [[VHALF]] done{{$}}
+; GCN: {{exp|export}} mrt7 [[VHALF]], [[VHALF]], [[VHALF]], [[VHALF]]{{$}}
+; GCN: {{exp|export}} mrt7 [[VHALF]], [[VHALF]], [[VHALF]], [[VHALF]] done{{$}}
define amdgpu_kernel void @test_export_mrt7_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 7, i32 15, float 0.5, float 0.5, float 0.5, float 0.5, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 7, i32 15, float 0.5, float 0.5, float 0.5, float 0.5, i1 true, i1 false)
@@ -125,8 +126,8 @@ define amdgpu_kernel void @test_export_mrt7_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp mrtz [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; GCN: exp mrtz [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; GCN: {{exp|export}} mrtz [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; GCN: {{exp|export}} mrtz [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_z_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 8, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 8, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
@@ -138,8 +139,8 @@ define amdgpu_kernel void @test_export_z_f32() #0 {
; PREGFX11-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; PREGFX11-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; PREGFX11-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; PREGFX11: exp null [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; PREGFX11: exp null [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; PREGFX11: {{exp|export}} null [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; PREGFX11: {{exp|export}} null [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_null_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 9, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 9, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
@@ -151,8 +152,8 @@ define amdgpu_kernel void @test_export_null_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp invalid_target_10 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; GCN: exp invalid_target_10 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; GCN: {{exp|export}} invalid_target_10 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; GCN: {{exp|export}} invalid_target_10 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_reserved10_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 10, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 10, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
@@ -164,8 +165,8 @@ define amdgpu_kernel void @test_export_reserved10_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp invalid_target_11 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; GCN: exp invalid_target_11 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; GCN: {{exp|export}} invalid_target_11 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; GCN: {{exp|export}} invalid_target_11 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_reserved11_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 11, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 11, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
@@ -177,8 +178,8 @@ define amdgpu_kernel void @test_export_reserved11_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp pos0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; GCN: exp pos0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; GCN: {{exp|export}} pos0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; GCN: {{exp|export}} pos0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_pos0_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 12, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 12, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
@@ -190,8 +191,8 @@ define amdgpu_kernel void @test_export_pos0_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; GCN: exp pos3 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; GCN: exp pos3 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; GCN: {{exp|export}} pos3 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; GCN: {{exp|export}} pos3 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_pos3_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 15, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 15, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
@@ -203,8 +204,8 @@ define amdgpu_kernel void @test_export_pos3_f32() #0 {
; PREGFX11-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; PREGFX11-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; PREGFX11-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; PREGFX11: exp param0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; PREGFX11: exp param0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; PREGFX11: {{exp|export}} param0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; PREGFX11: {{exp|export}} param0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_param0_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 32, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 32, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
@@ -216,8 +217,8 @@ define amdgpu_kernel void @test_export_param0_f32() #0 {
; PREGFX11-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; PREGFX11-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; PREGFX11-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; PREGFX11: exp param31 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; PREGFX11: exp param31 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; PREGFX11: {{exp|export}} param31 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; PREGFX11: {{exp|export}} param31 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_param31_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 63, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 false)
call void @llvm.amdgcn.exp.f32(i32 63, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 false)
@@ -229,10 +230,10 @@ define amdgpu_kernel void @test_export_param31_f32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2.0
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 0.5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4.0
-; PREGFX11: exp mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] vm{{$}}
-; PREGFX11: exp mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done vm{{$}}
-; GFX11: exp mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; GFX11: exp mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; PREGFX11: {{exp|export}} mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] vm{{$}}
+; PREGFX11: {{exp|export}} mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done vm{{$}}
+; GFX11: {{exp|export}} mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; GFX11: {{exp|export}} mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_vm_f32() #0 {
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 false, i1 true)
call void @llvm.amdgcn.exp.f32(i32 0, i32 15, float 1.0, float 2.0, float 0.5, float 4.0, i1 true, i1 true)
@@ -254,8 +255,8 @@ define amdgpu_kernel void @test_export_vm_f32() #0 {
; GCN-LABEL: {{^}}test_export_zeroes_i32:
-; GCN: exp mrt0 off, off, off, off{{$}}
-; GCN: exp mrt0 off, off, off, off done{{$}}
+; GCN: {{exp|export}} mrt0 off, off, off, off{{$}}
+; GCN: {{exp|export}} mrt0 off, off, off, off done{{$}}
define amdgpu_kernel void @test_export_zeroes_i32() #0 {
call void @llvm.amdgcn.exp.i32(i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i1 false, i1 false)
@@ -270,7 +271,7 @@ define amdgpu_kernel void @test_export_zeroes_i32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4
-; GCN: exp mrt0 [[SRC0]], off, off, off done{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], off, off, off done{{$}}
define amdgpu_kernel void @test_export_en_src0_i32() #0 {
call void @llvm.amdgcn.exp.i32(i32 0, i32 1, i32 1, i32 2, i32 5, i32 4, i1 true, i1 false)
ret void
@@ -281,7 +282,7 @@ define amdgpu_kernel void @test_export_en_src0_i32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4
-; GCN: exp mrt0 off, [[SRC1]], off, off done{{$}}
+; GCN: {{exp|export}} mrt0 off, [[SRC1]], off, off done{{$}}
define amdgpu_kernel void @test_export_en_src1_i32() #0 {
call void @llvm.amdgcn.exp.i32(i32 0, i32 2, i32 1, i32 2, i32 5, i32 4, i1 true, i1 false)
ret void
@@ -292,7 +293,7 @@ define amdgpu_kernel void @test_export_en_src1_i32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4
-; GCN: exp mrt0 off, off, [[SRC2]], off done{{$}}
+; GCN: {{exp|export}} mrt0 off, off, [[SRC2]], off done{{$}}
define amdgpu_kernel void @test_export_en_src2_i32() #0 {
call void @llvm.amdgcn.exp.i32(i32 0, i32 4, i32 1, i32 2, i32 5, i32 4, i1 true, i1 false)
ret void
@@ -303,7 +304,7 @@ define amdgpu_kernel void @test_export_en_src2_i32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4
-; GCN: exp mrt0 off, off, off, [[SRC3]] done{{$}}
+; GCN: {{exp|export}} mrt0 off, off, off, [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_en_src3_i32() #0 {
call void @llvm.amdgcn.exp.i32(i32 0, i32 8, i32 1, i32 2, i32 5, i32 4, i1 true, i1 false)
ret void
@@ -314,7 +315,7 @@ define amdgpu_kernel void @test_export_en_src3_i32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4
-; GCN: exp mrt0 [[SRC0]], [[SRC1]], off, off done{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], [[SRC1]], off, off done{{$}}
define amdgpu_kernel void @test_export_en_src0_src1_i32() #0 {
call void @llvm.amdgcn.exp.i32(i32 0, i32 3, i32 1, i32 2, i32 5, i32 4, i1 true, i1 false)
ret void
@@ -325,7 +326,7 @@ define amdgpu_kernel void @test_export_en_src0_src1_i32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4
-; GCN: exp mrt0 [[SRC0]], off, [[SRC2]], off done{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], off, [[SRC2]], off done{{$}}
define amdgpu_kernel void @test_export_en_src0_src2_i32() #0 {
call void @llvm.amdgcn.exp.i32(i32 0, i32 5, i32 1, i32 2, i32 5, i32 4, i1 true, i1 false)
ret void
@@ -336,8 +337,8 @@ define amdgpu_kernel void @test_export_en_src0_src2_i32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4
-; GCN: exp mrt0 [[SRC0]], off, off, [[SRC3]]{{$}}
-; GCN: exp mrt0 [[SRC0]], off, off, [[SRC3]] done{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], off, off, [[SRC3]]{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], off, off, [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_en_src0_src3_i32() #0 {
call void @llvm.amdgcn.exp.i32(i32 0, i32 9, i32 1, i32 2, i32 5, i32 4, i1 false, i1 false)
call void @llvm.amdgcn.exp.i32(i32 0, i32 9, i32 1, i32 2, i32 5, i32 4, i1 true, i1 false)
@@ -349,8 +350,8 @@ define amdgpu_kernel void @test_export_en_src0_src3_i32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4
-; GCN: exp mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; GCN: exp mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
+; GCN: {{exp|export}} mrt0 [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]] done{{$}}
define amdgpu_kernel void @test_export_en_src0_src1_src2_src3_i32() #0 {
call void @llvm.amdgcn.exp.i32(i32 0, i32 15, i32 1, i32 2, i32 5, i32 4, i1 false, i1 false)
call void @llvm.amdgcn.exp.i32(i32 0, i32 15, i32 1, i32 2, i32 5, i32 4, i1 true, i1 false)
@@ -359,8 +360,8 @@ define amdgpu_kernel void @test_export_en_src0_src1_src2_src3_i32() #0 {
; GCN-LABEL: {{^}}test_export_mrt7_i32:
; GCN-DAG: v_mov_b32_e32 [[VHALF:v[0-9]+]], 5
-; GCN: exp mrt7 [[VHALF]], [[VHALF]], [[VHALF]], [[VHALF]]{{$}}
-; GCN: exp mrt7 [[VHALF]], [[VHALF]], [[VHALF]], [[VHALF]] done{{$}}
+; GCN: {{exp|export}} mrt7 [[VHALF]], [[VHALF]], [[VHALF]], [[VHALF]]{{$}}
+; GCN: {{exp|export}} mrt7 [[VHALF]], [[VHALF]], [[VHALF]], [[VHALF]] done{{$}}
define amdgpu_kernel void @test_export_mrt7_i32() #0 {
call void @llvm.amdgcn.exp.i32(i32 7, i32 15, i32 5, i32 5, i32 5, i32 5, i1 false, i1 false)
call void @llvm.amdgcn.exp.i32(i32 7, i32 15, i32 5, i32 5, i32 5, i32 5, i1 true, i1 false)
@@ -372,8 +373,8 @@ define amdgpu_kernel void @test_export_mrt7_i32() #0 {
; GCN-DAG: v_mov_b32_e32 [[SRC1:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[SRC2:v[0-9]+]], 5
; GCN-DAG: v_mov_b32_e32 [[SRC3:v[0-9]+]], 4
-; GCN: exp mrtz [[SRC0]], [[SRC1]], [[SRC2]], [[SRC3]]{{$}}
-; GCN: exp mrtz [[SRC0]], [[S...
[truncated]
``````````
</details>
https://github.com/llvm/llvm-project/pull/100415
More information about the llvm-commits
mailing list