Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[AMDGPU] Fix hard clausing for image instructions on gfx12 #90221

Merged
merged 2 commits into from
Apr 29, 2024
Merged

[AMDGPU] Fix hard clausing for image instructions on gfx12 #90221

merged 2 commits into from
Apr 29, 2024

Conversation

dstutt
Copy link
Collaborator

@dstutt dstutt commented Apr 26, 2024 •
edited by jayfoad
Loading

Also updated hard-clauses.mir to have separate versions for gfx11 and gfx12 since
the MIR instructions are different for each of them.

@dstutt dstutt requested a review from jayfoad April 26, 2024 15:21
@llvmbot
Copy link
Collaborator

llvmbot commented Apr 26, 2024

@llvm/pr-subscribers-backend-amdgpu

Author: David Stuttard (dstutt)

Changes

Also updated hard-clause.mir to have separate versions for gfx11 and gfx12 since
the MIR instructions are different for each of them.

Patch is 49.08 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/90221.diff

5 Files Affected:

  • (modified) llvm/lib/Target/AMDGPU/SIInstrInfo.cpp (+4-2)
  • (added) llvm/test/CodeGen/AMDGPU/hard-clauses-img-gfx11.mir (+40)
  • (added) llvm/test/CodeGen/AMDGPU/hard-clauses-img-gfx12.mir (+40)
  • (added) llvm/test/CodeGen/AMDGPU/hard-clauses-img.mir (+34)
  • (modified) llvm/test/CodeGen/AMDGPU/hard-clauses.mir (+184-51)
diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
index e20fe1b716b647..76b90042d65f7d 100644
--- a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -461,8 +461,10 @@ bool SIInstrInfo::getMemOperandsWithOffsetWidth(
     return true;
   }
 
-  if (isMIMG(LdSt)) {
-    int SRsrcIdx = AMDGPU::getNamedOperandIdx(Opc, AMDGPU::OpName::srsrc);
+  if (isImage(LdSt)) {
+    auto RsrcOpName =
+        isMIMG(LdSt) ? AMDGPU::OpName::srsrc : AMDGPU::OpName::rsrc;
+    int SRsrcIdx = AMDGPU::getNamedOperandIdx(Opc, RsrcOpName);
     BaseOps.push_back(&LdSt.getOperand(SRsrcIdx));
     int VAddr0Idx = AMDGPU::getNamedOperandIdx(Opc, AMDGPU::OpName::vaddr0);
     if (VAddr0Idx >= 0) {
diff --git a/llvm/test/CodeGen/AMDGPU/hard-clauses-img-gfx11.mir b/llvm/test/CodeGen/AMDGPU/hard-clauses-img-gfx11.mir
new file mode 100644
index 00000000000000..b22de06e68a7f3
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/hard-clauses-img-gfx11.mir
@@ -0,0 +1,40 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -mtriple=amdgcn -mcpu=gfx1100 -verify-machineinstrs -run-pass si-insert-hard-clauses %s -o - | FileCheck %s
+
+---
+name: mimg_nsa
+tracksRegLiveness: true
+body: |
+  bb.0:
+    liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-LABEL: name: mimg_nsa
+    ; CHECK: liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-NEXT: {{  $}}
+    ; CHECK-NEXT: BUNDLE implicit-def $vgpr10_vgpr11_vgpr12, implicit-def $vgpr10, implicit-def $vgpr10_lo16, implicit-def $vgpr10_hi16, implicit-def $vgpr11, implicit-def $vgpr11_lo16, implicit-def $vgpr11_hi16, implicit-def $vgpr12, implicit-def $vgpr12_lo16, implicit-def $vgpr12_hi16, implicit-def $vgpr10_vgpr11, implicit-def $vgpr11_vgpr12, implicit-def $vgpr20_vgpr21_vgpr22, implicit-def $vgpr20, implicit-def $vgpr20_lo16, implicit-def $vgpr20_hi16, implicit-def $vgpr21, implicit-def $vgpr21_lo16, implicit-def $vgpr21_hi16, implicit-def $vgpr22, implicit-def $vgpr22_lo16, implicit-def $vgpr22_hi16, implicit-def $vgpr20_vgpr21, implicit-def $vgpr21_vgpr22, implicit $vgpr3, implicit $vgpr8, implicit $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, implicit $sgpr8_sgpr9_sgpr10_sgpr11, implicit $exec {
+    ; CHECK-NEXT:   S_CLAUSE 1
+    ; CHECK-NEXT:   $vgpr10_vgpr11_vgpr12 = IMAGE_SAMPLE_LZ_V3_V2_nsa_gfx11 $vgpr3, $vgpr8, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    ; CHECK-NEXT:   $vgpr20_vgpr21_vgpr22 = IMAGE_SAMPLE_LZ_V3_V2_nsa_gfx11 $vgpr3, $vgpr8, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    ; CHECK-NEXT: }
+    $vgpr10_vgpr11_vgpr12 = IMAGE_SAMPLE_LZ_V3_V2_nsa_gfx11 $vgpr3, $vgpr8, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+    $vgpr20_vgpr21_vgpr22 = IMAGE_SAMPLE_LZ_V3_V2_nsa_gfx11 $vgpr3, $vgpr8, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+...
+
+---
+name: mimg_nsa_mixed
+tracksRegLiveness: true
+body: |
+  bb.0:
+    liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-LABEL: name: mimg_nsa_mixed
+    ; CHECK: liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-NEXT: {{  $}}
+    ; CHECK-NEXT: BUNDLE implicit-def $vgpr10, implicit-def $vgpr10_lo16, implicit-def $vgpr10_hi16, implicit-def $vgpr14, implicit-def $vgpr14_lo16, implicit-def $vgpr14_hi16, implicit-def $vgpr20_vgpr21_vgpr22, implicit-def $vgpr20, implicit-def $vgpr20_lo16, implicit-def $vgpr20_hi16, implicit-def $vgpr21, implicit-def $vgpr21_lo16, implicit-def $vgpr21_hi16, implicit-def $vgpr22, implicit-def $vgpr22_lo16, implicit-def $vgpr22_hi16, implicit-def $vgpr20_vgpr21, implicit-def $vgpr21_vgpr22, implicit $vgpr3, implicit $vgpr8, implicit $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, implicit $sgpr8_sgpr9_sgpr10_sgpr11, implicit $exec, implicit $vgpr5_vgpr6 {
+    ; CHECK-NEXT:   S_CLAUSE 2
+    ; CHECK-NEXT:   $vgpr10 = IMAGE_SAMPLE_LZ_V1_V2_nsa_gfx11 $vgpr3, $vgpr8, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    ; CHECK-NEXT:   $vgpr14 = IMAGE_SAMPLE_LZ_V1_V2_gfx11 $vgpr5_vgpr6, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s128), addrspace 7)
+    ; CHECK-NEXT:   $vgpr20_vgpr21_vgpr22 = IMAGE_SAMPLE_LZ_V3_V2_nsa_gfx11 $vgpr3, $vgpr8, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    ; CHECK-NEXT: }
+    $vgpr10 = IMAGE_SAMPLE_LZ_V1_V2_nsa_gfx11 $vgpr3, $vgpr8, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+    $vgpr14 = IMAGE_SAMPLE_LZ_V1_V2_gfx11 $vgpr5_vgpr6, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s128), addrspace 7)
+    $vgpr20_vgpr21_vgpr22 = IMAGE_SAMPLE_LZ_V3_V2_nsa_gfx11 $vgpr3, $vgpr8, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+...
diff --git a/llvm/test/CodeGen/AMDGPU/hard-clauses-img-gfx12.mir b/llvm/test/CodeGen/AMDGPU/hard-clauses-img-gfx12.mir
new file mode 100644
index 00000000000000..243a84562ab305
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/hard-clauses-img-gfx12.mir
@@ -0,0 +1,40 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -mtriple=amdgcn -mcpu=gfx1100 -verify-machineinstrs -run-pass si-insert-hard-clauses %s -o - | FileCheck %s
+
+---
+name: mimg
+tracksRegLiveness: true
+body: |
+  bb.0:
+    liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-LABEL: name: mimg
+    ; CHECK: liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-NEXT: {{  $}}
+    ; CHECK-NEXT: BUNDLE implicit-def $vgpr10_vgpr11_vgpr12, implicit-def $vgpr10, implicit-def $vgpr10_lo16, implicit-def $vgpr10_hi16, implicit-def $vgpr11, implicit-def $vgpr11_lo16, implicit-def $vgpr11_hi16, implicit-def $vgpr12, implicit-def $vgpr12_lo16, implicit-def $vgpr12_hi16, implicit-def $vgpr10_vgpr11, implicit-def $vgpr11_vgpr12, implicit-def $vgpr20_vgpr21_vgpr22, implicit-def $vgpr20, implicit-def $vgpr20_lo16, implicit-def $vgpr20_hi16, implicit-def $vgpr21, implicit-def $vgpr21_lo16, implicit-def $vgpr21_hi16, implicit-def $vgpr22, implicit-def $vgpr22_lo16, implicit-def $vgpr22_hi16, implicit-def $vgpr20_vgpr21, implicit-def $vgpr21_vgpr22, implicit $vgpr3, implicit $vgpr4, implicit $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, implicit $sgpr8_sgpr9_sgpr10_sgpr11, implicit $exec {
+    ; CHECK-NEXT:   S_CLAUSE 1
+    ; CHECK-NEXT:   $vgpr10_vgpr11_vgpr12 = IMAGE_SAMPLE_LZ_V3_V2_gfx12 $vgpr3, $vgpr4, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    ; CHECK-NEXT:   $vgpr20_vgpr21_vgpr22 = IMAGE_SAMPLE_LZ_V3_V2_gfx12 $vgpr3, $vgpr4, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    ; CHECK-NEXT: }
+    $vgpr10_vgpr11_vgpr12 = IMAGE_SAMPLE_LZ_V3_V2_gfx12 $vgpr3, $vgpr4, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+    $vgpr20_vgpr21_vgpr22 = IMAGE_SAMPLE_LZ_V3_V2_gfx12 $vgpr3, $vgpr4, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+...
+
+---
+name: mimg_mixed
+tracksRegLiveness: true
+body: |
+  bb.0:
+    liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-LABEL: name: mimg_mixed
+    ; CHECK: liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-NEXT: {{  $}}
+    ; CHECK-NEXT: BUNDLE implicit-def $vgpr10, implicit-def $vgpr10_lo16, implicit-def $vgpr10_hi16, implicit-def $vgpr14, implicit-def $vgpr14_lo16, implicit-def $vgpr14_hi16, implicit-def $vgpr20_vgpr21_vgpr22, implicit-def $vgpr20, implicit-def $vgpr20_lo16, implicit-def $vgpr20_hi16, implicit-def $vgpr21, implicit-def $vgpr21_lo16, implicit-def $vgpr21_hi16, implicit-def $vgpr22, implicit-def $vgpr22_lo16, implicit-def $vgpr22_hi16, implicit-def $vgpr20_vgpr21, implicit-def $vgpr21_vgpr22, implicit $vgpr3, implicit $vgpr4, implicit $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, implicit $sgpr8_sgpr9_sgpr10_sgpr11, implicit $exec, implicit $vgpr5, implicit $vgpr6 {
+    ; CHECK-NEXT:   S_CLAUSE 2
+    ; CHECK-NEXT:   $vgpr10 = IMAGE_SAMPLE_LZ_V1_V2_gfx12 $vgpr3, $vgpr4, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    ; CHECK-NEXT:   $vgpr14 = IMAGE_SAMPLE_LZ_V1_V2_gfx12 $vgpr5, $vgpr6, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s128), addrspace 7)
+    ; CHECK-NEXT:   $vgpr20_vgpr21_vgpr22 = IMAGE_SAMPLE_LZ_V3_V2_gfx12 $vgpr3, $vgpr4, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    ; CHECK-NEXT: }
+    $vgpr10 = IMAGE_SAMPLE_LZ_V1_V2_gfx12 $vgpr3, $vgpr4, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+    $vgpr14 = IMAGE_SAMPLE_LZ_V1_V2_gfx12 $vgpr5, $vgpr6, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s128), addrspace 7)
+    $vgpr20_vgpr21_vgpr22 = IMAGE_SAMPLE_LZ_V3_V2_gfx12 $vgpr3, $vgpr4, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 14, 1, -1, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+...
diff --git a/llvm/test/CodeGen/AMDGPU/hard-clauses-img.mir b/llvm/test/CodeGen/AMDGPU/hard-clauses-img.mir
new file mode 100644
index 00000000000000..50eea4aebd5e96
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/hard-clauses-img.mir
@@ -0,0 +1,34 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -mtriple=amdgcn -mcpu=gfx1010 -verify-machineinstrs -run-pass si-insert-hard-clauses %s -o - | FileCheck %s
+
+---
+name: mimg_nsa
+tracksRegLiveness: true
+body: |
+  bb.0:
+    liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-LABEL: name: mimg_nsa
+    ; CHECK: liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-NEXT: {{  $}}
+    ; CHECK-NEXT: $vgpr10_vgpr11_vgpr12_vgpr13 = IMAGE_SAMPLE_D_V4_V9_nsa_gfx10 $vgpr3, $vgpr8, $vgpr7, $vgpr5, $vgpr4, $vgpr6, $vgpr0, $vgpr2, $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 15, 2, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    ; CHECK-NEXT: $vgpr20_vgpr21_vgpr22_vgpr23 = IMAGE_SAMPLE_D_V4_V9_nsa_gfx10 $vgpr3, $vgpr8, $vgpr7, $vgpr5, $vgpr4, $vgpr6, $vgpr0, $vgpr2, $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 15, 2, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    $vgpr10_vgpr11_vgpr12_vgpr13 = IMAGE_SAMPLE_D_V4_V9_nsa_gfx10 $vgpr3, $vgpr8, $vgpr7, $vgpr5, $vgpr4, $vgpr6, $vgpr0, $vgpr2, $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 15, 2, 0, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+    $vgpr20_vgpr21_vgpr22_vgpr23 = IMAGE_SAMPLE_D_V4_V9_nsa_gfx10 $vgpr3, $vgpr8, $vgpr7, $vgpr5, $vgpr4, $vgpr6, $vgpr0, $vgpr2, $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 15, 2, 0, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+...
+
+---
+name: mimg_nsa_mixed
+tracksRegLiveness: true
+body: |
+  bb.0:
+    liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-LABEL: name: mimg_nsa_mixed
+    ; CHECK: liveins: $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, $vgpr0, $vgpr1, $vgpr2, $vgpr3, $vgpr4, $vgpr5, $vgpr6, $vgpr7, $vgpr8
+    ; CHECK-NEXT: {{  $}}
+    ; CHECK-NEXT: $vgpr10_vgpr11_vgpr12_vgpr13 = IMAGE_SAMPLE_D_V4_V9_nsa_gfx10 $vgpr3, $vgpr8, $vgpr7, $vgpr5, $vgpr4, $vgpr6, $vgpr0, $vgpr2, $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 15, 2, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    ; CHECK-NEXT: $vgpr14 = IMAGE_SAMPLE_LZ_V1_V2_gfx10 $vgpr5_vgpr6, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s128), addrspace 7)
+    ; CHECK-NEXT: $vgpr20_vgpr21_vgpr22_vgpr23 = IMAGE_SAMPLE_D_V4_V9_nsa_gfx10 $vgpr3, $vgpr8, $vgpr7, $vgpr5, $vgpr4, $vgpr6, $vgpr0, $vgpr2, $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 15, 2, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (load (s128))
+    $vgpr10_vgpr11_vgpr12_vgpr13 = IMAGE_SAMPLE_D_V4_V9_nsa_gfx10 $vgpr3, $vgpr8, $vgpr7, $vgpr5, $vgpr4, $vgpr6, $vgpr0, $vgpr2, $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 15, 2, 0, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+    $vgpr14 = IMAGE_SAMPLE_LZ_V1_V2_gfx10 $vgpr5_vgpr6, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 1, 1, 0, 0, 0, 0, 0, 0, 0, implicit $exec :: (dereferenceable load (s128), addrspace 7)
+    $vgpr20_vgpr21_vgpr22_vgpr23 = IMAGE_SAMPLE_D_V4_V9_nsa_gfx10 $vgpr3, $vgpr8, $vgpr7, $vgpr5, $vgpr4, $vgpr6, $vgpr0, $vgpr2, $vgpr2, $sgpr0_sgpr1_sgpr2_sgpr3_sgpr4_sgpr5_sgpr6_sgpr7, $sgpr8_sgpr9_sgpr10_sgpr11, 15, 2, 0, 0, 0, 0, 0, 0, 0, implicit $exec  :: (load (s128))
+...
diff --git a/llvm/test/CodeGen/AMDGPU/hard-clauses.mir b/llvm/test/CodeGen/AMDGPU/hard-clauses.mir
index 1c6bdff51015ee..44b988a7121c79 100644
--- a/llvm/test/CodeGen/AMDGPU/hard-clauses.mir
+++ b/llvm/test/CodeGen/AMDGPU/hard-clauses.mir
@@ -1,6 +1,7 @@
 # NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
 # RUN: llc -mtriple=amdgcn -mcpu=gfx1010 -verify-machineinstrs -run-pass si-insert-hard-clauses %s -o - | FileCheck %s
 # RUN: llc -mtriple=amdgcn -mcpu=gfx1100 -verify-machineinstrs -run-pass si-insert-hard-clauses %s -o - | FileCheck %s -check-prefix=GFX11
+# RUN: llc -mtriple=amdgcn -mcpu=gfx1200 -verify-machineinstrs -run-pass si-insert-hard-clauses %s -o - | FileCheck %s -check-prefix=GFX12
 
 ---
 name: nop1
@@ -19,6 +20,12 @@ body: |
     ; GFX11-NEXT: {{  $}}
     ; GFX11-NEXT: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
     ; GFX11-NEXT: S_NOP 2
+    ;
+    ; GFX12-LABEL: name: nop1
+    ; GFX12: liveins: $sgpr0_sgpr1
+    ; GFX12-NEXT: {{  $}}
+    ; GFX12-NEXT: $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
+    ; GFX12-NEXT: S_NOP 2
     $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
     S_NOP 2
 ...
@@ -48,6 +55,16 @@ body: |
     ; GFX11-NEXT:   S_NOP 2
     ; GFX11-NEXT:   $sgpr3 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 4, 0
     ; GFX11-NEXT: }
+    ;
+    ; GFX12-LABEL: name: nop2
+    ; GFX12: liveins: $sgpr0_sgpr1
+    ; GFX12-NEXT: {{  $}}
+    ; GFX12-NEXT: BUNDLE implicit-def $sgpr2, implicit-def $sgpr2_lo16, implicit-def $sgpr2_hi16, implicit-def $sgpr3, implicit-def $sgpr3_lo16, implicit-def $sgpr3_hi16, implicit $sgpr0_sgpr1 {
+    ; GFX12-NEXT:   S_CLAUSE 2
+    ; GFX12-NEXT:   $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
+    ; GFX12-NEXT:   S_NOP 2
+    ; GFX12-NEXT:   $sgpr3 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 4, 0
+    ; GFX12-NEXT: }
     $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
     S_NOP 2
     $sgpr3 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 4, 0
@@ -80,6 +97,17 @@ body: |
     ; GFX11-NEXT:   $sgpr3 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 4, 0
     ; GFX11-NEXT: }
     ; GFX11-NEXT: S_NOP 2
+    ;
+    ; GFX12-LABEL: name: nop3
+    ; GFX12: liveins: $sgpr0_sgpr1
+    ; GFX12-NEXT: {{  $}}
+    ; GFX12-NEXT: BUNDLE implicit-def $sgpr2, implicit-def $sgpr2_lo16, implicit-def $sgpr2_hi16, implicit-def $sgpr3, implicit-def $sgpr3_lo16, implicit-def $sgpr3_hi16, implicit $sgpr0_sgpr1 {
+    ; GFX12-NEXT:   S_CLAUSE 2
+    ; GFX12-NEXT:   $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
+    ; GFX12-NEXT:   S_NOP 2
+    ; GFX12-NEXT:   $sgpr3 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 4, 0
+    ; GFX12-NEXT: }
+    ; GFX12-NEXT: S_NOP 2
     $sgpr2 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 0, 0
     S_NOP 2
     $sgpr3 = S_LOAD_DWORD_IMM $sgpr0_sgpr1, 4, 0
@@ -274,6 +302,99 @@ body: |
     ; GFX11-NEXT:   $vgpr79 = BUFFER_LOAD_DWORD_OFFEN $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 316, 0, 0, implicit $exec
     ; GFX11-NEXT:   $vgpr80 = BUFFER_LOAD_DWORD_OFFEN $vgpr0, $sgpr0_sgpr1_sgpr2_sgpr3, 0, 320, 0, 0, implicit $exec
     ; GFX11-NEXT: }
+    ;
+    ; GFX12-LABEL: name: long_clause
+    ; GFX12: liveins: $sgpr0_sgpr1_sgpr2_sgpr3, $vgpr0
+    ; GFX12-NEXT: {{  $}}
+    ; GFX12-NEXT: BUNDLE implicit-def $vgpr1, implicit-def $vgpr1_lo16, implicit-def $vgpr1_hi16, implicit-def $vgpr2, implicit-def $vgpr2_lo16, implicit-def $vgpr2_hi16, implicit-def $vgpr3, implicit-def $vgpr3_lo16, implicit-def $vgpr3_hi16, implicit-def $vgpr4, implicit-def $vgpr4_lo16, implicit-def $vgpr4_hi16, implicit-def $vgpr5, implicit-def $vgpr5_lo16, implicit-def $vgpr5_hi16, implicit-def $vgpr6, implicit-def $vgpr6_lo16, implicit-def $vgpr6_hi16, implicit-def $vgpr7, implicit-def $vgpr7_lo16, implicit-def $vgpr7_hi16, implicit-def $vgpr8, implicit-def $vgpr8_lo16, implicit-def $vgpr8_hi16, implicit-def $vgpr9, implicit-def $vgpr9_lo16, implicit-def $vgpr9_hi16, implicit-def $vgpr10, implicit-def $vgpr10_lo16, implicit-def $vgpr10_hi16, implicit-def $vgpr11, implicit-def $vgpr11_lo16, implicit-def $vgpr11_hi16, implicit-def $vgpr12, implicit-def $vgpr12_lo16, implicit-def $vgpr12_hi16, implicit-def $vgpr13, implicit-def $vgpr13_lo16, implicit-def $vgpr13_hi16, implicit-def $vgpr14, implicit-def $vgpr14_lo16, implicit-def $vgpr14_hi16, implicit-def $vgpr15, implicit-def $vgpr15_lo16, implicit-def $vgpr15_hi16, implicit-def $vgpr16, implicit-def $vgpr16_lo16, implicit-def $vgpr16_hi16, implicit-def $vgpr17, implicit-def $vgpr17_lo16, implicit-def $vgpr17_hi16, implicit-def $vgpr18, implicit-def $vgpr18_lo16, implicit-def $vgpr18_hi16, implicit-def $vgpr19, implicit-def $vgpr19_lo16, implicit-def $vgpr19_hi16, implicit-def $vgpr20, implicit-def $vgpr20_lo16, implicit-def $vgpr20_hi16, implicit-def $vgpr21, implicit-def $vgpr21_lo16, implicit-def $vgpr21_hi16, implicit-def $vgpr22, implicit-def $vgpr22_lo16, implicit-def $vgpr22_hi16, implicit-def $vgpr23, implicit-def $vgpr23_lo16, implicit-def $vgpr23_hi16, implicit-def $vgpr24, implicit-def $vgpr24_lo16, implicit-def $vgpr24_hi16, implicit-def $vgpr25, implicit-def $vgpr25_lo16, implicit-def $vgpr25_hi16, implicit-def $vgpr26, implicit-def $vgpr26_lo16, implicit-def $vgpr26_hi16, implicit-def $vgpr27, implicit-def $vgpr27_lo16, imp...
[truncated]

jayfoad
jayfoad approved these changes Apr 26, 2024
View reviewed changes
Copy link
Contributor

@jayfoad jayfoad left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM with nit

llvm/test/CodeGen/AMDGPU/hard-clauses-img.mir Outdated Show resolved Hide resolved
@dstutt
[AMDGPU] Fix hard clausing for image instructions on gfx12
be2cf64 
Also updated hard-clause.mir to have separate versions for gfx11 and gfx12 since
the MIR instructions are different for each of them.
@dstutt
Rename test in line with others in the group
544b541
@dstutt dstutt merged commit 2914a11 into llvm:main Apr 29, 2024
4 checks passed
@dstutt dstutt deleted the llvm-gfx12-img-clause branch April 29, 2024 10:42
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@jayfoad jayfoad jayfoad approved these changes

Assignees
No one assigned
Labels
backend:AMDGPU
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
3 participants
@dstutt @llvmbot @jayfoad