[llvm] [AMDGPU] Add & use amdgcn-verify-memory-legalizer (PR #78697)

via llvm-commits llvm-commits at lists.llvm.org
Fri Jan 19 03:22:37 PST 2024 

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-llvm-globalisel

Author: Pierre van Houtryve (Pierre-vh)

<details>
<summary>Changes</summary>

SIMemoryLegalizer tests were slow, with most of them taking 4.5 to 5.3s to complete and that's on a fast machine. I also  recall  seeing them in the slowest tests list on build bots.

I propose to only verify MIR before & after the memory legalizer runs in the memory legalizer tests. This greatly speeds things up, with the slowest test now taking barely 2.08s instead of 5.28s on my machine. I also don't think this has any significant drawbacks.

---

Patch is 84.90 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/78697.diff


28 Files Affected:

- (modified) llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp (+11) 
- (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/memory-legalizer-atomic-fence.ll (+7-7) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-fence.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-agent.ll (+10-11) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-nontemporal.ll (+10-10) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-singlethread.ll (+10-11) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-system.ll (+10-11) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-volatile.ll (+6-6) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-wavefront.ll (+10-11) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-workgroup.ll (+10-11) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-global-agent.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-global-nontemporal.ll (+11-11) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-global-singlethread.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-global-system.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-global-volatile.ll (+7-7) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-global-wavefront.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-global-workgroup.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-invalid-syncscope.ll (+4-4) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-local-agent.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-local-nontemporal.ll (+11-11) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-local-singlethread.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-local-system.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-local-volatile.ll (+7-7) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-local-wavefront.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-local-workgroup.ll (+11-12) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-private-nontemporal.ll (+13-13) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-private-volatile.ll (+9-9) 
- (modified) llvm/test/CodeGen/AMDGPU/memory-legalizer-store-infinite-loop.ll (+1-1) 


``````````diff
diff --git a/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp b/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp
index 84b9330ef9633eb..c9552d709931a87 100644
--- a/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp
+++ b/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp
@@ -34,6 +34,10 @@ static cl::opt<bool> AmdgcnSkipCacheInvalidations(
     "amdgcn-skip-cache-invalidations", cl::init(false), cl::Hidden,
     cl::desc("Use this to skip inserting cache invalidating instructions."));
 
+static cl::opt<bool> VerifyMemoryLegalizer(
+    "amdgcn-verify-memory-legalizer", cl::init(false), cl::Hidden,
+    cl::desc("Verify MIR before and after SIMemoryLegalizer"));
+
 namespace {
 
 LLVM_ENABLE_BITMASK_ENUMS_IN_NAMESPACE();
@@ -2567,6 +2571,9 @@ bool SIMemoryLegalizer::runOnMachineFunction(MachineFunction &MF) {
   SIMemOpAccess MOA(MF);
   CC = SICacheControl::create(MF.getSubtarget<GCNSubtarget>());
 
+  if (VerifyMemoryLegalizer)
+    MF.verify();
+
   for (auto &MBB : MF) {
     for (auto MI = MBB.begin(); MI != MBB.end(); ++MI) {
 
@@ -2601,6 +2608,10 @@ bool SIMemoryLegalizer::runOnMachineFunction(MachineFunction &MF) {
   }
 
   Changed |= removeAtomicPseudoMIs();
+
+  if (VerifyMemoryLegalizer)
+    MF.verify();
+
   return Changed;
 }
 
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/memory-legalizer-atomic-fence.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/memory-legalizer-atomic-fence.ll
index 4cb83040bfe738a..4966edfe8a59a75 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/memory-legalizer-atomic-fence.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/memory-legalizer-atomic-fence.ll
@@ -1,11 +1,11 @@
 ; NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 4
-; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd- -mcpu=gfx600 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX6 %s
-; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd- -mcpu=gfx803 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX8 %s
-; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd-amdhsa -mcpu=gfx803 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX8 %s
-; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10WGP %s
-; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -verify-machineinstrs < %s | FileCheck -check-prefix=GFX10CU %s
-; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -check-prefix=GFX11WGP %s
-; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -verify-machineinstrs < %s | FileCheck -check-prefix=GFX11CU %s
+; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd- -mcpu=gfx600 -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefix=GFX6 %s
+; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd- -mcpu=gfx803 -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefix=GFX8 %s
+; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd-amdhsa -mcpu=gfx803 -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefix=GFX8 %s
+; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefix=GFX10WGP %s
+; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefix=GFX10CU %s
+; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefix=GFX11WGP %s
+; RUN: llc -global-isel -stop-after=si-memory-legalizer -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefix=GFX11CU %s
 
 ; Note: we use MIR test checks + stop after legalizer to prevent
 ; tests from being optimized out.
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-fence.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-fence.ll
index d950c24192f39ca..c7eb53c13a97744 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-fence.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-fence.ll
@@ -1,15 +1,15 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx600 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX6 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX7 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX10-WGP %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX10-CU %s
-; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx700 -amdgcn-skip-cache-invalidations -verify-machineinstrs < %s | FileCheck --check-prefixes=SKIP-CACHE-INV %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX90A-NOTTGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -mattr=+tgsplit -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX90A-TGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX940-NOTTGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -mattr=+tgsplit -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX940-TGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX11-WGP %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX11-CU %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx600 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX6 %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX7 %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX10-WGP %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX10-CU %s
+; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx700 -amdgcn-skip-cache-invalidations -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=SKIP-CACHE-INV %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX90A-NOTTGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -mattr=+tgsplit -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX90A-TGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX940-NOTTGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -mattr=+tgsplit -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX940-TGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX11-WGP %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX11-CU %s
 
 define amdgpu_kernel void @singlethread_acquire_fence() {
 ; GFX6-LABEL: singlethread_acquire_fence:
@@ -2210,4 +2210,3 @@ entry:
   fence syncscope("one-as") seq_cst
   ret void
 }
-
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-agent.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-agent.ll
index 6989c068794eecd..28d159046e5c42c 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-agent.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-agent.ll
@@ -1,14 +1,14 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX7 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX10-WGP %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX10-CU %s
-; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx700 -amdgcn-skip-cache-invalidations -verify-machineinstrs < %s | FileCheck --check-prefixes=SKIP-CACHE-INV %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX90A-NOTTGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -mattr=+tgsplit -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX90A-TGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX940-NOTTGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -mattr=+tgsplit -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX940-TGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX11-WGP %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX11-CU %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX7 %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX10-WGP %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX10-CU %s
+; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx700 -amdgcn-skip-cache-invalidations -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=SKIP-CACHE-INV %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX90A-NOTTGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -mattr=+tgsplit -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX90A-TGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX940-NOTTGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -mattr=+tgsplit -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX940-TGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX11-WGP %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX11-CU %s
 
 define amdgpu_kernel void @flat_agent_unordered_load(
 ; GFX7-LABEL: flat_agent_unordered_load:
@@ -13140,4 +13140,3 @@ entry:
   store i32 %val0, ptr %out, align 4
   ret void
 }
-
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-nontemporal.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-nontemporal.ll
index 6d48e36d91b1457..57005ab21f663fe 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-nontemporal.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-nontemporal.ll
@@ -1,14 +1,14 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX7 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX10-WGP %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX10-CU %s
-; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx700 -amdgcn-skip-cache-invalidations -verify-machineinstrs < %s | FileCheck --check-prefixes=SKIP-CACHE-INV %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX90A-NOTTGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -mattr=+tgsplit -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX90A-TGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX940-NOTTGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -mattr=+tgsplit -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX940-TGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX11-WGP %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX11-CU %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX7 %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX10-WGP %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX10-CU %s
+; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx700 -amdgcn-skip-cache-invalidations -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=SKIP-CACHE-INV %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX90A-NOTTGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -mattr=+tgsplit -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX90A-TGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX940-NOTTGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -mattr=+tgsplit -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX940-TGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX11-WGP %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX11-CU %s
 
 define amdgpu_kernel void @flat_nontemporal_load_0(
 ; GFX7-LABEL: flat_nontemporal_load_0:
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-singlethread.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-singlethread.ll
index 8720186ab02f3c0..a01c12f81ccde32 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-singlethread.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-singlethread.ll
@@ -1,14 +1,14 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX7 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX10-WGP %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX10-CU %s
-; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx700 -amdgcn-skip-cache-invalidations -verify-machineinstrs < %s | FileCheck --check-prefixes=SKIP-CACHE-INV %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX90A-NOTTGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -mattr=+tgsplit -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX90A-TGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX940-NOTTGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -mattr=+tgsplit -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX940-TGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX11-WGP %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX11-CU %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX7 %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX10-WGP %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX10-CU %s
+; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx700 -amdgcn-skip-cache-invalidations -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=SKIP-CACHE-INV %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX90A-NOTTGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -mattr=+tgsplit -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX90A-TGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX940-NOTTGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -mattr=+tgsplit -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX940-TGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX11-WGP %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX11-CU %s
 
 define amdgpu_kernel void @flat_singlethread_unordered_load(
 ; GFX7-LABEL: flat_singlethread_unordered_load:
@@ -11615,4 +11615,3 @@ entry:
   store i32 %val0, ptr %out, align 4
   ret void
 }
-
diff --git a/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-system.ll b/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-system.ll
index a3d73bbd9ff004e..be1e84be2bd13a6 100644
--- a/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-system.ll
+++ b/llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-system.ll
@@ -1,14 +1,14 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX7 %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX10-WGP %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX10-CU %s
-; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx700 -amdgcn-skip-cache-invalidations -verify-machineinstrs < %s | FileCheck --check-prefixes=SKIP-CACHE-INV %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX90A-NOTTGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -mattr=+tgsplit -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX90A-TGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX940-NOTTGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -mattr=+tgsplit -verify-machineinstrs < %s | FileCheck -check-prefixes=GFX940-TGSPLIT %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX11-WGP %s
-; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -mattr=+cumode -verify-machineinstrs < %s | FileCheck --check-prefixes=GFX11-CU %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx700 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX7 %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX10-WGP %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -mattr=+cumode -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=GFX10-CU %s
+; RUN: llc -mtriple=amdgcn-amd-amdpal -mcpu=gfx700 -amdgcn-skip-cache-invalidations -amdgcn-verify-memory-legalizer < %s | FileCheck --check-prefixes=SKIP-CACHE-INV %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX90A-NOTTGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -mattr=+tgsplit -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX90A-TGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx940 -amdgcn-verify-memory-legalizer < %s | FileCheck -check-prefixes=GFX940-NOTTGSPLIT %s
+; RUN: llc -mtriple=amdgcn-amd-amdhs...
[truncated]

``````````

</details>


https://github.com/llvm/llvm-project/pull/78697

More information about the llvm-commits mailing list