[clang] [llvm] [AMDGPU] Emit a waitcnt instruction after each memory instruction (PR #79236)

[via cfe-commits]

github-actions[bot] wrote:

<!--LLVM CODE FORMAT COMMENT: {clang-format}-->


:warning: C/C++ code formatter, clang-format found issues in your code. :warning:

<details>
<summary>
You can test this locally with the following command:
</summary>

``````````bash
git-clang-format --diff 7fc25928233c133a4af1dadf0e060fb5d42ebd4e bc7d09dedd199eb2ae739d625183d9370cac6436 -- clang/test/Driver/amdgpu-features.c llvm/lib/Target/AMDGPU/GCNSubtarget.h llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp
``````````

</details>

<details>
<summary>
View the diff from clang-format here.
</summary>

``````````diff
diff --git a/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp b/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp
index c8f02b2482..0e4468a9ea 100644
--- a/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp
+++ b/llvm/lib/Target/AMDGPU/SIMemoryLegalizer.cpp
@@ -2419,22 +2419,22 @@ bool SIGfx6CacheControl::handleNonAtomicForPreciseMemory(
   if (TII->isSMRD(Inst)) { // scalar
     if (Inst.mayStore())
       return false;
-    Wait.DsCnt = 0;                   // LgkmCnt
-  } else {                            // vector
-    if (Inst.mayLoad()) {             // vector load
-      if (TII->isVMEM(Inst))          // VMEM load
-        Wait.LoadCnt = 0;             // VmCnt
-      else if (TII->isFLAT(Inst)) {   // Flat load
-        Wait.LoadCnt = 0;             // VmCnt
-        Wait.DsCnt = 0;               // LgkmCnt
-      } else                          // LDS load
-        Wait.DsCnt = 0;               // LgkmCnt
-    } else {                          // vector store
-      if (TII->isVMEM(Inst))          // VMEM store
-        Wait.LoadCnt = 0;             // VmCnt
-      else if (TII->isFLAT(Inst)) {   // Flat store
-        Wait.LoadCnt = 0;             // VmCnt
-        Wait.DsCnt = 0;               // LgkmCnt
+    Wait.DsCnt = 0;                 // LgkmCnt
+  } else {                          // vector
+    if (Inst.mayLoad()) {           // vector load
+      if (TII->isVMEM(Inst))        // VMEM load
+        Wait.LoadCnt = 0;           // VmCnt
+      else if (TII->isFLAT(Inst)) { // Flat load
+        Wait.LoadCnt = 0;           // VmCnt
+        Wait.DsCnt = 0;             // LgkmCnt
+      } else                        // LDS load
+        Wait.DsCnt = 0;             // LgkmCnt
+    } else {                        // vector store
+      if (TII->isVMEM(Inst))        // VMEM store
+        Wait.LoadCnt = 0;           // VmCnt
+      else if (TII->isFLAT(Inst)) { // Flat store
+        Wait.LoadCnt = 0;           // VmCnt
+        Wait.DsCnt = 0;             // LgkmCnt
       } else
         Wait.DsCnt = 0; // LDS store; LgkmCnt
     }
@@ -2476,16 +2476,16 @@ bool SIGfx10CacheControl::handleNonAtomicForPreciseMemory(
   if (TII->isSMRD(Inst)) { // scalar
     if (Inst.mayStore())
       return false;
-    Wait.DsCnt = 0;                   // LgkmCnt
-  } else {                            // vector
-    if (Inst.mayLoad()) {             // vector load
-      if (TII->isVMEM(Inst))          // VMEM load
-        Wait.LoadCnt = 0;             // VmCnt
-      else if (TII->isFLAT(Inst)) {   // Flat load
-        Wait.LoadCnt = 0;             // VmCnt
-        Wait.DsCnt = 0;               // LgkmCnt
-      } else                          // LDS load
-        Wait.DsCnt = 0;               // LgkmCnt
+    Wait.DsCnt = 0;                 // LgkmCnt
+  } else {                          // vector
+    if (Inst.mayLoad()) {           // vector load
+      if (TII->isVMEM(Inst))        // VMEM load
+        Wait.LoadCnt = 0;           // VmCnt
+      else if (TII->isFLAT(Inst)) { // Flat load
+        Wait.LoadCnt = 0;           // VmCnt
+        Wait.DsCnt = 0;             // LgkmCnt
+      } else                        // LDS load
+        Wait.DsCnt = 0;             // LgkmCnt
     }
 
     // For some vector instructions, mayLoad() and mayStore() can be both true.
@@ -2567,8 +2567,8 @@ bool SIGfx12CacheControl ::handleNonAtomicForPreciseMemory(
         WaitType = AMDGPU::S_WAIT_LOADCNT_DSCNT;
       else // LDS load
         WaitType = AMDGPU::S_WAIT_DSCNT;
-    } else {                   // vector store
-      if (TII->isVMEM(Inst))   // VMEM store
+    } else {                 // vector store
+      if (TII->isVMEM(Inst)) // VMEM store
         WaitType = AMDGPU::S_WAIT_STORECNT;
       else if (TII->isFLAT(Inst)) // Flat store
         WaitType = AMDGPU::S_WAIT_STORECNT_DSCNT;

``````````

</details>


https://github.com/llvm/llvm-project/pull/79236