[llvm] AMDGPU: Fix assert on physreg MUBUF rsrc operand (PR #120815)

[via llvm-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-amdgpu

Author: Matt Arsenault (arsenm)

<details>
<summary>Changes</summary>

The stack case uses a physical register and should not ordinarily
reach here, but strange things happen at -O0. The testcase still
errors because we do not yet attempt to handle arbitrary dynamic
sized allocas yet.

Fixes: SWDEV-503538

---
Full diff: https://github.com/llvm/llvm-project/pull/120815.diff


2 Files Affected:

- (modified) llvm/lib/Target/AMDGPU/SIInstrInfo.cpp (+1-2) 
- (added) llvm/test/CodeGen/AMDGPU/swdev503538-move-to-valu-stack-srd-physreg.ll (+23) 


``````````diff
diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
index f97ea40caa6704..7e61aa9cd1da6a 100644
--- a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -6866,9 +6866,8 @@ SIInstrInfo::legalizeOperands(MachineInstr &MI,
       AMDGPU::getNamedOperandIdx(MI.getOpcode(), AMDGPU::OpName::srsrc);
   if (RsrcIdx != -1) {
     MachineOperand *Rsrc = &MI.getOperand(RsrcIdx);
-    if (Rsrc->isReg() && !RI.isSGPRClass(MRI.getRegClass(Rsrc->getReg()))) {
+    if (Rsrc->isReg() && !RI.isSGPRReg(MRI, Rsrc->getReg()))
       isRsrcLegal = false;
-    }
   }
 
   // The operands are legal.
diff --git a/llvm/test/CodeGen/AMDGPU/swdev503538-move-to-valu-stack-srd-physreg.ll b/llvm/test/CodeGen/AMDGPU/swdev503538-move-to-valu-stack-srd-physreg.ll
new file mode 100644
index 00000000000000..6849c8b4e609ee
--- /dev/null
+++ b/llvm/test/CodeGen/AMDGPU/swdev503538-move-to-valu-stack-srd-physreg.ll
@@ -0,0 +1,23 @@
+; RUN: not llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx90a -O0 2> %t.err < %s | FileCheck %s
+; RUN: FileCheck -check-prefix=ERR %s < %t.err
+
+; FIXME: This error will be fixed by supporting arbitrary divergent
+; dynamic allocas by performing a wave umax of the size.
+
+; ERR: error: <unknown>:0:0: in function move_to_valu_assert_srd_is_physreg_swdev503538 i32 (ptr addrspace(1)): illegal VGPR to SGPR copy
+
+; CHECK: ; illegal copy v0 to s32
+
+define i32 @move_to_valu_assert_srd_is_physreg_swdev503538(ptr addrspace(1) %ptr) {
+entry:
+  %idx = load i32, ptr addrspace(1) %ptr, align 4
+  %zero = extractelement <4 x i32> zeroinitializer, i32 %idx
+  %alloca = alloca [2048 x i8], i32 %zero, align 8, addrspace(5)
+  %ld = load i32, ptr addrspace(5) %alloca, align 8
+  call void @llvm.memset.p5.i32(ptr addrspace(5) %alloca, i8 0, i32 2048, i1 false)
+  ret i32 %ld
+}
+
+declare void @llvm.memset.p5.i32(ptr addrspace(5) nocapture writeonly, i8, i32, i1 immarg) #0
+
+attributes #0 = { nocallback nofree nounwind willreturn memory(argmem: write) }

``````````

</details>


https://github.com/llvm/llvm-project/pull/120815