[PATCH] D86541: Preserve the vcc/vcc_lo when shrinking V_CNDMASK

[Piotr Sobczak via Phabricator via llvm-commits]

piotr created this revision.
Herald added subscribers: llvm-commits, kerbowa, hiraditya, nhaehnle, jvesely, arsenm.
Herald added a project: LLVM.
piotr requested review of this revision.

There is no justification for changing vcc_lo to vcc
when shrinking V_CNDMASK, and such a change could
later confuse live variable analysis.

Make sure the original register is preserved.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D86541

Files:
  llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
  llvm/test/CodeGen/AMDGPU/shrink-instructions-implicit-vcclo.mir


Index: llvm/test/CodeGen/AMDGPU/shrink-instructions-implicit-vcclo.mir
===================================================================
--- /dev/null
+++ llvm/test/CodeGen/AMDGPU/shrink-instructions-implicit-vcclo.mir
@@ -0,0 +1,22 @@
+# RUN: llc -march=amdgcn -mcpu=gfx1010 -run-pass=si-shrink-instructions --verify-machineinstrs %s -o - | FileCheck %s
+
+# Make sure the implicit vcc_lo of V_CNDMASK is preserved and not promoted to vcc.
+---
+
+name:            shrink_cndmask_implicit_vcc_lo
+tracksRegLiveness: true
+body:             |
+  bb.0:
+    liveins: $vgpr0, $vgpr1
+    ; CHECK-LABEL: name: shrink_cndmask_implicit_vcc_lo
+    ; CHECK: [[COPY:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+    ; CHECK: [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+    ; CHECK: V_CMP_LT_I32_e32 0, [[COPY]], implicit-def $vcc_lo, implicit $exec
+    ; CHECK: V_CNDMASK_B32_e32 0, [[COPY1]], implicit $vcc_lo, implicit $exec
+    %0:vgpr_32 = COPY $vgpr0
+    %1:vgpr_32 = COPY $vgpr0
+    V_CMP_LT_I32_e32 0, %0:vgpr_32, implicit-def $vcc_lo, implicit $exec, implicit-def $vcc
+    %2:vgpr_32 = V_CNDMASK_B32_e64 0, 0, 0, %1:vgpr_32, $vcc_lo, implicit $exec
+    S_NOP 0
+
+...
Index: llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
===================================================================
--- llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
+++ llvm/lib/Target/AMDGPU/SIInstrInfo.cpp
@@ -3427,6 +3427,9 @@
   for (MachineOperand &Use : MI.implicit_operands()) {
     if (Use.isUse() &&
         (Use.getReg() == AMDGPU::VCC || Use.getReg() == AMDGPU::VCC_LO)) {
+      assert(Orig.getReg() == AMDGPU::VCC || Orig.getReg() == AMDGPU::VCC_LO);
+
+      Use.setReg(Orig.getReg());
       Use.setIsUndef(Orig.isUndef());
       Use.setIsKill(Orig.isKill());
       return;


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D86541.287662.patch
Type: text/x-patch
Size: 1746 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200825/963dc58a/attachment.bin>