[PATCH] D149348: RFD: Do not CSE convergent calls in different basic blocks

[Jay Foad via Phabricator via llvm-commits]

foad updated this revision to Diff 517609.
foad added a comment.

Restrict the changes to EarlyCSE itself.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D149348/new/

https://reviews.llvm.org/D149348

Files:
  llvm/lib/Transforms/Scalar/EarlyCSE.cpp
  llvm/test/CodeGen/AMDGPU/cse-convergent.ll


Index: llvm/test/CodeGen/AMDGPU/cse-convergent.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/cse-convergent.ll
+++ llvm/test/CodeGen/AMDGPU/cse-convergent.ll
@@ -21,15 +21,25 @@
 ; GCN-NEXT:    s_or_saveexec_b32 s4, -1
 ; GCN-NEXT:    v_mov_b32_dpp v2, v3 row_xmask:1 row_mask:0xf bank_mask:0xf
 ; GCN-NEXT:    s_mov_b32 exec_lo, s4
-; GCN-NEXT:    v_mov_b32_e32 v4, 0
-; GCN-NEXT:    v_mov_b32_e32 v0, v2
+; GCN-NEXT:    v_mov_b32_e32 v5, 0
+; GCN-NEXT:    v_mov_b32_e32 v4, v2
 ; GCN-NEXT:    v_cmp_eq_u32_e32 vcc_lo, 0, v1
 ; GCN-NEXT:    s_and_saveexec_b32 s4, vcc_lo
 ; GCN-NEXT:  ; %bb.1: ; %if
-; GCN-NEXT:    v_mov_b32_e32 v4, v0
+; GCN-NEXT:    s_or_saveexec_b32 s5, -1
+; GCN-NEXT:    v_mov_b32_e32 v2, 0
+; GCN-NEXT:    s_mov_b32 exec_lo, s5
+; GCN-NEXT:    v_mov_b32_e32 v3, v0
+; GCN-NEXT:    s_not_b32 exec_lo, exec_lo
+; GCN-NEXT:    v_mov_b32_e32 v3, 0
+; GCN-NEXT:    s_not_b32 exec_lo, exec_lo
+; GCN-NEXT:    s_or_saveexec_b32 s5, -1
+; GCN-NEXT:    v_mov_b32_dpp v2, v3 row_xmask:1 row_mask:0xf bank_mask:0xf
+; GCN-NEXT:    s_mov_b32 exec_lo, s5
+; GCN-NEXT:    v_mov_b32_e32 v5, v2
 ; GCN-NEXT:  ; %bb.2: ; %end
 ; GCN-NEXT:    s_or_b32 exec_lo, exec_lo, s4
-; GCN-NEXT:    v_add_nc_u32_e32 v0, v0, v4
+; GCN-NEXT:    v_add_nc_u32_e32 v0, v4, v5
 ; GCN-NEXT:    s_xor_saveexec_b32 s4, -1
 ; GCN-NEXT:    s_clause 0x1
 ; GCN-NEXT:    buffer_load_dword v2, off, s[0:3], s32
Index: llvm/lib/Transforms/Scalar/EarlyCSE.cpp
===================================================================
--- llvm/lib/Transforms/Scalar/EarlyCSE.cpp
+++ llvm/lib/Transforms/Scalar/EarlyCSE.cpp
@@ -318,6 +318,14 @@
     return hash_combine(GCR->getOpcode(), GCR->getOperand(0),
                         GCR->getBasePtr(), GCR->getDerivedPtr());
 
+  // Don't CSE convergent calls in different basic blocks, because they
+  // implicitly depend on the set of threads that is currently executing.
+  if (CallInst *CI = dyn_cast<CallInst>(Inst); CI && CI->isConvergent()) {
+    return hash_combine(
+        Inst->getOpcode(), Inst->getParent(),
+        hash_combine_range(Inst->value_op_begin(), Inst->value_op_end()));
+  }
+
   // Mix in the opcode.
   return hash_combine(
       Inst->getOpcode(),
@@ -344,8 +352,16 @@
 
   if (LHSI->getOpcode() != RHSI->getOpcode())
     return false;
-  if (LHSI->isIdenticalToWhenDefined(RHSI))
+  if (LHSI->isIdenticalToWhenDefined(RHSI)) {
+    // Convergent calls implicitly depend on the set of threads that is
+    // currently executing, so conservatively return false if they are in
+    // different basic blocks.
+    if (CallInst *CI = dyn_cast<CallInst>(LHSI);
+        CI && CI->isConvergent() && LHSI->getParent() != RHSI->getParent())
+      return false;
+
     return true;
+  }
 
   // If we're not strictly identical, we still might be a commutable instruction
   if (BinaryOperator *LHSBinOp = dyn_cast<BinaryOperator>(LHSI)) {


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D149348.517609.patch
Type: text/x-patch
Size: 2945 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230427/4b29a4df/attachment.bin>