[PGO][PGSO] Update BFI in CodeGenPrepare::optimizeSelectInst.

[hjyamauchi]

Summary

Without the BFI update, some hot blocks are incorrectly treated as cold code.

This fixes a FDO perf regression in the TSVC benchmark from D71288.

[davidxl]

lgtm

llvm/lib/CodeGen/CodeGenPrepare.cpp

6141

I wonder how many more such bugs are there

[hjyamauchi]

llvm/lib/CodeGen/CodeGenPrepare.cpp

6141

So do I. Will check.

[hans]

Cherry-picked to 10.x in b5cdee9bdce49b31c74b0cf4f8231f66bbc2708e
Please let me know if there are any follow-ups.

llvm/lib/CodeGen/CodeGenPrepare.cpp

bool CodeGenPrepare::optimizeSelectInst(SelectInst *SI) {
   // block and its branch may be optimized away. In that case, one side of the
   // first branch will point directly to select.end, and the corresponding PHI
   // predecessor block will be the start block.
 
   // First, we split the block containing the select into 2 blocks.
   BasicBlock *StartBlock = SI->getParent();
   BasicBlock::iterator SplitPt = ++(BasicBlock::iterator(LastSI));
   BasicBlock *EndBlock = StartBlock->splitBasicBlock(SplitPt, "select.end");
+  BFI->setBlockFreq(EndBlock, BFI->getBlockFreq(StartBlock).getFrequency());
 
   // Delete the unconditional branch that was just created by the split.
   StartBlock->getTerminator()->eraseFromParent();
 
   // These are the new basic blocks for the conditional branch.
   // At least one will become an actual new basic block.
   BasicBlock *TrueBlock = nullptr;
   BasicBlock *FalseBlock = nullptr;

llvm/test/CodeGen/X86/cmov-into-branch.ll

 ; CHECK-NEXT:    testl %edi, %edi
 ; CHECK-NEXT:    cmovnel %edi, %eax
 ; CHECK-NEXT:    retq
   %cmp = icmp ne i32 %a, 0
   %sel = select i1 %cmp, i32 %a, i32 %b, !prof !16
   ret i32 %sel
 }
 
+; If two selects in a row are predictable, turn them into branches.
+define i32 @weighted_selects(i32 %a, i32 %b) !prof !19 {
+; CHECK-LABEL: weighted_selects:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    movl %esi, %eax
+; CHECK-NEXT:    testl %edi, %edi
+; CHECK-NEXT:    movl %edi, %ecx
+; CHECK-NEXT:    jne .LBB11_2
+; CHECK-NEXT:  # %bb.1: # %select.false
+; CHECK-NEXT:    movl %eax, %ecx
+; CHECK-NEXT:  .LBB11_2: # %select.end
+; CHECK-NEXT:    testl %ecx, %ecx
+; CHECK-NEXT:    jne .LBB11_4
+; CHECK-NEXT:  # %bb.3: # %select.false2
+; CHECK-NEXT:    movl %edi, %eax
+; CHECK-NEXT:  .LBB11_4: # %select.end1
+; CHECK-NEXT:    retq
+  %cmp = icmp ne i32 %a, 0
+  %sel = select i1 %cmp, i32 %a, i32 %b, !prof !16
+  %cmp1 = icmp ne i32 %sel, 0
+  %sel1 = select i1 %cmp1, i32 %b, i32 %a, !prof !16
+  ret i32 %sel1
+}
+
 !llvm.module.flags = !{!0}
 !0 = !{i32 1, !"ProfileSummary", !1}
 !1 = !{!2, !3, !4, !5, !6, !7, !8, !9}
 !2 = !{!"ProfileFormat", !"InstrProf"}
 !3 = !{!"TotalCount", i64 10000}
 !4 = !{!"MaxCount", i64 10}
 !5 = !{!"MaxInternalCount", i64 1}
 !6 = !{!"MaxFunctionCount", i64 1000}
 !7 = !{!"NumCounts", i64 3}
 !8 = !{!"NumFunctions", i64 3}
 !9 = !{!"DetailedSummary", !10}
 !10 = !{!11, !12, !13}
 !11 = !{i32 10000, i64 100, i32 1}
 !12 = !{i32 999000, i64 100, i32 1}
 !13 = !{i32 999999, i64 1, i32 2}
 !14 = !{!"function_entry_count", i64 0}
 !15 = !{!"branch_weights", i32 1, i32 99}
 !16 = !{!"branch_weights", i32 1, i32 100}
 !17 = !{!"branch_weights", i32 100, i32 1}
 !18 = !{!"branch_weights", i32 0, i32 0}
+!19 = !{!"function_entry_count", i64 100}