[llvm] 72faccc - [DAGCombine] Add more tests for cmp to sbb combination; NFC

[Paweł Bylica via llvm-commits]

Author: Paweł Bylica
Date: 2022-08-24T08:27:58+02:00
New Revision: 72faccc3f055fe1d570e54281f9b5d44eea2091e

URL: https://github.com/llvm/llvm-project/commit/72faccc3f055fe1d570e54281f9b5d44eea2091e
DIFF: https://github.com/llvm/llvm-project/commit/72faccc3f055fe1d570e54281f9b5d44eea2091e.diff

LOG: [DAGCombine] Add more tests for cmp to sbb combination; NFC

Add 2 more tests for potential DAG combine of cmp into sbb.

Differential Revision: https://reviews.llvm.org/D132463

Added: 
    

Modified: 
    llvm/test/CodeGen/X86/subcarry.ll

Removed: 
    


################################################################################
diff  --git a/llvm/test/CodeGen/X86/subcarry.ll b/llvm/test/CodeGen/X86/subcarry.ll
index 1ee37297dfeee..4dcdf56eae955 100644
--- a/llvm/test/CodeGen/X86/subcarry.ll
+++ b/llvm/test/CodeGen/X86/subcarry.ll
@@ -652,6 +652,8 @@ define void @sub_U256_without_i128_or_recursive(ptr sret(%uint256) %0, ptr %1, p
   ret void
 }
 
+; unsigned less than of two 2x64 integers
+; TODO: This should be optimized to cmp + sbb.
 define i1 @subcarry_ult_2x64(i64 %x0, i64 %x1, i64 %y0, i64 %y1) nounwind {
 ; CHECK-LABEL: subcarry_ult_2x64:
 ; CHECK:       # %bb.0:
@@ -670,3 +672,48 @@ define i1 @subcarry_ult_2x64(i64 %x0, i64 %x1, i64 %y0, i64 %y1) nounwind {
   %b1 = or i1 %b10, %b11
   ret i1 %b1
 }
+
+; New version of subcarry_ult_2x64 after the InstCombine change
+; https://github.com/llvm/llvm-project/commit/926e7312b2f20f2f7b0a3d5ddbd29da5625507f3
+; This is also the result of "naive" implementation (x1 < y1) | ((x0 < y0) & (x1 == y1)).
+; C source: https://godbolt.org/z/W1qqvqGbr
+; TODO: This should be optimized to cmp + sbb.
+define i1 @subcarry_ult_2x64_2(i64 %x0, i64 %x1, i64 %y0, i64 %y1) nounwind {
+; CHECK-LABEL: subcarry_ult_2x64_2:
+; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    cmpq %rdx, %rdi
+; CHECK-NEXT:    setb %dl
+; CHECK-NEXT:    cmpq %rcx, %rsi
+; CHECK-NEXT:    setb %cl
+; CHECK-NEXT:    sete %al
+; CHECK-NEXT:    andb %dl, %al
+; CHECK-NEXT:    orb %cl, %al
+; CHECK-NEXT:    retq
+entry:
+  %0 = icmp ult i64 %x0, %y0
+  %1 = icmp ult i64 %x1, %y1
+  %2 = icmp eq i64 %x1, %y1
+  %3 = and i1 %0, %2
+  %4 = or i1 %1, %3
+  ret i1 %4
+}
+
+; unsigned less than of 2x64 and i64 integers
+; The IR comes from C source that uses __builtin_subcl but also the naive version (x0 < y) & (x1 == 0).
+; https://godbolt.org/z/W1qqvqGbr
+; TODO: This should be optimized to cmp + sbb.
+define i1 @subcarry_ult_2x64_1x64(i64 %x0, i64 %x1, i64 %y) nounwind {
+; CHECK-LABEL: subcarry_ult_2x64_1x64:
+; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    cmpq %rdx, %rdi
+; CHECK-NEXT:    setb %cl
+; CHECK-NEXT:    testq %rsi, %rsi
+; CHECK-NEXT:    sete %al
+; CHECK-NEXT:    andb %cl, %al
+; CHECK-NEXT:    retq
+entry:
+  %0 = icmp ult i64 %x0, %y
+  %1 = icmp eq i64 %x1, 0
+  %2 = and i1 %1, %0
+  ret i1 %2
+}