Optimize a couple of common patterns involving conditional moves where the false
value is zero. Instead of a cmov + op, issue an conditional op instead. e.g.
cmp r9, r4
mov r4, #0
moveq r4, #1
orr lr, lr, r4
should be:
cmp r9, r4
orreq lr, lr, #1
That is, optimize (or x, (cmov 0, y, cond)) to (or.cond x, y). Similarly extend
this to xor as well as (and x, (cmov -1, y, cond)) => (and.cond x, y).
It's possible to extend this to ADD and SUB but I don't think they are common.
rdar://8659097
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@151224 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/Target/ARM/ARMISelLowering.h b/lib/Target/ARM/ARMISelLowering.h
index 527156a..a72a476 100644
--- a/lib/Target/ARM/ARMISelLowering.h
+++ b/lib/Target/ARM/ARMISelLowering.h
@@ -56,7 +56,11 @@
CMPFP, // ARM VFP compare instruction, sets FPSCR.
CMPFPw0, // ARM VFP compare against zero instruction, sets FPSCR.
FMSTAT, // ARM fmstat instruction.
+
CMOV, // ARM conditional move instructions.
+ CAND, // ARM conditional and instructions.
+ COR, // ARM conditional or instructions.
+ CXOR, // ARM conditional xor instructions.
BCC_i64,