[llvm] [CGP]: Optimize mul.overflow. (PR #148343)
Hassnaa Hamdi via llvm-commits
llvm-commits at lists.llvm.org
Sat Jul 12 01:17:46 PDT 2025
================
@@ -34,46 +68,133 @@ start:
define i128 @__muloti4(i128 %0, i128 %1, ptr nocapture nonnull writeonly align 4 %2) #2 {
; AARCH-LABEL: __muloti4:
-; AARCH: // %bb.0: // %Entry
-; AARCH-NEXT: asr x11, x1, #63
-; AARCH-NEXT: asr x9, x3, #63
-; AARCH-NEXT: umulh x12, x0, x2
-; AARCH-NEXT: mov x8, x1
+; AARCH: // %bb.0: // %overflow.entry
+; AARCH-NEXT: asr x8, x2, #63
+; AARCH-NEXT: cmp x1, x0, asr #63
; AARCH-NEXT: str wzr, [x4]
-; AARCH-NEXT: mul x13, x1, x2
-; AARCH-NEXT: umulh x10, x1, x2
-; AARCH-NEXT: mul x11, x11, x2
-; AARCH-NEXT: adds x12, x13, x12
-; AARCH-NEXT: mul x15, x0, x3
-; AARCH-NEXT: umulh x14, x0, x3
-; AARCH-NEXT: adc x10, x10, x11
-; AARCH-NEXT: mul x9, x0, x9
-; AARCH-NEXT: mul x16, x1, x3
-; AARCH-NEXT: adds x1, x15, x12
-; AARCH-NEXT: asr x12, x10, #63
-; AARCH-NEXT: smulh x11, x8, x3
-; AARCH-NEXT: adc x9, x14, x9
-; AARCH-NEXT: asr x13, x9, #63
-; AARCH-NEXT: adds x9, x10, x9
-; AARCH-NEXT: asr x10, x1, #63
+; AARCH-NEXT: b.eq .LBB1_3
+; AARCH-NEXT: // %bb.1: // %overflow.lhs
+; AARCH-NEXT: cmp x3, x8
+; AARCH-NEXT: b.eq .LBB1_5
+; AARCH-NEXT: // %bb.2: // %overflow
+; AARCH-NEXT: asr x9, x1, #63
+; AARCH-NEXT: umulh x10, x0, x2
+; AARCH-NEXT: asr x13, x3, #63
+; AARCH-NEXT: mul x11, x1, x2
+; AARCH-NEXT: umulh x8, x1, x2
+; AARCH-NEXT: mul x9, x9, x2
+; AARCH-NEXT: adds x10, x11, x10
+; AARCH-NEXT: mul x14, x0, x3
+; AARCH-NEXT: umulh x12, x0, x3
+; AARCH-NEXT: adc x9, x8, x9
+; AARCH-NEXT: mul x13, x0, x13
+; AARCH-NEXT: adds x8, x14, x10
+; AARCH-NEXT: mul x15, x1, x3
+; AARCH-NEXT: smulh x10, x1, x3
+; AARCH-NEXT: adc x11, x12, x13
+; AARCH-NEXT: asr x12, x9, #63
+; AARCH-NEXT: asr x13, x11, #63
; AARCH-NEXT: mul x0, x0, x2
+; AARCH-NEXT: adds x9, x9, x11
+; AARCH-NEXT: asr x11, x8, #63
; AARCH-NEXT: adc x12, x12, x13
-; AARCH-NEXT: adds x9, x16, x9
-; AARCH-NEXT: adc x11, x11, x12
-; AARCH-NEXT: cmp x9, x10
-; AARCH-NEXT: ccmp x11, x10, #0, eq
+; AARCH-NEXT: adds x9, x15, x9
+; AARCH-NEXT: adc x10, x10, x12
+; AARCH-NEXT: cmp x9, x11
+; AARCH-NEXT: ccmp x10, x11, #0, eq
+; AARCH-NEXT: b .LBB1_7
+; AARCH-NEXT: .LBB1_3: // %overflow.no.lhs
+; AARCH-NEXT: cmp x3, x8
+; AARCH-NEXT: b.eq .LBB1_8
+; AARCH-NEXT: // %bb.4: // %overflow.no.lhs.only
+; AARCH-NEXT: asr x8, x1, #63
+; AARCH-NEXT: asr x10, x3, #63
+; AARCH-NEXT: eor x9, x0, x8
+; AARCH-NEXT: eor x11, x1, x8
+; AARCH-NEXT: eor x12, x2, x10
+; AARCH-NEXT: subs x9, x9, x8
+; AARCH-NEXT: sbc x8, x11, x8
+; AARCH-NEXT: cmp x1, #0
+; AARCH-NEXT: eor x11, x3, x10
+; AARCH-NEXT: cset w13, lt
+; AARCH-NEXT: csel x8, x8, x1, lt
+; AARCH-NEXT: csel x9, x9, x0, lt
+; AARCH-NEXT: subs x12, x12, x10
+; AARCH-NEXT: sbc x10, x11, x10
+; AARCH-NEXT: cmp x3, #0
+; AARCH-NEXT: csel x11, x12, x2, lt
+; AARCH-NEXT: csel x10, x10, x3, lt
+; AARCH-NEXT: umulh x12, x9, x11
+; AARCH-NEXT: mul x15, x8, x10
+; AARCH-NEXT: madd x8, x8, x11, x12
+; AARCH-NEXT: cset w12, lt
+; AARCH-NEXT: mul x14, x9, x11
+; AARCH-NEXT: mul x11, x9, x10
+; AARCH-NEXT: umulh x9, x9, x10
+; AARCH-NEXT: eor w10, w12, w13
+; AARCH-NEXT: b .LBB1_6
+; AARCH-NEXT: .LBB1_5: // %overflow.no.rhs.only
+; AARCH-NEXT: asr x8, x3, #63
+; AARCH-NEXT: asr x10, x1, #63
+; AARCH-NEXT: eor x9, x2, x8
+; AARCH-NEXT: eor x11, x3, x8
+; AARCH-NEXT: eor x12, x0, x10
+; AARCH-NEXT: subs x9, x9, x8
+; AARCH-NEXT: sbc x8, x11, x8
+; AARCH-NEXT: cmp x3, #0
+; AARCH-NEXT: eor x11, x1, x10
+; AARCH-NEXT: cset w13, lt
+; AARCH-NEXT: csel x8, x8, x3, lt
+; AARCH-NEXT: csel x9, x9, x2, lt
+; AARCH-NEXT: subs x12, x12, x10
+; AARCH-NEXT: sbc x10, x11, x10
+; AARCH-NEXT: cmp x1, #0
+; AARCH-NEXT: csel x11, x12, x0, lt
+; AARCH-NEXT: csel x10, x10, x1, lt
+; AARCH-NEXT: umulh x12, x9, x11
+; AARCH-NEXT: mul x14, x9, x11
+; AARCH-NEXT: mul x15, x8, x10
+; AARCH-NEXT: madd x8, x8, x11, x12
+; AARCH-NEXT: cset w12, lt
+; AARCH-NEXT: mul x11, x9, x10
+; AARCH-NEXT: umulh x9, x9, x10
+; AARCH-NEXT: eor w10, w13, w12
+; AARCH-NEXT: .LBB1_6: // %overflow.res
+; AARCH-NEXT: sbfx x12, x10, #0, #1
+; AARCH-NEXT: adds x8, x8, x11
+; AARCH-NEXT: adc x9, x9, x15
+; AARCH-NEXT: eor x13, x14, x12
+; AARCH-NEXT: eor x8, x8, x12
+; AARCH-NEXT: eor x9, x9, x12
+; AARCH-NEXT: add x0, x13, x10
+; AARCH-NEXT: cmp x0, x10
+; AARCH-NEXT: cset w10, lo
+; AARCH-NEXT: cinc x8, x8, lo
+; AARCH-NEXT: cmp x8, x10
+; AARCH-NEXT: cinc x9, x9, lo
+; AARCH-NEXT: cmp x9, #0
+; AARCH-NEXT: .LBB1_7: // %overflow.res
; AARCH-NEXT: cset w9, ne
-; AARCH-NEXT: tbz x8, #63, .LBB1_2
-; AARCH-NEXT: // %bb.1: // %Entry
-; AARCH-NEXT: eor x8, x3, #0x8000000000000000
-; AARCH-NEXT: orr x8, x2, x8
-; AARCH-NEXT: cbz x8, .LBB1_3
-; AARCH-NEXT: .LBB1_2: // %Else2
-; AARCH-NEXT: cbz w9, .LBB1_4
-; AARCH-NEXT: .LBB1_3: // %Then7
-; AARCH-NEXT: mov w8, #1 // =0x1
-; AARCH-NEXT: str w8, [x4]
-; AARCH-NEXT: .LBB1_4: // %Block9
+; AARCH-NEXT: tbnz x1, #63, .LBB1_9
+; AARCH-NEXT: b .LBB1_10
+; AARCH-NEXT: .LBB1_8: // %overflow.no
+; AARCH-NEXT: umulh x8, x0, x2
+; AARCH-NEXT: mov w9, wzr
+; AARCH-NEXT: madd x8, x0, x3, x8
+; AARCH-NEXT: mul x0, x0, x2
+; AARCH-NEXT: madd x8, x1, x2, x8
----------------
hassnaaHamdi wrote:
For the 2 instruction of `madd`, there should be extra patch to fold them away as x3 and x1 are expected to be 0.
https://github.com/llvm/llvm-project/pull/148343
More information about the llvm-commits
mailing list