llvm-project/llvm/test/Transforms/InstCombine/abs-1.ll

391 lines
12 KiB
LLVM

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -instcombine -S | FileCheck %s
target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128"
declare i32 @abs(i32)
declare i64 @labs(i64)
declare i64 @llabs(i64)
; Test that the abs library call simplifier works correctly.
; abs(x) -> x <s 0 ? -x : x.
define i32 @test_abs(i32 %x) {
; CHECK-LABEL: @test_abs(
; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i32 [[X:%.*]], 0
; CHECK-NEXT: [[NEG:%.*]] = sub nsw i32 0, [[X]]
; CHECK-NEXT: [[TMP2:%.*]] = select i1 [[TMP1]], i32 [[NEG]], i32 [[X]]
; CHECK-NEXT: ret i32 [[TMP2]]
;
%ret = call i32 @abs(i32 %x)
ret i32 %ret
}
define i64 @test_labs(i64 %x) {
; CHECK-LABEL: @test_labs(
; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i64 [[X:%.*]], 0
; CHECK-NEXT: [[NEG:%.*]] = sub nsw i64 0, [[X]]
; CHECK-NEXT: [[TMP2:%.*]] = select i1 [[TMP1]], i64 [[NEG]], i64 [[X]]
; CHECK-NEXT: ret i64 [[TMP2]]
;
%ret = call i64 @labs(i64 %x)
ret i64 %ret
}
define i64 @test_llabs(i64 %x) {
; CHECK-LABEL: @test_llabs(
; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i64 [[X:%.*]], 0
; CHECK-NEXT: [[NEG:%.*]] = sub nsw i64 0, [[X]]
; CHECK-NEXT: [[TMP2:%.*]] = select i1 [[TMP1]], i64 [[NEG]], i64 [[X]]
; CHECK-NEXT: ret i64 [[TMP2]]
;
%ret = call i64 @llabs(i64 %x)
ret i64 %ret
}
; We have a canonical form of abs to make CSE easier.
define i8 @abs_canonical_1(i8 %x) {
; CHECK-LABEL: @abs_canonical_1(
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[NEG:%.*]] = sub i8 0, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[CMP]], i8 [[NEG]], i8 [[X]]
; CHECK-NEXT: ret i8 [[ABS]]
;
%cmp = icmp sgt i8 %x, 0
%neg = sub i8 0, %x
%abs = select i1 %cmp, i8 %x, i8 %neg
ret i8 %abs
}
; Vectors should work too.
define <2 x i8> @abs_canonical_2(<2 x i8> %x) {
; CHECK-LABEL: @abs_canonical_2(
; CHECK-NEXT: [[CMP:%.*]] = icmp slt <2 x i8> [[X:%.*]], zeroinitializer
; CHECK-NEXT: [[NEG:%.*]] = sub <2 x i8> zeroinitializer, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select <2 x i1> [[CMP]], <2 x i8> [[NEG]], <2 x i8> [[X]]
; CHECK-NEXT: ret <2 x i8> [[ABS]]
;
%cmp = icmp sgt <2 x i8> %x, <i8 -1, i8 -1>
%neg = sub <2 x i8> zeroinitializer, %x
%abs = select <2 x i1> %cmp, <2 x i8> %x, <2 x i8> %neg
ret <2 x i8> %abs
}
; NSW should not change.
define i8 @abs_canonical_3(i8 %x) {
; CHECK-LABEL: @abs_canonical_3(
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[NEG:%.*]] = sub nsw i8 0, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[CMP]], i8 [[NEG]], i8 [[X]]
; CHECK-NEXT: ret i8 [[ABS]]
;
%cmp = icmp slt i8 %x, 0
%neg = sub nsw i8 0, %x
%abs = select i1 %cmp, i8 %neg, i8 %x
ret i8 %abs
}
define i8 @abs_canonical_4(i8 %x) {
; CHECK-LABEL: @abs_canonical_4(
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[NEG:%.*]] = sub i8 0, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[CMP]], i8 [[NEG]], i8 [[X]]
; CHECK-NEXT: ret i8 [[ABS]]
;
%cmp = icmp slt i8 %x, 1
%neg = sub i8 0, %x
%abs = select i1 %cmp, i8 %neg, i8 %x
ret i8 %abs
}
define i32 @abs_canonical_5(i8 %x) {
; CHECK-LABEL: @abs_canonical_5(
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[CONV:%.*]] = sext i8 [[X]] to i32
; CHECK-NEXT: [[NEG:%.*]] = sub nsw i32 0, [[CONV]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[CMP]], i32 [[NEG]], i32 [[CONV]]
; CHECK-NEXT: ret i32 [[ABS]]
;
%cmp = icmp sgt i8 %x, 0
%conv = sext i8 %x to i32
%neg = sub i32 0, %conv
%abs = select i1 %cmp, i32 %conv, i32 %neg
ret i32 %abs
}
; We have a canonical form of nabs to make CSE easier.
define i8 @nabs_canonical_1(i8 %x) {
; CHECK-LABEL: @nabs_canonical_1(
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[NEG:%.*]] = sub i8 0, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[CMP]], i8 [[X]], i8 [[NEG]]
; CHECK-NEXT: ret i8 [[ABS]]
;
%cmp = icmp sgt i8 %x, 0
%neg = sub i8 0, %x
%abs = select i1 %cmp, i8 %neg, i8 %x
ret i8 %abs
}
; Vectors should work too.
define <2 x i8> @nabs_canonical_2(<2 x i8> %x) {
; CHECK-LABEL: @nabs_canonical_2(
; CHECK-NEXT: [[CMP:%.*]] = icmp slt <2 x i8> [[X:%.*]], zeroinitializer
; CHECK-NEXT: [[NEG:%.*]] = sub <2 x i8> zeroinitializer, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select <2 x i1> [[CMP]], <2 x i8> [[X]], <2 x i8> [[NEG]]
; CHECK-NEXT: ret <2 x i8> [[ABS]]
;
%cmp = icmp sgt <2 x i8> %x, <i8 -1, i8 -1>
%neg = sub <2 x i8> zeroinitializer, %x
%abs = select <2 x i1> %cmp, <2 x i8> %neg, <2 x i8> %x
ret <2 x i8> %abs
}
; NSW should not change.
define i8 @nabs_canonical_3(i8 %x) {
; CHECK-LABEL: @nabs_canonical_3(
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[NEG:%.*]] = sub nsw i8 0, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[CMP]], i8 [[X]], i8 [[NEG]]
; CHECK-NEXT: ret i8 [[ABS]]
;
%cmp = icmp slt i8 %x, 0
%neg = sub nsw i8 0, %x
%abs = select i1 %cmp, i8 %x, i8 %neg
ret i8 %abs
}
define i8 @nabs_canonical_4(i8 %x) {
; CHECK-LABEL: @nabs_canonical_4(
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[NEG:%.*]] = sub i8 0, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[CMP]], i8 [[X]], i8 [[NEG]]
; CHECK-NEXT: ret i8 [[ABS]]
;
%cmp = icmp slt i8 %x, 1
%neg = sub i8 0, %x
%abs = select i1 %cmp, i8 %x, i8 %neg
ret i8 %abs
}
define i32 @nabs_canonical_5(i8 %x) {
; CHECK-LABEL: @nabs_canonical_5(
; CHECK-NEXT: [[CMP:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[CONV:%.*]] = sext i8 [[X]] to i32
; CHECK-NEXT: [[NEG:%.*]] = sub nsw i32 0, [[CONV]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[CMP]], i32 [[CONV]], i32 [[NEG]]
; CHECK-NEXT: ret i32 [[ABS]]
;
%cmp = icmp sgt i8 %x, 0
%conv = sext i8 %x to i32
%neg = sub i32 0, %conv
%abs = select i1 %cmp, i32 %neg, i32 %conv
ret i32 %abs
}
; The following 5 tests use a shift+add+xor to implement abs():
; B = ashr i8 A, 7 -- smear the sign bit.
; xor (add A, B), B -- add -1 and flip bits if negative
define i8 @shifty_abs_commute0(i8 %x) {
; CHECK-LABEL: @shifty_abs_commute0(
; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[TMP2:%.*]] = sub i8 0, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[TMP1]], i8 [[TMP2]], i8 [[X]]
; CHECK-NEXT: ret i8 [[ABS]]
;
%signbit = ashr i8 %x, 7
%add = add i8 %signbit, %x
%abs = xor i8 %add, %signbit
ret i8 %abs
}
define i8 @shifty_abs_commute0_nsw(i8 %x) {
; CHECK-LABEL: @shifty_abs_commute0_nsw(
; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[TMP2:%.*]] = sub nsw i8 0, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[TMP1]], i8 [[TMP2]], i8 [[X]]
; CHECK-NEXT: ret i8 [[ABS]]
;
%signbit = ashr i8 %x, 7
%add = add nsw i8 %signbit, %x
%abs = xor i8 %add, %signbit
ret i8 %abs
}
; The nuw flag creates a contradiction. If the shift produces all 1s, the only
; way for the add to not wrap is for %x to be 0, but then the shift couldn't
; have produced all 1s. We partially optimize this.
define i8 @shifty_abs_commute0_nuw(i8 %x) {
; CHECK-LABEL: @shifty_abs_commute0_nuw(
; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt i8 [[X:%.*]], 0
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[TMP1]], i8 [[X]], i8 0
; CHECK-NEXT: ret i8 [[ABS]]
;
%signbit = ashr i8 %x, 7
%add = add nuw i8 %signbit, %x
%abs = xor i8 %add, %signbit
ret i8 %abs
}
define <2 x i8> @shifty_abs_commute1(<2 x i8> %x) {
; CHECK-LABEL: @shifty_abs_commute1(
; CHECK-NEXT: [[TMP1:%.*]] = icmp slt <2 x i8> [[X:%.*]], zeroinitializer
; CHECK-NEXT: [[TMP2:%.*]] = sub <2 x i8> zeroinitializer, [[X]]
; CHECK-NEXT: [[ABS:%.*]] = select <2 x i1> [[TMP1]], <2 x i8> [[TMP2]], <2 x i8> [[X]]
; CHECK-NEXT: ret <2 x i8> [[ABS]]
;
%signbit = ashr <2 x i8> %x, <i8 7, i8 7>
%add = add <2 x i8> %signbit, %x
%abs = xor <2 x i8> %signbit, %add
ret <2 x i8> %abs
}
define <2 x i8> @shifty_abs_commute2(<2 x i8> %x) {
; CHECK-LABEL: @shifty_abs_commute2(
; CHECK-NEXT: [[Y:%.*]] = mul <2 x i8> [[X:%.*]], <i8 3, i8 3>
; CHECK-NEXT: [[TMP1:%.*]] = icmp slt <2 x i8> [[Y]], zeroinitializer
; CHECK-NEXT: [[TMP2:%.*]] = sub <2 x i8> zeroinitializer, [[Y]]
; CHECK-NEXT: [[ABS:%.*]] = select <2 x i1> [[TMP1]], <2 x i8> [[TMP2]], <2 x i8> [[Y]]
; CHECK-NEXT: ret <2 x i8> [[ABS]]
;
%y = mul <2 x i8> %x, <i8 3, i8 3> ; extra op to thwart complexity-based canonicalization
%signbit = ashr <2 x i8> %y, <i8 7, i8 7>
%add = add <2 x i8> %y, %signbit
%abs = xor <2 x i8> %signbit, %add
ret <2 x i8> %abs
}
define i8 @shifty_abs_commute3(i8 %x) {
; CHECK-LABEL: @shifty_abs_commute3(
; CHECK-NEXT: [[Y:%.*]] = mul i8 [[X:%.*]], 3
; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i8 [[Y]], 0
; CHECK-NEXT: [[TMP2:%.*]] = sub i8 0, [[Y]]
; CHECK-NEXT: [[ABS:%.*]] = select i1 [[TMP1]], i8 [[TMP2]], i8 [[Y]]
; CHECK-NEXT: ret i8 [[ABS]]
;
%y = mul i8 %x, 3 ; extra op to thwart complexity-based canonicalization
%signbit = ashr i8 %y, 7
%add = add i8 %y, %signbit
%abs = xor i8 %add, %signbit
ret i8 %abs
}
; Negative test - don't transform if it would increase instruction count.
declare void @extra_use(i8)
define i8 @shifty_abs_too_many_uses(i8 %x) {
; CHECK-LABEL: @shifty_abs_too_many_uses(
; CHECK-NEXT: [[SIGNBIT:%.*]] = ashr i8 [[X:%.*]], 7
; CHECK-NEXT: [[ADD:%.*]] = add i8 [[SIGNBIT]], [[X]]
; CHECK-NEXT: [[ABS:%.*]] = xor i8 [[ADD]], [[SIGNBIT]]
; CHECK-NEXT: call void @extra_use(i8 [[SIGNBIT]])
; CHECK-NEXT: ret i8 [[ABS]]
;
%signbit = ashr i8 %x, 7
%add = add i8 %x, %signbit
%abs = xor i8 %add, %signbit
call void @extra_use(i8 %signbit)
ret i8 %abs
}
; There's another way to make abs() using shift, xor, and subtract.
; PR36036 - https://bugs.llvm.org/show_bug.cgi?id=36036
define i8 @shifty_sub(i8 %x) {
; CHECK-LABEL: @shifty_sub(
; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[TMP2:%.*]] = sub i8 0, [[X]]
; CHECK-NEXT: [[R:%.*]] = select i1 [[TMP1]], i8 [[TMP2]], i8 [[X]]
; CHECK-NEXT: ret i8 [[R]]
;
%sh = ashr i8 %x, 7
%xor = xor i8 %x, %sh
%r = sub i8 %xor, %sh
ret i8 %r
}
define i8 @shifty_sub_nsw_commute(i8 %x) {
; CHECK-LABEL: @shifty_sub_nsw_commute(
; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i8 [[X:%.*]], 0
; CHECK-NEXT: [[TMP2:%.*]] = sub nsw i8 0, [[X]]
; CHECK-NEXT: [[R:%.*]] = select i1 [[TMP1]], i8 [[TMP2]], i8 [[X]]
; CHECK-NEXT: ret i8 [[R]]
;
%sh = ashr i8 %x, 7
%xor = xor i8 %sh, %x
%r = sub nsw i8 %xor, %sh
ret i8 %r
}
define <4 x i32> @shifty_sub_nuw_vec_commute(<4 x i32> %x) {
; CHECK-LABEL: @shifty_sub_nuw_vec_commute(
; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt <4 x i32> [[X:%.*]], zeroinitializer
; CHECK-NEXT: [[R:%.*]] = select <4 x i1> [[TMP1]], <4 x i32> [[X]], <4 x i32> zeroinitializer
; CHECK-NEXT: ret <4 x i32> [[R]]
;
%sh = ashr <4 x i32> %x, <i32 31, i32 31, i32 31, i32 31>
%xor = xor <4 x i32> %sh, %x
%r = sub nuw <4 x i32> %xor, %sh
ret <4 x i32> %r
}
define i12 @shifty_sub_nsw_nuw(i12 %x) {
; CHECK-LABEL: @shifty_sub_nsw_nuw(
; CHECK-NEXT: [[TMP1:%.*]] = icmp sgt i12 [[X:%.*]], 0
; CHECK-NEXT: [[R:%.*]] = select i1 [[TMP1]], i12 [[X]], i12 0
; CHECK-NEXT: ret i12 [[R]]
;
%sh = ashr i12 %x, 11
%xor = xor i12 %x, %sh
%r = sub nsw nuw i12 %xor, %sh
ret i12 %r
}
define i8 @negate_abs(i8 %x) {
; CHECK-LABEL: @negate_abs(
; CHECK-NEXT: [[N:%.*]] = sub i8 0, [[X:%.*]]
; CHECK-NEXT: [[C:%.*]] = icmp slt i8 [[X]], 0
; CHECK-NEXT: [[S:%.*]] = select i1 [[C]], i8 [[X]], i8 [[N]]
; CHECK-NEXT: ret i8 [[S]]
;
%n = sub i8 0, %x
%c = icmp slt i8 %x, 0
%s = select i1 %c, i8 %n, i8 %x
%r = sub i8 0, %s
ret i8 %r
}
define <2 x i8> @negate_nabs(<2 x i8> %x) {
; CHECK-LABEL: @negate_nabs(
; CHECK-NEXT: [[N:%.*]] = sub <2 x i8> zeroinitializer, [[X:%.*]]
; CHECK-NEXT: [[C:%.*]] = icmp slt <2 x i8> [[X]], zeroinitializer
; CHECK-NEXT: [[S:%.*]] = select <2 x i1> [[C]], <2 x i8> [[N]], <2 x i8> [[X]]
; CHECK-NEXT: ret <2 x i8> [[S]]
;
%n = sub <2 x i8> zeroinitializer, %x
%c = icmp slt <2 x i8> %x, zeroinitializer
%s = select <2 x i1> %c, <2 x i8> %x, <2 x i8> %n
%r = sub <2 x i8> zeroinitializer, %s
ret <2 x i8> %r
}
define i1 @abs_must_be_positive(i32 %x) {
; CHECK-LABEL: @abs_must_be_positive(
; CHECK-NEXT: ret i1 true
;
%negx = sub nsw i32 0, %x
%c = icmp sge i32 %x, 0
%sel = select i1 %c, i32 %x, i32 %negx
%c2 = icmp sge i32 %sel, 0
ret i1 %c2
}