1 ; Tests to make sure intrinsics are automatically upgraded.
2 ; RUN: llvm-as < %s | llvm-dis | not grep {i32 @llvm\\.ct}
3 ; RUN: llvm-as < %s | llvm-dis | \
4 ; RUN: not grep {llvm\\.part\\.set\\.i\[0-9\]*\\.i\[0-9\]*\\.i\[0-9\]*}
5 ; RUN: llvm-as < %s | llvm-dis | \
6 ; RUN: not grep {llvm\\.part\\.select\\.i\[0-9\]*\\.i\[0-9\]*}
7 ; RUN: llvm-as < %s | llvm-dis | \
8 ; RUN: not grep {llvm\\.bswap\\.i\[0-9\]*\\.i\[0-9\]*}
9 ; RUN: llvm-as < %s | llvm-dis | \
10 ; RUN: not grep {llvm\\.x86\\.sse2\\.loadu}
11 ; RUN: llvm-as < %s | llvm-dis | \
12 ; RUN: grep {llvm\\.x86\\.mmx\\.ps} | grep {x86_mmx} | count 16
13 ; RUN: llvm-as < %s | llvm-dis | FileCheck %s
15 declare i32 @llvm.ctpop.i28(i28 %val)
16 declare i32 @llvm.cttz.i29(i29 %val)
17 declare i32 @llvm.ctlz.i30(i30 %val)
19 define i32 @test_ct(i32 %A) {
20 %c1 = call i32 @llvm.ctpop.i28(i28 1234)
21 %c2 = call i32 @llvm.cttz.i29(i29 2345)
22 %c3 = call i32 @llvm.ctlz.i30(i30 3456)
23 %r1 = add i32 %c1, %c2
24 %r2 = add i32 %r1, %c3
28 declare i32 @llvm.part.set.i32.i32.i32(i32 %x, i32 %rep, i32 %hi, i32 %lo)
29 declare i16 @llvm.part.set.i16.i16.i16(i16 %x, i16 %rep, i32 %hi, i32 %lo)
30 define i32 @test_part_set(i32 %A, i16 %B) {
31 %a = call i32 @llvm.part.set.i32.i32.i32(i32 %A, i32 27, i32 8, i32 0)
32 %b = call i16 @llvm.part.set.i16.i16.i16(i16 %B, i16 27, i32 8, i32 0)
33 %c = zext i16 %b to i32
38 declare i32 @llvm.part.select.i32.i32(i32 %x, i32 %hi, i32 %lo)
39 declare i16 @llvm.part.select.i16.i16(i16 %x, i32 %hi, i32 %lo)
40 define i32 @test_part_select(i32 %A, i16 %B) {
41 %a = call i32 @llvm.part.select.i32.i32(i32 %A, i32 8, i32 0)
42 %b = call i16 @llvm.part.select.i16.i16(i16 %B, i32 8, i32 0)
43 %c = zext i16 %b to i32
48 declare i32 @llvm.bswap.i32.i32(i32 %x)
49 declare i16 @llvm.bswap.i16.i16(i16 %x)
50 define i32 @test_bswap(i32 %A, i16 %B) {
51 %a = call i32 @llvm.bswap.i32.i32(i32 %A)
52 %b = call i16 @llvm.bswap.i16.i16(i16 %B)
53 %c = zext i16 %b to i32
58 declare <4 x i16> @llvm.x86.mmx.psra.w(<4 x i16>, <2 x i32>) nounwind readnone
59 declare <4 x i16> @llvm.x86.mmx.psll.w(<4 x i16>, <2 x i32>) nounwind readnone
60 declare <4 x i16> @llvm.x86.mmx.psrl.w(<4 x i16>, <2 x i32>) nounwind readnone
61 define void @sh16(<4 x i16> %A, <2 x i32> %B) {
62 %r1 = call <4 x i16> @llvm.x86.mmx.psra.w( <4 x i16> %A, <2 x i32> %B ) ; <<4 x i16>> [#uses=0]
63 %r2 = call <4 x i16> @llvm.x86.mmx.psll.w( <4 x i16> %A, <2 x i32> %B ) ; <<4 x i16>> [#uses=0]
64 %r3 = call <4 x i16> @llvm.x86.mmx.psrl.w( <4 x i16> %A, <2 x i32> %B ) ; <<4 x i16>> [#uses=0]
68 declare <2 x i32> @llvm.x86.mmx.psra.d(<2 x i32>, <2 x i32>) nounwind readnone
69 declare <2 x i32> @llvm.x86.mmx.psll.d(<2 x i32>, <2 x i32>) nounwind readnone
70 declare <2 x i32> @llvm.x86.mmx.psrl.d(<2 x i32>, <2 x i32>) nounwind readnone
71 define void @sh32(<2 x i32> %A, <2 x i32> %B) {
72 %r1 = call <2 x i32> @llvm.x86.mmx.psra.d( <2 x i32> %A, <2 x i32> %B ) ; <<2 x i32>> [#uses=0]
73 %r2 = call <2 x i32> @llvm.x86.mmx.psll.d( <2 x i32> %A, <2 x i32> %B ) ; <<2 x i32>> [#uses=0]
74 %r3 = call <2 x i32> @llvm.x86.mmx.psrl.d( <2 x i32> %A, <2 x i32> %B ) ; <<2 x i32>> [#uses=0]
78 declare <1 x i64> @llvm.x86.mmx.psll.q(<1 x i64>, <2 x i32>) nounwind readnone
79 declare <1 x i64> @llvm.x86.mmx.psrl.q(<1 x i64>, <2 x i32>) nounwind readnone
80 define void @sh64(<1 x i64> %A, <2 x i32> %B) {
81 %r1 = call <1 x i64> @llvm.x86.mmx.psll.q( <1 x i64> %A, <2 x i32> %B ) ; <<1 x i64>> [#uses=0]
82 %r2 = call <1 x i64> @llvm.x86.mmx.psrl.q( <1 x i64> %A, <2 x i32> %B ) ; <<1 x i64>> [#uses=0]
86 declare <4 x float> @llvm.x86.sse.loadu.ps(i8*) nounwind readnone
87 declare <16 x i8> @llvm.x86.sse2.loadu.dq(i8*) nounwind readnone
88 declare <2 x double> @llvm.x86.sse2.loadu.pd(double*) nounwind readnone
89 define void @test_loadu(i8* %a, double* %b) {
90 %v0 = call <4 x float> @llvm.x86.sse.loadu.ps(i8* %a)
91 %v1 = call <16 x i8> @llvm.x86.sse2.loadu.dq(i8* %a)
92 %v2 = call <2 x double> @llvm.x86.sse2.loadu.pd(double* %b)
96 declare void @llvm.x86.sse.movnt.ps(i8*, <4 x float>) nounwind readnone
97 declare void @llvm.x86.sse2.movnt.dq(i8*, <2 x double>) nounwind readnone
98 declare void @llvm.x86.sse2.movnt.pd(i8*, <2 x double>) nounwind readnone
99 declare void @llvm.x86.sse2.movnt.i(i8*, i32) nounwind readnone
101 define void @f(<4 x float> %A, i8* %B, <2 x double> %C, i32 %D) {
102 ; CHECK: store{{.*}}nontemporal
103 call void @llvm.x86.sse.movnt.ps(i8* %B, <4 x float> %A)
104 ; CHECK: store{{.*}}nontemporal
105 call void @llvm.x86.sse2.movnt.dq(i8* %B, <2 x double> %C)
106 ; CHECK: store{{.*}}nontemporal
107 call void @llvm.x86.sse2.movnt.pd(i8* %B, <2 x double> %C)
108 ; CHECK: store{{.*}}nontemporal
109 call void @llvm.x86.sse2.movnt.i(i8* %B, i32 %D)