Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(385)

Side by Side Diff: tests_lit/llvm2ice_tests/square.ll

Issue 1377213004: Subzero: Improve lowering sequence for "a=b*b". (Closed) Base URL: https://chromium.googlesource.com/native_client/pnacl-subzero.git@master
Patch Set: Fix comment Created 5 years, 2 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View unified diff | Download patch
« no previous file with comments | « src/IceTargetLoweringX86BaseImpl.h ('k') | no next file » | no next file with comments »
Toggle Intra-line Diffs ('i') | Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
OLDNEW
(Empty)
1 ; Test the a=b*b lowering sequence which can use a single temporary register
2 ; instead of two registers.
3
4 ; RUN: %if --need=target_X8632 --command %p2i --filetype=obj --disassemble \
5 ; RUN: --target x8632 -i %s --args -O2 -mattr=sse4.1 \
6 ; RUN: | %if --need=target_X8632 --command FileCheck %s
7
8 ; RUN: %if --need=target_X8632 --command %p2i --filetype=obj --disassemble \
9 ; RUN: --target x8632 -i %s --args -Om1 -mattr=sse4.1 \
10 ; RUN: | %if --need=target_X8632 --command FileCheck %s
11
12 define float @Square_float(float %a) {
13 entry:
14 %result = fmul float %a, %a
15 ret float %result
16 }
17 ; CHECK-LABEL: Square_float
18 ; CHECK: mulss [[REG:xmm.]],[[REG]]
19
20 define double @Square_double(double %a) {
21 entry:
22 %result = fmul double %a, %a
23 ret double %result
24 }
25 ; CHECK-LABEL: Square_double
26 ; CHECK: mulsd [[REG:xmm.]],[[REG]]
27
28 define i32 @Square_i32(i32 %a) {
29 entry:
30 %result = mul i32 %a, %a
31 ret i32 %result
32 }
33 ; CHECK-LABEL: Square_i32
34 ; CHECK: imul [[REG:e..]],[[REG]]
35
36 define i16 @Square_i16(i16 %a) {
37 entry:
38 %result = mul i16 %a, %a
39 ret i16 %result
40 }
41 ; CHECK-LABEL: Square_i16
42 ; CHECK: imul [[REG:..]],[[REG]]
43
44 define i8 @Square_i8(i8 %a) {
45 entry:
46 %result = mul i8 %a, %a
47 ret i8 %result
48 }
49 ; CHECK-LABEL: Square_i8
50 ; CHECK: imul al
51
52 define <4 x float> @Square_v4f32(<4 x float> %a) {
53 entry:
54 %result = fmul <4 x float> %a, %a
55 ret <4 x float> %result
56 }
57 ; CHECK-LABEL: Square_v4f32
58 ; CHECK: mulps [[REG:xmm.]],[[REG]]
59
60 define <4 x i32> @Square_v4i32(<4 x i32> %a) {
61 entry:
62 %result = mul <4 x i32> %a, %a
63 ret <4 x i32> %result
64 }
65 ; CHECK-LABEL: Square_v4i32
66 ; CHECK: pmulld [[REG:xmm.]],[[REG]]
67
68 define <8 x i16> @Square_v8i16(<8 x i16> %a) {
69 entry:
70 %result = mul <8 x i16> %a, %a
71 ret <8 x i16> %result
72 }
73 ; CHECK-LABEL: Square_v8i16
74 ; CHECK: pmullw [[REG:xmm.]],[[REG]]
75
76 define <16 x i8> @Square_v16i8(<16 x i8> %a) {
77 entry:
78 %result = mul <16 x i8> %a, %a
79 ret <16 x i8> %result
80 }
81 ; CHECK-LABEL: Square_v16i8
82 ; CHECK-NOT: pmul
OLDNEW
« no previous file with comments | « src/IceTargetLoweringX86BaseImpl.h ('k') | no next file » | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698