1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
|
; RUN: llc -mtriple=hexagon < %s | FileCheck %s
; Testing for these 6 variants of circular load:
; Q6_circ_load_update_B(inputLR, pDelay, -1, nConvLength, 4);
; Q6_circ_load_update_D(inputLR, pDelay, -1, nConvLength, 4);
; Q6_circ_load_update_H(inputLR, pDelay, -1, nConvLength, 4);
; Q6_circ_load_update_UB(inputLR, pDelay, -1, nConvLength, 4);
; Q6_circ_load_update_UH(inputLR, pDelay, -1, nConvLength, 4);
; Q6_circ_load_update_W(inputLR, pDelay, -1, nConvLength, 4);
; producing these:
; r0 = memb(r1++#-1:circ(m0))
; r3:2 = memd(r1++#-8:circ(m0))
; r0 = memh(r1++#-2:circ(m0))
; r0 = memub(r1++#-1:circ(m0))
; r0 = memuh(r1++#-2:circ(m0))
; r0 = memw(r1++#-4:circ(m0))
target datalayout = "e-p:32:32:32-i64:64:64-i32:32:32-i16:16:16-i1:32:32-f64:64:64-f32:32:32-v64:64:64-v32:32:32-a0:0-n16:32"
target triple = "hexagon"
define signext i8 @foo1(i16 zeroext %filtMemLen, ptr %filtMemLR, i16 signext %filtMemIndex) nounwind {
entry:
%inputLR = alloca i8, align 1
%conv = zext i16 %filtMemLen to i32
%shr1 = lshr i32 %conv, 1
%idxprom = sext i16 %filtMemIndex to i32
%arrayidx = getelementptr inbounds i16, ptr %filtMemLR, i32 %idxprom
%or = or i32 %shr1, 33554432
; CHECK: = memb(r{{[0-9]+}}++#-1:circ(m{{[0-1]}}))
%0 = call ptr @llvm.hexagon.circ.ldb(ptr %arrayidx, ptr %inputLR, i32 %or, i32 -1)
%1 = load i8, ptr %inputLR, align 1, !tbaa !0
ret i8 %1
}
declare ptr @llvm.hexagon.circ.ldb(ptr, ptr, i32, i32) nounwind
define i64 @foo2(i16 zeroext %filtMemLen, ptr %filtMemLR, i16 signext %filtMemIndex) nounwind {
entry:
%inputLR = alloca i64, align 8
%conv = zext i16 %filtMemLen to i32
%shr1 = lshr i32 %conv, 1
%idxprom = sext i16 %filtMemIndex to i32
%arrayidx = getelementptr inbounds i16, ptr %filtMemLR, i32 %idxprom
%shl = shl nuw nsw i32 %shr1, 3
%or = or i32 %shl, 83886080
; CHECK: = memd(r{{[0-9]+}}++#-8:circ(m{{[0-1]}}))
%0 = call ptr @llvm.hexagon.circ.ldd(ptr %arrayidx, ptr %inputLR, i32 %or, i32 -8)
%1 = load i64, ptr %inputLR, align 8, !tbaa !0
ret i64 %1
}
declare ptr @llvm.hexagon.circ.ldd(ptr, ptr, i32, i32) nounwind
define signext i16 @foo3(i16 zeroext %filtMemLen, ptr %filtMemLR, i16 signext %filtMemIndex) nounwind {
entry:
%inputLR = alloca i16, align 2
%conv = zext i16 %filtMemLen to i32
%shr1 = and i32 %conv, 65534
%idxprom = sext i16 %filtMemIndex to i32
%arrayidx = getelementptr inbounds i16, ptr %filtMemLR, i32 %idxprom
%or = or i32 %shr1, 50331648
; CHECK: = memh(r{{[0-9]+}}++#-2:circ(m{{[0-1]}}))
%0 = call ptr @llvm.hexagon.circ.ldh(ptr %arrayidx, ptr %inputLR, i32 %or, i32 -2)
%1 = load i16, ptr %inputLR, align 2, !tbaa !2
ret i16 %1
}
declare ptr @llvm.hexagon.circ.ldh(ptr, ptr, i32, i32) nounwind
define zeroext i8 @foo4(i16 zeroext %filtMemLen, ptr %filtMemLR, i16 signext %filtMemIndex) nounwind {
entry:
%inputLR = alloca i8, align 1
%conv = zext i16 %filtMemLen to i32
%shr1 = lshr i32 %conv, 1
%idxprom = sext i16 %filtMemIndex to i32
%arrayidx = getelementptr inbounds i16, ptr %filtMemLR, i32 %idxprom
%or = or i32 %shr1, 33554432
; CHECK: = memub(r{{[0-9]+}}++#-1:circ(m{{[0-1]}}))
%0 = call ptr @llvm.hexagon.circ.ldub(ptr %arrayidx, ptr %inputLR, i32 %or, i32 -1)
%1 = load i8, ptr %inputLR, align 1, !tbaa !0
ret i8 %1
}
declare ptr @llvm.hexagon.circ.ldub(ptr, ptr, i32, i32) nounwind
define zeroext i16 @foo5(i16 zeroext %filtMemLen, ptr %filtMemLR, i16 signext %filtMemIndex) nounwind {
entry:
%inputLR = alloca i16, align 2
%conv = zext i16 %filtMemLen to i32
%shr1 = and i32 %conv, 65534
%idxprom = sext i16 %filtMemIndex to i32
%arrayidx = getelementptr inbounds i16, ptr %filtMemLR, i32 %idxprom
%or = or i32 %shr1, 50331648
; CHECK: = memuh(r{{[0-9]+}}++#-2:circ(m{{[0-1]}}))
%0 = call ptr @llvm.hexagon.circ.lduh(ptr %arrayidx, ptr %inputLR, i32 %or, i32 -2)
%1 = load i16, ptr %inputLR, align 2, !tbaa !2
ret i16 %1
}
declare ptr @llvm.hexagon.circ.lduh(ptr, ptr, i32, i32) nounwind
define i32 @foo6(i16 zeroext %filtMemLen, ptr %filtMemLR, i16 signext %filtMemIndex) nounwind {
entry:
%inputLR = alloca i32, align 4
%conv = zext i16 %filtMemLen to i32
%shr1 = lshr i32 %conv, 1
%idxprom = sext i16 %filtMemIndex to i32
%arrayidx = getelementptr inbounds i16, ptr %filtMemLR, i32 %idxprom
%shl = shl nuw nsw i32 %shr1, 2
%or = or i32 %shl, 67108864
; CHECK: = memw(r{{[0-9]+}}++#-4:circ(m{{[0-1]}}))
%0 = call ptr @llvm.hexagon.circ.ldw(ptr %arrayidx, ptr %inputLR, i32 %or, i32 -4)
%1 = load i32, ptr %inputLR, align 4, !tbaa !3
ret i32 %1
}
declare ptr @llvm.hexagon.circ.ldw(ptr, ptr, i32, i32) nounwind
!0 = !{!"omnipotent char", !1}
!1 = !{!"Simple C/C++ TBAA"}
!2 = !{!"short", !0}
!3 = !{!"int", !0}
|
