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
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
|
From 9c93b4607adcf9b3efd53aba43e2d33bf5aef9df Mon Sep 17 00:00:00 2001
From: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Date: Sun, 4 Aug 2024 18:04:49 +0300
Subject: [PATCH] mpi/ec-inline: reduce register pressure on 32-bit ARM
* mpi/ec-inline.h [HAVE_COMPATIBLE_GCC_ARM_PLATFORM_AS] (ADD4_LIMB32)
(ADD6_LIMB32, SUB4_LIMB32, SUB6_LIMB32): Reuse input registers
as output (use just two unique operators).
--
This fixes building ec-nist.c with GCC-14 on 32-bit ARM.
GnuPG-bug-id: 7226
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
---
mpi/ec-inline.h | 63 ++++++++++++++++++++++++-------------------------
1 file changed, 31 insertions(+), 32 deletions(-)
diff --git a/mpi/ec-inline.h b/mpi/ec-inline.h
index c24d5352..3a526246 100644
--- a/mpi/ec-inline.h
+++ b/mpi/ec-inline.h
@@ -832,89 +832,88 @@ LIMB64_HILO(mpi_limb_t hi, mpi_limb_t lo)
#endif /* __i386__ */
/* ARM addition/subtraction helpers. */
#ifdef HAVE_COMPATIBLE_GCC_ARM_PLATFORM_AS
#define ADD4_LIMB32(A3, A2, A1, A0, B3, B2, B1, B0, C3, C2, C1, C0) \
- __asm__ ("adds %3, %7, %11\n" \
- "adcs %2, %6, %10\n" \
- "adcs %1, %5, %9\n" \
- "adc %0, %4, %8\n" \
+ __asm__ ("adds %3, %3, %11\n" \
+ "adcs %2, %2, %10\n" \
+ "adcs %1, %1, %9\n" \
+ "adc %0, %0, %8\n" \
: "=r" (A3), \
"=&r" (A2), \
"=&r" (A1), \
"=&r" (A0) \
- : "r" ((mpi_limb_t)(B3)), \
- "r" ((mpi_limb_t)(B2)), \
- "r" ((mpi_limb_t)(B1)), \
- "r" ((mpi_limb_t)(B0)), \
+ : "0" ((mpi_limb_t)(B3)), \
+ "1" ((mpi_limb_t)(B2)), \
+ "2" ((mpi_limb_t)(B1)), \
+ "3" ((mpi_limb_t)(B0)), \
"Ir" ((mpi_limb_t)(C3)), \
"Ir" ((mpi_limb_t)(C2)), \
"Ir" ((mpi_limb_t)(C1)), \
"Ir" ((mpi_limb_t)(C0)) \
: "cc")
#define ADD6_LIMB32(A5, A4, A3, A2, A1, A0, B5, B4, B3, B2, B1, B0, \
C5, C4, C3, C2, C1, C0) do { \
mpi_limb_t __carry6_32; \
- __asm__ ("adds %3, %7, %10\n" \
- "adcs %2, %6, %9\n" \
- "adcs %1, %5, %8\n" \
- "adc %0, %4, %4\n" \
+ __asm__ ("adds %3, %3, %10\n" \
+ "adcs %2, %2, %9\n" \
+ "adcs %1, %1, %8\n" \
+ "adc %0, %0, %0\n" \
: "=r" (__carry6_32), \
"=&r" (A2), \
"=&r" (A1), \
"=&r" (A0) \
- : "r" ((mpi_limb_t)(0)), \
- "r" ((mpi_limb_t)(B2)), \
- "r" ((mpi_limb_t)(B1)), \
- "r" ((mpi_limb_t)(B0)), \
+ : "0" ((mpi_limb_t)(0)), \
+ "1" ((mpi_limb_t)(B2)), \
+ "2" ((mpi_limb_t)(B1)), \
+ "3" ((mpi_limb_t)(B0)), \
"Ir" ((mpi_limb_t)(C2)), \
"Ir" ((mpi_limb_t)(C1)), \
"Ir" ((mpi_limb_t)(C0)) \
: "cc"); \
ADD4_LIMB32(A5, A4, A3, __carry6_32, B5, B4, B3, __carry6_32, \
C5, C4, C3, 0xffffffffU); \
} while (0)
#define SUB4_LIMB32(A3, A2, A1, A0, B3, B2, B1, B0, C3, C2, C1, C0) \
- __asm__ ("subs %3, %7, %11\n" \
- "sbcs %2, %6, %10\n" \
- "sbcs %1, %5, %9\n" \
- "sbc %0, %4, %8\n" \
+ __asm__ ("subs %3, %3, %11\n" \
+ "sbcs %2, %2, %10\n" \
+ "sbcs %1, %1, %9\n" \
+ "sbc %0, %0, %8\n" \
: "=r" (A3), \
"=&r" (A2), \
"=&r" (A1), \
"=&r" (A0) \
- : "r" ((mpi_limb_t)(B3)), \
- "r" ((mpi_limb_t)(B2)), \
- "r" ((mpi_limb_t)(B1)), \
- "r" ((mpi_limb_t)(B0)), \
+ : "0" ((mpi_limb_t)(B3)), \
+ "1" ((mpi_limb_t)(B2)), \
+ "2" ((mpi_limb_t)(B1)), \
+ "3" ((mpi_limb_t)(B0)), \
"Ir" ((mpi_limb_t)(C3)), \
"Ir" ((mpi_limb_t)(C2)), \
"Ir" ((mpi_limb_t)(C1)), \
"Ir" ((mpi_limb_t)(C0)) \
: "cc")
#define SUB6_LIMB32(A5, A4, A3, A2, A1, A0, B5, B4, B3, B2, B1, B0, \
C5, C4, C3, C2, C1, C0) do { \
mpi_limb_t __borrow6_32; \
- __asm__ ("subs %3, %7, %10\n" \
- "sbcs %2, %6, %9\n" \
- "sbcs %1, %5, %8\n" \
- "sbc %0, %4, %4\n" \
+ __asm__ ("subs %3, %3, %9\n" \
+ "sbcs %2, %2, %8\n" \
+ "sbcs %1, %1, %7\n" \
+ "sbc %0, %0, %0\n" \
: "=r" (__borrow6_32), \
"=&r" (A2), \
"=&r" (A1), \
"=&r" (A0) \
- : "r" ((mpi_limb_t)(0)), \
- "r" ((mpi_limb_t)(B2)), \
- "r" ((mpi_limb_t)(B1)), \
- "r" ((mpi_limb_t)(B0)), \
+ : "1" ((mpi_limb_t)(B2)), \
+ "2" ((mpi_limb_t)(B1)), \
+ "3" ((mpi_limb_t)(B0)), \
"Ir" ((mpi_limb_t)(C2)), \
"Ir" ((mpi_limb_t)(C1)), \
"Ir" ((mpi_limb_t)(C0)) \
: "cc"); \
SUB4_LIMB32(A5, A4, A3, __borrow6_32, B5, B4, B3, 0, \
C5, C4, C3, -__borrow6_32); \
} while (0)
--
2.43.0
|
