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
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
|
https://bugs.gentoo.org/915060
https://git.gnupg.org/cgi-bin/gitweb.cgi?p=libgcrypt.git;a=commit;h=07f0563d325c6589ca1560525d3b22291feec227
From 07f0563d325c6589ca1560525d3b22291feec227 Mon Sep 17 00:00:00 2001
From: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Date: Tue, 19 Dec 2023 20:23:47 +0200
Subject: [PATCH] mpi/ec-inline: refactor i386 assembly to reduce register
usage
* mpi/ec-inline.h [__i386__] (ADD2_LIMB32_CARRY_OUT)
(ADD2_LIMB32_CARRY_IN_OUT, ADD2_LIB32_CARRY_IN, SUB2_LIMB32_CARRY_OUT)
(SUB2_LIMB32_CARRY_IN_OUT, SUB2_LIB32_CARRY_IN, ADD8_LIMB32)
(ADD10_LIMB32, ADD14_LIMB32, SUB8_LIMB32, SUB10_LIMB32)
(SUB14_LIMB32): New.
[__i386__] (ADD4_LIMB32, ADD6_LIMB32, SUB4_LIMB32, SUB6_LIMB32): Rewrite
to use new *_CARRY_* macros.
[BYTES_PER_MPI_LIMB == 4] (ADD4_LIMB64): Use ADD8_LIMB32 if available.
[BYTES_PER_MPI_LIMB == 4] (ADD5_LIMB64): Use ADD10_LIMB32 if available.
[BYTES_PER_MPI_LIMB == 4] (ADD7_LIMB64): Use ADD14_LIMB32 if available.
[BYTES_PER_MPI_LIMB == 4] (SUB4_LIMB64): Use SUB8_LIMB32 if available.
[BYTES_PER_MPI_LIMB == 4] (SUB5_LIMB64): Use SUB10_LIMB32 if available.
[BYTES_PER_MPI_LIMB == 4] (SUB7_LIMB64): Use SUB14_LIMB32 if available.
--
Cherry pick of master commit:
956f1ed4ec6ead59dc56f574f943f1fe25dac723
This commit reduces number register operands and register pressure
at i386 assembly used in `ec-nist.c` NIST-P192, P224, P256, and P384.
Performance stays relatively same, with P192 being ~2% slower
than before and P384 being ~5% faster.
GnuPG-bug-id: T6892
Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
---
mpi/ec-inline.h | 327 +++++++++++++++++++++++++++++++++---------------
1 file changed, 229 insertions(+), 98 deletions(-)
diff --git a/mpi/ec-inline.h b/mpi/ec-inline.h
index a07826e3..0ffdf8eb 100644
--- a/mpi/ec-inline.h
+++ b/mpi/ec-inline.h
@@ -641,116 +641,192 @@ LIMB64_HILO(mpi_limb_t hi, mpi_limb_t lo)
/* i386 addition/subtraction helpers. */
#if defined (__i386__) && defined(HAVE_CPU_ARCH_X86) && __GNUC__ >= 4
-#define ADD4_LIMB32(a3, a2, a1, a0, b3, b2, b1, b0, c3, c2, c1, c0) \
- __asm__ ("addl %11, %3\n" \
- "adcl %10, %2\n" \
- "adcl %9, %1\n" \
- "adcl %8, %0\n" \
- : "=r" (a3), \
- "=&r" (a2), \
+#define ADD2_LIMB32_CARRY_OUT(carry, a1, a0, b1, b0, c1, c0) \
+ __asm__ ("addl %7, %2\n" \
+ "adcl %6, %1\n" \
+ "sbbl %0, %0\n" \
+ : "=r" (carry), \
"=&r" (a1), \
"=&r" (a0) \
- : "0" ((mpi_limb_t)(b3)), \
- "1" ((mpi_limb_t)(b2)), \
- "2" ((mpi_limb_t)(b1)), \
- "3" ((mpi_limb_t)(b0)), \
- "g" ((mpi_limb_t)(c3)), \
- "g" ((mpi_limb_t)(c2)), \
- "g" ((mpi_limb_t)(c1)), \
- "g" ((mpi_limb_t)(c0)) \
+ : "0" ((mpi_limb_t)(0)), \
+ "1" ((mpi_limb_t)(b1)), \
+ "2" ((mpi_limb_t)(b0)), \
+ "re" ((mpi_limb_t)(c1)), \
+ "re" ((mpi_limb_t)(c0)) \
: "cc")
+#define ADD2_LIMB32_CARRY_IN_OUT(a1, a0, b1, b0, c1, c0, carry) \
+ __asm__ ("addl $1, %0\n" \
+ "adcl %7, %2\n" \
+ "adcl %6, %1\n" \
+ "sbbl %0, %0\n" \
+ : "=r" (carry), \
+ "=&r" (a1), \
+ "=&r" (a0) \
+ : "0" ((mpi_limb_t)(carry)), \
+ "1" ((mpi_limb_t)(b1)), \
+ "2" ((mpi_limb_t)(b0)), \
+ "re" ((mpi_limb_t)(c1)), \
+ "re" ((mpi_limb_t)(c0)) \
+ : "cc")
+
+#define ADD2_LIMB32_CARRY_IN(a1, a0, b1, b0, c1, c0, carry) \
+ __asm__ ("addl $1, %2\n" \
+ "adcl %7, %1\n" \
+ "adcl %6, %0\n" \
+ : "=r" (a1), \
+ "=&r" (a0), \
+ "=&g" (carry) \
+ : "0" ((mpi_limb_t)(b1)), \
+ "1" ((mpi_limb_t)(b0)), \
+ "2" ((mpi_limb_t)(carry)), \
+ "re" ((mpi_limb_t)(c1)), \
+ "re" ((mpi_limb_t)(c0)) \
+ : "cc")
+
+#define ADD4_LIMB32(a3, a2, a1, a0, b3, b2, b1, b0, c3, c2, c1, c0) do { \
+ mpi_limb_t __carry4_32; \
+ ADD2_LIMB32_CARRY_OUT(__carry4_32, a1, a0, b1, b0, c1, c0); \
+ ADD2_LIMB32_CARRY_IN(a3, a2, b3, b2, c3, c2, __carry4_32); \
+ } while (0)
+
#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__ ("addl %10, %3\n" \
- "adcl %9, %2\n" \
- "adcl %8, %1\n" \
- "sbbl %0, %0\n" \
- : "=r" (__carry6_32), \
- "=&r" (a2), \
- "=&r" (a1), \
- "=&r" (a0) \
- : "0" ((mpi_limb_t)(0)), \
- "1" ((mpi_limb_t)(b2)), \
- "2" ((mpi_limb_t)(b1)), \
- "3" ((mpi_limb_t)(b0)), \
- "g" ((mpi_limb_t)(c2)), \
- "g" ((mpi_limb_t)(c1)), \
- "g" ((mpi_limb_t)(c0)) \
- : "cc"); \
- __asm__ ("addl $1, %3\n" \
- "adcl %10, %2\n" \
- "adcl %9, %1\n" \
- "adcl %8, %0\n" \
- : "=r" (a5), \
- "=&r" (a4), \
- "=&r" (a3), \
- "=&r" (__carry6_32) \
- : "0" ((mpi_limb_t)(b5)), \
- "1" ((mpi_limb_t)(b4)), \
- "2" ((mpi_limb_t)(b3)), \
- "3" ((mpi_limb_t)(__carry6_32)), \
- "g" ((mpi_limb_t)(c5)), \
- "g" ((mpi_limb_t)(c4)), \
- "g" ((mpi_limb_t)(c3)) \
- : "cc"); \
+ ADD2_LIMB32_CARRY_OUT(__carry6_32, a1, a0, b1, b0, c1, c0); \
+ ADD2_LIMB32_CARRY_IN_OUT(a3, a2, b3, b2, c3, c2, __carry6_32); \
+ ADD2_LIMB32_CARRY_IN(a5, a4, b5, b4, c5, c4, __carry6_32); \
+ } while (0)
+
+#define ADD8_LIMB32(a7, a6, a5, a4, a3, a2, a1, a0, \
+ b7, b6, b5, b4, b3, b2, b1, b0, \
+ c7, c6, c5, c4, c3, c2, c1, c0) do { \
+ mpi_limb_t __carry8_32; \
+ ADD2_LIMB32_CARRY_OUT(__carry8_32, a1, a0, b1, b0, c1, c0); \
+ ADD2_LIMB32_CARRY_IN_OUT(a3, a2, b3, b2, c3, c2, __carry8_32); \
+ ADD2_LIMB32_CARRY_IN_OUT(a5, a4, b5, b4, c5, c4, __carry8_32); \
+ ADD2_LIMB32_CARRY_IN(a7, a6, b7, b6, c7, c6, __carry8_32); \
} while (0)
-#define SUB4_LIMB32(a3, a2, a1, a0, b3, b2, b1, b0, c3, c2, c1, c0) \
- __asm__ ("subl %11, %3\n" \
- "sbbl %10, %2\n" \
- "sbbl %9, %1\n" \
- "sbbl %8, %0\n" \
- : "=r" (a3), \
- "=&r" (a2), \
+#define ADD10_LIMB32(a9, a8, a7, a6, a5, a4, a3, a2, a1, a0, \
+ b9, b8, b7, b6, b5, b4, b3, b2, b1, b0, \
+ c9, c8, c7, c6, c5, c4, c3, c2, c1, c0) do { \
+ mpi_limb_t __carry10_32; \
+ ADD2_LIMB32_CARRY_OUT(__carry10_32, a1, a0, b1, b0, c1, c0); \
+ ADD2_LIMB32_CARRY_IN_OUT(a3, a2, b3, b2, c3, c2, __carry10_32); \
+ ADD2_LIMB32_CARRY_IN_OUT(a5, a4, b5, b4, c5, c4, __carry10_32); \
+ ADD2_LIMB32_CARRY_IN_OUT(a7, a6, b7, b6, c7, c6, __carry10_32); \
+ ADD2_LIMB32_CARRY_IN(a9, a8, b9, b8, c9, c8, __carry10_32); \
+ } while (0)
+
+#define ADD14_LIMB32(a13, a12, a11, a10, a9, a8, a7, \
+ a6, a5, a4, a3, a2, a1, a0, \
+ b13, b12, b11, b10, b9, b8, b7, \
+ b6, b5, b4, b3, b2, b1, b0, \
+ c13, c12, c11, c10, c9, c8, c7, \
+ c6, c5, c4, c3, c2, c1, c0) do { \
+ mpi_limb_t __carry14_32; \
+ ADD2_LIMB32_CARRY_OUT(__carry14_32, a1, a0, b1, b0, c1, c0); \
+ ADD2_LIMB32_CARRY_IN_OUT(a3, a2, b3, b2, c3, c2, __carry14_32); \
+ ADD2_LIMB32_CARRY_IN_OUT(a5, a4, b5, b4, c5, c4, __carry14_32); \
+ ADD2_LIMB32_CARRY_IN_OUT(a7, a6, b7, b6, c7, c6, __carry14_32); \
+ ADD2_LIMB32_CARRY_IN_OUT(a9, a8, b9, b8, c9, c8, __carry14_32); \
+ ADD2_LIMB32_CARRY_IN_OUT(a11, a10, b11, b10, c11, c10, __carry14_32); \
+ ADD2_LIMB32_CARRY_IN(a13, a12, b13, b12, c13, c12, __carry14_32); \
+ } while (0)
+
+#define SUB2_LIMB32_CARRY_OUT(carry, a1, a0, b1, b0, c1, c0) \
+ __asm__ ("subl %7, %2\n" \
+ "sbbl %6, %1\n" \
+ "sbbl %0, %0\n" \
+ : "=r" (carry), \
"=&r" (a1), \
"=&r" (a0) \
- : "0" ((mpi_limb_t)(b3)), \
- "1" ((mpi_limb_t)(b2)), \
- "2" ((mpi_limb_t)(b1)), \
- "3" ((mpi_limb_t)(b0)), \
- "g" ((mpi_limb_t)(c3)), \
- "g" ((mpi_limb_t)(c2)), \
- "g" ((mpi_limb_t)(c1)), \
- "g" ((mpi_limb_t)(c0)) \
+ : "0" ((mpi_limb_t)(0)), \
+ "1" ((mpi_limb_t)(b1)), \
+ "2" ((mpi_limb_t)(b0)), \
+ "re" ((mpi_limb_t)(c1)), \
+ "re" ((mpi_limb_t)(c0)) \
+ : "cc")
+
+#define SUB2_LIMB32_CARRY_IN_OUT(a1, a0, b1, b0, c1, c0, carry) \
+ __asm__ ("addl $1, %0\n" \
+ "sbbl %7, %2\n" \
+ "sbbl %6, %1\n" \
+ "sbbl %0, %0\n" \
+ : "=r" (carry), \
+ "=&r" (a1), \
+ "=&r" (a0) \
+ : "0" ((mpi_limb_t)(carry)), \
+ "1" ((mpi_limb_t)(b1)), \
+ "2" ((mpi_limb_t)(b0)), \
+ "re" ((mpi_limb_t)(c1)), \
+ "re" ((mpi_limb_t)(c0)) \
+ : "cc")
+
+#define SUB2_LIMB32_CARRY_IN(a1, a0, b1, b0, c1, c0, carry) \
+ __asm__ ("addl $1, %2\n" \
+ "sbbl %7, %1\n" \
+ "sbbl %6, %0\n" \
+ : "=r" (a1), \
+ "=&r" (a0), \
+ "=&g" (carry) \
+ : "0" ((mpi_limb_t)(b1)), \
+ "1" ((mpi_limb_t)(b0)), \
+ "2" ((mpi_limb_t)(carry)), \
+ "re" ((mpi_limb_t)(c1)), \
+ "re" ((mpi_limb_t)(c0)) \
: "cc")
+#define SUB4_LIMB32(a3, a2, a1, a0, b3, b2, b1, b0, c3, c2, c1, c0) do { \
+ mpi_limb_t __carry4_32; \
+ SUB2_LIMB32_CARRY_OUT(__carry4_32, a1, a0, b1, b0, c1, c0); \
+ SUB2_LIMB32_CARRY_IN(a3, a2, b3, b2, c3, c2, __carry4_32); \
+ } while (0)
+
#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__ ("subl %10, %3\n" \
- "sbbl %9, %2\n" \
- "sbbl %8, %1\n" \
- "sbbl %0, %0\n" \
- : "=r" (__borrow6_32), \
- "=&r" (a2), \
- "=&r" (a1), \
- "=&r" (a0) \
- : "0" ((mpi_limb_t)(0)), \
- "1" ((mpi_limb_t)(b2)), \
- "2" ((mpi_limb_t)(b1)), \
- "3" ((mpi_limb_t)(b0)), \
- "g" ((mpi_limb_t)(c2)), \
- "g" ((mpi_limb_t)(c1)), \
- "g" ((mpi_limb_t)(c0)) \
- : "cc"); \
- __asm__ ("addl $1, %3\n" \
- "sbbl %10, %2\n" \
- "sbbl %9, %1\n" \
- "sbbl %8, %0\n" \
- : "=r" (a5), \
- "=&r" (a4), \
- "=&r" (a3), \
- "=&r" (__borrow6_32) \
- : "0" ((mpi_limb_t)(b5)), \
- "1" ((mpi_limb_t)(b4)), \
- "2" ((mpi_limb_t)(b3)), \
- "3" ((mpi_limb_t)(__borrow6_32)), \
- "g" ((mpi_limb_t)(c5)), \
- "g" ((mpi_limb_t)(c4)), \
- "g" ((mpi_limb_t)(c3)) \
- : "cc"); \
+ mpi_limb_t __carry6_32; \
+ SUB2_LIMB32_CARRY_OUT(__carry6_32, a1, a0, b1, b0, c1, c0); \
+ SUB2_LIMB32_CARRY_IN_OUT(a3, a2, b3, b2, c3, c2, __carry6_32); \
+ SUB2_LIMB32_CARRY_IN(a5, a4, b5, b4, c5, c4, __carry6_32); \
+ } while (0)
+
+#define SUB8_LIMB32(a7, a6, a5, a4, a3, a2, a1, a0, \
+ b7, b6, b5, b4, b3, b2, b1, b0, \
+ c7, c6, c5, c4, c3, c2, c1, c0) do { \
+ mpi_limb_t __carry8_32; \
+ SUB2_LIMB32_CARRY_OUT(__carry8_32, a1, a0, b1, b0, c1, c0); \
+ SUB2_LIMB32_CARRY_IN_OUT(a3, a2, b3, b2, c3, c2, __carry8_32); \
+ SUB2_LIMB32_CARRY_IN_OUT(a5, a4, b5, b4, c5, c4, __carry8_32); \
+ SUB2_LIMB32_CARRY_IN(a7, a6, b7, b6, c7, c6, __carry8_32); \
+ } while (0)
+
+#define SUB10_LIMB32(a9, a8, a7, a6, a5, a4, a3, a2, a1, a0, \
+ b9, b8, b7, b6, b5, b4, b3, b2, b1, b0, \
+ c9, c8, c7, c6, c5, c4, c3, c2, c1, c0) do { \
+ mpi_limb_t __carry10_32; \
+ SUB2_LIMB32_CARRY_OUT(__carry10_32, a1, a0, b1, b0, c1, c0); \
+ SUB2_LIMB32_CARRY_IN_OUT(a3, a2, b3, b2, c3, c2, __carry10_32); \
+ SUB2_LIMB32_CARRY_IN_OUT(a5, a4, b5, b4, c5, c4, __carry10_32); \
+ SUB2_LIMB32_CARRY_IN_OUT(a7, a6, b7, b6, c7, c6, __carry10_32); \
+ SUB2_LIMB32_CARRY_IN(a9, a8, b9, b8, c9, c8, __carry10_32); \
+ } while (0)
+
+#define SUB14_LIMB32(a13, a12, a11, a10, a9, a8, a7, \
+ a6, a5, a4, a3, a2, a1, a0, \
+ b13, b12, b11, b10, b9, b8, b7, \
+ b6, b5, b4, b3, b2, b1, b0, \
+ c13, c12, c11, c10, c9, c8, c7, \
+ c6, c5, c4, c3, c2, c1, c0) do { \
+ mpi_limb_t __carry14_32; \
+ SUB2_LIMB32_CARRY_OUT(__carry14_32, a1, a0, b1, b0, c1, c0); \
+ SUB2_LIMB32_CARRY_IN_OUT(a3, a2, b3, b2, c3, c2, __carry14_32); \
+ SUB2_LIMB32_CARRY_IN_OUT(a5, a4, b5, b4, c5, c4, __carry14_32); \
+ SUB2_LIMB32_CARRY_IN_OUT(a7, a6, b7, b6, c7, c6, __carry14_32); \
+ SUB2_LIMB32_CARRY_IN_OUT(a9, a8, b9, b8, c9, c8, __carry14_32); \
+ SUB2_LIMB32_CARRY_IN_OUT(a11, a10, b11, b10, c11, c10, __carry14_32); \
+ SUB2_LIMB32_CARRY_IN(a13, a12, b13, b12, c13, c12, __carry14_32); \
} while (0)
#endif /* __i386__ */
@@ -820,7 +896,6 @@ LIMB64_HILO(mpi_limb_t hi, mpi_limb_t lo)
"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; \
@@ -875,7 +950,13 @@ LIMB64_HILO(mpi_limb_t hi, mpi_limb_t lo)
C2.hi, C2.lo, C1.hi, C1.lo, C0.hi, C0.lo)
#endif
-#if defined(ADD6_LIMB32)
+#if defined(ADD8_LIMB32)
+/* A[0..3] = B[0..3] + C[0..3] */
+#define ADD4_LIMB64(A3, A2, A1, A0, B3, B2, B1, B0, C3, C2, C1, C0) \
+ ADD8_LIMB32(A3.hi, A3.lo, A2.hi, A2.lo, A1.hi, A1.lo, A0.hi, A0.lo, \
+ B3.hi, B3.lo, B2.hi, B2.lo, B1.hi, B1.lo, B0.hi, B0.lo, \
+ C3.hi, C3.lo, C2.hi, C2.lo, C1.hi, C1.lo, C0.hi, C0.lo)
+#elif defined(ADD6_LIMB32)
/* A[0..3] = B[0..3] + C[0..3] */
#define ADD4_LIMB64(A3, A2, A1, A0, B3, B2, B1, B0, C3, C2, C1, C0) do { \
mpi_limb_t __carry4; \
@@ -888,6 +969,28 @@ LIMB64_HILO(mpi_limb_t hi, mpi_limb_t lo)
} while (0)
#endif
+#if defined(ADD10_LIMB32)
+/* A[0..4] = B[0..4] + C[0..4] */
+#define ADD5_LIMB64(A4, A3, A2, A1, A0, B4, B3, B2, B1, B0, \
+ C4, C3, C2, C1, C0) \
+ ADD10_LIMB32(A4.hi, A4.lo, A3.hi, A3.lo, A2.hi, A2.lo, A1.hi, A1.lo, \
+ A0.hi, A0.lo, B4.hi, B4.lo, B3.hi, B3.lo, B2.hi, B2.lo, \
+ B1.hi, B1.lo, B0.hi, B0.lo, C4.hi, C4.lo, C3.hi, C3.lo, \
+ C2.hi, C2.lo, C1.hi, C1.lo, C0.hi, C0.lo)
+#endif
+
+#if defined(ADD14_LIMB32)
+/* A[0..6] = B[0..6] + C[0..6] */
+#define ADD7_LIMB64(A6, A5, A4, A3, A2, A1, A0, B6, B5, B4, B3, B2, B1, B0, \
+ C6, C5, C4, C3, C2, C1, C0) \
+ ADD14_LIMB32(A6.hi, A6.lo, A5.hi, A5.lo, A4.hi, A4.lo, A3.hi, A3.lo, \
+ A2.hi, A2.lo, A1.hi, A1.lo, A0.hi, A0.lo, B6.hi, B6.lo, \
+ B5.hi, B5.lo, B4.hi, B4.lo, B3.hi, B3.lo, B2.hi, B2.lo, \
+ B1.hi, B1.lo, B0.hi, B0.lo, C6.hi, C6.lo, C5.hi, C5.lo, \
+ C4.hi, C4.lo, C3.hi, C3.lo, C2.hi, C2.lo, C1.hi, C1.lo, \
+ C0.hi, C0.lo)
+#endif
+
#if defined(SUB4_LIMB32)
/* A[0..1] = B[0..1] - C[0..1] */
#define SUB2_LIMB64(A1, A0, B1, B0, C1, C0) \
@@ -914,7 +1017,13 @@ LIMB64_HILO(mpi_limb_t hi, mpi_limb_t lo)
C2.hi, C2.lo, C1.hi, C1.lo, C0.hi, C0.lo)
#endif
-#if defined(SUB6_LIMB32)
+#if defined(SUB8_LIMB32)
+/* A[0..3] = B[0..3] - C[0..3] */
+#define SUB4_LIMB64(A3, A2, A1, A0, B3, B2, B1, B0, C3, C2, C1, C0) \
+ SUB8_LIMB32(A3.hi, A3.lo, A2.hi, A2.lo, A1.hi, A1.lo, A0.hi, A0.lo, \
+ B3.hi, B3.lo, B2.hi, B2.lo, B1.hi, B1.lo, B0.hi, B0.lo, \
+ C3.hi, C3.lo, C2.hi, C2.lo, C1.hi, C1.lo, C0.hi, C0.lo)
+#elif defined(SUB6_LIMB32)
/* A[0..3] = B[0..3] - C[0..3] */
#define SUB4_LIMB64(A3, A2, A1, A0, B3, B2, B1, B0, C3, C2, C1, C0) do { \
mpi_limb_t __borrow4; \
@@ -927,6 +1036,28 @@ LIMB64_HILO(mpi_limb_t hi, mpi_limb_t lo)
} while (0)
#endif
+#if defined(SUB10_LIMB32)
+/* A[0..4] = B[0..4] - C[0..4] */
+#define SUB5_LIMB64(A4, A3, A2, A1, A0, B4, B3, B2, B1, B0, \
+ C4, C3, C2, C1, C0) \
+ SUB10_LIMB32(A4.hi, A4.lo, A3.hi, A3.lo, A2.hi, A2.lo, A1.hi, A1.lo, \
+ A0.hi, A0.lo, B4.hi, B4.lo, B3.hi, B3.lo, B2.hi, B2.lo, \
+ B1.hi, B1.lo, B0.hi, B0.lo, C4.hi, C4.lo, C3.hi, C3.lo, \
+ C2.hi, C2.lo, C1.hi, C1.lo, C0.hi, C0.lo)
+#endif
+
+#if defined(SUB14_LIMB32)
+/* A[0..6] = B[0..6] - C[0..6] */
+#define SUB7_LIMB64(A6, A5, A4, A3, A2, A1, A0, B6, B5, B4, B3, B2, B1, B0, \
+ C6, C5, C4, C3, C2, C1, C0) \
+ SUB14_LIMB32(A6.hi, A6.lo, A5.hi, A5.lo, A4.hi, A4.lo, A3.hi, A3.lo, \
+ A2.hi, A2.lo, A1.hi, A1.lo, A0.hi, A0.lo, B6.hi, B6.lo, \
+ B5.hi, B5.lo, B4.hi, B4.lo, B3.hi, B3.lo, B2.hi, B2.lo, \
+ B1.hi, B1.lo, B0.hi, B0.lo, C6.hi, C6.lo, C5.hi, C5.lo, \
+ C4.hi, C4.lo, C3.hi, C3.lo, C2.hi, C2.lo, C1.hi, C1.lo, \
+ C0.hi, C0.lo)
+#endif
+
#endif /* BYTES_PER_MPI_LIMB == 4 */
--
2.30.2
|
