block-sha1: support for architectures with memory alignment restrictions
This is needed on architectures with poor or non-existent unaligned memory support and/or no fast byte swap instruction (such as ARM) by using byte accesses to memory and shifting the result together. This also makes the code portable, therefore the byte access methods are the defaults. Any architecture that properly supports unaligned word accesses in hardware simply has to enable the alternative methods. Signed-off-by: Nicolas Pitre <nico@cam.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Nicolas Pitre committed
Aug 12, 2009 at 15:47 UTC
660231aa9727d29c7d2c16319bc6a3fa8bed3e0e
1 file changed
+30
-2
block-sha1/sha1.c
+30
-2
@@ -60,6 +60,34 @@
60
#define setW(x, val) (W(x) = (val))
61
#endif
62
63
+/*
64
+ * Performance might be improved if the CPU architecture is OK with
65
+ * unaligned 32-bit loads and a fast ntohl() is available.
66
+ * Otherwise fall back to byte loads and shifts which is portable,
67
+ * and is faster on architectures with memory alignment issues.
68
+ */
69
+
70
+#if defined(__i386__) || defined(__x86_64__)
71
+
72
+#define get_be32(p) ntohl(*(unsigned int *)(p))
73
+#define put_be32(p, v) do { *(unsigned int *)(p) = htonl(v); } while (0)
74
+
75
+#else
76
+
77
+#define get_be32(p) ( \
78
+ (*((unsigned char *)(p) + 0) << 24) | \
79
+ (*((unsigned char *)(p) + 1) << 16) | \
80
+ (*((unsigned char *)(p) + 2) << 8) | \
81
+ (*((unsigned char *)(p) + 3) << 0) )
82
+#define put_be32(p, v) do { \
83
+ unsigned int __v = (v); \
84
+ *((unsigned char *)(p) + 0) = __v >> 24; \
85
+ *((unsigned char *)(p) + 1) = __v >> 16; \
86
+ *((unsigned char *)(p) + 2) = __v >> 8; \
87
+ *((unsigned char *)(p) + 3) = __v >> 0; } while (0)
88
+
89
+#endif
90
+
91
/* This "rolls" over the 512-bit array */
92
#define W(x) (array[(x)&15])
93
@@ -67,7 +95,7 @@
95
* Where do we get the source from? The first 16 iterations get it from
96
* the input data, the next mix it from the 512-bit array.
97
*/
70
-#define SHA_SRC(t) htonl(data[t])
98
+#define SHA_SRC(t) get_be32(data + t)
99
#define SHA_MIX(t) SHA_ROL(W(t+13) ^ W(t+8) ^ W(t+2) ^ W(t), 1)
100
101
#define SHA_ROUND(t, input, fn, constant, A, B, C, D, E) do { \
@@ -245,5 +273,5 @@ void blk_SHA1_Final(unsigned char hashout[20], blk_SHA_CTX *ctx)
273
274
/* Output hash */
275
for (i = 0; i < 5; i++)
248
- ((unsigned int *)hashout)[i] = htonl(ctx->H[i]);
276
+ put_be32(hashout + i*4, ctx->H[i]);
277
}