diff options
| author | lloyd <[email protected]> | 2006-08-12 18:34:52 +0000 |
|---|---|---|
| committer | lloyd <[email protected]> | 2006-08-12 18:34:52 +0000 |
| commit | bae2a1447f6620d44a977e38587779b8e47be665 (patch) | |
| tree | bb72acd9ebdd7b0411a7332093fee783d916ce0e /modules/sha_x86/sha1core.S | |
| parent | b9f901807d9c6e2bd1e1d52d64fe327c9478acb9 (diff) | |
Entire SHA-1 compression function now implemented in assembly, just needs
instruction scheduling and code cleanup.
Diffstat (limited to 'modules/sha_x86/sha1core.S')
| -rw-r--r-- | modules/sha_x86/sha1core.S | 56 |
1 files changed, 44 insertions, 12 deletions
diff --git a/modules/sha_x86/sha1core.S b/modules/sha_x86/sha1core.S index 17352688d..acba9fb8d 100644 --- a/modules/sha_x86/sha1core.S +++ b/modules/sha_x86/sha1core.S @@ -54,7 +54,9 @@ sha160_core: #define MAGIC1 $0x5A827999 #define MAGIC2 $0x6ED9EBA1 - +#define MAGIC3 $0x8F1BBCDC +#define MAGIC4 $0xCA62C1D6 + #define FUNC1(B, C, D, TEMP) \ movl C, TEMP ; \ xorl D, TEMP ; \ @@ -66,16 +68,29 @@ sha160_core: xorl C, TEMP ; \ xorl D, TEMP +#define FUNC3(B, C, D, TEMP) \ + movl B, TEMP ; \ + orl C, TEMP ; \ + andl D, TEMP ; \ + movl B, (%edi) ; \ + andl C, (%edi) ; \ + orl (%edi), TEMP + +#define FUNC4(B, C, D, TEMP) \ + movl B, TEMP ; \ + xorl C, TEMP ; \ + xorl D, TEMP + #define F(A, B, C, D, E, TEMP, MAGIC, FUNC) \ + addl (%edi), E ; \ FUNC(B, C, D, TEMP) ; \ - addl 0(%edi), E ; \ addl $4, %edi ; \ addl TEMP, E ; \ addl MAGIC, E ; \ roll $5, A ; \ addl A, E ; \ rorl $5, A ; \ - roll $30, B + roll $30, B ; #define F1(A, B, C, D, E, TEMP) \ F(A, B, C, D, E, TEMP, MAGIC1, FUNC1) @@ -83,11 +98,17 @@ sha160_core: #define F2(A, B, C, D, E, TEMP) \ F(A, B, C, D, E, TEMP, MAGIC2, FUNC2) +#define F3(A, B, C, D, E, TEMP) \ + F(A, B, C, D, E, TEMP, MAGIC3, FUNC3) + +#define F4(A, B, C, D, E, TEMP) \ + F(A, B, C, D, E, TEMP, MAGIC4, FUNC4) + #define F_BLOCK(F) \ - F(%eax, %ebx, %ecx, %edx, %esi, %ebp) ; \ - F(%esi, %eax, %ebx, %ecx, %edx, %ebp) ; \ - F(%edx, %esi, %eax, %ebx, %ecx, %ebp) ; \ - F(%ecx, %edx, %esi, %eax, %ebx, %ebp) ; \ + F(%eax, %ebx, %ecx, %edx, %esi, %ebp) \ + F(%esi, %eax, %ebx, %ecx, %edx, %ebp) \ + F(%edx, %esi, %eax, %ebx, %ecx, %ebp) \ + F(%ecx, %edx, %esi, %eax, %ebx, %ebp) \ F(%ebx, %ecx, %edx, %esi, %eax, %ebp) /* F1(%eax, %ebx, %ecx, %edx, %esi, %ebp) @@ -96,6 +117,7 @@ sha160_core: F1(%ecx, %edx, %esi, %eax, %ebx, %ebp) F1(%ebx, %ecx, %edx, %esi, %eax, %ebp) */ + F_BLOCK(F1) F_BLOCK(F1) F_BLOCK(F1) @@ -106,12 +128,22 @@ sha160_core: F_BLOCK(F2) F_BLOCK(F2) + F_BLOCK(F3) + F_BLOCK(F3) + F_BLOCK(F3) + F_BLOCK(F3) + + F_BLOCK(F4) + F_BLOCK(F4) + F_BLOCK(F4) + F_BLOCK(F4) + movl 20(%esp), %ebp - movl %eax, 0(%ebp) - movl %ebx, 4(%ebp) - movl %ecx, 8(%ebp) - movl %edx, 12(%ebp) - movl %esi, 16(%ebp) + addl %eax, 0(%ebp) + addl %ebx, 4(%ebp) + addl %ecx, 8(%ebp) + addl %edx, 12(%ebp) + addl %esi, 16(%ebp) popl %ebx popl %esi |