(AES_FINAL_ROUND): New macro.

(_aes_crypt): Use AES_FINAL_ROUND for the first word if the final
round.
(_aes_crypt): And for the rest of the final round.

Rev: src/nettle/sparc/aes.asm:1.99

diff --git a/sparc/aes.asm b/sparc/aes.asm
index 63cdb31cfc15f6d02cb283aed5656d595d74fccd..51021a90690d29ef19dc6867143572385f342692 100644 (file)
--- a/sparc/aes.asm
+++ b/sparc/aes.asm
@@ -285,122 +285,125 @@ C       stb     t0, [dst]
 C      add     dst, 4, dst
        
        C i = 1
-       ld      [IDX1+4], t1    ! 1
-
-       ldub    [wtxt+t1], t1   ! 1
-
-       ld      [wtxt+4], t0    ! 0
-       ! IDX2(j) = j XOR 2
-       lduh    [wtxt+12], t2   ! 2
-       and     t0, 255, t0     ! 0
-       ld      [IDX3 + 4], t3  ! 3
-       
-       and     t2, 255, t2     ! 2
-       ldub    [T+t1], t1      ! 1
-       ldub    [T+t0], t0      ! 0
-       sll     t1, 8, t1       ! 1
-       
-       ldub    [wtxt+t3], t3   ! 3
-       or      t0, t1, t0      ! 0, 1
-       ldub    [T+t2], t2      ! 2
-       ldub    [T+t3], t3      ! 3
-       
-       sll     t2, 16, t2      ! 2
-       or      t0, t2, t0      ! 0, 1, 2
-       ld      [key + 4], t2
-       sll     t3, 24, t3      ! 3
-       
-       or      t0, t3, t0      ! 0, 1, 2, 3
-       xor     t0, t2, t0
-       
-       srl     t0, 24, t3
-       srl     t0, 16, t2
-       srl     t0, 8, t1
-       stb     t1, [dst+1]
-       
-       stb     t3, [dst+3]
-       stb     t2, [dst+2]
-       stb     t0, [dst]
-       add     dst, 4, dst
+       AES_FINAL_ROUND(4)      ! i = 1
+C      ld      [IDX1+4], t1    ! 1
+C 
+C      ldub    [wtxt+t1], t1   ! 1
+C 
+C      ld      [wtxt+4], t0    ! 0
+C      ! IDX2(j) = j XOR 2
+C      lduh    [wtxt+12], t2   ! 2
+C      and     t0, 255, t0     ! 0
+C      ld      [IDX3 + 4], t3  ! 3
+C      
+C      and     t2, 255, t2     ! 2
+C      ldub    [T+t1], t1      ! 1
+C      ldub    [T+t0], t0      ! 0
+C      sll     t1, 8, t1       ! 1
+C      
+C      ldub    [wtxt+t3], t3   ! 3
+C      or      t0, t1, t0      ! 0, 1
+C      ldub    [T+t2], t2      ! 2
+C      ldub    [T+t3], t3      ! 3
+C      
+C      sll     t2, 16, t2      ! 2
+C      or      t0, t2, t0      ! 0, 1, 2
+C      ld      [key + 4], t2
+C      sll     t3, 24, t3      ! 3
+C      
+C      or      t0, t3, t0      ! 0, 1, 2, 3
+C      xor     t0, t2, t0
+C      
+C      srl     t0, 24, t3
+C      srl     t0, 16, t2
+C      srl     t0, 8, t1
+C      stb     t1, [dst+1]
+C      
+C      stb     t3, [dst+3]
+C      stb     t2, [dst+2]
+C      stb     t0, [dst]
+C      add     dst, 4, dst
        
        C i = 2
-       ld      [IDX1+8], t1    ! 1
-       
-       ldub    [wtxt+t1], t1   ! 1
-
-       ld      [wtxt+8], t0    ! 0
-       ! IDX2(j) = j XOR 2
-       lduh    [wtxt+0], t2    ! 2
-       and     t0, 255, t0     ! 0
-       ld      [IDX3 + 8], t3  ! 3
-       
-       and     t2, 255, t2     ! 2
-       ldub    [T+t1], t1      ! 1
-       ldub    [T+t0], t0      ! 0
-       sll     t1, 8, t1       ! 1
-       
-       ldub    [wtxt+t3], t3   ! 3
-       or      t0, t1, t0      ! 0, 1
-       ldub    [T+t2], t2      ! 2
-       ldub    [T+t3], t3      ! 3
-       
-       sll     t2, 16, t2      ! 2
-       or      t0, t2, t0      ! 0, 1, 2
-       ld      [key + 8], t2
-       sll     t3, 24, t3      ! 3
-       
-       or      t0, t3, t0      ! 0, 1, 2, 3
-       xor     t0, t2, t0
-       
-       srl     t0, 24, t3
-       srl     t0, 16, t2
-       srl     t0, 8, t1
-       stb     t1, [dst+1]
-       
-       stb     t3, [dst+3]
-       stb     t2, [dst+2]
-       stb     t0, [dst]
-       add     dst, 4, dst
+       AES_FINAL_ROUND(8)      ! i = 2
+C      ld      [IDX1+8], t1    ! 1
+C      
+C      ldub    [wtxt+t1], t1   ! 1
+C 
+C      ld      [wtxt+8], t0    ! 0
+C      ! IDX2(j) = j XOR 2
+C      lduh    [wtxt+0], t2    ! 2
+C      and     t0, 255, t0     ! 0
+C      ld      [IDX3 + 8], t3  ! 3
+C      
+C      and     t2, 255, t2     ! 2
+C      ldub    [T+t1], t1      ! 1
+C      ldub    [T+t0], t0      ! 0
+C      sll     t1, 8, t1       ! 1
+C      
+C      ldub    [wtxt+t3], t3   ! 3
+C      or      t0, t1, t0      ! 0, 1
+C      ldub    [T+t2], t2      ! 2
+C      ldub    [T+t3], t3      ! 3
+C      
+C      sll     t2, 16, t2      ! 2
+C      or      t0, t2, t0      ! 0, 1, 2
+C      ld      [key + 8], t2
+C      sll     t3, 24, t3      ! 3
+C      
+C      or      t0, t3, t0      ! 0, 1, 2, 3
+C      xor     t0, t2, t0
+C      
+C      srl     t0, 24, t3
+C      srl     t0, 16, t2
+C      srl     t0, 8, t1
+C      stb     t1, [dst+1]
+C      
+C      stb     t3, [dst+3]
+C      stb     t2, [dst+2]
+C      stb     t0, [dst]
+C      add     dst, 4, dst
        
        C i = 3
-       ld      [IDX1+12], t1   ! 1
-
-       ldub    [wtxt+t1], t1   ! 1
+       AES_FINAL_ROUND(12)     ! i = 3
+C      ld      [IDX1+12], t1   ! 1
+C 
+C      ldub    [wtxt+t1], t1   ! 1
+C 
+C      ld      [wtxt+12], t0   ! 0
+C      ! IDX2(j) = j XOR 2
+C      lduh    [wtxt+4], t2    ! 2
+C      and     t0, 255, t0     ! 0
+C      ld      [IDX3 + 12], t3 ! 3
+C      
+C      and     t2, 255, t2     ! 2
+C      ldub    [T+t1], t1      ! 1
+C      ldub    [T+t0], t0      ! 0
+C      sll     t1, 8, t1       ! 1
+C      
+C      ldub    [wtxt+t3], t3   ! 3
+C      or      t0, t1, t0      ! 0, 1
+C      ldub    [T+t2], t2      ! 2
+C      ldub    [T+t3], t3      ! 3
+C      
+C      sll     t2, 16, t2      ! 2
+C      or      t0, t2, t0      ! 0, 1, 2
+C      ld      [key + 12], t2
+C      sll     t3, 24, t3      ! 3
+C      
+C      or      t0, t3, t0      ! 0, 1, 2, 3
+C      xor     t0, t2, t0
+C      
+C      srl     t0, 24, t3
+C      srl     t0, 16, t2
+C      srl     t0, 8, t1
+C      stb     t1, [dst+1]
+C      
+C      stb     t3, [dst+3]
+C      stb     t2, [dst+2]
+C      stb     t0, [dst]
+C      add     dst, 4, dst
 
-       ld      [wtxt+12], t0   ! 0
-       ! IDX2(j) = j XOR 2
-       lduh    [wtxt+4], t2    ! 2
-       and     t0, 255, t0     ! 0
-       ld      [IDX3 + 12], t3 ! 3
-       
-       and     t2, 255, t2     ! 2
-       ldub    [T+t1], t1      ! 1
-       ldub    [T+t0], t0      ! 0
-       sll     t1, 8, t1       ! 1
-       
-       ldub    [wtxt+t3], t3   ! 3
-       or      t0, t1, t0      ! 0, 1
-       ldub    [T+t2], t2      ! 2
-       ldub    [T+t3], t3      ! 3
-       
-       sll     t2, 16, t2      ! 2
-       or      t0, t2, t0      ! 0, 1, 2
-       ld      [key + 12], t2
-       sll     t3, 24, t3      ! 3
-       
-       or      t0, t3, t0      ! 0, 1, 2, 3
-       xor     t0, t2, t0
-       
-       srl     t0, 24, t3
-       srl     t0, 16, t2
-       srl     t0, 8, t1
-       stb     t1, [dst+1]
-       
-       stb     t3, [dst+3]
-       stb     t2, [dst+2]
-       stb     t0, [dst]
-       add     dst, 4, dst
-       
        C Unrolled final loop ends
        
        addcc   length, -16, length