some minor optimizations

diff --git a/camellia-asm.S b/camellia-asm.S
index f7688931efde6018efd3b79b651af015c0c08c3a..c743997518247b41bfeca78081016311b80baff7 100644 (file)
--- a/camellia-asm.S
+++ b/camellia-asm.S
@@ -209,36 +209,36 @@ camellia_s4:
 camellia_s:
        movw r26, r24   ; backup r24,r25 -> X
        clr r25
-       call camellia_s2
+       rcall camellia_s2
        mov r26, r24
 
        mov r24, r27
-       call camellia_s1
+       rcall camellia_s1
        mov r27, r24
         
        mov r24, r23
-       call camellia_s3
+       rcall camellia_s3
        mov r23, r24
        
        mov r24, r22
-       call camellia_s4
+       rcall camellia_s4
        mov r22, r24
 
        mov r24, r21
-       call camellia_s2
+       rcall camellia_s2
        mov r21, r24
        
        mov r24, r20
-       call camellia_s3
+       rcall camellia_s3
        mov r20, r24
        
        mov r24, r19
-       call camellia_s4
+       rcall camellia_s4
        mov r19, r24
 
 
        mov r24, r18
-       call camellia_s1
+       rcall camellia_s1
        mov r18, r24
                
        movw r24, r26
@@ -300,8 +300,8 @@ camellia_f:
        eor r23, r15
        eor r24, r16
        eor r25, r17
-       call camellia_s
-       call camellia_p
+       rcall camellia_s
+       rcall camellia_p
        ret
        
 ;##############################################################################
@@ -802,15 +802,15 @@ main_loop:
         neg r22
        SBRS xro_sec, 2 // KEY_DIR
         rjmp 2f
-       call camellia128_keyop_inv
+       rcall camellia128_keyop_inv
        rjmp 3f 
-2:     call camellia128_keyop
+2:     rcall camellia128_keyop
 3:     /* loop back */
        SWAP_R br1_sec, bl1_sec
        SWAP_R br2_sec, bl2_sec
        dec loop_cnt
        breq 2f
-       jmp main_loop
+       rjmp main_loop
 2:
        pop r7
        pop r8
@@ -937,18 +937,18 @@ camellia128_init:
 //     / * step 1 * /
        ldi r26, lo8(camellia_sigma)
        ldi r27, hi8(camellia_sigma)
-       call X64_xor_in
-       call camellia_s
-       call camellia_p // / * f(x,k) is done * /
+       rcall X64_xor_in
+       rcall camellia_s
+       rcall camellia_p        // / * f(x,k) is done * /
        sbiw r30, 128/8
        movw r28, r30 ; Z&Y point on kar now
        call Y64_load_xor_store
 
 //     / * step 2 now * /
-       call X64_xor_in
-       call camellia_s
-       call camellia_p // / * f(x,k) is done * /
-       call Y64_load_xor_store
+       rcall X64_xor_in
+       rcall camellia_s
+       rcall camellia_p        // / * f(x,k) is done * /
+       rcall Y64_load_xor_store
        
 //     / * now the xor part (kl and kr) * /
        sbiw r30, 128/8         ; Z points to klr
@@ -961,23 +961,23 @@ camellia128_init:
        brne 1b
        
 //     / * now s->kar ^= camellia_f(s->kal, camellia_sigma[2]); * /    
-       call X64_load           ; load sigma[2]
+       rcall X64_load          ; load sigma[2]
        movw r26, r28           ; X&Y point at kal
-       call X64_xor_in
-       call camellia_s
-       call camellia_p
+       rcall X64_xor_in
+       rcall camellia_s
+       rcall camellia_p
        sbiw r28, 128/8/2       ; Y points at kar
-       call Y64_load_xor_store
+       rcall Y64_load_xor_store
        
 //     / * now s->kal ^= camellia_f(s->kar, camellia_sigma[3]); * /
        sbiw r26, 128/8         ;
-       call X64_load           ; load kar
+       rcall X64_load          ; load kar
        ldi r26, lo8(camellia_sigma+3*8)
        ldi r27, hi8(camellia_sigma+3*8)
-       call X64_xor_in         ; xor sigma[3] in
-       call camellia_s
-       call camellia_p
-       call Y64_load_xor_store
+       rcall X64_xor_in                ; xor sigma[3] in
+       rcall camellia_s
+       rcall camellia_p
+       rcall Y64_load_xor_store
        
        pop r28
        pop r29
@@ -993,3 +993,4 @@ camellia128_init:
 
 
 
+       

diff --git a/sha1-asm.S b/sha1-asm.S
index 577d9e812f9896be6d54fe06083b531b7d3208c2..a7a33f7b512e6397eedc96772968f9c623b041d5 100644 (file)
--- a/sha1-asm.S
+++ b/sha1-asm.S
@@ -732,7 +732,7 @@ sha1_nextBlock_update_a:
        inc LoopC
        cpi LoopC, 80
        brge 1f
-       jmp sha1_nextBlock_mainloop
+       rjmp sha1_nextBlock_mainloop
 /**************************************/
 1:     
    /* littel patch */