bmw tiny is now 1910 bytes in size

diff --git a/bmw/bmw_small-tinyasm.S b/bmw/bmw_small-tinyasm.S
index 198aa7024688413f0bcb985079237e80bb3532b1..ddad3c9f68fdf9cadf96a36a08a8f9ce0e40b3ef 100644 (file)
--- a/bmw/bmw_small-tinyasm.S
+++ b/bmw/bmw_small-tinyasm.S
@@ -154,10 +154,9 @@ sn:
        movw r24, r4
        lpm r20, Z+
        rcall rotateleft32
-       eor r22, r16
-       eor r23, r17
-       eor r24, r18
-       eor r25, r19
+       rcall eor_r22_in_r16
+       movw r22, r16
+       movw r24, r18
        pop r19
        pop r17
        pop_range 2, 5
@@ -172,6 +171,7 @@ sn:
 memxor_short:
 ;      tst r20
 ;      breq memxor_exit
+       ldi r20, 64
 10: ld r21, X
        ld r22, Z+
        eor r21, r22
@@ -193,23 +193,14 @@ add_hx_to_w:
        movw r26, h0
        add r26, r16
        adc r27, r1
-       ld r22, Y
-       ldd r23, Y+1
-       ldd r24, Y+2
-       ldd r25, Y+3
+       rcall load32_from_Y
+       sbiw r28, 4
        lsl r20
        rol r21
        brcs 30f
        /* addition */
-       ld r0, X+
-       add r22, r0
-       ld r0, X+
-       adc r23, r0
-       ld r0, X+
-       adc r24, r0
-       ld r0, X+
-       adc r25, r0
-       rjmp 50f
+       rcall add_X_to_32
+       rjmp store32_to_Y;50f
 30: /* substract */
        ld r0, X+
        sub r22, r0
@@ -220,11 +211,9 @@ add_hx_to_w:
        ld r0, X+
        sbc r25, r0
 50:
-       st Y+, r22
-       st Y+, r23
-       st Y+, r24
-       st Y+, r25
-       ret
+       rjmp store32_to_Y
+;      rcall store32_to_Y
+;      ret
 
 /******************************************************************************/
 load32_from_X:
@@ -241,6 +230,13 @@ load32_from_Y:
        ld r25, Y+
        ret
 
+store32_to_Y:
+       st Y+, r22
+       st Y+, r23
+       st Y+, r24
+       st Y+, r25
+       ret
+
 add_X_to_32:
        ld r0, X+
        add r22, r0
@@ -292,7 +288,7 @@ f0:
 ;      pop_range  22, 25
 ;--- END DBG
        /* xor m into h */
-       ldi r20, 64
+;      ldi r20, 64
        rcall memxor_short
        movw r30, m0
        movw r26, h0
@@ -340,7 +336,7 @@ f0:
 ;      pop_range  22, 25
 ;--- END DBG
        /* xor m into h */
-       ldi r20, 64
+;      ldi r20, 64
        movw r26, h0
        movw r30, m0
        rcall memxor_short
@@ -351,38 +347,30 @@ f0:
        ldi r21, 15
        mov r8, r21
 50:
-       ldd r22, Y+0
-       ldd r23, Y+1
-       ldd r24, Y+2
-       ldd r25, Y+3
+       rcall load32_from_Y
+       sbiw r28, 4
        lpm r20, Z+
        movw r2, r30
        rcall sn
        movw r30, r2
 
        rcall add_X_to_32
+       rcall store32_to_Y
 
-       st Y+, r22
-       st Y+, r23
-       st Y+, r24
-       st Y+, r25
        dec r8
        brne 50b
 ;---
-       ldd r22, Y+0
-       ldd r23, Y+1
-       ldd r24, Y+2
-       ldd r25, Y+3
+       rcall load32_from_Y
        clr r20
        rcall sn
        movw r30, r2
        movw r26, h0
        rcall add_X_to_32
        sbiw r26, 4
-       std Y+0, r22
-       std Y+1, r23
-       std Y+2, r24
-       std Y+3, r25
+       st -Y, r25
+       st -Y, r24
+       st -Y, r23
+       st -Y, r22
        sbiw r28, 15*4
        movw r20, h0
        movw r22, m0
@@ -464,10 +452,7 @@ load_rotate_add_M:
        movw r26, m0
        add r26, r0
        adc r27, r1
-       ld r22, X+
-       ld r23, X+
-       ld r24, X+
-       ld r25, X+
+       rcall load32_from_X
        inc r20
        rcall rotateleft32
        brts 10f
@@ -670,35 +655,23 @@ f2:
        adiw r26, 63
        adiw r26,  1
        movw q16_0, r26
-       clr xl0
-       clr xl1
-       clr xl2
-       clr xl3
-       ldi r17, 8
-10:    ld r0, X+
-       eor xl0, r0
-       ld r0, X+
-       eor xl1, r0
-       ld r0, X+
-       eor xl2, r0
-       ld r0, X+
-       eor xl3, r0
-       dec r17
-       brne 10b
-;--- /* calc XH */
-       movw xh0, xl0
-       movw xh2, xl2
-       ldi r17, 8
-10:    ld r0, X+
-       eor xh0, r0
-       ld r0, X+
-       eor xh1, r0
-       ld r0, X+
-       eor xh2, r0
-       ld r0, X+
-       eor xh3, r0
+       movw h0, r20
+       movw r28, r22
+       rcall load32_from_X
+       movw acc0, r22
+       movw acc2, r24
+       ldi r17, 15
+10:    rcall load32_from_X
+       rcall eor32_to_acc
+       cpi r17, 9
+       brne 15f
+       movw xl0, acc0
+       movw xl2, acc2
+15:
        dec r17
        brne 10b
+       movw xh0, acc0
+       movw xh2, acc2
 ;--- DBG
 ;      push_range 22, 25
 ;      movw r22, xl0
@@ -711,8 +684,6 @@ f2:
 ;--- END DBG
 
 ;--- /* calc first half of h0..h15 */
-       movw h0, r20
-       movw r28, r22
        movw r26, q16_0
        ldi r17, 16
 10:
@@ -912,11 +883,6 @@ bmw256_nextBlock:
        ret
 
 /******************************************************************************/
-/*
-  param ctx:  r24:r25
-  param msg:  r22:r23
-  param len:  r20:r21
-*/
 ctx0 =  2
 ctx1 =  3
 blc0 =  4
@@ -926,6 +892,22 @@ len1 = 29
 buf0 =  6
 buf1 =  7
 
+load32_from_Z_stub:
+       movw r30, ctx0
+       adiw r30, 60
+       ldd r21, Z+4
+       ldd r22, Z+5
+       ldd r23, Z+6
+       ldd r24, Z+7
+       ret
+
+/******************************************************************************/
+/*
+  param ctx:  r24:r25
+  param msg:  r22:r23
+  param len:  r20:r21
+*/
+
 .global bmw_small_lastBlock
 .global bmw224_lastBlock
 .global bmw256_lastBlock
@@ -1020,12 +1002,7 @@ bmw256_lastBlock:
        st X+, r1
        dec r20
        brne 350b
-       movw r30, ctx0
-       adiw r30, 60
-       ldd r21, Z+4
-       ldd r22, Z+5
-       ldd r23, Z+6
-       ldd r24, Z+7
+       rcall load32_from_Z_stub
        subi r21, 1
        sbc r22, r1
        sbc r23, r1
@@ -1035,12 +1012,7 @@ bmw256_lastBlock:
        bmw_small_nextBlock(ctx, pctx.buffer);
 */
 400:
-       movw r30, ctx0
-       adiw r30, 60
-       ldd r21, Z+4
-       ldd r22, Z+5
-       ldd r23, Z+6
-       ldd r24, Z+7
+       rcall load32_from_Z_stub
 410:
        clr r25
        lsl r21
@@ -1195,19 +1167,20 @@ ctx0 =   2
 ctx1 =   3
 msg0 =   4
 msg1 =   5
-len0 =   6
-len1 =   7
+len0 =  28
+len1 =  29
 len2 =   8
 len3 =   9
-dst0 =  10
-dst1 =  11
+dst0 =   6
+dst1 =   7
 .global bmw224
 bmw224:
        push r16
        clr r16
 
 bmw_small_all:
-       push_range 2, 11
+       push_range 2, 9
+       push_range 28, 29
        stack_alloc_large 64+4
        adiw r30, 1
        movw ctx0, r30
@@ -1228,8 +1201,7 @@ bmw_small_all:
        movw r24, ctx0
        movw r22, msg0
        rcall bmw_small_nextBlock
-       ldi r20, 2
-       sub len1, r20
+       subi len1, 2
        sbc len2, r1
        sbc len3, r1
        ldi r20, 64
@@ -1249,7 +1221,8 @@ bmw_small_all:
        adc r31, r1
        icall
        stack_free_large 64+4
-       pop_range 2, 11
+       pop_range 28, 29
+       pop_range 2, 9
        pop r16
        ret