For longs swap alternate register usage so that low word is in main registers

diff --git a/sm3.asm b/sm3.asm
index 8fc3742..e213385 100644 (file)
--- a/sm3.asm
+++ b/sm3.asm
@@ -31,14 +31,21 @@ restarts:
        ret
        .db     0
        ; rst 0x30, pop hl:de'
-       pop     hl
-       exx
-       pop     de
-       exx
-       ex      (sp),hl
-       ret
+       ;pop    hl
+       ;exx
+       ;pop    de
+       ;exx
+       ;ex     (sp),hl
+       ;ret
+       .db     0,0,0,0,0,0
        .db     0,0
-       ; rst 0x38, immediate to hl:de'
+       ; rst 0x38, immediate to de':hl
+       ld      a,(bc)
+       inc     bc
+       ld      l,a
+       ld      a,(bc)
+       inc     bc
+       ld      h,a
        ld      a,(bc)
        inc     bc
        exx
@@ -49,12 +56,6 @@ restarts:
        exx
        ld      d,a
        exx
-       ld      a,(bc)
-       inc     bc
-       ld      l,a
-       ld      a,(bc)
-       inc     bc
-       ld      h,a
        ret
 restarts_end:
 
@@ -175,10 +176,10 @@ page0_imm_w:
 ; page 0 to 2
 
 page0_page2:
+       pop     de
        exx
        pop     hl
        exx
-       pop     de
        jr      page0_dispatch2
 
 page0_stkld_l:
@@ -451,10 +452,10 @@ page1_imm_slrev_l:
        jr      page1_dispatch2
 page1_sl_l:
        ;rst    0x30
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
        call    math_sl_l0
        jr      page1_dispatch2
 
@@ -465,10 +466,10 @@ page1_imm_srrev_sl:
        jr      page1_dispatch2
 page1_sr_sl:
        ;rst    0x30
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
        call    math_sr_sl0
        jr      page1_dispatch2
 
@@ -479,10 +480,10 @@ page1_imm_srrev_ul:
        jr      page1_dispatch2
 page1_sr_ul:
        ;rst    0x30
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
        call    math_sr_ul0
 page1_dispatch2:
        ld      a,(bc)
@@ -532,10 +533,10 @@ page2_trace:
 ; page 2 to 0
 
 page2_page0:
-       push    de
        exx
        push    hl
        exx
+       push    de
        jr      page2_dispatch0
 
 page2_stkst_l:
@@ -630,15 +631,15 @@ page2_imm_add_l:
        jr      add_l_entry
 page2_add_l:
        ;rst    0x30
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
-add_l_entry:
-       exx
+add_l_entry: ; optimize this
        add     hl,de
        exx
        adc     hl,de
+       exx
        jr      add_l_done
 
 page2_imm_subrev_l:
@@ -647,31 +648,31 @@ page2_imm_subrev_l:
        jr      sub_l_entry
 page2_sub_l:
        ;rst    0x30
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
-sub_l_entry:
-       exx
-       ex      de,hl
+sub_l_entry: ; optimize this
        or      a
        sbc     hl,de
        exx
+       ex      de,hl
        sbc     hl,de
+       exx
        jr      add_l_done
 
 ; use addition for page2_imm_sub_l
 page2_subrev_l:
        ;rst    0x30
-       exx
-       pop     de
-       exx
        pop     hl
-       exx
+       ;exx
+       ;pop    de
+       ;exx
+       ex      de,hl
        or      a
        sbc     hl,de
-       exx
-       ex      de,hl
+       exx 
+ pop de
        sbc     hl,de
        jr      add_l_done
 
@@ -708,10 +709,10 @@ page2_imm_mul_ul:
 
 page2_mul_l:
        ;rst    0x30
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
        ex      de,hl
        call    math_mul_l
        jr      mul_l_done
@@ -723,10 +724,10 @@ page2_imm_divrev_sl:
        jr      div_l_done
 page2_div_sl:
        ;rst    0x30
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
        call    math_div_sl0
        jr      div_l_done
 
@@ -738,10 +739,10 @@ page2_imm_div_sl:
        jr      div_l_done
 page2_divrev_sl:
        ;rst    0x30
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
        ex      de,hl
        call    math_div_sl
        jr      div_l_done
@@ -753,10 +754,10 @@ page2_imm_divrev_ul:
        jr      div_l_done
 page2_div_ul:
        ;rst    0x30
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
        call    math_div_ul0
        jr      div_l_done
 
@@ -768,17 +769,17 @@ page2_imm_div_ul:
        jr      div_l_done
 page2_divrev_ul:
        ;rst    0x30
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
        ex      de,hl
        call    math_div_ul
 div_l_done:
-       push    de
        exx
        push    de
        exx
+       push    de
 mul_l_done:
        ex      de,hl
 ;page2_dispatch2h:
@@ -831,7 +832,13 @@ math_imm_w: ; immediate to de
        ld      h,page1
        jp      (hl)
 
-math_imm_l: ; immediate to de:hl'
+math_imm_l: ; immediate to hl':de
+       ld      a,(bc)
+       inc     bc
+       ld      e,a
+       ld      a,(bc)
+       inc     bc
+       ld      d,a
        ld      a,(bc)
        inc     bc
        exx
@@ -842,13 +849,6 @@ math_imm_l: ; immediate to de:hl'
        exx
        ld      h,a
        exx
-;math_imm_w: ; immediate to de
-       ld      a,(bc)
-       inc     bc
-       ld      e,a
-       ld      a,(bc)
-       inc     bc
-       ld      d,a
        ;ret
        ld      a,(bc)
        inc     bc
@@ -883,20 +883,20 @@ math_stkld_l: ; sp(imm_w) to de:hl'
        inc     bc
        ld      h,a
        add     hl,sp
-math_ld_l: ; (hl) to de:hl'
+math_ld_l: ; (hl) to hl':de
+       ld      e,(hl)
+       inc     hl
+       ld      d,(hl)
+       inc     hl
        ld      a,(hl)
        inc     hl
        exx
        ld      l,a
        exx
        ld      a,(hl)
-       inc     hl
        exx
        ld      h,a
        exx
-       ld      e,(hl)
-       inc     hl
-       ld      d,(hl)
        ;ret
        ld      a,(bc)
        inc     bc
@@ -930,7 +930,7 @@ math_stkst_l: ; de:hl' to sp(imm_w)
        inc     bc
        ld      h,a
        add     hl,sp
-math_st_l: ; de:hl' to (hl)
+math_st_l: ; hl':de to (hl)
 .if 0
  call print_word
  ld a,':
@@ -948,6 +948,10 @@ math_st_l: ; de:hl' to (hl)
  ld a,' 
  call print_char
 .endif
+       ld      (hl),e
+       inc     hl
+       ld      (hl),d
+       inc     hl
        exx
        ld      a,l
        exx
@@ -957,10 +961,6 @@ math_st_l: ; de:hl' to (hl)
        ld      a,h
        exx
        ld      (hl),a
-       inc     hl
-       ld      (hl),e
-       inc     hl
-       ld      (hl),d
        ;ret
        ld      a,(bc)
        inc     bc
@@ -968,20 +968,23 @@ math_st_l: ; de:hl' to (hl)
        ld      h,page0
        jp      (hl)
 
-math_and_imm_l: ; de:hl' &= imm_l
+math_and_imm_w: ; de &= imm_w
        ld      a,(bc)
        inc     bc
-       exx
-       and     l
-       ld      l,a
-       exx
+       and     e
+       ld      e,a
        ld      a,(bc)
        inc     bc
-       exx
-       and     h
-       ld      h,a
-       exx
-math_and_imm_w: ; de &= imm_w
+       and     d
+       ld      d,a
+       ;ret
+       ld      a,(bc)
+       inc     bc
+       ld      l,a
+       ;ld     h,page1
+       jp      (hl)
+
+math_and_imm_l: ; hl':de &= imm_l
        ld      a,(bc)
        inc     bc
        and     e
@@ -990,11 +993,23 @@ math_and_imm_w: ; de &= imm_w
        inc     bc
        and     d
        ld      d,a
+       ld      a,(bc)
+       inc     bc
+       exx
+       and     l
+       ld      l,a
+       exx
+       ld      a,(bc)
+       inc     bc
+       exx
+       and     h
+       ld      h,a
+       exx
        ;ret
        ld      a,(bc)
        inc     bc
        ld      l,a
-       ;ld     h,page1
+       ;ld     h,page2
        jp      (hl)
 
 math_and_w: ; de &= hl
@@ -1012,11 +1027,17 @@ math_and_w: ; de &= hl
        jp      (hl)
 
 math_and_l0:
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
-math_and_l: ; de:hl' &= hl:de'
+math_and_l: ; hl':de &= de':hl
+       ld      a,e
+       and     l
+       ld      e,a
+       ld      a,d
+       and     h
+       ld      d,a
        exx
        ld      a,l
        and     e
@@ -1025,13 +1046,6 @@ math_and_l: ; de:hl' &= hl:de'
        and     d
        ld      h,a
        exx
-;math_and_imm_w: ; de &= imm_w
-       ld      a,e
-       and     l
-       ld      e,a
-       ld      a,d
-       and     h
-       ld      d,a
        ;ret
        ld      a,(bc)
        inc     bc
@@ -1052,10 +1066,18 @@ math_or_imm_w: ; de |= imm_w
        ld      a,(bc)
        inc     bc
        ld      l,a
-       ld      h,page1
+       ;ld     h,page1
        jp      (hl)
 
-math_or_imm_l: ; de:hl' |= imm_l
+math_or_imm_l: ; hl':de |= imm_l
+       ld      a,(bc)
+       inc     bc
+       or      e
+       ld      e,a
+       ld      a,(bc)
+       inc     bc
+       or      d
+       ld      d,a
        ld      a,(bc)
        inc     bc
        exx
@@ -1068,15 +1090,6 @@ math_or_imm_l: ; de:hl' |= imm_l
        or      h
        ld      h,a
        exx
-;math_or_imm_w: ; de |= imm_w
-       ld      a,(bc)
-       inc     bc
-       or      e
-       ld      e,a
-       ld      a,(bc)
-       inc     bc
-       or      d
-       ld      d,a
        ;ret
        ld      a,(bc)
        inc     bc
@@ -1099,11 +1112,17 @@ math_or_w: ; de |= hl
        jp      (hl)
 
 math_or_l0:
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
-math_or_l: ; de:hl' |= hl:de'
+math_or_l: ; hl':de |= de':hl
+       ld      a,e
+       or      l
+       ld      e,a
+       ld      a,d
+       or      h
+       ld      d,a
        exx
        ld      a,l
        or      e
@@ -1112,13 +1131,6 @@ math_or_l: ; de:hl' |= hl:de'
        or      d
        ld      h,a
        exx
-;math_or_w: ; de |= hl
-       ld      a,e
-       or      l
-       ld      e,a
-       ld      a,d
-       or      h
-       ld      d,a
        ;ret
        ld      a,(bc)
        inc     bc
@@ -1142,7 +1154,15 @@ math_xor_imm_w: ; de ^= imm_w
        ;ld     h,page1
        jp      (hl)
 
-math_xor_imm_l: ; de:hl' ^= imm_l
+math_xor_imm_l: ; hl':de ^= imm_l
+       ld      a,(bc)
+       inc     bc
+       xor     e
+       ld      e,a
+       ld      a,(bc)
+       inc     bc
+       xor     d
+       ld      d,a
        ld      a,(bc)
        inc     bc
        exx
@@ -1155,15 +1175,6 @@ math_xor_imm_l: ; de:hl' ^= imm_l
        xor     h
        ld      h,a
        exx
-;math_xor_imm_w: ; de ^= imm_w
-       ld      a,(bc)
-       inc     bc
-       xor     e
-       ld      e,a
-       ld      a,(bc)
-       inc     bc
-       xor     d
-       ld      d,a
        ;ret
        ld      a,(bc)
        inc     bc
@@ -1186,11 +1197,17 @@ math_xor_w: ; de ^= hl
        jp      (hl)
 
 math_xor_l0:
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
-math_xor_l: ; de:hl' ^= hl:de'
+math_xor_l: ; hl':de ^= de':hl
+       ld      a,e
+       xor     l
+       ld      e,a
+       ld      a,d
+       xor     h
+       ld      d,a
        exx
        ld      a,l
        xor     e
@@ -1199,13 +1216,6 @@ math_xor_l: ; de:hl' ^= hl:de'
        xor     d
        ld      h,a
        exx
-;math_xor_w: ; de ^= hl
-       ld      a,e
-       xor     l
-       ld      e,a
-       ld      a,d
-       xor     h
-       ld      d,a
        ;ret
        ld      a,(bc)
        inc     bc
@@ -1233,16 +1243,17 @@ math_cmprev_sw: ; cf=1 hl < de, zf=1 hl == de, signed
 ; use inline code for math_cmp_uw, math_cmprev_uw
 
 math_cmp_sl0:
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
-math_cmp_sl: ; cf=1 de:hl' < hl:de', zf=1 de:hl' == hl:de', signed
-       ex      de,hl
+math_cmp_sl: ; cf=1 hl':de < de':hl, zf=1 hl':de == de':hl, signed
+       exx
        or      a
        sbc     hl,de
-       jr      z,cmp_l_entry
        ld      a,h
+       exx ; optimize this
+       jr      z,cmp_l_entry
        rla
        ;ret    po
        jp      po,1$
@@ -1255,20 +1266,20 @@ math_cmp_sl: ; cf=1 de:hl' < hl:de', zf=1 de:hl' == hl:de', signed
        jp      (hl)
 
 math_cmp_ul0:
+       pop     hl
        exx
        pop     de
+       exx ; optimize this
+math_cmp_ul: ; cf=1 hl':de < de':hl, zf=1 hl':de == de':hl, unsigned
        exx
-       pop     hl
-math_cmp_ul: ; cf=1 de:hl' < hl:de', zf=1 de:hl' == hl:de', unsigned
-       ex      de,hl
        or      a
        sbc     hl,de
        ;ret    nz
+       exx
        jr      nz,cmp_l_dispatch
 cmp_l_entry:
-       exx
+       ex      de,hl
        sbc     hl,de
-       exx
        ;ret
 cmp_l_dispatch:
        ld      a,(bc)
@@ -1278,15 +1289,18 @@ cmp_l_dispatch:
        jp      (hl)
 
 math_cmprev_sl0:
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
-math_cmprev_sl: ; cf=1 hl:de' < de:hl', zf=1 hl:de' == de:hl', signed
+math_cmprev_sl: ; cf=1 de':hl < hl':de, zf=1 de':hl == hl':de, signed
+       exx
+       ex      de,hl
        or      a
        sbc     hl,de
-       jr      z,cmprev_l_entry
        ld      a,h
+       exx ; optimize this
+       jr      z,cmprev_l_entry
        rla
        ;ret    po
        jp      po,1$
@@ -1299,20 +1313,20 @@ math_cmprev_sl: ; cf=1 hl:de' < de:hl', zf=1 hl:de' == de:hl', signed
        jp      (hl)
 
 math_cmprev_ul0:
+       pop     hl
        exx
        pop     de
        exx
-       pop     hl
-math_cmprev_ul: ; cf=1 de:hl' < hl:de', zf=1 de:hl' == hl:de', unsigned
+math_cmprev_ul: ; cf=1 de':hl < hl':de, zf=1 de':hl == hl':de, signed
+       exx
+       ex      de,hl
        or      a
        sbc     hl,de
+       exx
        ;ret    nz
        jr      nz,cmprev_l_dispatch
 cmprev_l_entry:
-       exx
-       ex      de,hl
        sbc     hl,de
-       exx
        ;ret
 cmprev_l_dispatch:
        ld      a,(bc)
@@ -1373,22 +1387,21 @@ math_sl_w: ; hl <<= e & 0xf
        ld      l,0
        ret
 
-math_sl_l0: ; hl:de' <<= e & 0x1f
+math_sl_l0: ; de':hl <<= e & 0x1f
        exx
        ex      de,hl
        exx
-math_sl_l: ; hl:hl' <<= e & 0x1f
+math_sl_l: ; hl':hl <<= e & 0x1f
        ; by 1
        bit     0,e
        jr      z,1$
-       exx
        add     hl,hl
        exx
        adc     hl,hl
+       exx
 1$:    ; by 2
        bit     1,e
        jr      z,2$
-       exx
        add     hl,hl
        exx
        adc     hl,hl
@@ -1396,10 +1409,10 @@ math_sl_l: ; hl:hl' <<= e & 0x1f
        add     hl,hl
        exx
        adc     hl,hl
+       exx
 2$:    ; by 4
        bit     2,e
        jr      z,3$
-       exx
        add     hl,hl
        exx
        adc     hl,hl
@@ -1415,24 +1428,25 @@ math_sl_l: ; hl:hl' <<= e & 0x1f
        add     hl,hl
        exx
        adc     hl,hl
+       exx
 3$:    ; by 8
        bit     3,e
        jr      z,4$
-       exx
        ld      a,h
        ld      h,l
        ld      l,0
        exx
        ld      h,l
        ld      l,a
+       exx
 4$:    ; by 16
        bit     4,e
        ret     z
-       exx
        push    hl
        ld      hl,0
        exx
        pop     hl
+       exx
        ret
 
 math_sr_uw0: ; hl = de >> (l & 0xf), logical
@@ -1509,67 +1523,70 @@ math_sr_ul1: ; hl:hl' >>= a - 7, immediate a in [7, 0x27), logical
        sub     a
        jr      sr_l_entry
 
-math_sr_sl0: ; hl:de' >>= e & 0x1f, arithmetic
+math_sr_sl0: ; de':hl >>= e & 0x1f, arithmetic
        exx
        ex      de,hl
        exx
-math_sr_sl: ; hl:hl' >>= e & 0x1f, arithmetic
+math_sr_sl: ; hl':hl >>= e & 0x1f, arithmetic
        ld      a,e
        and     0xf
        add     7
-math_sr_sl1: ; hl:hl' >>= a - 7, immediate a in [7, 0x27), arithmetic
+math_sr_sl1: ; hl':hl >>= a - 7, immediate a in [7, 0x27), arithmetic
        ld      e,a
+       exx
        ld      a,h
+       exx
        rla
        sbc     a,a
 sr_l_entry:
        ; by -1
        bit     0,e
        jr      nz,1$
-       exx
        add     hl,hl
        exx
        adc     hl,hl
+       exx
        rla
 1$:    ; by -2
        bit     1,e
        jr      nz,2$
-       exx
        add     hl,hl
        exx
        adc     hl,hl
-       rla
        exx
+       rla
        add     hl,hl
        exx
        adc     hl,hl
+       exx
        rla
 2$:    ; by -4
        bit     2,e
        jr      nz,3$
-       exx
        add     hl,hl
        exx
        adc     hl,hl
-       rla
        exx
+       rla
        add     hl,hl
        exx
        adc     hl,hl
-       rla
        exx
+       rla
        add     hl,hl
        exx
        adc     hl,hl
-       rla
        exx
+       rla
        add     hl,hl
        exx
        adc     hl,hl
+       exx
        rla
 3$:    ; by 8
        bit     3,e
        jr      z,4$
+       exx
        ld      d,l
        ld      l,h
        ld      h,a
@@ -1578,11 +1595,11 @@ sr_l_entry:
        exx
        ld      l,h
        ld      h,a
-       exx
        sbc     a,a
 4$:    ; by 16
        bit     4,e
        jr      z,5$
+       exx
        push    hl
        ld      l,a
        rla
@@ -1590,12 +1607,10 @@ sr_l_entry:
        ld      h,a
        exx
        pop     hl
-       exx
        ret
 5$:    ; by 32 (can't occur simultaneously with by 16)
        bit     5,e
        ret     z
-       exx
        ld      l,a
        rla
        sbc     a,a
@@ -1603,6 +1618,7 @@ sr_l_entry:
        exx
        ld      l,a
        ld      h,a
+       exx
        ret
 
 ; this routine is just an optimization, therefore use interpreter registers
@@ -1667,15 +1683,15 @@ mul_w0: rla
        ret
 
 ; this routine is just an optimization, therefore use interpreter registers
-math_mul_imm_l: ; de:hl' *= imm_l, big-endian imm_l
-       exx     
-       ex      de,hl
+math_mul_imm_l: ; hl':de *= imm_l, big-endian imm_l
        sub     a
        ld      l,a
        ld      h,a
-       exx
+       exx     
+       ex      de,hl
        ld      l,a
        ld      h,a
+       exx
        ld      a,(bc)
        inc     bc
        call    mul_l0
@@ -1691,8 +1707,7 @@ math_mul_imm_l: ; de:hl' *= imm_l, big-endian imm_l
        ex      de,hl
        ret
 
-math_mul_l: ; hl:hl' *= de:de'
-       exx
+math_mul_l: ; hl':hl *= de':de
        ld      a,l
        push    af
        push    hl
@@ -1702,6 +1717,7 @@ math_mul_l: ; hl:hl' *= de:de'
        push    af
        ld      a,h
        ld      hl,0
+       exx
        call    mul_l0
        pop     af
        call    mul_l
@@ -1709,93 +1725,93 @@ math_mul_l: ; hl:hl' *= de:de'
        call    mul_l
        pop     af
 mul_l: ; bit 0
-       exx
        add     hl,hl
        exx
        adc     hl,hl
+       exx
 mul_l0:        rla
        jr      nc,1$
-       exx
        add     hl,de
        exx
        adc     hl,de
-1$:    ; bit 1
        exx
+1$:    ; bit 1
        add     hl,hl
        exx
        adc     hl,hl
+       exx
        rla
        jr      nc,2$
-       exx
        add     hl,de
        exx
        adc     hl,de
-2$:    ; bit 2
        exx
+2$:    ; bit 2
        add     hl,hl
        exx
        adc     hl,hl
+       exx
        rla
        jr      nc,3$
-       exx
        add     hl,de
        exx
        adc     hl,de
-3$:    ; bit 3
        exx
+3$:    ; bit 3
        add     hl,hl
        exx
        adc     hl,hl
+       exx
        rla
        jr      nc,4$
-       exx
        add     hl,de
        exx
        adc     hl,de
-4$:    ; bit 4
        exx
+4$:    ; bit 4
        add     hl,hl
        exx
        adc     hl,hl
+       exx
        rla
        jr      nc,5$
-       exx
        add     hl,de
        exx
        adc     hl,de
-5$:    ; bit 5
        exx
+5$:    ; bit 5
        add     hl,hl
        exx
        adc     hl,hl
+       exx
        rla
        jr      nc,6$
-       exx
        add     hl,de
        exx
        adc     hl,de
-6$:    ; bit 6
        exx
+6$:    ; bit 6
        add     hl,hl
        exx
        adc     hl,hl
+       exx
        rla
        jr      nc,7$
-       exx
        add     hl,de
        exx
        adc     hl,de
-7$:    ; bit 7
        exx
+7$:    ; bit 7
        add     hl,hl
        exx
        adc     hl,hl
+       exx
        rla
        ret     nc
-       exx
        add     hl,de
        exx
        adc     hl,de
+       exx
        ret
 
 math_div_sw0: ; hl, de = de % hl, de / hl, signed
@@ -2153,20 +2169,21 @@ div_w_n18: ; done, below
        scf
        ret
 
-math_div_sl0: ; hl:hl', de:de' = hl:de' % de:hl', hl:de' / de:hl', signed
+math_div_sl0: ; hl':hl, de':de = de':hl % hl':de, de':hl / hl':de, signed
        exx
        ex      de,hl
        exx
-math_div_sl: ; ; hl:hl', de:de' = hl:hl' % de:de', hl:hl' / de:de', signed
+math_div_sl: ; ; hl':hl, de':de = hl':hl % de':de, hl':hl / de':de, signed
        push    bc
+       exx
        ld      a,h
        or      a
        ld      a,d
        rla
+       exx
        jp      m,div_l_n               ; positive dividend
 
        ; positive dividend
-       exx
        ld      c,l
        ld      b,h
        ld      hl,0
@@ -2174,35 +2191,35 @@ math_div_sl: ; ; hl:hl', de:de' = hl:hl' % de:de', hl:hl' / de:de', signed
        ld      a,h
        ld      c,l
        ld      hl,0
+       exx
        jr      nc,div_l_pp             ; positive dividend, positive divisor
 
        ; positive dividend, negative divisor
        call    div_l_n1
+       exx
        ld      b,a
        ld      a,c
+       exx
        call    div_l_ncf
+       exx
        ld      c,a
        exx
        ld      a,b
-       exx
        call    div_l_ncf
-       exx
        ld      b,a
        ld      a,c
-       exx
        call    div_l_ncf
        inc     a
        jr      c,div_l_p_done
-       exx
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jr      div_l_p_done
 
 div_l_n:
        ; negative dividend
        call    dec_l                   ; reduces remainder by 1 (we inc later)
-       exx
        ld      c,l
        ld      b,h
        ld      hl,-1
@@ -2210,29 +2227,30 @@ div_l_n:
        ld      a,h
        ld      c,l
        ld      hl,-1
+       exx
        jr      c,div_l_nn              ; negative dividend, negative divisor 
 
        ; negative dividend, positive divisor
        call    div_l1
+       exx
        ld      b,a
        ld      a,c
+       exx
        call    div_lcf
+       exx
        ld      c,a
        exx
        ld      a,b
-       exx
        call    div_lcf
-       exx
        ld      b,a
        ld      a,c
-       exx
        call    div_lcf
        inc     a
        jr      c,div_l_n_done
-       exx
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jr      div_l_n_done
 
 math_div_ul0: ; hl:hl', de:de' = hl:de' % de:hl', hl:de' / de:hl', unsigned
@@ -2241,7 +2259,6 @@ math_div_ul0: ; hl:hl', de:de' = hl:de' % de:hl', hl:de' / de:hl', unsigned
        exx
 math_div_ul: ; ; hl:hl', de:de' = hl:hl' % de:de', hl:hl' / de:de', unsigned
        push    bc
-       exx
        ld      c,l
        ld      b,h
        ld      hl,0
@@ -2249,82 +2266,83 @@ math_div_ul: ; ; hl:hl', de:de' = hl:hl' % de:de', hl:hl' / de:de', unsigned
        ld      a,h
        ld      c,l
        ld      hl,0
+       exx
 div_l_pp: ; positive dividend, positive divisor
        call    div_l0
+       exx
        ld      b,a
        ld      a,c
+       exx
        call    div_lcf
+       exx
        ld      c,a
        exx
        ld      a,b
-       exx
        call    div_lcf
-       exx
        ld      b,a
        ld      a,c
-       exx
        call    div_lcf
        jr      nc,div_l_p_done
-       exx
        add     hl,de
        exx
        adc     hl,de
-div_l_p_done:
        exx
+div_l_p_done:
        ld      e,a
        ld      d,b
        exx
        ld      e,c
        ld      d,b
+       exx
        pop     bc
        ret
 
 div_l_nn: ; negative dividend, negative divisor
        call    div_l_n0
+       exx
        ld      b,a
        ld      a,c
+       exx
        call    div_l_ncf
+       exx
        ld      c,a
        exx
        ld      a,b
-       exx
        call    div_l_ncf
-       exx
        ld      b,a
        ld      a,c
-       exx
        call    div_l_ncf
        jr      nc,div_l_n_done
-       exx
        add     hl,de
        exx
        adc     hl,de
-div_l_n_done:
        exx
+div_l_n_done:
        ld      e,a
        ld      d,b
        exx
        ld      e,c
        ld      d,b
+       exx
        pop     bc
 inc_l:                                 ; get into range divisor+1..0
-       exx
        inc     hl
        ld      a,l
        or      h
-       exx
        ret     nz
+       exx
        inc     hl
+       exx
        ret
 
 dec_l:
-       exx
        ld      a,l
        or      h
        dec     hl
-       exx
        ret     nz
+       exx
        dec     hl
+       exx
        ret
 
 ; non-restoring division routine
@@ -2337,7 +2355,6 @@ div_lcf:
 div_l0: ; bit 0, above
        scf
        rla
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2345,11 +2362,11 @@ div_l0: ; bit 0, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      nc,div_l01
        dec     a
 div_l11: ; bit 1, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2357,11 +2374,11 @@ div_l11: ; bit 1, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      nc,div_l12
        inc     a
 div_l02: ; bit 2, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2369,11 +2386,11 @@ div_l02: ; bit 2, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      nc,div_l03
        dec     a
 div_l13: ; bit 3, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2381,11 +2398,11 @@ div_l13: ; bit 3, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      nc,div_l14
        inc     a
 div_l04: ; bit 4, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2393,11 +2410,11 @@ div_l04: ; bit 4, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      nc,div_l05
        dec     a
 div_l15: ; bit 5, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2405,11 +2422,11 @@ div_l15: ; bit 5, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      nc,div_l16
        inc     a
 div_l06: ; bit 6, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2417,11 +2434,11 @@ div_l06: ; bit 6, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      nc,div_l07
        dec     a
 div_l17: ; bit 7, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2429,6 +2446,7 @@ div_l17: ; bit 7, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      nc,div_l18
        inc     a
 div_l08: ; done, above
@@ -2439,7 +2457,6 @@ div_l08: ; done, above
 
 div_l1: ; bit 0, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2447,11 +2464,11 @@ div_l1: ; bit 0, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      nc,div_l11
        inc     a
 div_l01: ; bit 1, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2459,11 +2476,11 @@ div_l01: ; bit 1, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      nc,div_l02
        dec     a
 div_l12: ; bit 2, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2471,11 +2488,11 @@ div_l12: ; bit 2, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      nc,div_l13
        inc     a
 div_l03: ; bit 3, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2483,11 +2500,11 @@ div_l03: ; bit 3, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      nc,div_l04
        dec     a
 div_l14: ; bit 4, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2495,11 +2512,11 @@ div_l14: ; bit 4, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      nc,div_l15
        inc     a
 div_l05: ; bit 5, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2507,11 +2524,11 @@ div_l05: ; bit 5, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      nc,div_l06
        dec     a
 div_l16: ; bit 6, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2519,11 +2536,11 @@ div_l16: ; bit 6, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      nc,div_l17
        inc     a
 div_l07: ; bit 7, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2531,6 +2548,7 @@ div_l07: ; bit 7, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      nc,div_l08
        dec     a
 div_l18: ; done, below
@@ -2548,20 +2566,19 @@ div_l_ncf:
 div_l_n0: ; bit 0, above
        scf
        rla
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
-       or      a
        exx
+       or      a
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      c,div_l_n01
        dec     a
 div_l_n11: ; bit 1, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2569,24 +2586,24 @@ div_l_n11: ; bit 1, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      c,div_l_n12
        inc     a
 div_l_n02: ; bit 2, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
-       or      a
        exx
+       or      a
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      c,div_l_n03
        dec     a
 div_l_n13: ; bit 3, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2594,24 +2611,24 @@ div_l_n13: ; bit 3, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      c,div_l_n14
        inc     a
 div_l_n04: ; bit 4, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
-       or      a
        exx
+       or      a
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      c,div_l_n05
        dec     a
 div_l_n15: ; bit 5, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2619,11 +2636,11 @@ div_l_n15: ; bit 5, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      c,div_l_n16
        inc     a
 div_l_n06: ; bit 6, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2632,11 +2649,11 @@ div_l_n06: ; bit 6, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      c,div_l_n07
        dec     a
 div_l_n17: ; bit 7, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2644,6 +2661,7 @@ div_l_n17: ; bit 7, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      c,div_l_n18
        inc     a
 div_l_n08: ; done, above
@@ -2654,7 +2672,6 @@ div_l_n08: ; done, above
 
 div_l_n1: ; bit 0, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2662,11 +2679,11 @@ div_l_n1: ; bit 0, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      c,div_l_n11
        inc     a
 div_l_n01: ; bit 1, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2675,11 +2692,11 @@ div_l_n01: ; bit 1, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      c,div_l_n02
        dec     a
 div_l_n12: ; bit 2, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2687,11 +2704,11 @@ div_l_n12: ; bit 2, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      c,div_l_n13
        inc     a
 div_l_n03: ; bit 3, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2700,11 +2717,11 @@ div_l_n03: ; bit 3, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      c,div_l_n04
        dec     a
 div_l_n14: ; bit 4, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2712,11 +2729,11 @@ div_l_n14: ; bit 4, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      c,div_l_n15
        inc     a
 div_l_n05: ; bit 5, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2725,11 +2742,11 @@ div_l_n05: ; bit 5, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      c,div_l_n06
        dec     a
 div_l_n16: ; bit 6, below
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2737,11 +2754,11 @@ div_l_n16: ; bit 6, below
        add     hl,de
        exx
        adc     hl,de
+       exx
        jp      c,div_l_n17
        inc     a
 div_l_n07: ; bit 7, above
        add     a,a
-       exx
        adc     hl,hl
        exx
        adc     hl,hl
@@ -2750,6 +2767,7 @@ div_l_n07: ; bit 7, above
        sbc     hl,de
        exx
        sbc     hl,de
+       exx
        jp      c,div_l_n08
        dec     a
 div_l_n18: ; done, below
@@ -2761,7 +2779,7 @@ div_l_n18: ; done, below
 
 ; debugging
 
-print_trace: ; print af, bc, de:hl', hl:de', sp, (sp+2):(sp)
+print_trace: ; print af, bc, hl':de, de':hl, (sp+2):(sp), sp
        call    print_trace2
        ld      a,(bc)
        inc     bc
@@ -2780,34 +2798,29 @@ print_trace2:
        call    print_word
        ld      a,' 
        call    print_char
-       ld      l,e
-       ld      h,d
-       call    print_word
-       ld      a,':
-       call    print_char
        exx
        push    hl
        exx
        pop     hl
        call    print_word
-       ld      a,' 
+       ld      a,':
        call    print_char
-       pop     af
-       pop     hl
-       push    hl
-       push    af
+       ld      l,e
+       ld      h,d
        call    print_word
-       ld      a,':
+       ld      a,' 
        call    print_char
        exx
        push    de
        exx
        pop     hl
        call    print_word
-       ld      a,' 
+       ld      a,':
        call    print_char
-       ld      hl,6
-       add     hl,sp
+       pop     af
+       pop     hl
+       push    hl
+       push    af
        call    print_word
        ld      a,' 
        call    print_char
@@ -2827,6 +2840,11 @@ print_trace2:
        ld      h,(hl)
        ld      l,a
        call    print_word
+       ld      a,' 
+       call    print_char
+       ld      hl,6
+       add     hl,sp
+       call    print_word
        ld      a,0xd
        call    print_char
        ld      a,0xa