Minor code size optimizations.

README.md

@@ -41,7 +41,7 @@ please let me know under chrisly@platon42.de. Thank you.
 
 The new replayer comes as a drop-in binary replacement if you wish.
 In this case you will get faster sample generation (about 12%
-faster on 68000) and about 45% less CPU time spent. However, you
+faster on 68000) and about 45% less CPU time spent during playback. However, you
 won't get stuff as song-end detection and precalc progress this way.
 This mode uses the old CPU DMA wait that takes away 8 raster lines.
 
@@ -121,7 +121,7 @@ solve this problem.
 - Optimized base displacement by reordering variables.
 - Further optimized ADSR code.
 - Optimized wave loop code.
-- Bake in this strange vibrato speed multiplication to precalculated vibrato value (where possible).
+- Baked in this strange vibrato speed multiplication to precalculated vibrato value (where possible).
 - Various small optimizations.
 - Store instrument number * 4 on loading to avoid using two adds every frame.
 - Optimized speed/shuffle code. Idea of using xor turned out to make things too complicated for pattern breaks/jumps.

src/raspberry_casket.asm

@@ -1,5 +1,5 @@
 ;--------------------------------------------------------------------
-; Raspberry Casket Player V1.1 (28-Dec-2022)
+; Raspberry Casket Player V1.1 (20-May-2023)
 ; ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 ;
 ; Provided by Chris 'platon42' Hodges <chrisly@platon42.de>
@@ -97,8 +97,7 @@
 ; of the time is spent on muls operations, so this is the limiting
 ; factor.
 ;
-; Raspberry Casket is about twice as fast as the old replayer for
+; Raspberry Casket is about twice as fast as the old replayer for playback.
-; playback.
 ;
 ; Unfortunately, the replayer is still pretty slow and has high
 ; jitter compared to other standard music replayers.
@@ -814,17 +813,15 @@ pre_SongInit:
         moveq.l #0,d1
         move.b  (a0)+,d1                        ; ii_adsr_attack
         add.w   d1,d1
-        lea     pre_fast_roll_off_16(pc),a5
+        move.w  pre_fast_roll_off_16-pre_vib_delay_table(a5,d1.w),d1
-        move.w  (a5,d1.w),d1
         move.w  d1,uii_adsr_attack(a4)
 
         moveq.l #0,d1
         move.b  (a0)+,d1                        ; ii_adsr_decay
-        lea     pre_ramp_up_16(pc),a5
+        move.b  pre_ramp_up_16-pre_vib_delay_table(a5,d1.w),uii_adsr_decay+1(a4)
-        move.b  (a5,d1.w),uii_adsr_decay+1(a4)
 
         move.b  (a0)+,d1                        ; ii_adsr_sustain
-                    ; what is this? a patch?
+        ; what is this? a patch?
         cmp.b   #15,d1
         bne.s   .dont_patch_sustain
         moveq.l #16,d1
@@ -834,8 +831,7 @@ pre_SongInit:
 
         moveq.l #0,d1
         move.b  (a0)+,d1                        ; ii_adsr_release
-        lea     (pre_ramp_up_16,pc),a5
+        move.b  pre_ramp_up_16-pre_vib_delay_table(a5,d1.w),uii_adsr_release(a4)
-        move.b  (a5,d1.w),uii_adsr_release(a4)
 
         move.b  (a0)+,d1                        ; ii_pattern_steps
         move.b  d1,uii_pattern_steps(a4)
@@ -1257,7 +1253,6 @@ pre_PlayerInit:
 
         moveq.l #0,d3
         move.b  wi_osc_phase_min_b(a3),d3
-
         mulu    d5,d3
         lsl.l   #6,d3
 
@@ -1387,7 +1382,7 @@ pre_PlayerInit:
 
         cmpa.l  #$8000,a5       ; if symmetrical
         beq.s   .gen_noise_centered
-        ; what does this do? (a5 - $8000) (a5 +$7fff)&$8000
+        ; FIXME what does this do? d4 = (a5 - $8000) d1 = (a5 + $7fff)&$ffff8000 -> d4 - d1 == ((a5 - $8000) - ((a5 + $7fff)&$ffff8000)
         move.l  a5,d4
         addi.l  #$FFFF8000,d4
         move.l  a5,d1
@@ -1593,7 +1588,7 @@ pre_PlayerInit:
 .entry_to_filter_loop
         move.l  d0,a6
         move.l  d3,d1       ; flt_speed_b*128
-        adda.l  d1,a6	    ; suppress M68kUnexpectedConditionalInstruction
+        adda.l  d1,a6       ; suppress M68kUnexpectedConditionalInstruction
         bgt.s   .filter_speed_pos
 
 .filter_speed_neg
@@ -3535,7 +3530,7 @@ pre_PlayerTick:
 .dont_release_note
 
 ; ----------------------------------------
-; calculate final volume output = inst_vol * ADSR volume * pattern volume 
+; calculate final volume output = inst_vol * ADSR volume * pattern volume
 
         IFNE    PRETRACKER_VOLUME_TABLE
         lea     pv_volume_table(a4),a1
@@ -3550,7 +3545,7 @@ pre_PlayerTick:
         lsr.w   #4,d2
         mulu    d2,d1
         lsr.w   #6,d1
-        
+
         moveq.l #0,d2
         move.b  pcd_pat_vol_b(a5),d2
         mulu    d1,d2
@@ -3577,7 +3572,7 @@ pre_PlayerTick:
 
         clr.b   pcd_wave_offset_b(a5)
 
-        ; keep current direction of ping-pong unchanged 
+        ; keep current direction of ping-pong unchanged
         move.b  pcd_inst_ping_pong_dir_b(a5),d4
         bpl.s   .wave_move_one_step_ahead
         sub.w   d2,d1       ; go in reverse direction one step?
@@ -3851,7 +3846,7 @@ pre_PlayerTick:
         beq     .updatechannels         ; no track delay for last channel
 
         moveq.l #MAX_TRACK_DELAY-1,d0   ; load from last buffer
-        
+
         ; handle track delay
         cmpi.b  #$FF,d3
         beq.s   .clear_track_delay
@@ -3871,7 +3866,7 @@ pre_PlayerTick:
         move.l  (a1)+,(a3)+             ; ocd_length/ocd_loop_offset
         move.l  (a1)+,(a3)+             ; ocd_period/ocd_volume/ocd_trigger
         ;move.l  (a1)+,(a3)+             ; this is never used
-        
+
         move.b  -(a3),d2
         add.b   d2,d2               ; increment channel
         bne.s   .copy_trigger_for_delayed_channel
@@ -4129,6 +4124,13 @@ pre_vib_depth_table:
 pre_vib_delay_table:
         dc.b    0,4,8,10,12,14,16,18,20,24,32,40,56,96,150,255
 
+pre_ramp_up_16:
+        dc.b    0,1,3,6,7,9,10,11,12,13,14,16,19,35,55,143
+
+pre_fast_roll_off_16:
+        dc.w    $400,$200,$80,$64,$50,$40,$30,$20,$10,14,12,10,8
+        dc.w    4,2,1
+
 pre_roll_off_table:
         dc.w    $400,$200,$180,$140,$100,$C0,$A0,$80,$78,$74,$6E
         dc.w    $69,$64,$5A,$46,$40,$38,$30,$28,$20,$1F,$1E,$1D
@@ -4139,13 +4141,6 @@ pre_roll_off_table:
         dc.w    2,3,3,2,3,3,2,3,2,3,2,3,2,3,2,3,2,2,2,2,2,2,2,2,1
         dc.w    2,1,2,1,2,1,2,1,1,2,1,1,1,2,1
 
-pre_ramp_up_16:
-        dc.b    0,1,3,6,7,9,10,11,12,13,14,16,19,35,55,143
-
-pre_fast_roll_off_16:
-        dc.w    $400,$200,$80,$64,$50,$40,$30,$20,$10,14,12,10,8
-        dc.w    4,2,1
-
 pre_octave_note_offset_table:
         dc.b    1*NOTES_IN_OCTAVE*4,1*NOTES_IN_OCTAVE*4,1*NOTES_IN_OCTAVE*4
         dc.b    2*NOTES_IN_OCTAVE*4,2*NOTES_IN_OCTAVE*4,2*NOTES_IN_OCTAVE*4