Nosync/sync wave selection optimized among other smaller optimizations.

README.md

@@ -101,7 +101,7 @@ The original code compressed with *Blueberry's* Shrinkler goes from
 18052 bytes down to 9023 bytes.
 
 Raspberry Casket, depending on the features compiled in, is about
-5802 bytes and shrinkles down to ~4125 bytes (in isolation).
+5784 bytes and shrinkles down to ~4118 bytes (in isolation).
 
 So this means that the optimization is not just "on the outside".
 
@@ -165,10 +165,11 @@ solve this problem.
 - This also fixes usages of illegal period 0 in the lead-in that could cause the replay to miss the first trigger.
 - Moved pattern table init from PlayerInit to SongInit, optimized SongInit a bit.
 - Wave order table filling moved and optimized in SongInit.
-- Added Presto player draft.
 - Bugfix: Songend detection for back-jumps was broken since at least V1.1.
 - Optimized some more wave selection code.
-- Drop-in replacement code size: 5802 bytes.
+- Nosync/sync wave selection optimized.
+- Added Presto player draft.
+- Drop-in replacement code size: 5784 bytes.
 
 ### V1.x (unreleased)
 - Fixed a bug regarding the copper output mode with looping waves having a loop-offset.

src/raspberry_casket.asm

@@ -1,5 +1,5 @@
 ;--------------------------------------------------------------------
-; Raspberry Casket Player V2.x (19-Aug-2023)
+; Raspberry Casket Player V2.x (20-Aug-2023)
 ; ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 ;
 ; Provided by Chris 'platon42' Hodges <chrisly@platon42.de>
@@ -96,7 +96,7 @@
 ; 18052 bytes down to 9023 bytes.
 ;
 ; Raspberry Casket, depending on the features compiled in, is about
-; 5802 bytes and shrinkles down to ~4125 bytes (in isolation).
+; 5784 bytes and shrinkles down to ~4118 bytes (in isolation).
 ;
 ; So this means that the optimization is not just "on the outside".
 ;
@@ -433,7 +433,7 @@ pre_SongInit:
         dbra    d3,.waveorderloop
 
 .hasnowaveordering
-        
+
         lea     sv_pattern_table(a1),a0
 .pattableloop
         move.l  a4,(a0)+
@@ -1476,7 +1476,6 @@ pre_PlayerTick:
         IFNE    PRETRACKER_PARANOIA_MODE
         moveq.l #0,d2               ; default to not stitched
         move.b  (a0)+,d1            ; pdb_pitch_ctrl get pitch byte
-        move.b  d1,d6
         bpl.s   .inst_note_is_not_stitched  ; means that note is stitched
 
         tst.w   d4
@@ -1486,15 +1485,15 @@ pre_PlayerTick:
         moveq.l #1,d2               ; next note will be fetched immediately
 .inst_note_is_not_stitched
         ELSE
-        move.b  (a0)+,d1            ; pdb_pitch_ctrl get pitch byte
+        move.b  (a0)+,d6            ; pdb_pitch_ctrl get pitch byte
-        move.b  d1,d6
         smi     d2                  ; note stitched?
         ;neg.b   d2
         ENDC
 
         tst.b   d3
         bne.s   .skippitchloading
-        and.w   #$3f,d1
+        moveq.l #$3f,d1
+        and.w   d6,d1
         beq.s   .skippitchloading   ; no new note
         subq.w  #1,d1
         lsl.w   #4,d1
@@ -1508,9 +1507,9 @@ pre_PlayerTick:
         moveq.l #15,d6
         and.b   (a0)+,d6        ; pdb_effect_cmd command number
         add.w   d6,d6
-        moveq.l #0,d5
-        move.b  (a0)+,d5        ; pdb_effect_data command parameter byte
         move.w  .inst_command_jmptable(pc,d6.w),d3
+        moveq.l #0,d5
+        move.b  (a0)+,d5        ; pdb_effect_data command parameter byte, note that condition codes are used in inst_set_speed
         jmp     .inst_command_jmptable(pc,d3.w)
 
 .inst_command_jmptable
@@ -1534,6 +1533,7 @@ pre_PlayerTick:
 ; d0 = current step / next step
 ; d5 = command parameter data / scratch
 ; d2 = note stitched flag
+; d1 = scratch
 ; d3 = scratch
 ; d6 = scratch
 ; ----------------------------------------
@@ -1546,7 +1546,7 @@ pre_PlayerTick:
         add.w   d5,d5
         cmp.w   pcd_inst_wave_num4_w(a5),d5
         beq.s   .inst_set_wave_rts
-        
+
 .inst_select_wave_subroutine
         move.w  d5,pcd_inst_wave_num4_w(a5)
         move.l  sv_waveinfo_table(a6,d5.w),a3
@@ -1560,16 +1560,58 @@ pre_PlayerTick:
 
         moveq.l #0,d3
 
-        tst.w   d6
+        move.w  wi_loop_offset_w(a3),d1     ; is unlikely >= 32768 -- if it is, it will be past end of sample
-        bne.s   .inst_select_wave_nosync
 
-        move.w  wi_loop_offset_w(a3),d6     ; is unlikely >= 32768 -- if it is, it will be past end of sample
         tst.w   wi_subloop_len_w(a3)
-        bne.s   .inst_set_wave_has_subloop
+        beq.s   .inst_set_wave_has_no_subloop
-.inst_set_wave_has_no_subloop
-        adda.w  d6,a1                       ; add loop offset (which is actually not a loop offset for one-shot samples)
 
-        sub.w   d6,d5
+        tst.w   d6
+        beq.s   .inst_set_wave_has_subloop
+
+        ; nosync version
+.inst_set_wave_ns_has_subloop
+        move.w  d3,pcd_out_lof_w(a5)
+        move.l  a1,pcd_out_ptr_l(a5)
+
+        IFEQ    PRETRACKER_BUGFIX_CODE
+        move.w  pcd_inst_loop_offset_w(a5),d1
+        cmp.w   d1,d5
+        ELSE
+        cmp.w   pcd_inst_loop_offset_w(a5),d5
+        ENDC
+        bhs.s   .inst_set_wave_ns_keep_pp
+        st      pcd_inst_ping_pong_dir_b(a5)
+.inst_set_wave_ns_keep_pp
+        clr.w   pcd_inst_subloop_wait_w(a5)
+
+        IFEQ    PRETRACKER_BUGFIX_CODE          ; This adds an extra jump to the wave that's not desired
+        sub.w   wi_subloop_step_w(a3),d1
+        move.w  d1,pcd_inst_loop_offset_w(a5)
+        ENDC
+        rts
+
+.inst_set_wave_has_no_subloop
+        IFEQ    PRETRACKER_BUGFIX_CODE
+        tst.w   d6
+        beq.s   .inst_select_wave_nosync_no_subloop
+
+        ; note that nosync on non-looping waves doesn't effectivly do anything special
+        adda.w  d1,a1
+        sub.w   d1,d5
+        ;cmp.w   #1,d5 ; not necessary as increases in steps of 2
+        bhi.s   .inst_set_wave_ns_has_min_length
+        moveq.l #2,d5
+.inst_set_wave_ns_has_min_length
+        move.w  d5,pcd_out_len_w(a5)
+        moveq.l #-1,d3
+
+        bra.s   .inst_set_wave_ns_has_subloop
+
+.inst_select_wave_nosync_no_subloop
+        ENDC
+
+        adda.w  d1,a1                       ; add loop offset (which is actually not a loop offset for one-shot samples)
+        sub.w   d1,d5
         ;cmp.w   #1,d5 ; not necessary as increases in steps of 2
         bhi.s   .inst_set_wave_has_min_length
         moveq.l #2,d5
@@ -1583,7 +1625,7 @@ pre_PlayerTick:
 .inst_set_wave_has_subloop
         move.w  d3,pcd_out_lof_w(a5)
         move.l  a1,pcd_out_ptr_l(a5)
-        move.w  d6,pcd_inst_loop_offset_w(a5)
+        move.w  d1,pcd_inst_loop_offset_w(a5)
 
         st      pcd_inst_ping_pong_dir_b(a5)
         moveq.l #0,d5
@@ -1593,57 +1635,6 @@ pre_PlayerTick:
 .inst_set_wave_rts
         rts
 
-; ----------------------------------------
-.inst_select_wave_nosync
-        IFNE    PRETRACKER_BUGFIX_CODE
-        move.w  wi_loop_offset_w(a3),d6
-        tst.w   wi_subloop_len_w(a3)
-        beq.s   .inst_set_wave_has_no_subloop
-        ELSE
-        tst.w   wi_subloop_len_w(a3)
-        bne.s   .inst_set_wave_ns_has_subloop
-
-        ; note that nosync on non-looping waves doesn't effectivly do anything special
-        move.w  wi_loop_offset_w(a3),d6
-        adda.w  d6,a1
-
-        sub.w   d6,d5
-        ;cmp.w   #1,d5 ; not necessary as increases in steps of 2
-        bhi.s   .inst_set_wave_ns_has_min_length
-        moveq.l #2,d5
-.inst_set_wave_ns_has_min_length
-        move.w  d5,pcd_out_len_w(a5)
-
-        moveq.l #-1,d3
-        ENDC
-.inst_set_wave_ns_has_subloop
-        move.w  d3,pcd_out_lof_w(a5)
-        move.l  a1,pcd_out_ptr_l(a5)
-
-        IFEQ    PRETRACKER_BUGFIX_CODE
-        move.w  pcd_inst_loop_offset_w(a5),d6
-        cmp.w   d6,d5
-        ELSE
-        cmp.w   pcd_inst_loop_offset_w(a5),d5
-        ENDC
-        bhs.s   .inst_set_wave_ns_keep_pp
-        st      pcd_inst_ping_pong_dir_b(a5)
-.inst_set_wave_ns_keep_pp
-        clr.w   pcd_inst_subloop_wait_w(a5)
-
-        IFEQ    PRETRACKER_BUGFIX_CODE          ; This adds an extra jump to the wave that's not desired
-        sub.w   wi_subloop_step_w(a3),d6
-        move.w  d6,pcd_inst_loop_offset_w(a5)
-        ENDC
-        rts
-
-; ----------------------------------------
-.inst_slide_down
-        neg.w   d5
-.inst_slide_up
-        move.w  d5,pcd_inst_pitch_slide_w(a5)
-        bra     .inst_cmd_cont_next
-
 ; ----------------------------------------
 .inst_adsr
         subq.w  #1,d5
@@ -1681,7 +1672,7 @@ pre_PlayerTick:
         bra.s   .inst_cmd_cont_next
 
 .inst_vol_slide_up
-        lsr.b   #4,d5
+        lsr.w   #4,d5
         move.b  d5,pcd_inst_vol_slide_b(a5)
         bra.s   .inst_cmd_cont_next
 
@@ -1708,12 +1699,12 @@ pre_PlayerTick:
         movea.l sv_inst_patterns_table-4(a6,d1.w),a0
 
         add.w   d5,a0
-        add.w   d5,d5
+        add.w   d5,a0
         add.w   d5,a0
 
         bra.s   .inst_fetch_next
 .inst_we_were_stitched
-        move.b  pcd_inst_speed_stop_b(a5),d2
+        moveq.l #0,d2
         bra.s   .inst_cont_from_nasty_double_jump
 
 ; ----------------------------------------
@@ -1753,12 +1744,17 @@ pre_PlayerTick:
         addq.w  #1,d0
         bra.s   .inst_fetch_next
 
+; ----------------------------------------
+.inst_slide_down
+        neg.w   d5
+.inst_slide_up
+        move.w  d5,pcd_inst_pitch_slide_w(a5)
+        bra.s   .inst_cmd_cont_next
+
 ; ----------------------------------------
 .inst_set_speed
-        tst.w   d5
+        seq     d3
-        bne.s   .inst_set_speed_nostop
+        or.b    d3,d5
-        st      d5
-.inst_set_speed_nostop
         move.b  d5,pcd_inst_speed_stop_b(a5)
 
 ; ----------------------------------------
@@ -1777,8 +1773,8 @@ pre_PlayerTick:
 .inst_pat_loop_exit
         st      d2
 .inst_pat_loop_exit2
-        add.b   pcd_inst_speed_stop_b(a5),d2
 .inst_cont_from_nasty_double_jump
+        add.b   pcd_inst_speed_stop_b(a5),d2
         move.b  d2,pcd_inst_line_ticks_b(a5)
         move.b  d0,pcd_inst_step_pos_b(a5)  ; update inst step pos
 
@@ -1830,8 +1826,9 @@ pre_PlayerTick:
         ; same code for both release and decay
 .adsr_decay_and_release
         moveq.l #0,d4
+        moveq.l #-$71,d5        ; same as $8f, we only need the byte
         move.b  pcd_adsr_phase_speed_b(a5),d4
-        cmp.b   #$8f,d4
+        cmp.b   d5,d4
         bhs.s   .adsr_absurd_slow_release
         move.b  d4,d5
         addq.b  #1,d5
@@ -1842,7 +1839,6 @@ pre_PlayerTick:
 
 .adsr_absurd_slow_release
         moveq.l #2,d4           ; FIXME I guess this should be 1, if I look at the roll-off table
-        moveq.l #-$71,d5        ; same as $8f, we only need the byte
 .adsr_release_cont
         move.b  d5,pcd_adsr_phase_speed_b(a5)
 
@@ -1998,7 +1994,7 @@ pre_PlayerTick:
         move.w  wi_loop_start_w(a3),d4
 
         move.w  d4,d2
-        sub.w   d1,d2                   ; calc how many bytes we are part front
+        sub.w   d1,d2                   ; calc how many bytes we are past front
         bmi.s   .wave_new_loop_pos_fits
 
 .wave_hit_front_of_loop