08522: samesame: Continue music keeps playing after you select yes

ID	Category [?]	Severity [?]	Reproducibility	Date Submitted	Last Update
08522	Sound	Major	Always	Nov 30, 2022, 21:58	Dec 7, 2022, 17:22

Tester	Cow	View Status	Public	Platform	MAME (Official Binary)
Assigned To	hap	Resolution	Fixed	OS	Windows 10/11 (64-bit)
Status [?]	Resolved			Driver	toaplan/toaplan1.cpp
Version	0.250	Fixed in Version	0.251	Build	64-bit
		Fixed in Git Commit	b237736	Github Pull Request #

Summary	08522: samesame: Continue music keeps playing after you select yes
Description	After selecting "Yes" on the Continue screen, the continue screen theme will keep on playing while all other sound effects (ex. firing your gun) are M.I.A.
Steps To Reproduce	Get a game over (easiest is to just stand still) on the Continue screen insert a coin and select "Yes"
Additional Information
Github Commit

Flags
Regression Version
Affected Sets / Systems	samesame

Attached Files	Mame 2022.11.30 - 15.48.54.03-1-1.mp4 (3,278,865 bytes) Nov 30, 2022, 22:15 Uploaded by Cow

No.20879 kmg Senior Tester Nov 30, 2022, 23:06 edited on: Nov 30, 2022, 23:08	I see mostly the same thing. But the description is incomplete. It seems the coin up sound during the continue music is the culprit. Note in the video when Cow coins up on the continue screen the sound effect does not play, and the continue music is all that's played from that point onward. If you already have available credits and hit continue on the continue screen, the continue music stops and sounds proceed as normal. This behavior happens for me (on macos) on top of tree and on a 0.246 I have sitting around.
No.20880 kmg Senior Tester Dec 1, 2022, 01:20	A better way to repro: Flip the service mode DIP switch to on. Use stick and P2 buttons to navigate menus. Sound check #31 is the continue track. Sound check #6 and #7 are the end level/high score/etc tracks. Play any of those three tracks and play sound check #40, the coin sound (alternatively try to insert either coin or use the service button, all of which normally play the coin sound). The coin sound will not play and button 2 will no longer stop sounds in the sound check.
No.20881 kmg Senior Tester Dec 1, 2022, 01:37 edited on: Dec 1, 2022, 04:21	This is likely reproducible on all fireshrk clones. For instance if you beat level 1 and try inserting coins during the landing/level end music the coin sound will not play, that music will loop indefinitely and no other sounds will play. I haven't checked all clones, but you can also verify this as described in my previous comment via the service mode sound check.
No.20882 kmg Senior Tester Dec 1, 2022, 06:52	Note, this also repros on 0.229—before the YM rewrite.
No.20885 Robbbert Senior Tester Dec 1, 2022, 12:35	It works on 0.212, but is broken on 0.215
No.20927 hap Developer Dec 7, 2022, 00:45 edited on: Dec 7, 2022, 01:23	It appears to work if you (re)add this to toaplan1_samesame_state::hd647180_io_map: map(0x63, 0x63).r(FUNC(toaplan1_samesame_state::cmdavailable_r)); But I don't understand why that would be needed. Since AJR removed them deliberately in a lot of drivers, and added devcb. In samesame case, it's this one: audiocpu.in_pd_callback().set(FUNC(toaplan1_samesame_state::cmdavailable_r)); edit ah, it does: in a,($63) while A=0x08 z180 IN opcode is: OP(op,db) { unsigned n = ARG() \| (_A << 8); _A = IN(n ); } /* IN A,(n) */ so it tries reading from port 0x0863, which is not considered an internal port.
No.20928 dink Tester Dec 7, 2022, 01:27	maybe needs a 0xff mask on port#?
No.20930 hap Developer Dec 7, 2022, 02:19	No, 0x0100-0xff00 is external I/O. I suspect when it reads from 0x0863, it's a bug in the game. https://github.com/mamedev/mame/commit/b23773662401c335799ee5134a8b1963edb4272f Continue music now stops when selecting YES. It still does not play the coin sfx on the continue screen though.
No.20931 kmg Senior Tester Dec 7, 2022, 03:02	When I checked 0.214 and earlier (before the regression) they didn't play a coin sound during the continue screen music, though the music would stop when appropriate. Maybe it's like that on hardware and this is now resolved?
No.20933 Haze Senior Tester Dec 7, 2022, 16:19 edited on: Dec 7, 2022, 16:20	could it be an error in the internal ROM dump? remember the dumping process for these was very fragile, with a small chance of errors being introduced. I seem to recall this one didn't always give the same readout, but this was the most consistent readout.
No.20936 hap Developer Dec 7, 2022, 17:21 edited on: Dec 7, 2022, 17:29	It's a possibility, IMO not high enough chance to have to mark it bad dump. edit but if the 2nd and 3rd most consistent dumps only have a handful of bytes different from this one, it's worth looking at? That is, if the files are still somewhere.