02521: ichir and clones, tantr and clones: Some colors are missing and other graphics errors

ID	Category [?]	Severity [?]	Reproducibility	Date Submitted	Last Update
02521	Color/Palette	Minor	Always	Oct 10, 2008, 23:28	Oct 11, 2008, 13:14

Tester	shinji	View Status	Public	Platform	MAME (Self-compiled)
Assigned To	Haze	Resolution	Fixed	OS	Windows XP/Vista 32-bit
Status [?]	Resolved			Driver	sega/segac2.cpp
Version	0.127u7	Fixed in Version	0.127u8	Build	Normal
		Fixed in Git Commit		Github Pull Request #

Summary	02521: ichir and clones, tantr and clones: Some colors are missing and other graphics errors
Description	The graphics in ichir are missing some colors and the text is missing or invisible due to the same issue. In tantr there are scrolling issues on the bottom (i.e. the background for the score/lives section is scrolling when it shouldn't). This also happens to all clones of each game respectively. The color table in all games appears correct when I compared 0.127 against 0.127u7 so the offset may be wrong?
Steps To Reproduce	Load and start a game for each. The effects should be noticeable immediately when entering a game.
Additional Information	I have uploaded screenshots onto my webspace so that you may see a comparison between 0.127 and 0.127u7. Both are my own builds but I have made no customizations or any optimizations on the builds that these screenshots came from. They are from ichir (Ichidant-R (World)) but you get the picture. 0.127: http://www.elite-systems.org/mametester/ichir_0127.PNG 0.127u7: http://www.elite-systems.org/mametester/ichir_0127u7.PNG I do use Core 2 Duo optimized builds normally however I verified it on an unoptimized one and I am reporting it on that. It is being initially reported as minor and I will allow the developer to make a proper severity level choice. I do not know what revision it started occurring on.
Github Commit

Flags
Regression Version
Affected Sets / Systems	ichir and clones, tantr and clones

Attached Files	0000.png (6,218 bytes) Oct 11, 2008, 03:36
Attached Files	0001.png (5,545 bytes) Oct 11, 2008, 03:36

No.02750 Haze Senior Tester Oct 10, 2008, 23:36	segac2 is still using the old genesis code, the raster interrupts probably broke completely (they were never right anyway)
No.02751 shinji Tester Oct 10, 2008, 23:42 edited on: Oct 10, 2008, 23:42	Heh. Ok. I don't think the break is in the driver anyways. I just did a diff (should of done it before) and the only thing that changed was one line of code. Nothing that would affect this. I'll check the diffs and see what have changed since I was the one that applied them. ;) EDIT: Never expected one of the devs to see this report so soon.
No.02752 Fujix Administrator Oct 11, 2008, 03:37	Confirmed the color regression. Added some screen shots.
No.02753 Haze Senior Tester Oct 11, 2008, 08:22	right, the raster interrupt isn't being triggered, probably related to the 68k IRQ changes.
No.02757 Haze Senior Tester Oct 11, 2008, 12:27	i've fixed it, although the change is pretty extensive and ribbit has broken completely in the process.
No.02759 Haze Senior Tester Oct 11, 2008, 13:14	ok, restored ribbit to where it was before.. although this does render the 'ribbit crash' replay rather useless as timings have changed..