 No.19757
Haze Senior Tester
Feb 5, 2022, 22:32
|
if this started happening only in the latest version it's almost certainly https://github.com/mamedev/mame/commit/eb89e1ac3189495a3ee1ce4dccd22ab444b9b8b4 |
|---|---|
 No.19758
hap Developer
Feb 6, 2022, 00:38
edited on: Feb 6, 2022, 00:39
|
The equations converted to C++ are correct. The file was added here: https://github.com/mamedev/mame/commit/5ddd0069199e3a7bcb67f81cef15bd1ec76e8c1f Maybe the culprit is a bad PAL(GAL) file, I don't know where AJR obtained it from. Another one was added by Clawgrip later: ROM_LOAD( "u001_pal16r6cn.j2", 0x0000, 0x0104, CRC(84dbe498) SHA1(5863342b2db85ffef31b5e9ce26bfd8fca9923b0) ) jedutil shows this for Clawgrip's dump: weird, no equations for 13,14,15? jedutil -view u001_pal16r6cn.j2 pal16r6 Inputs: 2, 3, 4, 5, 6, 7, 8, 9, 12, 13, 14, 15, 16, 17, 18, 19 Outputs: 13 (Registered, Output feedback registered, Active low) 14 (Registered, Output feedback registered, Active low) 15 (Registered, Output feedback registered, Active low) 16 (Registered, Output feedback registered, Active low) 17 (Registered, Output feedback registered, Active low) 18 (Registered, Output feedback registered, Active low) Equations: /rf13 := rf13.oe = OE /rf14 := rf14.oe = OE /rf15 := rf15.oe = OE /rf16 := /rf16 + /rf16 + rf14 & rf17 + rf14 & /rf16 & rf17 + rf14 & /rf15 & rf17 & /rf18 + rf13 & rf14 & /rf15 & /rf16 & rf17 & /rf18 rf16.oe = OE /rf17 := i2 & /i5 + i2 & /i3 & i4 & /i5 & rf17 + i2 & i3 & /i4 & /i5 & /rf14 & /rf17 rf17.oe = OE /rf18 := /i7 & /i8 & /i9 & /rf15 & /rf18 + rf14 & rf17 + rf13 & rf14 & rf17 & /rf18 rf18.oe = OE |
|
No.19759
hap Developer
Feb 6, 2022, 00:54
edited on: Feb 6, 2022, 00:56
|
Quick log from game running in MAME: lastwritten data _ expected value _ m_pal_u001 value attract mode: A5_40_40 CD_16_16 36_7A_7A 6F_3E_3E A5_40_40 CD_16_16 36_7A_7A 6F_3E_3E etc sequence after EXTRA: C2_4C_7E NO MATCH 36_2A_7A NO MATCH 6F_66_3E NO MATCH The latter 2 are very suspicious, since the last written value is the same as the one during attract mode, but it expects a different return value from before. I expect it means that some of the PAL outputs are fed back to the inputs. |
|
No.19762
hap Developer
Feb 6, 2022, 11:29
|
The Taito sets nop out the EXTRA protection check. Maybe the PAL dump is correct for the Taito versions? |
|
No.19763
hap Developer
Feb 6, 2022, 12:25
|
I readded the crazy hack for mrdo, it should be working same like before |
|
No.19784
Steve Coomber Tester
Feb 12, 2022, 16:28
|
hap, I have just tested mrdo, mrdoy, mrdobl & yankeedo. I can confirm the issue has been corrected. |
|
No.21279
Haze Senior Tester
Apr 5, 2023, 07:40
edited on: Apr 5, 2023, 08:22
|
https://www.mameworld.info/ubbthreads/showthreaded.php?Cat=&Number=395929 is this related or something else? (actually it makes no sense, given the fix was to restore the old hack behavior?) |
|
No.21280
hap Developer
Apr 5, 2023, 08:50
|
Right, if it's a new regression, the regression commit is somewhere else. I won't be surprised if it's protection related in general. |
|
No.21281
Haze Senior Tester
Apr 5, 2023, 09:09
edited on: Apr 5, 2023, 09:10
|
It's definitely odd. I can't reproduce it either, I thought it could be Z80 adjustments breaking it somehow, but then I'd expect more failures. I wonder if it's something the user has done. Anyway, not related to this. |