- --
Viewing Issue Advanced Details
ID | Category [?] | Severity [?] | Reproducibility | Date Submitted | Last Update |
---|---|---|---|---|---|
03703 | Debugger | Minor | Always | Feb 1, 2010, 17:09 | Feb 17, 2010, 03:45 |
Tester | smkdan | View Status | Public | Platform | MAME (Official Binary) |
Assigned To | Resolution | Open | OS | Windows Vista/7 (32-bit) | |
Status [?] | Acknowledged | Driver | |||
Version | 0.136 | Fixed in Version | Build | Debug | |
Fixed in Git Commit | Github Pull Request # | ||||
Summary | 03703: The wpiset watchpoint will not snap the debugger unless the written data is 00 (Z80) | ||||
Description |
When running the Z80 debugger, wpiset will not snap the debugger unless wpdata is 00. It will simply ignore any others. For example: wpiset 4,1,w Only snaps when data=0 wpiset 4,1,w,wpdata>0 Will never snap. I ran into this when running Neo Geo Z80 program in the debugger and watching writes to YM2610 address port 0. Instances of OUT ($04),a where a != 0 will simply be ignored. |
||||
Steps To Reproduce |
Run lbowling.zip with -debug switch. Enter: Make sure you have (audiocpu) - Z80 focus in the debugger wpiset 4,1,w Get to the title screen and insert a coin (you may need to insert two to make the coin sound). The debugger should snap with PC=13BB. Note the OUT ($04),a directly above. Note PC=13C2 which is another OUT ($04),a. Step until this is reached, the debugger will not snap because A!=00. |
||||
Additional Information |
I do not know how broad this issue is but it seems very general so I just picked neodrv.c since that's what I do all my debugging in. The example applies to a write to YM2610 from the Z80 using the OUT instruction that is triggered by the insert coin sound, it's one of the only sounds that seem to cause the debugger to actually snap (since it writes 00 to the 2610). |
||||
Github Commit | |||||
Flags | |||||
Regression Version | |||||
Affected Sets / Systems | |||||
Attached Files
|
|||||
Relationships
There are no relationship linked to this issue. |
Notes
2
No.05606
Tafoid Administrator
Feb 1, 2010, 17:57
|
Adjusted steps to reproduce. I'll leave this open and ask for any Devs who are more knowledgeable about the code to critique. |
---|---|
No.05615
Kale Developer
Feb 3, 2010, 11:29
edited on: Feb 3, 2010, 11:29 |
Mmh ... maybe it's actually writing to the mirrors, hence it doesn't write to port 0x0004 but to 0x**04? AM_RANGE(0x04, 0x07) AM_MIRROR(0xff00) AM_DEVREADWRITE("ymsnd", ym2610_r, ym2610_w) |