[iOs] ROMs other than v1.3 and v2.04 casues RetroArch to Crash on Core Start #542
Labels
Comments
|
@sonninnos I was able to build & deploy retroarch and the up to date PUAE core from Xcode for debug, unfortunately problem seems to lies within core dynamic library (puae_libretro_ios.dylib) which I am unable to debug properly. If someone points me to the right direction about debugging dylib files I am willing to do so.
Cheers, |
|
Whoops forgot about this. Simply building the core with |
|
I forced the
|
|
A stack trace in text form would be more useful, but do try commenting out Also the regular RA debug level log to the point of crash might be useful. |
|
I can get a trace in text form, yes. The crash indicates that the
‘locklist’ pointer is null.
I did try commenting out that define and while it did not crash, it
resulted in hanging the execution upon starting content.
…On Sun, Jan 8, 2023 at 5:56 AM sonninnos ***@***.***> wrote:
A stack trace in text form would be more useful, but do try commenting out
UAE_FILESYS_THREADS define in retrodep/sysconfig.h.
—
Reply to this email directly, view it on GitHub
<#542 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAEJ4JUH6OOSUXJ5QXZXGKDWRLPTFANCNFSM52NT7DGQ>
.
You are receiving this because you commented.Message ID:
***@***.***>
|
|
@sonninnos Here's the text stack trace on the crashing thread: Here's the trace on the main thread at the time: |
|
Since it is indeed something something threading related, perhaps try adding this to makefile for iOS/tvOS, since it is already added and required for OSX: |
|
Ah that's a good callout, but adding |
|
Progress of some sort at least. How does the debug level libretro log look? |
|
Yes, progress :) Where do the libretro debug logs appear? They don't appear in stdout/console - are they logged to a file? |
|
Depends on the setting if they are written to a file or not. https://docs.libretro.com/guides/generating-retroarch-logs/#generating-logs-via-menu |
|
Wow, I totally did not know that, that's really good to know, thanks for that! I was wondering why I never saw anything logged to the console. Here's the dump of the log: |
|
Aargh, could you fix that to be more readable instead of showing everything twice for no reason..? Or I guess no need since it stops so suddenly anyway. And you might also want to try it with rewind disabled, since savestates are very sluggish with this core anyway, and might cause extra issues with filesystem threads. |
|
Hmm not sure why it's showing 2 log statements for each log entry - might be a libretro thing. I added more of my own logging and turned on filesystem logging ( It looks like it fails on handling the .info file of an LHA - I tried a couple LHA files and they seem to stop on the .info file - I still haven't tracked down where the exit is happening: |
|
Never seen that double logging before so can't be a libretro thing per se, and pretty sure the info file just happens to be the first file in the archive. |
|
Ok, I figured out the problem - the So it turns out that the PLATFLAG that was for macOS did the trick! I was able to open a couple LHA archives without any issue. Here's the PR to fix this: Thanks so much @sonninnos for your help on this! |
|
Thanks to both of you @yoshisuga & @sonninnos cheers. |
|
Huzzah! Pretty nasty debug feature that one, heh. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This new build crashes on core start if try to boot with roms other than 1.3 or 2.04 regardless of selected 'Model' in the settings, on iOs.
The text was updated successfully, but these errors were encountered: