http://bugs.winehq.org/show_bug.cgi?id=58508
Bug ID: 58508 Summary: Dn-FamiTracker 0.5.0.2: hangs on startup (regression, have bisected commit) Product: Wine Version: 10.11 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: winmm&mci Assignee: wine-bugs@winehq.org Reporter: zerica@callcc.eu Distribution: ---
Created attachment 78972 --> http://bugs.winehq.org/attachment.cgi?id=78972 program output before freezing
i believe this happens with any program that does audio input, but i'm focusing on dn-famitracker since it was where i first noticed the issue.
i have isolated the faulty commit to exactly https://gitlab.winehq.org/wine/wine/-/commit/f7044c4b8f3ab02ca8070cb37c7cc90.... reverting it in 10.11 fixes the freeze entirely.
http://bugs.winehq.org/show_bug.cgi?id=58508
Ken Sharp imwellcushtymelike@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Regression SHA1| |f7044c4b8f3ab02ca8070cb37c7 | |cc90e311d512f CC| |julliard@winehq.org Keywords| |regression Summary|Dn-FamiTracker 0.5.0.2: |Dn-FamiTracker 0.5.0.2: |hangs on startup |hangs on startup |(regression, have bisected | |commit) |
--- Comment #1 from Ken Sharp imwellcushtymelike@gmail.com ---
winmm: Always load mmdevapi as audio driver.
Is this the application? https://github.com/Dn-Programming-Core-Management/Dn-FamiTracker
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #2 from zerica@callcc.eu zerica@callcc.eu --- yes, that's correct
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #3 from zerica@callcc.eu zerica@callcc.eu --- after some further testing, it seems like while reverting the commit on 10.11 is a fix, reverting the commit on 10.12 is NOT. i will try doing some further bisecting...
http://bugs.winehq.org/show_bug.cgi?id=58508
Stian Low wineryyyyy@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |wineryyyyy@gmail.com
--- Comment #4 from Stian Low wineryyyyy@gmail.com --- Unable to reproduce for wine-10.11-217-g3c4cc9d04b6 which is my custom dev build that uses https://gitlab.winehq.org/wine/wine/-/wikis/Building-Wine#shared-wow64.
I successfully tested both the win32 and x64 versions of both 0.5.0.2 and the latest version 0.5.1.1 at: https://github.com/Dn-Programming-Core-Management/Dn-FamiTracker/releases
The extent of the test was creating a quick tracks and verifying playback. I've used Famitracker and it seems to be working as expected from these brief tests.
The regression commits are included in my wine build so maybe the issue has been patched by later commits.
Can you provide instructions on how to reproduce if this brief test isn't enough to trigger the bug?
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #5 from zerica@callcc.eu zerica@callcc.eu --- simply starting the program should be enough, it freezes before the main window has a chance to show up. i wouldn't be surprised if there is something specific to my setup influencing the failure, so for reference, here is the distro packaging script https://github.com/chimera-linux/cports/tree/master/main/wine. is there any debug flag that could provide potentially useful logs?
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #6 from Stian Low wineryyyyy@gmail.com --- WINEDEBUG=+timestamp,+pid,+loaddll,+seh minimally. Someone else may be able to recommend additional flags that may also help.
Comparing your output to mine looks similar except for the very first error in yours:
00c0:err:hid:udev_bus_init UDEV monitor creation failed
~/public_html/winehq_appdb_stianlow $ ~/tmp/wine_stianlow/install_32_64/bin/wine /media/any/store_0/wine_catalogue/Dn-FamiTracker_v0502_x64_Release/Dn-FamiTracker.exe Dn-FamiTracker 0.5.0.2 012c:fixme:nls:RtlGetThreadPreferredUILanguages 00000038, 00007FFFFE1FF9F4, 00007FFFFE1FFA10 00007FFFFE1FF9F0 012c:fixme:nls:get_dummy_preferred_ui_language (0x38 0x409 00007FFFFE1FF9F4 00007FFFFE1FFA10 00007FFFFE1FF9F0) returning a dummy value (current locale) 0130:fixme:kernelbase:AppPolicyGetThreadInitializationType FFFFFFFFFFFFFFFA, 00007FFFFF7EFF50 0140:fixme:avrt:AvSetMmThreadCharacteristicsW (L"Pro Audio",00007FEC0482FCC0): stub 0140:fixme:avrt:AvRevertMmThreadCharacteristics (0000000012345678): stub 0140:fixme:avrt:AvSetMmThreadCharacteristicsW (L"Pro Audio",00007FEC0482FCC0): stub killed
~/public_html/winehq_appdb_stianlow $ ~/tmp/wine_stianlow/install_32_64/bin/wine /media/any/store_0/wine_catalogue/Dn-FamiTracker_v0511_Win32_Release/Dn-FamiTracker.exe 0024:fixme:nls:RtlGetThreadPreferredUILanguages 00000038, 0094FAFC, 0094FB0C 0094FB00 0024:fixme:nls:get_dummy_preferred_ui_language (0x38 0x409 0094FAFC 0094FB0C 0094FB00) returning a dummy value (current locale) 0120:fixme:kernelbase:AppPolicyGetThreadInitializationType FFFFFFFA, 01CFFF18 0130:fixme:avrt:AvSetMmThreadCharacteristicsW (L"Pro Audio",02DAFD88): stub killed
~/public_html/winehq_appdb_stianlow $ ~/tmp/wine_stianlow/install_32_64/bin/wine /media/any/store_0/wine_catalogue/Dn-FamiTracker_v0502_Win32_Release/Dn-FamiTracker.exe 0024:fixme:nls:RtlGetThreadPreferredUILanguages 00000038, 0094FAFC, 0094FB0C 0094FB00 0024:fixme:nls:get_dummy_preferred_ui_language (0x38 0x409 0094FAFC 0094FB0C 0094FB00) returning a dummy value (current locale) 0120:fixme:kernelbase:AppPolicyGetThreadInitializationType FFFFFFFA, 01CFFF18 0130:fixme:avrt:AvSetMmThreadCharacteristicsW (L"Pro Audio",02DAFDA4): stub 0130:fixme:avrt:AvRevertMmThreadCharacteristics (12345678): stub 0024:fixme:kernelbase:AppPolicyGetProcessTerminationMethod FFFFFFFA, 0094FEB8
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #7 from zerica@callcc.eu zerica@callcc.eu --- Created attachment 78973 --> http://bugs.winehq.org/attachment.cgi?id=78973 debug output (+timestamp,+pid,+loaddll,+seh), wine 10.11
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #8 from zerica@callcc.eu zerica@callcc.eu --- note that line 1122 is in response to me stopping the process after ~24 seconds of inactivity
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #9 from Stian Low wineryyyyy@gmail.com --- Looks like WINEDEBUG=+mmdevapi is relevant per the faulty commit you reported.
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #10 from zerica@callcc.eu zerica@callcc.eu --- Created attachment 78974 --> http://bugs.winehq.org/attachment.cgi?id=78974 debug output (timestamp,+pid,+loaddll,+seh,+mmdevapi), wine 10.11
process stopped in line 505 after hanging
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #11 from Stian Low wineryyyyy@gmail.com --- Or rather WINEDEBUG=+winmm. Maybe both.
Here's the list associated with winmm: dlls/winmm: driver mci mmio mmtime winmm
Try for seemingly the most relevant debug options: WINEDEBUG=+driver,+mci,+mmio,+mmtime,+winmm
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #12 from zerica@callcc.eu zerica@callcc.eu --- Created attachment 78975 --> http://bugs.winehq.org/attachment.cgi?id=78975 debug output (+driver,+mci,+mmio,+mmtime,+winmm), wine 10.11
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #13 from zerica@callcc.eu zerica@callcc.eu --- it's possible this has to do with _midi_ specifically, and not audio input as i thought
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #14 from Stian Low wineryyyyy@gmail.com --- Created attachment 78976 --> http://bugs.winehq.org/attachment.cgi?id=78976 debug output (+driver,+mci,+mmio,+mmtime,+winmm), wine 10.11
Here's mine for comparison.
The error on your side is what stands out most which is missing at the beginning of my side: 00c0:err:hid:udev_bus_init UDEV monitor creation failed
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #15 from zerica@callcc.eu zerica@callcc.eu --- another point of evidence toward this being a midi issue: i am able to start up OpenMPT 1.32.02.00 fine.... until i open the setup dialog and switch to the MIDI settings tab, at which point the program completely freezes.
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #16 from zerica@callcc.eu zerica@callcc.eu --- Created attachment 78977 --> http://bugs.winehq.org/attachment.cgi?id=78977 debug output for OpenMPT 1.32.02.00 (+driver,+mci,+mmio,+mmtime,+winmm), wine 10.11
here's the equivalent log
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #17 from zerica@callcc.eu zerica@callcc.eu --- i'm not sure the udev line is necessarily relevant to this issue... it seems to pop up in every single wine invocation, including when using a version that runs the program just fine.
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #18 from Stian Low wineryyyyy@gmail.com --- I assume all of these types of apps result in the same problem? I have more experience with MilkyTracker which runs fine through Wine also:
https://github.com/milkytracker/MilkyTracker/releases/tag/v1.05.01
milkytracker-1.05.01-win64.zip
WINEDEBUG=+midi is another available channel.
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #19 from zerica@callcc.eu zerica@callcc.eu --- milktracker starts up with no issue.
http://bugs.winehq.org/show_bug.cgi?id=58508
Ken Sharp imwellcushtymelike@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download, source URL| |https://github.com/Dn-Progr | |amming-Core-Management/Dn-F | |amiTracker
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #20 from zerica@callcc.eu zerica@callcc.eu --- is there any way to tell exactly _where_ the program is freezing?
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #21 from Stian Low wineryyyyy@gmail.com --- (In reply to zerica@callcc.eu from comment #20)
is there any way to tell exactly _where_ the program is freezing?
From https://gitlab.winehq.org/wine/wine/-/wikis/Debugging-Hints: winedbg Dn-FamiTracker.exe
bt performs backtrace for unhandled exceptions
Based on (WINEDEBUG=+driver,+mci,+mmio,+mmtime,+winmm) log midi may indeed be an issue as my equivalent log is missing: 0130:warn:driver:DriverCallback Unknown callback type 7
It even happens for your Aux so it may be deeper than just midi.
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #22 from Stian Low wineryyyyy@gmail.com --- Created attachment 78982 --> http://bugs.winehq.org/attachment.cgi?id=78982 vkeybd virtual midi successfully controlling Wine Dn-Famitracker
Attached is a snapshot of my working system with vkeybd used as a virtual midi device controlling Dn-Famitracker.
http://bugs.winehq.org/show_bug.cgi?id=58508
--- Comment #23 from Stian Low wineryyyyy@gmail.com --- I had all the same successes testing all versions against the latest official release of wine-10.12 as well:
https://gitlab.winehq.org/wine/wine/-/releases/wine-10.12