View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0018586 | MMW 5 | Main Panel: Toolbars & Menus | public | 2021-11-25 07:21 | 2021-11-26 18:54 |
| Reporter | barrym | Assigned To | |||
| Priority | urgent | Severity | crash | Reproducibility | sometimes |
| Status | resolved | Resolution | unable to reproduce | ||
| Platform | Windows | OS | - | OS Version | 10 |
| Product Version | 5.0.2 | ||||
| Target Version | 5.0.3 | ||||
| Summary | 0018586: Release Candidate 2523 has crashed on me twice ... Find More From Same possibly implicated | ||||
| Description | Unable to replicate at will. Happened twice withing 1st 30 minutes using the new release candidate. Both times I had pressed Find More From same recently from the context menu. Other than that I was just navigating within MM5. The first time submitted dump 87990000 to you. I had debug running the 2nd time. The log is at https://www.dropbox.com/s/kaart58izjcivqn/crash%20_from_2523.LOG?dl=0 The run hung at about 00059015 Windows said the MM was not responding .. I allowed Windows to terminate MM5. The debug facility was still recording when I restarted MM5 to get the release #. Submitted Ticket 0003205: | ||||
| Tags | No tags attached. | ||||
| Fixed in build | |||||
|
|
Hi Barry, I see some 87990000 crashes here, but all are from older builds (and not from the past days), so your log is not here. can you please search your computer for MediaMonkeyEngine.el file and attach to ticket # 3205? It should be located in the same folder as MediaMonkey.exe or as subfolders of "%appdata%\EurekaLab s.a.s\EurekaLog\Bug Reports\" or at "%appdata%\EurekaLog\MediaMonkey" Thanks! |
|
|
In the DbgView log I see 00058941 380.22061157 [40616] [1125/164528.581:ERROR:gl_surface_egl.cc(778)] EGL Driver message (Error) eglPostSubBufferNV: Bad match. 00058942 380.22088623 [40616] [1125/164528.581:ERROR:skia_output_surface_impl_on_gpu.cc(1495)] Failed to make current. but the error is not from MM5 process, as MM5 porcess has id [44736] By searching the internet I found this: https://stackoverflow.com/questions/57090258/how-to-address-the-error-egl-driver-message-error-eglquerydeviceattribext-bad which indicates that the '--use-gl' Chromium attribute might be responsible for this. And I see that Petr added '--use-gl' in SVN revision 38862 (trying to fix 0018283), but subsequently reverted in SVN revision 38867 and it seems that --use-gl is now used just in case of Mac (OSX) version. So it doesn not look like the possible cause. Waiting for the MediaMonkeyEngine.el from Barry... |
|
|
--use-gl parameter can be used to define what renderer to use (GL, SwiftShader etc.) ... currently used only on MAC |
|
|
.el file now attached to my ticket. Failure to dispatch there error reports seems to be a recurrent weakness. Before dispatching the report I pressed the Details button. Maybe there is a weakness there? |
|
|
Analyzing the EL and seeing that it was debug assertion that there were two albums in album list with identical persistentID, this is unexpected. Unfortunatelly although in the DbgView part of the log should show which ID isn't unique it is missing there for some reason. I adjusted it and added the particular unique ID directly to the exception description (to have more info next time this happens). Let me know if you can replicate the issue and whether I should sent updated debug build to you. EDIT: As for the reproducibility: Can you share your MM5.DB with me via the ticket? I guess it might be reproducible like being in Entire LIbrary > All tracks with view 'Grid (by Album)' and using 'Find more from same' on the last album in the list. Maybe also 'Column Filter' would need to be enabled (hard to guess, but could be something like this). |
|
|
Barry is not longer able to replicate the crash, so resolving (to be opened once the crash re-appears in the future builds). |
