View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0019512 | MMW 5 | Casting (Google Cast / UPnP) | public | 2022-10-28 11:56 | 2022-11-03 23:46 |
| Reporter | rusty | Assigned To | |||
| Priority | immediate | Severity | crash | Reproducibility | sometimes |
| Status | closed | Resolution | fixed | ||
| Product Version | 5.0.4 | ||||
| Target Version | 5.0.4 | Fixed in Version | 5.0.4 | ||
| Summary | 0019512: Casting: Invalid pointer operation | ||||
| Description | While playing about 10 different tracks of different types to a chromecast MM generated an 'Invalid pointer operation' error (no crashlog) when about the 10th track started playing (I wasn't doing anything in MediaMonkey at the time). The error occurs infrequently, but I replicated it twice. Note: 1) In the course of testing this (i.e. prior to the error), I had also - Added tracks to Playing from another tab - Re-ordered tracks in Playing (in the main tracklist) - Used Play/Stop/Next/Previous/Seek/Volume controls However, I've been unable to determine what the specific cause is--it just seems to occur randomly. 2) although severity is labeled as 'crash' MM doesn't actually crash. The error occurs and it continues to function 3) Tested on build 2670 and 2678, and was only able to replicate on 2678, though it's possible that this was just a fluke (though the user that reported the casting bug indicated that 2670 solved it for him). | ||||
| Additional Information | Possibly related to: https://www.mediamonkey.com/forum/viewtopic.php?t=102837 0019494 (though doubtful, since the issue was reported by the user on build 2675) | ||||
| Tags | No tags attached. | ||||
| Fixed in build | 2680 | ||||
|
|
Unfortunatelly the logs are quite useless because: 1) In the "Invalid pointer operation" log I see the exception in the playback starter thread id [15116], but no crash log was generated by Eureka (for some reason) so I can't see callstack of the thread 15116 to see where and why the exception was raised. 2) Re the "out of memory" log with fastmm-enabled build: I don't know who created this build for you, but the log is missing the exception completelly and what's worse it is also missing the regular memory usage output (search MemRAM in the DbgView) , by evidence it hasn't been compiled with Eureka at all and is totally useless :-/ So please try to replicate with official 2679 debug build and once the "invalid pointer operation" occurs then try to wait for Eureka's crash log and if there really isn't any generated then go to menu > Debug > Send logs.. and submit a crash log manually. |
|
|
I was unable to trigger the exact same error, but was able to trigger crashlog 0AF10000 by playing several high bandwidth MP3/FLAC tracks and then playing another playlist of mixed formats. The error occurred on the track subsequent to a flac track that generated a codec error toast message. Note: before the error occurred, the track that was playing was playing silently, and was playing with an incorrect time indicator. Once the error occurred, MM was frozen so there was no way to submit an error through the UI. |
|
|
Based on the log it was instance of 0018860 where connection has been lost and MM was trying to re-connect, but stucked in a loop and resulted in "task queue deadlock" assertion. EDIT: Analyzing further and it was combination of 0019513 / 0018860. |
|
|
Fixed in 5.0.4.2680 |
|
|
Verified 2681. |
