View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0019052 | MMW 5 | Other | public | 2022-05-04 00:16 | 2022-05-06 10:11 |
| Reporter | peke | Assigned To | |||
| Priority | immediate | Severity | crash | Reproducibility | random |
| Status | closed | Resolution | fixed | ||
| Target Version | 5.0.3 | Fixed in Version | 5.0.3 | ||
| Summary | 0019052: MM5 2620 Crashed on number of ocasions | ||||
| Description | MM5 crashed on Number of occasions. Logs: 6A3707D5 6A370D9F 62D60000 -> 0018288 ? Logs and video of a possible crash point sent offline. It crashes random, but there was letovers from previous crash. | ||||
| Tags | No tags attached. | ||||
| Fixed in build | 2621 | ||||
|
|
Can you send me bugreports (if they're stored somewhere) by skype ? |
|
|
The crash logs have been submitted by Peke, but there has been a server issue (indicated by Petr) -- so we were not receiving ELFs for the last couple of days. I've fixed a debug info regression and send new MM.exe and MME.exe to Peke (to generate new logs to see more). |
|
|
Should be fixed in 2621 -- sent new EXE to Peke to confirm. It was a regression with signed versus unsigned integer usage in TFastReadWriteLock -- Petr made recently the change because of MAC OSX support - revision 39475 from 03/09/2022, i.e. two months old regression, but was tricky to replicate as it was timing issue (one thread must acquiring read lock while another already had a read lock). |
|
|
Re-opened: Peke is still observing the issue... |
|
|
Fixed in 2621 (confirmed by Peke using the supplied EXE files) |
|
|
Verified 2621 Tricky one. |
