Opened 12 years ago
Last modified 12 years ago
#195 new defect
ProjectLibre unstable and produces hs_err_pidXXX.logs
Reported by: | Joop | Owned by: | Joop |
---|---|---|---|
Priority: | major | Milestone: | GA6 |
Component: | general | Version: | 1.6.0 Build 25 GA3 |
Severity: | medium | Keywords: | ProjectLibre |
Cc: |
Description
I try to run ProjectLibre latest version in Java GA1 and Java GA3 in order to rule out development differences in Java, however behavior is same in both Java version. One time ProjectLibre is just starting up, you can load a datafile. Sometimes it hangs there. Other times it won't start at all and produce a hs_err_pidXXX.log. I'm unsure if this is a Java problem or if this is a ProjectLibre problem. If the latter istrue then I have to issue a ticket there.
Attachments (4)
Change History (9)
by , 12 years ago
Attachment: | hs_err_pid169.log added |
---|
comment:1 by , 12 years ago
Milestone: | Enhanced → GA5 |
---|---|
Priority: | blocker → major |
Please test with GA4 and Odin 0.8.8.
comment:2 by , 12 years ago
This is a strange one. With the latest Odin and Java and kernel32 and jawtos2 test fixes (from other tickets), I get similar results. Sometimes it starts and seems to work, other times it just crashes. When it crashes I get the identical POPUPLOG and error log every time (different from Joop though). I will upload those. Every time it starts (whether it crashes or not) I also see some cryptic messages in the console - not sure if they are relevant. I'll upload them too.
It's a shame this doesn't work - it seems like it could be very useful....
by , 12 years ago
Attachment: | daves_hs_err_pid78.log added |
---|
Error log using Odin 0.8.8 and Java 6_27 GA4
by , 12 years ago
Attachment: | daves_console.log added |
---|
Console message using Odin 0.8.8 and Java 6_27 GA4
comment:3 by , 12 years ago
Short report. It crashes at startup every second time here, the crash is exactly the same as David has. According to the map file, it crashes in some static code following Universe::check_mirror() in an attempt to read at the address 0xFFFFFFFF. I checked the code but didn't find anything at the first attempt.
comment:4 by , 12 years ago
It's something new. The debug version asserts in objectMonitor.cpp at line 912:
assert(false, "Non-balanced monitor enter/exit!");
Have no idea why the enter/exit call is unbalanced. I guess that this is a reason for the crash of the release build.
comment:5 by , 12 years ago
The call stack, to be exact:
InterpreterRuntime::monitorexit() ObjectSynchronizer::slow_exit() ObjectSynchronizer::fast_exit() ObjectMonitor::exit() <-- assertion s here
Laatste PID log met Java GA3.