Closed Bug 49129 Opened 25 years ago Closed 25 years ago

Memory Faults on OpenBSD Compile

Categories

(SeaMonkey :: Build Config, defect, P3)

x86
OpenBSD
defect

Tracking

(Not tracked)

VERIFIED DUPLICATE of bug 44426

People

(Reporter: chrisc, Assigned: cls)

Details

During the registration phase, it Memory faults and core dumps (Error code 139) just after "nsNativeComponentLoader: autoregistering succeeded"
I assume this is after the build has completed and when you tried running ./mozilla ? Please provide more info... OS version, compiler, linker, etc.
I think this is a duplicate of 49036. Could you do a "gdb -c mozilla-bin.core" or a backtrace from within gdb?
This is OpenBSD 2.7 i386. The compiler is GNU Make 3.78.1 The core file I found was regxpcom.core. It looks as though the build finished, and it was on the post build process.
Ok, this bug looks like its a duplicate of 44426. (i've finally figured out 49036 is a duplicate of it too). There is a patch for 44426, give it a try. I still get a segfault, but it gets a good bit past the nsNativeComponentLoader part.
I applied the patch and it helped get through the build process and it completed the install. However, I get a memory error when executing the program.
I think this bug should be marked a duplicate of 44426, as i'm fairly sure the initial problem was due to the JS_NewStringCopyZ / gcc 2.95.x bug. As for the current problem you are now having, it's probably the same as the one I have and is related to nsFSStringConversion::UCSToNewFS. However, without more information, i do not know. Regardless, i think a seperate bug report should be made for it. I am planning to make a bugreport on my problem after I look into things a bit more...
Changing OS to new OpenBSD field.
OS: other → OpenBSD
Based on Zach's comments, marking a dupe of 44426 *** This bug has been marked as a duplicate of 44426 ***
Status: UNCONFIRMED → RESOLVED
Closed: 25 years ago
Resolution: --- → DUPLICATE
verified.
Status: RESOLVED → VERIFIED
Product: Browser → Seamonkey
You need to log in before you can comment on or make changes to this bug.