I've seen it before--my understanding is that adding the debugging symbols changes the memory signature in the stack. So for example, what used to corrupt some crucial memory might now just overwrite text in a debugging symbol. Alternatively, the presence of debuggibg symbols might slightly alter the probability of encountering a race condition.
I could be totally wrong here, though.. what I've seen before was due to compiling with the -g flag vs without, I assume the dbgsym packages do essentially the same thing. On Sat, Mar 29, 2008 at 12:03 AM, James Westby <[EMAIL PROTECTED]> wrote: > On Sat, 2008-03-29 at 02:07 +0000, lcampagn wrote: > > I can't reproduce the bug with the dbgsym package, I left it running > > for 45 minutes or so with no problem. After removing dbgsym, the > > segfault is back again after ~5 minutes. > > > > Ok, that's quite unusual, are you sure no other package was > removed with the dbgsym packages? > > > > Thanks, > > James > > -- > segfault in hydrogen after ~5minutes of use > https://bugs.launchpad.net/bugs/197809 > You received this bug notification because you are a direct subscriber > of the bug. > -- segfault in hydrogen after ~5minutes of use https://bugs.launchpad.net/bugs/197809 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
