"Christopher C. Chimelis" wrote: >On Thu, 23 Nov 2000, Peter Maydell wrote: >> Is anybody else having trouble with Mozilla on Alpha?
>> Has anybody else encountered this? Anybody working on it, or >> should I try to build a debug version and play with it in gdb? >> [I can't find anything obviously this bug in Bugzilla.] > >Yes, yes, and please do :-P I saw the same problem, which is why M18 was >never uploaded to woody or potato. I'm not sure yet if this is a bug in >regxpcom or if we might've run into a bash problem. [Why do you think this might be a bash problem?] Well, I tried to do this. However, the gdb in potato (4.18.19990928-1) doesn't seem to work properly for threaded programs :-< If you take a simple thread-using program (eg, the example code in bug 74525) and run it under gdb, you immediately get: Program received signal ?, Unknown signal. __sigsuspend () at ../sysdeps/unix/sysv/linux/alpha/sigsuspend.S:42 42 ../sysdeps/unix/sysv/linux/alpha/sigsuspend.S: No such file or directory. Current language: auto; currently asm and if you try to continue it (eventually) hangs somehow so ctrl-C won't bring it back to the gdb prompt (you have to kill it from another xterm). Running the program not under gdb produces the expected results. [I know this works on intel.] Any ideas here? If I take the plunge and upgrade to unstable in the hopes that a newer gdb will fix this, am I likely to run into any nasty libc breakage or similar? (I'd really rather stay with potato or potato-plus-a-few-things if possible, and IME upgrades to unstable always seem to require upgrading most of the critical system components :-<) Peter Maydell
