> It seems the lock ups may have somehow been caused by the glibc 2.06 and
> 2.07-5. I have 2.07-6 (RedHat) and it seems to not lock up anymore.
> I will give it a few more trials. Seems it may have been an ld.so or
> weird libc/glibc interaction.
>
I'm running libc-5.4.33, and netscape has been unusable since the late
.80's :-(. Worse yet, it has resisted all my efforts at stracing from a
cold start. If attach the program when it's running, the Heisenberg
Uncertainty Principle takes effect and, while sluggish at redraws, it
refuses to freeze.
It does lots and lots of this, though:
sigreturn() = ? (mask now [])
gettimeofday({891646637, 643327}, NULL) = 0
gettimeofday({891646637, 643771}, NULL) = 0
sigprocmask(SIG_BLOCK, NULL, []) = 0
gettimeofday({891646637, 644724}, NULL) = 0
sigprocmask(SIG_SETMASK, [], NULL) = 0
gettimeofday({891646637, 645751}, NULL) = 0
gettimeofday({891646637, 646269}, NULL) = 0
ioctl(9, FIONREAD, [0]) = 0
write(9, "\2\0\4\0c\4\200\2\0@\0\0\0\0\0\0"..., 16) = 16
--- SIGALRM (Alarm clock) ---
Maybe 100-200 gettimeofday() calls every time it is deiconified. When
idle it's:
oldselect(23, [6 7 9 10 22], NULL, [6], {182, 916298}
If anyone has a suggestion for getting it to start under strace, let me
know. This is getting frustrating.
Steve
(Running netscape happily with strace attached <g>.)
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu