walt <w41ter@xxxxxxxxx> posted
455C8C8E.2030409@xxxxxxxxx, excerpted below, on Thu, 16 Nov 2006 08:06:38
> Duncan wrote:
>> FWIW, I returned to glibc-2.4-r4 (had to hack portage to get it to do it
>> since the package has a safety feature preventing downgrades, and my
>> non-pan issue is gone, as I expected it would be from an earlier test.
>> I'm not sure about the pan crashes as they were rather less frequent and
>> thus harder to duplicate, but so far, so good. =8^)...
> I'm afraid I'm seeing the same pan problem with glibc-2.4-r4, but it's
> rare, so I'm at a loss to trigger it on purpose. I guess running pan
> from gdb every time might catch it.
I still haven't seen it back, and I normally would have seen it a
couple times by now if I was still using glibc-2.5, so either it's two
different problems, or glibc-2.5 makes a bad thing worse.
I suspect the glibc-2.5 thing is either amd64 or SMP related, tho. I
asked on the amd64 group and several reported no problems with it, which
leans it toward SMP, or maybe both. Anyway, I doubt it's something
single-core single-CPU x86 will see.
Duncan - List replies preferred. No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master." Richard Stallman
Pan-users mailing list