[03:10] <jbailey> Joy, klibc segfaults from the shell but not from gdb
[03:12] <lifeless> funkay
[03:42] <jbailey> I don't mind debugging by printf, but this segfault is in the exit code, and it's already closed stdout.
[03:43] <jbailey> And when I strace, the bug doesn't show up either.
[03:43] <jbailey> Woot to looking at exit codes.
[03:44] <jbailey> Oooo..  Maybe process attach will work.
[03:56] <jbailey> Cool.  attach does work.
[02:31] <Mithrandir> Keybuk: I never had a T42.  I've had a T21, but that's years ago.
[02:31] <Keybuk> ahh
[02:31] <Keybuk> it's hard to keep track of them all
[02:31] <Keybuk> how was your holiday?
[02:31] <Mithrandir> nice, walked around a lot.
[02:31] <Mithrandir> you have weird beer in the UK.  Not bad, but a bit weird.
[02:32] <Mithrandir> I prefer it a bit more bubbly.
[02:33] <Keybuk> you aren't going to get any disagreement from me, I've never been a fan of UK beer
[04:34] <Mithrandir> Keybuk: is there any way I can detect if a drive is a CD/DVD drive in the initramfs?
[04:36] <jbailey> Mithrandir: /proc/ide/hda/media
[04:36] <Keybuk> what jbailey said
[04:36] <Mithrandir> and for SATA/SCSI drives?
[04:37] <jbailey> There's a type in sysfs.
[04:37] <jbailey> I don't remember that one off hand, since I got it right the first time.
[04:37] <jbailey> !@#$ing IDE.
[04:40] <Mithrandir> hmm, I could just look for the first removable block device and try mounting that, then continue.
[04:49] <jbailey> Use fstype on each device.
[04:49] <jbailey> It's be an iso-9660 won't it?
[04:50] <jbailey> That way you don't have any filesystem scans on devices that aren't the right one.
[05:05] <Mithrandir> fstype doesn't do iso9660, but that's easy enough to fix.
[05:12] <jbailey> Mithrandir: If you feed the patch to me, I'm working on a klibc update anyway.
[05:14] <Mithrandir> I'll do a byhand drive-by of the mountroot stuff in initramfs first, but sure, you'll have it.