On Mon, 3 Apr 2006, Ken Moffat wrote:
> On Mon, 3 Apr 2006, Alexander E. Patrakov wrote:
>
> > Yes, but waiting for rootfs is done in initramfs, not after mounting rootfs.
> > So the problem is "flash as /usr", not "flash as root". But, on the fast
> > machines with simple udev rules (like LFS ones), even the regular TTYs
> > trigger
> > the bug without $confirm because of the finite time when the uevent sits in
> > the kernel socket buffer.
>
Can't trigger this on my athlon64 4000+ with a test for 'test $confirm
-lt 1' : all 64 regular ttys and 4 ttyS on major 4, and 256 on major 3
(ttyp0-zf, ttya0-ef). And this is with a few extra simple rules for my
desktop. I'll just have to accept this part of it (and reduce the delay
locally when I remember). Or, maybe it's partly tied up with the exact
kernel version and config.
Ken
--
das eine Mal als Trag??die, das andere Mal als Farce
--
http://linuxfromscratch.org/mailman/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page
