I just finished helping someone (who was trying to build a bootable CD
using my BICK package) debug why they were getting a failure with
2.4.20-pre11, but things worked OK with 2.4.18. After several cycles
of "try this" and finally getting his kernel output captured via a
serial port, I found the problem. The initial ram disk mounted at /
was being mounted read/only in 2.4.20-pre11 whereas it was mounted
read/write in 2.4.18 (I haven't tested 2.4.19 yet because previously
upgrading the kernel on a rescue CD wasn't a priority).
The work around is to either specify "rw" in the boot parameters or to
do a remount to remove the read/only bit during initialization. The
next version of BICK will have this dealt with.
I suspect this was just a change being made to default to read/only for
hard disk mounted root filesystems. I can see maybe a few suprises it
might cause, but mostly no problems. But it could be more problematic,
or at least confusing, for initial ram disk situations. Maybe it should
test for whether / is initrd and turn off MS_RDONLY if so.
ref: http://freshmeat.net/projects/bick/
-- ----------------------------------------------------------------- | Phil Howard - KA9WGN | Dallas | http://linuxhomepage.com/ | | phil-nospam@ipal.net | Texas, USA | http://ka9wgn.ham.org/ | ----------------------------------------------------------------- - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Oct 31 2002 - 22:00:49 EST