On Friday 19 June 2009 03:40:29 Colin Watson wrote: >Historically /etc/mtab couldn't be made a symlink because some >information was missing from /proc/mounts, which was particularly >relevant for the handling of loop devices. It's only relatively recently >(Linux 2.6.26) that /proc/mounts has been useful. I think there may even >still be some problems with this: >http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=531371 has some more >information.
The busybox mount command got that one right back in 2006, using /proc/mounts. (Also, you don't need to specify -o loop with the busybox mount command, it autodetects when that's needed.) I know because I'm the one who wrote that code. :) > We probably will move to /etc/mtab as a symlink (cf. > http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=494001) but we won't do > it just because of a busybox bug. When you have production systems > depending on existing behaviour, a certain amount of caution and trying > to ensure that you've covered all the bases is appropriate. Sure, I was just surprised. I've just been a bit out of the loop since about 2006, and busybox's mtab support was already deprecated legacy code back then. I thought it would have gone away by now. (Then again, I see that busybox still has devfsd, and that was also deprecated back then. Sigh. We haven't got anything similar to feature-removal-schedule.txt...) Rob -- Latency is more important than throughput. It's that simple. - Linus Torvalds _______________________________________________ busybox mailing list [email protected] http://lists.busybox.net/mailman/listinfo/busybox
