On Sun, 16 Jan 2000, Linus Torvalds wrote:
>(The write-back handling will need some work at some point, we should be
>more aggressive about writing things out in a timely manner to floppies in
>particular. There's no reason to leave the drive idle but spinning and yet
>have a ton of dirty blocks around that we're just waiting to write out.
The only point of delaying flushes to disk is to merge two writes into one
single flush. This applyes even more strongly to floppies that are very
slow to write data. The slower the disk is, the more delay you want to
have between the flushes.
>That's independent of the media change, but would be another step into the
>direction of not making it such a big deal - people should be able to
>remove a floppy from a drive where the light is off without it being
>necessarily catastrophic).
It's not catastrophic for linux (or it's an implementation bug that has to
be fixed and not a design one as far I can tell) but you'll find your fd
corrupted when you'll read it from another computer, and that's expected.
Personally I don't want to fallback in luser-windows-mode just because
soembody doesn't know what unmount is good for, so the kernel I run on my
machine will continue to cache writes on floppy like with (or more then)
IDE/SCSI harddisk.
Setting a lower flushtime won't fix the hard-floppy-unplug problem either,
it may only happen to reduce the damage.
Andrea
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sun Jan 23 2000 - 21:00:14 EST