On Mon, Jul 11, 2011 at 04:45:48PM +0100, Bob Dunlop wrote: > Hi, > > I've been chasing an intermittent problem with busybox's fdisk > implementation. About one command in ten was failing on me when used > with slow external USB disks on a 320MHz ARM system. The BLKRRPART > ioctl was returning device busy. > > I traced it to a commented out sleep(2) in fdisk.c. With the sleep > reinstated fdisk is reliable. > > I know that linux sync() is documented as not returning until the buffers > have been flushed but I think something is delaying them, possibly in the > USB stack.
In my opinion the patch is unacceptable as-is. Any sleeps need to be optional and off-by-default, and at least someone should investigate the time actually needed. Or it could just loop retrying until it's no longer busy. This would be much more robust. Rich _______________________________________________ busybox mailing list [email protected] http://lists.busybox.net/mailman/listinfo/busybox
