On Tue, 22 Sep 2009 13:40:52 +0200 Christoph Hellwig <hch@xxxxxx> wrote: > On Mon, Sep 21, 2009 at 01:15:19PM -0700, Andrew Morton wrote: > > The syscall is all about dirty memory management and userspace control > > over IO timing/volume. Trying to use it for data integrity purposes is a > > bit perverse. > > Yes, its is. But that needs to be clearly documented. If you have > better wording to make it crystal clear you're welcome. > > Additionally I have a half-backed patch to allow it calling into ->fsync > to actually make it usable as a fdatasync_range/fsync_range, but I > wonder if we should just add those as syscalls instead. hm, hard call. I guess we should avoid permitting people to sync the metadata before the data, because that would provide an easy way to expose huge amounts of unwritten blocks if the machine crashes at the right time. Or perhaps it's not worth bothering about that. -- To unsubscribe from this list: send the line "unsubscribe linux-man" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
