On Fri, May 5, 2017 at 1:01 PM, Christoph Hellwig <hch@xxxxxxxxxxxxx> wrote: > On Fri, May 05, 2017 at 12:44:28PM +0300, Amir Goldstein wrote: >> This is used by overlayfs to encode intrasystem unique file handles. > > I think generate_random_uuid is the wrong interface to spread. > The right one would be uuid_be_gen, although I'll still need > to audit why one uses get_random_bytes and the other prandom_bytes. > Christoph, You're keep getting in my way with this uuid business ;-) Seriously, one has to do things in certain order. There is no reason to wait for the common uuid code to settle down before fixing something as trivial as this and it makes very little sense IMO to use uuid_be_gen() for sb->s_uuid as it is now. I am working actively with you and Andy to sort out the uuid.h mess and it even seems we are getting closer to consensus. There is no real 'cost' in 'spreading' generate_random_uuid() to tmpfs and I promise to post a patch to replace generate_random_uuid() in tmpfs to the new API once it is ready. Cheers, Amir. -- To unsubscribe from this list: send the line "unsubscribe linux-unionfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
