On Tue, Dec 19, 2023 at 09:45:05AM -0800, Darrick J. Wong wrote: > Eh, there's lots of, uh, cleanup opportunities in the xattr code. ;) > > The changes below look reasonable, but I wonder -- the leaf and node add > functions do a similar thing; can they go too? > > I'm assuming those can't go away because they actually set @args->index > and @args->rmt* and we might've blown that away after the initial lookup > in xfs_attr_set? But maybe they can? Insofar as figuring all that out > is probably an entire campaign on its own. Yeah, this looks pretty scary to touch for a cleanup series that's already gone kinda out of bounds.. > > > So.. I'm tempted to just turn these checks into asserts with something > > like the below on top of this patch, I'll just need to see if it survives > > testing: > > I'll await your return then. :) It has been surviving testing just fine over night.
