Hi,On Fri, 2003-03-21 at 00:12, Andrew Morton wrote: > > Well, there's still the > > if (err) > > __ext3_std_error(inode->i_sb, where, err); > > case in ext3_journal_stop() to worry about > > We already have that. Only if we fix the underlying problem --- I was only pointing out that even if we drop the setting of s_dirt entirely, which was what we were trying to fix, we can't avoid having to find the sb. > But I'm not particularly fussed either way - it will only be 100-200 bytes of > code saved. Yep, but there are probably other places we can find where we can avoid passing the sb around too if we have the back-pointer. I guess it makes sense to go ahead with that. > The journal and the superblock have a definite one-to-one relationship - I think the > backpointer makes sense. But whatever - I'll let you flip that coin. OK, go for it and I'll merge for 2.4. Cheers, Stephen _______________________________________________ Ext3-users@redhat.com https://listman.redhat.com/mailman/listinfo/ext3-users
