Hi Nico, I'd suggest making the patch file name more descriptive and suffixing it with '.patch'. 'Makefile.in.patch' would be ok particularly if there are multiple changes in the file. If the change is to address a specific issue then you could name the patch file according to what it fixes. At the very least this makes sense if you need to add another patch in the future.
Otherwise, with this and the changes Paul suggested it looks fine. Thanks Amanda Nicolas Williams wrote: > I need a code review, though I think I can count Bill Sommerfeld as > having done part of it: > > 6755391 sqlite3 should not be built with SQLITE_DEBUG; uses access(2) too > often, slows things down > 6762083 sqlite3 default pagesize selection could be better (SQLite3 ticket > #3452) > > http://cr.opensolaris.org/~nico/sqlite3-big-pages/ > > I'll be sending a separate code review request to sqlite-users at sqlite.org, > though I've not seen code review requests there. > > One regular participant from the sqlite-users list commented already in > the zfs-discuss thread about this problem, suggesting that we make > adjust the default cache_size selection to match the changes to the > default page_size selection. Bill and I talked about this and I made > the change, which Bill tested. > > All in all Bill says that this change significantly improves performance > of Evolution, which now uses SQLite3. > > Nico >
