From: Junio C Hamano Date: Mon, 13 May 2013 14:46:09 GMT Subject: Re: [PATCH] patch-ids.c: cache patch IDs in a notes tree Message-ID: <7v38tros3y.fsf@alter.siamese.dyndns.org> In-Reply-To: <20130513140243.GQ2299@serenity.lan> John Keeping writes: > On Mon, May 13, 2013 at 06:53:29AM -0700, Junio C Hamano wrote: >> John Keeping writes: >> >> > The caching layer could also introduce false positives though, which is >> > more serious. If you cache patch IDs with a pathspec restriction ... >> >> What? What business does patch-id have with pathspec-limited diff >> generation? You do not rebase or cherry-pick with pathspec, so >> unless you are populating the patch-id cache at a wrong point (like, >> say whenevern "git show $commit" is run), I am not sure why pathspec >> limit becomes even an issue. > > revision.c::cherry_pick_list() sets the pathspec to what was specified > in the revision options. It's done that since commit 36d56de (Fix > --cherry-pick with given paths, 2007-07-10) and t6007 tests that it > works. Then the caching should be automatically turned off when pathspec is given.