Re: performance problem: "git commit filename"

From: Junio C Hamano <gitster@pobox.com>
Date: Jan 13, 2008, 08:12 UTC
Message-ID: <7vtzliqh3u.fsf@gitster.siamese.dyndns.org>
In-Reply-To: <alpine.LFD.1.00.0801121949180.2806@woody.linux-foundation.org>

Linus Torvalds <torvalds@linux-foundation.org> writes:

> HOWEVER. This was just a quick hack, and while it all looks sane, this is 
> some damn core code. Somebody else should double- and triple-check this.

Double-checked.  The patch looks sane.

> [ That 4x lstat thing bothers me. I think we should add a flag to the 
>   index saying "we checked this once already, it's clean", so that if we 
>   do multiple passes over the index, we can still do just a single lstat() 
>   on just the first pass. But that's a separate issue.

I've thought about this in a different context before, but it seemed quite tricky, as some codepaths in more complex commands (commit being one of them) tend to use the cache and discard to use it for different purpose (like creating temporary index and then reading the real index). Besides, I had an impression that we ran out of the bits of ce_flags in the cache entry, although we could shorten the maximum path we support from 4k down to 2k bytes. I'll have to think about this a bit more.