Re: [PATCH] Avoid running lstat(2) on the same cache entry.

Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

From: Johannes Schindelin <Johannes.Schindelin@...>

To: Steffen Prohaska <prohaska@...>

Cc: Linus Torvalds <torvalds@...>, Junio C Hamano <gitster@...>, Git Mailing List <git@...>

Subject: Re: [PATCH] Avoid running lstat(2) on the same cache entry.

Date: Sunday, January 20, 2008 - 11:18 am

Hi,

On Sun, 20 Jan 2008, Steffen Prohaska wrote:

quoted text> On Jan 20, 2008, at 2:48 AM, Johannes Schindelin wrote:
> 
> > On Sat, 19 Jan 2008, Linus Torvalds wrote:
> > 
> > > Best time before:
> > > 
> > > 	[torvalds@woody linux]$ time git commit > /dev/null
> > > 	real    0m0.399s
> > > 	user    0m0.232s
> > > 	sys     0m0.164s
> > > 
> > > Best time after:
> > > 
> > > 	[torvalds@woody linux]$ time git commit > /dev/null
> > > 	real    0m0.254s
> > > 	user    0m0.140s
> > > 	sys     0m0.112s
> > 
> > Wow.
> > 
> > > I bet you'll see a much bigger performance improvement from this on
> > > Windows in particular.
> > 
> > I bet so, too.  Traditionally, filesystem calls are painfully slow on
> > Windows.
> > 
> > But I cannot test before Monday, so I would not be mad if somebody else
> > could perform some tests on Windows.
> 
> Here are timings for Windows XP running in a virtual machine on a
> Laptop.  The work tree contains 7k files.  I stripped user and
> sys times because they are apparently meaningless for MINGW.
> 
> Best time before:
> 
>    $ time git commit >/dev/null
>    real    0m1.662s
> 
> Best time after:
> 
>    $ time git commit >/dev/null
>    real    0m1.196s
> 
> The absolute time improvement is obviously larger, although the
> relative improvement is slightly smaller than in Linus' example.
> 
> 
> And here are the timings for the host system, which is Mac OS X,
> on the same work tree.
> 
> Best time before:
> 
>    $ time git commit >/dev/null
>    real    0m0.571s
>    user    0m0.332s
>    sys     0m0.237s
> 
> Best time after:
> 
>    $ time git commit >/dev/null
>    real    0m0.463s
>    user    0m0.273s
>    sys     0m0.186s
> 
> Interestingly, the relative improvements are even smaller on Mac
> OS X.

Still, nothing to spit at.

FWIW I have made two patches which should be probably folded into Linus' 
patch; I will post them as a reply.

Ciao,
Dscho

-
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:

Updated in-memory index cleanup, Linus Torvalds, (Fri Jan 18, 11:25 pm)

[PATCH] Avoid running lstat(2) on the same cache entry., Junio C Hamano, (Sat Jan 19, 3:45 am)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Linus Torvalds, (Sat Jan 19, 3:47 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Linus Torvalds, (Sat Jan 19, 9:22 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Johannes Schindelin, (Sat Jan 19, 10:42 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Johannes Schindelin, (Sat Jan 19, 9:48 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Steffen Prohaska, (Sun Jan 20, 11:10 am)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Linus Torvalds, (Sun Jan 20, 2:20 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Steffen Prohaska, (Sun Jan 20, 5:40 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Linus Torvalds, (Sun Jan 20, 6:09 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Brian Downing, (Sun Jan 20, 4:03 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Johannes Schindelin, (Sun Jan 20, 11:18 am)

[PATCH] Try to resurrect the handling for 'diff-index -m', Johannes Schindelin, (Sun Jan 20, 11:21 am)

[PATCH] Also use unpack_trees() in do_diff_cache(), Johannes Schindelin, (Sun Jan 20, 11:19 am)

Re: [PATCH] Also use unpack_trees() in do_diff_cache(), Linus Torvalds, (Sun Jan 20, 4:32 pm)

Re: [PATCH] Also use unpack_trees() in do_diff_cache(), Linus Torvalds, (Sun Jan 20, 5:53 pm)

Re: [PATCH] Also use unpack_trees() in do_diff_cache(), Johannes Schindelin, (Sun Jan 20, 7:34 pm)

Re: [PATCH] Also use unpack_trees() in do_diff_cache(), Linus Torvalds, (Sun Jan 20, 7:58 pm)

Re: [PATCH] Also use unpack_trees() in do_diff_cache(), Johannes Schindelin, (Sun Jan 20, 8:19 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Steffen Prohaska, (Sun Jan 20, 6:33 am)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Junio C Hamano, (Sun Jan 20, 8:18 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Johannes Schindelin, (Sun Jan 20, 10:15 am)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Linus Torvalds, (Sat Jan 19, 10:02 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Linus Torvalds, (Sat Jan 19, 9:34 pm)

Re: [PATCH] Avoid running lstat(2) on the same cache entry., Linus Torvalds, (Sat Jan 19, 3:58 pm)

[PATCH] index: be careful when handling long names, Junio C Hamano, (Sat Jan 19, 3:42 am)


linux-kernel:
Trent Piepho	[PATCH] [POWERPC] Improve (in\|out)_beXX() asm code
Stoyan Gaydarov	From 2.4 to 2.6 to 2.7?
Andi Kleen	[PATCH] [4/50] x86: add cpu codenames for Kconfig.cpu
Greg Kroah-Hartman	[PATCH 013/196] Documentation: Replace obsolete "driverfs" with "sysfs".
git:

openbsd-misc:

linux-netdev:
Gerrit Renker	[PATCH 27/37] dccp: Integration of dynamic feature activation - part 2 (server side)
Jarek Poplawski	Re: HTB accuracy for high speed
Jarek Poplawski	[PATCH] pkt_sched: Destroy gen estimators under rtnl_lock().
David Miller	Re: [GIT]: Networking

Re: [PATCH] Avoid running lstat(2) on the same cache entry.

Online users