Re: [PATCH v4 00/13] New remote-hg helper

From: Felipe Contreras <felipe.contreras@gmail.com>
To: Michael J Gruber <git@drmicha.warpmail.net>
Cc: Johannes Schindelin <Johannes.Schindelin@gmx.de>,
	Jeff King <peff@peff.net>,
	git@vger.kernel.org, Junio C Hamano <gitster@pobox.com>,
	Sverre Rabbelier <srabbelier@gmail.com>,
	Ilari Liusvaara <ilari.liusvaara@elisanet.fi>,
	Daniel Barkalow <barkalow@iabervon.org>
Subject: Re: [PATCH v4 00/13] New remote-hg helper
Date: Thu, 1 Nov 2012 05:08:52 +0100	[thread overview]
Message-ID: <CAMP44s3UHQE69O__EVK29uN_VPdZN=a0-Gczeh-Tbjp1ZAAbJw@mail.gmail.com> (raw)
In-Reply-To: <CAMP44s0KFJW2F3gbO_Xd9QKrZ1OoxvUCvecU084-zH2UDqXKag@mail.gmail.com>

On Wed, Oct 31, 2012 at 8:47 PM, Felipe Contreras
<felipe.contreras@gmail.com> wrote:
> On Wed, Oct 31, 2012 at 7:04 PM, Felipe Contreras
> <felipe.contreras@gmail.com> wrote:
>
>> How about performance?
>
>>      174.038642915 seconds time elapsed
>>
>> Compared to;
>
>>      340.976843750 seconds time elapsed
>>
>> Looks like there's something to improve in this area, but I wouldn't
>> be surprised if the reason for the better performance is that
>> something is not being done. I'll investigate.
>
> Turns out msysgit's remote-hg is not exporting the whole repository,
> that's why it's faster =/

It seems the reason is that it would only export to the point where
the branch is checked out. After updating the to the tip I noticed
there was a performance difference.

I investigated and found two reasons:

1) msysgit's version doesn't export files twice, I've now implemented the same
2) msysgit's version uses a very simple algorithm to find out file changes

This second point causes msysgit to miss some file changes. Using the
same algorithm I get the same performance, but the output is not
correct.

Here's after the latest updates:

 Performance counter stats for 'git clone hg::~/dev/hg-clean hg-clean-5':

     288338.286545 task-clock                #    1.299 CPUs utilized
            46,441 context-switches          #    0.161 K/sec
             6,098 CPU-migrations            #    0.021 K/sec
           509,600 page-faults               #    0.002 M/sec
   518,370,729,897 cycles                    #    1.798 GHz
   204,476,102,906 stalled-cycles-frontend   #   39.45% frontend cycles idle
   <not supported> stalled-cycles-backend
   726,005,034,102 instructions              #    1.40  insns per cycle
                                             #    0.28  stalled cycles per insn
   127,662,400,651 branches                  #  442.752 M/sec
     6,758,976,722 branch-misses             #    5.29% of all branches

     222.020233009 seconds time elapsed

Which is taking roughly 60% more time than msysgit, but the output is
actually correct.

Cheers.

-- 
Felipe Contreras