performance regression in git diffs

[Wagner Bruna]

Hello,

Steve Borho wrote:
> On Thu, Jun 10, 2010 at 4:22 PM, Benoit Boissinot <bboissin at gmail.com> wrote:
>> On Thu, Jun 10, 2010 at 11:11 PM, Steve Borho <steve at borho.org> wrote:
>>> I'm afraid I don't have any accurate dates on when this started to
>>> become a problem, but people have been reporting a drastic performance
>>> regression in TortoiseHg when viewing large repositories.
>>>
>>> [snip]
>>> It seems your repository needs a number of binary files in order to
>>> really expose the problem.
>>>
>> Since you can reproduce it, any chance you could bisect it? Shouldn't take long.
> 
> Bisect points at:
> 
> changeset:   10262:eb243551cbd8
> branch:      stable
> user:        Matt Mackall <mpm at selenic.com>
> date:        Tue Jan 19 22:20:05 2010 -0600
> summary:     copies: speed up copy detection

That changeset sounds familiar...

http://mercurial.selenic.com/bts/issue1994

http://mercurial.selenic.com/bts/issue2089

BTW, I remember mpm talking about reworking the copy detection logic
for 1.6 at the time, but I don't know if he could make progress on
that front.

Regards,
Wagner