Re: Drastic jump in the time required for the test suite

From: Duy Nguyen <pclouds@gmail.com>
Date: Oct 21, 2016, 11:03 UTC
Message-ID: <CACsJy8DDO_k-gtSj-Nbc9=DvTC1FFQRmc=OtTmh-c4rc=xC_+w@mail.gmail.com>
In-Reply-To: <1476996017.28685.10.camel@kaarsemaker.net>

On Fri, Oct 21, 2016 at 3:40 AM, Dennis Kaarsemaker <dennis@kaarsemaker.net> wrote:

Show 11 quoted lines

> On Thu, 2016-10-20 at 08:31 -0400, Jeff King wrote:
>
>> I'm also not entirely convinced that the test suite being a shell script
>> is the main culprit for its slowness. We run git a lot of times, and
>> that's inherent in testing it. I ran the whole test suite under
>> "strace -f -e execve". There are ~335K execs. Here's the breakdown of
>> the top ones:
>
> You're measuring execve's, but fork (well, fork emulation. There's no
> actual fork) is also expensive on windows iirc, so subshells add a lot
> to this cost.

shells fork on piping as well, and redirection and some other construct if I remember correctly (I attempted to port busybox ash to windows and had to find and "fix" all the forks)

> That said, strace -eclone says that a 'make test' forks
> ~408k times, and while this is significantly more than the amount of
> execs in your example, this does include cvs and svn tests and it's
> still in the same ballpark.

-- 
Duy