Front page | perl.perl5.porters |
Postings from September 2010
Re: All gone
From: David Golden
September 16, 2010 03:33
Re: All gone
Message ID: AANLkTimgxD6jQ89GZtvhMj5K+25wRUrTGqwpgx9Vh9kM@mail.gmail.com
On Thu, Sep 16, 2010 at 6:04 AM, Steffen Mueller <firstname.lastname@example.org> wrote:
> Getting back on topic: Even with Reini's benchmark showing a slowdown, it's
> entirely not clear that overall in actual, real production code, there'd be
> a slowdown or even an improvement in run-time. Previously, we were shown it
> was possible to produce micro-benchmarks showing that the change doesn't
> adversely affect them. Now we have one that shows that there are occasions
> in which the patch does have a net negative effect. We will have to
> investigate more to arrive at a reasonable conclusion. Ideally with more
> realistic code samples.
At the risk of throwing gasoline on a fire, could we start by
profiling these benchmarks on "fast" and "slow" perls and seeing
what's actually eating up the time?
We need to establish whether the slowdowns are due to patches being
unoptimized or doing some unintentional deoptimization or whether the
slowdown is due to being more correct. For example, perl is now more
careful about $@ -- I'm sure that has extra costs somewhere (whether
this benchmark shows it or not) but I wouldn't want to revert it.
I'm very supportive of the idea of creating a more general
benchmarking framework (if it doesn't already exist) that lets people
easily add small scripts to check certain code of interest.