develooper Front page | perl.perl5.porters | Postings from June 2016

Re: Inefficient stat-ing of files?

Thread Previous | Thread Next
From:
Sawyer X
Date:
June 29, 2016 21:04
Subject:
Re: Inefficient stat-ing of files?
Message ID:
577437B3.3060500@gmail.com


On 06/29/2016 09:51 PM, Craig A. Berry wrote:
> On Wed, Jun 29, 2016 at 2:24 PM, Sawyer X <xsawyerx@gmail.com> wrote:
>> On 06/24/2016 07:05 PM, Zefram wrote:
>>> Craig A. Berry wrote:
>>>
>>>> It looks like we could cut the number of stat calls when loading
>>>> modules in about half by removing the checks for .pmc files.
>>> That's a viable suggestion.  It would be a visible behaviour change,
>>> not a pure optimisation, but likely to be quite acceptable.
>> Perhaps we should revisit the idea?
> As bulk88 pointed out up-thread, you can already do this by
> configuring with -DPERL_DISABLE_PMC, so the question would be whether
> that became the default.  As Dave pointed out, there is no measurable
> benefit to skipping .pmc lookups, at least on Linux.  On systems with
> more expensive stat calls, it's possible it could make a difference,
> in which case we'd need to evaluate the risk of changing the default
> behavior.  Somebody somewhere has probably used it as a generic
> override mechanism rather than for its intended purpose.

With the insignificance of the stat() commands, I'd rather keep .pmc
loading the default for now.

>
>> Having .pmc files that work could be a much greater optimization...
> My impression is that's just not gonna happen, but I'd be happy to be
> proven wrong.

When you wish upon a star... :)

Thread Previous | Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About