in reply to Re^13: Avoid Locking Entire Hashes
in thread Avoid Locking Entire Hashes

I meant anything that has specifically stood out for you.

With regard to the disparity between our timings, they aren't down to 5.14 per se.

They are noticeably quicker on 5.14, but the delta remains the same (for me):

C:\test>\perl5.14-64\bin\perl.exe 909437-1.pl
Lock hash took 0.904 seconds

C:\test>\perl5.14-64\bin\perl.exe 909437-2.pl
Lock scalar ref took 2.482 seconds

C:\test>\perl5.14-64\bin\perl.exe 909437-3.pl
Lock scalars ref took 2.951 seconds

[download]

Are you using AS, or other?

Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
"Science is about questioning the status quo. Questioning authority".
In the absence of evidence, opinion is indistinguishable from prejudice.

Replies are listed 'Best First'.
Re^15: Avoid Locking Entire Hashes
by ikegami (Patriarch) on Jun 15, 2011 at 20:41 UTC

    I haven't really gone over the list of changes.

    I'm using ActivePerl build 1400 for x86.

    My CPU is 2 years newer. And while the clock is slower, it has a "turbo" frequency of 2.8 GHz (whatever that means) which is faster. (Presumably, it speeds up when needed, saving power the rest of the time. This is a laptop CPU. M = Mobile)

    Core i7 720QM
    Core 2 Q6600

[reply]
      And while the clock is slower, it has a "turbo" frequency of 2.8 GHz

      Hm. Still, if this was down to just a faster cpu then I would have expected both methods to have been sped up roughly in equal proportion.

      The last set of numbers I posted were also for AS 1400, though 64-bit. I still am not seeing causality for the indirected scalar locking to be so expensive on my system.

      Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
      "Science is about questioning the status quo. Questioning authority".
      In the absence of evidence, opinion is indistinguishable from prejudice.
[reply]

In Section Seekers of Perl Wisdom