Re: why is imp_xxh stored in an SV?

On Wed, Apr 18, 2012 at 04:35:07PM +0100, Dave Mitchell wrote:
> > Note that this and the just-submitted DBIS patch represent the end of
> > my work on improving performance on DBI, (not counting any remedial work
> > that may be required).
> 
> PS - here are some final timings, comparing my
>     'while ($sth->fetch()) {$c++}'
> test loop (with unchanged DBD-mysql-4.020), against
> (a) DBI r15098, just before the start of my changes;
> (b) DBI r15266 plus my recent DBIS and mg_ptr tweaks:
> 
> Times are in CPU seconds, lower is better:
> 
> 	         unthreaded        |      threaded 
>             5.8.1   5.8.9  5.15.6  |  5.8.1   5.8.9  5.15.7
> 	    -----   -----  ------ -+  -----   -----  ------
> (a)         15.28   14.84   15.11  |  42.86   48.92   47.72
> (b)         13.01   12.44   12.68  |  27.81   30.19   17.67
> 
> 
> Note especially how on modern threaded perls (i.e. >= 5.10), the timing
> is approaching that of unthreaded. (I have a separate patch pending
> against DBD-mysql which brings that 17.67 down by another 2.5 seconds)
> 
> All very satisfying :-)

Indeed!  Great work.  Many thanks again Dave.

Tim.