Re: COPY_TWO_BYTES and C strict aliasing

Philip Martin <philip.martin_at_wandisco.com> writes:

> Is COPY_TWO_BYTES a significant optimisation?

r1325896 implies this is most significant for ra_svn. So I compared the
testsuite over ra_svn built with the "fast" code and the "slow" code.
The CPU used by the testsuite and svnserve was:

   "fast" code

   749.93user 404.98system 9:09.25elapsed testsuite
   0:32.43 svnserve

   742.84user 400.00system 8:59.75elapsed testsuite
   0:32.70 svnserve

   739.51user 400.87system 8:58.91elapsed testsuite
   0:32.66 svnserve

   "slow" code

   745.06user 402.36system 9:04.21elapsed testsuite
   0:32.73 svnserve

   739.94user 401.86system 8:57.13elapsed testsuite
   0:32.58 svnserve

On my Linux box this optimisation isn't significant. Perhaps it makes a
difference on some other platform or with some other test scenario, but
at the moment it's not clear why we would want with this code.

-- 
uberSVN: Apache Subversion Made Easy
http://www.uberSVN.com