Re: pgsql: Speed up conversion of signed integers to C strings.

On 11/20/2010 01:10 AM, Robert Haas wrote:
> On Fri, Nov 19, 2010 at 11:47 PM, Tom Lane<tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> I wrote:
>>> Robert Haas<rhaas(at)postgresql(dot)org> writes:
>>>> Speed up conversion of signed integers to C strings.
>>> This patch breaks the build here:
>> ... and while I'm looking at it, the added int8 regression test cases
>> definitely merit a WTF. They aren't testing what I would expect.
>> What they are testing is platform-specific behavior of excessive
>> shifting, which is why some of the buildfarm members are pink.
> Ugh. I made an attempt at a fix to both of these issues, but I'm not
> totally sure I got it right, and I'm too tired to stay up any later to
> see what happens. I'll check the BF in the morning.

If you change a test that has alternative result files, you need to
change all the result files. See src/test/regress/resultmap.

In this case you missed out changing int8-exp-three-digits.out - see
<http://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=dawn_bat&dt=2010-11-20%2005%3A30%3A07>

cheers

andrew