Re: Tuplesort merge pre-reading

On 09/12/2016 06:47 PM, Claudio Freire wrote:
> On Mon, Sep 12, 2016 at 12:02 PM, Claudio Freire <klaussfreire(at)gmail(dot)com> wrote:
>> On Sun, Sep 11, 2016 at 12:47 PM, Heikki Linnakangas <hlinnaka(at)iki(dot)fi> wrote:
>>> Here's a new version of these patches, rebased over current master. I
>>> squashed the two patches into one, there's not much point to keep them
>>> separate.
>>
>> I don't know what was up with the other ones, but this one works fine.
>>
>> Benchmarking now.
>
> I spoke too soon, git AM had failed and I didn't notice.
>
> regression.diffs attached
>
> Built with
>
> ./configure --enable-debug --enable-cassert && make clean && make -j7
> && make check

Ah, of course! I had been building without assertions, as I was doing
performance testing. With --enable-cassert, it failed for me as well
(and there was even a compiler warning pointing out one of the issues).
Sorry about that.

Here's a fixed version. I'll go through Peter's comments and address
those, but I don't think there was anything there that should affect
performance much, so I think you can proceed with your benchmarking with
this version. (You'll also need to turn off assertions for that!)

- Heikki