Re: Slowdown around 5.0.5-dev-0

[Sebastian Gniazdowski <sgniazdowski@xxxxxxxxx>]

On 24 October 2015 at 08:09, Sebastian Gniazdowski
<sgniazdowski@xxxxxxxxx> wrote:
> On 24 October 2015 at 01:50, Bart Schaefer <schaefer@xxxxxxxxxxxxxxxx> wrote:
>> On Oct 23, 12:26pm, Bart Schaefer wrote:
>> }
>> } Try the below patch against 8e9a68ad1 ?
>>
>> However, if the results can be considered reliable anyway, note that
>> 36926 retains most of the search_test speed, which I think is an
>> improvement over backing out all of 36834 (Sebastian's results for
>> zsh-newheaps-zhalloc, if I'm reading his email correctly).
>
> I ran quick tests yesterday and obtained this too. Good news that not
> all gains of 36834 will be withdrawn! :) The newheaps-zhalloc is what
> you write, two following patches without 36834.

I misread this. What I wanted to state, is that your patch from
yesterday retained search_test speed.

The fact that 36836 has its contribution in speeding up search_test is
was described by me in previous email:

> newheaps is responsible for
> instant-responsivity of my script, 36836 for much faster searching –
> that's a good compromise

So the thing is that your patch from yesterday retains search_test
speed up of 36834, while fixing the memory issues of 36834. Together
with 36836, this gives full retain of search_test speedup, nothing is
lost, speed of pattern matching is as high as with no revert from
current head state. What's lost is string_test speedup 36834 gave,
however, it did so with cost of increased memory usage

Best regards,
Sebastian Gniazdowski