Speed up trace mode

[Mihir Luthra]

Hi Clemens,

I have added the functionality of utilising the wasted memory in shared
memory and as per the tests that I made till now, it works fine and drops
memory usage to half. [1]
Although I haven’t updated readme to contain info about this new feature.
This is more of a highly separate module of the code which I believe I have
explained mostly by comments.

Also, I have commented the code to contain almost each and every detail.
To make it easier, I will try making it in a presentation form instead of a
readme and attach that link in the readme.

I still have to add:
1) reference counting to unmap old mapping after expanding memory.
2) pthread thread local storage to store manager
3) Currently code can only insert and search for strings with ascii
letters, should I try to make it for unicode?

I think I should add these 3 once the existing code has been reviewed. Or
should I continue to add?

[1] https://github.com/MihirLuthra/sharedMemory

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.macports.org/pipermail/macports-dev/attachments/20190521/de472cbd/attachment.html>