Re: [[RFC][PATCH] v1 0/2] malloc/realloc with transparent huge page support

[Norbert Manthey <nmanthey@conp-solutions.com>]

On 04.05.20 23:07, DJ Delorie wrote:

> Norbert Manthey <nmanthey@conp-solutions.com> writes:
>> I have been looking for a way to enable huge pages in a user land process
>> without privileges, and without modifying the source of the process.
> In addition to the other feedback you've gotten, I'll point out some
> administrivia...
>
> * Non-trivial patches require a copyright assignment before we can act
>   on them (really, before we should even look at them).  The process is
>   easy and we can help you though it, but it does take time.
I understand, and am willing to move forward with this. I reached out to
assign@gnu.org.
>
> * Patches should be against current master and use the latest
>   conventions (like tunables) - at least, by the time you want them
>   included.  2.23 is "ancient" for us and a lot has changed in the
>   malloc code since then.  Your findings from that version may not be
>   the same as findings against master.

I understand. I will follow the provided feedback (migrate to tunables;
use a macro for the huge page size to be used), and can test whether the
speedup obtained for version 2.23 is comparable to what we would get
when starting from the current HEAD commit

Best,
Norbert