[updated patch] malloc per-thread cache ready for review

[DJ Delorie]

Will Newton <will.newton@gmail.com> writes:
> It would be good to get wider testing of the performance on
> performance critical workloads that other people may have.

I've been collecting workloads for exactly that reason, using the tools
in the dj/malloc branch.  With those, I can capture a site-specific
application's calls and repeat them elsewhere, in a generic environment.
My intention there is to create a corpus of such simulations, in order
to "represent" all these performance-critical apps during future malloc
work.

Now, if people would only send them to me, or at least tell me what
their performance critical workloads are ;-)