This is the mail archive of the
mailing list for the glibc project.
Re: TSX lock elision for glibc v2
On Fri, 2013-01-11 at 18:41 +0100, Andi Kleen wrote:
> > Have you started to look at potential candidates in detail yet?
> I considered malloc at some point, but it's not clear if there is
> a lot of potential even. The per thread malloc that is now in
> newer releases seems to do reasonably well on its own.
What about the internal locks used in the condvar and rwlock
implementations? We might perhaps be able to build better concurrent
algorithms for these first that do better than the internal locks, but
until somebody actually tries that, HLE could get rid of some of the
overheads I'd guess.