On Tue, May 31, 2011 at 4:25 PM, Ingo Molnar <mingo@xxxxxxx> wrote: > > * Sasha Levin <levinsasha928@xxxxxxxxx> wrote: > >> On Tue, 2011-05-31 at 15:09 +0200, Ingo Molnar wrote: >> > * Sasha Levin <levinsasha928@xxxxxxxxx> wrote: >> > >> > > I've started working on converting our MMIO code to use RCU rbtree. >> > >> > Well, why would we want to switch the MMIO code away from the brlock >> > right now? mmio tree reconfigurations are very, very rare. >> > >> > Won't something like the qcow cache be more suitable for that? That >> > has both frequent read and write activities. >> >> We don't have a qcow cache at the moment :) > > Wasnt one in the works by Prasad? I have the code changes which work. I think I had send patches few days back as well there were few improvements. The only reason I did not send the v2 of the patch is, I am not seeing performance improvement by adding cache. May be the DD of 1G file was a wrong test to calculate the performance. Sasha had asked me to run boniee++ for performance numbers. I will ensure the patches can be applied on the latest code base and send the patches. Sorry for the delay. Thanks and Regards, Prasad > >> It was either the MMIO or the ioport tree. We don't have to pull >> them into our master tree either - it's mostly a proof of concept >> I'm doing on a separate tree. > > Sure. > > Thanks, > > Ingo > -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
