On Tue, Oct 27, 2009 at 10:40 PM, Darin Fisher <da...@chromium.org> wrote:
> On Tue, Oct 27, 2009 at 10:35 PM, Adam Barth <aba...@chromium.org> wrote: > >> On Tue, Oct 27, 2009 at 10:27 PM, Darin Fisher <da...@chromium.org> >> wrote: >> > What version of Windows are you using? I find the double-buffering on >> Vista >> > and Win7 to have a big negative impact on performance as compared to >> WinXP. >> > I'm always delighted to run Chrome on my old WinXP laptop. It seems so >> > much faster there. >> >> I lied. I actually have four laptops. So both Vista and XP. >> However, the Vista one has worse specs so I wasn't counting it. >> >> > On X-windows, the renderer backingstores are managed by the X server, >> and >> > the transport DIBs are also managed by the X server. So, we avoid a lot >> of >> > memcpy costs incurred on Windows due to keeping the backingstores in >> main >> > memory there. >> >> We don't draw into a device dependent bitmap on Windows? Is that not >> similar? I was wondering if core IPC latency was lower on Linux. >> That number bleeds into a lot of other times. >> > > We do not. We once did, but DDBs are a very limited resource on Windows. > They get charged against the desktop process, and if you exceed the > seemingly artificial cap, then the system will start having serious > problems. New apps will fail to run properly. No remote desktop for you, > etc. > > So, we switched away from DDBs and just use DIBs. (We use a pixel depth to > match your display--kind of.) > > > >> >> > I suspect this is at least one of the bigger issues. >> > I also suspect that process creation is a problem on Windows. We should >> > probably look into having a spare child process on Windows to minimize >> new >> > tab jank. Maybe there is a bug on this already? >> >> If we're not doing that already, that seems like it might be a big win. >> > > We are most definitely not doing that yet. We could also just move process > creation to a background thread. An unused process might just get swapped > out and be no cheaper to "make live" than it would be to create a new > process. > There's a bug open on this: http://code.google.com/p/chromium/issues/detail?id=6844, but no owner. Any volunteers? :) > > -Darin > > > > --~--~---------~--~----~------------~-------~--~----~ Chromium Developers mailing list: chromium-dev@googlegroups.com View archives, change email options, or unsubscribe: http://groups.google.com/group/chromium-dev -~----------~----~----~----~------~----~------~--~---