On Tue, 31 Mar 2009 12:39:17 -0700 Eric Anholt <e...@anholt.net> wrote:
> On Fri, 2009-03-27 at 15:48 -0700, Jesse Barnes wrote: > > On Fri, 27 Mar 2009 22:53:00 +0100 > > Jakob Bornecrantz <wallbra...@gmail.com> wrote: > > > > > On Fri, Mar 27, 2009 at 8:58 PM, Jesse Barnes > > > <jbar...@virtuousgeek.org> wrote: > > > > This patch speeds up drmModeGetConnector by pre-allocating mode > > > > & property info space before calling into the kernel. In many > > > > cases this pre-allocation will be sufficient to hold the > > > > returned values (it's easy enough to tweak if the common case > > > > becomes larger), which means we don't have to make the second > > > > call, which saves a lot of time. > > > > > > > > Any comments or problems with the patch? > > > > > > Looks good, I do wonder how much time do we save on doing pre > > > allocation, just curious? Anyways the patch is > > > Acked-by: Jakob Bornecrantz <wallbra...@gmail.com> > > > > > > Cheers Jakob. > > > > Some of my testing showed it took about .3s for a > > drmModeGetConnector call and my testing showed this decrease a lot > > with the pre-allocation (conveniently lost the test results > > though). Theoretically it should be about half the cost with > > pre-allocation, which adds up if you multiply by the number of > > outputs. > > Seems like the ioctl needs a flag for "did the user want me to go > reprobe things?" like we did in the X Server. Yeah or a new ioctl... we'll need a new libdrm function for that at the very least for compatibility. -- Jesse Barnes, Intel Open Source Technology Center ------------------------------------------------------------------------------ -- _______________________________________________ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel
