Perfect. Thanks a lot.
On Fri, Nov 6, 2015 at 1:28 AM, Xiang, Haihao <haihao.xi...@intel.com> wrote: > > Hi Perry, > > The current X11 connection on your system works with your NVIDIA GPU, > you can use VAAPI on drm mode with your Intel GPU. I sent a patch to > this list to add an option to vainfo tool, you can use 'vainfo -- > display drm' to show available profile/entrypoint pairs with i965 > driver. > > Thanks > Haihao > > > >> Hello, >> >> Would anyone know a more appropriate place to ask? >> >> I poked around on #intel-gfx and was told most of the VAAPI guys >> frequent this list. >> >> many thanks. >> >> On Tue, Nov 3, 2015 at 5:42 PM, Perry Hung <pe...@leaflabs.com> >> wrote: >> > I spent some more time playing with this and it appears that DRI2 >> > is >> > enabled on the Intel driver. I have marked the driver as inactive >> > on >> > Layout0. It appears to me that everything is configured correctly >> > and >> > that DRI is enabled. I have atteched a copy of xorg log. >> > >> > Any idea what this assertion means? >> > >> > >> > On Tue, Nov 3, 2015 at 4:38 PM, Perry Hung <pe...@leaflabs.com> >> > wrote: >> > > Hi, >> > > >> > > I'm trying to use VAAPI for HEVC and h264 decode on a Devil's >> > > Canyon >> > > system running the latest 4.2.5 kernel running Arch Linux. >> > > >> > > There is an NVIDIA GPU driving 3 monitors and doing some compute, >> > > and >> > > I'd like to simultaneously use Intel quicksync. The integrated >> > > GPU has >> > > no monitors connected to it, but as far as I can tell has kernel >> > > support, and the i915 module is loaded: >> > > >> > > $d mesg | grep i915 >> > > [ 6.869230] i915 0000:00:02.0: enabling device (0000 -> 0003) >> > > [ 6.937495] snd_hda_intel 0000:00:03.0: bound 0000:00:02.0 >> > > (ops >> > > i915_audio_component_bind_ops [i915]) >> > > [ 6.937499] [drm] Initialized i915 1.6.0 20150522 for >> > > 0000:00:02.0 on minor 0 >> > > [ 6.948090] i915 0000:00:02.0: No connectors reported >> > > connected with modes >> > > [ 6.948336] i915 0000:00:02.0: fb1: inteldrmfb frame buffer >> > > device >> > > [ 6.948338] i915 0000:00:02.0: registered panic notifier >> > > >> > > >> > > $ lsmod | grep i915 >> > > i915 1056768 2 >> > > drm_kms_helper 102400 1 i915 >> > > intel_gtt 20480 1 i915 >> > > i2c_algo_bit 16384 1 i915 >> > > drm 286720 7 i915,drm_kms_helper,nvidia >> > > video 32768 1 i915 >> > > button 16384 1 i915 >> > > >> > > >> > > When I attempt to check the vaapi capabilities of the Intel >> > > driver >> > > with vainfo, I get the following assertion: >> > > >> > > $ LIBVA_DRIVER_NAME=i965 vainfo >> > > libva info: VA-API version 0.38.0 >> > > libva info: va_getDriverName() returns 0 >> > > libva info: User requested driver 'i965' >> > > libva info: Trying to open /usr/lib/dri/i965_drv_video.so >> > > libva info: Found init function __vaDriverInit_0_38 >> > > vainfo: intel_driver.c:100: intel_driver_init: Assertion >> > > `(((struct >> > > drm_state *)(ctx)->drm_state)->auth_type == (VA_DRM_AUTH_DRI1)) >> > > || >> > > (((struct drm_state *)(ctx)->drm_state)->auth_type == >> > > (VA_DRM_AUTH_DRI2)) || (((struct drm_state >> > > *)(ctx)->drm_state)->auth_type == (VA_DRM_AUTH_CUSTOM))' failed. >> > > zsh: abort (core dumped) LIBVA_DRIVER_NAME=i965 vainfo >> > > >> > > My question is, is this a supported configuration? Do I need to >> > > reconfigure xorg, or connect the iGPU up to a screen? >> _______________________________________________ >> Libva mailing list >> Libva@lists.freedesktop.org >> http://lists.freedesktop.org/mailman/listinfo/libva _______________________________________________ Libva mailing list Libva@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libva
