On Tue, Jun 23, 2015 at 09:58:37PM +0100, Russell King - ARM Linux wrote: > On Tue, Jun 23, 2015 at 05:48:51PM -0300, Fabio Estevam wrote: > > On Tue, Jun 23, 2015 at 5:41 PM, Russell King - ARM Linux > > <linux@xxxxxxxxxxxxxxxx> wrote: > > > > > I've just tried this (the HBi1 was booted previously with the HDMI socket > > > disconnected). Just now, I turned the TV on, and then connected it to > > > the HDMI. > > > > Thanks for testing it. > > > > > The TV reported a resolution of 1024x768, and the kernel log was silent. > > > > That's the point that makes me confused: shouldn't it report a higher > > resolution like 1080p? > > No - it won't go higher than the initial mode that was set. See > drm_fb_helper_hotplug_event(): > > max_width = fb_helper->fb->width; > max_height = fb_helper->fb->height; > > drm_fb_helper_probe_connector_modes(fb_helper, max_width, max_height); > > This calls connector->funcs->fill_modes with the max width/height, > which calls down into drm_helper_probe_single_connector_modes_merge_bits() > and there's limitations in there which means that we won't try to > increase the size of the framebuffer. > > You do get that behaviour if you have the X server running, and that's > all down to the X server itself doing that (reacting to the hotplug > event which DRM generates.) The initial fb allocated for fbdev emulation does indeed restrict what fbdev/con will show. But it should in now way affect the mode list (that's only the fbdev code doing the filtering there) or X. You should be able to list/set 1080p. Might be useful to figure out with drm.debug whether the mode was kicked out for some driver/hw restriction reason, that should get logged. -Daniel -- Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/dri-devel
