From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ville =?iso-8859-1?Q?Syrj=E4l=E4?= Subject: Re: [PATCH v3] drm: Only create a cmdline mode if no probed modes match Date: Wed, 1 Jun 2016 12:56:54 +0300 Message-ID: <20160601095654.GD4329@intel.com> References: <1429532795-3922-1-git-send-email-chris@chris-wilson.co.uk> <1464773676-20152-1-git-send-email-chris@chris-wilson.co.uk> <20160601094353.GC4329@intel.com> <20160601094751.GD10319@nuc-i3427.alporthouse.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Content-Disposition: inline In-Reply-To: <20160601094751.GD10319@nuc-i3427.alporthouse.com> Sender: stable-owner@vger.kernel.org To: Chris Wilson , dri-devel@lists.freedesktop.org, Radek =?iso-8859-1?Q?Dost=E1l?= , Jesse Barnes , Daniel Vetter , Julia Lemire , Dave Airlie , stable@vger.kernel.org List-Id: dri-devel@lists.freedesktop.org On Wed, Jun 01, 2016 at 10:47:51AM +0100, Chris Wilson wrote: > On Wed, Jun 01, 2016 at 12:43:53PM +0300, Ville Syrj=E4l=E4 wrote: > > On Wed, Jun 01, 2016 at 10:34:36AM +0100, Chris Wilson wrote: > > > The intention of using video=3D: is primarily to= select > > > the user's preferred resolution at startup. Currently we always c= reate a > > > new mode irrespective of whether the monitor has a native mode at= the > > > desired resolution. This has the issue that we may then select th= e fake > > > mode rather the native mode during fb_helper->inital_config() and= so > > > if the fake mode is invalid we then end up with a loss of signal.= Oops. > > > This invalid fake mode would also be exported to userspace, who > > > potentially may make the same mistake. > > >=20 > > > To avoid this issue, we filter out the added command line mode if= we > > > detect the desired resolution (and clock if specified) amongst th= e > > > probed modes. This fixes the immediate problem of adding a duplic= ate > > > mode, but perhaps more generically we should avoid adding a GTF m= ode if > > > the monitor has an EDID that is not GTF-compatible, or similarly = for > > > CVT. > > >=20 > > > Fixes regression from > > >=20 > > > commit eaf99c749d43ae74ac7ffece5512f3c73f01dfd2 > > > Author: Chris Wilson > > > Date: Wed Aug 6 10:08:32 2014 +0200 > > >=20 > > > drm: Perform cmdline mode parsing during connector initialisa= tion > > >=20 > > > that breaks HDMI output on BeagleBone Black with LG TV (model 19L= S4R-ZA). > > >=20 > > > v2: Explicitly delete our earlier cmdline mode > > > v3: Mode pruning should now be sufficient to delete stale cmdline= modes > > >=20 > > > Reported-by: Radek Dost=E1l > > > Signed-off-by: Chris Wilson > > > Cc: Radek Dost=E1l > > > Cc: Jesse Barnes > > > Cc: Ville Syrj=E4l=E4 > > > Cc: Daniel Vetter > > > Cc: dri-devel@lists.freedesktop.org > > > Cc: Julia Lemire > > > Cc: Dave Airlie > > > Cc: stable@vger.kernel.org > > > --- > > > drivers/gpu/drm/drm_probe_helper.c | 19 +++++++++++++++++-- > > > 1 file changed, 17 insertions(+), 2 deletions(-) > > >=20 > > > diff --git a/drivers/gpu/drm/drm_probe_helper.c b/drivers/gpu/drm= /drm_probe_helper.c > > > index 0329080d7f7c..a705ed12c062 100644 > > > --- a/drivers/gpu/drm/drm_probe_helper.c > > > +++ b/drivers/gpu/drm/drm_probe_helper.c > > > @@ -82,13 +82,28 @@ drm_mode_validate_flag(const struct drm_displ= ay_mode *mode, > > > =20 > > > static int drm_helper_probe_add_cmdline_mode(struct drm_connecto= r *connector) > > > { > > > + struct drm_cmdline_mode *cmdline_mode; > > > struct drm_display_mode *mode; > > > =20 > > > - if (!connector->cmdline_mode.specified) > > > + cmdline_mode =3D &connector->cmdline_mode; > > > + if (!cmdline_mode->specified) > > > return 0; > > > =20 > > > + /* Only add a GTF mode if we find no matching probed modes */ > > > + list_for_each_entry(mode, &connector->probed_modes, head) { > > > + if (mode->hdisplay !=3D cmdline_mode->xres || > > > + mode->vdisplay !=3D cmdline_mode->yres) > > > + continue; > > > + > > > + if (cmdline_mode->refresh_specified && > > > + mode->vrefresh !=3D cmdline_mode->refresh) > >=20 > > I think we might not have .vrefresh populated for the probed modes. > > We update .vrefresh only for the modes left on the real mode list i= n the > > end. >=20 > Or drm_mode_vrefresh() ? That should work. --=20 Ville Syrj=E4l=E4 Intel OTC