On Tue, Mar 19, 2019 at 12:48:19PM +0000, Måns Rullgård wrote: > Maxime Ripard writes: > > > On Mon, Mar 18, 2019 at 04:23:56PM +0000, Måns Rullgård wrote: > >> Maxime Ripard writes: > >> > >> > On Thu, Mar 14, 2019 at 04:09:13PM +0000, Måns Rullgård wrote: > >> >> Maxime Ripard writes: > >> >> > >> >> > On Mon, Mar 11, 2019 at 04:11:06PM +0000, Måns Rullgård wrote: > >> >> >> Maxime Ripard writes: > >> >> >> > >> >> >> > Hi! > >> >> >> > > >> >> >> > On Mon, Mar 11, 2019 at 01:47:13PM +0000, Mans Rullgard wrote: > >> >> >> >> Sometimes it is desirabled to use a separate i2c controller for ddc > >> >> >> >> access. This adds support for the ddc-i2c-bus property of the > >> >> >> >> hdmi-connector node, using the specified controller if provided. > >> >> >> >> > >> >> >> >> Signed-off-by: Mans Rullgard > >> >> >> >> --- > >> >> >> >> drivers/gpu/drm/sun4i/sun4i_hdmi.h | 1 + > >> >> >> >> drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c | 37 +++++++++++++++++++++++--- > >> >> >> >> 2 files changed, 35 insertions(+), 3 deletions(-) > >> >> >> >> > >> >> >> >> diff --git a/drivers/gpu/drm/sun4i/sun4i_hdmi.h b/drivers/gpu/drm/sun4i/sun4i_hdmi.h > >> >> >> >> index b685ee11623d..b08c4453d47c 100644 > >> >> >> >> --- a/drivers/gpu/drm/sun4i/sun4i_hdmi.h > >> >> >> >> +++ b/drivers/gpu/drm/sun4i/sun4i_hdmi.h > >> >> >> >> @@ -269,6 +269,7 @@ struct sun4i_hdmi { > >> >> >> >> struct clk *tmds_clk; > >> >> >> >> > >> >> >> >> struct i2c_adapter *i2c; > >> >> >> >> + struct i2c_adapter *ddc_i2c; > >> >> >> >> > >> >> >> >> /* Regmap fields for I2C adapter */ > >> >> >> >> struct regmap_field *field_ddc_en; > >> >> >> >> diff --git a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c > >> >> >> >> index 061d2e0d9011..5b2fac79f5d6 100644 > >> >> >> >> --- a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c > >> >> >> >> +++ b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c > >> >> >> >> @@ -212,7 +212,7 @@ static int sun4i_hdmi_get_modes(struct drm_connector *connector) > >> >> >> >> struct edid *edid; > >> >> >> >> int ret; > >> >> >> >> > >> >> >> >> - edid = drm_get_edid(connector, hdmi->i2c); > >> >> >> >> + edid = drm_get_edid(connector, hdmi->ddc_i2c ?: hdmi->i2c); > >> >> >> > > >> >> >> > You can't test whether ddc_i2c is NULL or not... > >> >> >> > > >> >> >> >> if (!edid) > >> >> >> >> return 0; > >> >> >> >> > >> >> >> >> @@ -228,6 +228,28 @@ static int sun4i_hdmi_get_modes(struct drm_connector *connector) > >> >> >> >> return ret; > >> >> >> >> } > >> >> >> >> > >> >> >> >> +static struct i2c_adapter *sun4i_hdmi_get_ddc(struct device *dev) > >> >> >> >> +{ > >> >> >> >> + struct device_node *phandle, *remote; > >> >> >> >> + struct i2c_adapter *ddc; > >> >> >> >> + > >> >> >> >> + remote = of_graph_get_remote_node(dev->of_node, 1, -1); > >> >> >> >> + if (!remote) > >> >> >> >> + return ERR_PTR(-EINVAL); > >> >> >> >> + > >> >> >> >> + phandle = of_parse_phandle(remote, "ddc-i2c-bus", 0); > >> >> >> >> + of_node_put(remote); > >> >> >> >> + if (!phandle) > >> >> >> >> + return NULL; > >> >> >> >> + > >> >> >> >> + ddc = of_get_i2c_adapter_by_node(phandle); > >> >> >> >> + of_node_put(phandle); > >> >> >> >> + if (!ddc) > >> >> >> >> + return ERR_PTR(-EPROBE_DEFER); > >> >> >> >> + > >> >> >> >> + return ddc; > >> >> >> > > >> >> >> > ... Since even in (most) error cases you're returning a !NULL pointer. > >> >> >> > > >> >> >> >> +} > >> >> >> >> + > >> >> >> >> static const struct drm_connector_helper_funcs sun4i_hdmi_connector_helper_funcs = { > >> >> >> >> .get_modes = sun4i_hdmi_get_modes, > >> >> >> >> }; > >> >> >> >> @@ -575,6 +597,12 @@ static int sun4i_hdmi_bind(struct device *dev, struct device *master, > >> >> >> >> goto err_disable_mod_clk; > >> >> >> >> } > >> >> >> >> > >> >> >> >> + hdmi->ddc_i2c = sun4i_hdmi_get_ddc(dev); > >> >> >> >> + if (IS_ERR(hdmi->ddc_i2c)) { > >> >> >> > >> >> >> ... which is checked here. > >> >> >> > >> >> >> The property is optional, so the idea was to return null in that case > >> >> >> and use the built-in controller. If the property exists but some error > >> >> >> occurs, we want to abort rather than proceed with the fallback which > >> >> >> almost certainly won't work. > >> >> >> > >> >> >> Maybe I got something wrong in that logic. > >> >> > > >> >> > Indeed, I just got confused. I guess returning ENODEV in such a case, > >> >> > and testing for that, would make things more obvious. > >> >> > >> >> There's also a case I hadn't thought of: property exists but isn't a > >> >> valid phandle. What do you think is the correct action in that case? > >> > > >> > I think we would have that one covered. of_parse_phandle will return > >> > !NULL, but then of_get_i2c_adapter_by_node will return NULL since we > >> > wouldn't have an associated i2c adapter to the bogus phandle, and you > >> > are checking for that already. > >> > >> of_parse_phandle() doesn't differentiate between a missing property and > >> an existing non-phandle value. The following cases are possible with > >> this patch: > >> > >> - ddc-i2c-bus points to valid i2c controller node: use this for ddc > >> - no ddc-i2c-bus property: return NULL, use internal i2c > >> - ddc-i2c-bus exists but isn't a phandle: likewise > >> - ddc-i2c-bus points to a non-i2c-controller node: EPROBE_DEFER > >> > >> The last two cases obviously mean the devicetree is invalid, so perhaps > >> it doesn't matter much what happens then. I don't think it's possible > >> to distinguish between a well-formed phandle pointing to some bogus node > >> and a good one where the i2c driver hasn't been probed yet. > > > > Ah, I see what you mean now. Yeah, there's not much we can do against > > a wrong / corrupted DT. The DT validation would help prevent the third > > case, and possibly the fourth, but that's basically the only thing we > > can do. > > We need to return -EPROBE_DEFER in the case that everything is fine but > the i2c driver hasn't been probed yet. From here, that is > indistinguishable from of_parse_phandle() returning a completely bogus > node. That's unfortunate, but if we start to not trust the DT content, we have far worse to deal with. > If the ddc-i2c-bus property doesn't contain a phandle at all, we could > either return an error or fall back to the internal i2c. The patch does > the latter because that's less code. I don't think that's any worse > than aborting entirely in terms of user experience. I'm totally fine with the latter behaviour as well. And like I said, the DT validation can help us prevent that case from happening entirely at compilation time. Maxime -- Maxime Ripard, Bootlin Embedded Linux and Kernel engineering https://bootlin.com