On Mon, Jan 28, 2019 at 11:53:38AM +0100, Andrzej Hajda wrote:
> On 25.01.2019 02:33, Ronald Tschalär wrote:
> > commit d6abe6df706c (drm/bridge: sil_sii8620: do not have a dependency
> > of RC_CORE) changed the driver to select both RC_CORE and INPUT.
> > However, this causes problems with other drivers, in particular an input
> > driver that depends on MFD_INTEL_LPSS_PCI (to be added in a separate
> > commit):
> >
> > drivers/clk/Kconfig:9:error: recursive dependency detected!
> > drivers/clk/Kconfig:9: symbol COMMON_CLK is selected by MFD_INTEL_LPSS
> > drivers/mfd/Kconfig:566: symbol MFD_INTEL_LPSS is selected by MFD_INTEL_LPSS_PCI
> > drivers/mfd/Kconfig:580: symbol MFD_INTEL_LPSS_PCI is implied by KEYBOARD_APPLESPI
> > drivers/input/keyboard/Kconfig:73: symbol KEYBOARD_APPLESPI depends on INPUT
> > drivers/input/Kconfig:8: symbol INPUT is selected by DRM_SIL_SII8620
> > drivers/gpu/drm/bridge/Kconfig:83: symbol DRM_SIL_SII8620 depends on DRM_BRIDGE
> > drivers/gpu/drm/bridge/Kconfig:1: symbol DRM_BRIDGE is selected by DRM_PL111
> > drivers/gpu/drm/pl111/Kconfig:1: symbol DRM_PL111 depends on COMMON_CLK
> >
> > According to the docs and general consensus, select should only be used
> > for non user-visible symbols, but both RC_CORE and INPUT are
> > user-visible. Furthermore almost all other references to INPUT
> > throughout the kernel config are depends, not selects. For this reason
> > the first part of this change reverts commit d6abe6df706c.
> >
> > In order to address the original reason for commit d6abe6df706c, namely
> > that not all boards use the remote controller functionality and hence
> > should not need have to deal with RC_CORE, the second part of this
> > change now makes the remote control support in the driver optional and
> > contingent on RC_CORE being defined. And with this the hard dependency
> > on INPUT also goes away as that is only needed if RC_CORE is defined
> > (which in turn already depends on INPUT).
>
>
> Thanks for fixing it, this seems to be a best way to deal with it, more
> comments below.
Sorry I didn't respond to this earlier, but since I wasn't on the
to/cc list and I neglected to check the archives I completely missed
it.
> > CC: Inki Dae <inki.dae@xxxxxxxxxxx>
> > CC: Andrzej Hajda <a.hajda@xxxxxxxxxxx>
> > CC: Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx>
> > CC: Dmitry Torokhov <dmitry.torokhov@xxxxxxxxx>
> > Signed-off-by: Ronald Tschalär <ronald@xxxxxxxxxxxxx>
> > ---
> > Resending this, as I somehow managed to forget to cc dri-devel.
> > Apologies for the duplication.
> >
> > Changes in v2:
> > - completely remove dependencies on both RC_CORE and INPUT in Kconfig,
> > - make remote control functionality in driver contingent on RC_CORE
> > being defined
> >
> > drivers/gpu/drm/bridge/Kconfig | 2 --
> > drivers/gpu/drm/bridge/sil-sii8620.c | 17 +++++++++++++++++
> > 2 files changed, 17 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/bridge/Kconfig b/drivers/gpu/drm/bridge/Kconfig
> > index 2fee47b0d50b..a11198a36005 100644
> > --- a/drivers/gpu/drm/bridge/Kconfig
> > +++ b/drivers/gpu/drm/bridge/Kconfig
> > @@ -85,8 +85,6 @@ config DRM_SIL_SII8620
> > depends on OF
> > select DRM_KMS_HELPER
> > imply EXTCON
> > - select INPUT
> > - select RC_CORE
>
> Shouldn't you put here "imply RC_CORE"? To avoid RC_CORE as module and
> sii8620 built-in.
Good point - fixed.
> > help
> > Silicon Image SII8620 HDMI/MHL bridge chip driver.
> >
> > diff --git a/drivers/gpu/drm/bridge/sil-sii8620.c b/drivers/gpu/drm/bridge/sil-sii8620.c
> > index 0cc293a6ac24..dee47752791e 100644
> > --- a/drivers/gpu/drm/bridge/sil-sii8620.c
> > +++ b/drivers/gpu/drm/bridge/sil-sii8620.c
> > @@ -66,7 +66,9 @@ enum sii8620_mt_state {
> > struct sii8620 {
> > struct drm_bridge bridge;
> > struct device *dev;
> > +#if IS_ENABLED(CONFIG_RC_CORE)
> > struct rc_dev *rc_dev;
> > +#endif
> > struct clk *clk_xtal;
> > struct gpio_desc *gpio_reset;
> > struct gpio_desc *gpio_int;
> > @@ -1756,6 +1758,7 @@ static void sii8620_send_features(struct sii8620 *ctx)
> > sii8620_write_buf(ctx, REG_MDT_XMIT_WRITE_PORT, buf, ARRAY_SIZE(buf));
> > }
> >
> > +#if IS_ENABLED(CONFIG_RC_CORE)
> > static bool sii8620_rcp_consume(struct sii8620 *ctx, u8 scancode)
> > {
> > bool pressed = !(scancode & MHL_RCP_KEY_RELEASED_MASK);
> > @@ -1774,6 +1777,12 @@ static bool sii8620_rcp_consume(struct sii8620 *ctx, u8 scancode)
> >
> > return true;
> > }
> > +#else
> > +static bool sii8620_rcp_consume(struct sii8620 *ctx, u8 scancode)
> > +{
> > + return false;
> > +}
> > +#endif
> >
> > static void sii8620_msc_mr_set_int(struct sii8620 *ctx)
> > {
> > @@ -2097,6 +2106,7 @@ static void sii8620_cable_in(struct sii8620 *ctx)
> > enable_irq(to_i2c_client(ctx->dev)->irq);
> > }
> >
> > +#if IS_ENABLED(CONFIG_RC_CORE)
> > static void sii8620_init_rcp_input_dev(struct sii8620 *ctx)
> > {
> > struct rc_dev *rc_dev;
> > @@ -2126,6 +2136,11 @@ static void sii8620_init_rcp_input_dev(struct sii8620 *ctx)
> > }
> > ctx->rc_dev = rc_dev;
> > }
> > +#else
> > +static void sii8620_init_rcp_input_dev(struct sii8620 *ctx)
> > +{
> > +}
> > +#endif
> >
> > static void sii8620_cable_out(struct sii8620 *ctx)
> > {
> > @@ -2214,9 +2229,11 @@ static int sii8620_attach(struct drm_bridge *bridge)
> >
> > static void sii8620_detach(struct drm_bridge *bridge)
> > {
> > +#if IS_ENABLED(CONFIG_RC_CORE)
> > struct sii8620 *ctx = bridge_to_sii8620(bridge);
> >
> > rc_unregister_device(ctx->rc_dev);
> > +#endif
> > }
>
> In two cases you create stub functions, in the third you put ifdefs
> inside function, some lack of consistency.
Agreed - I'll make a stub for sii8620_detach() too.
> I wonder if it wouldn't be better to create stubs just for rc_core
> functions, the best would be to put stubs into rc-core.h, but if the
> case of 'optional rc-core' is too rare you can put stubs into the driver:
>
> diff --git a/drivers/gpu/drm/bridge/sil-sii8620.c
> b/drivers/gpu/drm/bridge/sil-sii8620.c
>
> index a6e8f4591e63..6ca838d30f93 100644
> --- a/drivers/gpu/drm/bridge/sil-sii8620.c
> +++ b/drivers/gpu/drm/bridge/sil-sii8620.c
> @@ -35,6 +35,13 @@
>
> #include "sil-sii8620.h"
>
> +#if !IS_ENABLED(CONFIG_RC_CORE)
> +#define rc_unregister_device(dev) ({ (void)(dev); })
> +#define rc_allocate_device(type) ({ NULL; })
> +#define rc_keydown(...) ({ })
> +#define rc_keyup(...) ({ })
> +#endif
> +
> #define SII8620_BURST_BUF_LEN 288
> #define VAL_RX_HDMI_CTRL2_DEFVAL VAL_RX_HDMI_CTRL2_IDLE_CNT(3)
The reason I didn't do this was because it leads to unwanted log
messages: in particular in sii8620_init_rcp_input_dev() the failed
allocation results in an ugly dev_err() message; in the case of
sii8620_rcp_consume() it's a much more benign dev_dbg() though.
But if stubs for rc-core is the preferred approach, I'll certainly
do that instead.
As to how rare 'optional rc-core' is, the only other place I've found
is in drivers/media/usb/dvb-usb-v2/, and there the approach is to #if
out the whole rc related code and stub out the single entry point to
it (get_rc_config). So for now it looks like use for rc-core stubs
would be restricted to this module (if we go that route).
Cheers,
Ronald
