Re: [PATCH v4 6/9] drm: rcar-du: crtc: Enable and disable CMMs

[Laurent Pinchart]

Hi Kieran,

On Thu, Sep 19, 2019 at 09:08:18AM +0100, Kieran Bingham wrote:
> On 19/09/2019 00:23, Laurent Pinchart wrote:
> > On Thu, Sep 12, 2019 at 10:19:30AM +0100, Kieran Bingham wrote:
> >> On 12/09/2019 09:07, Jacopo Mondi wrote:
> >>> On Wed, Sep 11, 2019 at 07:40:27PM +0100, Kieran Bingham wrote:
> >>>> On 06/09/2019 14:54, Jacopo Mondi wrote:
> >>>>> Enable/disable the CMM associated with a CRTC at CRTC start and stop
> >>>>> time and enable the CMM unit through the Display Extensional Functions
> >>>>> register at group setup time.
> >>>>>
> >>>>> Reviewed-by: Ulrich Hecht <uli+renesas@xxxxxxxx>
> >>>>> Reviewed-by: Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx>
> >>>>> Signed-off-by: Jacopo Mondi <jacopo+renesas@xxxxxxxxxx>
> >>>>> ---
> >>>>>  drivers/gpu/drm/rcar-du/rcar_du_crtc.c  | 7 +++++++
> >>>>>  drivers/gpu/drm/rcar-du/rcar_du_group.c | 8 ++++++++
> >>>>>  drivers/gpu/drm/rcar-du/rcar_du_regs.h  | 5 +++++
> >>>>>  3 files changed, 20 insertions(+)
> >>>>>
> >>>>> diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
> >>>>> index 23f1d6cc1719..3dac605c3a67 100644
> >>>>> --- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
> >>>>> +++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
> >>>>> @@ -21,6 +21,7 @@
> >>>>>  #include <drm/drm_plane_helper.h>
> >>>>>  #include <drm/drm_vblank.h>
> >>>>>
> >>>>> +#include "rcar_cmm.h"
> >>>>>  #include "rcar_du_crtc.h"
> >>>>>  #include "rcar_du_drv.h"
> >>>>>  #include "rcar_du_encoder.h"
> >>>>> @@ -619,6 +620,9 @@ static void rcar_du_crtc_stop(struct rcar_du_crtc *rcrtc)
> >>>>>  	if (rcar_du_has(rcrtc->dev, RCAR_DU_FEATURE_VSP1_SOURCE))
> >>>>>  		rcar_du_vsp_disable(rcrtc);
> >>>>>
> >>>>> +	if (rcrtc->cmm)
> >>>>> +		rcar_cmm_disable(rcrtc->cmm);
> >>>>> +
> >>>>>  	/*
> >>>>>  	 * Select switch sync mode. This stops display operation and configures
> >>>>>  	 * the HSYNC and VSYNC signals as inputs.
> >>>>> @@ -686,6 +690,9 @@ static void rcar_du_crtc_atomic_enable(struct drm_crtc *crtc,
> >>>>>  	}
> >>>>>
> >>>>>  	rcar_du_crtc_start(rcrtc);
> >>>>> +
> >>>>> +	if (rcrtc->cmm)
> >>>>> +		rcar_cmm_enable(rcrtc->cmm);
> >>>>>  }
> >>>>>
> >>>>>  static void rcar_du_crtc_atomic_disable(struct drm_crtc *crtc,
> >>>>> diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.c b/drivers/gpu/drm/rcar-du/rcar_du_group.c
> >>>>> index 9eee47969e77..25d0fc125d7a 100644
> >>>>> --- a/drivers/gpu/drm/rcar-du/rcar_du_group.c
> >>>>> +++ b/drivers/gpu/drm/rcar-du/rcar_du_group.c
> >>>>> @@ -147,6 +147,14 @@ static void rcar_du_group_setup(struct rcar_du_group *rgrp)
> >>>>>
> >>>>>  	rcar_du_group_setup_pins(rgrp);
> >>>>>
> >>>>> +	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_CMM)) {
> >>>>> +		u32 defr7 = DEFR7_CODE
> >>>>> +			  | (rgrp->cmms_mask & BIT(1) ? DEFR7_CMME1 : 0)
> >>>>> +			  | (rgrp->cmms_mask & BIT(0) ? DEFR7_CMME0 : 0);
> >>>>> +
> >>>>> +		rcar_du_group_write(rgrp, DEFR7, defr7);
> >>>>> +	}
> >>>>> +
> >>>>
> >>>> What's the effect here on platforms with a CMM, but with
> >>>> CONFIG_DRM_RCAR_CMM unset?
> >>>>
> >>>> Will this incorrectly configure the DU ?
> >>>>
> >>>> Will it stall the display if the DU tries to interact with another
> >>>> module which is not enabled?
> >>>
> >>> I recall I tested that (that's why I had to add stubs for CMM
> >>> functions, as I had linkage errors otherwise) and thing seems to be
> >>> fine as the CMM configuration/enblement resolve to an empty function.
> >>
> >> Yes, I see the stubs to allow for linkage, but it's the hardware I'm
> >> concerned about. If it passes the tests and doesn't break then that's
> >> probably ok ... but I'm really weary that we're enabling a hardware
> >> pipeline with a disabled component in the middle.
> >>
> >>> Would you prefer to have this guarded by an #if IS_ENABLED() ?
> >>
> >> I don't think we need a compile time conditional, but I'd say it
> >> probably needs to be more about whether the CMM has actually probed or not
> >>
> >> Aha, and I see that in rcar_du_cmm_init() we already do a
> >> call to rcar_cmm_init(), which if fails will leave rcdu->cmms[i] as
> >> NULL. So that's catered for, which results in the rgrp->cmms_mask being
> >> correctly representative of whether there is a CMM connected or not.
> > 
> > Doesn't this result in probe failure ?
> 
> I think I mis-spoke above, I didn't mean "if rcar_cmm_init() fails" I
> meant "if rcar_du_cmm_init() determines there are no connected CMM's or
> if they are disabled."
> 
> If rcar_cmm_init() returns a failure, then yes we will fail to probe.
> 
> But I think it's up to rcar_du_cmm_init() to determine if the CMM exists
> or not (or is enabled) and if that's not a failure case then it should
> not prevent the probing of the DU.
> 
> In fact, I've now seen that if CONFIG_DRM_RCAR_CMM is not enabled,
> rcar_cmm_init() returns 0, and I think in fact it should return -ENODEV,
> with an exception on that return value in rcar_du_cmm_init() so that the
> DU continues with no CMM attached there.

I've replied to your other e-mail regarding this, and I agree with you.

> >>  ... so I think that means the ...
> >>  "if (rcar_du_has(rcdu, RCAR_DU_FEATURE_CMM))" is somewhat redundant:
> >>
> >>
> >> This could be:
> >>
> >>   if (rgrp->cmms_mask) {
> >> 	u32 defr7 = DEFR7_CODE
> >> 		  | (rgrp->cmms_mask & BIT(1) ? DEFR7_CMME1 : 0)
> >> 		  | (rgrp->cmms_mask & BIT(0) ? DEFR7_CMME0 : 0);
> >>
> >>   rcar_du_group_write(rgrp, DEFR7, defr7);
> >>
> >> Or in fact, if we don't mind writing 0 to DEFR7 when there is no CMM
> >> (which is safe by the looks of things as DEFR7 is available on all
> >> platforms), then we can even remove the outer conditional, and leave
> >> this all up to the ternary operators to write the correct value to the
> >> defr7.
> >>
> >> Phew ... net result - your current code *is* safe with the
> >> CONFIG_DRM_RCAR_CMM option disabled. I'll leave it up to you if you want
> >> to simplify the code here and remove the RCAR_DU_FEATURE_CMM.
> >>
> >> As this RCAR_DU_FEATURE_CMM flag is only checked here, removing it would
> >> however simplify all of the rcar_du_device_info structures.
> >>
> >> So - with or without the _FEATURE_CMM" simplification, this patch looks
> >> functional and safe so:
> >>
> >>
> >> Reviewed-by: Kieran Bingham <kieran.bingham+renesas@xxxxxxxxxxxxxxxx>
> >>
> >>>>>  	if (rcdu->info->gen >= 2) {
> >>>>>  		rcar_du_group_setup_defr8(rgrp);
> >>>>>  		rcar_du_group_setup_didsr(rgrp);
> >>>>> diff --git a/drivers/gpu/drm/rcar-du/rcar_du_regs.h b/drivers/gpu/drm/rcar-du/rcar_du_regs.h
> >>>>> index bc87f080b170..fb9964949368 100644
> >>>>> --- a/drivers/gpu/drm/rcar-du/rcar_du_regs.h
> >>>>> +++ b/drivers/gpu/drm/rcar-du/rcar_du_regs.h
> >>>>> @@ -197,6 +197,11 @@
> >>>>>  #define DEFR6_MLOS1		(1 << 2)
> >>>>>  #define DEFR6_DEFAULT		(DEFR6_CODE | DEFR6_TCNE1)
> >>>>>
> >>>>> +#define DEFR7			0x000ec
> >>>>> +#define DEFR7_CODE		(0x7779 << 16)
> >>>>> +#define DEFR7_CMME1		BIT(6)
> >>>>> +#define DEFR7_CMME0		BIT(4)
> >>>>> +
> >>>>>  /* -----------------------------------------------------------------------------
> >>>>>   * R8A7790-only Control Registers
> >>>>>   */

-- 
Regards,

Laurent Pinchart