Re: [RFC 3/5] media: adv748x: Make lanes number depend on routing

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Jacopo,

On 16/03/2019 15:47, Jacopo Mondi wrote:
> Use the TXA routing information to configure the number of active CSI-2
> data lanes. When routing AFE through TXA limit the number of data lanes
> to 1, while in the canonical HDMI->TXA routing use all the physically
> available ones.
> 
> The number of lanes collected from the 'data-lanes' DT property is now
> used as a the number of physically available data lanes, while the
> 'num_lanes' variable contains the number of active ones.

Could the variable perhaps be called active_lanes rather than num_lanes
if it represents how many are active rather than how many there are?


> 
> Signed-off-by: Jacopo Mondi <jacopo+renesas@xxxxxxxxxx>
> ---
>  drivers/media/i2c/adv748x/adv748x-core.c | 19 ++++++++++++++++++-
>  drivers/media/i2c/adv748x/adv748x.h      |  1 +
>  2 files changed, 19 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/media/i2c/adv748x/adv748x-core.c b/drivers/media/i2c/adv748x/adv748x-core.c
> index 02135741b1a6..f91c7b83f1bf 100644
> --- a/drivers/media/i2c/adv748x/adv748x-core.c
> +++ b/drivers/media/i2c/adv748x/adv748x-core.c
> @@ -350,6 +350,8 @@ static int adv748x_link_setup(struct media_entity *entity,
>  	struct adv748x_csi2 *tx = adv748x_sd_to_csi2(sd);
>  	bool enable = flags & MEDIA_LNK_FL_ENABLED;
>  	u8 csi4_in_sel = 0;
> +	u8 num_lanes;
> +	int ret;
>  
>  	/* Refuse to enable multiple links to the same TX at the same time. */
>  	if (enable && tx->src)
> @@ -373,10 +375,23 @@ static int adv748x_link_setup(struct media_entity *entity,
>  			csi4_in_sel |= ADV748X_IO_10_CSI4_IN_SEL_AFE;
>  		else
>  			csi4_in_sel |= ADV748X_IO_10_CSI1_EN;
> +
> +		num_lanes = 1;
>  	}
>  
> -	if (state->hdmi.tx)
> +	if (state->hdmi.tx) {
>  		csi4_in_sel |= ADV748X_IO_10_CSI4_EN;
> +		num_lanes = tx->available_lanes;
> +	}
> +
> +	/* Update the number of active lanes if it has changed. */
> +	if (num_lanes != tx->num_lanes) {
> +		tx->num_lanes = num_lanes;
> +		ret = adv748x_write(state, tx->page, 0x00,
> +				    0x80 | tx->num_lanes);
> +		if (ret)
> +			return ret;
> +	}
>  
>  	state->csi4_in_sel = csi4_in_sel;
>  
> @@ -608,6 +623,7 @@ static int adv748x_parse_csi2_lanes(struct adv748x_state *state,
>  		}
>  
>  		state->txa.num_lanes = num_lanes;
> +		state->txa.available_lanes = num_lanes;
>  		adv_dbg(state, "TXA: using %u lanes\n", state->txa.num_lanes);
>  	}
>  
> @@ -619,6 +635,7 @@ static int adv748x_parse_csi2_lanes(struct adv748x_state *state,
>  		}
>  
>  		state->txb.num_lanes = num_lanes;
> +		state->txb.available_lanes = num_lanes;
>  		adv_dbg(state, "TXB: using %u lanes\n", state->txb.num_lanes);
>  	}
>  
> diff --git a/drivers/media/i2c/adv748x/adv748x.h b/drivers/media/i2c/adv748x/adv748x.h
> index 27c116d09284..6e5c2cb421fe 100644
> --- a/drivers/media/i2c/adv748x/adv748x.h
> +++ b/drivers/media/i2c/adv748x/adv748x.h
> @@ -80,6 +80,7 @@ struct adv748x_csi2 {
>  	unsigned int page;
>  	unsigned int port;
>  	unsigned int num_lanes;

Aha, except num_lanes looks like its' from an earlier patch...

I wonder if it's worth renaming...

> +	unsigned int available_lanes;
>  
>  	struct media_pad pads[ADV748X_CSI2_NR_PADS];
>  	struct v4l2_ctrl_handler ctrl_hdl;
> 

-- 
Regards
--
Kieran



[Index of Archives]     [Linux Input]     [Video for Linux]     [Gstreamer Embedded]     [Mplayer Users]     [Linux USB Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]

  Powered by Linux