Re: [PATCH v3 11/15] media: qcom: camss: Functionally decompose CSIPHY clock lookups

[Konrad Dybcio <konrad.dybcio@linaro.org>]

On 23.08.2023 12:44, Bryan O'Donoghue wrote:
> The csiphyX_timer and csiX_phy values need not be hard-coded. We can
> functionally decompose the string matching inside of a loop.
> 
> Static string values are brittle, difficult to extend and not required
> anyway since the camss->res->csiphy_num value informs us of the number
> of CSIPHYs and hence the set of potential clocks for a given CSIPHY.
> 
> In simple terms if we have five CSIPHYs we can have no more and no less
> than five csiphy_timer clocks. Similarly csi_phy core clocks have a 1:1
> relationship with the PHY they clock.
> 
> Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
> ---
>  .../media/platform/qcom/camss/camss-csiphy.c  | 37 ++++++++++++-------
>  1 file changed, 23 insertions(+), 14 deletions(-)
> 
> diff --git a/drivers/media/platform/qcom/camss/camss-csiphy.c b/drivers/media/platform/qcom/camss/camss-csiphy.c
> index 0e8c2a59ea241..baf78c525fbfc 100644
> --- a/drivers/media/platform/qcom/camss/camss-csiphy.c
> +++ b/drivers/media/platform/qcom/camss/camss-csiphy.c
> @@ -536,6 +536,15 @@ static int csiphy_init_formats(struct v4l2_subdev *sd,
>  	return csiphy_set_format(sd, fh ? fh->state : NULL, &format);
>  }
>  
> +static bool csiphy_match_clock_name(const char *clock_name, const char *format,
> +				    int index)
> +{
> +	char name[CAMSS_RES_MAX];
similar comment to the previous patch about the buf size

> +
> +	snprintf(name, sizeof(name), format, index);
> +	return !strcmp(clock_name, name);
> +}
> +
>  /*
>   * msm_csiphy_subdev_init - Initialize CSIPHY device structure and resources
>   * @csiphy: CSIPHY device
> @@ -550,7 +559,7 @@ int msm_csiphy_subdev_init(struct camss *camss,
>  {
>  	struct device *dev = camss->dev;
>  	struct platform_device *pdev = to_platform_device(dev);
> -	int i, j;
> +	int i, j, k;
>  	int ret;
>  
>  	csiphy->camss = camss;
> @@ -666,19 +675,19 @@ int msm_csiphy_subdev_init(struct camss *camss,
>  		for (j = 0; j < clock->nfreqs; j++)
>  			clock->freq[j] = res->clock_rate[i][j];
>  
> -		if (!strcmp(clock->name, "csiphy0_timer") ||
> -		    !strcmp(clock->name, "csiphy1_timer") ||
> -		    !strcmp(clock->name, "csiphy2_timer") ||
> -		    !strcmp(clock->name, "csiphy3_timer") ||
> -		    !strcmp(clock->name, "csiphy4_timer") ||
> -		    !strcmp(clock->name, "csiphy5_timer"))
> -			csiphy->rate_set[i] = true;
> -
> -		if (camss->res->version == CAMSS_660 &&
> -		    (!strcmp(clock->name, "csi0_phy") ||
> -		     !strcmp(clock->name, "csi1_phy") ||
> -		     !strcmp(clock->name, "csi2_phy")))
> -			csiphy->rate_set[i] = true;
> +		for (k = 0; k < camss->res->csiphy_num; k++) {
> +			csiphy->rate_set[i] = csiphy_match_clock_name(clock->name,
> +								      "csiphy%d_timer", k);
This entire functions is like.. soooo over-engineered

adding something like csiphy_timer_clks and cisphy_clks and stuff
would make this string comparison mess unnecessary

> +			if (csiphy->rate_set[i])
> +				break;
> +
> +			if (camss->res->version == CAMSS_660) {
> +				csiphy->rate_set[i] = csiphy_match_clock_name(clock->name,
> +									      "csi%d_phy", k);
hm, only ratesetting on 660 sounds very sus

Konrad