Re: [PATCH V4 3/9] soc: qcom: geni: Support for ICC voting

[Matthias Kaehlcke <mka@xxxxxxxxxxxx>]

Hi Akash,

On Wed, Apr 15, 2020 at 03:53:12PM +0530, Akash Asthana wrote:
> Add necessary macros and structure variables to support ICC BW
> voting from individual SE drivers.
> 
> Signed-off-by: Akash Asthana <akashast@xxxxxxxxxxxxxx>
> ---
> Changes in V2:
>  - As per Bjorn's comment dropped enums for ICC paths, given the three
>    paths individual members
> 
> Changes in V3:
>  - Add geni_icc_get, geni_icc_vote_on and geni_icc_vote_off as helper API.
>  - Add geni_icc_path structure in common header
> 
> Changes in V4:
>  - As per Bjorn's comment print error message in geni_icc_get if return
>    value is not -EPROBE_DEFER.
>  - As per Bjorn's comment remove NULL on path before calling icc_set_bw
>    API.
>  - As per Bjorn's comment drop __func__ print.
>  - As per Matthias's comment, make ICC path a array instead of individual
>    member entry in geni_se struct.
> 
> Note: I have ignored below check patch suggestion because it was throwing
>       compilation error as 'icc_ddr' is not compile time comstant.
> 
> WARNING: char * array declaration might be better as static const
>  - FILE: drivers/soc/qcom/qcom-geni-se.c:726:
>  - const char *icc_names[] = {"qup-core", "qup-config", icc_ddr};
> 
> 
>  drivers/soc/qcom/qcom-geni-se.c | 61 +++++++++++++++++++++++++++++++++++++++++
>  include/linux/qcom-geni-se.h    | 31 +++++++++++++++++++++
>  2 files changed, 92 insertions(+)
> 
> diff --git a/drivers/soc/qcom/qcom-geni-se.c b/drivers/soc/qcom/qcom-geni-se.c
> index 7d622ea..1527bc4 100644
> --- a/drivers/soc/qcom/qcom-geni-se.c
> +++ b/drivers/soc/qcom/qcom-geni-se.c
> @@ -720,6 +720,67 @@ void geni_se_rx_dma_unprep(struct geni_se *se, dma_addr_t iova, size_t len)
>  }
>  EXPORT_SYMBOL(geni_se_rx_dma_unprep);
>  
> +int geni_icc_get(struct geni_se *se, const char *icc_ddr)
> +{
> +	int i, icc_err;
> +	const char *icc_names[] = {"qup-core", "qup-config", icc_ddr};
> +
> +	for (i = 0; i < ARRAY_SIZE(se->icc_paths); i++) {
> +		if (!icc_names[i])
> +			continue;
> +
> +		se->icc_paths[i].path = devm_of_icc_get(se->dev, icc_names[i]);
> +		if (IS_ERR(se->icc_paths[i].path))
> +			goto icc_get_failure;
> +	}
> +
> +	return 0;
> +
> +icc_get_failure:
> +	icc_err = PTR_ERR(se->icc_paths[i].path);
> +	if (icc_err != -EPROBE_DEFER)
> +		dev_err_ratelimited(se->dev, "Failed to get path:%d, ret:%d\n",

Better be explicit that it's an ICC path and log icc_names[i] instead of i.

> +					i, icc_err);
> +	return icc_err;
> +
> +}
> +EXPORT_SYMBOL(geni_icc_get);
> +
> +int geni_icc_vote_on(struct geni_se *se)
> +{
> +	int i, ret;
> +
> +	for (i = 0; i < ARRAY_SIZE(se->icc_paths); i++) {
> +		ret = icc_set_bw(se->icc_paths[i].path,
> +			se->icc_paths[i].avg_bw, se->icc_paths[i].peak_bw);

I'll leave it to others to decide whether it's ok to leave the
implementation of the icc_enable/disable() APIs suggested on
https://patchwork.kernel.org/patch/11467511/#23269555 for later.

> +		if (ret) {
> +			dev_err_ratelimited(se->dev, "ICC BW voting failed on path:%d, ret:%d\n",
> +					i, ret);

Instead of logging the index, which isn't very expressive, you could have
a static string array of the path names ({"core", "config", "ddr"} or
similar) that is used when logging errors in _vote_on/off().

> +			return ret;
> +		}
> +	}
> +
> +	return 0;
> +}
> +EXPORT_SYMBOL(geni_icc_vote_on);
> +
> +int geni_icc_vote_off(struct geni_se *se)
> +{
> +	int i, ret;
> +
> +	for (i = 0; i < ARRAY_SIZE(se->icc_paths); i++) {
> +		ret = icc_set_bw(se->icc_paths[i].path, 0, 0);
> +		if (ret) {
> +			dev_err_ratelimited(se->dev, "ICC BW remove failed on path:%d, ret:%d\n",
> +					i, ret);
> +			return ret;
> +		}
> +	}
> +
> +	return 0;
> +}
> +EXPORT_SYMBOL(geni_icc_vote_off);
> +
>  static int geni_se_probe(struct platform_device *pdev)
>  {
>  	struct device *dev = &pdev->dev;
> diff --git a/include/linux/qcom-geni-se.h b/include/linux/qcom-geni-se.h
> index dd46494..b5b9316 100644
> --- a/include/linux/qcom-geni-se.h
> +++ b/include/linux/qcom-geni-se.h
> @@ -6,6 +6,8 @@
>  #ifndef _LINUX_QCOM_GENI_SE
>  #define _LINUX_QCOM_GENI_SE
>  
> +#include <linux/interconnect.h>
> +
>  /* Transfer mode supported by GENI Serial Engines */
>  enum geni_se_xfer_mode {
>  	GENI_SE_INVALID,
> @@ -25,6 +27,12 @@ enum geni_se_protocol_type {
>  struct geni_wrapper;
>  struct clk;
>  
> +struct geni_icc_path {
> +	struct icc_path *path;
> +	unsigned int avg_bw;
> +	unsigned int peak_bw;
> +};
> +
>  /**
>   * struct geni_se - GENI Serial Engine
>   * @base:		Base Address of the Serial Engine's register block
> @@ -33,6 +41,7 @@ struct clk;
>   * @clk:		Handle to the core serial engine clock
>   * @num_clk_levels:	Number of valid clock levels in clk_perf_tbl
>   * @clk_perf_tbl:	Table of clock frequency input to serial engine clock
> + * @icc_paths:		Array of ICC paths for SE
>   */
>  struct geni_se {
>  	void __iomem *base;
> @@ -41,6 +50,7 @@ struct geni_se {
>  	struct clk *clk;
>  	unsigned int num_clk_levels;
>  	unsigned long *clk_perf_tbl;
> +	struct geni_icc_path icc_paths[3];

You also need enums for the paths, otherwise you end up with code like
this, which isn't really self-explanatory:

  gi2c->se.icc_paths[0].avg_bw = GENI_DEFAULT_BW;
  gi2c->se.icc_paths[1].avg_bw = GENI_DEFAULT_BW;
  gi2c->se.icc_paths[2].avg_bw = Bps_to_icc(gi2c->clk_freq_out);

  (from "[V4,5/9] i2c: i2c-qcom-geni: Add interconnect support")

I know Bjorn asked in v1 to remove the enums you had, however it was
a slightly different context. If we are sticking to use an array of
'struct geni_icc_path' (which reduces redundant code) the enums are
'needed'.