[PATCH v2] PM: AVS: qcom-cpr: Use nvmem_cell_read_variable_le_u32()

[PATCH v2] PM: AVS: qcom-cpr: Use nvmem_cell_read_variable_le_u32()

[Douglas Anderson]

Let's delete the private function cpr_read_efuse() since it does the
basically the same thing as the new API call
nvmem_cell_read_variable_le_u32().

Differences between the new API call and the old private function:
* less error printing (I assume this is OK).
* will give an error if the value doesn't fit in 32-bits (the old code
  would have truncated silently).

Signed-off-by: Douglas Anderson <dianders@chromium.org>
---
I haven't done any more than compile-test this. Mostly I'm just
writing this patch because it helped provide inspiration for the
general API function.

Changes in v2:
- Resending v1 as a singleton patch; dependency is merged in mainline.

 drivers/soc/qcom/cpr.c | 43 +++++-------------------------------------
 1 file changed, 5 insertions(+), 38 deletions(-)

diff --git a/drivers/soc/qcom/cpr.c b/drivers/soc/qcom/cpr.c
index b24cc77d1889..4ce8e816154f 100644
--- a/drivers/soc/qcom/cpr.c
+++ b/drivers/soc/qcom/cpr.c
@@ -801,38 +801,6 @@ static int cpr_set_performance_state(struct generic_pm_domain *domain,
 	return ret;
 }
 
-static int cpr_read_efuse(struct device *dev, const char *cname, u32 *data)
-{
-	struct nvmem_cell *cell;
-	ssize_t len;
-	char *ret;
-	int i;
-
-	*data = 0;
-
-	cell = nvmem_cell_get(dev, cname);
-	if (IS_ERR(cell)) {
-		if (PTR_ERR(cell) != -EPROBE_DEFER)
-			dev_err(dev, "undefined cell %s\n", cname);
-		return PTR_ERR(cell);
-	}
-
-	ret = nvmem_cell_read(cell, &len);
-	nvmem_cell_put(cell);
-	if (IS_ERR(ret)) {
-		dev_err(dev, "can't read cell %s\n", cname);
-		return PTR_ERR(ret);
-	}
-
-	for (i = 0; i < len; i++)
-		*data |= ret[i] << (8 * i);
-
-	kfree(ret);
-	dev_dbg(dev, "efuse read(%s) = %x, bytes %zd\n", cname, *data, len);
-
-	return 0;
-}
-
 static int
 cpr_populate_ring_osc_idx(struct cpr_drv *drv)
 {
@@ -843,8 +811,7 @@ cpr_populate_ring_osc_idx(struct cpr_drv *drv)
 	int ret;
 
 	for (; fuse < end; fuse++, fuses++) {
-		ret = cpr_read_efuse(drv->dev, fuses->ring_osc,
-				     &data);
+		ret = nvmem_cell_read_variable_le_u32(drv->dev, fuses->ring_osc, &data);
 		if (ret)
 			return ret;
 		fuse->ring_osc_idx = data;
@@ -863,7 +830,7 @@ static int cpr_read_fuse_uV(const struct cpr_desc *desc,
 	u32 bits = 0;
 	int ret;
 
-	ret = cpr_read_efuse(drv->dev, init_v_efuse, &bits);
+	ret = nvmem_cell_read_variable_le_u32(drv->dev, init_v_efuse, &bits);
 	if (ret)
 		return ret;
 
@@ -932,7 +899,7 @@ static int cpr_fuse_corner_init(struct cpr_drv *drv)
 		}
 
 		/* Populate target quotient by scaling */
-		ret = cpr_read_efuse(drv->dev, fuses->quotient, &fuse->quot);
+		ret = nvmem_cell_read_variable_le_u32(drv->dev, fuses->quotient, &fuse->quot);
 		if (ret)
 			return ret;
 
@@ -1001,7 +968,7 @@ static int cpr_calculate_scaling(const char *quot_offset,
 	prev_fuse = fuse - 1;
 
 	if (quot_offset) {
-		ret = cpr_read_efuse(drv->dev, quot_offset, &quot_diff);
+		ret = nvmem_cell_read_variable_le_u32(drv->dev, quot_offset, &quot_diff);
 		if (ret)
 			return ret;
 
@@ -1701,7 +1668,7 @@ static int cpr_probe(struct platform_device *pdev)
 	 * initialized after attaching to the power domain,
 	 * since it depends on the CPU's OPP table.
 	 */
-	ret = cpr_read_efuse(dev, "cpr_fuse_revision", &cpr_rev);
+	ret = nvmem_cell_read_variable_le_u32(dev, "cpr_fuse_revision", &cpr_rev);
 	if (ret)
 		return ret;
 
-- 
2.31.1.818.g46aad6cb9e-goog

Re: [PATCH v2] PM: AVS: qcom-cpr: Use nvmem_cell_read_variable_le_u32()

[Niklas Cassel]

On Fri, May 21, 2021 at 01:44:53PM -0700, Douglas Anderson wrote:
> Let's delete the private function cpr_read_efuse() since it does the
> basically the same thing as the new API call
> nvmem_cell_read_variable_le_u32().
> 
> Differences between the new API call and the old private function:
> * less error printing (I assume this is OK).
> * will give an error if the value doesn't fit in 32-bits (the old code
>   would have truncated silently).
> 
> Signed-off-by: Douglas Anderson <dianders@chromium.org>
> ---
> I haven't done any more than compile-test this. Mostly I'm just
> writing this patch because it helped provide inspiration for the
> general API function.
> 
> Changes in v2:
> - Resending v1 as a singleton patch; dependency is merged in mainline.
> 
>  drivers/soc/qcom/cpr.c | 43 +++++-------------------------------------
>  1 file changed, 5 insertions(+), 38 deletions(-)

Acked-by: Niklas Cassel <nks@flawful.org>

Re: [PATCH v2] PM: AVS: qcom-cpr: Use nvmem_cell_read_variable_le_u32()

[Bjorn Andersson]

On Fri 21 May 15:44 CDT 2021, Douglas Anderson wrote:

> Let's delete the private function cpr_read_efuse() since it does the
> basically the same thing as the new API call
> nvmem_cell_read_variable_le_u32().
> 
> Differences between the new API call and the old private function:
> * less error printing (I assume this is OK).
> * will give an error if the value doesn't fit in 32-bits (the old code
>   would have truncated silently).
> 

Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>

Regards,
Bjorn

> Signed-off-by: Douglas Anderson <dianders@chromium.org>
> ---
> I haven't done any more than compile-test this. Mostly I'm just
> writing this patch because it helped provide inspiration for the
> general API function.
> 
> Changes in v2:
> - Resending v1 as a singleton patch; dependency is merged in mainline.
> 
>  drivers/soc/qcom/cpr.c | 43 +++++-------------------------------------
>  1 file changed, 5 insertions(+), 38 deletions(-)
> 
> diff --git a/drivers/soc/qcom/cpr.c b/drivers/soc/qcom/cpr.c
> index b24cc77d1889..4ce8e816154f 100644
> --- a/drivers/soc/qcom/cpr.c
> +++ b/drivers/soc/qcom/cpr.c
> @@ -801,38 +801,6 @@ static int cpr_set_performance_state(struct generic_pm_domain *domain,
>  	return ret;
>  }
>  
> -static int cpr_read_efuse(struct device *dev, const char *cname, u32 *data)
> -{
> -	struct nvmem_cell *cell;
> -	ssize_t len;
> -	char *ret;
> -	int i;
> -
> -	*data = 0;
> -
> -	cell = nvmem_cell_get(dev, cname);
> -	if (IS_ERR(cell)) {
> -		if (PTR_ERR(cell) != -EPROBE_DEFER)
> -			dev_err(dev, "undefined cell %s\n", cname);
> -		return PTR_ERR(cell);
> -	}
> -
> -	ret = nvmem_cell_read(cell, &len);
> -	nvmem_cell_put(cell);
> -	if (IS_ERR(ret)) {
> -		dev_err(dev, "can't read cell %s\n", cname);
> -		return PTR_ERR(ret);
> -	}
> -
> -	for (i = 0; i < len; i++)
> -		*data |= ret[i] << (8 * i);
> -
> -	kfree(ret);
> -	dev_dbg(dev, "efuse read(%s) = %x, bytes %zd\n", cname, *data, len);
> -
> -	return 0;
> -}
> -
>  static int
>  cpr_populate_ring_osc_idx(struct cpr_drv *drv)
>  {
> @@ -843,8 +811,7 @@ cpr_populate_ring_osc_idx(struct cpr_drv *drv)
>  	int ret;
>  
>  	for (; fuse < end; fuse++, fuses++) {
> -		ret = cpr_read_efuse(drv->dev, fuses->ring_osc,
> -				     &data);
> +		ret = nvmem_cell_read_variable_le_u32(drv->dev, fuses->ring_osc, &data);
>  		if (ret)
>  			return ret;
>  		fuse->ring_osc_idx = data;
> @@ -863,7 +830,7 @@ static int cpr_read_fuse_uV(const struct cpr_desc *desc,
>  	u32 bits = 0;
>  	int ret;
>  
> -	ret = cpr_read_efuse(drv->dev, init_v_efuse, &bits);
> +	ret = nvmem_cell_read_variable_le_u32(drv->dev, init_v_efuse, &bits);
>  	if (ret)
>  		return ret;
>  
> @@ -932,7 +899,7 @@ static int cpr_fuse_corner_init(struct cpr_drv *drv)
>  		}
>  
>  		/* Populate target quotient by scaling */
> -		ret = cpr_read_efuse(drv->dev, fuses->quotient, &fuse->quot);
> +		ret = nvmem_cell_read_variable_le_u32(drv->dev, fuses->quotient, &fuse->quot);
>  		if (ret)
>  			return ret;
>  
> @@ -1001,7 +968,7 @@ static int cpr_calculate_scaling(const char *quot_offset,
>  	prev_fuse = fuse - 1;
>  
>  	if (quot_offset) {
> -		ret = cpr_read_efuse(drv->dev, quot_offset, &quot_diff);
> +		ret = nvmem_cell_read_variable_le_u32(drv->dev, quot_offset, &quot_diff);
>  		if (ret)
>  			return ret;
>  
> @@ -1701,7 +1668,7 @@ static int cpr_probe(struct platform_device *pdev)
>  	 * initialized after attaching to the power domain,
>  	 * since it depends on the CPU's OPP table.
>  	 */
> -	ret = cpr_read_efuse(dev, "cpr_fuse_revision", &cpr_rev);
> +	ret = nvmem_cell_read_variable_le_u32(dev, "cpr_fuse_revision", &cpr_rev);
>  	if (ret)
>  		return ret;
>  
> -- 
> 2.31.1.818.g46aad6cb9e-goog
> 

Re: [PATCH v2] PM: AVS: qcom-cpr: Use nvmem_cell_read_variable_le_u32()

[Doug Anderson]

Rafael / Bjorn,

On Fri, May 21, 2021 at 1:45 PM Douglas Anderson <dianders@chromium.org> wrote:
>
> Let's delete the private function cpr_read_efuse() since it does the
> basically the same thing as the new API call
> nvmem_cell_read_variable_le_u32().
>
> Differences between the new API call and the old private function:
> * less error printing (I assume this is OK).
> * will give an error if the value doesn't fit in 32-bits (the old code
>   would have truncated silently).
>
> Signed-off-by: Douglas Anderson <dianders@chromium.org>
> ---
> I haven't done any more than compile-test this. Mostly I'm just
> writing this patch because it helped provide inspiration for the
> general API function.
>
> Changes in v2:
> - Resending v1 as a singleton patch; dependency is merged in mainline.
>
>  drivers/soc/qcom/cpr.c | 43 +++++-------------------------------------
>  1 file changed, 5 insertions(+), 38 deletions(-)

Are either of you interested in landing this? I guess Rafael landed
most of the recent changes to this driver, but it used to be under the
'power' directory. Now that it's under 'drivers/soc/qcom' maybe it
should go through Bjorn's tree?

-Doug

Re: [PATCH v2] PM: AVS: qcom-cpr: Use nvmem_cell_read_variable_le_u32()

[Ulf Hansson]

On Sat, 24 Jul 2021 at 01:05, Doug Anderson <dianders@chromium.org> wrote:
>
> Rafael / Bjorn,
>
> On Fri, May 21, 2021 at 1:45 PM Douglas Anderson <dianders@chromium.org> wrote:
> >
> > Let's delete the private function cpr_read_efuse() since it does the
> > basically the same thing as the new API call
> > nvmem_cell_read_variable_le_u32().
> >
> > Differences between the new API call and the old private function:
> > * less error printing (I assume this is OK).
> > * will give an error if the value doesn't fit in 32-bits (the old code
> >   would have truncated silently).
> >
> > Signed-off-by: Douglas Anderson <dianders@chromium.org>
> > ---
> > I haven't done any more than compile-test this. Mostly I'm just
> > writing this patch because it helped provide inspiration for the
> > general API function.
> >
> > Changes in v2:
> > - Resending v1 as a singleton patch; dependency is merged in mainline.
> >
> >  drivers/soc/qcom/cpr.c | 43 +++++-------------------------------------
> >  1 file changed, 5 insertions(+), 38 deletions(-)
>
> Are either of you interested in landing this? I guess Rafael landed
> most of the recent changes to this driver, but it used to be under the
> 'power' directory. Now that it's under 'drivers/soc/qcom' maybe it
> should go through Bjorn's tree?

Yes, that was exactly the intent of moving the files. I guess Bjorn
has simply been busy.

Kind regards
Uffe

Re: [PATCH v2] PM: AVS: qcom-cpr: Use nvmem_cell_read_variable_le_u32()

[Bjorn Andersson]

On Fri 23 Jul 18:05 CDT 2021, Doug Anderson wrote:

> Rafael / Bjorn,
> 
> On Fri, May 21, 2021 at 1:45 PM Douglas Anderson <dianders@chromium.org> wrote:
> >
> > Let's delete the private function cpr_read_efuse() since it does the
> > basically the same thing as the new API call
> > nvmem_cell_read_variable_le_u32().
> >
> > Differences between the new API call and the old private function:
> > * less error printing (I assume this is OK).
> > * will give an error if the value doesn't fit in 32-bits (the old code
> >   would have truncated silently).
> >
> > Signed-off-by: Douglas Anderson <dianders@chromium.org>
> > ---
> > I haven't done any more than compile-test this. Mostly I'm just
> > writing this patch because it helped provide inspiration for the
> > general API function.
> >
> > Changes in v2:
> > - Resending v1 as a singleton patch; dependency is merged in mainline.
> >
> >  drivers/soc/qcom/cpr.c | 43 +++++-------------------------------------
> >  1 file changed, 5 insertions(+), 38 deletions(-)
> 
> Are either of you interested in landing this? I guess Rafael landed
> most of the recent changes to this driver, but it used to be under the
> 'power' directory. Now that it's under 'drivers/soc/qcom' maybe it
> should go through Bjorn's tree?
> 

Yes, it makes sense to pick it up through my tree, unfortunately I
forgot about this move and ignored the patch in my queue based on the
$subject.

I've picked it up now.

Regards,
Bjorn