| Message ID | ab6dd34ce0236de12ec1e9984931780c3df643f9.1549525708.git.amit.kucheria@linaro.org (mailing list archive) |
|---|---|
| State | Changes Requested |
| Delegated to: | Eduardo Valentin |
| Headers | show |
| Series | thermal: tsens: Refactor to use regmap_field | expand |
On Thu, Feb 07, 2019 at 04:19:31PM +0530, Amit Kucheria wrote: > On some TSENS IP, version is stored. Print that version at init. > > Signed-off-by: Amit Kucheria <amit.kucheria@linaro.org> > --- > drivers/thermal/qcom/tsens-common.c | 23 ++++++++++++++++++++++- > 1 file changed, 22 insertions(+), 1 deletion(-) > > diff --git a/drivers/thermal/qcom/tsens-common.c b/drivers/thermal/qcom/tsens-common.c > index aae3d71d7eed..39cd5733fd44 100644 > --- a/drivers/thermal/qcom/tsens-common.c > +++ b/drivers/thermal/qcom/tsens-common.c > @@ -121,7 +121,7 @@ int __init init_common(struct tsens_priv *priv) > void __iomem *tm_base, *srot_base; > struct device *dev = priv->dev; > struct resource *res; > - u32 enabled; > + u32 enabled, maj_ver, min_ver; > int ret, i, j; > struct platform_device *op = of_find_device_by_node(priv->dev->of_node); > > @@ -155,6 +155,27 @@ int __init init_common(struct tsens_priv *priv) > if (IS_ERR(priv->tm_map)) > return PTR_ERR(priv->tm_map); > > + for (i = 0; i < MAX_REGFIELDS; i++) { > + priv->rf[i] = NULL; > + } > + > + /* alloc regmap_fields in srot_map */ > + if (priv->feat->ver_info) { > + for (i = 0, j = VER_MAJOR; i < 2; i++, j++) { > + priv->rf[j] = devm_regmap_field_alloc(dev, priv->srot_map, > + priv->fields[j]); > + if (IS_ERR(priv->rf[j])) > + return PTR_ERR(priv->rf[j]); > + } > + ret = regmap_field_read(priv->rf[VER_MAJOR], &maj_ver); > + if (ret) > + return ret; > + ret = regmap_field_read(priv->rf[VER_MINOR], &min_ver); > + if (ret) > + return ret; > + dev_info(dev, "version: %d.%d\n", maj_ver, min_ver); Is this really needed? Printing stuff into kernel log during boot can just add a lot of info that ends up just slowing down things. Besides, if this info is important shouldnt be available somewhere else as the kernel log is a circular buffer and eventually that line will get wrapped up? > + } > + > priv->rf[TSENS_EN] = devm_regmap_field_alloc(dev, priv->srot_map, > priv->fields[TSENS_EN]); > if (IS_ERR(priv->rf[TSENS_EN])) > -- > 2.17.1 >
On Wed, Feb 20, 2019 at 6:47 AM Eduardo Valentin <edubezval@gmail.com> wrote: > > On Thu, Feb 07, 2019 at 04:19:31PM +0530, Amit Kucheria wrote: > > On some TSENS IP, version is stored. Print that version at init. > > > > Signed-off-by: Amit Kucheria <amit.kucheria@linaro.org> > > --- > > drivers/thermal/qcom/tsens-common.c | 23 ++++++++++++++++++++++- > > 1 file changed, 22 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/thermal/qcom/tsens-common.c b/drivers/thermal/qcom/tsens-common.c > > index aae3d71d7eed..39cd5733fd44 100644 > > --- a/drivers/thermal/qcom/tsens-common.c > > +++ b/drivers/thermal/qcom/tsens-common.c > > @@ -121,7 +121,7 @@ int __init init_common(struct tsens_priv *priv) > > void __iomem *tm_base, *srot_base; > > struct device *dev = priv->dev; > > struct resource *res; > > - u32 enabled; > > + u32 enabled, maj_ver, min_ver; > > int ret, i, j; > > struct platform_device *op = of_find_device_by_node(priv->dev->of_node); > > > > @@ -155,6 +155,27 @@ int __init init_common(struct tsens_priv *priv) > > if (IS_ERR(priv->tm_map)) > > return PTR_ERR(priv->tm_map); > > > > + for (i = 0; i < MAX_REGFIELDS; i++) { > > + priv->rf[i] = NULL; > > + } > > + > > + /* alloc regmap_fields in srot_map */ > > + if (priv->feat->ver_info) { > > + for (i = 0, j = VER_MAJOR; i < 2; i++, j++) { > > + priv->rf[j] = devm_regmap_field_alloc(dev, priv->srot_map, > > + priv->fields[j]); > > + if (IS_ERR(priv->rf[j])) > > + return PTR_ERR(priv->rf[j]); > > + } > > + ret = regmap_field_read(priv->rf[VER_MAJOR], &maj_ver); > > + if (ret) > > + return ret; > > + ret = regmap_field_read(priv->rf[VER_MINOR], &min_ver); > > + if (ret) > > + return ret; > > + dev_info(dev, "version: %d.%d\n", maj_ver, min_ver); > > Is this really needed? Printing stuff into kernel log during boot can > just add a lot of info that ends up just slowing down things. Besides, > if this info is important shouldnt be available somewhere else as the > kernel log is a circular buffer and eventually that line will get > wrapped up? Fair point, I'll remove this statement and dump this into debugfs for which I'll send a separate series adding debugfs support.
diff --git a/drivers/thermal/qcom/tsens-common.c b/drivers/thermal/qcom/tsens-common.c index aae3d71d7eed..39cd5733fd44 100644 --- a/drivers/thermal/qcom/tsens-common.c +++ b/drivers/thermal/qcom/tsens-common.c @@ -121,7 +121,7 @@ int __init init_common(struct tsens_priv *priv) void __iomem *tm_base, *srot_base; struct device *dev = priv->dev; struct resource *res; - u32 enabled; + u32 enabled, maj_ver, min_ver; int ret, i, j; struct platform_device *op = of_find_device_by_node(priv->dev->of_node); @@ -155,6 +155,27 @@ int __init init_common(struct tsens_priv *priv) if (IS_ERR(priv->tm_map)) return PTR_ERR(priv->tm_map); + for (i = 0; i < MAX_REGFIELDS; i++) { + priv->rf[i] = NULL; + } + + /* alloc regmap_fields in srot_map */ + if (priv->feat->ver_info) { + for (i = 0, j = VER_MAJOR; i < 2; i++, j++) { + priv->rf[j] = devm_regmap_field_alloc(dev, priv->srot_map, + priv->fields[j]); + if (IS_ERR(priv->rf[j])) + return PTR_ERR(priv->rf[j]); + } + ret = regmap_field_read(priv->rf[VER_MAJOR], &maj_ver); + if (ret) + return ret; + ret = regmap_field_read(priv->rf[VER_MINOR], &min_ver); + if (ret) + return ret; + dev_info(dev, "version: %d.%d\n", maj_ver, min_ver); + } + priv->rf[TSENS_EN] = devm_regmap_field_alloc(dev, priv->srot_map, priv->fields[TSENS_EN]); if (IS_ERR(priv->rf[TSENS_EN]))
On some TSENS IP, version is stored. Print that version at init. Signed-off-by: Amit Kucheria <amit.kucheria@linaro.org> --- drivers/thermal/qcom/tsens-common.c | 23 ++++++++++++++++++++++- 1 file changed, 22 insertions(+), 1 deletion(-)