Message ID | 20221108075051.5139-4-rui.zhang@intel.com (mailing list archive) |
---|---|
State | Changes Requested |
Headers | show |
Series | hwmon (coretemp): Add support for dynamic tjmax/ttarget | expand |
On Tue, Nov 08, 2022 at 03:50:51PM +0800, Zhang Rui wrote: > Tjmax value retrieved from MSR_IA32_TEMPERATURE_TARGET can be changed at > runtime when the Intel SST-PP (Intel Speed Select Technology - > Performance Profile) level is changed. As a result, the ttarget value > also becomes dyamic. > > Improve the code to always get updated ttarget value. > > Signed-off-by: Zhang Rui <rui.zhang@intel.com> > --- > drivers/hwmon/coretemp.c | 69 ++++++++++++++++++++++++++++------------ > 1 file changed, 48 insertions(+), 21 deletions(-) > > diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c > index 5292f7844860..d6084600862f 100644 > --- a/drivers/hwmon/coretemp.c > +++ b/drivers/hwmon/coretemp.c > @@ -69,7 +69,6 @@ MODULE_PARM_DESC(tjmax, "TjMax value in degrees Celsius"); > */ > struct temp_data { > int temp; > - int ttarget; > int tjmax; > unsigned long last_updated; > unsigned int cpu; > @@ -96,6 +95,7 @@ struct platform_data { > }; > > static int get_tjmax(struct temp_data *tdata, struct device *dev); > +static int get_ttarget(struct temp_data *tdata, struct device *dev); Please rearrange to code to avoid forward declarations. > > /* Keep track of how many zone pointers we allocated in init() */ > static int max_zones __read_mostly; > @@ -150,8 +150,17 @@ static ssize_t show_ttarget(struct device *dev, > { > struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr); > struct platform_data *pdata = dev_get_drvdata(dev); > + struct temp_data *tdata = pdata->core_data[attr->index]; > + int ttarget; > + > + mutex_lock(&tdata->update_lock); Is that mutex really necessary ? I don't immediately see the need for it. > + ttarget = get_ttarget(tdata, dev); > + mutex_unlock(&tdata->update_lock); > > - return sprintf(buf, "%d\n", pdata->core_data[attr->index]->ttarget); > + if (ttarget >= 0) > + return sprintf(buf, "%d\n", ttarget); > + else > + return ttarget; else after return is unnecessary. Much better would be if (ttarget < 0) return ttarget; return sprintf(buf, "%d\n", ttarget); > } > > static ssize_t show_temp(struct device *dev, > @@ -393,6 +402,38 @@ static int get_tjmax(struct temp_data *tdata, struct device *dev) > return tdata->tjmax; > } > > +static int get_ttarget(struct temp_data *tdata, struct device *dev) > +{ > + u32 eax, edx; > + struct cpuinfo_x86 *c = &cpu_data(tdata->cpu); > + int tj_max, ttarget_offset, ret; Please use tjmax for consistency. > + > + /* > + * ttarget is valid only if tjmax can be retrieved from > + * MSR_IA32_TEMPERATURE_TARGET > + */ > + if (tdata->tjmax) > + return -ENODEV; > + > + if (c->x86_model <= 0xe || c->x86_model == 0x1c) > + return -ENODEV; > + Does it really make sense to re-check this each time the target temperature is read ? > + /* > + * Read the still undocumented bits 8:15 of IA32_TEMPERATURE_TARGET. > + * The target temperature is available on older CPUs but not in this > + * register. Atoms don't have the register at all. > + */ > + ret = rdmsr_safe_on_cpu(tdata->cpu, MSR_IA32_TEMPERATURE_TARGET, > + &eax, &edx); Please watch multi-line alignment. > + if (ret) > + return ret; > + > + tj_max = (eax >> 16) & 0xff; > + ttarget_offset = (eax >> 8) & 0xff; > + > + return (tj_max - ttarget_offset) * 1000; > +} > + > static int create_core_attrs(struct temp_data *tdata, struct device *dev, > int attr_no) > { > @@ -468,9 +509,8 @@ static int create_core_data(struct platform_device *pdev, unsigned int cpu, > { > struct temp_data *tdata; > struct platform_data *pdata = platform_get_drvdata(pdev); > - struct cpuinfo_x86 *c = &cpu_data(cpu); > u32 eax, edx; > - int err, index, attr_no, tjmax; > + int err, index, attr_no; > > /* > * Find attr number for sysfs: > @@ -504,23 +544,10 @@ static int create_core_data(struct platform_device *pdev, unsigned int cpu, > if (err) > goto exit_free; > > - /* We can access status register. Get Critical Temperature */ > - tjmax = get_tjmax(tdata, &pdev->dev); > - > - /* > - * Read the still undocumented bits 8:15 of IA32_TEMPERATURE_TARGET. > - * The target temperature is available on older CPUs but not in this > - * register. Atoms don't have the register at all. > - */ > - if (c->x86_model > 0xe && c->x86_model != 0x1c) { > - err = rdmsr_safe_on_cpu(cpu, MSR_IA32_TEMPERATURE_TARGET, > - &eax, &edx); > - if (!err) { > - tdata->ttarget > - = tjmax - ((eax >> 8) & 0xff) * 1000; > - tdata->attr_size++; > - } > - } > + /* Make sure tdata->tjmax is a valid indicator for dynamic/static tjmax */ > + get_tjmax(tdata, &pdev->dev); > + if (get_ttarget(tdata, &pdev->dev) >= 0) > + tdata->attr_size++; > > pdata->core_data[attr_no] = tdata; >
Hi, Guenter, Thanks for reviewing the patches. I will address the comments in next version. And also some comments below. On Fri, 2022-11-11 at 13:34 -0800, Guenter Roeck wrote: > > > > /* Keep track of how many zone pointers we allocated in init() */ > > static int max_zones __read_mostly; > > @@ -150,8 +150,17 @@ static ssize_t show_ttarget(struct device > > *dev, > > { > > struct sensor_device_attribute *attr = > > to_sensor_dev_attr(devattr); > > struct platform_data *pdata = dev_get_drvdata(dev); > > + struct temp_data *tdata = pdata->core_data[attr->index]; > > + int ttarget; > > + > > + mutex_lock(&tdata->update_lock); > > Is that mutex really necessary ? I don't immediately see the need > for it. I just followed the same pattern as show_crit_alarm(). I checked the history and it was introduced by commit 723f573433b2 ("hwmon: (coretemp) Fixup target cpu for package when cpu is offlined"), to make sure the msr is not running on an offlined cpu. > > + > > + /* > > + * ttarget is valid only if tjmax can be retrieved from > > + * MSR_IA32_TEMPERATURE_TARGET > > + */ > > + if (tdata->tjmax) > > + return -ENODEV; > > + > > + if (c->x86_model <= 0xe || c->x86_model == 0x1c) > > + return -ENODEV; > > + > > Does it really make sense to re-check this each time the target > temperature > is read ? You're right. We can keep this as it was, when deciding whether to create this sysfs attr or not. Then the check in get_ttarget() can be removed. thanks, rui
On Sat, Nov 12, 2022 at 04:37:50PM +0800, Zhang Rui wrote: > Hi, Guenter, > > Thanks for reviewing the patches. > I will address the comments in next version. And also some comments > below. > > On Fri, 2022-11-11 at 13:34 -0800, Guenter Roeck wrote: > > > > > > /* Keep track of how many zone pointers we allocated in init() */ > > > static int max_zones __read_mostly; > > > @@ -150,8 +150,17 @@ static ssize_t show_ttarget(struct device > > > *dev, > > > { > > > struct sensor_device_attribute *attr = > > > to_sensor_dev_attr(devattr); > > > struct platform_data *pdata = dev_get_drvdata(dev); > > > + struct temp_data *tdata = pdata->core_data[attr->index]; > > > + int ttarget; > > > + > > > + mutex_lock(&tdata->update_lock); > > > > Is that mutex really necessary ? I don't immediately see the need > > for it. > > I just followed the same pattern as show_crit_alarm(). > I checked the history and it was introduced by commit 723f573433b2 > ("hwmon: (coretemp) Fixup target cpu for package when cpu is > offlined"), to make sure the msr is not running on an offlined cpu. > Good point. I am not sure if it matters at that point if the code uses the old or the new CPU, but I guess it is safer. Thanks for the clarification, Guenter > > > + > > > + /* > > > + * ttarget is valid only if tjmax can be retrieved from > > > + * MSR_IA32_TEMPERATURE_TARGET > > > + */ > > > + if (tdata->tjmax) > > > + return -ENODEV; > > > + > > > + if (c->x86_model <= 0xe || c->x86_model == 0x1c) > > > + return -ENODEV; > > > + > > > > Does it really make sense to re-check this each time the target > > temperature > > is read ? > > You're right. We can keep this as it was, when deciding whether to > create this sysfs attr or not. Then the check in get_ttarget() can be > removed. > > thanks, > rui >
diff --git a/drivers/hwmon/coretemp.c b/drivers/hwmon/coretemp.c index 5292f7844860..d6084600862f 100644 --- a/drivers/hwmon/coretemp.c +++ b/drivers/hwmon/coretemp.c @@ -69,7 +69,6 @@ MODULE_PARM_DESC(tjmax, "TjMax value in degrees Celsius"); */ struct temp_data { int temp; - int ttarget; int tjmax; unsigned long last_updated; unsigned int cpu; @@ -96,6 +95,7 @@ struct platform_data { }; static int get_tjmax(struct temp_data *tdata, struct device *dev); +static int get_ttarget(struct temp_data *tdata, struct device *dev); /* Keep track of how many zone pointers we allocated in init() */ static int max_zones __read_mostly; @@ -150,8 +150,17 @@ static ssize_t show_ttarget(struct device *dev, { struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr); struct platform_data *pdata = dev_get_drvdata(dev); + struct temp_data *tdata = pdata->core_data[attr->index]; + int ttarget; + + mutex_lock(&tdata->update_lock); + ttarget = get_ttarget(tdata, dev); + mutex_unlock(&tdata->update_lock); - return sprintf(buf, "%d\n", pdata->core_data[attr->index]->ttarget); + if (ttarget >= 0) + return sprintf(buf, "%d\n", ttarget); + else + return ttarget; } static ssize_t show_temp(struct device *dev, @@ -393,6 +402,38 @@ static int get_tjmax(struct temp_data *tdata, struct device *dev) return tdata->tjmax; } +static int get_ttarget(struct temp_data *tdata, struct device *dev) +{ + u32 eax, edx; + struct cpuinfo_x86 *c = &cpu_data(tdata->cpu); + int tj_max, ttarget_offset, ret; + + /* + * ttarget is valid only if tjmax can be retrieved from + * MSR_IA32_TEMPERATURE_TARGET + */ + if (tdata->tjmax) + return -ENODEV; + + if (c->x86_model <= 0xe || c->x86_model == 0x1c) + return -ENODEV; + + /* + * Read the still undocumented bits 8:15 of IA32_TEMPERATURE_TARGET. + * The target temperature is available on older CPUs but not in this + * register. Atoms don't have the register at all. + */ + ret = rdmsr_safe_on_cpu(tdata->cpu, MSR_IA32_TEMPERATURE_TARGET, + &eax, &edx); + if (ret) + return ret; + + tj_max = (eax >> 16) & 0xff; + ttarget_offset = (eax >> 8) & 0xff; + + return (tj_max - ttarget_offset) * 1000; +} + static int create_core_attrs(struct temp_data *tdata, struct device *dev, int attr_no) { @@ -468,9 +509,8 @@ static int create_core_data(struct platform_device *pdev, unsigned int cpu, { struct temp_data *tdata; struct platform_data *pdata = platform_get_drvdata(pdev); - struct cpuinfo_x86 *c = &cpu_data(cpu); u32 eax, edx; - int err, index, attr_no, tjmax; + int err, index, attr_no; /* * Find attr number for sysfs: @@ -504,23 +544,10 @@ static int create_core_data(struct platform_device *pdev, unsigned int cpu, if (err) goto exit_free; - /* We can access status register. Get Critical Temperature */ - tjmax = get_tjmax(tdata, &pdev->dev); - - /* - * Read the still undocumented bits 8:15 of IA32_TEMPERATURE_TARGET. - * The target temperature is available on older CPUs but not in this - * register. Atoms don't have the register at all. - */ - if (c->x86_model > 0xe && c->x86_model != 0x1c) { - err = rdmsr_safe_on_cpu(cpu, MSR_IA32_TEMPERATURE_TARGET, - &eax, &edx); - if (!err) { - tdata->ttarget - = tjmax - ((eax >> 8) & 0xff) * 1000; - tdata->attr_size++; - } - } + /* Make sure tdata->tjmax is a valid indicator for dynamic/static tjmax */ + get_tjmax(tdata, &pdev->dev); + if (get_ttarget(tdata, &pdev->dev) >= 0) + tdata->attr_size++; pdata->core_data[attr_no] = tdata;
Tjmax value retrieved from MSR_IA32_TEMPERATURE_TARGET can be changed at runtime when the Intel SST-PP (Intel Speed Select Technology - Performance Profile) level is changed. As a result, the ttarget value also becomes dyamic. Improve the code to always get updated ttarget value. Signed-off-by: Zhang Rui <rui.zhang@intel.com> --- drivers/hwmon/coretemp.c | 69 ++++++++++++++++++++++++++++------------ 1 file changed, 48 insertions(+), 21 deletions(-)