From: Andreas Herrmann <[email protected]>
This CPU family provides NB register values to gather following
TDP information
* ProcessorPwrWatts: Specifies in Watts the maximum amount of power
the processor can support.
* CurrPwrWatts: Specifies in Watts the current amount of power being
consumed by the processor.
This driver provides
* power1_cap (ProcessorPwrWatts)
* power1_input (CurrPwrWatts)
Changes from v1::
- removed deprecated code line,
- fixed comment,
- report power only once per socket (Power information is provided
for the entire package. On multi-node processors it should only be
reported on internal node 0.)
Signed-off-by: Andreas Herrmann <[email protected]>
---
drivers/hwmon/Kconfig | 10 ++
drivers/hwmon/Makefile | 1 +
drivers/hwmon/f15h_power.c | 216 ++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 227 insertions(+), 0 deletions(-)
create mode 100644 drivers/hwmon/f15h_power.c
diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
index 8e84b31..0268623 100644
--- a/drivers/hwmon/Kconfig
+++ b/drivers/hwmon/Kconfig
@@ -249,6 +249,16 @@ config SENSORS_K10TEMP
This driver can also be built as a module. If so, the module
will be called k10temp.
+config SENSORS_F15H_POWER
+ tristate "AMD Family 15h processor power"
+ depends on X86 && PCI
+ help
+ If you say yes here you get support for processor power
+ information of your AMD family 15h CPU.
+
+ This driver can also be built as a module. If so, the module
+ will be called f15hpower.
+
config SENSORS_ASB100
tristate "Asus ASB100 Bach"
depends on X86 && I2C && EXPERIMENTAL
diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
index dde02d9..e554b88 100644
--- a/drivers/hwmon/Makefile
+++ b/drivers/hwmon/Makefile
@@ -62,6 +62,7 @@ obj-$(CONFIG_SENSORS_JC42) += jc42.o
obj-$(CONFIG_SENSORS_JZ4740) += jz4740-hwmon.o
obj-$(CONFIG_SENSORS_K8TEMP) += k8temp.o
obj-$(CONFIG_SENSORS_K10TEMP) += k10temp.o
+obj-$(CONFIG_SENSORS_F15H_POWER) += f15h_power.o
obj-$(CONFIG_SENSORS_LIS3LV02D) += lis3lv02d.o hp_accel.o
obj-$(CONFIG_SENSORS_LIS3_SPI) += lis3lv02d.o lis3lv02d_spi.o
obj-$(CONFIG_SENSORS_LIS3_I2C) += lis3lv02d.o lis3lv02d_i2c.o
diff --git a/drivers/hwmon/f15h_power.c b/drivers/hwmon/f15h_power.c
new file mode 100644
index 0000000..40e3cbd
--- /dev/null
+++ b/drivers/hwmon/f15h_power.c
@@ -0,0 +1,216 @@
+/*
+ * f15h_power.c - AMD Family 15h processor power monitoring
+ *
+ * Copyright (c) 2011 Advanced Micro Devices, Inc.
+ * Author: Andreas Herrmann <[email protected]>
+ *
+ *
+ * This driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this driver; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/err.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/bitops.h>
+#include <asm/processor.h>
+
+MODULE_DESCRIPTION("AMD Family 15h CPU processor power monitor");
+MODULE_AUTHOR("Andreas Herrmann <[email protected]>");
+MODULE_LICENSE("GPL");
+
+/* D18F3 */
+#define REG_NORTHBRIDGE_CAP 0xe8
+
+/* D18F4 */
+#define REG_PROCESSOR_TDP 0x1b8
+
+/* D18F5 */
+#define REG_TDP_RUNNING_AVERAGE 0xe0
+#define REG_TDP_LIMIT3 0xe8
+
+static ssize_t show_power(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ u32 val, btdp, tdpl, tdp2w, arange;
+ s32 acap;
+ u64 ctdp;
+ struct pci_dev *f4 = to_pci_dev(dev);
+ struct pci_dev *f5;
+
+ pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
+ btdp = (val >> 16) & 0xffff;
+
+ f5 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5));
+ if (!f5) {
+ dev_err(dev, "no function 5 available on this slot\n");
+ return 0;
+ }
+
+ pci_read_config_dword(f5, REG_TDP_RUNNING_AVERAGE, &val);
+ acap = (val >> 4) & 0x3fffff;
+ acap = sign_extend32(acap, 22);
+ arange = val & 0xf;
+
+ pci_read_config_dword(f5, REG_TDP_LIMIT3, &val);
+ pci_dev_put(f5);
+
+ tdpl = (val >> 16) & 0x1fff;
+ tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
+ ctdp = tdpl - (s32)(acap >> (arange + 1)) + btdp;
+ ctdp *= tdp2w;
+
+ /*
+ * Convert to microWatt
+ *
+ * power is in Watt provided as fixed point integer with
+ * scaling factor 1/(2^16). For conversion we use
+ * (10^6)/(2^16) = 15625/(2^10)
+ */
+ ctdp = (ctdp * 15625) >> 10;
+ return sprintf(buf, "%d\n", (u32) ctdp);
+}
+static DEVICE_ATTR(power1_input, S_IRUGO, show_power, NULL);
+
+static ssize_t show_power_cap(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ u32 val, tdp2w;
+ u64 ptdp;
+ struct pci_dev *f4 = to_pci_dev(dev);
+ struct pci_dev *f5;
+
+ pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
+ ptdp = val & 0xffff;
+
+ f5 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5));
+ if (!f5) {
+ dev_err(dev, "no function 5 available on this slot\n");
+ return 0;
+ }
+
+ pci_read_config_dword(f5, REG_TDP_LIMIT3, &val);
+ pci_dev_put(f5);
+
+ tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
+ ptdp *= tdp2w;
+
+ /* result not allowed to be >= 256W */
+ WARN_ON(ptdp>>16 >= 256);
+
+ /* convert to microWatt */
+ ptdp = (ptdp * 15625) >> 10;
+ return sprintf(buf, "%d\n", (u32) ptdp);
+}
+static DEVICE_ATTR(power1_cap, S_IRUGO, show_power_cap, NULL);
+
+static ssize_t show_name(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ return sprintf(buf, "f15h_power\n");
+}
+static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
+
+static int __devinit f15h_power_is_internal_node0(struct pci_dev *f4)
+{
+ u32 val;
+ struct pci_dev *f3;
+
+ f3 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3));
+ if (!f3) {
+ dev_err(&f4->dev, "no function 3 available on this slot\n");
+ return 0;
+ }
+
+ pci_read_config_dword(f3, REG_NORTHBRIDGE_CAP, &val);
+ pci_dev_put(f3);
+
+ if ((val & BIT(29)) && ((val >> 30) & 3))
+ return 0;
+
+ return 1;
+}
+
+static int __devinit f15h_power_probe(struct pci_dev *pdev,
+ const struct pci_device_id *id)
+{
+ struct device *hwmon_dev;
+ int err = -ENODEV;
+
+ if (!f15h_power_is_internal_node0(pdev))
+ goto exit;
+
+ err = device_create_file(&pdev->dev, &dev_attr_power1_input);
+ if (err)
+ goto exit;
+ err = device_create_file(&pdev->dev, &dev_attr_power1_cap);
+ if (err)
+ goto exit_remove;
+
+ err = device_create_file(&pdev->dev, &dev_attr_name);
+ if (err)
+ goto exit_remove;
+
+ hwmon_dev = hwmon_device_register(&pdev->dev);
+ if (IS_ERR(hwmon_dev)) {
+ err = PTR_ERR(hwmon_dev);
+ goto exit_remove;
+ }
+ dev_set_drvdata(&pdev->dev, hwmon_dev);
+
+ return 0;
+
+exit_remove:
+ device_remove_file(&pdev->dev, &dev_attr_name);
+ device_remove_file(&pdev->dev, &dev_attr_power1_input);
+ device_remove_file(&pdev->dev, &dev_attr_power1_cap);
+exit:
+ return err;
+}
+
+static void __devexit f15h_power_remove(struct pci_dev *pdev)
+{
+ hwmon_device_unregister(dev_get_drvdata(&pdev->dev));
+ device_remove_file(&pdev->dev, &dev_attr_name);
+ device_remove_file(&pdev->dev, &dev_attr_power1_input);
+ device_remove_file(&pdev->dev, &dev_attr_power1_cap);
+ dev_set_drvdata(&pdev->dev, NULL);
+}
+
+static const struct pci_device_id f15h_power_id_table[] = {
+ { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_LINK) },
+ {}
+};
+MODULE_DEVICE_TABLE(pci, f15h_power_id_table);
+
+static struct pci_driver f15h_power_driver = {
+ .name = "f15h_power",
+ .id_table = f15h_power_id_table,
+ .probe = f15h_power_probe,
+ .remove = __devexit_p(f15h_power_remove),
+};
+
+static int __init f15h_power_init(void)
+{
+ return pci_register_driver(&f15h_power_driver);
+}
+
+static void __exit f15h_power_exit(void)
+{
+ pci_unregister_driver(&f15h_power_driver);
+}
+
+module_init(f15h_power_init)
+module_exit(f15h_power_exit)
--
1.7.3.1
On Tue, Apr 05, 2011 at 10:45:36AM -0400, Andreas Herrmann wrote:
> From: Andreas Herrmann <[email protected]>
>
> This CPU family provides NB register values to gather following
> TDP information
>
> * ProcessorPwrWatts: Specifies in Watts the maximum amount of power
> the processor can support.
> * CurrPwrWatts: Specifies in Watts the current amount of power being
> consumed by the processor.
>
> This driver provides
>
> * power1_cap (ProcessorPwrWatts)
Hi Andreas,
What does the CPU do if ProcessorPwrWatts is reached ? Does it start to limit
power consumption, ie does it enforce the limit ?
If not you might want to use power1_max instead.
> * power1_input (CurrPwrWatts)
>
> Changes from v1::
> - removed deprecated code line,
> - fixed comment,
> - report power only once per socket (Power information is provided
> for the entire package. On multi-node processors it should only be
> reported on internal node 0.)
>
> Signed-off-by: Andreas Herrmann <[email protected]>
> ---
> drivers/hwmon/Kconfig | 10 ++
> drivers/hwmon/Makefile | 1 +
> drivers/hwmon/f15h_power.c | 216 ++++++++++++++++++++++++++++++++++++++++++++
Please also add Documentation/hwmon/f15h_power.
> 3 files changed, 227 insertions(+), 0 deletions(-)
> create mode 100644 drivers/hwmon/f15h_power.c
>
> diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
> index 8e84b31..0268623 100644
> --- a/drivers/hwmon/Kconfig
> +++ b/drivers/hwmon/Kconfig
> @@ -249,6 +249,16 @@ config SENSORS_K10TEMP
> This driver can also be built as a module. If so, the module
> will be called k10temp.
>
> +config SENSORS_F15H_POWER
> + tristate "AMD Family 15h processor power"
> + depends on X86 && PCI
> + help
> + If you say yes here you get support for processor power
> + information of your AMD family 15h CPU.
> +
> + This driver can also be built as a module. If so, the module
> + will be called f15hpower.
f15h_power
> +
> config SENSORS_ASB100
> tristate "Asus ASB100 Bach"
> depends on X86 && I2C && EXPERIMENTAL
> diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
> index dde02d9..e554b88 100644
> --- a/drivers/hwmon/Makefile
> +++ b/drivers/hwmon/Makefile
> @@ -62,6 +62,7 @@ obj-$(CONFIG_SENSORS_JC42) += jc42.o
> obj-$(CONFIG_SENSORS_JZ4740) += jz4740-hwmon.o
> obj-$(CONFIG_SENSORS_K8TEMP) += k8temp.o
> obj-$(CONFIG_SENSORS_K10TEMP) += k10temp.o
> +obj-$(CONFIG_SENSORS_F15H_POWER) += f15h_power.o
> obj-$(CONFIG_SENSORS_LIS3LV02D) += lis3lv02d.o hp_accel.o
> obj-$(CONFIG_SENSORS_LIS3_SPI) += lis3lv02d.o lis3lv02d_spi.o
> obj-$(CONFIG_SENSORS_LIS3_I2C) += lis3lv02d.o lis3lv02d_i2c.o
> diff --git a/drivers/hwmon/f15h_power.c b/drivers/hwmon/f15h_power.c
> new file mode 100644
> index 0000000..40e3cbd
> --- /dev/null
> +++ b/drivers/hwmon/f15h_power.c
> @@ -0,0 +1,216 @@
> +/*
> + * f15h_power.c - AMD Family 15h processor power monitoring
> + *
> + * Copyright (c) 2011 Advanced Micro Devices, Inc.
> + * Author: Andreas Herrmann <[email protected]>
> + *
> + *
> + * This driver is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This driver is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
> + * See the GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this driver; if not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include <linux/err.h>
> +#include <linux/hwmon.h>
> +#include <linux/hwmon-sysfs.h>
> +#include <linux/init.h>
> +#include <linux/module.h>
> +#include <linux/pci.h>
> +#include <linux/bitops.h>
> +#include <asm/processor.h>
> +
> +MODULE_DESCRIPTION("AMD Family 15h CPU processor power monitor");
> +MODULE_AUTHOR("Andreas Herrmann <[email protected]>");
> +MODULE_LICENSE("GPL");
> +
> +/* D18F3 */
> +#define REG_NORTHBRIDGE_CAP 0xe8
> +
> +/* D18F4 */
> +#define REG_PROCESSOR_TDP 0x1b8
> +
> +/* D18F5 */
> +#define REG_TDP_RUNNING_AVERAGE 0xe0
> +#define REG_TDP_LIMIT3 0xe8
> +
> +static ssize_t show_power(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + u32 val, btdp, tdpl, tdp2w, arange;
> + s32 acap;
> + u64 ctdp;
> + struct pci_dev *f4 = to_pci_dev(dev);
> + struct pci_dev *f5;
> +
> + pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
> + btdp = (val >> 16) & 0xffff;
> +
> + f5 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5));
> + if (!f5) {
> + dev_err(dev, "no function 5 available on this slot\n");
> + return 0;
> + }
Unless this is a temporary error condition, you should check if this function
exists in the probe function, and not create a sysfs entry if it doesn't.
If it is a temporary error which needs a runtime check, you should return
a negative value to indicate that there was an error.
> +
> + pci_read_config_dword(f5, REG_TDP_RUNNING_AVERAGE, &val);
> + acap = (val >> 4) & 0x3fffff;
> + acap = sign_extend32(acap, 22);
> + arange = val & 0xf;
> +
> + pci_read_config_dword(f5, REG_TDP_LIMIT3, &val);
> + pci_dev_put(f5);
> +
> + tdpl = (val >> 16) & 0x1fff;
> + tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
> + ctdp = tdpl - (s32)(acap >> (arange + 1)) + btdp;
> + ctdp *= tdp2w;
> +
> + /*
> + * Convert to microWatt
> + *
> + * power is in Watt provided as fixed point integer with
> + * scaling factor 1/(2^16). For conversion we use
> + * (10^6)/(2^16) = 15625/(2^10)
> + */
> + ctdp = (ctdp * 15625) >> 10;
> + return sprintf(buf, "%d\n", (u32) ctdp);
%u
> +}
> +static DEVICE_ATTR(power1_input, S_IRUGO, show_power, NULL);
> +
> +static ssize_t show_power_cap(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + u32 val, tdp2w;
> + u64 ptdp;
> + struct pci_dev *f4 = to_pci_dev(dev);
> + struct pci_dev *f5;
> +
> + pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
> + ptdp = val & 0xffff;
> +
> + f5 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5));
> + if (!f5) {
> + dev_err(dev, "no function 5 available on this slot\n");
> + return 0;
> + }
> +
Same as above - should be checked in probe function.
If it is a temporary error which needs a runtime check, you should return
a negative value to indicate that there was an error.
> + pci_read_config_dword(f5, REG_TDP_LIMIT3, &val);
> + pci_dev_put(f5);
> +
> + tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
> + ptdp *= tdp2w;
> +
> + /* result not allowed to be >= 256W */
> + WARN_ON(ptdp>>16 >= 256);
> +
Does this really ask for such drastic measures, or would dev_warn() be sufficient ?
> + /* convert to microWatt */
> + ptdp = (ptdp * 15625) >> 10;
> + return sprintf(buf, "%d\n", (u32) ptdp);
%u
> +}
> +static DEVICE_ATTR(power1_cap, S_IRUGO, show_power_cap, NULL);
> +
> +static ssize_t show_name(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + return sprintf(buf, "f15h_power\n");
> +}
> +static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
> +
> +static int __devinit f15h_power_is_internal_node0(struct pci_dev *f4)
> +{
> + u32 val;
> + struct pci_dev *f3;
> +
> + f3 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3));
> + if (!f3) {
> + dev_err(&f4->dev, "no function 3 available on this slot\n");
> + return 0;
It is a common practice to return a negative value on errors. Why not here ?
Also, is this really an error which asks for an error message, or just a CPU
or system which does not support the attribute ? In the latter case, you should
not display an error message.
> + }
> +
> + pci_read_config_dword(f3, REG_NORTHBRIDGE_CAP, &val);
> + pci_dev_put(f3);
> +
> + if ((val & BIT(29)) && ((val >> 30) & 3))
> + return 0;
> +
> + return 1;
> +}
> +
> +static int __devinit f15h_power_probe(struct pci_dev *pdev,
> + const struct pci_device_id *id)
> +{
> + struct device *hwmon_dev;
> + int err = -ENODEV;
> +
> + if (!f15h_power_is_internal_node0(pdev))
> + goto exit;
> +
> + err = device_create_file(&pdev->dev, &dev_attr_power1_input);
> + if (err)
> + goto exit;
> + err = device_create_file(&pdev->dev, &dev_attr_power1_cap);
> + if (err)
> + goto exit_remove;
> +
> + err = device_create_file(&pdev->dev, &dev_attr_name);
> + if (err)
> + goto exit_remove;
> +
> + hwmon_dev = hwmon_device_register(&pdev->dev);
> + if (IS_ERR(hwmon_dev)) {
> + err = PTR_ERR(hwmon_dev);
> + goto exit_remove;
> + }
> + dev_set_drvdata(&pdev->dev, hwmon_dev);
> +
> + return 0;
> +
> +exit_remove:
> + device_remove_file(&pdev->dev, &dev_attr_name);
> + device_remove_file(&pdev->dev, &dev_attr_power1_input);
> + device_remove_file(&pdev->dev, &dev_attr_power1_cap);
> +exit:
> + return err;
> +}
> +
> +static void __devexit f15h_power_remove(struct pci_dev *pdev)
> +{
> + hwmon_device_unregister(dev_get_drvdata(&pdev->dev));
> + device_remove_file(&pdev->dev, &dev_attr_name);
> + device_remove_file(&pdev->dev, &dev_attr_power1_input);
> + device_remove_file(&pdev->dev, &dev_attr_power1_cap);
> + dev_set_drvdata(&pdev->dev, NULL);
> +}
> +
> +static const struct pci_device_id f15h_power_id_table[] = {
checkpatch says:
WARNING: Use DEFINE_PCI_DEVICE_TABLE for struct pci_device_id
#326: FILE: drivers/hwmon/f15h_power.c:192:
Please fix.
> + { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_LINK) },
> + {}
> +};
> +MODULE_DEVICE_TABLE(pci, f15h_power_id_table);
> +
> +static struct pci_driver f15h_power_driver = {
> + .name = "f15h_power",
> + .id_table = f15h_power_id_table,
> + .probe = f15h_power_probe,
> + .remove = __devexit_p(f15h_power_remove),
> +};
> +
> +static int __init f15h_power_init(void)
> +{
> + return pci_register_driver(&f15h_power_driver);
> +}
> +
> +static void __exit f15h_power_exit(void)
> +{
> + pci_unregister_driver(&f15h_power_driver);
> +}
> +
> +module_init(f15h_power_init)
> +module_exit(f15h_power_exit)
> --
> 1.7.3.1
>
Guenter Roeck wrote:
> On Tue, Apr 05, 2011 at 10:45:36AM -0400, Andreas Herrmann wrote:
> > +static int __devinit f15h_power_is_internal_node0(struct pci_dev *f4)
> > +{
> > + u32 val;
> > + struct pci_dev *f3;
> > +
> > + f3 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3));
> > + if (!f3) {
> > + dev_err(&f4->dev, "no function 3 available on this slot\n");
> > + return 0;
>
> It is a common practice to return a negative value on errors. Why not here ?
Apparently, this function returns a boolean value. Using "bool"/"true"/
"false" would have made this more obvious.
> Also, is this really an error which asks for an error message, or just a CPU
> or system which does not support the attribute ?
AFAICT all F15h CPUs are _known_ to have all these PCI functions; the
error should never occur in practice.
What I do in the k10temp driver in this situation is to trust the CPU to
be there, omitting the pci_get_slot/pci_dev_put and just replacing this:
> > + pci_read_config_dword(f3, REG_NORTHBRIDGE_CAP, &val);
> > + pci_dev_put(f3);
with the equivalent of:
pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3),
REG_NORTHBRIDGE_CAP, &val);
BTW: The family 15h CPUs have the same temperature sensor registers
(D18F3xA4 and D18F3x64) as the earlier families, haven't they?
Regards,
Clemens
On Tue, Apr 05, 2011 at 01:12:38PM -0700, Guenter Roeck wrote:
> On Tue, Apr 05, 2011 at 10:45:36AM -0400, Andreas Herrmann wrote:
> > From: Andreas Herrmann <[email protected]>
> >
> > This CPU family provides NB register values to gather following
> > TDP information
> >
> > * ProcessorPwrWatts: Specifies in Watts the maximum amount of power
> > the processor can support.
> > * CurrPwrWatts: Specifies in Watts the current amount of power being
> > consumed by the processor.
> >
> > This driver provides
> >
> > * power1_cap (ProcessorPwrWatts)
>
> Hi Andreas,
>
> What does the CPU do if ProcessorPwrWatts is reached ? Does it start to limit
> power consumption, ie does it enforce the limit ?
No.
> If not you might want to use power1_max instead.
Ok, so I picked the wrong attribute. Will correct this.
> > * power1_input (CurrPwrWatts)
> >
> > Changes from v1::
> > - removed deprecated code line,
> > - fixed comment,
> > - report power only once per socket (Power information is provided
> > for the entire package. On multi-node processors it should only be
> > reported on internal node 0.)
> >
> > Signed-off-by: Andreas Herrmann <[email protected]>
> > ---
> > drivers/hwmon/Kconfig | 10 ++
> > drivers/hwmon/Makefile | 1 +
> > drivers/hwmon/f15h_power.c | 216 ++++++++++++++++++++++++++++++++++++++++++++
>
> Please also add Documentation/hwmon/f15h_power.
Ok.
> > 3 files changed, 227 insertions(+), 0 deletions(-)
> > create mode 100644 drivers/hwmon/f15h_power.c
> >
> > diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
> > index 8e84b31..0268623 100644
> > --- a/drivers/hwmon/Kconfig
> > +++ b/drivers/hwmon/Kconfig
> > @@ -249,6 +249,16 @@ config SENSORS_K10TEMP
> > This driver can also be built as a module. If so, the module
> > will be called k10temp.
> >
> > +config SENSORS_F15H_POWER
> > + tristate "AMD Family 15h processor power"
> > + depends on X86 && PCI
> > + help
> > + If you say yes here you get support for processor power
> > + information of your AMD family 15h CPU.
> > +
> > + This driver can also be built as a module. If so, the module
> > + will be called f15hpower.
>
> f15h_power
Good catch.
> > +
> > config SENSORS_ASB100
> > tristate "Asus ASB100 Bach"
> > depends on X86 && I2C && EXPERIMENTAL
> > diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
> > index dde02d9..e554b88 100644
> > --- a/drivers/hwmon/Makefile
> > +++ b/drivers/hwmon/Makefile
> > @@ -62,6 +62,7 @@ obj-$(CONFIG_SENSORS_JC42) += jc42.o
> > obj-$(CONFIG_SENSORS_JZ4740) += jz4740-hwmon.o
> > obj-$(CONFIG_SENSORS_K8TEMP) += k8temp.o
> > obj-$(CONFIG_SENSORS_K10TEMP) += k10temp.o
> > +obj-$(CONFIG_SENSORS_F15H_POWER) += f15h_power.o
> > obj-$(CONFIG_SENSORS_LIS3LV02D) += lis3lv02d.o hp_accel.o
> > obj-$(CONFIG_SENSORS_LIS3_SPI) += lis3lv02d.o lis3lv02d_spi.o
> > obj-$(CONFIG_SENSORS_LIS3_I2C) += lis3lv02d.o lis3lv02d_i2c.o
> > diff --git a/drivers/hwmon/f15h_power.c b/drivers/hwmon/f15h_power.c
> > new file mode 100644
> > index 0000000..40e3cbd
> > --- /dev/null
> > +++ b/drivers/hwmon/f15h_power.c
> > @@ -0,0 +1,216 @@
> > +/*
> > + * f15h_power.c - AMD Family 15h processor power monitoring
> > + *
> > + * Copyright (c) 2011 Advanced Micro Devices, Inc.
> > + * Author: Andreas Herrmann <[email protected]>
> > + *
> > + *
> > + * This driver is free software; you can redistribute it and/or
> > + * modify it under the terms of the GNU General Public License; either
> > + * version 2 of the License, or (at your option) any later version.
> > + *
> > + * This driver is distributed in the hope that it will be useful,
> > + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
> > + * See the GNU General Public License for more details.
> > + *
> > + * You should have received a copy of the GNU General Public License
> > + * along with this driver; if not, see <http://www.gnu.org/licenses/>.
> > + */
> > +
> > +#include <linux/err.h>
> > +#include <linux/hwmon.h>
> > +#include <linux/hwmon-sysfs.h>
> > +#include <linux/init.h>
> > +#include <linux/module.h>
> > +#include <linux/pci.h>
> > +#include <linux/bitops.h>
> > +#include <asm/processor.h>
> > +
> > +MODULE_DESCRIPTION("AMD Family 15h CPU processor power monitor");
> > +MODULE_AUTHOR("Andreas Herrmann <[email protected]>");
> > +MODULE_LICENSE("GPL");
> > +
> > +/* D18F3 */
> > +#define REG_NORTHBRIDGE_CAP 0xe8
> > +
> > +/* D18F4 */
> > +#define REG_PROCESSOR_TDP 0x1b8
> > +
> > +/* D18F5 */
> > +#define REG_TDP_RUNNING_AVERAGE 0xe0
> > +#define REG_TDP_LIMIT3 0xe8
> > +
> > +static ssize_t show_power(struct device *dev,
> > + struct device_attribute *attr, char *buf)
> > +{
> > + u32 val, btdp, tdpl, tdp2w, arange;
> > + s32 acap;
> > + u64 ctdp;
> > + struct pci_dev *f4 = to_pci_dev(dev);
> > + struct pci_dev *f5;
> > +
> > + pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
> > + btdp = (val >> 16) & 0xffff;
> > +
> > + f5 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5));
> > + if (!f5) {
> > + dev_err(dev, "no function 5 available on this slot\n");
> > + return 0;
> > + }
>
> Unless this is a temporary error condition, you should check if this function
> exists in the probe function, and not create a sysfs entry if it doesn't.
>
> If it is a temporary error which needs a runtime check, you should return
> a negative value to indicate that there was an error.
Probably I'll follow Clemens' advice to ignore the theoretic case that
the device doesn't exist. (AFAIK such a scenario is only imaginable
under virtualization if only some PCI devices are passed into a
guest.)
> > + pci_read_config_dword(f5, REG_TDP_RUNNING_AVERAGE, &val);
> > + acap = (val >> 4) & 0x3fffff;
> > + acap = sign_extend32(acap, 22);
> > + arange = val & 0xf;
> > +
> > + pci_read_config_dword(f5, REG_TDP_LIMIT3, &val);
> > + pci_dev_put(f5);
> > +
> > + tdpl = (val >> 16) & 0x1fff;
> > + tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
> > + ctdp = tdpl - (s32)(acap >> (arange + 1)) + btdp;
> > + ctdp *= tdp2w;
> > +
> > + /*
> > + * Convert to microWatt
> > + *
> > + * power is in Watt provided as fixed point integer with
> > + * scaling factor 1/(2^16). For conversion we use
> > + * (10^6)/(2^16) = 15625/(2^10)
> > + */
> > + ctdp = (ctdp * 15625) >> 10;
> > + return sprintf(buf, "%d\n", (u32) ctdp);
>
> %u
Yep.
> > +}
> > +static DEVICE_ATTR(power1_input, S_IRUGO, show_power, NULL);
> > +
> > +static ssize_t show_power_cap(struct device *dev,
> > + struct device_attribute *attr, char *buf)
> > +{
> > + u32 val, tdp2w;
> > + u64 ptdp;
> > + struct pci_dev *f4 = to_pci_dev(dev);
> > + struct pci_dev *f5;
> > +
> > + pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
> > + ptdp = val & 0xffff;
> > +
> > + f5 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5));
> > + if (!f5) {
> > + dev_err(dev, "no function 5 available on this slot\n");
> > + return 0;
> > + }
> > +
> Same as above - should be checked in probe function.
>
> If it is a temporary error which needs a runtime check, you should return
> a negative value to indicate that there was an error.
>
> > + pci_read_config_dword(f5, REG_TDP_LIMIT3, &val);
> > + pci_dev_put(f5);
> > +
> > + tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
> > + ptdp *= tdp2w;
> > +
> > + /* result not allowed to be >= 256W */
> > + WARN_ON(ptdp>>16 >= 256);
> > +
> Does this really ask for such drastic measures, or would dev_warn() be sufficient ?
dev_warn() should be sufficient.
> > + /* convert to microWatt */
> > + ptdp = (ptdp * 15625) >> 10;
> > + return sprintf(buf, "%d\n", (u32) ptdp);
>
> %u
> > +}
> > +static DEVICE_ATTR(power1_cap, S_IRUGO, show_power_cap, NULL);
> > +
> > +static ssize_t show_name(struct device *dev,
> > + struct device_attribute *attr, char *buf)
> > +{
> > + return sprintf(buf, "f15h_power\n");
> > +}
> > +static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
> > +
> > +static int __devinit f15h_power_is_internal_node0(struct pci_dev *f4)
> > +{
> > + u32 val;
> > + struct pci_dev *f3;
> > +
> > + f3 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3));
> > + if (!f3) {
> > + dev_err(&f4->dev, "no function 3 available on this slot\n");
> > + return 0;
>
> It is a common practice to return a negative value on errors. Why not here ?
> Also, is this really an error which asks for an error message, or just a CPU
> or system which does not support the attribute ? In the latter case, you should
> not display an error message.
This paranoid check will be removed.
> > + }
> > +
> > + pci_read_config_dword(f3, REG_NORTHBRIDGE_CAP, &val);
> > + pci_dev_put(f3);
> > +
> > + if ((val & BIT(29)) && ((val >> 30) & 3))
> > + return 0;
> > +
> > + return 1;
> > +}
> > +
> > +static int __devinit f15h_power_probe(struct pci_dev *pdev,
> > + const struct pci_device_id *id)
> > +{
> > + struct device *hwmon_dev;
> > + int err = -ENODEV;
> > +
> > + if (!f15h_power_is_internal_node0(pdev))
> > + goto exit;
> > +
> > + err = device_create_file(&pdev->dev, &dev_attr_power1_input);
> > + if (err)
> > + goto exit;
> > + err = device_create_file(&pdev->dev, &dev_attr_power1_cap);
> > + if (err)
> > + goto exit_remove;
> > +
> > + err = device_create_file(&pdev->dev, &dev_attr_name);
> > + if (err)
> > + goto exit_remove;
> > +
> > + hwmon_dev = hwmon_device_register(&pdev->dev);
> > + if (IS_ERR(hwmon_dev)) {
> > + err = PTR_ERR(hwmon_dev);
> > + goto exit_remove;
> > + }
> > + dev_set_drvdata(&pdev->dev, hwmon_dev);
> > +
> > + return 0;
> > +
> > +exit_remove:
> > + device_remove_file(&pdev->dev, &dev_attr_name);
> > + device_remove_file(&pdev->dev, &dev_attr_power1_input);
> > + device_remove_file(&pdev->dev, &dev_attr_power1_cap);
> > +exit:
> > + return err;
> > +}
> > +
> > +static void __devexit f15h_power_remove(struct pci_dev *pdev)
> > +{
> > + hwmon_device_unregister(dev_get_drvdata(&pdev->dev));
> > + device_remove_file(&pdev->dev, &dev_attr_name);
> > + device_remove_file(&pdev->dev, &dev_attr_power1_input);
> > + device_remove_file(&pdev->dev, &dev_attr_power1_cap);
> > + dev_set_drvdata(&pdev->dev, NULL);
> > +}
> > +
> > +static const struct pci_device_id f15h_power_id_table[] = {
>
> checkpatch says:
>
> WARNING: Use DEFINE_PCI_DEVICE_TABLE for struct pci_device_id
> #326: FILE: drivers/hwmon/f15h_power.c:192:
>
> Please fix.
Ok.
> > + { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_LINK) },
> > + {}
> > +};
> > +MODULE_DEVICE_TABLE(pci, f15h_power_id_table);
> > +
> > +static struct pci_driver f15h_power_driver = {
> > + .name = "f15h_power",
> > + .id_table = f15h_power_id_table,
> > + .probe = f15h_power_probe,
> > + .remove = __devexit_p(f15h_power_remove),
> > +};
> > +
> > +static int __init f15h_power_init(void)
> > +{
> > + return pci_register_driver(&f15h_power_driver);
> > +}
> > +
> > +static void __exit f15h_power_exit(void)
> > +{
> > + pci_unregister_driver(&f15h_power_driver);
> > +}
> > +
> > +module_init(f15h_power_init)
> > +module_exit(f15h_power_exit)
Thanks for the review.
Andreas
On Wed, Apr 06, 2011 at 09:14:07AM +0200, Clemens Ladisch wrote:
> Guenter Roeck wrote:
> > On Tue, Apr 05, 2011 at 10:45:36AM -0400, Andreas Herrmann wrote:
> > > +static int __devinit f15h_power_is_internal_node0(struct pci_dev *f4)
> > > +{
> > > + u32 val;
> > > + struct pci_dev *f3;
> > > +
> > > + f3 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3));
> > > + if (!f3) {
> > > + dev_err(&f4->dev, "no function 3 available on this slot\n");
> > > + return 0;
> >
> > It is a common practice to return a negative value on errors. Why not here ?
>
> Apparently, this function returns a boolean value. Using "bool"/"true"/
> "false" would have made this more obvious.
Agreed, using bool as return type would have been the better choice.
I'll adapt it.
> > Also, is this really an error which asks for an error message, or just a CPU
> > or system which does not support the attribute ?
>
> AFAICT all F15h CPUs are _known_ to have all these PCI functions; the
> error should never occur in practice.
Except under virtualization. But common practice is to not pass any
such devices into a guest. So either all CPU northbridge functions are
there or no function at all. I'll adapt the code to your suggestion.
> What I do in the k10temp driver in this situation is to trust the CPU to
> be there, omitting the pci_get_slot/pci_dev_put and just replacing this:
>
> > > + pci_read_config_dword(f3, REG_NORTHBRIDGE_CAP, &val);
> > > + pci_dev_put(f3);
>
> with the equivalent of:
>
> pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3),
> REG_NORTHBRIDGE_CAP, &val);
>
>
> BTW: The family 15h CPUs have the same temperature sensor registers
> (D18F3xA4 and D18F3x64) as the earlier families, haven't they?
Yes and I think Andre has just sent a patch to add support for 15h to
k10temp.
Thanks,
Andreas
From: Andreas Herrmann <[email protected]>
This CPU family provides NB register values to gather following
TDP information
* ProcessorPwrWatts: Specifies in Watts the maximum amount of power
the processor can support.
* CurrPwrWatts: Specifies in Watts the current amount of power being
consumed by the processor.
This driver provides
* power1_max (ProcessorPwrWatts)
* power1_input (CurrPwrWatts)
Changes from v2::
- fix format strings
- removed paranoid checks for existense of functions 3 and 5
- changed return type of function f15h_power_is_internal_node0
- provide power1_max instead of power1_cap
- use dev_warn instead of WARN_ON
- rebased against 2.6.39-rc2
- added Documentation/hwmon/f15h_power
Signed-off-by: Andreas Herrmann <[email protected]>
---
Documentation/hwmon/f15h_power | 37 ++++++++
drivers/hwmon/Kconfig | 10 ++
drivers/hwmon/Makefile | 1 +
drivers/hwmon/f15h_power.c | 197 ++++++++++++++++++++++++++++++++++++++++
drivers/hwmon/k10temp.c | 2 +-
5 files changed, 246 insertions(+), 1 deletions(-)
create mode 100644 Documentation/hwmon/f15h_power
create mode 100644 drivers/hwmon/f15h_power.c
diff --git a/Documentation/hwmon/f15h_power b/Documentation/hwmon/f15h_power
new file mode 100644
index 0000000..5e990d3
--- /dev/null
+++ b/Documentation/hwmon/f15h_power
@@ -0,0 +1,37 @@
+Kernel driver f15h_power
+========================
+
+Supported chips:
+* AMD Family 15h Processors
+
+ Prefix: 'f15h_power'
+ Addresses scanned: PCI space
+ Datasheets:
+ BIOS and Kernel Developer's Guide (BKDG) For AMD Family 15h Processors
+ (not yet published)
+
+Author: Andreas Herrmann <[email protected]>
+
+Description
+-----------
+
+This driver permits reading of registers providing power information
+of AMD Family 15h processors.
+
+For AMD Family 15h processors the following power values can be
+calculated using different processor northbridge function registers
+
+* BasePwrWatts: Specifies in watts the maximum amount of power
+ consumed by the processor for NB and logic external to the core.
+* ProcessorPwrWatts: Specifies in watts the maximum amount of power
+ the processor can support.
+* CurrPwrWatts: Specifies in watts the current amount of power being
+ consumed by the processor.
+
+This driver provides ProcessorPwrWatts and CurrPwrWatts:
+* power1_max (ProcessorPwrWatts)
+* power1_input (CurrPwrWatts)
+
+On multi-node processors the calculated value is for the entire
+package and not for a single node. Thus the driver creates sysfs
+attributes only for internal node0 of a multi-node processor.
diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
index 060ef63..36ae085 100644
--- a/drivers/hwmon/Kconfig
+++ b/drivers/hwmon/Kconfig
@@ -249,6 +249,16 @@ config SENSORS_K10TEMP
This driver can also be built as a module. If so, the module
will be called k10temp.
+config SENSORS_F15H_POWER
+ tristate "AMD Family 15h processor power"
+ depends on X86 && PCI
+ help
+ If you say yes here you get support for processor power
+ information of your AMD family 15h CPU.
+
+ This driver can also be built as a module. If so, the module
+ will be called f15h_power.
+
config SENSORS_ASB100
tristate "Asus ASB100 Bach"
depends on X86 && I2C && EXPERIMENTAL
diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
index 967d0ea..a8d03d5 100644
--- a/drivers/hwmon/Makefile
+++ b/drivers/hwmon/Makefile
@@ -63,6 +63,7 @@ obj-$(CONFIG_SENSORS_JC42) += jc42.o
obj-$(CONFIG_SENSORS_JZ4740) += jz4740-hwmon.o
obj-$(CONFIG_SENSORS_K8TEMP) += k8temp.o
obj-$(CONFIG_SENSORS_K10TEMP) += k10temp.o
+obj-$(CONFIG_SENSORS_F15H_POWER) += f15h_power.o
obj-$(CONFIG_SENSORS_LINEAGE) += lineage-pem.o
obj-$(CONFIG_SENSORS_LM63) += lm63.o
obj-$(CONFIG_SENSORS_LM70) += lm70.o
diff --git a/drivers/hwmon/f15h_power.c b/drivers/hwmon/f15h_power.c
new file mode 100644
index 0000000..648c21d
--- /dev/null
+++ b/drivers/hwmon/f15h_power.c
@@ -0,0 +1,197 @@
+/*
+ * f15h_power.c - AMD Family 15h processor power monitoring
+ *
+ * Copyright (c) 2011 Advanced Micro Devices, Inc.
+ * Author: Andreas Herrmann <[email protected]>
+ *
+ *
+ * This driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this driver; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/err.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/bitops.h>
+#include <asm/processor.h>
+
+MODULE_DESCRIPTION("AMD Family 15h CPU processor power monitor");
+MODULE_AUTHOR("Andreas Herrmann <[email protected]>");
+MODULE_LICENSE("GPL");
+
+/* D18F3 */
+#define REG_NORTHBRIDGE_CAP 0xe8
+
+/* D18F4 */
+#define REG_PROCESSOR_TDP 0x1b8
+
+/* D18F5 */
+#define REG_TDP_RUNNING_AVERAGE 0xe0
+#define REG_TDP_LIMIT3 0xe8
+
+static ssize_t show_power(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ u32 val, btdp, tdpl, tdp2w, arange;
+ s32 acap;
+ u64 ctdp;
+ struct pci_dev *f4 = to_pci_dev(dev);
+
+ pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
+ btdp = (val >> 16) & 0xffff;
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
+ REG_TDP_RUNNING_AVERAGE, &val);
+ acap = (val >> 4) & 0x3fffff;
+ acap = sign_extend32(acap, 22);
+ arange = val & 0xf;
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
+ REG_TDP_LIMIT3, &val);
+
+ tdpl = (val >> 16) & 0x1fff;
+ tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
+ ctdp = tdpl - (s32)(acap >> (arange + 1)) + btdp;
+ ctdp *= tdp2w;
+
+ /*
+ * Convert to microWatt
+ *
+ * power is in Watt provided as fixed point integer with
+ * scaling factor 1/(2^16). For conversion we use
+ * (10^6)/(2^16) = 15625/(2^10)
+ */
+ ctdp = (ctdp * 15625) >> 10;
+ return sprintf(buf, "%u\n", (u32) ctdp);
+}
+static DEVICE_ATTR(power1_input, S_IRUGO, show_power, NULL);
+
+static ssize_t show_power_max(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ u32 val, tdp2w;
+ u64 ptdp;
+ struct pci_dev *f4 = to_pci_dev(dev);
+
+ pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
+ ptdp = val & 0xffff;
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
+ REG_TDP_LIMIT3, &val);
+
+ tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
+ ptdp *= tdp2w;
+
+ /* result not allowed to be >= 256W */
+ if ((ptdp>>16) >= 256)
+ dev_warn(dev, "Bogus value for ProcessorPwrWatts (ptdp>=%u)\n",
+ (u32) (ptdp >> 16));
+
+ /* convert to microWatt */
+ ptdp = (ptdp * 15625) >> 10;
+ return sprintf(buf, "%u\n", (u32) ptdp);
+}
+static DEVICE_ATTR(power1_max, S_IRUGO, show_power_max, NULL);
+
+static ssize_t show_name(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ return sprintf(buf, "f15h_power\n");
+}
+static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
+
+static bool __devinit f15h_power_is_internal_node0(struct pci_dev *f4)
+{
+ u32 val;
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3),
+ REG_NORTHBRIDGE_CAP, &val);
+ if ((val & BIT(29)) && ((val >> 30) & 3))
+ return false;
+
+ return true;
+}
+
+static int __devinit f15h_power_probe(struct pci_dev *pdev,
+ const struct pci_device_id *id)
+{
+ struct device *hwmon_dev;
+ int err = -ENODEV;
+
+ if (!f15h_power_is_internal_node0(pdev))
+ goto exit;
+
+ err = device_create_file(&pdev->dev, &dev_attr_power1_input);
+ if (err)
+ goto exit;
+ err = device_create_file(&pdev->dev, &dev_attr_power1_max);
+ if (err)
+ goto exit_remove;
+
+ err = device_create_file(&pdev->dev, &dev_attr_name);
+ if (err)
+ goto exit_remove;
+
+ hwmon_dev = hwmon_device_register(&pdev->dev);
+ if (IS_ERR(hwmon_dev)) {
+ err = PTR_ERR(hwmon_dev);
+ goto exit_remove;
+ }
+ dev_set_drvdata(&pdev->dev, hwmon_dev);
+
+ return 0;
+
+exit_remove:
+ device_remove_file(&pdev->dev, &dev_attr_name);
+ device_remove_file(&pdev->dev, &dev_attr_power1_input);
+ device_remove_file(&pdev->dev, &dev_attr_power1_max);
+exit:
+ return err;
+}
+
+static void __devexit f15h_power_remove(struct pci_dev *pdev)
+{
+ hwmon_device_unregister(dev_get_drvdata(&pdev->dev));
+ device_remove_file(&pdev->dev, &dev_attr_name);
+ device_remove_file(&pdev->dev, &dev_attr_power1_input);
+ device_remove_file(&pdev->dev, &dev_attr_power1_max);
+ dev_set_drvdata(&pdev->dev, NULL);
+}
+
+static DEFINE_PCI_DEVICE_TABLE(f15h_power_id_table) = {
+ { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_F4) },
+ {}
+};
+MODULE_DEVICE_TABLE(pci, f15h_power_id_table);
+
+static struct pci_driver f15h_power_driver = {
+ .name = "f15h_power",
+ .id_table = f15h_power_id_table,
+ .probe = f15h_power_probe,
+ .remove = __devexit_p(f15h_power_remove),
+};
+
+static int __init f15h_power_init(void)
+{
+ return pci_register_driver(&f15h_power_driver);
+}
+
+static void __exit f15h_power_exit(void)
+{
+ pci_unregister_driver(&f15h_power_driver);
+}
+
+module_init(f15h_power_init)
+module_exit(f15h_power_exit)
diff --git a/drivers/hwmon/k10temp.c b/drivers/hwmon/k10temp.c
index 82bf65a..2cbc1ab 100644
--- a/drivers/hwmon/k10temp.c
+++ b/drivers/hwmon/k10temp.c
@@ -205,7 +205,7 @@ static void __devexit k10temp_remove(struct pci_dev *pdev)
dev_set_drvdata(&pdev->dev, NULL);
}
-static const struct pci_device_id k10temp_id_table[] = {
+static DEFINE_PCI_DEVICE_TABLE(k10temp_id_table) = {
{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_10H_NB_MISC) },
{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_11H_NB_MISC) },
{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_CNB17H_F3) },
--
1.7.4.1
On Tue, 5 Apr 2011 16:45:36 +0200, Andreas Herrmann wrote:
> From: Andreas Herrmann <[email protected]>
>
> This CPU family provides NB register values to gather following
> TDP information
>
> * ProcessorPwrWatts: Specifies in Watts the maximum amount of power
> the processor can support.
> * CurrPwrWatts: Specifies in Watts the current amount of power being
> consumed by the processor.
>
> This driver provides
>
> * power1_cap (ProcessorPwrWatts)
> * power1_input (CurrPwrWatts)
Just a few random comments as I don't have the time for a complete
review:
>
> Changes from v1::
> - removed deprecated code line,
> - fixed comment,
> - report power only once per socket (Power information is provided
> for the entire package. On multi-node processors it should only be
> reported on internal node 0.)
>
> Signed-off-by: Andreas Herrmann <[email protected]>
> ---
> drivers/hwmon/Kconfig | 10 ++
> drivers/hwmon/Makefile | 1 +
> drivers/hwmon/f15h_power.c | 216 ++++++++++++++++++++++++++++++++++++++++++++
I would prefer fam15h to f15h. The latter is a little too cryptic for
me.
> 3 files changed, 227 insertions(+), 0 deletions(-)
> create mode 100644 drivers/hwmon/f15h_power.c
>
> diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
> index 8e84b31..0268623 100644
> --- a/drivers/hwmon/Kconfig
> +++ b/drivers/hwmon/Kconfig
> @@ -249,6 +249,16 @@ config SENSORS_K10TEMP
> This driver can also be built as a module. If so, the module
> will be called k10temp.
>
> +config SENSORS_F15H_POWER
> + tristate "AMD Family 15h processor power"
> + depends on X86 && PCI
> + help
> + If you say yes here you get support for processor power
> + information of your AMD family 15h CPU.
> +
> + This driver can also be built as a module. If so, the module
> + will be called f15hpower.
> +
> config SENSORS_ASB100
> tristate "Asus ASB100 Bach"
> depends on X86 && I2C && EXPERIMENTAL
> diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
> index dde02d9..e554b88 100644
> --- a/drivers/hwmon/Makefile
> +++ b/drivers/hwmon/Makefile
> @@ -62,6 +62,7 @@ obj-$(CONFIG_SENSORS_JC42) += jc42.o
> obj-$(CONFIG_SENSORS_JZ4740) += jz4740-hwmon.o
> obj-$(CONFIG_SENSORS_K8TEMP) += k8temp.o
> obj-$(CONFIG_SENSORS_K10TEMP) += k10temp.o
> +obj-$(CONFIG_SENSORS_F15H_POWER) += f15h_power.o
Alphabetic order please.
> obj-$(CONFIG_SENSORS_LIS3LV02D) += lis3lv02d.o hp_accel.o
> obj-$(CONFIG_SENSORS_LIS3_SPI) += lis3lv02d.o lis3lv02d_spi.o
> obj-$(CONFIG_SENSORS_LIS3_I2C) += lis3lv02d.o lis3lv02d_i2c.o
> diff --git a/drivers/hwmon/f15h_power.c b/drivers/hwmon/f15h_power.c
> new file mode 100644
> index 0000000..40e3cbd
> --- /dev/null
> +++ b/drivers/hwmon/f15h_power.c
> @@ -0,0 +1,216 @@
> +/*
> + * f15h_power.c - AMD Family 15h processor power monitoring
> + *
> + * Copyright (c) 2011 Advanced Micro Devices, Inc.
> + * Author: Andreas Herrmann <[email protected]>
> + *
> + *
> + * This driver is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This driver is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
> + * See the GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this driver; if not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include <linux/err.h>
> +#include <linux/hwmon.h>
> +#include <linux/hwmon-sysfs.h>
> +#include <linux/init.h>
> +#include <linux/module.h>
> +#include <linux/pci.h>
> +#include <linux/bitops.h>
> +#include <asm/processor.h>
> +
> +MODULE_DESCRIPTION("AMD Family 15h CPU processor power monitor");
> +MODULE_AUTHOR("Andreas Herrmann <[email protected]>");
> +MODULE_LICENSE("GPL");
> +
> +/* D18F3 */
> +#define REG_NORTHBRIDGE_CAP 0xe8
> +
> +/* D18F4 */
> +#define REG_PROCESSOR_TDP 0x1b8
> +
> +/* D18F5 */
> +#define REG_TDP_RUNNING_AVERAGE 0xe0
> +#define REG_TDP_LIMIT3 0xe8
> +
> +static ssize_t show_power(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + u32 val, btdp, tdpl, tdp2w, arange;
> + s32 acap;
> + u64 ctdp;
These variable names aren't easy to understand.
> + struct pci_dev *f4 = to_pci_dev(dev);
> + struct pci_dev *f5;
> +
> + pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
> + btdp = (val >> 16) & 0xffff;
Useless masking.
> +
> + f5 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5));
> + if (!f5) {
> + dev_err(dev, "no function 5 available on this slot\n");
> + return 0;
> + }
> +
> + pci_read_config_dword(f5, REG_TDP_RUNNING_AVERAGE, &val);
Does this value change over time?
> + acap = (val >> 4) & 0x3fffff;
> + acap = sign_extend32(acap, 22);
> + arange = val & 0xf;
> +
> + pci_read_config_dword(f5, REG_TDP_LIMIT3, &val);
Same question here. If these values don't change over time, they should
be stored in a per-device structure to avoid repeated PCI configuration
space access.
> + pci_dev_put(f5);
> +
> + tdpl = (val >> 16) & 0x1fff;
> + tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
> + ctdp = tdpl - (s32)(acap >> (arange + 1)) + btdp;
> + ctdp *= tdp2w;
> +
> + /*
> + * Convert to microWatt
> + *
> + * power is in Watt provided as fixed point integer with
> + * scaling factor 1/(2^16). For conversion we use
> + * (10^6)/(2^16) = 15625/(2^10)
> + */
> + ctdp = (ctdp * 15625) >> 10;
> + return sprintf(buf, "%d\n", (u32) ctdp);
%d and u32 aren't compatible. %lu and long unsigned int would be OK.
> +}
> +static DEVICE_ATTR(power1_input, S_IRUGO, show_power, NULL);
> +
> +static ssize_t show_power_cap(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + u32 val, tdp2w;
> + u64 ptdp;
> + struct pci_dev *f4 = to_pci_dev(dev);
> + struct pci_dev *f5;
> +
> + pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
> + ptdp = val & 0xffff;
> +
> + f5 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5));
> + if (!f5) {
> + dev_err(dev, "no function 5 available on this slot\n");
> + return 0;
> + }
> +
> + pci_read_config_dword(f5, REG_TDP_LIMIT3, &val);
Same question as above, can this value change over time?
> + pci_dev_put(f5);
> +
> + tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
> + ptdp *= tdp2w;
> +
> + /* result not allowed to be >= 256W */
> + WARN_ON(ptdp>>16 >= 256);
> +
> + /* convert to microWatt */
> + ptdp = (ptdp * 15625) >> 10;
> + return sprintf(buf, "%d\n", (u32) ptdp);
%d and u32 aren't compatible. %lu and long unsigned int would be OK.
> +}
> +static DEVICE_ATTR(power1_cap, S_IRUGO, show_power_cap, NULL);
> +
> +static ssize_t show_name(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + return sprintf(buf, "f15h_power\n");
> +}
> +static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
> +
> +static int __devinit f15h_power_is_internal_node0(struct pci_dev *f4)
> +{
> + u32 val;
> + struct pci_dev *f3;
> +
> + f3 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3));
> + if (!f3) {
> + dev_err(&f4->dev, "no function 3 available on this slot\n");
Please use %d and pass the number as a parameter, and do the same for
the 2 similar messages above. This will let the compiler reuse the
string.
> + return 0;
> + }
> +
> + pci_read_config_dword(f3, REG_NORTHBRIDGE_CAP, &val);
> + pci_dev_put(f3);
> +
> + if ((val & BIT(29)) && ((val >> 30) & 3))
> + return 0;
> +
> + return 1;
> +}
> +
> +static int __devinit f15h_power_probe(struct pci_dev *pdev,
> + const struct pci_device_id *id)
> +{
> + struct device *hwmon_dev;
> + int err = -ENODEV;
You should move this initialization to the only case which needs it.
Otherwise you're slowing down the success path with no good reason.
> +
> + if (!f15h_power_is_internal_node0(pdev))
> + goto exit;
> +
> + err = device_create_file(&pdev->dev, &dev_attr_power1_input);
> + if (err)
> + goto exit;
> + err = device_create_file(&pdev->dev, &dev_attr_power1_cap);
> + if (err)
> + goto exit_remove;
> +
> + err = device_create_file(&pdev->dev, &dev_attr_name);
> + if (err)
> + goto exit_remove;
Please consider defining and using an attribute group, it makes the
code more simple.
> +
> + hwmon_dev = hwmon_device_register(&pdev->dev);
> + if (IS_ERR(hwmon_dev)) {
> + err = PTR_ERR(hwmon_dev);
> + goto exit_remove;
> + }
> + dev_set_drvdata(&pdev->dev, hwmon_dev);
pci_set_drvdata()
> +
> + return 0;
> +
> +exit_remove:
> + device_remove_file(&pdev->dev, &dev_attr_name);
> + device_remove_file(&pdev->dev, &dev_attr_power1_input);
> + device_remove_file(&pdev->dev, &dev_attr_power1_cap);
> +exit:
> + return err;
> +}
> +
> +static void __devexit f15h_power_remove(struct pci_dev *pdev)
> +{
> + hwmon_device_unregister(dev_get_drvdata(&pdev->dev));
> + device_remove_file(&pdev->dev, &dev_attr_name);
> + device_remove_file(&pdev->dev, &dev_attr_power1_input);
> + device_remove_file(&pdev->dev, &dev_attr_power1_cap);
> + dev_set_drvdata(&pdev->dev, NULL);
pci_set_drvdata()
> +}
> +
> +static const struct pci_device_id f15h_power_id_table[] = {
> + { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_LINK) },
Not defined anywhere, causing the driver to fail building:
drivers/hwmon/f15h_power.c:193:1: error: 'PCI_DEVICE_ID_AMD_15H_NB_LINK' undeclared here (not in a function)
make[2]: *** [drivers/hwmon/f15h_power.o] Error 1
make[1]: *** [drivers/hwmon] Error 2
make[1]: *** Waiting for unfinished jobs....
make: *** [drivers] Error 2
> + {}
> +};
> +MODULE_DEVICE_TABLE(pci, f15h_power_id_table);
> +
> +static struct pci_driver f15h_power_driver = {
> + .name = "f15h_power",
> + .id_table = f15h_power_id_table,
> + .probe = f15h_power_probe,
> + .remove = __devexit_p(f15h_power_remove),
> +};
> +
> +static int __init f15h_power_init(void)
> +{
> + return pci_register_driver(&f15h_power_driver);
> +}
> +
> +static void __exit f15h_power_exit(void)
> +{
> + pci_unregister_driver(&f15h_power_driver);
> +}
> +
> +module_init(f15h_power_init)
> +module_exit(f15h_power_exit)
--
Jean Delvare
On Wed, Apr 06, 2011 at 04:14:01PM +0200, Jean Delvare wrote:
> On Tue, 5 Apr 2011 16:45:36 +0200, Andreas Herrmann wrote:
> > From: Andreas Herrmann <[email protected]>
> >
> > This CPU family provides NB register values to gather following
> > TDP information
> >
> > * ProcessorPwrWatts: Specifies in Watts the maximum amount of power
> > the processor can support.
> > * CurrPwrWatts: Specifies in Watts the current amount of power being
> > consumed by the processor.
> >
> > This driver provides
> >
> > * power1_cap (ProcessorPwrWatts)
> > * power1_input (CurrPwrWatts)
>
> Just a few random comments as I don't have the time for a complete
> review:
>
> >
> > Changes from v1::
> > - removed deprecated code line,
> > - fixed comment,
> > - report power only once per socket (Power information is provided
> > for the entire package. On multi-node processors it should only be
> > reported on internal node 0.)
> >
> > Signed-off-by: Andreas Herrmann <[email protected]>
> > ---
> > drivers/hwmon/Kconfig | 10 ++
> > drivers/hwmon/Makefile | 1 +
> > drivers/hwmon/f15h_power.c | 216 ++++++++++++++++++++++++++++++++++++++++++++
>
> I would prefer fam15h to f15h. The latter is a little too cryptic for
> me.
Hmm, (unfortunately;-) seconded, e.g.
linux-2.6 $ grep -rs fam10h arch/x86 | wc
71 450 7930
linux-2.6 $ grep -rs f10h arch/x86 | wc
0 0 0
Will change it.
[snip]
> > --- a/drivers/hwmon/Makefile
> > +++ b/drivers/hwmon/Makefile
> > @@ -62,6 +62,7 @@ obj-$(CONFIG_SENSORS_JC42) += jc42.o
> > obj-$(CONFIG_SENSORS_JZ4740) += jz4740-hwmon.o
> > obj-$(CONFIG_SENSORS_K8TEMP) += k8temp.o
> > obj-$(CONFIG_SENSORS_K10TEMP) += k10temp.o
> > +obj-$(CONFIG_SENSORS_F15H_POWER) += f15h_power.o
>
> Alphabetic order please.
ok.
[snip]
> > +static ssize_t show_power(struct device *dev,
> > + struct device_attribute *attr, char *buf)
> > +{
> > + u32 val, btdp, tdpl, tdp2w, arange;
> > + s32 acap;
> > + u64 ctdp;
>
> These variable names aren't easy to understand.
Just random names which eventually map to the spec:
btdp - base_tdp
tdpl - tdp_limit
tdp2w - tdp_to_watt
acap - average_accumulator_capture (or even worse how about "processor_tdp_running_average_accumulator":(
arange - average_range
I don't think that changing the names make it much easier to
reconstruct the calculation but if you insist in changing it I'll
adapt it.
> > + struct pci_dev *f4 = to_pci_dev(dev);
> > + struct pci_dev *f5;
> > +
> > + pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
> > + btdp = (val >> 16) & 0xffff;
>
> Useless masking.
To be removed.
> > +
> > + f5 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5));
> > + if (!f5) {
> > + dev_err(dev, "no function 5 available on this slot\n");
> > + return 0;
> > + }
> > +
> > + pci_read_config_dword(f5, REG_TDP_RUNNING_AVERAGE, &val);
>
> Does this value change over time?
Yes.
> > + acap = (val >> 4) & 0x3fffff;
> > + acap = sign_extend32(acap, 22);
> > + arange = val & 0xf;
> > +
> > + pci_read_config_dword(f5, REG_TDP_LIMIT3, &val);
>
> Same question here. If these values don't change over time, they should
> be stored in a per-device structure to avoid repeated PCI configuration
> space access.
The parts I use from this register shouldn't change => code to be modified.
> > + pci_dev_put(f5);
> > +
> > + tdpl = (val >> 16) & 0x1fff;
> > + tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
> > + ctdp = tdpl - (s32)(acap >> (arange + 1)) + btdp;
> > + ctdp *= tdp2w;
> > +
> > + /*
> > + * Convert to microWatt
> > + *
> > + * power is in Watt provided as fixed point integer with
> > + * scaling factor 1/(2^16). For conversion we use
> > + * (10^6)/(2^16) = 15625/(2^10)
> > + */
> > + ctdp = (ctdp * 15625) >> 10;
> > + return sprintf(buf, "%d\n", (u32) ctdp);
>
> %d and u32 aren't compatible. %lu and long unsigned int would be OK.
already fixed in v3
> > +}
> > +static DEVICE_ATTR(power1_input, S_IRUGO, show_power, NULL);
> > +
> > +static ssize_t show_power_cap(struct device *dev,
> > + struct device_attribute *attr, char *buf)
> > +{
> > + u32 val, tdp2w;
> > + u64 ptdp;
> > + struct pci_dev *f4 = to_pci_dev(dev);
> > + struct pci_dev *f5;
> > +
> > + pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
> > + ptdp = val & 0xffff;
> > +
> > + f5 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5));
> > + if (!f5) {
> > + dev_err(dev, "no function 5 available on this slot\n");
> > + return 0;
> > + }
> > +
> > + pci_read_config_dword(f5, REG_TDP_LIMIT3, &val);
>
> Same question as above, can this value change over time?
Needs to be modified.
> > + pci_dev_put(f5);
> > +
> > + tdp2w = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
> > + ptdp *= tdp2w;
> > +
> > + /* result not allowed to be >= 256W */
> > + WARN_ON(ptdp>>16 >= 256);
> > +
> > + /* convert to microWatt */
> > + ptdp = (ptdp * 15625) >> 10;
> > + return sprintf(buf, "%d\n", (u32) ptdp);
>
> %d and u32 aren't compatible. %lu and long unsigned int would be OK.
fixed in v3.
> > +}
> > +static DEVICE_ATTR(power1_cap, S_IRUGO, show_power_cap, NULL);
> > +
> > +static ssize_t show_name(struct device *dev,
> > + struct device_attribute *attr, char *buf)
> > +{
> > + return sprintf(buf, "f15h_power\n");
> > +}
> > +static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
> > +
> > +static int __devinit f15h_power_is_internal_node0(struct pci_dev *f4)
> > +{
> > + u32 val;
> > + struct pci_dev *f3;
> > +
> > + f3 = pci_get_slot(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3));
> > + if (!f3) {
> > + dev_err(&f4->dev, "no function 3 available on this slot\n");
>
> Please use %d and pass the number as a parameter, and do the same for
> the 2 similar messages above. This will let the compiler reuse the
> string.
message removed in v3
> > + return 0;
> > + }
> > +
> > + pci_read_config_dword(f3, REG_NORTHBRIDGE_CAP, &val);
> > + pci_dev_put(f3);
> > +
> > + if ((val & BIT(29)) && ((val >> 30) & 3))
> > + return 0;
> > +
> > + return 1;
> > +}
> > +
> > +static int __devinit f15h_power_probe(struct pci_dev *pdev,
> > + const struct pci_device_id *id)
> > +{
> > + struct device *hwmon_dev;
> > + int err = -ENODEV;
>
> You should move this initialization to the only case which needs it.
> Otherwise you're slowing down the success path with no good reason.
to be done
> > +
> > + if (!f15h_power_is_internal_node0(pdev))
> > + goto exit;
> > +
> > + err = device_create_file(&pdev->dev, &dev_attr_power1_input);
> > + if (err)
> > + goto exit;
> > + err = device_create_file(&pdev->dev, &dev_attr_power1_cap);
> > + if (err)
> > + goto exit_remove;
> > +
> > + err = device_create_file(&pdev->dev, &dev_attr_name);
> > + if (err)
> > + goto exit_remove;
>
> Please consider defining and using an attribute group, it makes the
> code more simple.
Will think about it.
> > +
> > + hwmon_dev = hwmon_device_register(&pdev->dev);
> > + if (IS_ERR(hwmon_dev)) {
> > + err = PTR_ERR(hwmon_dev);
> > + goto exit_remove;
> > + }
> > + dev_set_drvdata(&pdev->dev, hwmon_dev);
>
> pci_set_drvdata()
Are you sure?
No single hwmon driver is using this wrapper around dev_set_drvdata() so far. (Or I just missed it.)
At the moment only dev_set_drvdata() is used by
drivers/hwmon/adcxx.c
drivers/hwmon/lm70.c
drivers/hwmon/ultra45_env.c
drivers/hwmon/ibmpex.c
drivers/hwmon/k10temp.c
drivers/hwmon/ibmaem.c
drivers/hwmon/k8temp.c
> > +
> > + return 0;
> > +
> > +exit_remove:
> > + device_remove_file(&pdev->dev, &dev_attr_name);
> > + device_remove_file(&pdev->dev, &dev_attr_power1_input);
> > + device_remove_file(&pdev->dev, &dev_attr_power1_cap);
> > +exit:
> > + return err;
> > +}
> > +
> > +static void __devexit f15h_power_remove(struct pci_dev *pdev)
> > +{
> > + hwmon_device_unregister(dev_get_drvdata(&pdev->dev));
> > + device_remove_file(&pdev->dev, &dev_attr_name);
> > + device_remove_file(&pdev->dev, &dev_attr_power1_input);
> > + device_remove_file(&pdev->dev, &dev_attr_power1_cap);
> > + dev_set_drvdata(&pdev->dev, NULL);
>
> pci_set_drvdata()
>
> > +}
> > +
> > +static const struct pci_device_id f15h_power_id_table[] = {
> > + { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_LINK) },
>
> Not defined anywhere, causing the driver to fail building:
>
> drivers/hwmon/f15h_power.c:193:1: error: 'PCI_DEVICE_ID_AMD_15H_NB_LINK' undeclared here (not in a function)
> make[2]: *** [drivers/hwmon/f15h_power.o] Error 1
> make[1]: *** [drivers/hwmon] Error 2
> make[1]: *** Waiting for unfinished jobs....
> make: *** [drivers] Error 2
The macro changed with .39-rc2
Already fixed in v3
Thanks for your review,
Please let me know whether changing the variable names and using
pci_set_drvdata() is strictly required.
Andreas
On Wed, 6 Apr 2011 17:19:01 +0200, Andreas Herrmann wrote:
> On Wed, Apr 06, 2011 at 04:14:01PM +0200, Jean Delvare wrote:
> > On Tue, 5 Apr 2011 16:45:36 +0200, Andreas Herrmann wrote:
> > > +static ssize_t show_power(struct device *dev,
> > > + struct device_attribute *attr, char *buf)
> > > +{
> > > + u32 val, btdp, tdpl, tdp2w, arange;
> > > + s32 acap;
> > > + u64 ctdp;
> >
> > These variable names aren't easy to understand.
>
> Just random names which eventually map to the spec:
>
> btdp - base_tdp
> tdpl - tdp_limit
> tdp2w - tdp_to_watt
> acap - average_accumulator_capture (or even worse how about "processor_tdp_running_average_accumulator":(
> arange - average_range
avg_cap and avg_range would do, respectively, for the last two.
> I don't think that changing the names make it much easier to
> reconstruct the calculation but if you insist in changing it I'll
> adapt it.
I do prefer the "extended" names, really. Sure, this doesn't change the
calculations, but it helps the reader understand what's going on. Which
will be useful if one ever has to fix a bug in the code or extend it
for a different CPU family.
But maybe this is just me. Guenter, do you have an opinion?
> > > (...)
> > > + dev_set_drvdata(&pdev->dev, hwmon_dev);
> >
> > pci_set_drvdata()
>
> Are you sure?
Yes, I'm sure. This function exists, so there is no reason not to use
it.
> No single hwmon driver is using this wrapper around dev_set_drvdata() so far. (Or I just missed it.)
$ grep i2c_set_clientdata drivers/hwmon/*.c | wc -l
67
$ grep spi_set_drvdata drivers/hwmon/*.c | wc -l
2
$ grep platform_set_drvdata drivers/hwmon/*.c | wc -l
66
$
> At the moment only dev_set_drvdata() is used by
>
> drivers/hwmon/adcxx.c
> drivers/hwmon/lm70.c
> drivers/hwmon/ultra45_env.c
> drivers/hwmon/ibmpex.c
> drivers/hwmon/k10temp.c
> drivers/hwmon/ibmaem.c
> drivers/hwmon/k8temp.c
Most of these (at least spi and pci drivers) should be fixed. I'll send
a patch later.
--
Jean Delvare
On Wed, Apr 06, 2011 at 11:35:43AM -0400, Jean Delvare wrote:
> On Wed, 6 Apr 2011 17:19:01 +0200, Andreas Herrmann wrote:
> > On Wed, Apr 06, 2011 at 04:14:01PM +0200, Jean Delvare wrote:
> > > On Tue, 5 Apr 2011 16:45:36 +0200, Andreas Herrmann wrote:
> > > > +static ssize_t show_power(struct device *dev,
> > > > + struct device_attribute *attr, char *buf)
> > > > +{
> > > > + u32 val, btdp, tdpl, tdp2w, arange;
> > > > + s32 acap;
> > > > + u64 ctdp;
> > >
> > > These variable names aren't easy to understand.
> >
> > Just random names which eventually map to the spec:
> >
> > btdp - base_tdp
> > tdpl - tdp_limit
> > tdp2w - tdp_to_watt
> > acap - average_accumulator_capture (or even worse how about "processor_tdp_running_average_accumulator":(
> > arange - average_range
>
> avg_cap and avg_range would do, respectively, for the last two.
>
> > I don't think that changing the names make it much easier to
> > reconstruct the calculation but if you insist in changing it I'll
> > adapt it.
>
> I do prefer the "extended" names, really. Sure, this doesn't change the
> calculations, but it helps the reader understand what's going on. Which
> will be useful if one ever has to fix a bug in the code or extend it
> for a different CPU family.
>
> But maybe this is just me. Guenter, do you have an opinion?
>
I agree. base_tdp is definitely much better than btdp. Same for the others.
Thanks,
Guenter
On Wed, Apr 06, 2011 at 09:14:45AM -0700, Guenter Roeck wrote:
> On Wed, Apr 06, 2011 at 11:35:43AM -0400, Jean Delvare wrote:
> > On Wed, 6 Apr 2011 17:19:01 +0200, Andreas Herrmann wrote:
> > > On Wed, Apr 06, 2011 at 04:14:01PM +0200, Jean Delvare wrote:
> > > > On Tue, 5 Apr 2011 16:45:36 +0200, Andreas Herrmann wrote:
> > > > > +static ssize_t show_power(struct device *dev,
> > > > > + struct device_attribute *attr, char *buf)
> > > > > +{
> > > > > + u32 val, btdp, tdpl, tdp2w, arange;
> > > > > + s32 acap;
> > > > > + u64 ctdp;
> > > >
> > > > These variable names aren't easy to understand.
> > >
> > > Just random names which eventually map to the spec:
> > >
> > > btdp - base_tdp
> > > tdpl - tdp_limit
> > > tdp2w - tdp_to_watt
> > > acap - average_accumulator_capture (or even worse how about "processor_tdp_running_average_accumulator":(
> > > arange - average_range
> >
> > avg_cap and avg_range would do, respectively, for the last two.
> >
> > > I don't think that changing the names make it much easier to
> > > reconstruct the calculation but if you insist in changing it I'll
> > > adapt it.
> >
> > I do prefer the "extended" names, really. Sure, this doesn't change the
> > calculations, but it helps the reader understand what's going on. Which
> > will be useful if one ever has to fix a bug in the code or extend it
> > for a different CPU family.
> >
> > But maybe this is just me. Guenter, do you have an opinion?
> >
> I agree. base_tdp is definitely much better than btdp. Same for the others.
>
> Thanks,
> Guenter
Ok.
Will update the patch soon.
Andreas
Hi Andreas,
On Wed, 6 Apr 2011 15:54:24 +0200, Andreas Herrmann wrote:
> --- /dev/null
> +++ b/Documentation/hwmon/f15h_power
> @@ -0,0 +1,37 @@
> +Kernel driver f15h_power
> +========================
> +
> +Supported chips:
> +* AMD Family 15h Processors
> +
> + Prefix: 'f15h_power'
> + Addresses scanned: PCI space
> + Datasheets:
> + BIOS and Kernel Developer's Guide (BKDG) For AMD Family 15h Processors
> + (not yet published)
> +
> +Author: Andreas Herrmann <[email protected]>
BTW, please consider adding an entry in MAINTAINERS.
> +
> +Description
> +-----------
> +
> +This driver permits reading of registers providing power information
> +of AMD Family 15h processors.
> +
> +For AMD Family 15h processors the following power values can be
> +calculated using different processor northbridge function registers
A trailing ":" would be nice.
> +
> +* BasePwrWatts: Specifies in watts the maximum amount of power
> + consumed by the processor for NB and logic external to the core.
> +* ProcessorPwrWatts: Specifies in watts the maximum amount of power
> + the processor can support.
> +* CurrPwrWatts: Specifies in watts the current amount of power being
> + consumed by the processor.
> +
> +This driver provides ProcessorPwrWatts and CurrPwrWatts:
> +* power1_max (ProcessorPwrWatts)
I see you changed this name once already, but... What is expected to
happen if the power consumed exceeds this limit? If you expect the CPU
to get damaged, then power1_crit would be more appropriate.
Another way to decide if _max or _crit is more appropriate is by
answering the question: if a second limit was added in the future,
would it more likely be below or above this one?
> +* power1_input (CurrPwrWatts)
> +
> +On multi-node processors the calculated value is for the entire
> +package and not for a single node. Thus the driver creates sysfs
> +attributes only for internal node0 of a multi-node processor.
> (...)
> + return sprintf(buf, "%u\n", (u32) ptdp);
Maybe I'm just nitpicking, but I still don't think it's correct. %u
wants unsigned int, not u32. It might be the same but that's by luck.
> --- a/drivers/hwmon/k10temp.c
> +++ b/drivers/hwmon/k10temp.c
> @@ -205,7 +205,7 @@ static void __devexit k10temp_remove(struct pci_dev *pdev)
> dev_set_drvdata(&pdev->dev, NULL);
> }
>
> -static const struct pci_device_id k10temp_id_table[] = {
> +static DEFINE_PCI_DEVICE_TABLE(k10temp_id_table) = {
> { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_10H_NB_MISC) },
> { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_11H_NB_MISC) },
> { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_CNB17H_F3) },
Such cleanups should go to separate patches, as they don't have
anything to do with your initial effort. And that way you can fix
k8temp i5k_amb too.
--
Jean Delvare
On Wed, Apr 06, 2011 at 06:50:44PM +0200, Jean Delvare wrote:
> Hi Andreas,
>
> On Wed, 6 Apr 2011 15:54:24 +0200, Andreas Herrmann wrote:
> > --- /dev/null
> > +++ b/Documentation/hwmon/f15h_power
> > @@ -0,0 +1,37 @@
> > +Kernel driver f15h_power
> > +========================
> > +
> > +Supported chips:
> > +* AMD Family 15h Processors
> > +
> > + Prefix: 'f15h_power'
> > + Addresses scanned: PCI space
> > + Datasheets:
> > + BIOS and Kernel Developer's Guide (BKDG) For AMD Family 15h Processors
> > + (not yet published)
> > +
> > +Author: Andreas Herrmann <[email protected]>
>
> BTW, please consider adding an entry in MAINTAINERS.
Yep.
> > +
> > +Description
> > +-----------
> > +
> > +This driver permits reading of registers providing power information
> > +of AMD Family 15h processors.
> > +
> > +For AMD Family 15h processors the following power values can be
> > +calculated using different processor northbridge function registers
>
> A trailing ":" would be nice.
>
> > +
> > +* BasePwrWatts: Specifies in watts the maximum amount of power
> > + consumed by the processor for NB and logic external to the core.
> > +* ProcessorPwrWatts: Specifies in watts the maximum amount of power
> > + the processor can support.
> > +* CurrPwrWatts: Specifies in watts the current amount of power being
> > + consumed by the processor.
> > +
> > +This driver provides ProcessorPwrWatts and CurrPwrWatts:
> > +* power1_max (ProcessorPwrWatts)
>
> I see you changed this name once already, but... What is expected to
> happen if the power consumed exceeds this limit? If you expect the CPU
> to get damaged, then power1_crit would be more appropriate.
>
> Another way to decide if _max or _crit is more appropriate is by
> answering the question: if a second limit was added in the future,
> would it more likely be below or above this one?
It might make sense to use it as _crit and the interim result of
(tdp_limit + base_tdp) * tdp_to_watts
being less than ProcessorPwrWatts as _max.
> > +* power1_input (CurrPwrWatts)
> > +
> > +On multi-node processors the calculated value is for the entire
> > +package and not for a single node. Thus the driver creates sysfs
> > +attributes only for internal node0 of a multi-node processor.
>
> > (...)
> > + return sprintf(buf, "%u\n", (u32) ptdp);
>
> Maybe I'm just nitpicking, but I still don't think it's correct. %u
> wants unsigned int, not u32. It might be the same but that's by luck.
>
> > --- a/drivers/hwmon/k10temp.c
> > +++ b/drivers/hwmon/k10temp.c
> > @@ -205,7 +205,7 @@ static void __devexit k10temp_remove(struct pci_dev *pdev)
> > dev_set_drvdata(&pdev->dev, NULL);
> > }
> >
> > -static const struct pci_device_id k10temp_id_table[] = {
> > +static DEFINE_PCI_DEVICE_TABLE(k10temp_id_table) = {
> > { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_10H_NB_MISC) },
> > { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_11H_NB_MISC) },
> > { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_CNB17H_F3) },
>
> Such cleanups should go to separate patches, as they don't have
> anything to do with your initial effort. And that way you can fix
> k8temp i5k_amb too.
This slipped in by accident. I might send a separate patch to clean this
up in several hwmon drivers.
Andreas
From: Andreas Herrmann <[email protected]>
This CPU family provides NB register values to gather following
TDP information
* ProcessorPwrWatts: Specifies in Watts the maximum amount of power
the processor can support.
* CurrPwrWatts: Specifies in Watts the current amount of power being
consumed by the processor.
This driver provides
* power1_max (ProcessorPwrWatts)
* power1_input (CurrPwrWatts)
Changes from v2:
- fix format strings
- removed paranoid checks for existense of functions 3 and 5
- changed return type of function f15h_power_is_internal_node0
- use power1_max instead of power1_cap
- use dev_warn instead of WARN_ON
- rebased against 2.6.39-rc2
- added Documentation/hwmon/f15h_power
Changes from v3:
- read static power information only once (during driver initialization)
- made use of attribute groups
- renamed f15h_power to fam15h_power
Signed-off-by: Andreas Herrmann <[email protected]>
---
Documentation/hwmon/fam15h_power | 37 ++++++
drivers/hwmon/Kconfig | 10 ++
drivers/hwmon/Makefile | 1 +
drivers/hwmon/fam15h_power.c | 229 ++++++++++++++++++++++++++++++++++++++
4 files changed, 277 insertions(+), 0 deletions(-)
create mode 100644 Documentation/hwmon/fam15h_power
create mode 100644 drivers/hwmon/fam15h_power.c
At this stage I consider the _max attribute as the right one for
reporting ProcessorPwrWatts.
Hopefully I've addressed all your coments.
Please apply.
Thanks,
Andreas
diff --git a/Documentation/hwmon/fam15h_power b/Documentation/hwmon/fam15h_power
new file mode 100644
index 0000000..2f4d291
--- /dev/null
+++ b/Documentation/hwmon/fam15h_power
@@ -0,0 +1,37 @@
+Kernel driver fam15h_power
+==========================
+
+Supported chips:
+* AMD Family 15h Processors
+
+ Prefix: 'fam15h_power'
+ Addresses scanned: PCI space
+ Datasheets:
+ BIOS and Kernel Developer's Guide (BKDG) For AMD Family 15h Processors
+ (not yet published)
+
+Author: Andreas Herrmann <[email protected]>
+
+Description
+-----------
+
+This driver permits reading of registers providing power information
+of AMD Family 15h processors.
+
+For AMD Family 15h processors the following power values can be
+calculated using different processor northbridge function registers:
+
+* BasePwrWatts: Specifies in watts the maximum amount of power
+ consumed by the processor for NB and logic external to the core.
+* ProcessorPwrWatts: Specifies in watts the maximum amount of power
+ the processor can support.
+* CurrPwrWatts: Specifies in watts the current amount of power being
+ consumed by the processor.
+
+This driver provides ProcessorPwrWatts and CurrPwrWatts:
+* power1_max (ProcessorPwrWatts)
+* power1_input (CurrPwrWatts)
+
+On multi-node processors the calculated value is for the entire
+package and not for a single node. Thus the driver creates sysfs
+attributes only for internal node0 of a multi-node processor.
diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
index 060ef63..fb3e334 100644
--- a/drivers/hwmon/Kconfig
+++ b/drivers/hwmon/Kconfig
@@ -249,6 +249,16 @@ config SENSORS_K10TEMP
This driver can also be built as a module. If so, the module
will be called k10temp.
+config SENSORS_FAM15H_POWER
+ tristate "AMD Family 15h processor power"
+ depends on X86 && PCI
+ help
+ If you say yes here you get support for processor power
+ information of your AMD family 15h CPU.
+
+ This driver can also be built as a module. If so, the module
+ will be called fam15h_power.
+
config SENSORS_ASB100
tristate "Asus ASB100 Bach"
depends on X86 && I2C && EXPERIMENTAL
diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
index 967d0ea..236d3f9 100644
--- a/drivers/hwmon/Makefile
+++ b/drivers/hwmon/Makefile
@@ -49,6 +49,7 @@ obj-$(CONFIG_SENSORS_EMC2103) += emc2103.o
obj-$(CONFIG_SENSORS_F71805F) += f71805f.o
obj-$(CONFIG_SENSORS_F71882FG) += f71882fg.o
obj-$(CONFIG_SENSORS_F75375S) += f75375s.o
+obj-$(CONFIG_SENSORS_FAM15H_POWER) += fam15h_power.o
obj-$(CONFIG_SENSORS_FSCHMD) += fschmd.o
obj-$(CONFIG_SENSORS_G760A) += g760a.o
obj-$(CONFIG_SENSORS_GL518SM) += gl518sm.o
diff --git a/drivers/hwmon/fam15h_power.c b/drivers/hwmon/fam15h_power.c
new file mode 100644
index 0000000..cb6eb99
--- /dev/null
+++ b/drivers/hwmon/fam15h_power.c
@@ -0,0 +1,229 @@
+/*
+ * fam15h_power.c - AMD Family 15h processor power monitoring
+ *
+ * Copyright (c) 2011 Advanced Micro Devices, Inc.
+ * Author: Andreas Herrmann <[email protected]>
+ *
+ *
+ * This driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this driver; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/err.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/bitops.h>
+#include <asm/processor.h>
+
+MODULE_DESCRIPTION("AMD Family 15h CPU processor power monitor");
+MODULE_AUTHOR("Andreas Herrmann <[email protected]>");
+MODULE_LICENSE("GPL");
+
+/* D18F3 */
+#define REG_NORTHBRIDGE_CAP 0xe8
+
+/* D18F4 */
+#define REG_PROCESSOR_TDP 0x1b8
+
+/* D18F5 */
+#define REG_TDP_RUNNING_AVERAGE 0xe0
+#define REG_TDP_LIMIT3 0xe8
+
+struct fam15h_power_data {
+ struct device *hwmon_dev;
+ unsigned int tdp_to_watt;
+ unsigned int base_tdp;
+ unsigned int processor_pwr_watts;
+};
+
+static ssize_t show_power(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ u32 val, tdp_limit, running_avg_range;
+ s32 running_avg_capture;
+ u64 curr_pwr_watts;
+ struct pci_dev *f4 = to_pci_dev(dev);
+ struct fam15h_power_data *data = pci_get_drvdata(f4);
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
+ REG_TDP_RUNNING_AVERAGE, &val);
+ running_avg_capture = (val >> 4) & 0x3fffff;
+ running_avg_capture = sign_extend32(running_avg_capture, 22);
+ running_avg_range = val & 0xf;
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
+ REG_TDP_LIMIT3, &val);
+
+ tdp_limit = val >> 16;
+ curr_pwr_watts = tdp_limit + data->base_tdp -
+ (s32)(running_avg_capture >> (running_avg_range + 1));
+ curr_pwr_watts *= data->tdp_to_watt;
+
+ /*
+ * Convert to microWatt
+ *
+ * power is in Watt provided as fixed point integer with
+ * scaling factor 1/(2^16). For conversion we use
+ * (10^6)/(2^16) = 15625/(2^10)
+ */
+ curr_pwr_watts = (curr_pwr_watts * 15625) >> 10;
+ return sprintf(buf, "%u\n", (unsigned int) curr_pwr_watts);
+}
+static DEVICE_ATTR(power1_input, S_IRUGO, show_power, NULL);
+
+static ssize_t show_power_max(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ struct pci_dev *f4 = to_pci_dev(dev);
+ struct fam15h_power_data *data = pci_get_drvdata(f4);
+
+ return sprintf(buf, "%u\n", data->processor_pwr_watts);
+}
+static DEVICE_ATTR(power1_max, S_IRUGO, show_power_max, NULL);
+
+static ssize_t show_name(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ return sprintf(buf, "fam15h_power\n");
+}
+static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
+
+static struct attribute *fam15h_power_attrs[] = {
+ &dev_attr_power1_input.attr,
+ &dev_attr_power1_max.attr,
+ &dev_attr_name.attr,
+ NULL
+};
+
+static struct attribute_group fam15h_power_attr_group = {
+ .attrs = fam15h_power_attrs,
+};
+
+static bool __devinit fam15h_power_is_internal_node0(struct pci_dev *f4)
+{
+ u32 val;
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3),
+ REG_NORTHBRIDGE_CAP, &val);
+ if ((val & BIT(29)) && ((val >> 30) & 3))
+ return false;
+
+ return true;
+}
+
+static void __devinit fam15h_power_init_data(struct pci_dev *f4,
+ struct fam15h_power_data *data)
+{
+ u32 val;
+ u64 tmp;
+
+ pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
+ data->base_tdp = val >> 16;
+ tmp = val & 0xffff;
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
+ REG_TDP_LIMIT3, &val);
+
+ data->tdp_to_watt = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
+ tmp *= data->tdp_to_watt;
+
+ /* result not allowed to be >= 256W */
+ if ((tmp>>16) >= 256)
+ dev_warn(&f4->dev, "Bogus value for ProcessorPwrWatts "
+ "(processor_pwr_watts>=%u)\n",
+ (unsigned int) (tmp >> 16));
+
+ /* convert to microWatt */
+ data->processor_pwr_watts = (tmp * 15625) >> 10;
+}
+
+static int __devinit fam15h_power_probe(struct pci_dev *pdev,
+ const struct pci_device_id *id)
+{
+ struct fam15h_power_data *data;
+ struct device *dev;
+ int err;
+
+ if (!fam15h_power_is_internal_node0(pdev)) {
+ err = -ENODEV;
+ goto exit;
+ }
+
+ data = kzalloc(sizeof(struct fam15h_power_data), GFP_KERNEL);
+ if (!data) {
+ err = -ENOMEM;
+ goto exit;
+ }
+ fam15h_power_init_data(pdev, data);
+
+ dev = &pdev->dev;
+ err = sysfs_create_group(&dev->kobj, &fam15h_power_attr_group);
+ if (err)
+ goto exit_free_data;
+
+ data->hwmon_dev = hwmon_device_register(&pdev->dev);
+ if (IS_ERR(data->hwmon_dev)) {
+ err = PTR_ERR(data->hwmon_dev);
+ goto exit_remove_group;
+ }
+ pci_set_drvdata(pdev, data);
+
+ return 0;
+
+exit_remove_group:
+ sysfs_remove_group(&dev->kobj, &fam15h_power_attr_group);
+exit_free_data:
+ kfree(data);
+exit:
+ return err;
+}
+
+static void __devexit fam15h_power_remove(struct pci_dev *pdev)
+{
+ struct device *dev;
+ struct fam15h_power_data *data = pci_get_drvdata(pdev);
+
+ dev = &pdev->dev;
+ hwmon_device_unregister(data->hwmon_dev);
+ sysfs_remove_group(&dev->kobj, &fam15h_power_attr_group);
+ pci_set_drvdata(pdev, NULL);
+ kfree(data);
+}
+
+static DEFINE_PCI_DEVICE_TABLE(fam15h_power_id_table) = {
+ { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_F4) },
+ {}
+};
+MODULE_DEVICE_TABLE(pci, fam15h_power_id_table);
+
+static struct pci_driver fam15h_power_driver = {
+ .name = "fam15h_power",
+ .id_table = fam15h_power_id_table,
+ .probe = fam15h_power_probe,
+ .remove = __devexit_p(fam15h_power_remove),
+};
+
+static int __init fam15h_power_init(void)
+{
+ return pci_register_driver(&fam15h_power_driver);
+}
+
+static void __exit fam15h_power_exit(void)
+{
+ pci_unregister_driver(&fam15h_power_driver);
+}
+
+module_init(fam15h_power_init)
+module_exit(fam15h_power_exit)
--
1.7.4.1
From: Andreas Herrmann <[email protected]>
Signed-off-by: Andreas Herrmann <[email protected]>
---
MAINTAINERS | 7 +++++++
1 files changed, 7 insertions(+), 0 deletions(-)
As suggested by Jean, here's a patch to add a maintainers
entry for the new driver.
Regards,
Andreas
diff --git a/MAINTAINERS b/MAINTAINERS
index 6b4b9cd..af8421d 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -483,6 +483,13 @@ F: drivers/tty/serial/altera_jtaguart.c
F: include/linux/altera_uart.h
F: include/linux/altera_jtaguart.h
+AMD FAM15H PROCESSOR POWER HARDWARE MONITORING DRIVER
+M: Andreas Herrmann <[email protected]>
+L: [email protected]
+S: Maintained
+F: Documentation/hwmon/fam15h_power
+F: drivers/hwmon/fam15h_power.c
+
AMD GEODE CS5536 USB DEVICE CONTROLLER DRIVER
M: Thomas Dahlmann <[email protected]>
L: [email protected] (moderated for non-subscribers)
--
1.7.4.1
Hi Andreas,
On Fri, 8 Apr 2011 15:54:07 +0200, Andreas Herrmann wrote:
> From: Andreas Herrmann <[email protected]>
>
> This CPU family provides NB register values to gather following
> TDP information
>
> * ProcessorPwrWatts: Specifies in Watts the maximum amount of power
> the processor can support.
> * CurrPwrWatts: Specifies in Watts the current amount of power being
> consumed by the processor.
>
> This driver provides
>
> * power1_max (ProcessorPwrWatts)
Still questionable whether power1_crit would be more appropriate...
> * power1_input (CurrPwrWatts)
>
> Changes from v2:
> - fix format strings
> - removed paranoid checks for existense of functions 3 and 5
> - changed return type of function f15h_power_is_internal_node0
> - use power1_max instead of power1_cap
> - use dev_warn instead of WARN_ON
> - rebased against 2.6.39-rc2
> - added Documentation/hwmon/f15h_power
>
> Changes from v3:
> - read static power information only once (during driver initialization)
> - made use of attribute groups
> - renamed f15h_power to fam15h_power
>
> Signed-off-by: Andreas Herrmann <[email protected]>
> ---
> Documentation/hwmon/fam15h_power | 37 ++++++
> drivers/hwmon/Kconfig | 10 ++
> drivers/hwmon/Makefile | 1 +
> drivers/hwmon/fam15h_power.c | 229 ++++++++++++++++++++++++++++++++++++++
> 4 files changed, 277 insertions(+), 0 deletions(-)
> create mode 100644 Documentation/hwmon/fam15h_power
> create mode 100644 drivers/hwmon/fam15h_power.c
>
> At this stage I consider the _max attribute as the right one for
> reporting ProcessorPwrWatts.
>
> Hopefully I've addressed all your coments.
> Please apply.
It seems I have some more comments, but these are very small things you
should have no problem addressing quickly:
> diff --git a/Documentation/hwmon/fam15h_power b/Documentation/hwmon/fam15h_power
> new file mode 100644
> index 0000000..2f4d291
> --- /dev/null
> +++ b/Documentation/hwmon/fam15h_power
> @@ -0,0 +1,37 @@
> +Kernel driver fam15h_power
> +==========================
> +
> +Supported chips:
> +* AMD Family 15h Processors
> +
> + Prefix: 'fam15h_power'
> + Addresses scanned: PCI space
> + Datasheets:
> + BIOS and Kernel Developer's Guide (BKDG) For AMD Family 15h Processors
> + (not yet published)
> +
> +Author: Andreas Herrmann <[email protected]>
> +
> +Description
> +-----------
> +
> +This driver permits reading of registers providing power information
> +of AMD Family 15h processors.
> +
> +For AMD Family 15h processors the following power values can be
> +calculated using different processor northbridge function registers:
> +
> +* BasePwrWatts: Specifies in watts the maximum amount of power
> + consumed by the processor for NB and logic external to the core.
> +* ProcessorPwrWatts: Specifies in watts the maximum amount of power
> + the processor can support.
> +* CurrPwrWatts: Specifies in watts the current amount of power being
> + consumed by the processor.
> +
> +This driver provides ProcessorPwrWatts and CurrPwrWatts:
> +* power1_max (ProcessorPwrWatts)
> +* power1_input (CurrPwrWatts)
> +
> +On multi-node processors the calculated value is for the entire
> +package and not for a single node. Thus the driver creates sysfs
> +attributes only for internal node0 of a multi-node processor.
> diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
> index 060ef63..fb3e334 100644
> --- a/drivers/hwmon/Kconfig
> +++ b/drivers/hwmon/Kconfig
> @@ -249,6 +249,16 @@ config SENSORS_K10TEMP
> This driver can also be built as a module. If so, the module
> will be called k10temp.
>
> +config SENSORS_FAM15H_POWER
> + tristate "AMD Family 15h processor power"
> + depends on X86 && PCI
> + help
> + If you say yes here you get support for processor power
> + information of your AMD family 15h CPU.
> +
> + This driver can also be built as a module. If so, the module
> + will be called fam15h_power.
> +
> config SENSORS_ASB100
> tristate "Asus ASB100 Bach"
> depends on X86 && I2C && EXPERIMENTAL
> diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
> index 967d0ea..236d3f9 100644
> --- a/drivers/hwmon/Makefile
> +++ b/drivers/hwmon/Makefile
> @@ -49,6 +49,7 @@ obj-$(CONFIG_SENSORS_EMC2103) += emc2103.o
> obj-$(CONFIG_SENSORS_F71805F) += f71805f.o
> obj-$(CONFIG_SENSORS_F71882FG) += f71882fg.o
> obj-$(CONFIG_SENSORS_F75375S) += f75375s.o
> +obj-$(CONFIG_SENSORS_FAM15H_POWER) += fam15h_power.o
> obj-$(CONFIG_SENSORS_FSCHMD) += fschmd.o
> obj-$(CONFIG_SENSORS_G760A) += g760a.o
> obj-$(CONFIG_SENSORS_GL518SM) += gl518sm.o
> diff --git a/drivers/hwmon/fam15h_power.c b/drivers/hwmon/fam15h_power.c
> new file mode 100644
> index 0000000..cb6eb99
> --- /dev/null
> +++ b/drivers/hwmon/fam15h_power.c
> @@ -0,0 +1,229 @@
> +/*
> + * fam15h_power.c - AMD Family 15h processor power monitoring
> + *
> + * Copyright (c) 2011 Advanced Micro Devices, Inc.
> + * Author: Andreas Herrmann <[email protected]>
> + *
> + *
> + * This driver is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This driver is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
> + * See the GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this driver; if not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include <linux/err.h>
> +#include <linux/hwmon.h>
> +#include <linux/hwmon-sysfs.h>
> +#include <linux/init.h>
> +#include <linux/module.h>
> +#include <linux/pci.h>
> +#include <linux/bitops.h>
> +#include <asm/processor.h>
> +
> +MODULE_DESCRIPTION("AMD Family 15h CPU processor power monitor");
> +MODULE_AUTHOR("Andreas Herrmann <[email protected]>");
> +MODULE_LICENSE("GPL");
> +
> +/* D18F3 */
> +#define REG_NORTHBRIDGE_CAP 0xe8
> +
> +/* D18F4 */
> +#define REG_PROCESSOR_TDP 0x1b8
> +
> +/* D18F5 */
> +#define REG_TDP_RUNNING_AVERAGE 0xe0
> +#define REG_TDP_LIMIT3 0xe8
> +
> +struct fam15h_power_data {
> + struct device *hwmon_dev;
> + unsigned int tdp_to_watt;
> + unsigned int base_tdp;
> + unsigned int processor_pwr_watts;
watt vs. watts is inconsistent, isn't it?
> +};
> +
> +static ssize_t show_power(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + u32 val, tdp_limit, running_avg_range;
> + s32 running_avg_capture;
> + u64 curr_pwr_watts;
> + struct pci_dev *f4 = to_pci_dev(dev);
> + struct fam15h_power_data *data = pci_get_drvdata(f4);
dev_get_drvdata(dev) would be more efficient.
> +
> + pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
> + REG_TDP_RUNNING_AVERAGE, &val);
> + running_avg_capture = (val >> 4) & 0x3fffff;
> + running_avg_capture = sign_extend32(running_avg_capture, 22);
> + running_avg_range = val & 0xf;
> +
> + pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
> + REG_TDP_LIMIT3, &val);
> +
> + tdp_limit = val >> 16;
> + curr_pwr_watts = tdp_limit + data->base_tdp -
Doubled space.
> + (s32)(running_avg_capture >> (running_avg_range + 1));
> + curr_pwr_watts *= data->tdp_to_watt;
> +
> + /*
> + * Convert to microWatt
> + *
> + * power is in Watt provided as fixed point integer with
> + * scaling factor 1/(2^16). For conversion we use
> + * (10^6)/(2^16) = 15625/(2^10)
> + */
> + curr_pwr_watts = (curr_pwr_watts * 15625) >> 10;
> + return sprintf(buf, "%u\n", (unsigned int) curr_pwr_watts);
> +}
> +static DEVICE_ATTR(power1_input, S_IRUGO, show_power, NULL);
> +
> +static ssize_t show_power_max(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + struct pci_dev *f4 = to_pci_dev(dev);
> + struct fam15h_power_data *data = pci_get_drvdata(f4);
dev_get_drvdata(dev) would be more efficient (you don't even need f4
then.)
> + return sprintf(buf, "%u\n", data->processor_pwr_watts);
> +}
> +static DEVICE_ATTR(power1_max, S_IRUGO, show_power_max, NULL);
> +
> +static ssize_t show_name(struct device *dev,
> + struct device_attribute *attr, char *buf)
> +{
> + return sprintf(buf, "fam15h_power\n");
> +}
> +static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
> +
> +static struct attribute *fam15h_power_attrs[] = {
> + &dev_attr_power1_input.attr,
> + &dev_attr_power1_max.attr,
> + &dev_attr_name.attr,
> + NULL
> +};
> +
> +static struct attribute_group fam15h_power_attr_group = {
Can be made const.
> + .attrs = fam15h_power_attrs,
> +};
> +
> +static bool __devinit fam15h_power_is_internal_node0(struct pci_dev *f4)
> +{
> + u32 val;
> +
> + pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3),
> + REG_NORTHBRIDGE_CAP, &val);
> + if ((val & BIT(29)) && ((val >> 30) & 3))
> + return false;
> +
> + return true;
> +}
> +
> +static void __devinit fam15h_power_init_data(struct pci_dev *f4,
> + struct fam15h_power_data *data)
> +{
> + u32 val;
> + u64 tmp;
> +
> + pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
> + data->base_tdp = val >> 16;
> + tmp = val & 0xffff;
> +
> + pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
> + REG_TDP_LIMIT3, &val);
> +
> + data->tdp_to_watt = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
> + tmp *= data->tdp_to_watt;
> +
> + /* result not allowed to be >= 256W */
> + if ((tmp>>16) >= 256)
Please add spaces around ">>" for consistency.
> + dev_warn(&f4->dev, "Bogus value for ProcessorPwrWatts "
> + "(processor_pwr_watts>=%u)\n",
> + (unsigned int) (tmp >> 16));
> +
> + /* convert to microWatt */
> + data->processor_pwr_watts = (tmp * 15625) >> 10;
> +}
> +
> +static int __devinit fam15h_power_probe(struct pci_dev *pdev,
> + const struct pci_device_id *id)
> +{
> + struct fam15h_power_data *data;
> + struct device *dev;
> + int err;
> +
> + if (!fam15h_power_is_internal_node0(pdev)) {
> + err = -ENODEV;
> + goto exit;
> + }
> +
> + data = kzalloc(sizeof(struct fam15h_power_data), GFP_KERNEL);
> + if (!data) {
> + err = -ENOMEM;
> + goto exit;
> + }
> + fam15h_power_init_data(pdev, data);
> +
> + dev = &pdev->dev;
> + err = sysfs_create_group(&dev->kobj, &fam15h_power_attr_group);
> + if (err)
> + goto exit_free_data;
> +
> + data->hwmon_dev = hwmon_device_register(&pdev->dev);
> + if (IS_ERR(data->hwmon_dev)) {
> + err = PTR_ERR(data->hwmon_dev);
> + goto exit_remove_group;
> + }
> + pci_set_drvdata(pdev, data);
This is racy. pci_set_drvdata() should be called before creating the
sysfs attributes, because the sysfs callbacks need it.
> +
> + return 0;
> +
> +exit_remove_group:
> + sysfs_remove_group(&dev->kobj, &fam15h_power_attr_group);
> +exit_free_data:
> + kfree(data);
> +exit:
> + return err;
> +}
> +
> +static void __devexit fam15h_power_remove(struct pci_dev *pdev)
> +{
> + struct device *dev;
> + struct fam15h_power_data *data = pci_get_drvdata(pdev);
> +
> + dev = &pdev->dev;
> + hwmon_device_unregister(data->hwmon_dev);
> + sysfs_remove_group(&dev->kobj, &fam15h_power_attr_group);
> + pci_set_drvdata(pdev, NULL);
> + kfree(data);
> +}
> +
> +static DEFINE_PCI_DEVICE_TABLE(fam15h_power_id_table) = {
> + { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_F4) },
> + {}
> +};
> +MODULE_DEVICE_TABLE(pci, fam15h_power_id_table);
> +
> +static struct pci_driver fam15h_power_driver = {
> + .name = "fam15h_power",
> + .id_table = fam15h_power_id_table,
> + .probe = fam15h_power_probe,
> + .remove = __devexit_p(fam15h_power_remove),
> +};
> +
> +static int __init fam15h_power_init(void)
> +{
> + return pci_register_driver(&fam15h_power_driver);
> +}
> +
> +static void __exit fam15h_power_exit(void)
> +{
> + pci_unregister_driver(&fam15h_power_driver);
> +}
> +
> +module_init(fam15h_power_init)
> +module_exit(fam15h_power_exit)
Thanks,
--
Jean Delvare
Hi Andreas,
On Fri, 8 Apr 2011 15:57:46 +0200, Andreas Herrmann wrote:
> From: Andreas Herrmann <[email protected]>
>
> Signed-off-by: Andreas Herrmann <[email protected]>
> ---
> MAINTAINERS | 7 +++++++
> 1 files changed, 7 insertions(+), 0 deletions(-)
>
> As suggested by Jean, here's a patch to add a maintainers
> entry for the new driver.
>
>
> Regards,
>
> Andreas
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 6b4b9cd..af8421d 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -483,6 +483,13 @@ F: drivers/tty/serial/altera_jtaguart.c
> F: include/linux/altera_uart.h
> F: include/linux/altera_jtaguart.h
>
> +AMD FAM15H PROCESSOR POWER HARDWARE MONITORING DRIVER
I'd remove "HARDWARE" as the target hardware is already described.
> +M: Andreas Herrmann <[email protected]>
> +L: [email protected]
> +S: Maintained
> +F: Documentation/hwmon/fam15h_power
> +F: drivers/hwmon/fam15h_power.c
> +
> AMD GEODE CS5536 USB DEVICE CONTROLLER DRIVER
> M: Thomas Dahlmann <[email protected]>
> L: [email protected] (moderated for non-subscribers)
Other than this small detail, which I can fix myself if you agree with
me, this looks good, thanks for stepping in!
--
Jean Delvare
On Wed, Apr 13, 2011 at 03:08:55PM +0200, Jean Delvare wrote:
> Hi Andreas,
>
> On Fri, 8 Apr 2011 15:57:46 +0200, Andreas Herrmann wrote:
> > From: Andreas Herrmann <[email protected]>
> >
> > Signed-off-by: Andreas Herrmann <[email protected]>
> > ---
> > MAINTAINERS | 7 +++++++
> > 1 files changed, 7 insertions(+), 0 deletions(-)
> >
> > As suggested by Jean, here's a patch to add a maintainers
> > entry for the new driver.
> >
> >
> > Regards,
> >
> > Andreas
> >
> > diff --git a/MAINTAINERS b/MAINTAINERS
> > index 6b4b9cd..af8421d 100644
> > --- a/MAINTAINERS
> > +++ b/MAINTAINERS
> > @@ -483,6 +483,13 @@ F: drivers/tty/serial/altera_jtaguart.c
> > F: include/linux/altera_uart.h
> > F: include/linux/altera_jtaguart.h
> >
> > +AMD FAM15H PROCESSOR POWER HARDWARE MONITORING DRIVER
>
> I'd remove "HARDWARE" as the target hardware is already described.
>
> > +M: Andreas Herrmann <[email protected]>
> > +L: [email protected]
> > +S: Maintained
> > +F: Documentation/hwmon/fam15h_power
> > +F: drivers/hwmon/fam15h_power.c
> > +
> > AMD GEODE CS5536 USB DEVICE CONTROLLER DRIVER
> > M: Thomas Dahlmann <[email protected]>
> > L: [email protected] (moderated for non-subscribers)
>
> Other than this small detail, which I can fix myself if you agree with
> me, this looks good, thanks for stepping in!
I agree.
Thanks,
Andreas
On Wed, Apr 13, 2011 at 03:06:33PM +0200, Jean Delvare wrote:
> Hi Andreas,
>
> On Fri, 8 Apr 2011 15:54:07 +0200, Andreas Herrmann wrote:
> > From: Andreas Herrmann <[email protected]>
> >
> > This CPU family provides NB register values to gather following
> > TDP information
> >
> > * ProcessorPwrWatts: Specifies in Watts the maximum amount of power
> > the processor can support.
> > * CurrPwrWatts: Specifies in Watts the current amount of power being
> > consumed by the processor.
> >
> > This driver provides
> >
> > * power1_max (ProcessorPwrWatts)
>
> Still questionable whether power1_crit would be more appropriate...
Finally, I've changed it.
> > * power1_input (CurrPwrWatts)
> >
> > Changes from v2:
> > - fix format strings
> > - removed paranoid checks for existense of functions 3 and 5
> > - changed return type of function f15h_power_is_internal_node0
> > - use power1_max instead of power1_cap
> > - use dev_warn instead of WARN_ON
> > - rebased against 2.6.39-rc2
> > - added Documentation/hwmon/f15h_power
> >
> > Changes from v3:
> > - read static power information only once (during driver initialization)
> > - made use of attribute groups
> > - renamed f15h_power to fam15h_power
> >
> > Signed-off-by: Andreas Herrmann <[email protected]>
> > ---
> > Documentation/hwmon/fam15h_power | 37 ++++++
> > drivers/hwmon/Kconfig | 10 ++
> > drivers/hwmon/Makefile | 1 +
> > drivers/hwmon/fam15h_power.c | 229 ++++++++++++++++++++++++++++++++++++++
> > 4 files changed, 277 insertions(+), 0 deletions(-)
> > create mode 100644 Documentation/hwmon/fam15h_power
> > create mode 100644 drivers/hwmon/fam15h_power.c
> >
> > At this stage I consider the _max attribute as the right one for
> > reporting ProcessorPwrWatts.
> >
> > Hopefully I've addressed all your coments.
> > Please apply.
>
> It seems I have some more comments, but these are very small things you
> should have no problem addressing quickly:
>
> > diff --git a/Documentation/hwmon/fam15h_power b/Documentation/hwmon/fam15h_power
> > new file mode 100644
> > index 0000000..2f4d291
> > --- /dev/null
> > +++ b/Documentation/hwmon/fam15h_power
> > @@ -0,0 +1,37 @@
> > +Kernel driver fam15h_power
> > +==========================
> > +
> > +Supported chips:
> > +* AMD Family 15h Processors
> > +
> > + Prefix: 'fam15h_power'
> > + Addresses scanned: PCI space
> > + Datasheets:
> > + BIOS and Kernel Developer's Guide (BKDG) For AMD Family 15h Processors
> > + (not yet published)
> > +
> > +Author: Andreas Herrmann <[email protected]>
> > +
> > +Description
> > +-----------
> > +
> > +This driver permits reading of registers providing power information
> > +of AMD Family 15h processors.
> > +
> > +For AMD Family 15h processors the following power values can be
> > +calculated using different processor northbridge function registers:
> > +
> > +* BasePwrWatts: Specifies in watts the maximum amount of power
> > + consumed by the processor for NB and logic external to the core.
> > +* ProcessorPwrWatts: Specifies in watts the maximum amount of power
> > + the processor can support.
> > +* CurrPwrWatts: Specifies in watts the current amount of power being
> > + consumed by the processor.
> > +
> > +This driver provides ProcessorPwrWatts and CurrPwrWatts:
> > +* power1_max (ProcessorPwrWatts)
> > +* power1_input (CurrPwrWatts)
> > +
> > +On multi-node processors the calculated value is for the entire
> > +package and not for a single node. Thus the driver creates sysfs
> > +attributes only for internal node0 of a multi-node processor.
> > diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
> > index 060ef63..fb3e334 100644
> > --- a/drivers/hwmon/Kconfig
> > +++ b/drivers/hwmon/Kconfig
> > @@ -249,6 +249,16 @@ config SENSORS_K10TEMP
> > This driver can also be built as a module. If so, the module
> > will be called k10temp.
> >
> > +config SENSORS_FAM15H_POWER
> > + tristate "AMD Family 15h processor power"
> > + depends on X86 && PCI
> > + help
> > + If you say yes here you get support for processor power
> > + information of your AMD family 15h CPU.
> > +
> > + This driver can also be built as a module. If so, the module
> > + will be called fam15h_power.
> > +
> > config SENSORS_ASB100
> > tristate "Asus ASB100 Bach"
> > depends on X86 && I2C && EXPERIMENTAL
> > diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
> > index 967d0ea..236d3f9 100644
> > --- a/drivers/hwmon/Makefile
> > +++ b/drivers/hwmon/Makefile
> > @@ -49,6 +49,7 @@ obj-$(CONFIG_SENSORS_EMC2103) += emc2103.o
> > obj-$(CONFIG_SENSORS_F71805F) += f71805f.o
> > obj-$(CONFIG_SENSORS_F71882FG) += f71882fg.o
> > obj-$(CONFIG_SENSORS_F75375S) += f75375s.o
> > +obj-$(CONFIG_SENSORS_FAM15H_POWER) += fam15h_power.o
> > obj-$(CONFIG_SENSORS_FSCHMD) += fschmd.o
> > obj-$(CONFIG_SENSORS_G760A) += g760a.o
> > obj-$(CONFIG_SENSORS_GL518SM) += gl518sm.o
> > diff --git a/drivers/hwmon/fam15h_power.c b/drivers/hwmon/fam15h_power.c
> > new file mode 100644
> > index 0000000..cb6eb99
> > --- /dev/null
> > +++ b/drivers/hwmon/fam15h_power.c
> > @@ -0,0 +1,229 @@
> > +/*
> > + * fam15h_power.c - AMD Family 15h processor power monitoring
> > + *
> > + * Copyright (c) 2011 Advanced Micro Devices, Inc.
> > + * Author: Andreas Herrmann <[email protected]>
> > + *
> > + *
> > + * This driver is free software; you can redistribute it and/or
> > + * modify it under the terms of the GNU General Public License; either
> > + * version 2 of the License, or (at your option) any later version.
> > + *
> > + * This driver is distributed in the hope that it will be useful,
> > + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
> > + * See the GNU General Public License for more details.
> > + *
> > + * You should have received a copy of the GNU General Public License
> > + * along with this driver; if not, see <http://www.gnu.org/licenses/>.
> > + */
> > +
> > +#include <linux/err.h>
> > +#include <linux/hwmon.h>
> > +#include <linux/hwmon-sysfs.h>
> > +#include <linux/init.h>
> > +#include <linux/module.h>
> > +#include <linux/pci.h>
> > +#include <linux/bitops.h>
> > +#include <asm/processor.h>
> > +
> > +MODULE_DESCRIPTION("AMD Family 15h CPU processor power monitor");
> > +MODULE_AUTHOR("Andreas Herrmann <[email protected]>");
> > +MODULE_LICENSE("GPL");
> > +
> > +/* D18F3 */
> > +#define REG_NORTHBRIDGE_CAP 0xe8
> > +
> > +/* D18F4 */
> > +#define REG_PROCESSOR_TDP 0x1b8
> > +
> > +/* D18F5 */
> > +#define REG_TDP_RUNNING_AVERAGE 0xe0
> > +#define REG_TDP_LIMIT3 0xe8
> > +
> > +struct fam15h_power_data {
> > + struct device *hwmon_dev;
> > + unsigned int tdp_to_watt;
> > + unsigned int base_tdp;
> > + unsigned int processor_pwr_watts;
>
> watt vs. watts is inconsistent, isn't it?
Changed.
> > +};
> > +
> > +static ssize_t show_power(struct device *dev,
> > + struct device_attribute *attr, char *buf)
> > +{
> > + u32 val, tdp_limit, running_avg_range;
> > + s32 running_avg_capture;
> > + u64 curr_pwr_watts;
> > + struct pci_dev *f4 = to_pci_dev(dev);
> > + struct fam15h_power_data *data = pci_get_drvdata(f4);
>
> dev_get_drvdata(dev) would be more efficient.
>
> > +
> > + pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
> > + REG_TDP_RUNNING_AVERAGE, &val);
> > + running_avg_capture = (val >> 4) & 0x3fffff;
> > + running_avg_capture = sign_extend32(running_avg_capture, 22);
> > + running_avg_range = val & 0xf;
> > +
> > + pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
> > + REG_TDP_LIMIT3, &val);
> > +
> > + tdp_limit = val >> 16;
> > + curr_pwr_watts = tdp_limit + data->base_tdp -
>
> Doubled space.
Whoops!
> > + (s32)(running_avg_capture >> (running_avg_range + 1));
> > + curr_pwr_watts *= data->tdp_to_watt;
> > +
> > + /*
> > + * Convert to microWatt
> > + *
> > + * power is in Watt provided as fixed point integer with
> > + * scaling factor 1/(2^16). For conversion we use
> > + * (10^6)/(2^16) = 15625/(2^10)
> > + */
> > + curr_pwr_watts = (curr_pwr_watts * 15625) >> 10;
> > + return sprintf(buf, "%u\n", (unsigned int) curr_pwr_watts);
> > +}
> > +static DEVICE_ATTR(power1_input, S_IRUGO, show_power, NULL);
> > +
> > +static ssize_t show_power_max(struct device *dev,
> > + struct device_attribute *attr, char *buf)
> > +{
> > + struct pci_dev *f4 = to_pci_dev(dev);
> > + struct fam15h_power_data *data = pci_get_drvdata(f4);
>
> dev_get_drvdata(dev) would be more efficient (you don't even need f4
> then.)
>
> > + return sprintf(buf, "%u\n", data->processor_pwr_watts);
> > +}
> > +static DEVICE_ATTR(power1_max, S_IRUGO, show_power_max, NULL);
> > +
> > +static ssize_t show_name(struct device *dev,
> > + struct device_attribute *attr, char *buf)
> > +{
> > + return sprintf(buf, "fam15h_power\n");
> > +}
> > +static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
> > +
> > +static struct attribute *fam15h_power_attrs[] = {
> > + &dev_attr_power1_input.attr,
> > + &dev_attr_power1_max.attr,
> > + &dev_attr_name.attr,
> > + NULL
> > +};
> > +
> > +static struct attribute_group fam15h_power_attr_group = {
>
> Can be made const.
Done this ....
> > + .attrs = fam15h_power_attrs,
> > +};
> > +
> > +static bool __devinit fam15h_power_is_internal_node0(struct pci_dev *f4)
> > +{
> > + u32 val;
> > +
> > + pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3),
> > + REG_NORTHBRIDGE_CAP, &val);
> > + if ((val & BIT(29)) && ((val >> 30) & 3))
> > + return false;
> > +
> > + return true;
> > +}
> > +
> > +static void __devinit fam15h_power_init_data(struct pci_dev *f4,
> > + struct fam15h_power_data *data)
> > +{
> > + u32 val;
> > + u64 tmp;
> > +
> > + pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
> > + data->base_tdp = val >> 16;
> > + tmp = val & 0xffff;
> > +
> > + pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
> > + REG_TDP_LIMIT3, &val);
> > +
> > + data->tdp_to_watt = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
> > + tmp *= data->tdp_to_watt;
> > +
> > + /* result not allowed to be >= 256W */
> > + if ((tmp>>16) >= 256)
>
> Please add spaces around ">>" for consistency.
... and that.
> > + dev_warn(&f4->dev, "Bogus value for ProcessorPwrWatts "
> > + "(processor_pwr_watts>=%u)\n",
> > + (unsigned int) (tmp >> 16));
> > +
> > + /* convert to microWatt */
> > + data->processor_pwr_watts = (tmp * 15625) >> 10;
> > +}
> > +
> > +static int __devinit fam15h_power_probe(struct pci_dev *pdev,
> > + const struct pci_device_id *id)
> > +{
> > + struct fam15h_power_data *data;
> > + struct device *dev;
> > + int err;
> > +
> > + if (!fam15h_power_is_internal_node0(pdev)) {
> > + err = -ENODEV;
> > + goto exit;
> > + }
> > +
> > + data = kzalloc(sizeof(struct fam15h_power_data), GFP_KERNEL);
> > + if (!data) {
> > + err = -ENOMEM;
> > + goto exit;
> > + }
> > + fam15h_power_init_data(pdev, data);
> > +
> > + dev = &pdev->dev;
> > + err = sysfs_create_group(&dev->kobj, &fam15h_power_attr_group);
> > + if (err)
> > + goto exit_free_data;
> > +
> > + data->hwmon_dev = hwmon_device_register(&pdev->dev);
> > + if (IS_ERR(data->hwmon_dev)) {
> > + err = PTR_ERR(data->hwmon_dev);
> > + goto exit_remove_group;
> > + }
> > + pci_set_drvdata(pdev, data);
>
> This is racy. pci_set_drvdata() should be called before creating the
> sysfs attributes, because the sysfs callbacks need it.
Oops -- fixed.
> > +
> > + return 0;
[snip]
Thanks again for the review
Andreas
From: Andreas Herrmann <[email protected]>
This CPU family provides NB register values to gather following
TDP information
* ProcessorPwrWatts: Specifies in Watts the maximum amount of power
the processor can support.
* CurrPwrWatts: Specifies in Watts the current amount of power being
consumed by the processor.
This driver provides
* power1_crit (ProcessorPwrWatts)
* power1_input (CurrPwrWatts)
Changes from v2:
- fix format strings
- removed paranoid checks for existense of functions 3 and 5
- changed return type of function f15h_power_is_internal_node0
- use power1_max instead of power1_cap
- use dev_warn instead of WARN_ON
- rebased against 2.6.39-rc2
- added Documentation/hwmon/f15h_power
Changes from v3:
- read static power information only once (during driver initialization)
- made use of attribute groups
- renamed f15h_power to fam15h_power
Changes from v4:
- fixed race in probe function
- use power1_crit instead of powe1_max
- coding style changes
Signed-off-by: Andreas Herrmann <[email protected]>
---
Documentation/hwmon/fam15h_power | 37 ++++++
drivers/hwmon/Kconfig | 10 ++
drivers/hwmon/Makefile | 1 +
drivers/hwmon/fam15h_power.c | 229 ++++++++++++++++++++++++++++++++++++++
4 files changed, 277 insertions(+), 0 deletions(-)
create mode 100644 Documentation/hwmon/fam15h_power
create mode 100644 drivers/hwmon/fam15h_power.c
Patch is rebased on v2.6.39-rc3-82-g85f2e68
Andreas
diff --git a/Documentation/hwmon/fam15h_power b/Documentation/hwmon/fam15h_power
new file mode 100644
index 0000000..a92918e
--- /dev/null
+++ b/Documentation/hwmon/fam15h_power
@@ -0,0 +1,37 @@
+Kernel driver fam15h_power
+==========================
+
+Supported chips:
+* AMD Family 15h Processors
+
+ Prefix: 'fam15h_power'
+ Addresses scanned: PCI space
+ Datasheets:
+ BIOS and Kernel Developer's Guide (BKDG) For AMD Family 15h Processors
+ (not yet published)
+
+Author: Andreas Herrmann <[email protected]>
+
+Description
+-----------
+
+This driver permits reading of registers providing power information
+of AMD Family 15h processors.
+
+For AMD Family 15h processors the following power values can be
+calculated using different processor northbridge function registers:
+
+* BasePwrWatts: Specifies in watts the maximum amount of power
+ consumed by the processor for NB and logic external to the core.
+* ProcessorPwrWatts: Specifies in watts the maximum amount of power
+ the processor can support.
+* CurrPwrWatts: Specifies in watts the current amount of power being
+ consumed by the processor.
+
+This driver provides ProcessorPwrWatts and CurrPwrWatts:
+* power1_crit (ProcessorPwrWatts)
+* power1_input (CurrPwrWatts)
+
+On multi-node processors the calculated value is for the entire
+package and not for a single node. Thus the driver creates sysfs
+attributes only for internal node0 of a multi-node processor.
diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
index 060ef63..fb3e334 100644
--- a/drivers/hwmon/Kconfig
+++ b/drivers/hwmon/Kconfig
@@ -249,6 +249,16 @@ config SENSORS_K10TEMP
This driver can also be built as a module. If so, the module
will be called k10temp.
+config SENSORS_FAM15H_POWER
+ tristate "AMD Family 15h processor power"
+ depends on X86 && PCI
+ help
+ If you say yes here you get support for processor power
+ information of your AMD family 15h CPU.
+
+ This driver can also be built as a module. If so, the module
+ will be called fam15h_power.
+
config SENSORS_ASB100
tristate "Asus ASB100 Bach"
depends on X86 && I2C && EXPERIMENTAL
diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
index 967d0ea..236d3f9 100644
--- a/drivers/hwmon/Makefile
+++ b/drivers/hwmon/Makefile
@@ -49,6 +49,7 @@ obj-$(CONFIG_SENSORS_EMC2103) += emc2103.o
obj-$(CONFIG_SENSORS_F71805F) += f71805f.o
obj-$(CONFIG_SENSORS_F71882FG) += f71882fg.o
obj-$(CONFIG_SENSORS_F75375S) += f75375s.o
+obj-$(CONFIG_SENSORS_FAM15H_POWER) += fam15h_power.o
obj-$(CONFIG_SENSORS_FSCHMD) += fschmd.o
obj-$(CONFIG_SENSORS_G760A) += g760a.o
obj-$(CONFIG_SENSORS_GL518SM) += gl518sm.o
diff --git a/drivers/hwmon/fam15h_power.c b/drivers/hwmon/fam15h_power.c
new file mode 100644
index 0000000..523f8fb
--- /dev/null
+++ b/drivers/hwmon/fam15h_power.c
@@ -0,0 +1,229 @@
+/*
+ * fam15h_power.c - AMD Family 15h processor power monitoring
+ *
+ * Copyright (c) 2011 Advanced Micro Devices, Inc.
+ * Author: Andreas Herrmann <[email protected]>
+ *
+ *
+ * This driver is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This driver is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this driver; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/err.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/bitops.h>
+#include <asm/processor.h>
+
+MODULE_DESCRIPTION("AMD Family 15h CPU processor power monitor");
+MODULE_AUTHOR("Andreas Herrmann <[email protected]>");
+MODULE_LICENSE("GPL");
+
+/* D18F3 */
+#define REG_NORTHBRIDGE_CAP 0xe8
+
+/* D18F4 */
+#define REG_PROCESSOR_TDP 0x1b8
+
+/* D18F5 */
+#define REG_TDP_RUNNING_AVERAGE 0xe0
+#define REG_TDP_LIMIT3 0xe8
+
+struct fam15h_power_data {
+ struct device *hwmon_dev;
+ unsigned int tdp_to_watts;
+ unsigned int base_tdp;
+ unsigned int processor_pwr_watts;
+};
+
+static ssize_t show_power(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ u32 val, tdp_limit, running_avg_range;
+ s32 running_avg_capture;
+ u64 curr_pwr_watts;
+ struct pci_dev *f4 = to_pci_dev(dev);
+ struct fam15h_power_data *data = dev_get_drvdata(dev);
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
+ REG_TDP_RUNNING_AVERAGE, &val);
+ running_avg_capture = (val >> 4) & 0x3fffff;
+ running_avg_capture = sign_extend32(running_avg_capture, 22);
+ running_avg_range = val & 0xf;
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
+ REG_TDP_LIMIT3, &val);
+
+ tdp_limit = val >> 16;
+ curr_pwr_watts = tdp_limit + data->base_tdp -
+ (s32)(running_avg_capture >> (running_avg_range + 1));
+ curr_pwr_watts *= data->tdp_to_watts;
+
+ /*
+ * Convert to microWatt
+ *
+ * power is in Watt provided as fixed point integer with
+ * scaling factor 1/(2^16). For conversion we use
+ * (10^6)/(2^16) = 15625/(2^10)
+ */
+ curr_pwr_watts = (curr_pwr_watts * 15625) >> 10;
+ return sprintf(buf, "%u\n", (unsigned int) curr_pwr_watts);
+}
+static DEVICE_ATTR(power1_input, S_IRUGO, show_power, NULL);
+
+static ssize_t show_power_crit(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ struct fam15h_power_data *data = dev_get_drvdata(dev);
+
+ return sprintf(buf, "%u\n", data->processor_pwr_watts);
+}
+static DEVICE_ATTR(power1_crit, S_IRUGO, show_power_crit, NULL);
+
+static ssize_t show_name(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ return sprintf(buf, "fam15h_power\n");
+}
+static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
+
+static struct attribute *fam15h_power_attrs[] = {
+ &dev_attr_power1_input.attr,
+ &dev_attr_power1_crit.attr,
+ &dev_attr_name.attr,
+ NULL
+};
+
+static const struct attribute_group fam15h_power_attr_group = {
+ .attrs = fam15h_power_attrs,
+};
+
+static bool __devinit fam15h_power_is_internal_node0(struct pci_dev *f4)
+{
+ u32 val;
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 3),
+ REG_NORTHBRIDGE_CAP, &val);
+ if ((val & BIT(29)) && ((val >> 30) & 3))
+ return false;
+
+ return true;
+}
+
+static void __devinit fam15h_power_init_data(struct pci_dev *f4,
+ struct fam15h_power_data *data)
+{
+ u32 val;
+ u64 tmp;
+
+ pci_read_config_dword(f4, REG_PROCESSOR_TDP, &val);
+ data->base_tdp = val >> 16;
+ tmp = val & 0xffff;
+
+ pci_bus_read_config_dword(f4->bus, PCI_DEVFN(PCI_SLOT(f4->devfn), 5),
+ REG_TDP_LIMIT3, &val);
+
+ data->tdp_to_watts = ((val & 0x3ff) << 6) | ((val >> 10) & 0x3f);
+ tmp *= data->tdp_to_watts;
+
+ /* result not allowed to be >= 256W */
+ if ((tmp >> 16) >= 256)
+ dev_warn(&f4->dev, "Bogus value for ProcessorPwrWatts "
+ "(processor_pwr_watts>=%u)\n",
+ (unsigned int) (tmp >> 16));
+
+ /* convert to microWatt */
+ data->processor_pwr_watts = (tmp * 15625) >> 10;
+}
+
+static int __devinit fam15h_power_probe(struct pci_dev *pdev,
+ const struct pci_device_id *id)
+{
+ struct fam15h_power_data *data;
+ struct device *dev;
+ int err;
+
+ if (!fam15h_power_is_internal_node0(pdev)) {
+ err = -ENODEV;
+ goto exit;
+ }
+
+ data = kzalloc(sizeof(struct fam15h_power_data), GFP_KERNEL);
+ if (!data) {
+ err = -ENOMEM;
+ goto exit;
+ }
+ fam15h_power_init_data(pdev, data);
+ dev = &pdev->dev;
+
+ dev_set_drvdata(dev, data);
+ err = sysfs_create_group(&dev->kobj, &fam15h_power_attr_group);
+ if (err)
+ goto exit_free_data;
+
+ data->hwmon_dev = hwmon_device_register(dev);
+ if (IS_ERR(data->hwmon_dev)) {
+ err = PTR_ERR(data->hwmon_dev);
+ goto exit_remove_group;
+ }
+
+ return 0;
+
+exit_remove_group:
+ sysfs_remove_group(&dev->kobj, &fam15h_power_attr_group);
+exit_free_data:
+ kfree(data);
+exit:
+ return err;
+}
+
+static void __devexit fam15h_power_remove(struct pci_dev *pdev)
+{
+ struct device *dev;
+ struct fam15h_power_data *data;
+
+ dev = &pdev->dev;
+ data = dev_get_drvdata(dev);
+ hwmon_device_unregister(data->hwmon_dev);
+ sysfs_remove_group(&dev->kobj, &fam15h_power_attr_group);
+ dev_set_drvdata(dev, NULL);
+ kfree(data);
+}
+
+static DEFINE_PCI_DEVICE_TABLE(fam15h_power_id_table) = {
+ { PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_15H_NB_F4) },
+ {}
+};
+MODULE_DEVICE_TABLE(pci, fam15h_power_id_table);
+
+static struct pci_driver fam15h_power_driver = {
+ .name = "fam15h_power",
+ .id_table = fam15h_power_id_table,
+ .probe = fam15h_power_probe,
+ .remove = __devexit_p(fam15h_power_remove),
+};
+
+static int __init fam15h_power_init(void)
+{
+ return pci_register_driver(&fam15h_power_driver);
+}
+
+static void __exit fam15h_power_exit(void)
+{
+ pci_unregister_driver(&fam15h_power_driver);
+}
+
+module_init(fam15h_power_init)
+module_exit(fam15h_power_exit)
--
1.7.5.rc1
On Thu, 14 Apr 2011 21:20:11 +0200, Andreas Herrmann wrote:
> From: Andreas Herrmann <[email protected]>
>
> This CPU family provides NB register values to gather following
> TDP information
>
> * ProcessorPwrWatts: Specifies in Watts the maximum amount of power
> the processor can support.
> * CurrPwrWatts: Specifies in Watts the current amount of power being
> consumed by the processor.
>
> This driver provides
>
> * power1_crit (ProcessorPwrWatts)
> * power1_input (CurrPwrWatts)
> (...)
Applied, thanks.
--
Jean Delvare