| Message ID | 20210405204810.339763-1-linux@weissschuh.net (mailing list archive) |
|---|---|
| State | Superseded, archived |
| Headers | show |
| Series | [v2] platform/x86: add Gigabyte WMI temperature driver | expand |
Hi Thomas, Thank you for your new driver and thank you for the quick respin addressing Barnabás' request to make it a WMI driver. The code looks good, so merging this should be a no-brainer, yet I'm not sure if I should merge this driver as-is, let me explain. The problem is that I assume that this is based on reverse-engineering? We have some mixes experiences with reverse-engineered WMI drivers, sometimes a feature is not supported yet the wmi_evaluate_method() call happily succeeds. One example of this causing trouble is: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=1797d588af15174d4a4e7159dac8c800538e4f8c At a minimum I think your driver should check in its probe function that gigabyte_wmi_query_integer(GIGABYTE_WMI_TEMPERATURE_QUERY, ...) actually succeeds on the machine the driver is running on chances are that Gigabyte has been using the DEADBEEF-2001-0000-00A0-C90629100000 GUID for ages, where as the 0x125 value for GIGABYTE_WMI_TEMPERATURE_QUERY suggests that this is a pretty new API. It would be good if you can see if you can find some DSDT-s for older gigabyte motherboards attached to various distro's bug-tracking systems or forum posts and see how those respond to an unknown gigabyte_wmi_commandtype. Another open question to make sure this driver is suitable as a generic driver (and does not misbehave) is how to figure out how many temperature sensors there actually are. Perhaps the WMI interface returns an error when you query an out-of-range temperature channel? One option here might be to add a DMI matching table and only load on systems on that table for now. That table could then perhaps also provide labels for each of the temperature channels, which is something which would be nice to have regardless of my worries about how well this driver will work on motherboards on which it has not been tested. You could combine this DMI matching table with a "force" module option to continue with probing on boards which are not on the table to allow users to test and report their results to you. And hopefully after a while, when we're confident that the code works well on most gigabyte boards we can drop the DMI table, or at least only use it for the channel labels. Please don't take this the wrong way; I think it is great that you are working on this. And the quick turnaround of the v2 of this drivers makes me pretty certain that we can figure something out and get this merged. Have you tried contacting Gigabyte about this? I don't think the WMI interface is something which they need to keep secret for competitive reasons, so maybe they can help? Note if they want you to sign a NDA of sorts to view docs, then make sure that it contains some language about them allowing you to release an opensource driver for their hardware based on the "protected" information. Regards, Hans On 4/5/21 10:48 PM, Thomas Weißschuh wrote: > Changes since v1: > * Incorporate feedback from Barnabás Pőcze > * Use a WMI driver instead of a platform driver > * Let the kernel manage the driver lifecycle > * Fix errno/ACPI error confusion > * Fix resource cleanup > * Document reason for integer casting > > Thank you Barnabás for your review, it is much appreciated. > > -- >8 -- > > Tested with a X570 I Aorus Pro Wifi. > The mainboard contains an ITE IT8688E chip for management. > This chips is also handled by drivers/hwmon/i87.c but as it is also used > by the firmware itself it needs an ACPI driver. > > Unfortunately not all sensor registers are handled by the firmware and even > less are exposed via WMI. > > Signed-off-by: Thomas Weißschuh <linux@weissschuh.net> > --- > drivers/platform/x86/Kconfig | 11 +++ > drivers/platform/x86/Makefile | 1 + > drivers/platform/x86/gigabyte-wmi.c | 138 ++++++++++++++++++++++++++++ > 3 files changed, 150 insertions(+) > create mode 100644 drivers/platform/x86/gigabyte-wmi.c > > diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig > index ad4e630e73e2..96622a2106f7 100644 > --- a/drivers/platform/x86/Kconfig > +++ b/drivers/platform/x86/Kconfig > @@ -123,6 +123,17 @@ config XIAOMI_WMI > To compile this driver as a module, choose M here: the module will > be called xiaomi-wmi. > > +config GIGABYTE_WMI > + tristate "Gigabyte WMI temperature driver" > + depends on ACPI_WMI > + depends on HWMON > + help > + Say Y here if you want to support WMI-based temperature reporting on > + Gigabyte mainboards. > + > + To compile this driver as a module, choose M here: the module will > + be called gigabyte-wmi. > + > config ACERHDF > tristate "Acer Aspire One temperature and fan driver" > depends on ACPI && THERMAL > diff --git a/drivers/platform/x86/Makefile b/drivers/platform/x86/Makefile > index 60d554073749..1621ebfd04fd 100644 > --- a/drivers/platform/x86/Makefile > +++ b/drivers/platform/x86/Makefile > @@ -15,6 +15,7 @@ obj-$(CONFIG_INTEL_WMI_THUNDERBOLT) += intel-wmi-thunderbolt.o > obj-$(CONFIG_MXM_WMI) += mxm-wmi.o > obj-$(CONFIG_PEAQ_WMI) += peaq-wmi.o > obj-$(CONFIG_XIAOMI_WMI) += xiaomi-wmi.o > +obj-$(CONFIG_GIGABYTE_WMI) += gigabyte-wmi.o > > # Acer > obj-$(CONFIG_ACERHDF) += acerhdf.o > diff --git a/drivers/platform/x86/gigabyte-wmi.c b/drivers/platform/x86/gigabyte-wmi.c > new file mode 100644 > index 000000000000..8618363e3ccf > --- /dev/null > +++ b/drivers/platform/x86/gigabyte-wmi.c > @@ -0,0 +1,138 @@ > +// SPDX-License-Identifier: GPL-2.0-or-later > +/* > + * Copyright (C) 2021 Thomas Weißschuh <thomas@weissschuh.net> > + */ > +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt > + > +#include <linux/acpi.h> > +#include <linux/hwmon.h> > +#include <linux/module.h> > +#include <linux/wmi.h> > + > +#define GIGABYTE_WMI_GUID "DEADBEEF-2001-0000-00A0-C90629100000" > + > +enum gigabyte_wmi_commandtype { > + GIGABYTE_WMI_BUILD_DATE_QUERY = 0x1, > + GIGABYTE_WMI_MAINBOARD_TYPE_QUERY = 0x2, > + GIGABYTE_WMI_FIRMWARE_VERSION_QUERY = 0x4, > + GIGABYTE_WMI_MAINBOARD_NAME_QUERY = 0x5, > + GIGABYTE_WMI_TEMPERATURE_QUERY = 0x125, > +}; > + > +struct gigabyte_wmi_args { > + u32 arg1; > +}; > + > +static int gigabyte_wmi_perform_query(enum gigabyte_wmi_commandtype command, > + struct gigabyte_wmi_args *args, struct acpi_buffer *out) > +{ > + const struct acpi_buffer in = { > + .length = sizeof(*args), > + .pointer = args, > + }; > + > + acpi_status ret = wmi_evaluate_method(GIGABYTE_WMI_GUID, 0x0, command, &in, out); > + if (ret == AE_OK) { > + return 0; > + } else { > + return -EIO; > + }; > +} > + > +static int gigabyte_wmi_query_integer(enum gigabyte_wmi_commandtype command, > + struct gigabyte_wmi_args *args, u64 *res) > +{ > + union acpi_object *obj; > + struct acpi_buffer result = { ACPI_ALLOCATE_BUFFER, NULL }; > + int ret; > + > + ret = gigabyte_wmi_perform_query(command, args, &result); > + if (ret) { > + goto out; > + } > + obj = result.pointer; > + if (obj && obj->type == ACPI_TYPE_INTEGER) { > + *res = obj->integer.value; > + ret = 0; > + } else { > + ret = -EIO; > + } > +out: > + kfree(result.pointer); > + return ret; > +} > + > +static int gigabyte_wmi_temperature(u8 sensor, long *res) > +{ > + struct gigabyte_wmi_args args = { > + .arg1 = sensor, > + }; > + u64 temp; > + acpi_status ret; > + > + ret = gigabyte_wmi_query_integer(GIGABYTE_WMI_TEMPERATURE_QUERY, &args, &temp); > + if (ret == 0) > + *res = (s8) temp * 1000; // value is a signed 8-bit integer > + return ret; > +} > + > +static int gigabyte_wmi_hwmon_read(struct device *dev, enum hwmon_sensor_types type, > + u32 attr, int channel, long *val) > +{ > + return gigabyte_wmi_temperature(channel, val); > +} > + > +static umode_t gigabyte_wmi_hwmon_is_visible(const void *data, enum hwmon_sensor_types type, > + u32 attr, int channel) > +{ > + return 0444; > +} > + > +static const struct hwmon_channel_info *gigabyte_wmi_hwmon_info[] = { > + HWMON_CHANNEL_INFO(temp, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT), > + NULL, > +}; > + > +static const struct hwmon_ops gigabyte_wmi_hwmon_ops = { > + .read = gigabyte_wmi_hwmon_read, > + .is_visible = gigabyte_wmi_hwmon_is_visible, > +}; > + > +static const struct hwmon_chip_info gigabyte_wmi_hwmon_chip_info = { > + .ops = &gigabyte_wmi_hwmon_ops, > + .info = gigabyte_wmi_hwmon_info, > +}; > + > +static int gigabyte_wmi_probe(struct wmi_device *wdev, const void *context) > +{ > + struct device *hwmon_dev = devm_hwmon_device_register_with_info(&wdev->dev, > + "gigabyte_wmi", NULL, > + &gigabyte_wmi_hwmon_chip_info, NULL); > + > + return PTR_ERR_OR_ZERO(hwmon_dev); > +} > + > +static const struct wmi_device_id gigabyte_wmi_id_table[] = { > + { GIGABYTE_WMI_GUID, NULL }, > + { }, > +}; > + > +static struct wmi_driver gigabyte_wmi_driver = { > + .driver = { > + .name = "gigabyte-wmi", > + }, > + .id_table = gigabyte_wmi_id_table, > + .probe = gigabyte_wmi_probe, > +}; > +module_wmi_driver(gigabyte_wmi_driver); > + > +MODULE_DEVICE_TABLE(wmi, gigabyte_wmi_id_table); > +MODULE_AUTHOR("Thomas Weißschuh <thomas@weissschuh.net>"); > +MODULE_DESCRIPTION("Gigabyte Temperature WMI Driver"); > +MODULE_LICENSE("GPL"); > > base-commit: 144c79ef33536b4ecb4951e07dbc1f2b7fa99d32 >
Hi 2021. április 5., hétfő 22:48 keltezéssel, Thomas Weißschuh írta: > Changes since v1: > * Incorporate feedback from Barnabás Pőcze > * Use a WMI driver instead of a platform driver > * Let the kernel manage the driver lifecycle > * Fix errno/ACPI error confusion > * Fix resource cleanup > * Document reason for integer casting > > Thank you Barnabás for your review, it is much appreciated. > > -- >8 -- > > Tested with a X570 I Aorus Pro Wifi. > The mainboard contains an ITE IT8688E chip for management. > This chips is also handled by drivers/hwmon/i87.c but as it is also used > by the firmware itself it needs an ACPI driver. I gather this means you're getting the ACPI Warning: SystemIO range ... conflicts with ... ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver warning? > > Unfortunately not all sensor registers are handled by the firmware and even > less are exposed via WMI. > > Signed-off-by: Thomas Weißschuh <linux@weissschuh.net> > --- > drivers/platform/x86/Kconfig | 11 +++ > drivers/platform/x86/Makefile | 1 + > drivers/platform/x86/gigabyte-wmi.c | 138 ++++++++++++++++++++++++++++ > 3 files changed, 150 insertions(+) > create mode 100644 drivers/platform/x86/gigabyte-wmi.c > > diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig > index ad4e630e73e2..96622a2106f7 100644 > --- a/drivers/platform/x86/Kconfig > +++ b/drivers/platform/x86/Kconfig > @@ -123,6 +123,17 @@ config XIAOMI_WMI > To compile this driver as a module, choose M here: the module will > be called xiaomi-wmi. > > +config GIGABYTE_WMI > + tristate "Gigabyte WMI temperature driver" > + depends on ACPI_WMI > + depends on HWMON > + help > + Say Y here if you want to support WMI-based temperature reporting on > + Gigabyte mainboards. > + > + To compile this driver as a module, choose M here: the module will > + be called gigabyte-wmi. > + > config ACERHDF > tristate "Acer Aspire One temperature and fan driver" > depends on ACPI && THERMAL > diff --git a/drivers/platform/x86/Makefile b/drivers/platform/x86/Makefile > index 60d554073749..1621ebfd04fd 100644 > --- a/drivers/platform/x86/Makefile > +++ b/drivers/platform/x86/Makefile > @@ -15,6 +15,7 @@ obj-$(CONFIG_INTEL_WMI_THUNDERBOLT) += intel-wmi-thunderbolt.o > obj-$(CONFIG_MXM_WMI) += mxm-wmi.o > obj-$(CONFIG_PEAQ_WMI) += peaq-wmi.o > obj-$(CONFIG_XIAOMI_WMI) += xiaomi-wmi.o > +obj-$(CONFIG_GIGABYTE_WMI) += gigabyte-wmi.o > > # Acer > obj-$(CONFIG_ACERHDF) += acerhdf.o > diff --git a/drivers/platform/x86/gigabyte-wmi.c b/drivers/platform/x86/gigabyte-wmi.c > new file mode 100644 > index 000000000000..8618363e3ccf > --- /dev/null > +++ b/drivers/platform/x86/gigabyte-wmi.c > @@ -0,0 +1,138 @@ > +// SPDX-License-Identifier: GPL-2.0-or-later > +/* > + * Copyright (C) 2021 Thomas Weißschuh <thomas@weissschuh.net> > + */ > +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt > + > +#include <linux/acpi.h> > +#include <linux/hwmon.h> > +#include <linux/module.h> > +#include <linux/wmi.h> > + > +#define GIGABYTE_WMI_GUID "DEADBEEF-2001-0000-00A0-C90629100000" > + > +enum gigabyte_wmi_commandtype { > + GIGABYTE_WMI_BUILD_DATE_QUERY = 0x1, > + GIGABYTE_WMI_MAINBOARD_TYPE_QUERY = 0x2, > + GIGABYTE_WMI_FIRMWARE_VERSION_QUERY = 0x4, > + GIGABYTE_WMI_MAINBOARD_NAME_QUERY = 0x5, > + GIGABYTE_WMI_TEMPERATURE_QUERY = 0x125, > +}; > + > +struct gigabyte_wmi_args { > + u32 arg1; > +}; > + > +static int gigabyte_wmi_perform_query(enum gigabyte_wmi_commandtype command, > + struct gigabyte_wmi_args *args, struct acpi_buffer *out) > +{ > + const struct acpi_buffer in = { > + .length = sizeof(*args), > + .pointer = args, > + }; > + > + acpi_status ret = wmi_evaluate_method(GIGABYTE_WMI_GUID, 0x0, command, &in, out); Ideally, you'd use the WMI device that was passed to the probe method to do the query using `wmidev_evaluate_method()`. You can pass the WMI device pointer to `devm_hwmon_device_register_with_info()` in the `drvdata` argument, then in the ->read() callback you can retrieve it: static int gigabyte_wmi_hwmon_read(struct device *dev, ...) { struct wmi_device *wdev = dev_get_drvdata(dev); and then you can pass that to the other functions. > + if (ret == AE_OK) { > + return 0; > + } else { > + return -EIO; > + }; The `;` is not needed. And please use `ACPI_FAILURE()` or `ACPI_SUCCESS()` to check the returned value. For example: acpi_status ret = ...; if (ACPI_FAILURE(ret)) return -EIO; return 0; > +} > + > +static int gigabyte_wmi_query_integer(enum gigabyte_wmi_commandtype command, > + struct gigabyte_wmi_args *args, u64 *res) > +{ > + union acpi_object *obj; > + struct acpi_buffer result = { ACPI_ALLOCATE_BUFFER, NULL }; > + int ret; > + > + ret = gigabyte_wmi_perform_query(command, args, &result); > + if (ret) { > + goto out; I believe if this branch is taken, no buffer is allocated (due to the failure), so you can just `return ret;` here and do away with the goto completely - if I'm not mistaken. > + } > + obj = result.pointer; > + if (obj && obj->type == ACPI_TYPE_INTEGER) { > + *res = obj->integer.value; > + ret = 0; > + } else { > + ret = -EIO; > + } > +out: > + kfree(result.pointer); > + return ret; > +} > + > +static int gigabyte_wmi_temperature(u8 sensor, long *res) > +{ > + struct gigabyte_wmi_args args = { > + .arg1 = sensor, > + }; > + u64 temp; > + acpi_status ret; > + > + ret = gigabyte_wmi_query_integer(GIGABYTE_WMI_TEMPERATURE_QUERY, &args, &temp); > + if (ret == 0) > + *res = (s8) temp * 1000; // value is a signed 8-bit integer > + return ret; > +} > + > +static int gigabyte_wmi_hwmon_read(struct device *dev, enum hwmon_sensor_types type, > + u32 attr, int channel, long *val) > +{ > + return gigabyte_wmi_temperature(channel, val); > +} > + > +static umode_t gigabyte_wmi_hwmon_is_visible(const void *data, enum hwmon_sensor_types type, > + u32 attr, int channel) > +{ > + return 0444; > +} > + > +static const struct hwmon_channel_info *gigabyte_wmi_hwmon_info[] = { > + HWMON_CHANNEL_INFO(temp, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT), > + NULL, ^ Minor thing: usually commas after sentinel values are omitted. > +}; > + > +static const struct hwmon_ops gigabyte_wmi_hwmon_ops = { > + .read = gigabyte_wmi_hwmon_read, > + .is_visible = gigabyte_wmi_hwmon_is_visible, > +}; > + > +static const struct hwmon_chip_info gigabyte_wmi_hwmon_chip_info = { > + .ops = &gigabyte_wmi_hwmon_ops, > + .info = gigabyte_wmi_hwmon_info, > +}; > + > +static int gigabyte_wmi_probe(struct wmi_device *wdev, const void *context) > +{ > + struct device *hwmon_dev = devm_hwmon_device_register_with_info(&wdev->dev, > + "gigabyte_wmi", NULL, > + &gigabyte_wmi_hwmon_chip_info, NULL); > + > + return PTR_ERR_OR_ZERO(hwmon_dev); > +} > + > +static const struct wmi_device_id gigabyte_wmi_id_table[] = { > + { GIGABYTE_WMI_GUID, NULL }, > + { }, ^ Same here. > +}; > + > +static struct wmi_driver gigabyte_wmi_driver = { > + .driver = { > + .name = "gigabyte-wmi", > + }, > + .id_table = gigabyte_wmi_id_table, > + .probe = gigabyte_wmi_probe, > +}; > +module_wmi_driver(gigabyte_wmi_driver); > + > +MODULE_DEVICE_TABLE(wmi, gigabyte_wmi_id_table); > +MODULE_AUTHOR("Thomas Weißschuh <thomas@weissschuh.net>"); > +MODULE_DESCRIPTION("Gigabyte Temperature WMI Driver"); It's a very minor thing, but could you please synchronize this description with the Kconfig? > +MODULE_LICENSE("GPL"); > > base-commit: 144c79ef33536b4ecb4951e07dbc1f2b7fa99d32 > -- > 2.31.1 Regards, Barnabás Pőcze
Hi Hans, On Mi, 2021-04-07T17:54+0200, Hans de Goede wrote: > Thank you for your new driver and thank you for the quick respin > addressing Barnabás' request to make it a WMI driver. > > The code looks good, so merging this should be a no-brainer, > yet I'm not sure if I should merge this driver as-is, let me > explain. thanks for the encouraging words. > The problem is that I assume that this is based on reverse-engineering? Yes, it is completely reverse-engineered. Essentially I stumbled upon Matthews comment at https://bugzilla.kernel.org/show_bug.cgi?id=204807#c37 and went from there. > We have some mixes experiences with reverse-engineered WMI drivers, > sometimes a feature is not supported yet the wmi_evaluate_method() > call happily succeeds. One example of this causing trouble is: > > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=1797d588af15174d4a4e7159dac8c800538e4f8c There actually are reports of recent, similar mainboards with recent firmware and similar sensor chips that do not support the temperature query. (https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/3 and https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/2) Unfortunately for unknown WMI queries the firmware does not return any value. This ends up as an ACPI integer with value 0 on the driver side. (Which I could not yet find documentation for if that is expected) In the current version of the driver EIO is returned for 0 values which get translated to N/A by lm-sensors. > At a minimum I think your driver should check in its > probe function that > > gigabyte_wmi_query_integer(GIGABYTE_WMI_TEMPERATURE_QUERY, ...) > > actually succeeds on the machine the driver is running on chances > are that Gigabyte has been using the DEADBEEF-2001-0000-00A0-C90629100000 > GUID for ages, where as the 0x125 value for GIGABYTE_WMI_TEMPERATURE_QUERY > suggests that this is a pretty new API. Would it be enough to probe all six sensors and check if all return 0? > It would be good if you can see if you can find some DSDT-s for older > gigabyte motherboards attached to various distro's bug-tracking systems > or forum posts and see how those respond to an unknown gigabyte_wmi_commandtype. Will do. > Another open question to make sure this driver is suitable > as a generic driver (and does not misbehave) is how to figure out > how many temperature sensors there actually are. So far I could not find out how to query this from the firmware. The IT8688 chip can report the state of each sensor but that is not exposed by the firmware. But even the state information from the IT8688 is not accurate as is. One of the sensors that is reported as being active (directly via it87) on my machine always reports -55°C (yes, negative). > Perhaps the WMI interface returns an error when you query an out-of-range > temperature channel? Also "0" as mentioned above. > One option here might be to add a DMI matching table and only load on > systems on that table for now. That table could then perhaps also provide > labels for each of the temperature channels, which is something which > would be nice to have regardless of my worries about how well this driver > will work on motherboards on which it has not been tested. I am collecting reports for working motherboards at https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/1 . > You could combine this DMI matching table with a "force" module option to > continue with probing on boards which are not on the table to allow users > to test and report their results to you. > > And hopefully after a while, when we're confident that the code works > well on most gigabyte boards we can drop the DMI table, or at least > only use it for the channel labels. That sounds good. > Please don't take this the wrong way; I think it is great that you are > working on this. And the quick turnaround of the v2 of this drivers makes > me pretty certain that we can figure something out and get this merged. Thank you for the feedback! > Have you tried contacting Gigabyte about this? I don't think the WMI > interface is something which they need to keep secret for competitive > reasons, so maybe they can help? Note if they want you to sign a NDA > of sorts to view docs, then make sure that it contains some language > about them allowing you to release an opensource driver for their > hardware based on the "protected" information. I have not contacted them yet, will do. As mentioned in the initial patch submission there would be different ways to access this information firmware: * Directly call the underlying ACPI methods (these are present in all so far observed firmwares, even if not exposed via WMI). * Directly access the ACPI IndexField representing the it87 chip. * Directly access the it87 registers while holding the relevant locks via ACPI. I assume all of those mechanisms have no place in a proper kernel driver but would like to get your opinion on it. Thomas
Hi, On 4/7/21 9:43 PM, Thomas Weißschuh wrote: > Hi Hans, > > On Mi, 2021-04-07T17:54+0200, Hans de Goede wrote: >> Thank you for your new driver and thank you for the quick respin >> addressing Barnabás' request to make it a WMI driver. >> >> The code looks good, so merging this should be a no-brainer, >> yet I'm not sure if I should merge this driver as-is, let me >> explain. > > thanks for the encouraging words. > >> The problem is that I assume that this is based on reverse-engineering? > > Yes, it is completely reverse-engineered. > Essentially I stumbled upon Matthews comment at > https://bugzilla.kernel.org/show_bug.cgi?id=204807#c37 and went from there. > >> We have some mixes experiences with reverse-engineered WMI drivers, >> sometimes a feature is not supported yet the wmi_evaluate_method() >> call happily succeeds. One example of this causing trouble is: >> >> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=1797d588af15174d4a4e7159dac8c800538e4f8c > > There actually are reports of recent, similar mainboards with recent firmware and > similar sensor chips that do not support the temperature query. > (https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/3 and > https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/2) > > Unfortunately for unknown WMI queries the firmware does not return any value. > This ends up as an ACPI integer with value 0 on the driver side. > (Which I could not yet find documentation for if that is expected) > In the current version of the driver EIO is returned for 0 values which > get translated to N/A by lm-sensors. > >> At a minimum I think your driver should check in its >> probe function that >> >> gigabyte_wmi_query_integer(GIGABYTE_WMI_TEMPERATURE_QUERY, ...) >> >> actually succeeds on the machine the driver is running on chances >> are that Gigabyte has been using the DEADBEEF-2001-0000-00A0-C90629100000 >> GUID for ages, where as the 0x125 value for GIGABYTE_WMI_TEMPERATURE_QUERY >> suggests that this is a pretty new API. > > Would it be enough to probe all six sensors and check if all return 0? I think that failing the probe with -ENODEV, with a dev_info() explaining why when all six sensors return 0 would be good yes, that should also fix those 2 issues on https://github.com/t-8ch/linux-gigabyte-wmi-driver/. >> It would be good if you can see if you can find some DSDT-s for older >> gigabyte motherboards attached to various distro's bug-tracking systems >> or forum posts and see how those respond to an unknown gigabyte_wmi_commandtype. > > Will do. Since you alreayd have bugreports of boards where this does not work, please don't spend too much time on this. I guess those older DSDT-s will also just return an integer with value 0. >> Another open question to make sure this driver is suitable >> as a generic driver (and does not misbehave) is how to figure out >> how many temperature sensors there actually are. > > So far I could not find out how to query this from the firmware. > The IT8688 chip can report the state of each sensor but that is not exposed by > the firmware. > But even the state information from the IT8688 is not accurate as is. > One of the sensors that is reported as being active (directly via it87) on my > machine always reports -55°C (yes, negative). Ok. >> Perhaps the WMI interface returns an error when you query an out-of-range >> temperature channel? > > Also "0" as mentioned above. Hmm, so maybe this can be used to limit the amount of reported temperature sensors, IOW if sensors 5 and 6 report 0, only register 4 sensors ? > >> One option here might be to add a DMI matching table and only load on >> systems on that table for now. That table could then perhaps also provide >> labels for each of the temperature channels, which is something which >> would be nice to have regardless of my worries about how well this driver >> will work on motherboards on which it has not been tested. > > I am collecting reports for working motherboards at > https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/1 . Good, you should probably ask reporters there to provide the output of: grep . /sys/class/dmi/id/* 2> /dev/null Ran as a normal user (so that the serial-numbers will be skipped) so that you will have DMI strings to match on if you decide to go that route. > >> You could combine this DMI matching table with a "force" module option to >> continue with probing on boards which are not on the table to allow users >> to test and report their results to you. >> >> And hopefully after a while, when we're confident that the code works >> well on most gigabyte boards we can drop the DMI table, or at least >> only use it for the channel labels. > > That sounds good. > >> Please don't take this the wrong way; I think it is great that you are >> working on this. And the quick turnaround of the v2 of this drivers makes >> me pretty certain that we can figure something out and get this merged. > > Thank you for the feedback! > >> Have you tried contacting Gigabyte about this? I don't think the WMI >> interface is something which they need to keep secret for competitive >> reasons, so maybe they can help? Note if they want you to sign a NDA >> of sorts to view docs, then make sure that it contains some language >> about them allowing you to release an opensource driver for their >> hardware based on the "protected" information. > > I have not contacted them yet, will do. > > As mentioned in the initial patch submission there would be different ways to > access this information firmware: > > * Directly call the underlying ACPI methods (these are present in all so far > observed firmwares, even if not exposed via WMI). > * Directly access the ACPI IndexField representing the it87 chip. > * Directly access the it87 registers while holding the relevant locks via ACPI. > > I assume all of those mechanisms have no place in a proper kernel driver but > would like to get your opinion on it. Actually the "Directly access the it87 registers" option is potentially interesting since it will allow using the it87 driver which gives a lot more features. I actually wrote a rough outline of how something like this could work here: https://bugzilla.kernel.org/show_bug.cgi?id=204807#c47 Note I'm not sure if that is the right approach, but it definitely is an option. It seems that this one might also solve the X470-AORUS-ULTRA-GAMING case (https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/2) Hopefully the direct-register ACPI/WMI access methods will also allow reading the super-io vendor and product ids so that we can be reasonably sure that we are not loading the wrong driver on a board. OTOH the WMI-temp method approach may also work on boards where the sensors (or some of the sensors) are not supported. I'm afraid there is no obviously correct answer here. If you feel like it experimenting with the "Directly access the it87 registers" option would certainly be interesting IMHO. It might be good to get hwmon subsystems maintainer's opinion on this before sinking a lot of time into this though (added to the Cc). Jean, Guenter, Thomas has been working on a WMI driver to expose various motherboard temperatures on a gigabyte board where the IO-addresses for the it87 chip are reserved by ACPI. We are discussing how best to deal with this, there are some ACPI methods to directly access the super-IO registers (with locking to protect against other ACPI accesses). This reminded me of an idea I had a while ago to solve a similar issue with an other superIO chip, abstract the superIO register access-es using some reg_ops struct and allow an ACPI/WMI driver to provide alternative reg_ops: https://bugzilla.kernel.org/show_bug.cgi?id=204807#c47 Do you think this is a good idea (or a bad one)? And would something like that be acceptable to you ? Regards, Hans
Hi, On Mi, 2021-04-07T18:27+0000, Barnabás Pőcze wrote: > 2021. április 5., hétfő 22:48 keltezéssel, Thomas Weißschuh írta: > > Tested with a X570 I Aorus Pro Wifi. > > The mainboard contains an ITE IT8688E chip for management. > > This chips is also handled by drivers/hwmon/i87.c but as it is also used > > by the firmware itself it needs an ACPI driver. > > I gather this means you're getting the > > ACPI Warning: SystemIO range ... conflicts with ... > ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver > > warning? Exactly. > > +struct gigabyte_wmi_args { > > + u32 arg1; > > +}; > > + > > +static int gigabyte_wmi_perform_query(enum gigabyte_wmi_commandtype command, > > + struct gigabyte_wmi_args *args, struct acpi_buffer *out) > > +{ > > + const struct acpi_buffer in = { > > + .length = sizeof(*args), > > + .pointer = args, > > + }; > > + > > + acpi_status ret = wmi_evaluate_method(GIGABYTE_WMI_GUID, 0x0, command, &in, out); > > Ideally, you'd use the WMI device that was passed to the probe method to do the query > using `wmidev_evaluate_method()`. You can pass the WMI device pointer > to `devm_hwmon_device_register_with_info()` in the `drvdata` argument, > then in the ->read() callback you can retrieve it: > > static int gigabyte_wmi_hwmon_read(struct device *dev, ...) > { > struct wmi_device *wdev = dev_get_drvdata(dev); > > and then you can pass that to the other functions. Done. > > + if (ret == AE_OK) { > > + return 0; > > + } else { > > + return -EIO; > > + }; > > The `;` is not needed. And please use `ACPI_FAILURE()` or `ACPI_SUCCESS()` > to check the returned value. For example: > > acpi_status ret = ...; > if (ACPI_FAILURE(ret)) > return -EIO; > > return 0; Done. > > +} > > + > > +static int gigabyte_wmi_query_integer(enum gigabyte_wmi_commandtype command, > > + struct gigabyte_wmi_args *args, u64 *res) > > +{ > > + union acpi_object *obj; > > + struct acpi_buffer result = { ACPI_ALLOCATE_BUFFER, NULL }; > > + int ret; > > + > > + ret = gigabyte_wmi_perform_query(command, args, &result); > > + if (ret) { > > + goto out; > > I believe if this branch is taken, no buffer is allocated (due to the failure), > so you can just `return ret;` here and do away with the goto completely - if I'm not mistaken. Done. > > +static const struct hwmon_channel_info *gigabyte_wmi_hwmon_info[] = { > > + HWMON_CHANNEL_INFO(temp, > > + HWMON_T_INPUT, > > + HWMON_T_INPUT, > > + HWMON_T_INPUT, > > + HWMON_T_INPUT, > > + HWMON_T_INPUT, > > + HWMON_T_INPUT), > > + NULL, > ^ > Minor thing: usually commas after sentinel values are omitted. Done. > > +static const struct wmi_device_id gigabyte_wmi_id_table[] = { > > + { GIGABYTE_WMI_GUID, NULL }, > > + { }, > ^ > Same here. Done. > > > +}; > > + > > +static struct wmi_driver gigabyte_wmi_driver = { > > + .driver = { > > + .name = "gigabyte-wmi", > > + }, > > + .id_table = gigabyte_wmi_id_table, > > + .probe = gigabyte_wmi_probe, > > +}; > > +module_wmi_driver(gigabyte_wmi_driver); > > + > > +MODULE_DEVICE_TABLE(wmi, gigabyte_wmi_id_table); > > +MODULE_AUTHOR("Thomas Weißschuh <thomas@weissschuh.net>"); > > +MODULE_DESCRIPTION("Gigabyte Temperature WMI Driver"); > > It's a very minor thing, but could you please > synchronize this description with the Kconfig? Of course. Thanks again for the review! Thomas
Hi Hans, On Do, 2021-04-08T11:36+0200, Hans de Goede wrote: > On 4/7/21 9:43 PM, Thomas Weißschuh wrote: > > Hi Hans, > > > > On Mi, 2021-04-07T17:54+0200, Hans de Goede wrote: > >> Thank you for your new driver and thank you for the quick respin > >> addressing Barnabás' request to make it a WMI driver. > >> > >> The code looks good, so merging this should be a no-brainer, > >> yet I'm not sure if I should merge this driver as-is, let me > >> explain. > > > > thanks for the encouraging words. > > > >> The problem is that I assume that this is based on reverse-engineering? > > > > Yes, it is completely reverse-engineered. > > Essentially I stumbled upon Matthews comment at > > https://bugzilla.kernel.org/show_bug.cgi?id=204807#c37 and went from there. > > > >> We have some mixes experiences with reverse-engineered WMI drivers, > >> sometimes a feature is not supported yet the wmi_evaluate_method() > >> call happily succeeds. One example of this causing trouble is: > >> > >> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=1797d588af15174d4a4e7159dac8c800538e4f8c > > > > There actually are reports of recent, similar mainboards with recent firmware and > > similar sensor chips that do not support the temperature query. > > (https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/3 and > > https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/2) > > > > Unfortunately for unknown WMI queries the firmware does not return any value. > > This ends up as an ACPI integer with value 0 on the driver side. > > (Which I could not yet find documentation for if that is expected) > > In the current version of the driver EIO is returned for 0 values which > > get translated to N/A by lm-sensors. > > > >> At a minimum I think your driver should check in its > >> probe function that > >> > >> gigabyte_wmi_query_integer(GIGABYTE_WMI_TEMPERATURE_QUERY, ...) > >> > >> actually succeeds on the machine the driver is running on chances > >> are that Gigabyte has been using the DEADBEEF-2001-0000-00A0-C90629100000 > >> GUID for ages, where as the 0x125 value for GIGABYTE_WMI_TEMPERATURE_QUERY > >> suggests that this is a pretty new API. > > > > Would it be enough to probe all six sensors and check if all return 0? > > I think that failing the probe with -ENODEV, with a dev_info() explaining why when > all six sensors return 0 would be good yes, that should also fix those 2 > issues on https://github.com/t-8ch/linux-gigabyte-wmi-driver/. I added such a validation step. > >> It would be good if you can see if you can find some DSDT-s for older > >> gigabyte motherboards attached to various distro's bug-tracking systems > >> or forum posts and see how those respond to an unknown gigabyte_wmi_commandtype. > > > > Will do. > > Since you alreayd have bugreports of boards where this does not work, > please don't spend too much time on this. I guess those older DSDT-s will > also just return an integer with value 0. Ok. > >> Another open question to make sure this driver is suitable > >> as a generic driver (and does not misbehave) is how to figure out > >> how many temperature sensors there actually are. > > > > So far I could not find out how to query this from the firmware. > > The IT8688 chip can report the state of each sensor but that is not exposed by > > the firmware. > > But even the state information from the IT8688 is not accurate as is. > > One of the sensors that is reported as being active (directly via it87) on my > > machine always reports -55°C (yes, negative). > > Ok. > > >> Perhaps the WMI interface returns an error when you query an out-of-range > >> temperature channel? > > > > Also "0" as mentioned above. > > Hmm, so maybe this can be used to limit the amount of reported temperature > sensors, IOW if sensors 5 and 6 report 0, only register 4 sensors ? So far the 0-returning sensors have not been at the end of the list but in the middle. Is it worth building logic to properly probe a bitmask of useful sensors? > >> One option here might be to add a DMI matching table and only load on > >> systems on that table for now. That table could then perhaps also provide > >> labels for each of the temperature channels, which is something which > >> would be nice to have regardless of my worries about how well this driver > >> will work on motherboards on which it has not been tested. > > > > I am collecting reports for working motherboards at > > https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/1 . > > Good, you should probably ask reporters there to provide the output of: > > grep . /sys/class/dmi/id/* 2> /dev/null I added a DMI-based whitelist and asked users to submit their DMI information. > Ran as a normal user (so that the serial-numbers will be skipped) so that > you will have DMI strings to match on if you decide to go that route. The serials seem not to be too critical on these boards: /sys/class/dmi/id/board_serial:Default string /sys/class/dmi/id/chassis_serial:Default string /sys/class/dmi/id/product_serial:Default string > > As mentioned in the initial patch submission there would be different ways to > > access this information firmware: > > > > * Directly call the underlying ACPI methods (these are present in all so far > > observed firmwares, even if not exposed via WMI). > > * Directly access the ACPI IndexField representing the it87 chip. > > * Directly access the it87 registers while holding the relevant locks via ACPI. > > > > I assume all of those mechanisms have no place in a proper kernel driver but > > would like to get your opinion on it. > > Actually the "Directly access the it87 registers" option is potentially interesting > since it will allow using the it87 driver which gives a lot more features. > > I actually wrote a rough outline of how something like this could work here: > > https://bugzilla.kernel.org/show_bug.cgi?id=204807#c47 I must have overread this one, but yes that's also what I had in mind. > Note I'm not sure if that is the right approach, but it definitely is an > option. It seems that this one might also solve the X470-AORUS-ULTRA-GAMING > case (https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/2) > > Hopefully the direct-register ACPI/WMI access methods will also allow > reading the super-io vendor and product ids so that we can be reasonably > sure that we are not loading the wrong driver on a board. > > OTOH the WMI-temp method approach may also work on boards where the sensors > (or some of the sensors) are not supported. > > I'm afraid there is no obviously correct answer here. If you feel like it > experimenting with the "Directly access the it87 registers" option would certainly > be interesting IMHO. > > It might be good to get hwmon subsystems maintainer's opinion on this > before sinking a lot of time into this though (added to the Cc). Sounds good.
On 4/8/21 2:36 AM, Hans de Goede wrote: > Hi, > > On 4/7/21 9:43 PM, Thomas Weißschuh wrote: >> Hi Hans, >> >> On Mi, 2021-04-07T17:54+0200, Hans de Goede wrote: >>> Thank you for your new driver and thank you for the quick respin >>> addressing Barnabás' request to make it a WMI driver. >>> >>> The code looks good, so merging this should be a no-brainer, >>> yet I'm not sure if I should merge this driver as-is, let me >>> explain. >> >> thanks for the encouraging words. >> >>> The problem is that I assume that this is based on reverse-engineering? >> >> Yes, it is completely reverse-engineered. >> Essentially I stumbled upon Matthews comment at >> https://bugzilla.kernel.org/show_bug.cgi?id=204807#c37 and went from there. >> >>> We have some mixes experiences with reverse-engineered WMI drivers, >>> sometimes a feature is not supported yet the wmi_evaluate_method() >>> call happily succeeds. One example of this causing trouble is: >>> >>> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=1797d588af15174d4a4e7159dac8c800538e4f8c >> >> There actually are reports of recent, similar mainboards with recent firmware and >> similar sensor chips that do not support the temperature query. >> (https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/3 and >> https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/2) >> >> Unfortunately for unknown WMI queries the firmware does not return any value. >> This ends up as an ACPI integer with value 0 on the driver side. >> (Which I could not yet find documentation for if that is expected) >> In the current version of the driver EIO is returned for 0 values which >> get translated to N/A by lm-sensors. >> >>> At a minimum I think your driver should check in its >>> probe function that >>> >>> gigabyte_wmi_query_integer(GIGABYTE_WMI_TEMPERATURE_QUERY, ...) >>> >>> actually succeeds on the machine the driver is running on chances >>> are that Gigabyte has been using the DEADBEEF-2001-0000-00A0-C90629100000 >>> GUID for ages, where as the 0x125 value for GIGABYTE_WMI_TEMPERATURE_QUERY >>> suggests that this is a pretty new API. >> >> Would it be enough to probe all six sensors and check if all return 0? > > I think that failing the probe with -ENODEV, with a dev_info() explaining why when > all six sensors return 0 would be good yes, that should also fix those 2 > issues on https://github.com/t-8ch/linux-gigabyte-wmi-driver/. > >>> It would be good if you can see if you can find some DSDT-s for older >>> gigabyte motherboards attached to various distro's bug-tracking systems >>> or forum posts and see how those respond to an unknown gigabyte_wmi_commandtype. >> >> Will do. > > Since you alreayd have bugreports of boards where this does not work, > please don't spend too much time on this. I guess those older DSDT-s will > also just return an integer with value 0. > >>> Another open question to make sure this driver is suitable >>> as a generic driver (and does not misbehave) is how to figure out >>> how many temperature sensors there actually are. >> >> So far I could not find out how to query this from the firmware. >> The IT8688 chip can report the state of each sensor but that is not exposed by >> the firmware. >> But even the state information from the IT8688 is not accurate as is. >> One of the sensors that is reported as being active (directly via it87) on my >> machine always reports -55°C (yes, negative). > > Ok. > >>> Perhaps the WMI interface returns an error when you query an out-of-range >>> temperature channel? >> >> Also "0" as mentioned above. > > Hmm, so maybe this can be used to limit the amount of reported temperature > sensors, IOW if sensors 5 and 6 report 0, only register 4 sensors ? > >> >>> One option here might be to add a DMI matching table and only load on >>> systems on that table for now. That table could then perhaps also provide >>> labels for each of the temperature channels, which is something which >>> would be nice to have regardless of my worries about how well this driver >>> will work on motherboards on which it has not been tested. >> >> I am collecting reports for working motherboards at >> https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/1 . > > Good, you should probably ask reporters there to provide the output of: > > grep . /sys/class/dmi/id/* 2> /dev/null > > Ran as a normal user (so that the serial-numbers will be skipped) so that > you will have DMI strings to match on if you decide to go that route. > >> >>> You could combine this DMI matching table with a "force" module option to >>> continue with probing on boards which are not on the table to allow users >>> to test and report their results to you. >>> >>> And hopefully after a while, when we're confident that the code works >>> well on most gigabyte boards we can drop the DMI table, or at least >>> only use it for the channel labels. >> >> That sounds good. >> >>> Please don't take this the wrong way; I think it is great that you are >>> working on this. And the quick turnaround of the v2 of this drivers makes >>> me pretty certain that we can figure something out and get this merged. >> >> Thank you for the feedback! >> >>> Have you tried contacting Gigabyte about this? I don't think the WMI >>> interface is something which they need to keep secret for competitive >>> reasons, so maybe they can help? Note if they want you to sign a NDA >>> of sorts to view docs, then make sure that it contains some language >>> about them allowing you to release an opensource driver for their >>> hardware based on the "protected" information. >> >> I have not contacted them yet, will do. >> >> As mentioned in the initial patch submission there would be different ways to >> access this information firmware: >> >> * Directly call the underlying ACPI methods (these are present in all so far >> observed firmwares, even if not exposed via WMI). >> * Directly access the ACPI IndexField representing the it87 chip. >> * Directly access the it87 registers while holding the relevant locks via ACPI. >> >> I assume all of those mechanisms have no place in a proper kernel driver but >> would like to get your opinion on it. > > Actually the "Directly access the it87 registers" option is potentially interesting > since it will allow using the it87 driver which gives a lot more features. > > I actually wrote a rough outline of how something like this could work here: > > https://bugzilla.kernel.org/show_bug.cgi?id=204807#c47 > > Note I'm not sure if that is the right approach, but it definitely is an > option. It seems that this one might also solve the X470-AORUS-ULTRA-GAMING > case (https://github.com/t-8ch/linux-gigabyte-wmi-driver/issues/2) > > Hopefully the direct-register ACPI/WMI access methods will also allow > reading the super-io vendor and product ids so that we can be reasonably > sure that we are not loading the wrong driver on a board. > > OTOH the WMI-temp method approach may also work on boards where the sensors > (or some of the sensors) are not supported. > > I'm afraid there is no obviously correct answer here. If you feel like it > experimenting with the "Directly access the it87 registers" option would certainly > be interesting IMHO. > > It might be good to get hwmon subsystems maintainer's opinion on this > before sinking a lot of time into this though (added to the Cc). > > Jean, Guenter, > > Thomas has been working on a WMI driver to expose various motherboard > temperatures on a gigabyte board where the IO-addresses for the it87 chip > are reserved by ACPI. We are discussing how best to deal with this, there > are some ACPI methods to directly access the super-IO registers (with locking > to protect against other ACPI accesses). This reminded me of an idea I had > a while ago to solve a similar issue with an other superIO chip, abstract > the superIO register access-es using some reg_ops struct and allow an ACPI/WMI > driver to provide alternative reg_ops: > https://bugzilla.kernel.org/show_bug.cgi?id=204807#c47 > > Do you think this is a good idea (or a bad one)? And would something like that > be acceptable to you ? > The upstream it87 driver is severely out of date. I had an out-of-tree driver with various improvements which I didn't upstream, first because no one was willing to review changes and then because it had deviated too much. I pulled it from public view because I got pounded for not upstreaming it, because people started demanding support (not asking, demanding) for it, and because Gigabyte stopped providing datasheets for the more recent ITE chips and it became effectively unmaintainable. Some ITE chips have issues which can cause system hangs if accessed directly. I put some work to remedy that into the non-upstream driver, but that was all just guesswork. Gigabyte knows about the problem (or so I was told from someone who has an NDA with them), but I didn't get them or ITE to even acknowledge it to me. I even had a support case open with Gigabyte for a while, but all I could get out of them is that they don't support Linux and what I would have to reproduce the problem with Windows for them to provide assistance (even though, again, they knew about it). As for using ACPI locks or WMI to ensure that ACPI leaves the chip alone while the driver accesses chips directly: That is an option, but it has (at least) two problems. First, ACPI access methods are not well documented or standardized. I had tried to find useful means to do that some time ago, but I gave up because each board (even from the same vendor) handles locking and accesses differently. We would end up with lots of board specific code. Coincidentally, that was for ASUS boards and the nct6775 driver. Second, access through ACPI is only one of the issues. Turns out there are two ITE chips on many of the Gigabyte boards nowadays, and the two chips talk to each other using I2C. My out-of-tree driver tried to remedy that by blocking those accesses while the driver used the chip, but, again, without Gigabyte / ITE support this was never a perfect solution, and there was always the risk that the board ended up hanging because that access was blocked for too long. Recent ITE chips solve that problem by providing memory mapped access to the chip registers, but that is only useful if one has a datasheet. Overall, I don't think it makes much sense trying to make significant changes to the it87 driver without pulling in all the changes I had made, and without finding a better fix for the cross-chip access problems. I for sure won't have time for that (and getting hwmon patches reviewed is still very much an issue). Having said that, I am of course open to adding WMI/ACPI drivers for the various boards. Good luck getting support from Gigabyte, though. Or from ASUS, for that matter. Guenter
On Mon, Apr 05, 2021 at 10:48:10PM +0200, Thomas Weißschuh wrote: > Changes since v1: > * Incorporate feedback from Barnabás Pőcze > * Use a WMI driver instead of a platform driver > * Let the kernel manage the driver lifecycle > * Fix errno/ACPI error confusion > * Fix resource cleanup > * Document reason for integer casting > > Thank you Barnabás for your review, it is much appreciated. > > -- >8 -- > > Tested with a X570 I Aorus Pro Wifi. > The mainboard contains an ITE IT8688E chip for management. > This chips is also handled by drivers/hwmon/i87.c but as it is also used > by the firmware itself it needs an ACPI driver. > > Unfortunately not all sensor registers are handled by the firmware and even > less are exposed via WMI. > > Signed-off-by: Thomas Weißschuh <linux@weissschuh.net> > --- > drivers/platform/x86/Kconfig | 11 +++ > drivers/platform/x86/Makefile | 1 + > drivers/platform/x86/gigabyte-wmi.c | 138 ++++++++++++++++++++++++++++ Originally drivers/platform was supposed to be used for platform specific code. Not that I have control over it, but I really dislike that more and more hwmon drivers end up there. At least hwmon is in good company - I see drivers for various other subsystems there as well. I just wonder if that is such a good idea. That entire directory is bypassing subsystem maintainer reviews. Guenter > 3 files changed, 150 insertions(+) > create mode 100644 drivers/platform/x86/gigabyte-wmi.c > > diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig > index ad4e630e73e2..96622a2106f7 100644 > --- a/drivers/platform/x86/Kconfig > +++ b/drivers/platform/x86/Kconfig > @@ -123,6 +123,17 @@ config XIAOMI_WMI > To compile this driver as a module, choose M here: the module will > be called xiaomi-wmi. > > +config GIGABYTE_WMI > + tristate "Gigabyte WMI temperature driver" > + depends on ACPI_WMI > + depends on HWMON > + help > + Say Y here if you want to support WMI-based temperature reporting on > + Gigabyte mainboards. > + > + To compile this driver as a module, choose M here: the module will > + be called gigabyte-wmi. > + > config ACERHDF > tristate "Acer Aspire One temperature and fan driver" > depends on ACPI && THERMAL > diff --git a/drivers/platform/x86/Makefile b/drivers/platform/x86/Makefile > index 60d554073749..1621ebfd04fd 100644 > --- a/drivers/platform/x86/Makefile > +++ b/drivers/platform/x86/Makefile > @@ -15,6 +15,7 @@ obj-$(CONFIG_INTEL_WMI_THUNDERBOLT) += intel-wmi-thunderbolt.o > obj-$(CONFIG_MXM_WMI) += mxm-wmi.o > obj-$(CONFIG_PEAQ_WMI) += peaq-wmi.o > obj-$(CONFIG_XIAOMI_WMI) += xiaomi-wmi.o > +obj-$(CONFIG_GIGABYTE_WMI) += gigabyte-wmi.o > > # Acer > obj-$(CONFIG_ACERHDF) += acerhdf.o > diff --git a/drivers/platform/x86/gigabyte-wmi.c b/drivers/platform/x86/gigabyte-wmi.c > new file mode 100644 > index 000000000000..8618363e3ccf > --- /dev/null > +++ b/drivers/platform/x86/gigabyte-wmi.c > @@ -0,0 +1,138 @@ > +// SPDX-License-Identifier: GPL-2.0-or-later > +/* > + * Copyright (C) 2021 Thomas Weißschuh <thomas@weissschuh.net> > + */ > +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt > + > +#include <linux/acpi.h> > +#include <linux/hwmon.h> > +#include <linux/module.h> > +#include <linux/wmi.h> > + > +#define GIGABYTE_WMI_GUID "DEADBEEF-2001-0000-00A0-C90629100000" > + > +enum gigabyte_wmi_commandtype { > + GIGABYTE_WMI_BUILD_DATE_QUERY = 0x1, > + GIGABYTE_WMI_MAINBOARD_TYPE_QUERY = 0x2, > + GIGABYTE_WMI_FIRMWARE_VERSION_QUERY = 0x4, > + GIGABYTE_WMI_MAINBOARD_NAME_QUERY = 0x5, > + GIGABYTE_WMI_TEMPERATURE_QUERY = 0x125, > +}; > + > +struct gigabyte_wmi_args { > + u32 arg1; > +}; > + > +static int gigabyte_wmi_perform_query(enum gigabyte_wmi_commandtype command, > + struct gigabyte_wmi_args *args, struct acpi_buffer *out) > +{ > + const struct acpi_buffer in = { > + .length = sizeof(*args), > + .pointer = args, > + }; > + > + acpi_status ret = wmi_evaluate_method(GIGABYTE_WMI_GUID, 0x0, command, &in, out); > + if (ret == AE_OK) { > + return 0; > + } else { > + return -EIO; > + }; > +} > + > +static int gigabyte_wmi_query_integer(enum gigabyte_wmi_commandtype command, > + struct gigabyte_wmi_args *args, u64 *res) > +{ > + union acpi_object *obj; > + struct acpi_buffer result = { ACPI_ALLOCATE_BUFFER, NULL }; > + int ret; > + > + ret = gigabyte_wmi_perform_query(command, args, &result); > + if (ret) { > + goto out; > + } > + obj = result.pointer; > + if (obj && obj->type == ACPI_TYPE_INTEGER) { > + *res = obj->integer.value; > + ret = 0; > + } else { > + ret = -EIO; > + } > +out: > + kfree(result.pointer); > + return ret; > +} > + > +static int gigabyte_wmi_temperature(u8 sensor, long *res) > +{ > + struct gigabyte_wmi_args args = { > + .arg1 = sensor, > + }; > + u64 temp; > + acpi_status ret; > + > + ret = gigabyte_wmi_query_integer(GIGABYTE_WMI_TEMPERATURE_QUERY, &args, &temp); > + if (ret == 0) > + *res = (s8) temp * 1000; // value is a signed 8-bit integer > + return ret; > +} > + > +static int gigabyte_wmi_hwmon_read(struct device *dev, enum hwmon_sensor_types type, > + u32 attr, int channel, long *val) > +{ > + return gigabyte_wmi_temperature(channel, val); > +} > + > +static umode_t gigabyte_wmi_hwmon_is_visible(const void *data, enum hwmon_sensor_types type, > + u32 attr, int channel) > +{ > + return 0444; > +} > + > +static const struct hwmon_channel_info *gigabyte_wmi_hwmon_info[] = { > + HWMON_CHANNEL_INFO(temp, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT, > + HWMON_T_INPUT), > + NULL, > +}; > + > +static const struct hwmon_ops gigabyte_wmi_hwmon_ops = { > + .read = gigabyte_wmi_hwmon_read, > + .is_visible = gigabyte_wmi_hwmon_is_visible, > +}; > + > +static const struct hwmon_chip_info gigabyte_wmi_hwmon_chip_info = { > + .ops = &gigabyte_wmi_hwmon_ops, > + .info = gigabyte_wmi_hwmon_info, > +}; > + > +static int gigabyte_wmi_probe(struct wmi_device *wdev, const void *context) > +{ > + struct device *hwmon_dev = devm_hwmon_device_register_with_info(&wdev->dev, > + "gigabyte_wmi", NULL, > + &gigabyte_wmi_hwmon_chip_info, NULL); > + > + return PTR_ERR_OR_ZERO(hwmon_dev); > +} > + > +static const struct wmi_device_id gigabyte_wmi_id_table[] = { > + { GIGABYTE_WMI_GUID, NULL }, > + { }, > +}; > + > +static struct wmi_driver gigabyte_wmi_driver = { > + .driver = { > + .name = "gigabyte-wmi", > + }, > + .id_table = gigabyte_wmi_id_table, > + .probe = gigabyte_wmi_probe, > +}; > +module_wmi_driver(gigabyte_wmi_driver); > + > +MODULE_DEVICE_TABLE(wmi, gigabyte_wmi_id_table); > +MODULE_AUTHOR("Thomas Weißschuh <thomas@weissschuh.net>"); > +MODULE_DESCRIPTION("Gigabyte Temperature WMI Driver"); > +MODULE_LICENSE("GPL"); > > base-commit: 144c79ef33536b4ecb4951e07dbc1f2b7fa99d32 > -- > 2.31.1 >
Hi Guenter, On 4/8/21 5:08 PM, Guenter Roeck wrote: > On Mon, Apr 05, 2021 at 10:48:10PM +0200, Thomas Weißschuh wrote: >> Changes since v1: >> * Incorporate feedback from Barnabás Pőcze >> * Use a WMI driver instead of a platform driver >> * Let the kernel manage the driver lifecycle >> * Fix errno/ACPI error confusion >> * Fix resource cleanup >> * Document reason for integer casting >> >> Thank you Barnabás for your review, it is much appreciated. >> >> -- >8 -- >> >> Tested with a X570 I Aorus Pro Wifi. >> The mainboard contains an ITE IT8688E chip for management. >> This chips is also handled by drivers/hwmon/i87.c but as it is also used >> by the firmware itself it needs an ACPI driver. >> >> Unfortunately not all sensor registers are handled by the firmware and even >> less are exposed via WMI. >> >> Signed-off-by: Thomas Weißschuh <linux@weissschuh.net> >> --- >> drivers/platform/x86/Kconfig | 11 +++ >> drivers/platform/x86/Makefile | 1 + >> drivers/platform/x86/gigabyte-wmi.c | 138 ++++++++++++++++++++++++++++ > > Originally drivers/platform was supposed to be used for platform specific > code. Not that I have control over it, but I really dislike that more and > more hwmon drivers end up there. > > At least hwmon is in good company - I see drivers for various other subsystems > there as well. I just wonder if that is such a good idea. That entire directory > is bypassing subsystem maintainer reviews. In case you are not aware I've recent(ish) taken over the drivers/platform/x86 maintainership from Andy Shevchenko. Yes it is a bit of an odd grab-bag it mostly deals with vendor specific ACPI / WMI interfaces which often more or less require using a single driver while offering multiple functionalities. These firmware interfaces do not really lend themselves to demultiplexing through something like MFD. These are mostly found on laptops where they deal with some or all of: - Hotkeys for brightness adjust / wlan-on/off toggle, touchpad on/off toggle, etc. (input subsystem stuff) - Mic. / Speaker mute LEDS (and other special LEDs) found on some laptops (LED subsystem stuff) - Enabling/disabling radios (rfkill stuff) - Controlling the DPTF performance profile (ACPI stuff) - Various sensors, some hwmon, some IIO - Backlight control (drm/kms subsys) - Enabling/disabling of LCD-builtin privacy filters (requires KMS/DRM subsys integration, pending) - Fan control (hwmon subsys) And often all of this in a single driver. This is all "stuff" for which there are no standard APIs shared between vendors, so it is a free for all and often it is all stuffed behind a single WMI or ACPI object, because that is how the vendor's drivers under Windows work. It certainly is not my intention to bypass review by other subsystem maintainers and when there are significant questions I do try to always get other subsys maintainers involved. See e.g. this thread, but also the "[PATCH 1/3] thinkpad_acpi: add support for force_discharge" thread where I asked for input from sre for the power-supply aspects of that. The WMI code was reworked a while back to make WMI be a bus and have individual WMI objects be devices on that bus. version 2 of this driver has been reworked to use this. Since this new driver is just a hwmon driver and as this is for a desktop I expect it will stay that way, I'm fine with moving this one over to drivers/hwmon if that has your preference. As for other cases then this driver, if you want to make sure you are at least Cc-ed on all hwmon related changes I'm fine with adding you as a reviewer to the pdx86 MAINTAINERS entry. Regards, Hans > > Guenter > >> 3 files changed, 150 insertions(+) >> create mode 100644 drivers/platform/x86/gigabyte-wmi.c >> >> diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig >> index ad4e630e73e2..96622a2106f7 100644 >> --- a/drivers/platform/x86/Kconfig >> +++ b/drivers/platform/x86/Kconfig >> @@ -123,6 +123,17 @@ config XIAOMI_WMI >> To compile this driver as a module, choose M here: the module will >> be called xiaomi-wmi. >> >> +config GIGABYTE_WMI >> + tristate "Gigabyte WMI temperature driver" >> + depends on ACPI_WMI >> + depends on HWMON >> + help >> + Say Y here if you want to support WMI-based temperature reporting on >> + Gigabyte mainboards. >> + >> + To compile this driver as a module, choose M here: the module will >> + be called gigabyte-wmi. >> + >> config ACERHDF >> tristate "Acer Aspire One temperature and fan driver" >> depends on ACPI && THERMAL >> diff --git a/drivers/platform/x86/Makefile b/drivers/platform/x86/Makefile >> index 60d554073749..1621ebfd04fd 100644 >> --- a/drivers/platform/x86/Makefile >> +++ b/drivers/platform/x86/Makefile >> @@ -15,6 +15,7 @@ obj-$(CONFIG_INTEL_WMI_THUNDERBOLT) += intel-wmi-thunderbolt.o >> obj-$(CONFIG_MXM_WMI) += mxm-wmi.o >> obj-$(CONFIG_PEAQ_WMI) += peaq-wmi.o >> obj-$(CONFIG_XIAOMI_WMI) += xiaomi-wmi.o >> +obj-$(CONFIG_GIGABYTE_WMI) += gigabyte-wmi.o >> >> # Acer >> obj-$(CONFIG_ACERHDF) += acerhdf.o >> diff --git a/drivers/platform/x86/gigabyte-wmi.c b/drivers/platform/x86/gigabyte-wmi.c >> new file mode 100644 >> index 000000000000..8618363e3ccf >> --- /dev/null >> +++ b/drivers/platform/x86/gigabyte-wmi.c >> @@ -0,0 +1,138 @@ >> +// SPDX-License-Identifier: GPL-2.0-or-later >> +/* >> + * Copyright (C) 2021 Thomas Weißschuh <thomas@weissschuh.net> >> + */ >> +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt >> + >> +#include <linux/acpi.h> >> +#include <linux/hwmon.h> >> +#include <linux/module.h> >> +#include <linux/wmi.h> >> + >> +#define GIGABYTE_WMI_GUID "DEADBEEF-2001-0000-00A0-C90629100000" >> + >> +enum gigabyte_wmi_commandtype { >> + GIGABYTE_WMI_BUILD_DATE_QUERY = 0x1, >> + GIGABYTE_WMI_MAINBOARD_TYPE_QUERY = 0x2, >> + GIGABYTE_WMI_FIRMWARE_VERSION_QUERY = 0x4, >> + GIGABYTE_WMI_MAINBOARD_NAME_QUERY = 0x5, >> + GIGABYTE_WMI_TEMPERATURE_QUERY = 0x125, >> +}; >> + >> +struct gigabyte_wmi_args { >> + u32 arg1; >> +}; >> + >> +static int gigabyte_wmi_perform_query(enum gigabyte_wmi_commandtype command, >> + struct gigabyte_wmi_args *args, struct acpi_buffer *out) >> +{ >> + const struct acpi_buffer in = { >> + .length = sizeof(*args), >> + .pointer = args, >> + }; >> + >> + acpi_status ret = wmi_evaluate_method(GIGABYTE_WMI_GUID, 0x0, command, &in, out); >> + if (ret == AE_OK) { >> + return 0; >> + } else { >> + return -EIO; >> + }; >> +} >> + >> +static int gigabyte_wmi_query_integer(enum gigabyte_wmi_commandtype command, >> + struct gigabyte_wmi_args *args, u64 *res) >> +{ >> + union acpi_object *obj; >> + struct acpi_buffer result = { ACPI_ALLOCATE_BUFFER, NULL }; >> + int ret; >> + >> + ret = gigabyte_wmi_perform_query(command, args, &result); >> + if (ret) { >> + goto out; >> + } >> + obj = result.pointer; >> + if (obj && obj->type == ACPI_TYPE_INTEGER) { >> + *res = obj->integer.value; >> + ret = 0; >> + } else { >> + ret = -EIO; >> + } >> +out: >> + kfree(result.pointer); >> + return ret; >> +} >> + >> +static int gigabyte_wmi_temperature(u8 sensor, long *res) >> +{ >> + struct gigabyte_wmi_args args = { >> + .arg1 = sensor, >> + }; >> + u64 temp; >> + acpi_status ret; >> + >> + ret = gigabyte_wmi_query_integer(GIGABYTE_WMI_TEMPERATURE_QUERY, &args, &temp); >> + if (ret == 0) >> + *res = (s8) temp * 1000; // value is a signed 8-bit integer >> + return ret; >> +} >> + >> +static int gigabyte_wmi_hwmon_read(struct device *dev, enum hwmon_sensor_types type, >> + u32 attr, int channel, long *val) >> +{ >> + return gigabyte_wmi_temperature(channel, val); >> +} >> + >> +static umode_t gigabyte_wmi_hwmon_is_visible(const void *data, enum hwmon_sensor_types type, >> + u32 attr, int channel) >> +{ >> + return 0444; >> +} >> + >> +static const struct hwmon_channel_info *gigabyte_wmi_hwmon_info[] = { >> + HWMON_CHANNEL_INFO(temp, >> + HWMON_T_INPUT, >> + HWMON_T_INPUT, >> + HWMON_T_INPUT, >> + HWMON_T_INPUT, >> + HWMON_T_INPUT, >> + HWMON_T_INPUT), >> + NULL, >> +}; >> + >> +static const struct hwmon_ops gigabyte_wmi_hwmon_ops = { >> + .read = gigabyte_wmi_hwmon_read, >> + .is_visible = gigabyte_wmi_hwmon_is_visible, >> +}; >> + >> +static const struct hwmon_chip_info gigabyte_wmi_hwmon_chip_info = { >> + .ops = &gigabyte_wmi_hwmon_ops, >> + .info = gigabyte_wmi_hwmon_info, >> +}; >> + >> +static int gigabyte_wmi_probe(struct wmi_device *wdev, const void *context) >> +{ >> + struct device *hwmon_dev = devm_hwmon_device_register_with_info(&wdev->dev, >> + "gigabyte_wmi", NULL, >> + &gigabyte_wmi_hwmon_chip_info, NULL); >> + >> + return PTR_ERR_OR_ZERO(hwmon_dev); >> +} >> + >> +static const struct wmi_device_id gigabyte_wmi_id_table[] = { >> + { GIGABYTE_WMI_GUID, NULL }, >> + { }, >> +}; >> + >> +static struct wmi_driver gigabyte_wmi_driver = { >> + .driver = { >> + .name = "gigabyte-wmi", >> + }, >> + .id_table = gigabyte_wmi_id_table, >> + .probe = gigabyte_wmi_probe, >> +}; >> +module_wmi_driver(gigabyte_wmi_driver); >> + >> +MODULE_DEVICE_TABLE(wmi, gigabyte_wmi_id_table); >> +MODULE_AUTHOR("Thomas Weißschuh <thomas@weissschuh.net>"); >> +MODULE_DESCRIPTION("Gigabyte Temperature WMI Driver"); >> +MODULE_LICENSE("GPL"); >> >> base-commit: 144c79ef33536b4ecb4951e07dbc1f2b7fa99d32 >> -- >> 2.31.1 >> >
On Do, 2021-04-08T08:00-0700, Guenter Roeck wrote: > On 4/8/21 2:36 AM, Hans de Goede wrote: > > On 4/7/21 9:43 PM, Thomas Weißschuh wrote: > >> On Mi, 2021-04-07T17:54+0200, Hans de Goede wrote: > > Jean, Guenter, > > > > Thomas has been working on a WMI driver to expose various motherboard > > temperatures on a gigabyte board where the IO-addresses for the it87 chip > > are reserved by ACPI. We are discussing how best to deal with this, there > > are some ACPI methods to directly access the super-IO registers (with locking > > to protect against other ACPI accesses). This reminded me of an idea I had > > a while ago to solve a similar issue with an other superIO chip, abstract > > the superIO register access-es using some reg_ops struct and allow an ACPI/WMI > > driver to provide alternative reg_ops: > > https://bugzilla.kernel.org/show_bug.cgi?id=204807#c47 > > > > Do you think this is a good idea (or a bad one)? And would something like that > > be acceptable to you ? > > > > The upstream it87 driver is severely out of date. I had an out-of-tree driver > with various improvements which I didn't upstream, first because no one was willing > to review changes and then because it had deviated too much. I pulled it from > public view because I got pounded for not upstreaming it, because people started > demanding support (not asking, demanding) for it, and because Gigabyte stopped > providing datasheets for the more recent ITE chips and it became effectively > unmaintainable. > > Some ITE chips have issues which can cause system hangs if accessed directly. > I put some work to remedy that into the non-upstream driver, but that was all > just guesswork. Gigabyte knows about the problem (or so I was told from someone > who has an NDA with them), but I didn't get them or ITE to even acknowledge it > to me. I even had a support case open with Gigabyte for a while, but all I could > get out of them is that they don't support Linux and what I would have to reproduce > the problem with Windows for them to provide assistance (even though, again, > they knew about it). > > As for using ACPI locks or WMI to ensure that ACPI leaves the chip alone while > the driver accesses chips directly: That is an option, but it has (at least) > two problems. > > First, ACPI access methods are not well documented or standardized. I had tried > to find useful means to do that some time ago, but I gave up because each board > (even from the same vendor) handles locking and accesses differently. We would > end up with lots of board specific code. Coincidentally, that was for ASUS boards > and the nct6775 driver. At least for all the Gigabyte ACPI tables I have looked at all access is done via two-byte "OperationRegion" over the Index/Data addresses, a "Field" with two entries for these and an "IndexField" that is actually used to perform the accesses. As the IndexField is synchronized via "Lock" it should take a lock on the OperationRegion itself. So I think we should be technically fine with validating these assumption and then also taking locks on the OperationRegion. If it is reasonable to do so is another question. > Second, access through ACPI is only one of the issues. Turns out there are two > ITE chips on many of the Gigabyte boards nowadays, and the two chips talk to each > other using I2C. My out-of-tree driver tried to remedy that by blocking those > accesses while the driver used the chip, but, again, without Gigabyte / ITE > support this was never a perfect solution, and there was always the risk that > the board ended up hanging because that access was blocked for too long. > Recent ITE chips solve that problem by providing memory mapped access to the > chip registers, but that is only useful if one has a datasheet. Are both of these chips available at the two well-known registers 0x2e and 0x4e? Would this too-long blocking also occur when only accessing single registers for read-only access? Any write access would probably have to be blocked anyways. > Overall, I don't think it makes much sense trying to make significant changes > to the it87 driver without pulling in all the changes I had made, and without > finding a better fix for the cross-chip access problems. I for sure won't have > time for that (and getting hwmon patches reviewed is still very much an issue). > > Having said that, I am of course open to adding WMI/ACPI drivers for the various > boards. Good luck getting support from Gigabyte, though. Or from ASUS, for that > matter. Thomas
On 4/8/21 9:07 AM, Hans de Goede wrote: > Hi Guenter, > > On 4/8/21 5:08 PM, Guenter Roeck wrote: >> On Mon, Apr 05, 2021 at 10:48:10PM +0200, Thomas Weißschuh wrote: >>> Changes since v1: >>> * Incorporate feedback from Barnabás Pőcze >>> * Use a WMI driver instead of a platform driver >>> * Let the kernel manage the driver lifecycle >>> * Fix errno/ACPI error confusion >>> * Fix resource cleanup >>> * Document reason for integer casting >>> >>> Thank you Barnabás for your review, it is much appreciated. >>> >>> -- >8 -- >>> >>> Tested with a X570 I Aorus Pro Wifi. >>> The mainboard contains an ITE IT8688E chip for management. >>> This chips is also handled by drivers/hwmon/i87.c but as it is also used >>> by the firmware itself it needs an ACPI driver. >>> >>> Unfortunately not all sensor registers are handled by the firmware and even >>> less are exposed via WMI. >>> >>> Signed-off-by: Thomas Weißschuh <linux@weissschuh.net> >>> --- >>> drivers/platform/x86/Kconfig | 11 +++ >>> drivers/platform/x86/Makefile | 1 + >>> drivers/platform/x86/gigabyte-wmi.c | 138 ++++++++++++++++++++++++++++ >> >> Originally drivers/platform was supposed to be used for platform specific >> code. Not that I have control over it, but I really dislike that more and >> more hwmon drivers end up there. >> >> At least hwmon is in good company - I see drivers for various other subsystems >> there as well. I just wonder if that is such a good idea. That entire directory >> is bypassing subsystem maintainer reviews. > > In case you are not aware I've recent(ish) taken over the drivers/platform/x86 > maintainership from Andy Shevchenko. > > Yes it is a bit of an odd grab-bag it mostly deals with vendor specific > ACPI / WMI interfaces which often more or less require using a single > driver while offering multiple functionalities. These firmware interfaces > do not really lend themselves to demultiplexing through something like > MFD. These are mostly found on laptops where they deal with some or all of: > > - Hotkeys for brightness adjust / wlan-on/off toggle, touchpad on/off toggle, etc. > (input subsystem stuff) > - Mic. / Speaker mute LEDS (and other special LEDs) found on some laptops > (LED subsystem stuff) > - Enabling/disabling radios > (rfkill stuff) > - Controlling the DPTF performance profile > (ACPI stuff) > - Various sensors, some hwmon, some IIO > - Backlight control (drm/kms subsys) > - Enabling/disabling of LCD-builtin privacy filters (requires KMS/DRM subsys integration, pending) > - Fan control (hwmon subsys) > > And often all of this in a single driver. This is all "stuff" for which > there are no standard APIs shared between vendors, so it is a free for > all and often it is all stuffed behind a single WMI or ACPI object, > because that is how the vendor's drivers under Windows work. > > It certainly is not my intention to bypass review by other subsystem > maintainers and when there are significant questions I do try to always > get other subsys maintainers involved. See e.g. this thread, but also the > "[PATCH 1/3] thinkpad_acpi: add support for force_discharge" thread > where I asked for input from sre for the power-supply aspects of that. > > The WMI code was reworked a while back to make WMI be a bus and have > individual WMI objects be devices on that bus. version 2 of this > driver has been reworked to use this. Since this new driver is just a hwmon > driver and as this is for a desktop I expect it will stay that way, > I'm fine with moving this one over to drivers/hwmon if that has your > preference. > I thought about it, but I don't think it makes much sense since all other WMI drivers are in drivers/platform. > As for other cases then this driver, if you want to make sure you are at > least Cc-ed on all hwmon related changes I'm fine with adding you as a > reviewer to the pdx86 MAINTAINERS entry. > I think I have a better idea: I'll add a regex pattern into the MAINTAINERS entry for hardware monitoring drivers. This way we should get copied whenever anyone adds a hardware monitoring driver into the tree. Thanks, Guenter
On 4/8/21 11:02 PM, Thomas Weißschuh wrote: > On Do, 2021-04-08T08:00-0700, Guenter Roeck wrote: >> On 4/8/21 2:36 AM, Hans de Goede wrote: >>> On 4/7/21 9:43 PM, Thomas Weißschuh wrote: >>>> On Mi, 2021-04-07T17:54+0200, Hans de Goede wrote: >>> Jean, Guenter, >>> >>> Thomas has been working on a WMI driver to expose various motherboard >>> temperatures on a gigabyte board where the IO-addresses for the it87 chip >>> are reserved by ACPI. We are discussing how best to deal with this, there >>> are some ACPI methods to directly access the super-IO registers (with locking >>> to protect against other ACPI accesses). This reminded me of an idea I had >>> a while ago to solve a similar issue with an other superIO chip, abstract >>> the superIO register access-es using some reg_ops struct and allow an ACPI/WMI >>> driver to provide alternative reg_ops: >>> https://bugzilla.kernel.org/show_bug.cgi?id=204807#c47 >>> >>> Do you think this is a good idea (or a bad one)? And would something like that >>> be acceptable to you ? >>> >> >> The upstream it87 driver is severely out of date. I had an out-of-tree driver >> with various improvements which I didn't upstream, first because no one was willing >> to review changes and then because it had deviated too much. I pulled it from >> public view because I got pounded for not upstreaming it, because people started >> demanding support (not asking, demanding) for it, and because Gigabyte stopped >> providing datasheets for the more recent ITE chips and it became effectively >> unmaintainable. >> >> Some ITE chips have issues which can cause system hangs if accessed directly. >> I put some work to remedy that into the non-upstream driver, but that was all >> just guesswork. Gigabyte knows about the problem (or so I was told from someone >> who has an NDA with them), but I didn't get them or ITE to even acknowledge it >> to me. I even had a support case open with Gigabyte for a while, but all I could >> get out of them is that they don't support Linux and what I would have to reproduce >> the problem with Windows for them to provide assistance (even though, again, >> they knew about it). >> >> As for using ACPI locks or WMI to ensure that ACPI leaves the chip alone while >> the driver accesses chips directly: That is an option, but it has (at least) >> two problems. >> >> First, ACPI access methods are not well documented or standardized. I had tried >> to find useful means to do that some time ago, but I gave up because each board >> (even from the same vendor) handles locking and accesses differently. We would >> end up with lots of board specific code. Coincidentally, that was for ASUS boards >> and the nct6775 driver. > > At least for all the Gigabyte ACPI tables I have looked at all access is done > via two-byte "OperationRegion" over the Index/Data addresses, a "Field" with > two entries for these and an "IndexField" that is actually used to perform the > accesses. > As the IndexField is synchronized via "Lock" it should take a lock on the > OperationRegion itself. > > So I think we should be technically fine with validating these assumption and > then also taking locks on the OperationRegion. > > If it is reasonable to do so is another question. > You'd still have to validate this for each individual board unless you get confirmation from Gigabyte that the mechanism is consistent on their boards. Then you'd have to handle other vendors using it87 chips, and those are just as close-lipped as Gigabyte. Ultimately it would require acpi match tables to match the various boards and access methods. I had experimented with this this a long time ago but gave up on it after concluding that it was unmaintainable. >> Second, access through ACPI is only one of the issues. Turns out there are two >> ITE chips on many of the Gigabyte boards nowadays, and the two chips talk to each >> other using I2C. My out-of-tree driver tried to remedy that by blocking those >> accesses while the driver used the chip, but, again, without Gigabyte / ITE >> support this was never a perfect solution, and there was always the risk that >> the board ended up hanging because that access was blocked for too long. >> Recent ITE chips solve that problem by providing memory mapped access to the >> chip registers, but that is only useful if one has a datasheet. > > Are both of these chips available at the two well-known registers 0x2e and > 0x4e? > The ones I know of are, yes. Oh, that reminds me, there is another bug. Here are my comments about that: /* * On various Gigabyte AM4 boards (AB350, AX370), the second Super-IO chip * (IT8792E) needs to be in configuration mode before accessing the first * due to a bug in IT8792E which otherwise results in LPC bus access errors. * This needs to be done before accessing the first Super-IO chip since * the second chip may have been accessed prior to loading this driver. * * The problem is also reported to affect IT8795E, which is used on X299 boards * and has the same chip ID as IT8792E (0x8733). It also appears to affect * systems with IT8790E, which is used on some Z97X-Gaming boards as well as * Z87X-OC. * DMI entries for those systems will be added as they become available and * as the problem is confirmed to affect those boards. */ > Would this too-long blocking also occur when only accessing single registers > for read-only access? I don't know. Remember, zero support from Gigabyte / ITE. Guenter
Hi, On 4/10/21 8:56 AM, Guenter Roeck wrote: > On 4/8/21 11:02 PM, Thomas Weißschuh wrote: >> On Do, 2021-04-08T08:00-0700, Guenter Roeck wrote: >>> On 4/8/21 2:36 AM, Hans de Goede wrote: >>>> On 4/7/21 9:43 PM, Thomas Weißschuh wrote: >>>>> On Mi, 2021-04-07T17:54+0200, Hans de Goede wrote: >>>> Jean, Guenter, >>>> >>>> Thomas has been working on a WMI driver to expose various motherboard >>>> temperatures on a gigabyte board where the IO-addresses for the it87 chip >>>> are reserved by ACPI. We are discussing how best to deal with this, there >>>> are some ACPI methods to directly access the super-IO registers (with locking >>>> to protect against other ACPI accesses). This reminded me of an idea I had >>>> a while ago to solve a similar issue with an other superIO chip, abstract >>>> the superIO register access-es using some reg_ops struct and allow an ACPI/WMI >>>> driver to provide alternative reg_ops: >>>> https://bugzilla.kernel.org/show_bug.cgi?id=204807#c47 >>>> >>>> Do you think this is a good idea (or a bad one)? And would something like that >>>> be acceptable to you ? >>>> >>> >>> The upstream it87 driver is severely out of date. I had an out-of-tree driver >>> with various improvements which I didn't upstream, first because no one was willing >>> to review changes and then because it had deviated too much. I pulled it from >>> public view because I got pounded for not upstreaming it, because people started >>> demanding support (not asking, demanding) for it, and because Gigabyte stopped >>> providing datasheets for the more recent ITE chips and it became effectively >>> unmaintainable. >>> >>> Some ITE chips have issues which can cause system hangs if accessed directly. >>> I put some work to remedy that into the non-upstream driver, but that was all >>> just guesswork. Gigabyte knows about the problem (or so I was told from someone >>> who has an NDA with them), but I didn't get them or ITE to even acknowledge it >>> to me. I even had a support case open with Gigabyte for a while, but all I could >>> get out of them is that they don't support Linux and what I would have to reproduce >>> the problem with Windows for them to provide assistance (even though, again, >>> they knew about it). >>> >>> As for using ACPI locks or WMI to ensure that ACPI leaves the chip alone while >>> the driver accesses chips directly: That is an option, but it has (at least) >>> two problems. >>> >>> First, ACPI access methods are not well documented or standardized. I had tried >>> to find useful means to do that some time ago, but I gave up because each board >>> (even from the same vendor) handles locking and accesses differently. We would >>> end up with lots of board specific code. Coincidentally, that was for ASUS boards >>> and the nct6775 driver. >> >> At least for all the Gigabyte ACPI tables I have looked at all access is done >> via two-byte "OperationRegion" over the Index/Data addresses, a "Field" with >> two entries for these and an "IndexField" that is actually used to perform the >> accesses. >> As the IndexField is synchronized via "Lock" it should take a lock on the >> OperationRegion itself. >> >> So I think we should be technically fine with validating these assumption and >> then also taking locks on the OperationRegion. >> >> If it is reasonable to do so is another question. >> > You'd still have to validate this for each individual board unless you get > confirmation from Gigabyte that the mechanism is consistent on their boards. > Then you'd have to handle other vendors using it87 chips, and those are > just as close-lipped as Gigabyte. Ultimately it would require acpi match > tables to match the various boards and access methods. I had experimented > with this this a long time ago but gave up on it after concluding that it was > unmaintainable. > >>> Second, access through ACPI is only one of the issues. Turns out there are two >>> ITE chips on many of the Gigabyte boards nowadays, and the two chips talk to each >>> other using I2C. My out-of-tree driver tried to remedy that by blocking those >>> accesses while the driver used the chip, but, again, without Gigabyte / ITE >>> support this was never a perfect solution, and there was always the risk that >>> the board ended up hanging because that access was blocked for too long. >>> Recent ITE chips solve that problem by providing memory mapped access to the >>> chip registers, but that is only useful if one has a datasheet. >> >> Are both of these chips available at the two well-known registers 0x2e and >> 0x4e? >> > > The ones I know of are, yes. > > Oh, that reminds me, there is another bug. Here are my comments about that: > > /* > * On various Gigabyte AM4 boards (AB350, AX370), the second Super-IO chip > * (IT8792E) needs to be in configuration mode before accessing the first > * due to a bug in IT8792E which otherwise results in LPC bus access errors. > * This needs to be done before accessing the first Super-IO chip since > * the second chip may have been accessed prior to loading this driver. > * > * The problem is also reported to affect IT8795E, which is used on X299 boards > * and has the same chip ID as IT8792E (0x8733). It also appears to affect > * systems with IT8790E, which is used on some Z97X-Gaming boards as well as > * Z87X-OC. > * DMI entries for those systems will be added as they become available and > * as the problem is confirmed to affect those boards. > */ > >> Would this too-long blocking also occur when only accessing single registers >> for read-only access? > > I don't know. Remember, zero support from Gigabyte / ITE. So this all sounds like just using the WMI temperature functions as v2 of this driver does, does sound best overall. Presumably those are also used by Gigabyte's own Windows tool. Although even there we have the issue of the interface possibly changing from board to board. So even there I think we should start with a DMI based allow-list approach for now; we can revisit this when we have a better picture of things. Regards, Hans
diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig index ad4e630e73e2..96622a2106f7 100644 --- a/drivers/platform/x86/Kconfig +++ b/drivers/platform/x86/Kconfig @@ -123,6 +123,17 @@ config XIAOMI_WMI To compile this driver as a module, choose M here: the module will be called xiaomi-wmi. +config GIGABYTE_WMI + tristate "Gigabyte WMI temperature driver" + depends on ACPI_WMI + depends on HWMON + help + Say Y here if you want to support WMI-based temperature reporting on + Gigabyte mainboards. + + To compile this driver as a module, choose M here: the module will + be called gigabyte-wmi. + config ACERHDF tristate "Acer Aspire One temperature and fan driver" depends on ACPI && THERMAL diff --git a/drivers/platform/x86/Makefile b/drivers/platform/x86/Makefile index 60d554073749..1621ebfd04fd 100644 --- a/drivers/platform/x86/Makefile +++ b/drivers/platform/x86/Makefile @@ -15,6 +15,7 @@ obj-$(CONFIG_INTEL_WMI_THUNDERBOLT) += intel-wmi-thunderbolt.o obj-$(CONFIG_MXM_WMI) += mxm-wmi.o obj-$(CONFIG_PEAQ_WMI) += peaq-wmi.o obj-$(CONFIG_XIAOMI_WMI) += xiaomi-wmi.o +obj-$(CONFIG_GIGABYTE_WMI) += gigabyte-wmi.o # Acer obj-$(CONFIG_ACERHDF) += acerhdf.o diff --git a/drivers/platform/x86/gigabyte-wmi.c b/drivers/platform/x86/gigabyte-wmi.c new file mode 100644 index 000000000000..8618363e3ccf --- /dev/null +++ b/drivers/platform/x86/gigabyte-wmi.c @@ -0,0 +1,138 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (C) 2021 Thomas Weißschuh <thomas@weissschuh.net> + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/acpi.h> +#include <linux/hwmon.h> +#include <linux/module.h> +#include <linux/wmi.h> + +#define GIGABYTE_WMI_GUID "DEADBEEF-2001-0000-00A0-C90629100000" + +enum gigabyte_wmi_commandtype { + GIGABYTE_WMI_BUILD_DATE_QUERY = 0x1, + GIGABYTE_WMI_MAINBOARD_TYPE_QUERY = 0x2, + GIGABYTE_WMI_FIRMWARE_VERSION_QUERY = 0x4, + GIGABYTE_WMI_MAINBOARD_NAME_QUERY = 0x5, + GIGABYTE_WMI_TEMPERATURE_QUERY = 0x125, +}; + +struct gigabyte_wmi_args { + u32 arg1; +}; + +static int gigabyte_wmi_perform_query(enum gigabyte_wmi_commandtype command, + struct gigabyte_wmi_args *args, struct acpi_buffer *out) +{ + const struct acpi_buffer in = { + .length = sizeof(*args), + .pointer = args, + }; + + acpi_status ret = wmi_evaluate_method(GIGABYTE_WMI_GUID, 0x0, command, &in, out); + if (ret == AE_OK) { + return 0; + } else { + return -EIO; + }; +} + +static int gigabyte_wmi_query_integer(enum gigabyte_wmi_commandtype command, + struct gigabyte_wmi_args *args, u64 *res) +{ + union acpi_object *obj; + struct acpi_buffer result = { ACPI_ALLOCATE_BUFFER, NULL }; + int ret; + + ret = gigabyte_wmi_perform_query(command, args, &result); + if (ret) { + goto out; + } + obj = result.pointer; + if (obj && obj->type == ACPI_TYPE_INTEGER) { + *res = obj->integer.value; + ret = 0; + } else { + ret = -EIO; + } +out: + kfree(result.pointer); + return ret; +} + +static int gigabyte_wmi_temperature(u8 sensor, long *res) +{ + struct gigabyte_wmi_args args = { + .arg1 = sensor, + }; + u64 temp; + acpi_status ret; + + ret = gigabyte_wmi_query_integer(GIGABYTE_WMI_TEMPERATURE_QUERY, &args, &temp); + if (ret == 0) + *res = (s8) temp * 1000; // value is a signed 8-bit integer + return ret; +} + +static int gigabyte_wmi_hwmon_read(struct device *dev, enum hwmon_sensor_types type, + u32 attr, int channel, long *val) +{ + return gigabyte_wmi_temperature(channel, val); +} + +static umode_t gigabyte_wmi_hwmon_is_visible(const void *data, enum hwmon_sensor_types type, + u32 attr, int channel) +{ + return 0444; +} + +static const struct hwmon_channel_info *gigabyte_wmi_hwmon_info[] = { + HWMON_CHANNEL_INFO(temp, + HWMON_T_INPUT, + HWMON_T_INPUT, + HWMON_T_INPUT, + HWMON_T_INPUT, + HWMON_T_INPUT, + HWMON_T_INPUT), + NULL, +}; + +static const struct hwmon_ops gigabyte_wmi_hwmon_ops = { + .read = gigabyte_wmi_hwmon_read, + .is_visible = gigabyte_wmi_hwmon_is_visible, +}; + +static const struct hwmon_chip_info gigabyte_wmi_hwmon_chip_info = { + .ops = &gigabyte_wmi_hwmon_ops, + .info = gigabyte_wmi_hwmon_info, +}; + +static int gigabyte_wmi_probe(struct wmi_device *wdev, const void *context) +{ + struct device *hwmon_dev = devm_hwmon_device_register_with_info(&wdev->dev, + "gigabyte_wmi", NULL, + &gigabyte_wmi_hwmon_chip_info, NULL); + + return PTR_ERR_OR_ZERO(hwmon_dev); +} + +static const struct wmi_device_id gigabyte_wmi_id_table[] = { + { GIGABYTE_WMI_GUID, NULL }, + { }, +}; + +static struct wmi_driver gigabyte_wmi_driver = { + .driver = { + .name = "gigabyte-wmi", + }, + .id_table = gigabyte_wmi_id_table, + .probe = gigabyte_wmi_probe, +}; +module_wmi_driver(gigabyte_wmi_driver); + +MODULE_DEVICE_TABLE(wmi, gigabyte_wmi_id_table); +MODULE_AUTHOR("Thomas Weißschuh <thomas@weissschuh.net>"); +MODULE_DESCRIPTION("Gigabyte Temperature WMI Driver"); +MODULE_LICENSE("GPL");
Changes since v1: * Incorporate feedback from Barnabás Pőcze * Use a WMI driver instead of a platform driver * Let the kernel manage the driver lifecycle * Fix errno/ACPI error confusion * Fix resource cleanup * Document reason for integer casting Thank you Barnabás for your review, it is much appreciated. -- >8 -- Tested with a X570 I Aorus Pro Wifi. The mainboard contains an ITE IT8688E chip for management. This chips is also handled by drivers/hwmon/i87.c but as it is also used by the firmware itself it needs an ACPI driver. Unfortunately not all sensor registers are handled by the firmware and even less are exposed via WMI. Signed-off-by: Thomas Weißschuh <linux@weissschuh.net> --- drivers/platform/x86/Kconfig | 11 +++ drivers/platform/x86/Makefile | 1 + drivers/platform/x86/gigabyte-wmi.c | 138 ++++++++++++++++++++++++++++ 3 files changed, 150 insertions(+) create mode 100644 drivers/platform/x86/gigabyte-wmi.c base-commit: 144c79ef33536b4ecb4951e07dbc1f2b7fa99d32