kernel/linux.git/drivers/acpi/device_sysfs.c, branch v6.12.80

ACPI: sysfs: Use ACPI_FREE() for freeing an ACPI object

2025-11-13T20:34:03+00:00

[ Upstream commit 149139ddcb99583fdec8d1eaf7dada41e5896101 ] Since str_obj is allocated by ACPICA in acpi_evaluate_object_typed(), it should be free with ACPI_FREE() rather than with kfree(), so use the former instead of the latter for freeing it. Signed-off-by: Kaushlendra Kumar Link: https://patch.msgid.link/20250822061946.472594-1-kaushlendra.kumar@intel.com [ rjw: Subject and changelog rewrite ] Signed-off-by: Rafael J. Wysocki Signed-off-by: Sasha Levin

ACPI: sysfs: remove return value of acpi_device_setup_files()

2024-08-02T14:44:22+00:00

The function can not fail anymore, so drop its return value. Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20240709-acpi-sysfs-groups-v2-5-058ab0667fa8@weissschuh.net Signed-off-by: Rafael J. Wysocki

ACPI: sysfs: manage sysfs attributes through device core

2024-08-02T14:44:22+00:00

Now that the ACPI sysfs attributes are organized around an attribute_group, the device core can manage them. Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20240709-acpi-sysfs-groups-v2-4-058ab0667fa8@weissschuh.net [ rjw: Changelog edits ] Signed-off-by: Rafael J. Wysocki

ACPI: sysfs: manage attributes as attribute_group

2024-08-02T14:44:22+00:00

The current manual attribute management is inconsistent and brittle. Not all return values of device_create_file() are checked and the cleanup logic needs to be kept up to date manually. Moving all attributes into an attribute_group and using the is_visible() callback allows the management of all attributes as a single unit. Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20240709-acpi-sysfs-groups-v2-3-058ab0667fa8@weissschuh.net Signed-off-by: Rafael J. Wysocki

ACPI: sysfs: evaluate _STR on each sysfs access

2024-08-02T14:44:22+00:00

The handling of the _STR method is inconsistent with the other method evaluations. It is the only method which is cached. The cached value stored in 'struct acpi_device_pnp' has a different lifetime than the other struct members. Commit d1efe3c324ea ("ACPI: Add new sysfs interface to export device description") does not explain this difference. Evaluating the method every time also removes the necessity to manage the lifetime of the cached value, which would be a problem when managing the sysfs attributes through the device core. Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20240709-acpi-sysfs-groups-v2-2-058ab0667fa8@weissschuh.net Signed-off-by: Rafael J. Wysocki

ACPI: sysfs: validate return type of _STR method

2024-08-02T14:44:22+00:00

Only buffer objects are valid return values of _STR. If something else is returned description_show() will access invalid memory. Fixes: d1efe3c324ea ("ACPI: Add new sysfs interface to export device description") Cc: All applicable Signed-off-by: Thomas Weißschuh Link: https://patch.msgid.link/20240709-acpi-sysfs-groups-v2-1-058ab0667fa8@weissschuh.net Signed-off-by: Rafael J. Wysocki

Merge branches 'acpi-ec', 'acpi-sysfs', 'acpi-misc' and 'acpi-uid'

2023-10-26T12:59:52+00:00

Merge ACPI EC driver updates, ACPI sysfs interface updates, misc updates related to ACPI and changes related to ACPI _UID handling for 6.7-rc1: - Add EC GPE detection quirk for HP 250 G7 Notebook PC (Jonathan Denose). - Fix and clean up create_pnp_modalias() and create_of_modalias() (Christophe JAILLET). - Modify 2 pieces of code to use acpi_evaluate_dsm_typed() (Andy Shevchenko). - Define acpi_dev_uid_match() for matching _UID and use it in several places (Raag Jadav). - Use acpi_device_uid() for fetching _UID in 2 places (Raag Jadav). * acpi-ec: ACPI: EC: Add quirk for HP 250 G7 Notebook PC * acpi-sysfs: ACPI: sysfs: Clean up create_pnp_modalias() and create_of_modalias() ACPI: sysfs: Fix create_pnp_modalias() and create_of_modalias() * acpi-misc: ACPI: x86: s2idle: Switch to use acpi_evaluate_dsm_typed() ACPI: PCI: Switch to use acpi_evaluate_dsm_typed() * acpi-uid: perf: arm_cspmu: use acpi_dev_hid_uid_match() for matching _HID and _UID ACPI: x86: use acpi_dev_uid_match() for matching _UID ACPI: utils: use acpi_dev_uid_match() for matching _UID pinctrl: intel: use acpi_dev_uid_match() for matching _UID ACPI: utils: Introduce acpi_dev_uid_match() for matching _UID perf: qcom: use acpi_device_uid() for fetching _UID ACPI: sysfs: use acpi_device_uid() for fetching _UID

ACPI: sysfs: Clean up create_pnp_modalias() and create_of_modalias()

2023-10-24T19:31:26+00:00

'modalias' is only written with snprintf() and it is already guaranteed to be nul-terminated, so remove the unneeded (but harmless) writes of a trailing '\0' to it. Also snprintf() never returns negative values, so remove redundant (but harmless) checks for it. Signed-off-by: Christophe JAILLET [ rjw: Merge two patches into one, combine changelogs, add subject ] Signed-off-by: Rafael J. Wysocki

ACPI: sysfs: Fix create_pnp_modalias() and create_of_modalias()

2023-10-24T19:31:26+00:00

snprintf() does not return negative values on error. To know if the buffer was too small, the returned value needs to be compared with the length of the passed buffer. If it is greater or equal, the output has been truncated, so add checks for the truncation to create_pnp_modalias() and create_of_modalias(). Also make them return -ENOMEM in that case, as they already do that elsewhere. Moreover, the remaining size of the buffer used by snprintf() needs to be updated after the first write to avoid out-of-bounds access as already done correctly in create_pnp_modalias(), but not in create_of_modalias(), so change the latter accordingly. Fixes: 8765c5ba1949 ("ACPI / scan: Rework modalias creation when "compatible" is present") Signed-off-by: Christophe JAILLET [ rjw: Merge two patches into one, combine changelogs, add subject ] Signed-off-by: Rafael J. Wysocki

ACPI: sysfs: use acpi_device_uid() for fetching _UID

2023-10-20T17:17:41+00:00

Convert manual _UID references to use the standard ACPI helper. Signed-off-by: Raag Jadav Signed-off-by: Rafael J. Wysocki