[PATCH v2 0/4] platform/x86: wmi: Prepare for future changes

[PATCH v2 0/4] platform/x86: wmi: Prepare for future changes

[Armin Wolf]

After over a year of reverse engineering, i am finally ready to
introduce support for WMI-ACPI marshaling inside the WMI driver core.
Since the resulting patch series is quite large, i am planning to
submit the necessary patches as three separate patch series.

This is supposed to be the first of the three patch series. Its main
purpose is to prepare the WMI driver core for the upcoming changes.
The first patch fixes an issue inside the nls utf16 to utf8 conversion
code, while the next two patches fix some minor issues inside the WMI
driver core itself. The last patch finally moves the code of the WMI
driver core into a separate repository to allow for future additions
without cluttering the main directory.

Changes since v1:
- move WMI driver core to drivers/platoform/wmi to prepare for future
AArch64 support

Armin Wolf (4):
  fs/nls: Fix utf16 to utf8 conversion
  platform/x86: wmi: Use correct type when populating ACPI objects
  platform/x86: wmi: Remove extern keyword from prototypes
  platform/x86: wmi: Move WMI core code into a separate directory

 Documentation/driver-api/wmi.rst           |  2 +-
 MAINTAINERS                                |  2 +-
 drivers/platform/Kconfig                   |  2 ++
 drivers/platform/Makefile                  |  1 +
 drivers/platform/wmi/Kconfig               | 34 ++++++++++++++++++++++
 drivers/platform/wmi/Makefile              |  8 +++++
 drivers/platform/{x86/wmi.c => wmi/core.c} | 34 +++++++++++++---------
 drivers/platform/x86/Kconfig               | 30 -------------------
 drivers/platform/x86/Makefile              |  1 -
 fs/nls/nls_base.c                          | 16 +++++++---
 include/linux/wmi.h                        | 15 ++++------
 11 files changed, 85 insertions(+), 60 deletions(-)
 create mode 100644 drivers/platform/wmi/Kconfig
 create mode 100644 drivers/platform/wmi/Makefile
 rename drivers/platform/{x86/wmi.c => wmi/core.c} (98%)

-- 
2.39.5

[PATCH v2 1/4] fs/nls: Fix utf16 to utf8 conversion

[Armin Wolf]

Currently the function responsible for converting between utf16 and
utf8 strings will ignore any characters that cannot be converted. This
however also includes multi-byte characters that do not fit into the
provided string buffer.

This can cause problems if such a multi-byte character is followed by
a single-byte character. In such a case the multi-byte character might
be ignored when the provided string buffer is too small, but the
single-byte character might fit and is thus still copied into the
resulting string.

Fix this by stop filling the provided string buffer once a character
does not fit. In order to be able to do this extend utf32_to_utf8()
to return useful errno codes instead of -1.

Fixes: 74675a58507e ("NLS: update handling of Unicode")
Signed-off-by: Armin Wolf <W_Armin@gmx.de>
---
 fs/nls/nls_base.c | 16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)

diff --git a/fs/nls/nls_base.c b/fs/nls/nls_base.c
index 18d597e49a19..d434c4463a8f 100644
--- a/fs/nls/nls_base.c
+++ b/fs/nls/nls_base.c
@@ -94,7 +94,7 @@ int utf32_to_utf8(unicode_t u, u8 *s, int maxout)
 
 	l = u;
 	if (l > UNICODE_MAX || (l & SURROGATE_MASK) == SURROGATE_PAIR)
-		return -1;
+		return -EILSEQ;
 
 	nc = 0;
 	for (t = utf8_table; t->cmask && maxout; t++, maxout--) {
@@ -110,7 +110,7 @@ int utf32_to_utf8(unicode_t u, u8 *s, int maxout)
 			return nc;
 		}
 	}
-	return -1;
+	return -EOVERFLOW;
 }
 EXPORT_SYMBOL(utf32_to_utf8);
 
@@ -217,8 +217,16 @@ int utf16s_to_utf8s(const wchar_t *pwcs, int inlen, enum utf16_endian endian,
 				inlen--;
 			}
 			size = utf32_to_utf8(u, op, maxout);
-			if (size == -1) {
-				/* Ignore character and move on */
+			if (size < 0) {
+				if (size == -EILSEQ) {
+					/* Ignore character and move on */
+					continue;
+				}
+				/*
+				 * Stop filling the buffer with data once a character
+				 * does not fit anymore.
+				 */
+				break;
 			} else {
 				op += size;
 				maxout -= size;
-- 
2.39.5

[PATCH v2 2/4] platform/x86: wmi: Use correct type when populating ACPI objects

[Armin Wolf]

When evaluating a WMxx/WSxx ACPI control method, the data buffer
either needs to be passed as an ACPI buffer or as an ACPI string.
Use the correct type (buffer/string) when populating the associated
ACPI object.

Signed-off-by: Armin Wolf <W_Armin@gmx.de>
---
 drivers/platform/x86/wmi.c | 34 ++++++++++++++++++++--------------
 1 file changed, 20 insertions(+), 14 deletions(-)

diff --git a/drivers/platform/x86/wmi.c b/drivers/platform/x86/wmi.c
index 4e86a422f05f..6878c4fcb0b5 100644
--- a/drivers/platform/x86/wmi.c
+++ b/drivers/platform/x86/wmi.c
@@ -142,14 +142,6 @@ static inline void get_acpi_method_name(const struct wmi_block *wblock,
 	buffer[4] = '\0';
 }
 
-static inline acpi_object_type get_param_acpi_type(const struct wmi_block *wblock)
-{
-	if (wblock->gblock.flags & ACPI_WMI_STRING)
-		return ACPI_TYPE_STRING;
-	else
-		return ACPI_TYPE_BUFFER;
-}
-
 static int wmidev_match_guid(struct device *dev, const void *data)
 {
 	struct wmi_block *wblock = dev_to_wblock(dev);
@@ -351,9 +343,16 @@ acpi_status wmidev_evaluate_method(struct wmi_device *wdev, u8 instance, u32 met
 	params[0].integer.value = instance;
 	params[1].type = ACPI_TYPE_INTEGER;
 	params[1].integer.value = method_id;
-	params[2].type = get_param_acpi_type(wblock);
-	params[2].buffer.length = in->length;
-	params[2].buffer.pointer = in->pointer;
+
+	if (wblock->gblock.flags & ACPI_WMI_STRING) {
+		params[2].type = ACPI_TYPE_STRING;
+		params[2].string.length = in->length;
+		params[2].string.pointer = in->pointer;
+	} else {
+		params[2].type = ACPI_TYPE_BUFFER;
+		params[2].buffer.length = in->length;
+		params[2].buffer.pointer = in->pointer;
+	}
 
 	get_acpi_method_name(wblock, 'M', method);
 
@@ -519,9 +518,16 @@ acpi_status wmidev_block_set(struct wmi_device *wdev, u8 instance, const struct
 	input.pointer = params;
 	params[0].type = ACPI_TYPE_INTEGER;
 	params[0].integer.value = instance;
-	params[1].type = get_param_acpi_type(wblock);
-	params[1].buffer.length = in->length;
-	params[1].buffer.pointer = in->pointer;
+
+	if (wblock->gblock.flags & ACPI_WMI_STRING) {
+		params[1].type = ACPI_TYPE_STRING;
+		params[1].string.length = in->length;
+		params[1].string.pointer = in->pointer;
+	} else {
+		params[1].type = ACPI_TYPE_BUFFER;
+		params[1].buffer.length = in->length;
+		params[1].buffer.pointer = in->pointer;
+	}
 
 	get_acpi_method_name(wblock, 'S', method);
 
-- 
2.39.5

[PATCH v2 3/4] platform/x86: wmi: Remove extern keyword from prototypes

[Armin Wolf]

The external function definitions do not need the "extern" keyword.
Remove it to silence the associated checkpatch warnings.

Signed-off-by: Armin Wolf <W_Armin@gmx.de>
---
 include/linux/wmi.h | 15 ++++++---------
 1 file changed, 6 insertions(+), 9 deletions(-)

diff --git a/include/linux/wmi.h b/include/linux/wmi.h
index 10751c8e5e6a..665ea7dc8a92 100644
--- a/include/linux/wmi.h
+++ b/include/linux/wmi.h
@@ -36,13 +36,10 @@ struct wmi_device {
  */
 #define to_wmi_device(device)	container_of_const(device, struct wmi_device, dev)
 
-extern acpi_status wmidev_evaluate_method(struct wmi_device *wdev,
-					  u8 instance, u32 method_id,
-					  const struct acpi_buffer *in,
-					  struct acpi_buffer *out);
+acpi_status wmidev_evaluate_method(struct wmi_device *wdev, u8 instance, u32 method_id,
+				   const struct acpi_buffer *in, struct acpi_buffer *out);
 
-extern union acpi_object *wmidev_block_query(struct wmi_device *wdev,
-					     u8 instance);
+union acpi_object *wmidev_block_query(struct wmi_device *wdev, u8 instance);
 
 acpi_status wmidev_block_set(struct wmi_device *wdev, u8 instance, const struct acpi_buffer *in);
 
@@ -81,9 +78,9 @@ struct wmi_driver {
  */
 #define to_wmi_driver(drv)	container_of_const(drv, struct wmi_driver, driver)
 
-extern int __must_check __wmi_driver_register(struct wmi_driver *driver,
-					      struct module *owner);
-extern void wmi_driver_unregister(struct wmi_driver *driver);
+int __must_check __wmi_driver_register(struct wmi_driver *driver, struct module *owner);
+
+void wmi_driver_unregister(struct wmi_driver *driver);
 
 /**
  * wmi_driver_register() - Helper macro to register a WMI driver
-- 
2.39.5

[PATCH v2 4/4] platform/x86: wmi: Move WMI core code into a separate directory

[Armin Wolf]

Move the WMI core code into a separate directory to prepare for
future additions to the WMI driver. Also update the description
of the Kconfig entry to better fit with the other subsystem
Kconfig entries.

Signed-off-by: Armin Wolf <W_Armin@gmx.de>
---
 Documentation/driver-api/wmi.rst           |  2 +-
 MAINTAINERS                                |  2 +-
 drivers/platform/Kconfig                   |  2 ++
 drivers/platform/Makefile                  |  1 +
 drivers/platform/wmi/Kconfig               | 34 ++++++++++++++++++++++
 drivers/platform/wmi/Makefile              |  8 +++++
 drivers/platform/{x86/wmi.c => wmi/core.c} |  0
 drivers/platform/x86/Kconfig               | 30 -------------------
 drivers/platform/x86/Makefile              |  1 -
 9 files changed, 47 insertions(+), 33 deletions(-)
 create mode 100644 drivers/platform/wmi/Kconfig
 create mode 100644 drivers/platform/wmi/Makefile
 rename drivers/platform/{x86/wmi.c => wmi/core.c} (100%)

diff --git a/Documentation/driver-api/wmi.rst b/Documentation/driver-api/wmi.rst
index 4e8dbdb1fc67..db835b43c937 100644
--- a/Documentation/driver-api/wmi.rst
+++ b/Documentation/driver-api/wmi.rst
@@ -16,5 +16,5 @@ which will be bound to compatible WMI devices by the driver core.
 .. kernel-doc:: include/linux/wmi.h
    :internal:
 
-.. kernel-doc:: drivers/platform/x86/wmi.c
+.. kernel-doc:: drivers/platform/wmi/core.c
    :export:
diff --git a/MAINTAINERS b/MAINTAINERS
index 46126ce2f968..2dc2e6cd58df 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -402,7 +402,7 @@ S:	Maintained
 F:	Documentation/ABI/testing/sysfs-bus-wmi
 F:	Documentation/driver-api/wmi.rst
 F:	Documentation/wmi/
-F:	drivers/platform/x86/wmi.c
+F:	drivers/platform/wmi/
 F:	include/uapi/linux/wmi.h
 
 ACRN HYPERVISOR SERVICE MODULE
diff --git a/drivers/platform/Kconfig b/drivers/platform/Kconfig
index 960fd6a82450..6bb645aed3d1 100644
--- a/drivers/platform/Kconfig
+++ b/drivers/platform/Kconfig
@@ -18,3 +18,5 @@ source "drivers/platform/surface/Kconfig"
 source "drivers/platform/x86/Kconfig"
 
 source "drivers/platform/arm64/Kconfig"
+
+source "drivers/platform/wmi/Kconfig"
diff --git a/drivers/platform/Makefile b/drivers/platform/Makefile
index 19ac54648586..533f500dfcff 100644
--- a/drivers/platform/Makefile
+++ b/drivers/platform/Makefile
@@ -13,3 +13,4 @@ obj-$(CONFIG_CHROME_PLATFORMS)	+= chrome/
 obj-$(CONFIG_CZNIC_PLATFORMS)	+= cznic/
 obj-$(CONFIG_SURFACE_PLATFORMS)	+= surface/
 obj-$(CONFIG_ARM64_PLATFORM_DEVICES)	+= arm64/
+obj-$(CONFIG_ACPI_WMI)		+= wmi/
diff --git a/drivers/platform/wmi/Kconfig b/drivers/platform/wmi/Kconfig
new file mode 100644
index 000000000000..77fcbb18746b
--- /dev/null
+++ b/drivers/platform/wmi/Kconfig
@@ -0,0 +1,34 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+#
+# ACPI WMI Core
+#
+
+menuconfig ACPI_WMI
+	tristate "ACPI-WMI support"
+	depends on ACPI && X86
+	help
+	  This option enables support for the ACPI-WMI driver core.
+
+	  The ACPI-WMI interface is a proprietary extension of ACPI allowing
+	  the platform firmware to expose WMI (Windows Management Instrumentation)
+	  objects used for managing various aspects of the underlying system.
+	  Mapping between ACPI control methods and WMI objects happens through
+	  special mapper devices (PNP0C14) defined inside the ACPI tables.
+
+	  Enabling this option is necessary for building the vendor specific
+	  ACPI-WMI client drivers for Acer, Dell an HP machines (among others).
+
+	  It is safe to enable this option even for machines that do not contain
+	  any ACPI-WMI mapper devices at all.
+
+if ACPI_WMI
+
+config ACPI_WMI_LEGACY_DEVICE_NAMES
+	bool "Use legacy WMI device naming scheme"
+	help
+	  Say Y here to force the WMI driver core to use the old WMI device naming
+	  scheme when creating WMI devices. Doing so might be necessary for some
+	  userspace applications but will cause the registration of WMI devices with
+	  the same GUID to fail in some corner cases.
+
+endif # ACPI_WMI
diff --git a/drivers/platform/wmi/Makefile b/drivers/platform/wmi/Makefile
new file mode 100644
index 000000000000..98393d7391ec
--- /dev/null
+++ b/drivers/platform/wmi/Makefile
@@ -0,0 +1,8 @@
+# SPDX-License-Identifier: GPL-2.0-or-later
+#
+# Makefile for linux/drivers/platform/wmi
+# ACPI WMI core
+#
+
+wmi-y			:= core.o
+obj-$(CONFIG_ACPI_WMI)	+= wmi.o
diff --git a/drivers/platform/x86/wmi.c b/drivers/platform/wmi/core.c
similarity index 100%
rename from drivers/platform/x86/wmi.c
rename to drivers/platform/wmi/core.c
diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig
index 46e62feeda3c..8c13430366ae 100644
--- a/drivers/platform/x86/Kconfig
+++ b/drivers/platform/x86/Kconfig
@@ -16,36 +16,6 @@ menuconfig X86_PLATFORM_DEVICES
 
 if X86_PLATFORM_DEVICES
 
-config ACPI_WMI
-	tristate "WMI"
-	depends on ACPI
-	help
-	  This driver adds support for the ACPI-WMI (Windows Management
-	  Instrumentation) mapper device (PNP0C14) found on some systems.
-
-	  ACPI-WMI is a proprietary extension to ACPI to expose parts of the
-	  ACPI firmware to userspace - this is done through various vendor
-	  defined methods and data blocks in a PNP0C14 device, which are then
-	  made available for userspace to call.
-
-	  The implementation of this in Linux currently only exposes this to
-	  other kernel space drivers.
-
-	  This driver is a required dependency to build the firmware specific
-	  drivers needed on many machines, including Acer and HP laptops.
-
-	  It is safe to enable this driver even if your DSDT doesn't define
-	  any ACPI-WMI devices.
-
-config ACPI_WMI_LEGACY_DEVICE_NAMES
-	bool "Use legacy WMI device naming scheme"
-	depends on ACPI_WMI
-	help
-	  Say Y here to force the WMI driver core to use the old WMI device naming
-	  scheme when creating WMI devices. Doing so might be necessary for some
-	  userspace applications but will cause the registration of WMI devices with
-	  the same GUID to fail in some corner cases.
-
 config WMI_BMOF
 	tristate "WMI embedded Binary MOF driver"
 	depends on ACPI_WMI
diff --git a/drivers/platform/x86/Makefile b/drivers/platform/x86/Makefile
index c7db2a88c11a..978cb55672da 100644
--- a/drivers/platform/x86/Makefile
+++ b/drivers/platform/x86/Makefile
@@ -5,7 +5,6 @@
 #
 
 # Windows Management Interface
-obj-$(CONFIG_ACPI_WMI)		+= wmi.o
 obj-$(CONFIG_WMI_BMOF)		+= wmi-bmof.o
 
 # WMI drivers
-- 
2.39.5

Re: [PATCH v2 0/4] platform/x86: wmi: Prepare for future changes

[Ilpo Järvinen]

On Tue, 11 Nov 2025 14:11:21 +0100, Armin Wolf wrote:

> After over a year of reverse engineering, i am finally ready to
> introduce support for WMI-ACPI marshaling inside the WMI driver core.
> Since the resulting patch series is quite large, i am planning to
> submit the necessary patches as three separate patch series.
> 
> This is supposed to be the first of the three patch series. Its main
> purpose is to prepare the WMI driver core for the upcoming changes.
> The first patch fixes an issue inside the nls utf16 to utf8 conversion
> code, while the next two patches fix some minor issues inside the WMI
> driver core itself. The last patch finally moves the code of the WMI
> driver core into a separate repository to allow for future additions
> without cluttering the main directory.
> 
> [...]


Thank you for your contribution, it has been applied to my local
review-ilpo-next branch. Note it will show up in the public
platform-drivers-x86/review-ilpo-next branch only once I've pushed my
local branch there, which might take a while.

The list of commits applied:
[1/4] fs/nls: Fix utf16 to utf8 conversion
      commit: 25524b6190295577e4918c689644451365e6466d
[2/4] platform/x86: wmi: Use correct type when populating ACPI objects
      commit: c209195a2a4ad920ff481b56628ca942d62e01b1
[3/4] platform/x86: wmi: Remove extern keyword from prototypes
      commit: 32e3fee88a4ac183541b478f5bc94084ea76436c
[4/4] platform/x86: wmi: Move WMI core code into a separate directory
      commit: e2c1b56f3dfa014128e775e898774c0356e3ff05

--
 i.

Re: [PATCH v2 1/4] fs/nls: Fix utf16 to utf8 conversion

[Andy Shevchenko]

On Tue, Nov 11, 2025 at 02:11:22PM +0100, Armin Wolf wrote:
> Currently the function responsible for converting between utf16 and
> utf8 strings will ignore any characters that cannot be converted. This
> however also includes multi-byte characters that do not fit into the
> provided string buffer.
> 
> This can cause problems if such a multi-byte character is followed by
> a single-byte character. In such a case the multi-byte character might
> be ignored when the provided string buffer is too small, but the
> single-byte character might fit and is thus still copied into the
> resulting string.
> 
> Fix this by stop filling the provided string buffer once a character
> does not fit. In order to be able to do this extend utf32_to_utf8()
> to return useful errno codes instead of -1.

Can you also update utf8_to_utf32() to return meaningful error codes?
Without that done we have inconsistent APIs.

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v2 1/4] fs/nls: Fix utf16 to utf8 conversion

[Armin Wolf]

Am 26.11.25 um 16:18 schrieb Andy Shevchenko:

> On Tue, Nov 11, 2025 at 02:11:22PM +0100, Armin Wolf wrote:
>> Currently the function responsible for converting between utf16 and
>> utf8 strings will ignore any characters that cannot be converted. This
>> however also includes multi-byte characters that do not fit into the
>> provided string buffer.
>>
>> This can cause problems if such a multi-byte character is followed by
>> a single-byte character. In such a case the multi-byte character might
>> be ignored when the provided string buffer is too small, but the
>> single-byte character might fit and is thus still copied into the
>> resulting string.
>>
>> Fix this by stop filling the provided string buffer once a character
>> does not fit. In order to be able to do this extend utf32_to_utf8()
>> to return useful errno codes instead of -1.
> Can you also update utf8_to_utf32() to return meaningful error codes?
> Without that done we have inconsistent APIs.
>
Sure, i will send a separate patch for that.

Thanks,
Armin Wolf