* Re: [regression] 1a8953f4f774 ("HID: Add IGNORE quirk for SMARTLINKTECHNOLOGY") causes issue with ID 4c4a:4155 Jieli Technology USB Composite Device
From: Salvatore Bonaccorso @ 2025-09-13 8:55 UTC (permalink / raw)
To: Staffan Melin
Cc: zhangheng, Jiri Kosina, Benjamin Tissoires, linux-input,
linux-kernel, regressions, stable, 1114557
In-Reply-To: <aMUxHZF-7p7--1qS@eldamar.lan>
[-- Attachment #1: Type: text/plain, Size: 1010 bytes --]
Hi Staffan,
chiming in hopefully it is of help.
Now really with the patch ...
On Fri, Sep 12, 2025 at 09:57:04PM +0200, Staffan Melin wrote:
> Thank you,
>
> I tried to apply this patch to 6.12.39, the first problematic kernel, as
> well as 6.12.41, the first bad I tried, and on both I got an error message:
>
> Applying: HID: quirks: Add device descriptor for 4c4a:4155
> error: patch failed: drivers/hid/hid-quirks.c:1068
> error: drivers/hid/hid-quirks.c: patch does not apply
> Patch failed at 0001 HID: quirks: Add device descriptor for 4c4a:4155
>
> To which kernel version should I apply the patch?
As the deveopment goes from mainline then down to stable series, the
fix needs to be developed first for mainline. So the patch is targeted
there.
But please find attached an updated patch which hopefully should work
which resolved the context changes on top of 6.12.47.
But ideally you can provide a Tested-by on zhangheng's mainline patch
to get things rolling as needed.
Regards,
Salvatore
[-- Attachment #2: 0001-HID-quirks-Add-device-descriptor-for-4c4a-4155.patch --]
[-- Type: text/x-diff, Size: 2181 bytes --]
From 16b03e802a858ebf3eafcfdb22e3025e1a840afd Mon Sep 17 00:00:00 2001
From: Zhang Heng <zhangheng@kylinos.cn>
Date: Fri, 12 Sep 2025 20:38:18 +0800
Subject: [PATCH] HID: quirks: Add device descriptor for 4c4a:4155
Two USB devices use the same ID. To avoid affecting functionality,
device descriptors are added to distinguish between them
Signed-off-by: Zhang Heng <zhangheng@kylinos.cn>
---
drivers/hid/hid-quirks.c | 17 ++++++++++++++++-
1 file changed, 16 insertions(+), 1 deletion(-)
diff --git a/drivers/hid/hid-quirks.c b/drivers/hid/hid-quirks.c
index 64f9728018b8..cbb658c2c60d 100644
--- a/drivers/hid/hid-quirks.c
+++ b/drivers/hid/hid-quirks.c
@@ -898,6 +898,17 @@ static const struct hid_device_id hid_ignore_list[] = {
#endif
{ HID_USB_DEVICE(USB_VENDOR_ID_YEALINK, USB_DEVICE_ID_YEALINK_P1K_P4K_B2K) },
{ HID_USB_DEVICE(USB_VENDOR_ID_QUANTA, USB_DEVICE_ID_QUANTA_HP_5MP_CAMERA_5473) },
+ { }
+};
+/*
+ * hid_ignore_mic - Microphone devices do not require HID core processing
+ *
+ * Now there are two USB devices using the same ID, one is the microphone and the other
+ * is the touch screen. The touch screen requires hid core processing, but the
+ * microphone does not. The two have different bcdIDs, which will be used to
+ * distinguish them in the future
+ */
+static const struct hid_device_id hid_ignore_mic[] = {
{ HID_USB_DEVICE(USB_VENDOR_ID_SMARTLINKTECHNOLOGY, USB_DEVICE_ID_SMARTLINKTECHNOLOGY_4155) },
{ }
};
@@ -1061,6 +1072,9 @@ bool hid_ignore(struct hid_device *hdev)
hid_match_id(hdev, hid_mouse_ignore_list))
return true;
+ if(hid_match_id(hdev, hid_ignore_mic) && (hdev->version > 1.1))
+ return true;
+
return !!hid_match_id(hdev, hid_ignore_list);
}
EXPORT_SYMBOL_GPL(hid_ignore);
@@ -1259,7 +1273,8 @@ static unsigned long hid_gets_squirk(const struct hid_device *hdev)
const struct hid_device_id *bl_entry;
unsigned long quirks = hdev->initial_quirks;
- if (hid_match_id(hdev, hid_ignore_list))
+ if (hid_match_id(hdev, hid_ignore_list) ||
+ (hid_match_id(hdev, hid_ignore_mic) && (hdev->version > 1.1)))
quirks |= HID_QUIRK_IGNORE;
if (hid_match_id(hdev, hid_have_special_driver))
--
2.51.0
^ permalink raw reply related
* [PATCH v2 5/5] dt-bindings: touchscreen: remove touchscreen.txt
From: Dario Binacchi @ 2025-09-13 9:26 UTC (permalink / raw)
To: linux-kernel
Cc: linux-amarula, Dario Binacchi, Rob Herring, Conor Dooley,
Dmitry Torokhov, Krzysztof Kozlowski, devicetree, linux-input
In-Reply-To: <20250913092707.1005616-1-dario.binacchi@amarulasolutions.com>
With commit 1d6204e2f51f ("dt-bindings: touchscreen: Add touchscreen
schema") touchscreen.txt is no longer needed, and since no other file
refers to it, it can be safely removed.
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Acked-by: Rob Herring (Arm) <robh@kernel.org>
---
Changes in v2:
- Update the commit message
- Add Acked-by tag of Rob Herring
- Add patches:
- 1/5 dt-bindings: touchscreen: convert bu21013 bindings to json schema
- 2/5 dt-bindings: touchscreen: convert eeti bindings to json schema
- 3/5 dt-bindings: touchscreen: convert raspberrypi,firmware-ts bindings
to json schema
- 4/5 dt-bindings: touchscreen: convert zet6223 bindings to json schema
.../devicetree/bindings/input/touchscreen/touchscreen.txt | 1 -
1 file changed, 1 deletion(-)
delete mode 100644 Documentation/devicetree/bindings/input/touchscreen/touchscreen.txt
diff --git a/Documentation/devicetree/bindings/input/touchscreen/touchscreen.txt b/Documentation/devicetree/bindings/input/touchscreen/touchscreen.txt
deleted file mode 100644
index e1adb902d503..000000000000
--- a/Documentation/devicetree/bindings/input/touchscreen/touchscreen.txt
+++ /dev/null
@@ -1 +0,0 @@
-See touchscreen.yaml
--
2.43.0
base-commit: 22f20375f5b71f30c0d6896583b93b6e4bba7279
branch: drop-touchscreen.txt
^ permalink raw reply related
* [PATCH v2 4/5] dt-bindings: touchscreen: convert zet6223 bindings to json schema
From: Dario Binacchi @ 2025-09-13 9:26 UTC (permalink / raw)
To: linux-kernel
Cc: linux-amarula, Dario Binacchi, Conor Dooley, Dmitry Torokhov,
Krzysztof Kozlowski, Rob Herring, devicetree, linux-input
In-Reply-To: <20250913092707.1005616-1-dario.binacchi@amarulasolutions.com>
Convert Zeitec ZET6223 touchscreen controller device tree binding to
json-schema.
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
---
Changes in v2:
- Added in v2
.../bindings/input/touchscreen/zet6223.txt | 30 ---------
.../bindings/input/touchscreen/zet6223.yaml | 62 +++++++++++++++++++
2 files changed, 62 insertions(+), 30 deletions(-)
delete mode 100644 Documentation/devicetree/bindings/input/touchscreen/zet6223.txt
create mode 100644 Documentation/devicetree/bindings/input/touchscreen/zet6223.yaml
diff --git a/Documentation/devicetree/bindings/input/touchscreen/zet6223.txt b/Documentation/devicetree/bindings/input/touchscreen/zet6223.txt
deleted file mode 100644
index 27d55a506f18..000000000000
--- a/Documentation/devicetree/bindings/input/touchscreen/zet6223.txt
+++ /dev/null
@@ -1,30 +0,0 @@
-Zeitec ZET6223 I2C touchscreen controller
-
-Required properties:
-- compatible : "zeitec,zet6223"
-- reg : I2C slave address of the chip (0x76)
-- interrupts : interrupt specification for the zet6223 interrupt
-
-Optional properties:
-
-- vio-supply : Specification for VIO supply (1.8V or 3.3V,
- depending on system interface needs).
-- vcc-supply : Specification for 3.3V VCC supply.
-- touchscreen-size-x : See touchscreen.txt
-- touchscreen-size-y : See touchscreen.txt
-- touchscreen-inverted-x : See touchscreen.txt
-- touchscreen-inverted-y : See touchscreen.txt
-- touchscreen-swapped-x-y : See touchscreen.txt
-
-Example:
-
-i2c@00000000 {
-
- zet6223: touchscreen@76 {
- compatible = "zeitec,zet6223";
- reg = <0x76>;
- interrupt-parent = <&pio>;
- interrupts = <6 11 IRQ_TYPE_EDGE_FALLING>
- };
-
-};
diff --git a/Documentation/devicetree/bindings/input/touchscreen/zet6223.yaml b/Documentation/devicetree/bindings/input/touchscreen/zet6223.yaml
new file mode 100644
index 000000000000..a126f3f369ea
--- /dev/null
+++ b/Documentation/devicetree/bindings/input/touchscreen/zet6223.yaml
@@ -0,0 +1,62 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/input/touchscreen/zet6223.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Zeitec ZET6223 touchscreen controller
+
+description:
+ Zeitec ZET6223 I2C driven touchscreen controller.
+
+maintainers:
+ - Dario Binacchi <dario.binacchi@amarulasolutions.com>
+
+allOf:
+ - $ref: touchscreen.yaml#
+
+properties:
+ compatible:
+ enum:
+ - zeitec,zet6223
+
+ reg:
+ maxItems: 1
+
+ interrupts:
+ maxItems: 1
+
+ vio-supply:
+ description: Phandle to the 1.8V or 3.3V VIO supply.
+
+ vcc-supply:
+ description: Phandle to the 3.3V VCC supply.
+
+ touchscreen-inverted-x: true
+ touchscreen-inverted-y: true
+ touchscreen-size-x: true
+ touchscreen-size-y: true
+ touchscreen-swapped-x-y: true
+
+additionalProperties: false
+
+required:
+ - compatible
+ - reg
+ - interrupts
+
+examples:
+ - |
+ #include <dt-bindings/gpio/gpio.h>
+ #include <dt-bindings/interrupt-controller/irq.h>
+ i2c {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ touchscreen@76 {
+ compatible = "zeitec,zet6223";
+ reg = <0x76>;
+ interrupt-parent = <&pio>;
+ interrupts = <6 11 IRQ_TYPE_EDGE_FALLING>;
+ };
+ };
--
2.43.0
base-commit: 22f20375f5b71f30c0d6896583b93b6e4bba7279
branch: drop-touchscreen.txt
^ permalink raw reply related
* [PATCH v2 3/5] dt-bindings: touchscreen: convert raspberrypi,firmware-ts bindings to json schema
From: Dario Binacchi @ 2025-09-13 9:26 UTC (permalink / raw)
To: linux-kernel
Cc: linux-amarula, Dario Binacchi,
Broadcom internal kernel review list, Conor Dooley,
Dmitry Torokhov, Florian Fainelli, Krzysztof Kozlowski,
Rob Herring, devicetree, linux-arm-kernel, linux-input,
linux-rpi-kernel
In-Reply-To: <20250913092707.1005616-1-dario.binacchi@amarulasolutions.com>
Convert Raspberry Pi firmware 7" touchscreen controller device tree
binding to json-schema.
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
---
Changes in v2:
- Added in v2
.../touchscreen/raspberrypi,firmware-ts.txt | 26 -----------
.../touchscreen/raspberrypi,firmware-ts.yaml | 45 +++++++++++++++++++
2 files changed, 45 insertions(+), 26 deletions(-)
delete mode 100644 Documentation/devicetree/bindings/input/touchscreen/raspberrypi,firmware-ts.txt
create mode 100644 Documentation/devicetree/bindings/input/touchscreen/raspberrypi,firmware-ts.yaml
diff --git a/Documentation/devicetree/bindings/input/touchscreen/raspberrypi,firmware-ts.txt b/Documentation/devicetree/bindings/input/touchscreen/raspberrypi,firmware-ts.txt
deleted file mode 100644
index 2a1af240ccc3..000000000000
--- a/Documentation/devicetree/bindings/input/touchscreen/raspberrypi,firmware-ts.txt
+++ /dev/null
@@ -1,26 +0,0 @@
-Raspberry Pi firmware based 7" touchscreen
-=====================================
-
-Required properties:
- - compatible: "raspberrypi,firmware-ts"
-
-Optional properties:
- - firmware: Reference to RPi's firmware device node
- - touchscreen-size-x: See touchscreen.txt
- - touchscreen-size-y: See touchscreen.txt
- - touchscreen-inverted-x: See touchscreen.txt
- - touchscreen-inverted-y: See touchscreen.txt
- - touchscreen-swapped-x-y: See touchscreen.txt
-
-Example:
-
-firmware: firmware-rpi {
- compatible = "raspberrypi,bcm2835-firmware";
- mboxes = <&mailbox>;
-
- ts: touchscreen {
- compatible = "raspberrypi,firmware-ts";
- touchscreen-size-x = <800>;
- touchscreen-size-y = <480>;
- };
-};
diff --git a/Documentation/devicetree/bindings/input/touchscreen/raspberrypi,firmware-ts.yaml b/Documentation/devicetree/bindings/input/touchscreen/raspberrypi,firmware-ts.yaml
new file mode 100644
index 000000000000..8faa796fb7dd
--- /dev/null
+++ b/Documentation/devicetree/bindings/input/touchscreen/raspberrypi,firmware-ts.yaml
@@ -0,0 +1,45 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/input/touchscreen/raspberrypi,firmware-ts.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Raspberry Pi firmware based 7" touchscreen
+
+maintainers:
+ - Dario Binacchi <dario.binacchi@amarulasolutions.com>
+
+allOf:
+ - $ref: touchscreen.yaml#
+
+properties:
+ compatible:
+ enum:
+ - raspberrypi,firmware-ts
+
+ firmware:
+ description: Phandle to RPi's firmware device node.
+
+ touchscreen-size-x: true
+ touchscreen-size-y: true
+ touchscreen-inverted-x: true
+ touchscreen-inverted-y: true
+ touchscreen-swapped-x-y: true
+
+additionalProperties: false
+
+required:
+ - compatible
+
+examples:
+ - |
+ firmware: firmware-rpi {
+ compatible = "raspberrypi,bcm2835-firmware";
+ mboxes = <&mailbox>;
+
+ ts: touchscreen {
+ compatible = "raspberrypi,firmware-ts";
+ touchscreen-size-x = <800>;
+ touchscreen-size-y = <480>;
+ };
+ };
--
2.43.0
base-commit: 22f20375f5b71f30c0d6896583b93b6e4bba7279
branch: drop-touchscreen.txt
^ permalink raw reply related
* [PATCH v2 2/5] dt-bindings: touchscreen: convert eeti bindings to json schema
From: Dario Binacchi @ 2025-09-13 9:26 UTC (permalink / raw)
To: linux-kernel
Cc: linux-amarula, Dario Binacchi, Conor Dooley, Dmitry Torokhov,
Krzysztof Kozlowski, Rob Herring, devicetree, linux-input
In-Reply-To: <20250913092707.1005616-1-dario.binacchi@amarulasolutions.com>
Convert EETI touchscreen controller device tree binding to json-schema.
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
---
Changes in v2:
- Added in v2
.../bindings/input/touchscreen/eeti.txt | 30 ---------
.../bindings/input/touchscreen/eeti.yaml | 62 +++++++++++++++++++
2 files changed, 62 insertions(+), 30 deletions(-)
delete mode 100644 Documentation/devicetree/bindings/input/touchscreen/eeti.txt
create mode 100644 Documentation/devicetree/bindings/input/touchscreen/eeti.yaml
diff --git a/Documentation/devicetree/bindings/input/touchscreen/eeti.txt b/Documentation/devicetree/bindings/input/touchscreen/eeti.txt
deleted file mode 100644
index 32b3712c916e..000000000000
--- a/Documentation/devicetree/bindings/input/touchscreen/eeti.txt
+++ /dev/null
@@ -1,30 +0,0 @@
-Bindings for EETI touchscreen controller
-
-Required properties:
-- compatible: should be "eeti,exc3000-i2c"
-- reg: I2C address of the chip. Should be set to <0xa>
-- interrupts: interrupt to which the chip is connected
-
-Optional properties:
-- attn-gpios: A handle to a GPIO to check whether interrupt is still
- latched. This is necessary for platforms that lack
- support for level-triggered IRQs.
-
-The following optional properties described in touchscreen.txt are
-also supported:
-
-- touchscreen-inverted-x
-- touchscreen-inverted-y
-- touchscreen-swapped-x-y
-
-Example:
-
-i2c-master {
- touchscreen@a {
- compatible = "eeti,exc3000-i2c";
- reg = <0xa>;
- interrupt-parent = <&gpio>;
- interrupts = <123 IRQ_TYPE_EDGE_RISING>;
- attn-gpios = <&gpio 123 GPIO_ACTIVE_HIGH>;
- };
-};
diff --git a/Documentation/devicetree/bindings/input/touchscreen/eeti.yaml b/Documentation/devicetree/bindings/input/touchscreen/eeti.yaml
new file mode 100644
index 000000000000..97027914d6a6
--- /dev/null
+++ b/Documentation/devicetree/bindings/input/touchscreen/eeti.yaml
@@ -0,0 +1,62 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/input/touchscreen/eeti.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: EETI touchscreen controller
+
+description:
+ EETI I2C driven touchscreen controller.
+
+maintainers:
+ - Dario Binacchi <dario.binacchi@amarulasolutions.com>
+
+allOf:
+ - $ref: touchscreen.yaml#
+
+properties:
+ compatible:
+ enum:
+ - eeti,exc3000-i2c
+
+ reg:
+ maxItems: 1
+
+ interrupts:
+ maxItems: 1
+
+ attn-gpios:
+ maxItems: 1
+ description: Phandle to a GPIO to check whether interrupt is still
+ latched. This is necessary for platforms that lack
+ support for level-triggered IRQs.
+
+ touchscreen-inverted-x: true
+ touchscreen-inverted-y: true
+ touchscreen-swapped-x-y: true
+
+additionalProperties: false
+
+required:
+ - compatible
+ - reg
+ - interrupts
+
+examples:
+ - |
+ #include <dt-bindings/gpio/gpio.h>
+ #include <dt-bindings/interrupt-controller/irq.h>
+ i2c {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ touchscreen@a {
+ compatible = "eeti,exc3000-i2c";
+ reg = <0xa>;
+
+ interrupt-parent = <&gpio>;
+ interrupts = <123 IRQ_TYPE_EDGE_RISING>;
+ attn-gpios = <&gpio 123 GPIO_ACTIVE_HIGH>;
+ };
+ };
--
2.43.0
base-commit: 22f20375f5b71f30c0d6896583b93b6e4bba7279
branch: drop-touchscreen.txt
^ permalink raw reply related
* [PATCH v2 1/5] dt-bindings: touchscreen: convert bu21013 bindings to json schema
From: Dario Binacchi @ 2025-09-13 9:26 UTC (permalink / raw)
To: linux-kernel
Cc: linux-amarula, Dario Binacchi, Conor Dooley, Dmitry Torokhov,
Krzysztof Kozlowski, Rob Herring, devicetree, linux-input
Convert Rohm BU21013 I2C touchscreen controller device tree binding to
json-schema.
Additional changes:
- Replace <supply_name>-supply with avdd-supply to match example and
existing DTS.
- Add reset-gpios in the example because it is required.
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
---
Changes in v2:
- Added in v2
.../bindings/input/touchscreen/bu21013.txt | 43 ---------
.../bindings/input/touchscreen/bu21013.yaml | 96 +++++++++++++++++++
2 files changed, 96 insertions(+), 43 deletions(-)
delete mode 100644 Documentation/devicetree/bindings/input/touchscreen/bu21013.txt
create mode 100644 Documentation/devicetree/bindings/input/touchscreen/bu21013.yaml
diff --git a/Documentation/devicetree/bindings/input/touchscreen/bu21013.txt b/Documentation/devicetree/bindings/input/touchscreen/bu21013.txt
deleted file mode 100644
index da4c9d8b99b1..000000000000
--- a/Documentation/devicetree/bindings/input/touchscreen/bu21013.txt
+++ /dev/null
@@ -1,43 +0,0 @@
-* Rohm BU21013 Touch Screen
-
-Required properties:
- - compatible : "rohm,bu21013_tp"
- - reg : I2C device address
- - reset-gpios : GPIO pin enabling (selecting) chip (CS)
- - interrupt-parent : the phandle for the gpio controller
- - interrupts : (gpio) interrupt to which the chip is connected
-
-Optional properties:
- - touch-gpios : GPIO pin registering a touch event
- - <supply_name>-supply : Phandle to a regulator supply
- - touchscreen-size-x : General touchscreen binding, see [1].
- - touchscreen-size-y : General touchscreen binding, see [1].
- - touchscreen-inverted-x : General touchscreen binding, see [1].
- - touchscreen-inverted-y : General touchscreen binding, see [1].
- - touchscreen-swapped-x-y : General touchscreen binding, see [1].
-
-[1] All general touchscreen properties are described in
- Documentation/devicetree/bindings/input/touchscreen/touchscreen.txt.
-
-Deprecated properties:
- - rohm,touch-max-x : Maximum outward permitted limit in the X axis
- - rohm,touch-max-y : Maximum outward permitted limit in the Y axis
- - rohm,flip-x : Flip touch coordinates on the X axis
- - rohm,flip-y : Flip touch coordinates on the Y axis
-
-Example:
-
- i2c@80110000 {
- bu21013_tp@5c {
- compatible = "rohm,bu21013_tp";
- reg = <0x5c>;
- interrupt-parent = <&gpio2>;
- interrupts <&20 IRQ_TYPE_LEVEL_LOW>;
- touch-gpio = <&gpio2 20 GPIO_ACTIVE_LOW>;
- avdd-supply = <&ab8500_ldo_aux1_reg>;
-
- touchscreen-size-x = <384>;
- touchscreen-size-y = <704>;
- touchscreen-inverted-y;
- };
- };
diff --git a/Documentation/devicetree/bindings/input/touchscreen/bu21013.yaml b/Documentation/devicetree/bindings/input/touchscreen/bu21013.yaml
new file mode 100644
index 000000000000..aeb581fcaf29
--- /dev/null
+++ b/Documentation/devicetree/bindings/input/touchscreen/bu21013.yaml
@@ -0,0 +1,96 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/input/touchscreen/bu21013.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Rohm BU21013 touchscreen
+
+description:
+ Rohm BU21013 I2C driven touchscreen controller.
+
+maintainers:
+ - Dario Binacchi <dario.binacchi@amarulasolutions.com>
+
+allOf:
+ - $ref: touchscreen.yaml#
+
+properties:
+ compatible:
+ enum:
+ - rohm,bu21013_tp
+
+ reg:
+ maxItems: 1
+
+ interrupts:
+ maxItems: 1
+
+ reset-gpios:
+ maxItems: 1
+ description: GPIO resetting the chip
+
+ touch-gpios:
+ maxItems: 1
+ description: GPIO registering a touch event.
+
+ avdd-supply:
+ description: Phandle to the regulator supplying the analog circuit.
+
+ rohm,touch-max-x:
+ deprecated: true
+ description: Maximum value on the X axis.
+ $ref: /schemas/types.yaml#/definitions/uint32
+
+ rohm,touch-max-y:
+ deprecated: true
+ description: Maximum value on the Y axis.
+ $ref: /schemas/types.yaml#/definitions/uint32
+
+ rohm,flip-x:
+ deprecated: true
+ description: Flip touch coordinates on the X axis
+ type: boolean
+
+ rohm,flip-y:
+ deprecated: true
+ description: Flip touch coordinates on the Y axis
+ type: boolean
+
+ touchscreen-inverted-x: true
+ touchscreen-inverted-y: true
+ touchscreen-size-x: true
+ touchscreen-size-y: true
+ touchscreen-swapped-x-y: true
+
+additionalProperties: false
+
+required:
+ - compatible
+ - reg
+ - reset-gpios
+ - interrupts
+
+examples:
+ - |
+ #include <dt-bindings/gpio/gpio.h>
+ #include <dt-bindings/interrupt-controller/irq.h>
+ i2c {
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ touchscreen@5c {
+ compatible = "rohm,bu21013_tp";
+ reg = <0x5c>;
+
+ interrupt-parent = <&gpio2>;
+ interrupts = <0x20 IRQ_TYPE_LEVEL_LOW>;
+ reset-gpios = <&gpio2 19 GPIO_ACTIVE_LOW>;
+ touch-gpios = <&gpio2 20 GPIO_ACTIVE_LOW>;
+ avdd-supply = <&ab8500_ldo_aux1_reg>;
+
+ touchscreen-size-x = <384>;
+ touchscreen-size-y = <704>;
+ touchscreen-inverted-y;
+ };
+ };
--
2.43.0
base-commit: 22f20375f5b71f30c0d6896583b93b6e4bba7279
branch: drop-touchscreen.txt
^ permalink raw reply related
* Re: [regression] 1a8953f4f774 ("HID: Add IGNORE quirk for SMARTLINKTECHNOLOGY") causes issue with ID 4c4a:4155 Jieli Technology USB Composite Device
From: Salvatore Bonaccorso @ 2025-09-13 8:53 UTC (permalink / raw)
To: Staffan Melin
Cc: zhangheng, Jiri Kosina, Benjamin Tissoires, linux-input,
linux-kernel, regressions, stable, 1114557
In-Reply-To: <e81e8d68cb33c7de7b0e353791e21e53@oscillator.se>
Hi Staffan,
chiming in hopefully it is of help.
On Fri, Sep 12, 2025 at 09:57:04PM +0200, Staffan Melin wrote:
> Thank you,
>
> I tried to apply this patch to 6.12.39, the first problematic kernel, as
> well as 6.12.41, the first bad I tried, and on both I got an error message:
>
> Applying: HID: quirks: Add device descriptor for 4c4a:4155
> error: patch failed: drivers/hid/hid-quirks.c:1068
> error: drivers/hid/hid-quirks.c: patch does not apply
> Patch failed at 0001 HID: quirks: Add device descriptor for 4c4a:4155
>
> To which kernel version should I apply the patch?
As the deveopment goes from mainline then down to stable series, the
fix needs to be developed first for mainline. So the patch is targeted
there.
But please find attached an updated patch which hopefully should work
which resolved the context changes.
But ideally you can provide a Tested-by on zhangheng's mainline patch
to get things rolling as needed.
Regards,
Salvatore
^ permalink raw reply
* Re: [syzbot] [usb?] [input?] WARNING in bcm5974_start_traffic/usb_submit_urb (2)
From: syzbot @ 2025-09-13 5:52 UTC (permalink / raw)
To: dmitry.torokhov, gregkh, javier.carrasco, linux-input,
linux-kernel, linux-usb, mukattreyee, oneukum, rydberg, stern,
syzkaller-bugs
In-Reply-To: <0000000000008acb1e061618e68e@google.com>
syzbot suspects this issue was fixed by commit:
commit 503bbde34cc3dd2acd231f277ba70c3f9ed22e59
Author: Oliver Neukum <oneukum@suse.com>
Date: Thu Jun 12 12:20:25 2025 +0000
usb: core: usb_submit_urb: downgrade type check
bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=16c6eb62580000
start commit: e8ab83e34bdc Merge tag 'arm64-fixes' of git://git.kernel.o..
git tree: upstream
kernel config: https://syzkaller.appspot.com/x/.config?x=a9a25b7a36123454
dashboard link: https://syzkaller.appspot.com/bug?extid=b064b5599f18f7ebb1e1
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=1514f0f4580000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=11699a70580000
If the result looks correct, please mark the issue as fixed by replying with:
#syz fix: usb: core: usb_submit_urb: downgrade type check
For information about bisection process see: https://goo.gl/tpsmEJ#bisection
^ permalink raw reply
* Re: [PATCH v2 05/12] dt-bindings: display: mediatek,od: Add mediatek,gce-client-reg property
From: Rob Herring (Arm) @ 2025-09-12 22:37 UTC (permalink / raw)
To: Ariel D'Alessandro
Cc: luiz.dentz, kuba, airlied, mripard, angelogioacchino.delregno,
linux-arm-kernel, maarten.lankhorst, tiffany.lin, houlong.wei,
minghsiu.tsai, lgirdwood, louisalexis.eyraud, linus.walleij,
devicetree, linux-input, yunfei.dong, edumazet, linux-bluetooth,
tzimmermann, broonie, andrew+netdev, kernel, chunkuang.hu,
amergnat, conor+dt, matthias.bgg, support.opensource,
linux-rockchip, davem, andrew-ct.chen, krzk+dt, p.zabel,
sean.wang, linux-kernel, simona, linux-mediatek, marcel,
dmitry.torokhov, dri-devel, pabeni, jeesw, mchehab, linux-media,
flora.fu, linux-gpio, heiko, linux-sound, netdev
In-Reply-To: <20250911151001.108744-6-ariel.dalessandro@collabora.com>
On Thu, 11 Sep 2025 12:09:54 -0300, Ariel D'Alessandro wrote:
> Currently, users of Mediatek OD (display overdrive) DT bindings set
> mediatek,gce-client-reg node property, which is missing from the DT schema.
>
> For example, device tree arch/arm64/boot/dts/mediatek/mt8173.dtsi is
> causing the following dtb check error:
>
> ```
> $ make CHECK_DTBS=y mediatek/mt8173-elm.dtb
> SCHEMA Documentation/devicetree/bindings/processed-schema.json
> DTC [C] arch/arm64/boot/dts/mediatek/mt8173-elm.dtb
> [...]
> arch/arm64/boot/dts/mediatek/mt8173-elm.dtb: od@14023000
> (mediatek,mt8173-disp-od): 'mediatek,gce-client-reg' does not match
> any of the regexes: '^pinctrl-[0-9]+$'
> ```
>
> This commit adds the missing node property in the DT schema and updates the
> example as well.
>
> Signed-off-by: Ariel D'Alessandro <ariel.dalessandro@collabora.com>
> ---
> .../bindings/display/mediatek/mediatek,od.yaml | 14 ++++++++++++++
> 1 file changed, 14 insertions(+)
>
I fixed up the commit msg with Krzysztof's comments and applied both
display patches. Thanks!
Rob
^ permalink raw reply
* Re: Adding Kernel xpad support for Flydigi Apex 5
From: Vicki Pfau @ 2025-09-12 21:31 UTC (permalink / raw)
To: Jiri Kosina, brandon; +Cc: linux-input
In-Reply-To: <r856p96o-850o-47sn-rsps-04n94875o975@xreary.bet>
Hi Jiri and Brandon,
On 9/12/25 7:14 AM, Jiri Kosina wrote:
> On Fri, 12 Sep 2025, Jiri Kosina wrote:
>
>> On Fri, 29 Aug 2025, brandon wrote:
>>
>>> Hello,
>>>
>>> This adds support for the Flydigi Apex 5 controller to the xpad kernel module. This has been tested as part of: https://github.com/paroj/xpad/pull/328
>>>
>>> This patch is very small and simple, it simply white lists the controller so it can be used. The trigger and stick mappings are none standard, but SDL has the correct mappings already shipped so it's not a problem in Steam or games I've tried.
>>>
>>> Here is the patch:
>>>
>>> --- linux/drivers/input/joystick/xpad.c.orig 2025-08-29 18:20:56.157442704 +0100
>>> +++ linux/drivers/input/joystick/xpad.c 2025-08-29 18:19:29.539174760 +0100
>>> @@ -422,6 +422,7 @@ static const struct xpad_device {
>>> { 0x3537, 0x1010, "GameSir G7 SE", 0, XTYPE_XBOXONE },
>>> { 0x366c, 0x0005, "ByoWave Proteus Controller", MAP_SHARE_BUTTON, XTYPE_XBOXONE, FLAG_DELAY_INIT },
>>> { 0x3767, 0x0101, "Fanatec Speedster 3 Forceshock Wheel", 0, XTYPE_XBOX },
>>> + { 0x37d7, 0x2501, "Flydigi Apex 5", 0, XTYPE_XBOX },
>>> { 0x413d, 0x2104, "Black Shark Green Ghost Gamepad", 0, XTYPE_XBOX360 },
>>> { 0xffff, 0xffff, "Chinese-made Xbox Controller", 0, XTYPE_XBOX },
>>> { 0x0000, 0x0000, "Generic X-Box pad", 0, XTYPE_UNKNOWN }
>>> @@ -578,6 +579,7 @@ static const struct usb_device_id xpad_t
>>> XPAD_XBOX360_VENDOR(0x3537), /* GameSir Controllers */
>>> XPAD_XBOXONE_VENDOR(0x3537), /* GameSir Controllers */
>>> XPAD_XBOXONE_VENDOR(0x366c), /* ByoWave controllers */
>>> + XPAD_XBOXONE_VENDOR(0x37d7), /* Flydigi Controllers */
>>> XPAD_XBOX360_VENDOR(0x413d), /* Black Shark Green Ghost Controller */
>>> { }
>>> };
>>
>> Thanks for the patch.
>>
>> It's however whitespace-damaged, and also is missing a signoff. Could you
>> please fix that and resubmit, so that I could merge it?
>
> (sorry, it's not me who should be merging it, but Dmitry Torokhov as an
> input maintainer -- please CC him as well on the next iteration)
>
A corrected patch was already submitted by someone else: https://lore.kernel.org/linux-input/20250903165114.2987905-1-lkml@antheas.dev/
Vicki
^ permalink raw reply
* Re: [regression] 1a8953f4f774 ("HID: Add IGNORE quirk for SMARTLINKTECHNOLOGY") causes issue with ID 4c4a:4155 Jieli Technology USB Composite Device
From: Staffan Melin @ 2025-09-12 19:57 UTC (permalink / raw)
To: zhangheng
Cc: Salvatore Bonaccorso, Jiri Kosina, Benjamin Tissoires,
linux-input, linux-kernel, regressions, stable, 1114557
In-Reply-To: <c8f3d402-e0ec-4767-b925-d7764aec3d93@kylinos.cn>
Thank you,
I tried to apply this patch to 6.12.39, the first problematic kernel, as
well as 6.12.41, the first bad I tried, and on both I got an error
message:
Applying: HID: quirks: Add device descriptor for 4c4a:4155
error: patch failed: drivers/hid/hid-quirks.c:1068
error: drivers/hid/hid-quirks.c: patch does not apply
Patch failed at 0001 HID: quirks: Add device descriptor for 4c4a:4155
To which kernel version should I apply the patch?
I used the command
git am < <file>
Best regards,
Staffan
On 2025-09-12 14:49, zhangheng wrote:
> I am currently working on resolving this issue by limiting the original
> patch
>
> based on the information in the device descriptor to avoid affecting
> your device,
>
> You can try this patch and look forward to your reply.
>
>
> 在 2025/9/7 23:10, Salvatore Bonaccorso 写道:
>> Hi Zhang, hi Jiri,
>>
>> In Debian Staffan Melin reported that after an update containing the
>> commit 1a8953f4f774 ("HID: Add IGNORE quirk for SMARTLINKTECHNOLOGY"),
>> the input device with same idVendor and idProduct, the Jieli
>> Technology USB Composite Device, does not get recognized anymore.
>>
>> The full Debian report is at: https://bugs.debian.org/1114557
>>
>> The issue is not specific to the 6.12.y series and confirmed in 6.16.3
>> as well.
>>
>> Staffan Melin did bisect the kernels between 6.12.38 (which was still
>> working) and 6.1.41 (which was not), confirming by bisection that the
>> offending commit is
>>
>> 1a8953f4f774 ("HID: Add IGNORE quirk for SMARTLINKTECHNOLOGY")
>>
>> #regzbot introduced: 1a8953f4f774
>> #regzbot monitor: https://bugs.debian.org/1114557
>>
>> So it looks that the quirk applied is unfortunately affecting
>> negatively as well Staffan Melin case.
>>
>> Can you have a look?
>>
>> Regards,
>> Salvatore
^ permalink raw reply
* Re: [PATCH v1 2/2] Drivers: hv: Make CONFIG_HYPERV bool
From: Mukesh R @ 2025-09-12 18:10 UTC (permalink / raw)
To: Greg KH
Cc: dri-devel, linux-kernel, linux-input, linux-hyperv, netdev,
linux-pci, linux-scsi, linux-fbdev, linux-arch, virtualization,
maarten.lankhorst, mripard, tzimmermann, airlied, simona, jikos,
bentiss, kys, haiyangz, wei.liu, decui, dmitry.torokhov,
andrew+netdev, davem, edumazet, kuba, pabeni, bhelgaas,
James.Bottomley, martin.petersen, deller, arnd, sgarzare, horms
In-Reply-To: <2025091253-overwrite-carol-b197@gregkh>
On 9/12/25 04:43, Greg KH wrote:
> On Mon, Sep 08, 2025 at 02:01:34PM -0700, Mukesh R wrote:
>> On 9/6/25 04:36, Greg KH wrote:
>>> On Fri, Sep 05, 2025 at 06:09:52PM -0700, Mukesh Rathor wrote:
>>>> With CONFIG_HYPERV and CONFIG_HYPERV_VMBUS separated, change CONFIG_HYPERV
>>>> to bool from tristate. CONFIG_HYPERV now becomes the core Hyper-V
>>>> hypervisor support, such as hypercalls, clocks/timers, Confidential
>>>> Computing setup, PCI passthru, etc. that doesn't involve VMBus or VMBus
>>>> devices.
>>>
>>> But why are you making it so that this can not be a module anymore? You
>>> are now forcing ALL Linux distro users to always have this code in their
>>> system, despite not ever using the feature. That feels like a waste to
>>> me.
>>>
>>> What is preventing this from staying as a module? Why must you always
>>> have this code loaded at all times for everyone?
>>
>> This is currently not a module. I assume it was at the beginning. In
>> drivers/Makefile today:
>>
>> obj-$(subst m,y,$(CONFIG_HYPERV)) += hv/
>>
>>
>> More context: CONFIG_HYPERV doesn't really reflect one module. It is
>> both for kernel built in code and building of stuff in drivers/hv.
>>
>> drivers/hv then builds 4 modules:
>>
>> obj-$(CONFIG_HYPERV) += hv_vmbus.o
>> obj-$(CONFIG_HYPERV_UTILS) += hv_utils.o
>> obj-$(CONFIG_HYPERV_BALLOON) += hv_balloon.o
>> obj-$(CONFIG_MSHV_ROOT) += mshv_root.o
>>
>> Notice vmbus is using CONFIG_HYPERV because there is no
>> CONFIG_HYPERV_VMBUS. We are trying to fix that here.
>
> This series does not apply to my tree:
>
> checking file drivers/gpu/drm/Kconfig
> checking file drivers/hid/Kconfig
> checking file drivers/hv/Kconfig
> Hunk #2 FAILED at 82.
> 1 out of 2 hunks FAILED
> checking file drivers/hv/Makefile
> checking file drivers/input/serio/Kconfig
> checking file drivers/net/hyperv/Kconfig
> checking file drivers/pci/Kconfig
> checking file drivers/scsi/Kconfig
> checking file drivers/uio/Kconfig
> checking file drivers/video/fbdev/Kconfig
> checking file include/asm-generic/mshyperv.h
> Hunk #1 succeeded at 162 with fuzz 2 (offset -3 lines).
> Hunk #2 succeeded at 198 (offset -3 lines).
> Hunk #3 succeeded at 215 (offset -3 lines).
> checking file net/vmw_vsock/Kconfig
>
> What was it made against?
>
Sorry to hear that. It was built against hyper-next, but perhaps I
accidentally used our internal mirror. Let me rebase and send V2
right away.
Thanks,
-Mukesh
^ permalink raw reply
* Re: [PATCH v2 00/11] HID: playstation: Add support for audio jack handling on DualSense
From: Benjamin Tissoires @ 2025-09-12 17:04 UTC (permalink / raw)
To: Jiri Kosina
Cc: Cristian Ciocaltea, Roderick Colenbrander, Roderick Colenbrander,
Henrik Rydberg, kernel, linux-input, linux-kernel
In-Reply-To: <s3tus6usbokl5hpwlbzbxfqdqwnyxqqnjiwhzdbd5obvfxavvf@cwopa6fpjctb>
On Sep 12 2025, Benjamin Tissoires wrote:
> On Sep 12 2025, Jiri Kosina wrote:
> > On Wed, 27 Aug 2025, Cristian Ciocaltea wrote:
> >
> > > It's been over a month now since this was kind of blocked without any clear
> > > reason, and by the end of next week I'll be on leave, which means we're
> > > close to missing the merge window once again.
> > >
> > > Considering the counterpart quirk in the generic USB audio driver has been
> > > already merged since v6.17, I kindly ask for your support in getting this
> > > into v6.18.
> >
> > Roderick, do you have any word on this, please?
>
> If this can help moving forward:
> Patches 1-9 are:
> Reviewed-by: Benjamin Tissoires <bentiss@kernel.org>
> and can be merged right away
Also, to add a little bit more of confidence, I've just ran the
test_sony.py regression tests on the full series and no regressions have
been detected. (those will be run anyway after the series is accepted)
It would be nice to have new tests for this feature but I'm not sure
it'll be easy to add given that we are talking about new interaction
with other parts of the kernel.
Cheers,
Benjamin
^ permalink raw reply
* [PATCH v3 3/3] HID: hidraw: tighten ioctl command parsing
From: Benjamin Tissoires @ 2025-09-12 16:58 UTC (permalink / raw)
To: Jiri Kosina, Shuah Khan, Arnd Bergmann
Cc: linux-input, linux-kselftest, linux-kernel, Benjamin Tissoires,
Arnd Bergmann
In-Reply-To: <20250912-b4-hidraw-ioctls-v3-0-cd2c6efd8c20@kernel.org>
The handling for variable-length ioctl commands in hidraw_ioctl() is
rather complex and the check for the data direction is incomplete.
Simplify this code by factoring out the various ioctls grouped by dir
and size, and using a switch() statement with the size masked out, to
ensure the rest of the command is correctly matched.
Fixes: 9188e79ec3fd ("HID: add phys and name ioctls to hidraw")
Reported-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Benjamin Tissoires <bentiss@kernel.org>
---
drivers/hid/hidraw.c | 224 ++++++++++++++++++++++++--------------------
include/uapi/linux/hidraw.h | 2 +
2 files changed, 124 insertions(+), 102 deletions(-)
diff --git a/drivers/hid/hidraw.c b/drivers/hid/hidraw.c
index c887f48756f4be2a4bac03128f2885bde96c1e39..bbd6f23bce78951c7d667ff5c1c923cee3509e3f 100644
--- a/drivers/hid/hidraw.c
+++ b/drivers/hid/hidraw.c
@@ -394,27 +394,15 @@ static int hidraw_revoke(struct hidraw_list *list)
return 0;
}
-static long hidraw_ioctl(struct file *file, unsigned int cmd,
- unsigned long arg)
+static long hidraw_fixed_size_ioctl(struct file *file, struct hidraw *dev, unsigned int cmd,
+ void __user *arg)
{
- struct inode *inode = file_inode(file);
- unsigned int minor = iminor(inode);
- long ret = 0;
- struct hidraw *dev;
- struct hidraw_list *list = file->private_data;
- void __user *user_arg = (void __user*) arg;
-
- down_read(&minors_rwsem);
- dev = hidraw_table[minor];
- if (!dev || !dev->exist || hidraw_is_revoked(list)) {
- ret = -ENODEV;
- goto out;
- }
+ struct hid_device *hid = dev->hid;
switch (cmd) {
case HIDIOCGRDESCSIZE:
- if (put_user(dev->hid->rsize, (int __user *)arg))
- ret = -EFAULT;
+ if (put_user(hid->rsize, (int __user *)arg))
+ return -EFAULT;
break;
case HIDIOCGRDESC:
@@ -422,113 +410,145 @@ static long hidraw_ioctl(struct file *file, unsigned int cmd,
__u32 len;
if (get_user(len, (int __user *)arg))
- ret = -EFAULT;
- else if (len > HID_MAX_DESCRIPTOR_SIZE - 1)
- ret = -EINVAL;
- else if (copy_to_user(user_arg + offsetof(
- struct hidraw_report_descriptor,
- value[0]),
- dev->hid->rdesc,
- min(dev->hid->rsize, len)))
- ret = -EFAULT;
+ return -EFAULT;
+
+ if (len > HID_MAX_DESCRIPTOR_SIZE - 1)
+ return -EINVAL;
+
+ if (copy_to_user(arg + offsetof(
+ struct hidraw_report_descriptor,
+ value[0]),
+ hid->rdesc,
+ min(hid->rsize, len)))
+ return -EFAULT;
+
break;
}
case HIDIOCGRAWINFO:
{
struct hidraw_devinfo dinfo;
- dinfo.bustype = dev->hid->bus;
- dinfo.vendor = dev->hid->vendor;
- dinfo.product = dev->hid->product;
- if (copy_to_user(user_arg, &dinfo, sizeof(dinfo)))
- ret = -EFAULT;
+ dinfo.bustype = hid->bus;
+ dinfo.vendor = hid->vendor;
+ dinfo.product = hid->product;
+ if (copy_to_user(arg, &dinfo, sizeof(dinfo)))
+ return -EFAULT;
break;
}
case HIDIOCREVOKE:
{
- if (user_arg)
- ret = -EINVAL;
- else
- ret = hidraw_revoke(list);
- break;
+ struct hidraw_list *list = file->private_data;
+
+ if (arg)
+ return -EINVAL;
+
+ return hidraw_revoke(list);
}
default:
- {
- struct hid_device *hid = dev->hid;
- if (_IOC_TYPE(cmd) != 'H') {
- ret = -EINVAL;
- break;
- }
+ /*
+ * None of the above ioctls can return -EAGAIN, so
+ * use it as a marker that we need to check variable
+ * length ioctls.
+ */
+ return -EAGAIN;
+ }
- if (_IOC_NR(cmd) == _IOC_NR(HIDIOCSFEATURE(0))) {
- int len = _IOC_SIZE(cmd);
- ret = hidraw_send_report(file, user_arg, len, HID_FEATURE_REPORT);
- break;
- }
- if (_IOC_NR(cmd) == _IOC_NR(HIDIOCGFEATURE(0))) {
- int len = _IOC_SIZE(cmd);
- ret = hidraw_get_report(file, user_arg, len, HID_FEATURE_REPORT);
- break;
- }
+ return 0;
+}
- if (_IOC_NR(cmd) == _IOC_NR(HIDIOCSINPUT(0))) {
- int len = _IOC_SIZE(cmd);
- ret = hidraw_send_report(file, user_arg, len, HID_INPUT_REPORT);
- break;
- }
- if (_IOC_NR(cmd) == _IOC_NR(HIDIOCGINPUT(0))) {
- int len = _IOC_SIZE(cmd);
- ret = hidraw_get_report(file, user_arg, len, HID_INPUT_REPORT);
- break;
- }
+static long hidraw_rw_variable_size_ioctl(struct file *file, struct hidraw *dev, unsigned int cmd,
+ void __user *user_arg)
+{
+ int len = _IOC_SIZE(cmd);
+
+ switch (cmd & ~IOCSIZE_MASK) {
+ case HIDIOCSFEATURE(0):
+ return hidraw_send_report(file, user_arg, len, HID_FEATURE_REPORT);
+ case HIDIOCGFEATURE(0):
+ return hidraw_get_report(file, user_arg, len, HID_FEATURE_REPORT);
+ case HIDIOCSINPUT(0):
+ return hidraw_send_report(file, user_arg, len, HID_INPUT_REPORT);
+ case HIDIOCGINPUT(0):
+ return hidraw_get_report(file, user_arg, len, HID_INPUT_REPORT);
+ case HIDIOCSOUTPUT(0):
+ return hidraw_send_report(file, user_arg, len, HID_OUTPUT_REPORT);
+ case HIDIOCGOUTPUT(0):
+ return hidraw_get_report(file, user_arg, len, HID_OUTPUT_REPORT);
+ }
- if (_IOC_NR(cmd) == _IOC_NR(HIDIOCSOUTPUT(0))) {
- int len = _IOC_SIZE(cmd);
- ret = hidraw_send_report(file, user_arg, len, HID_OUTPUT_REPORT);
- break;
- }
- if (_IOC_NR(cmd) == _IOC_NR(HIDIOCGOUTPUT(0))) {
- int len = _IOC_SIZE(cmd);
- ret = hidraw_get_report(file, user_arg, len, HID_OUTPUT_REPORT);
- break;
- }
+ return -EINVAL;
+}
- /* Begin Read-only ioctls. */
- if (_IOC_DIR(cmd) != _IOC_READ) {
- ret = -EINVAL;
- break;
- }
+static long hidraw_ro_variable_size_ioctl(struct file *file, struct hidraw *dev, unsigned int cmd,
+ void __user *user_arg)
+{
+ struct hid_device *hid = dev->hid;
+ int len = _IOC_SIZE(cmd);
+ int field_len;
+
+ switch (cmd & ~IOCSIZE_MASK) {
+ case HIDIOCGRAWNAME(0):
+ field_len = strlen(hid->name) + 1;
+ if (len > field_len)
+ len = field_len;
+ return copy_to_user(user_arg, hid->name, len) ? -EFAULT : len;
+ case HIDIOCGRAWPHYS(0):
+ field_len = strlen(hid->phys) + 1;
+ if (len > field_len)
+ len = field_len;
+ return copy_to_user(user_arg, hid->phys, len) ? -EFAULT : len;
+ case HIDIOCGRAWUNIQ(0):
+ field_len = strlen(hid->uniq) + 1;
+ if (len > field_len)
+ len = field_len;
+ return copy_to_user(user_arg, hid->uniq, len) ? -EFAULT : len;
+ }
- if (_IOC_NR(cmd) == _IOC_NR(HIDIOCGRAWNAME(0))) {
- int len = strlen(hid->name) + 1;
- if (len > _IOC_SIZE(cmd))
- len = _IOC_SIZE(cmd);
- ret = copy_to_user(user_arg, hid->name, len) ?
- -EFAULT : len;
- break;
- }
+ return -EINVAL;
+}
- if (_IOC_NR(cmd) == _IOC_NR(HIDIOCGRAWPHYS(0))) {
- int len = strlen(hid->phys) + 1;
- if (len > _IOC_SIZE(cmd))
- len = _IOC_SIZE(cmd);
- ret = copy_to_user(user_arg, hid->phys, len) ?
- -EFAULT : len;
- break;
- }
+static long hidraw_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
+{
+ struct inode *inode = file_inode(file);
+ unsigned int minor = iminor(inode);
+ struct hidraw *dev;
+ struct hidraw_list *list = file->private_data;
+ void __user *user_arg = (void __user *)arg;
+ int ret;
- if (_IOC_NR(cmd) == _IOC_NR(HIDIOCGRAWUNIQ(0))) {
- int len = strlen(hid->uniq) + 1;
- if (len > _IOC_SIZE(cmd))
- len = _IOC_SIZE(cmd);
- ret = copy_to_user(user_arg, hid->uniq, len) ?
- -EFAULT : len;
- break;
- }
- }
+ down_read(&minors_rwsem);
+ dev = hidraw_table[minor];
+ if (!dev || !dev->exist || hidraw_is_revoked(list)) {
+ ret = -ENODEV;
+ goto out;
+ }
+
+ if (_IOC_TYPE(cmd) != 'H') {
+ ret = -EINVAL;
+ goto out;
+ }
+ if (_IOC_NR(cmd) > HIDIOCTL_LAST || _IOC_NR(cmd) == 0) {
ret = -ENOTTY;
+ goto out;
}
+
+ ret = hidraw_fixed_size_ioctl(file, dev, cmd, user_arg);
+ if (ret != -EAGAIN)
+ goto out;
+
+ switch (_IOC_DIR(cmd)) {
+ case (_IOC_READ | _IOC_WRITE):
+ ret = hidraw_rw_variable_size_ioctl(file, dev, cmd, user_arg);
+ break;
+ case _IOC_READ:
+ ret = hidraw_ro_variable_size_ioctl(file, dev, cmd, user_arg);
+ break;
+ default:
+ /* Any other IOC_DIR is wrong */
+ ret = -EINVAL;
+ }
+
out:
up_read(&minors_rwsem);
return ret;
diff --git a/include/uapi/linux/hidraw.h b/include/uapi/linux/hidraw.h
index d5ee269864e07fcaba481fa285bacbd98739e44f..ebd701b3c18d9d7465880199091933f13f2e1128 100644
--- a/include/uapi/linux/hidraw.h
+++ b/include/uapi/linux/hidraw.h
@@ -48,6 +48,8 @@ struct hidraw_devinfo {
#define HIDIOCGOUTPUT(len) _IOC(_IOC_WRITE|_IOC_READ, 'H', 0x0C, len)
#define HIDIOCREVOKE _IOW('H', 0x0D, int) /* Revoke device access */
+#define HIDIOCTL_LAST _IOC_NR(HIDIOCREVOKE)
+
#define HIDRAW_FIRST_MINOR 0
#define HIDRAW_MAX_DEVICES 64
/* number of reports to buffer */
--
2.51.0
^ permalink raw reply related
* [PATCH v3 2/3] selftests/hid: hidraw: forge wrong ioctls and tests them
From: Benjamin Tissoires @ 2025-09-12 16:58 UTC (permalink / raw)
To: Jiri Kosina, Shuah Khan, Arnd Bergmann
Cc: linux-input, linux-kselftest, linux-kernel, Benjamin Tissoires
In-Reply-To: <20250912-b4-hidraw-ioctls-v3-0-cd2c6efd8c20@kernel.org>
We also need coverage for when the malicious user is not using the
proper ioctls definitions and tries to work around the driver.
Most of the scaffholding has been generated by claude-4-sonnet and then
carefully reviewed.
Suggested-by: Arnd Bergmann <arnd@kernel.org>
Signed-off-by: Benjamin Tissoires <bentiss@kernel.org>
---
tools/testing/selftests/hid/hidraw.c | 127 +++++++++++++++++++++++++++++++++++
1 file changed, 127 insertions(+)
diff --git a/tools/testing/selftests/hid/hidraw.c b/tools/testing/selftests/hid/hidraw.c
index 6d61d03e2ef05e1900fe5a3938d93421717b2621..d625772f8b7cf71fd94956d3a49d54ff44e2b34d 100644
--- a/tools/testing/selftests/hid/hidraw.c
+++ b/tools/testing/selftests/hid/hidraw.c
@@ -332,6 +332,133 @@ TEST_F(hidraw, ioctl_gfeature_invalid)
ASSERT_EQ(errno, EIO) TH_LOG("expected EIO, got errno %d", errno);
}
+/*
+ * Test ioctl with incorrect nr bits
+ */
+TEST_F(hidraw, ioctl_invalid_nr)
+{
+ char buf[256] = {0};
+ int err;
+ unsigned int bad_cmd;
+
+ /*
+ * craft an ioctl command with wrong _IOC_NR bits
+ */
+ bad_cmd = _IOC(_IOC_WRITE|_IOC_READ, 'H', 0x00, sizeof(buf)); /* 0 is not valid */
+
+ /* test the ioctl */
+ err = ioctl(self->hidraw_fd, bad_cmd, buf);
+ ASSERT_LT(err, 0) TH_LOG("ioctl read-write with wrong _IOC_NR (0) should have failed");
+ ASSERT_EQ(errno, ENOTTY)
+ TH_LOG("expected ENOTTY for wrong read-write _IOC_NR (0), got errno %d", errno);
+
+ /*
+ * craft an ioctl command with wrong _IOC_NR bits
+ */
+ bad_cmd = _IOC(_IOC_READ, 'H', 0x00, sizeof(buf)); /* 0 is not valid */
+
+ /* test the ioctl */
+ err = ioctl(self->hidraw_fd, bad_cmd, buf);
+ ASSERT_LT(err, 0) TH_LOG("ioctl read-only with wrong _IOC_NR (0) should have failed");
+ ASSERT_EQ(errno, ENOTTY)
+ TH_LOG("expected ENOTTY for wrong read-only _IOC_NR (0), got errno %d", errno);
+
+ /* also test with bigger number */
+ bad_cmd = _IOC(_IOC_READ, 'H', 0x42, sizeof(buf)); /* 0x42 is not valid as well */
+
+ err = ioctl(self->hidraw_fd, bad_cmd, buf);
+ ASSERT_LT(err, 0) TH_LOG("ioctl read-only with wrong _IOC_NR (0x42) should have failed");
+ ASSERT_EQ(errno, ENOTTY)
+ TH_LOG("expected ENOTTY for wrong read-only _IOC_NR (0x42), got errno %d", errno);
+
+ /* also test with bigger number: 0x42 is not valid as well */
+ bad_cmd = _IOC(_IOC_WRITE|_IOC_READ, 'H', 0x42, sizeof(buf));
+
+ err = ioctl(self->hidraw_fd, bad_cmd, buf);
+ ASSERT_LT(err, 0) TH_LOG("ioctl read-write with wrong _IOC_NR (0x42) should have failed");
+ ASSERT_EQ(errno, ENOTTY)
+ TH_LOG("expected ENOTTY for wrong read-write _IOC_NR (0x42), got errno %d", errno);
+}
+
+/*
+ * Test ioctl with incorrect type bits
+ */
+TEST_F(hidraw, ioctl_invalid_type)
+{
+ char buf[256] = {0};
+ int err;
+ unsigned int bad_cmd;
+
+ /*
+ * craft an ioctl command with wrong _IOC_TYPE bits
+ */
+ bad_cmd = _IOC(_IOC_WRITE|_IOC_READ, 'I', 0x01, sizeof(buf)); /* 'I' should be 'H' */
+
+ /* test the ioctl */
+ err = ioctl(self->hidraw_fd, bad_cmd, buf);
+ ASSERT_LT(err, 0) TH_LOG("ioctl with wrong _IOC_TYPE (I) should have failed");
+ ASSERT_EQ(errno, EINVAL) TH_LOG("expected EINVAL for wrong _IOC_NR, got errno %d", errno);
+}
+
+/*
+ * Test HIDIOCGFEATURE ioctl with incorrect _IOC_DIR bits
+ */
+TEST_F(hidraw, ioctl_gfeature_invalid_dir)
+{
+ __u8 buf[10] = {0};
+ int err;
+ unsigned int bad_cmd;
+
+ /* set report ID 1 in first byte */
+ buf[0] = 1;
+
+ /*
+ * craft an ioctl command with wrong _IOC_DIR bits
+ * HIDIOCGFEATURE should have _IOC_WRITE|_IOC_READ, let's use only _IOC_WRITE
+ */
+ bad_cmd = _IOC(_IOC_WRITE, 'H', 0x07, sizeof(buf)); /* should be _IOC_WRITE|_IOC_READ */
+
+ /* try to get feature report with wrong direction bits */
+ err = ioctl(self->hidraw_fd, bad_cmd, buf);
+ ASSERT_LT(err, 0) TH_LOG("HIDIOCGFEATURE with wrong _IOC_DIR should have failed");
+ ASSERT_EQ(errno, EINVAL) TH_LOG("expected EINVAL for wrong _IOC_DIR, got errno %d", errno);
+
+ /* also test with only _IOC_READ */
+ bad_cmd = _IOC(_IOC_READ, 'H', 0x07, sizeof(buf)); /* should be _IOC_WRITE|_IOC_READ */
+
+ err = ioctl(self->hidraw_fd, bad_cmd, buf);
+ ASSERT_LT(err, 0) TH_LOG("HIDIOCGFEATURE with wrong _IOC_DIR should have failed");
+ ASSERT_EQ(errno, EINVAL) TH_LOG("expected EINVAL for wrong _IOC_DIR, got errno %d", errno);
+}
+
+/*
+ * Test read-only ioctl with incorrect _IOC_DIR bits
+ */
+TEST_F(hidraw, ioctl_readonly_invalid_dir)
+{
+ char buf[256] = {0};
+ int err;
+ unsigned int bad_cmd;
+
+ /*
+ * craft an ioctl command with wrong _IOC_DIR bits
+ * HIDIOCGRAWNAME should have _IOC_READ, let's use _IOC_WRITE
+ */
+ bad_cmd = _IOC(_IOC_WRITE, 'H', 0x04, sizeof(buf)); /* should be _IOC_READ */
+
+ /* try to get device name with wrong direction bits */
+ err = ioctl(self->hidraw_fd, bad_cmd, buf);
+ ASSERT_LT(err, 0) TH_LOG("HIDIOCGRAWNAME with wrong _IOC_DIR should have failed");
+ ASSERT_EQ(errno, EINVAL) TH_LOG("expected EINVAL for wrong _IOC_DIR, got errno %d", errno);
+
+ /* also test with _IOC_WRITE|_IOC_READ */
+ bad_cmd = _IOC(_IOC_WRITE|_IOC_READ, 'H', 0x04, sizeof(buf)); /* should be only _IOC_READ */
+
+ err = ioctl(self->hidraw_fd, bad_cmd, buf);
+ ASSERT_LT(err, 0) TH_LOG("HIDIOCGRAWNAME with wrong _IOC_DIR should have failed");
+ ASSERT_EQ(errno, EINVAL) TH_LOG("expected EINVAL for wrong _IOC_DIR, got errno %d", errno);
+}
+
/*
* Test HIDIOCSFEATURE ioctl to set feature report
*/
--
2.51.0
^ permalink raw reply related
* [PATCH v3 1/3] selftests/hid: hidraw: add more coverage for hidraw ioctls
From: Benjamin Tissoires @ 2025-09-12 16:58 UTC (permalink / raw)
To: Jiri Kosina, Shuah Khan, Arnd Bergmann
Cc: linux-input, linux-kselftest, linux-kernel, Benjamin Tissoires
In-Reply-To: <20250912-b4-hidraw-ioctls-v3-0-cd2c6efd8c20@kernel.org>
Try to ensure all ioctls are having at least one test.
Most of the scaffholding has been generated by claude-4-sonnet and then
carefully reviewed.
Signed-off-by: Benjamin Tissoires <bentiss@kernel.org>
---
tools/testing/selftests/hid/hid_common.h | 6 +
tools/testing/selftests/hid/hidraw.c | 346 +++++++++++++++++++++++++++++++
2 files changed, 352 insertions(+)
diff --git a/tools/testing/selftests/hid/hid_common.h b/tools/testing/selftests/hid/hid_common.h
index f77f69c6657d0f0f66beb3b50bf4b126f6f63348..8085519c47cb505b901ac80f2087dc9a1aa2b9c0 100644
--- a/tools/testing/selftests/hid/hid_common.h
+++ b/tools/testing/selftests/hid/hid_common.h
@@ -230,6 +230,12 @@ static int uhid_event(struct __test_metadata *_metadata, int fd)
break;
case UHID_SET_REPORT:
UHID_LOG("UHID_SET_REPORT from uhid-dev");
+
+ answer.type = UHID_SET_REPORT_REPLY;
+ answer.u.set_report_reply.id = ev.u.set_report.id;
+ answer.u.set_report_reply.err = 0; /* success */
+
+ uhid_write(_metadata, fd, &answer);
break;
default:
TH_LOG("Invalid event from uhid-dev: %u", ev.type);
diff --git a/tools/testing/selftests/hid/hidraw.c b/tools/testing/selftests/hid/hidraw.c
index 821db37ba4bbef82e5cf4b44b6675666f87a12ad..6d61d03e2ef05e1900fe5a3938d93421717b2621 100644
--- a/tools/testing/selftests/hid/hidraw.c
+++ b/tools/testing/selftests/hid/hidraw.c
@@ -2,6 +2,9 @@
/* Copyright (c) 2022-2024 Red Hat */
#include "hid_common.h"
+#include <linux/input.h>
+#include <string.h>
+#include <sys/ioctl.h>
/* for older kernels */
#ifndef HIDIOCREVOKE
@@ -215,6 +218,349 @@ TEST_F(hidraw, write_event_revoked)
pthread_mutex_unlock(&uhid_output_mtx);
}
+/*
+ * Test HIDIOCGRDESCSIZE ioctl to get report descriptor size
+ */
+TEST_F(hidraw, ioctl_rdescsize)
+{
+ int desc_size = 0;
+ int err;
+
+ /* call HIDIOCGRDESCSIZE ioctl */
+ err = ioctl(self->hidraw_fd, HIDIOCGRDESCSIZE, &desc_size);
+ ASSERT_EQ(err, 0) TH_LOG("HIDIOCGRDESCSIZE ioctl failed");
+
+ /* verify the size matches our test report descriptor */
+ ASSERT_EQ(desc_size, sizeof(rdesc))
+ TH_LOG("expected size %zu, got %d", sizeof(rdesc), desc_size);
+}
+
+/*
+ * Test HIDIOCGRDESC ioctl to get report descriptor data
+ */
+TEST_F(hidraw, ioctl_rdesc)
+{
+ struct hidraw_report_descriptor desc;
+ int err;
+
+ /* get the full report descriptor */
+ desc.size = sizeof(rdesc);
+ err = ioctl(self->hidraw_fd, HIDIOCGRDESC, &desc);
+ ASSERT_EQ(err, 0) TH_LOG("HIDIOCGRDESC ioctl failed");
+
+ /* verify the descriptor data matches our test descriptor */
+ ASSERT_EQ(memcmp(desc.value, rdesc, sizeof(rdesc)), 0)
+ TH_LOG("report descriptor data mismatch");
+}
+
+/*
+ * Test HIDIOCGRDESC ioctl with smaller buffer size
+ */
+TEST_F(hidraw, ioctl_rdesc_small_buffer)
+{
+ struct hidraw_report_descriptor desc;
+ int err;
+ size_t small_size = sizeof(rdesc) / 2; /* request half the descriptor size */
+
+ /* get partial report descriptor */
+ desc.size = small_size;
+ err = ioctl(self->hidraw_fd, HIDIOCGRDESC, &desc);
+ ASSERT_EQ(err, 0) TH_LOG("HIDIOCGRDESC ioctl failed with small buffer");
+
+ /* verify we got the first part of the descriptor */
+ ASSERT_EQ(memcmp(desc.value, rdesc, small_size), 0)
+ TH_LOG("partial report descriptor data mismatch");
+}
+
+/*
+ * Test HIDIOCGRAWINFO ioctl to get device information
+ */
+TEST_F(hidraw, ioctl_rawinfo)
+{
+ struct hidraw_devinfo devinfo;
+ int err;
+
+ /* get device info */
+ err = ioctl(self->hidraw_fd, HIDIOCGRAWINFO, &devinfo);
+ ASSERT_EQ(err, 0) TH_LOG("HIDIOCGRAWINFO ioctl failed");
+
+ /* verify device info matches our test setup */
+ ASSERT_EQ(devinfo.bustype, BUS_USB)
+ TH_LOG("expected bustype 0x03, got 0x%x", devinfo.bustype);
+ ASSERT_EQ(devinfo.vendor, 0x0001)
+ TH_LOG("expected vendor 0x0001, got 0x%x", devinfo.vendor);
+ ASSERT_EQ(devinfo.product, 0x0a37)
+ TH_LOG("expected product 0x0a37, got 0x%x", devinfo.product);
+}
+
+/*
+ * Test HIDIOCGFEATURE ioctl to get feature report
+ */
+TEST_F(hidraw, ioctl_gfeature)
+{
+ __u8 buf[10] = {0};
+ int err;
+
+ /* set report ID 1 in first byte */
+ buf[0] = 1;
+
+ /* get feature report */
+ err = ioctl(self->hidraw_fd, HIDIOCGFEATURE(sizeof(buf)), buf);
+ ASSERT_EQ(err, sizeof(feature_data)) TH_LOG("HIDIOCGFEATURE ioctl failed, got %d", err);
+
+ /* verify we got the expected feature data */
+ ASSERT_EQ(buf[0], feature_data[0])
+ TH_LOG("expected feature_data[0] = %d, got %d", feature_data[0], buf[0]);
+ ASSERT_EQ(buf[1], feature_data[1])
+ TH_LOG("expected feature_data[1] = %d, got %d", feature_data[1], buf[1]);
+}
+
+/*
+ * Test HIDIOCGFEATURE ioctl with invalid report ID
+ */
+TEST_F(hidraw, ioctl_gfeature_invalid)
+{
+ __u8 buf[10] = {0};
+ int err;
+
+ /* set invalid report ID (not 1) */
+ buf[0] = 2;
+
+ /* try to get feature report */
+ err = ioctl(self->hidraw_fd, HIDIOCGFEATURE(sizeof(buf)), buf);
+ ASSERT_LT(err, 0) TH_LOG("HIDIOCGFEATURE should have failed with invalid report ID");
+ ASSERT_EQ(errno, EIO) TH_LOG("expected EIO, got errno %d", errno);
+}
+
+/*
+ * Test HIDIOCSFEATURE ioctl to set feature report
+ */
+TEST_F(hidraw, ioctl_sfeature)
+{
+ __u8 buf[10] = {0};
+ int err;
+
+ /* prepare feature report data */
+ buf[0] = 1; /* report ID */
+ buf[1] = 0x42;
+ buf[2] = 0x24;
+
+ /* set feature report */
+ err = ioctl(self->hidraw_fd, HIDIOCSFEATURE(3), buf);
+ ASSERT_EQ(err, 3) TH_LOG("HIDIOCSFEATURE ioctl failed, got %d", err);
+
+ /*
+ * Note: The uhid mock doesn't validate the set report data,
+ * so we just verify the ioctl succeeds
+ */
+}
+
+/*
+ * Test HIDIOCGINPUT ioctl to get input report
+ */
+TEST_F(hidraw, ioctl_ginput)
+{
+ __u8 buf[10] = {0};
+ int err;
+
+ /* set report ID 1 in first byte */
+ buf[0] = 1;
+
+ /* get input report */
+ err = ioctl(self->hidraw_fd, HIDIOCGINPUT(sizeof(buf)), buf);
+ ASSERT_EQ(err, sizeof(feature_data)) TH_LOG("HIDIOCGINPUT ioctl failed, got %d", err);
+
+ /* verify we got the expected input data */
+ ASSERT_EQ(buf[0], feature_data[0])
+ TH_LOG("expected feature_data[0] = %d, got %d", feature_data[0], buf[0]);
+ ASSERT_EQ(buf[1], feature_data[1])
+ TH_LOG("expected feature_data[1] = %d, got %d", feature_data[1], buf[1]);
+}
+
+/*
+ * Test HIDIOCGINPUT ioctl with invalid report ID
+ */
+TEST_F(hidraw, ioctl_ginput_invalid)
+{
+ __u8 buf[10] = {0};
+ int err;
+
+ /* set invalid report ID (not 1) */
+ buf[0] = 2;
+
+ /* try to get input report */
+ err = ioctl(self->hidraw_fd, HIDIOCGINPUT(sizeof(buf)), buf);
+ ASSERT_LT(err, 0) TH_LOG("HIDIOCGINPUT should have failed with invalid report ID");
+ ASSERT_EQ(errno, EIO) TH_LOG("expected EIO, got errno %d", errno);
+}
+
+/*
+ * Test HIDIOCSINPUT ioctl to set input report
+ */
+TEST_F(hidraw, ioctl_sinput)
+{
+ __u8 buf[10] = {0};
+ int err;
+
+ /* prepare input report data */
+ buf[0] = 1; /* report ID */
+ buf[1] = 0x55;
+ buf[2] = 0xAA;
+
+ /* set input report */
+ err = ioctl(self->hidraw_fd, HIDIOCSINPUT(3), buf);
+ ASSERT_EQ(err, 3) TH_LOG("HIDIOCSINPUT ioctl failed, got %d", err);
+
+ /*
+ * Note: The uhid mock doesn't validate the set report data,
+ * so we just verify the ioctl succeeds
+ */
+}
+
+/*
+ * Test HIDIOCGOUTPUT ioctl to get output report
+ */
+TEST_F(hidraw, ioctl_goutput)
+{
+ __u8 buf[10] = {0};
+ int err;
+
+ /* set report ID 1 in first byte */
+ buf[0] = 1;
+
+ /* get output report */
+ err = ioctl(self->hidraw_fd, HIDIOCGOUTPUT(sizeof(buf)), buf);
+ ASSERT_EQ(err, sizeof(feature_data)) TH_LOG("HIDIOCGOUTPUT ioctl failed, got %d", err);
+
+ /* verify we got the expected output data */
+ ASSERT_EQ(buf[0], feature_data[0])
+ TH_LOG("expected feature_data[0] = %d, got %d", feature_data[0], buf[0]);
+ ASSERT_EQ(buf[1], feature_data[1])
+ TH_LOG("expected feature_data[1] = %d, got %d", feature_data[1], buf[1]);
+}
+
+/*
+ * Test HIDIOCGOUTPUT ioctl with invalid report ID
+ */
+TEST_F(hidraw, ioctl_goutput_invalid)
+{
+ __u8 buf[10] = {0};
+ int err;
+
+ /* set invalid report ID (not 1) */
+ buf[0] = 2;
+
+ /* try to get output report */
+ err = ioctl(self->hidraw_fd, HIDIOCGOUTPUT(sizeof(buf)), buf);
+ ASSERT_LT(err, 0) TH_LOG("HIDIOCGOUTPUT should have failed with invalid report ID");
+ ASSERT_EQ(errno, EIO) TH_LOG("expected EIO, got errno %d", errno);
+}
+
+/*
+ * Test HIDIOCSOUTPUT ioctl to set output report
+ */
+TEST_F(hidraw, ioctl_soutput)
+{
+ __u8 buf[10] = {0};
+ int err;
+
+ /* prepare output report data */
+ buf[0] = 1; /* report ID */
+ buf[1] = 0x33;
+ buf[2] = 0xCC;
+
+ /* set output report */
+ err = ioctl(self->hidraw_fd, HIDIOCSOUTPUT(3), buf);
+ ASSERT_EQ(err, 3) TH_LOG("HIDIOCSOUTPUT ioctl failed, got %d", err);
+
+ /*
+ * Note: The uhid mock doesn't validate the set report data,
+ * so we just verify the ioctl succeeds
+ */
+}
+
+/*
+ * Test HIDIOCGRAWNAME ioctl to get device name string
+ */
+TEST_F(hidraw, ioctl_rawname)
+{
+ char name[256] = {0};
+ char expected_name[64];
+ int err;
+
+ /* get device name */
+ err = ioctl(self->hidraw_fd, HIDIOCGRAWNAME(sizeof(name)), name);
+ ASSERT_GT(err, 0) TH_LOG("HIDIOCGRAWNAME ioctl failed, got %d", err);
+
+ /* construct expected name based on device id */
+ snprintf(expected_name, sizeof(expected_name), "test-uhid-device-%d", self->hid.dev_id);
+
+ /* verify the name matches expected pattern */
+ ASSERT_EQ(strcmp(name, expected_name), 0)
+ TH_LOG("expected name '%s', got '%s'", expected_name, name);
+}
+
+/*
+ * Test HIDIOCGRAWPHYS ioctl to get device physical address string
+ */
+TEST_F(hidraw, ioctl_rawphys)
+{
+ char phys[256] = {0};
+ char expected_phys[64];
+ int err;
+
+ /* get device physical address */
+ err = ioctl(self->hidraw_fd, HIDIOCGRAWPHYS(sizeof(phys)), phys);
+ ASSERT_GT(err, 0) TH_LOG("HIDIOCGRAWPHYS ioctl failed, got %d", err);
+
+ /* construct expected phys based on device id */
+ snprintf(expected_phys, sizeof(expected_phys), "%d", self->hid.dev_id);
+
+ /* verify the phys matches expected value */
+ ASSERT_EQ(strcmp(phys, expected_phys), 0)
+ TH_LOG("expected phys '%s', got '%s'", expected_phys, phys);
+}
+
+/*
+ * Test HIDIOCGRAWUNIQ ioctl to get device unique identifier string
+ */
+TEST_F(hidraw, ioctl_rawuniq)
+{
+ char uniq[256] = {0};
+ int err;
+
+ /* get device unique identifier */
+ err = ioctl(self->hidraw_fd, HIDIOCGRAWUNIQ(sizeof(uniq)), uniq);
+ ASSERT_GE(err, 0) TH_LOG("HIDIOCGRAWUNIQ ioctl failed, got %d", err);
+
+ /* uniq is typically empty in our test setup */
+ ASSERT_EQ(strlen(uniq), 0) TH_LOG("expected empty uniq, got '%s'", uniq);
+}
+
+/*
+ * Test device string ioctls with small buffer sizes
+ */
+TEST_F(hidraw, ioctl_strings_small_buffer)
+{
+ char small_buf[8] = {0};
+ char expected_name[64];
+ int err;
+
+ /* test HIDIOCGRAWNAME with small buffer */
+ err = ioctl(self->hidraw_fd, HIDIOCGRAWNAME(sizeof(small_buf)), small_buf);
+ ASSERT_EQ(err, sizeof(small_buf))
+ TH_LOG("HIDIOCGRAWNAME with small buffer failed, got %d", err);
+
+ /* construct expected truncated name */
+ snprintf(expected_name, sizeof(expected_name), "test-uhid-device-%d", self->hid.dev_id);
+
+ /* verify we got truncated name (first 8 chars, no null terminator guaranteed) */
+ ASSERT_EQ(strncmp(small_buf, expected_name, sizeof(small_buf)), 0)
+ TH_LOG("expected truncated name to match first %zu chars", sizeof(small_buf));
+
+ /* Note: hidraw driver doesn't guarantee null termination when buffer is too small */
+}
+
int main(int argc, char **argv)
{
return test_harness_run(argc, argv);
--
2.51.0
^ permalink raw reply related
* [PATCH v3 0/3] HID: hidraw: rework ioctls
From: Benjamin Tissoires @ 2025-09-12 16:58 UTC (permalink / raw)
To: Jiri Kosina, Shuah Khan, Arnd Bergmann
Cc: linux-input, linux-kselftest, linux-kernel, Benjamin Tissoires,
Arnd Bergmann
Arnd sent the v1 of the series in July, and it was bogus. So with a
little help from claude-sonnet I built up the missing ioctls tests and
tried to figure out a way to apply Arnd's logic without breaking the
existing ioctls.
The end result is in patch 3/3, which makes use of subfunctions to keep
the main ioctl code path clean.
Signed-off-by: Benjamin Tissoires <bentiss@kernel.org>
---
Changes in v3:
- dropped the co-developed-by tag and put a blurb instead
- change the attribution of patch 3/3 to me as requested by Arnd.
- Link to v2: https://lore.kernel.org/r/20250826-b4-hidraw-ioctls-v2-0-c7726b236719@kernel.org
changes in v2:
- add new hidraw ioctls tests
- refactor Arnd's patch to keep the existing error path logic
- link to v1: https://lore.kernel.org/linux-input/20250711072847.2836962-1-arnd@kernel.org/
---
Benjamin Tissoires (3):
selftests/hid: hidraw: add more coverage for hidraw ioctls
selftests/hid: hidraw: forge wrong ioctls and tests them
HID: hidraw: tighten ioctl command parsing
drivers/hid/hidraw.c | 224 ++++++++-------
include/uapi/linux/hidraw.h | 2 +
tools/testing/selftests/hid/hid_common.h | 6 +
tools/testing/selftests/hid/hidraw.c | 473 +++++++++++++++++++++++++++++++
4 files changed, 603 insertions(+), 102 deletions(-)
---
base-commit: 02d6eeedbc36d4b309d5518778071a749ef79c4e
change-id: 20250825-b4-hidraw-ioctls-66f34297032a
Best regards,
--
Benjamin Tissoires <bentiss@kernel.org>
^ permalink raw reply
* Re: [PATCH v2 00/11] HID: playstation: Add support for audio jack handling on DualSense
From: Benjamin Tissoires @ 2025-09-12 16:14 UTC (permalink / raw)
To: Jiri Kosina
Cc: Cristian Ciocaltea, Roderick Colenbrander, Roderick Colenbrander,
Henrik Rydberg, kernel, linux-input, linux-kernel
In-Reply-To: <r8qr0nrn-0n5r-6r96-7p26-q22ns73484np@xreary.bet>
On Sep 12 2025, Jiri Kosina wrote:
> On Wed, 27 Aug 2025, Cristian Ciocaltea wrote:
>
> > It's been over a month now since this was kind of blocked without any clear
> > reason, and by the end of next week I'll be on leave, which means we're
> > close to missing the merge window once again.
> >
> > Considering the counterpart quirk in the generic USB audio driver has been
> > already merged since v6.17, I kindly ask for your support in getting this
> > into v6.18.
>
> Roderick, do you have any word on this, please?
If this can help moving forward:
Patches 1-9 are:
Reviewed-by: Benjamin Tissoires <bentiss@kernel.org>
and can be merged right away
Cheers,
Benjamin
>
> --
> Jiri Kosina
> SUSE Labs
>
^ permalink raw reply
* Re: [PATCH v2 00/11] HID: playstation: Add support for audio jack handling on DualSense
From: Jiri Kosina @ 2025-09-12 15:38 UTC (permalink / raw)
To: Cristian Ciocaltea
Cc: Roderick Colenbrander, Roderick Colenbrander, Benjamin Tissoires,
Henrik Rydberg, kernel, linux-input, linux-kernel
In-Reply-To: <c6a16e71-e431-47dd-a3d1-6a79fd7e4a37@collabora.com>
On Wed, 27 Aug 2025, Cristian Ciocaltea wrote:
> It's been over a month now since this was kind of blocked without any clear
> reason, and by the end of next week I'll be on leave, which means we're
> close to missing the merge window once again.
>
> Considering the counterpart quirk in the generic USB audio driver has been
> already merged since v6.17, I kindly ask for your support in getting this
> into v6.18.
Roderick, do you have any word on this, please?
--
Jiri Kosina
SUSE Labs
^ permalink raw reply
* Re: [PATCH] HID: pidff: Fix possible null pointer dereference
From: Jiri Kosina @ 2025-09-12 15:29 UTC (permalink / raw)
To: Tomasz Pakuła; +Cc: bentiss, oleg, linux-input, dan.carpenter
In-Reply-To: <20250823194516.2425885-1-tomasz.pakula.oficjalny@gmail.com>
On Sat, 23 Aug 2025, Tomasz Pakuła wrote:
> As reported by Dan Carpenter, if the axes_enable field wasn't found,
> trying to find the axes themselves will result in a null pointer
> dereference. This could only occur with a broken PID descriptor, but
> it's worth protecting from.
>
> Exit early if the axes_enable wasn't found AND add a gate to the
> pidff_find_special_keys to exit early if the passed HID field is null.
> This will protect again null dereferencing in the future and properly
> return 0 found special keys.
>
> Fixes:
> Commit 1d72e7bd340b ("HID: pidff: Add support for AXES_ENABLE field")
> Made on top of for-6.18/pidff
>
> Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
> Signed-off-by: Tomasz Pakuła <tomasz.pakula.oficjalny@gmail.com>
Applied to hid.git#for-6.18/pidff, thanks.
--
Jiri Kosina
SUSE Labs
^ permalink raw reply
* Re: [PATCH v3 00/11] HID: Implement haptic touchpad support
From: Jiri Kosina @ 2025-09-12 15:28 UTC (permalink / raw)
To: Dmitry Torokhov
Cc: Jonathan Denose, Jonathan Corbet, Henrik Rydberg, linux-input,
linux-kernel, linux-doc, Angela Czubak, Sean O'Brien,
Benjamin Tissoires
In-Reply-To: <vyhhm3x6nfdfw6gbgluq3sjr6bzamhear7nec6xdi5wfxq7wcz@cx2egj4yr5sp>
On Thu, 4 Sep 2025, Benjamin Tissoires wrote:
> > > Angela Czubak (11):
> > > HID: add haptics page defines
> > > Input: add FF_HAPTIC effect type
> > > Input: add INPUT_PROP_HAPTIC_TOUCHPAD
> > > HID: haptic: introduce hid_haptic_device
> > > HID: input: allow mapping of haptic output
> > > HID: haptic: initialize haptic device
> > > HID: input: calculate resolution for pressure
> > > HID: haptic: add functions handling events
> > > Input: MT - add INPUT_MT_TOTAL_FORCE flags
> > > HID: haptic: add hid_haptic_switch_mode
> > > HID: multitouch: add haptic multitouch support
> > >
> > > Documentation/input/event-codes.rst | 14 +
> > > drivers/hid/Kconfig | 11 +
> > > drivers/hid/Makefile | 1 +
> > > drivers/hid/hid-haptic.c | 580 +++++++++++++++++++++++++++++++++
> > > drivers/hid/hid-haptic.h | 127 ++++++++
> > > drivers/hid/hid-input.c | 18 +-
> > > drivers/hid/hid-multitouch.c | 47 +++
> > > drivers/input/input-mt.c | 14 +-
> > > include/linux/hid.h | 29 ++
> > > include/linux/input/mt.h | 1 +
> > > include/uapi/linux/input-event-codes.h | 1 +
> > > include/uapi/linux/input.h | 22 +-
> > > 12 files changed, 858 insertions(+), 7 deletions(-)
> > > ---
> > > base-commit: 86731a2a651e58953fc949573895f2fa6d456841
> > > change-id: 20250625-support-forcepads-0b4f74fd3d0a
> > >
> > > Best regards,
> > > --
> > > Jonathan Denose <jdenose@google.com>
> > >
> > Hi all,
> >
> > Please let me know if there is anything else needed from me.
> >
>
> Dmitry, I've just re-reviewed and tested this series. I'm fine with it.
> Can you give us your ack on the input bits?
Dmitry, did you have time to review the input bits, please?
Thanks,
--
Jiri Kosina
SUSE Labs
^ permalink raw reply
* Re: [PATCH] HID: intel-ish-ipc: Remove redundant ready check after timeout function
From: Jiri Kosina @ 2025-09-12 15:23 UTC (permalink / raw)
To: Zhang Lixu; +Cc: linux-input, srinivas.pandruvada, benjamin.tissoires
In-Reply-To: <20250821020609.1947237-1-lixu.zhang@intel.com>
On Thu, 21 Aug 2025, Zhang Lixu wrote:
> timed_wait_for_timeout() internally checks for ish_is_input_ready() and
> ishtp_fw_is_ready() based on the provided parameters. If
> timed_wait_for_timeout() returns 0, it indicates the status is ready. In
> rare cases, another thread may send a message immediately after
> timed_wait_for_timeout() returns, causing a subsequent ish_is_input_ready()
> check to fail. Since the return value of timed_wait_for_timeout() is
> sufficient to determine readiness, the additional ready check is
> unnecessary and may introduce issues.
>
> This patch removes the redundant check and relies solely on the return
> value of timed_wait_for_timeout().
>
> Fixes: ae02e5d40d5f ("HID: intel-ish-hid: ipc layer")
> Signed-off-by: Zhang Lixu <lixu.zhang@intel.com>
> Acked-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Applied, thanks.
--
Jiri Kosina
SUSE Labs
^ permalink raw reply
* Re: [PATCH] hid-asus: add support for missing PX series fn keys
From: Jiri Kosina @ 2025-09-12 15:21 UTC (permalink / raw)
To: Amit Chaudhari; +Cc: bentiss, linux-input
In-Reply-To: <20250819214919.78610-1-amitchaudhari@mac.com>
On Tue, 19 Aug 2025, Amit Chaudhari wrote:
> Add support for missing hotkey keycodes affecting Asus PX13 and PX16
> families so userspace can use them.
Applied, thanks.
Next time, please try to follow the shortlog conventions of the subsystem,
prefixing it with 'HID: '.
I've fixed that now manually before applying.
--
Jiri Kosina
SUSE Labs
^ permalink raw reply
* Re: [PATCH] HID: cp2112: fix setter callbacks return value
From: Jiri Kosina @ 2025-09-12 15:15 UTC (permalink / raw)
To: Sébastien Szymanski
Cc: Benjamin Tissoires, Linus Walleij, Bartosz Golaszewski,
Bartosz Golaszewski, linux-input, linux-kernel, linux-gpio
In-Reply-To: <20250904-hid-cp2112-fix-set-value-v1-1-17d2e26dc8c9@armadeus.com>
On Thu, 4 Sep 2025, Sébastien Szymanski wrote:
> Since commit 6485543488a6 ("HID: cp2112: use new line value setter
> callbacks"), setting a GPIO value always fails with error -EBADE.
>
> That's because the returned value by the setter callbacks is the
> returned value by the hid_hw_raw_request() function which is the number of
> bytes sent on success or a negative value on error. The function
> gpiochip_set() returns -EBADE if the setter callbacks return a value >
> 0.
>
> Fix this by making the setter callbacks return 0 on success or a negative
> value on error.
>
> While at it, use the returned value by cp2112_gpio_set_unlocked() in the
> direction_output callback.
>
> Fixes: 6485543488a6 ("HID: cp2112: use new line value setter callbacks")
> Signed-off-by: Sébastien Szymanski <sebastien.szymanski@armadeus.com>
Applied, thank you!
--
Jiri Kosina
SUSE Labs
^ permalink raw reply
* Re: [PATCH 0/2] HID: bpf: allow bpf to rebind a driver to hid-mutltiouch
From: Jiri Kosina @ 2025-09-12 15:12 UTC (permalink / raw)
To: Benjamin Tissoires; +Cc: Peter Hutterer, linux-input, linux-kernel
In-Reply-To: <20250821-bpf-rescan-v1-0-08f9e2bc01bb@kernel.org>
On Thu, 21 Aug 2025, Benjamin Tissoires wrote:
> This happened while Peter was trying to fix a Viewsonic device: the HID
> device sending multiotuch data through a proprietary collection was
> handled by hid-generic, and we don't have any way of attaching it to
> hid-multitouch because the pre-scanning wasn't able to see the Contact
> ID HID usage.
>
> After a little of back and forth, it turns out that the best solution is
> to re-scan the device when a report descriptor is changed from the BPF
> point of view.
>
> Signed-off-by: Benjamin Tissoires <bentiss@kernel.org>
> ---
> Benjamin Tissoires (2):
> HID: core: factor out hid_set_group()
> HID: bpf: rescan the device for the group after a load/unload
>
> drivers/hid/hid-core.c | 44 ++++++++++++++++++++++++++++++++++++--------
> 1 file changed, 36 insertions(+), 8 deletions(-)
Now in hid.git#for-6.18/core. Thanks,
--
Jiri Kosina
SUSE Labs
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox