Skip to contentSkip to conversation

HID: evision: Add preliminary support for EVision keyboards

PATCH Superseded Philippe Valembois 3 participants 8 messages Dec 24, 2022 – Feb 6, 2023 HID

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 HID: evision: Add preliminary support for EVision keyboards

From: Philippe Valembois <[email protected]>

For now only supports one model and only filters out bogus reports sent when the keyboard has been configured through hidraw. Without this, as events are not released, soft repeat floods userspace with unknown key events.

---
 drivers/hid/Kconfig       |  7 ++++
 drivers/hid/Makefile      |  1 +
 drivers/hid/hid-evision.c | 79 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 87 insertions(+)
 create mode 100644 drivers/hid/hid-evision.c
drivers/hid/Kconfig +7 -0
index e2a5d30c8..1320ea75c 100644
@@ -329,6 +329,13 @@ config HID_ELO
 	Support for the ELO USB 4000/4500 touchscreens. Note that this is for
 	different devices than those handled by CONFIG_TOUCHSCREEN_USB_ELO.
 
+config HID_EVISION
+	tristate "EVision Keyboards Support"
+	depends on USB_HID
+	help
+	Support for some EVision keyboards. Note that this is needed only when
+	applying customization using userspace programs.
+
 config HID_EZKEY
 	tristate "Ezkey BTC 8193 keyboard"
 	default !EXPERT
drivers/hid/Makefile +1 -0
index e8014c1a2..bd01571dd 100644
@@ -45,6 +45,7 @@ obj-$(CONFIG_HID_EMS_FF)	+= hid-emsff.o
 obj-$(CONFIG_HID_ELAN)		+= hid-elan.o
 obj-$(CONFIG_HID_ELECOM)	+= hid-elecom.o
 obj-$(CONFIG_HID_ELO)		+= hid-elo.o
+obj-$(CONFIG_HID_EVISION)	+= hid-evision.o
 obj-$(CONFIG_HID_EZKEY)		+= hid-ezkey.o
 obj-$(CONFIG_HID_FT260)		+= hid-ft260.o
 obj-$(CONFIG_HID_GEMBIRD)	+= hid-gembird.o
drivers/hid/hid-evision.c new file +79 -0

79 lines not on this page · raw message

Signature · 1 line
2.38.2
#1 lore raw
Quoted 2 lines · Philippe Valembois

From: Philippe Valembois <[email protected]>

Jiri, I have a doubt. Do we accept emails from users.noreply.github.com?

Quoted 19 lines

For now only supports one model and only filters out bogus reports sent when the keyboard has been configured through hidraw. Without this, as events are not released, soft repeat floods userspace with unknown key events.

---
 drivers/hid/Kconfig       |  7 ++++
 drivers/hid/Makefile      |  1 +
 drivers/hid/hid-evision.c | 79 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 87 insertions(+)
 create mode 100644 drivers/hid/hid-evision.c
drivers/hid/Kconfig +0 -0
index e2a5d30c8..1320ea75c 100644

drivers/hid/Kconfig config HID_ELO · line 334

Quoted 6 lines
drivers/hid/Kconfig +3 -0
@@ -329,6 +329,13 @@ config HID_ELO
        Support for the ELO USB 4000/4500 touchscreens. Note that this is for
        different devices than those handled by CONFIG_TOUCHSCREEN_USB_ELO.
 
+config HID_EVISION
+       tristate "EVision Keyboards Support"
+       depends on USB_HID

AFAICT, the driver only uses pure HID API, so you should be able to depend on HID, not just USB_HID.

Quoted 29 lines
drivers/hid/Kconfig +4 -0
+       help
+       Support for some EVision keyboards. Note that this is needed only when
+       applying customization using userspace programs.
+
 config HID_EZKEY
        tristate "Ezkey BTC 8193 keyboard"
        default !EXPERT
drivers/hid/Makefile +1 -0
index e8014c1a2..bd01571dd 100644
@@ -45,6 +45,7 @@ obj-$(CONFIG_HID_EMS_FF)      += hid-emsff.o
 obj-$(CONFIG_HID_ELAN)         += hid-elan.o
 obj-$(CONFIG_HID_ELECOM)       += hid-elecom.o
 obj-$(CONFIG_HID_ELO)          += hid-elo.o
+obj-$(CONFIG_HID_EVISION)      += hid-evision.o
 obj-$(CONFIG_HID_EZKEY)                += hid-ezkey.o
 obj-$(CONFIG_HID_FT260)                += hid-ft260.o
 obj-$(CONFIG_HID_GEMBIRD)      += hid-gembird.o
drivers/hid/hid-evision.c new file +4 -0
new file mode 100644
index 000000000..6ea331575
@@ -0,0 +1,79 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ *  HID driver for EVision devices
+ *  For now, only ignore bogus consumer reports

drivers/hid/hid-evision.c line 14

Quoted 10 lines
drivers/hid/hid-evision.c new file +10 -0
+ *  sent after the keyboard has been configured
+ *
+ *  Copyright (c) 2022 Philippe Valembois
+ */
+
+#include <linux/device.h>
+#include <linux/input.h>
+#include <linux/hid.h>
+#include <linux/module.h>
+#include <linux/usb.h>

Outside of hid_is_usb(), you are not using anything USB related, so this can be dropped

drivers/hid/hid-evision.c line 18

Quoted 4 lines
drivers/hid/hid-evision.c new file +4 -0
+
+
+#define USB_VENDOR_ID_EVISION       0x320f
+#define USB_DEVICE_ID_EVISION_ICL01 0x5041

We tend to add those variables in drivers/hid/hid-ids.h

drivers/hid/hid-evision.c line 27

Quoted 9 lines
drivers/hid/hid-evision.c new file +9 -0
+
+static int evision_input_mapping(struct hid_device *hdev, struct hid_input *hi,
+               struct hid_field *field, struct hid_usage *usage,
+               unsigned long **bit, int *max)
+{
+       if ((usage->hid & HID_USAGE_PAGE) != HID_UP_CONSUMER)
+               return 0;
+
+       // Ignore key down event

No C++ comments style please, use /* */ instead

drivers/hid/hid-evision.c line 30

Quoted 3 lines
drivers/hid/hid-evision.c new file +3 -0
+       if ((usage->hid & HID_USAGE) >> 8 == 0x05)
+               return -1;
+       // Ignore key up event

Same (and for any other // below).

Quoted 8 lines
drivers/hid/hid-evision.c new file +8 -0
+       if ((usage->hid & HID_USAGE) >> 8 == 0x06)
+               return -1;
+
+       switch (usage->hid & HID_USAGE) {
+       // Ignore configuration saved event
+       case 0x0401: return -1;
+       // Ignore reset event
+       case 0x0402: return -1;

drivers/hid/hid-evision.c line 48

Quoted 10 lines
drivers/hid/hid-evision.c new file +10 -0
+       }
+       return 0;
+}
+
+static int evision_probe(struct hid_device *hdev, const struct hid_device_id *id)
+{
+       int ret;
+
+       if (!hid_is_usb(hdev))
+               return -EINVAL;

This can be dropped...

Quoted 4 lines
drivers/hid/hid-evision.c new file +4 -0
+
+       ret = hid_parse(hdev);
+       if (ret) {
+               hid_err(hdev, "EVision hid parse failed: %d\n", ret);

drivers/hid/hid-evision.c line 62

Quoted 10 lines
drivers/hid/hid-evision.c new file +10 -0
+               return ret;
+       }
+
+       ret = hid_hw_start(hdev, HID_CONNECT_DEFAULT);
+       if (ret) {
+               hid_err(hdev, "EVision hw start failed: %d\n", ret);
+               return ret;
+       }
+
+       return 0;

... which means the probe is the default one, meaning it can also be dropped from the patch :)

drivers/hid/hid-evision.c line 75

Quoted 13 lines
drivers/hid/hid-evision.c new file +13 -0
+}
+
+static const struct hid_device_id evision_devices[] = {
+       { HID_USB_DEVICE(USB_VENDOR_ID_EVISION, USB_DEVICE_ID_EVISION_ICL01) },
+       { }
+};
+MODULE_DEVICE_TABLE(hid, evision_devices);
+
+static struct hid_driver evision_driver = {
+       .name = "evision",
+       .id_table = evision_devices,
+       .input_mapping = evision_input_mapping,
+       .probe = evision_probe,

Just for completeness, remove that .probe line and your driver will behave the same and be smaller :)

Quoted 7 lines
drivers/hid/hid-evision.c new file +4 -1
+};
+module_hid_driver(evision_driver);
+
+MODULE_LICENSE("GPL");
--

2.38.2

Cheers,
Benjamin

#2 lore raw

PATCH v2 1/1 HID: evision: Add preliminary support for EVision keyboards

From: Philippe Valembois <[email protected]>

For now only supports one model and only filters out bogus reports sent when the keyboard has been configured through hidraw. Without this, as events are not released, soft repeat floods userspace with unknown key events.

---
 drivers/hid/Kconfig       |  7 ++++++
 drivers/hid/Makefile      |  1 +
 drivers/hid/hid-evision.c | 53 +++++++++++++++++++++++++++++++++++++++
 drivers/hid/hid-ids.h     |  3 +++
 4 files changed, 64 insertions(+)
 create mode 100644 drivers/hid/hid-evision.c
drivers/hid/Kconfig +7 -0
index e2a5d30c8..35f214773 100644
@@ -329,6 +329,13 @@ config HID_ELO
 	Support for the ELO USB 4000/4500 touchscreens. Note that this is for
 	different devices than those handled by CONFIG_TOUCHSCREEN_USB_ELO.
 
+config HID_EVISION
+	tristate "EVision Keyboards Support"
+	depends on HID
+	help
+	Support for some EVision keyboards. Note that this is needed only when
+	applying customization using userspace programs.
+
 config HID_EZKEY
 	tristate "Ezkey BTC 8193 keyboard"
 	default !EXPERT
drivers/hid/Makefile +1 -0
index e8014c1a2..bd01571dd 100644
@@ -45,6 +45,7 @@ obj-$(CONFIG_HID_EMS_FF)	+= hid-emsff.o
 obj-$(CONFIG_HID_ELAN)		+= hid-elan.o
 obj-$(CONFIG_HID_ELECOM)	+= hid-elecom.o
 obj-$(CONFIG_HID_ELO)		+= hid-elo.o
+obj-$(CONFIG_HID_EVISION)	+= hid-evision.o
 obj-$(CONFIG_HID_EZKEY)		+= hid-ezkey.o
 obj-$(CONFIG_HID_FT260)		+= hid-ft260.o
 obj-$(CONFIG_HID_GEMBIRD)	+= hid-gembird.o
drivers/hid/hid-evision.c new file +53 -0

53 lines not on this page · raw message

drivers/hid/hid-ids.h +3 -0
index 0f8c11842..7ae89aebd 100644
@@ -445,6 +445,9 @@
 #define USB_VENDOR_ID_EMS		0x2006
 #define USB_DEVICE_ID_EMS_TRIO_LINKER_PLUS_II 0x0118
 
+#define USB_VENDOR_ID_EVISION           0x320f
+#define USB_DEVICE_ID_EVISION_ICL01     0x5041
+
 #define USB_VENDOR_ID_FLATFROG		0x25b5
 #define USB_DEVICE_ID_MULTITOUCH_3200	0x0002
 
Signature · 1 line
2.39.1
#3 lore raw

PATCH v3 1/1 HID: evision: Add preliminary support for EVision keyboards

For now only supports one model and only filters out bogus reports sent when the keyboard has been configured through hidraw. Without this, as events are not released, soft repeat floods userspace with unknown key events.

---
 drivers/hid/Kconfig       |  7 ++++++
 drivers/hid/Makefile      |  1 +
 drivers/hid/hid-evision.c | 53 +++++++++++++++++++++++++++++++++++++++
 drivers/hid/hid-ids.h     |  3 +++
 4 files changed, 64 insertions(+)
 create mode 100644 drivers/hid/hid-evision.c
drivers/hid/Kconfig +7 -0
index e2a5d30c8..35f214773 100644
@@ -329,6 +329,13 @@ config HID_ELO
 	Support for the ELO USB 4000/4500 touchscreens. Note that this is for
 	different devices than those handled by CONFIG_TOUCHSCREEN_USB_ELO.
 
+config HID_EVISION
+	tristate "EVision Keyboards Support"
+	depends on HID
+	help
+	Support for some EVision keyboards. Note that this is needed only when
+	applying customization using userspace programs.
+
 config HID_EZKEY
 	tristate "Ezkey BTC 8193 keyboard"
 	default !EXPERT
drivers/hid/Makefile +1 -0
index e8014c1a2..bd01571dd 100644
@@ -45,6 +45,7 @@ obj-$(CONFIG_HID_EMS_FF)	+= hid-emsff.o
 obj-$(CONFIG_HID_ELAN)		+= hid-elan.o
 obj-$(CONFIG_HID_ELECOM)	+= hid-elecom.o
 obj-$(CONFIG_HID_ELO)		+= hid-elo.o
+obj-$(CONFIG_HID_EVISION)	+= hid-evision.o
 obj-$(CONFIG_HID_EZKEY)		+= hid-ezkey.o
 obj-$(CONFIG_HID_FT260)		+= hid-ft260.o
 obj-$(CONFIG_HID_GEMBIRD)	+= hid-gembird.o
drivers/hid/hid-evision.c new file +53 -0

53 lines not on this page · raw message

drivers/hid/hid-ids.h +3 -0
index 0f8c11842..7ae89aebd 100644
@@ -445,6 +445,9 @@
 #define USB_VENDOR_ID_EMS		0x2006
 #define USB_DEVICE_ID_EMS_TRIO_LINKER_PLUS_II 0x0118
 
+#define USB_VENDOR_ID_EVISION           0x320f
+#define USB_DEVICE_ID_EVISION_ICL01     0x5041
+
 #define USB_VENDOR_ID_FLATFROG		0x25b5
 #define USB_DEVICE_ID_MULTITOUCH_3200	0x0002
 
Signature · 1 line
2.39.1
#4 lore raw
Quoted 6 lines · Philippe Valembois

For now only supports one model and only filters out bogus reports sent when the keyboard has been configured through hidraw. Without this, as events are not released, soft repeat floods userspace with unknown key events.

Applied to https://git.kernel.org/pub/scm/linux/kernel/git/hid/hid.git (for-6.3/evision), thanks!

[1/1] HID: evision: Add preliminary support for EVision keyboards

      https://git.kernel.org/hid/hid/c/f5cd71cfdb5c

Cheers,

Signature · 1 line
Benjamin Tissoires <[email protected]>
#5 lore raw
In reply to Benjamin Tissoires

Hello,

Thanks for the review.
I just sent an updated patch taking your comments into account.

Regards,
Philippe Valembois

Quoted 172 lines · Benjamin Tissoires
Quoted 2 lines · Philippe Valembois

From: Philippe Valembois <[email protected]>

Jiri, I have a doubt. Do we accept emails from users.noreply.github.com?

Quoted 19 lines

For now only supports one model and only filters out bogus reports sent when the keyboard has been configured through hidraw. Without this, as events are not released, soft repeat floods userspace with unknown key events.

---
 drivers/hid/Kconfig       |  7 ++++
 drivers/hid/Makefile      |  1 +
 drivers/hid/hid-evision.c | 79 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 87 insertions(+)
 create mode 100644 drivers/hid/hid-evision.c
drivers/hid/Kconfig +0 -0
index e2a5d30c8..1320ea75c 100644

drivers/hid/Kconfig config HID_ELO · line 334

Quoted 6 lines
drivers/hid/Kconfig +3 -0
@@ -329,6 +329,13 @@ config HID_ELO
        Support for the ELO USB 4000/4500 touchscreens. Note that this is for
        different devices than those handled by CONFIG_TOUCHSCREEN_USB_ELO.
 
+config HID_EVISION
+       tristate "EVision Keyboards Support"
+       depends on USB_HID

AFAICT, the driver only uses pure HID API, so you should be able to depend on HID, not just USB_HID.

Quoted 29 lines
drivers/hid/Kconfig +4 -0
+       help
+       Support for some EVision keyboards. Note that this is needed only when
+       applying customization using userspace programs.
+
 config HID_EZKEY
        tristate "Ezkey BTC 8193 keyboard"
        default !EXPERT
drivers/hid/Makefile +1 -0
index e8014c1a2..bd01571dd 100644
@@ -45,6 +45,7 @@ obj-$(CONFIG_HID_EMS_FF)      += hid-emsff.o
 obj-$(CONFIG_HID_ELAN)         += hid-elan.o
 obj-$(CONFIG_HID_ELECOM)       += hid-elecom.o
 obj-$(CONFIG_HID_ELO)          += hid-elo.o
+obj-$(CONFIG_HID_EVISION)      += hid-evision.o
 obj-$(CONFIG_HID_EZKEY)                += hid-ezkey.o
 obj-$(CONFIG_HID_FT260)                += hid-ft260.o
 obj-$(CONFIG_HID_GEMBIRD)      += hid-gembird.o
drivers/hid/hid-evision.c new file +4 -0
new file mode 100644
index 000000000..6ea331575
@@ -0,0 +1,79 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ *  HID driver for EVision devices
+ *  For now, only ignore bogus consumer reports

drivers/hid/hid-evision.c line 14

Quoted 10 lines
drivers/hid/hid-evision.c new file +10 -0
+ *  sent after the keyboard has been configured
+ *
+ *  Copyright (c) 2022 Philippe Valembois
+ */
+
+#include <linux/device.h>
+#include <linux/input.h>
+#include <linux/hid.h>
+#include <linux/module.h>
+#include <linux/usb.h>

Outside of hid_is_usb(), you are not using anything USB related, so this can be dropped

drivers/hid/hid-evision.c line 18

Quoted 4 lines
drivers/hid/hid-evision.c new file +4 -0
+
+
+#define USB_VENDOR_ID_EVISION       0x320f
+#define USB_DEVICE_ID_EVISION_ICL01 0x5041

We tend to add those variables in drivers/hid/hid-ids.h

drivers/hid/hid-evision.c line 27

Quoted 9 lines
drivers/hid/hid-evision.c new file +9 -0
+
+static int evision_input_mapping(struct hid_device *hdev, struct hid_input *hi,
+               struct hid_field *field, struct hid_usage *usage,
+               unsigned long **bit, int *max)
+{
+       if ((usage->hid & HID_USAGE_PAGE) != HID_UP_CONSUMER)
+               return 0;
+
+       // Ignore key down event

No C++ comments style please, use /* */ instead

drivers/hid/hid-evision.c line 30

Quoted 3 lines
drivers/hid/hid-evision.c new file +3 -0
+       if ((usage->hid & HID_USAGE) >> 8 == 0x05)
+               return -1;
+       // Ignore key up event

Same (and for any other // below).

Quoted 8 lines
drivers/hid/hid-evision.c new file +8 -0
+       if ((usage->hid & HID_USAGE) >> 8 == 0x06)
+               return -1;
+
+       switch (usage->hid & HID_USAGE) {
+       // Ignore configuration saved event
+       case 0x0401: return -1;
+       // Ignore reset event
+       case 0x0402: return -1;

drivers/hid/hid-evision.c line 48

Quoted 10 lines
drivers/hid/hid-evision.c new file +10 -0
+       }
+       return 0;
+}
+
+static int evision_probe(struct hid_device *hdev, const struct hid_device_id *id)
+{
+       int ret;
+
+       if (!hid_is_usb(hdev))
+               return -EINVAL;

This can be dropped...

Quoted 4 lines
drivers/hid/hid-evision.c new file +4 -0
+
+       ret = hid_parse(hdev);
+       if (ret) {
+               hid_err(hdev, "EVision hid parse failed: %d\n", ret);

drivers/hid/hid-evision.c line 62

Quoted 10 lines
drivers/hid/hid-evision.c new file +10 -0
+               return ret;
+       }
+
+       ret = hid_hw_start(hdev, HID_CONNECT_DEFAULT);
+       if (ret) {
+               hid_err(hdev, "EVision hw start failed: %d\n", ret);
+               return ret;
+       }
+
+       return 0;

... which means the probe is the default one, meaning it can also be dropped from the patch :)

drivers/hid/hid-evision.c line 75

Quoted 13 lines
drivers/hid/hid-evision.c new file +13 -0
+}
+
+static const struct hid_device_id evision_devices[] = {
+       { HID_USB_DEVICE(USB_VENDOR_ID_EVISION, USB_DEVICE_ID_EVISION_ICL01) },
+       { }
+};
+MODULE_DEVICE_TABLE(hid, evision_devices);
+
+static struct hid_driver evision_driver = {
+       .name = "evision",
+       .id_table = evision_devices,
+       .input_mapping = evision_input_mapping,
+       .probe = evision_probe,

Just for completeness, remove that .probe line and your driver will behave the same and be smaller :)

Quoted 7 lines
drivers/hid/hid-evision.c new file +4 -1
+};
+module_hid_driver(evision_driver);
+
+MODULE_LICENSE("GPL");
--

2.38.2

Cheers,
Benjamin

maintainer
#6 lore raw
In reply to Benjamin Tissoires
Quoted 3 lines · Benjamin Tissoires
Quoted 1 line

From: Philippe Valembois <[email protected]>

Jiri, I have a doubt. Do we accept emails from users.noreply.github.com?

From quick search in github documentation, this seems to be a blackhole email address, right?

If so, I'd much prefer to have a real e-mail address there, in order to make sure the author / signoff holders can be reached in the future if there are any concerns regarding the patch.

The e-mail address disclosure concern is moot as soon as the emaill with the patch hits LKML anyway.

Thanks,

Signature · 2 lines
Jiri Kosina
SUSE Labs
#7 lore raw

Hello,

It is a blackhole address indeed.
I just sent a v3 with my corrected email address.

Thanks,
Regards,
Philippe Valembois

Quoted 23 lines · Jiri Kosina
Quoted 3 lines · Benjamin Tissoires
Quoted 1 line

From: Philippe Valembois <[email protected]>

Jiri, I have a doubt. Do we accept emails from users.noreply.github.com?

From quick search in github documentation, this seems to be a blackhole email address, right?

If so, I'd much prefer to have a real e-mail address there, in order to make sure the author / signoff holders can be reached in the future if there are any concerns regarding the patch.

The e-mail address disclosure concern is moot as soon as the emaill with the patch hits LKML anyway.

Thanks,

--
Jiri Kosina
SUSE Labs