2014-03-05 22:15:03

by Petri Gynther

Subject: [PATCH] input: Control HID encryption from input.conf

Add support for optional config item "Encryption=<true|false>" in input.conf.
Allows HID encryption to be disabled for keyboards and keyboard-like devices
that don't work properly with encryption due to firmware bugs.
---
profiles/input/device.c | 8 +++++++-
profiles/input/device.h | 1 +
profiles/input/manager.c | 21 ++++++++++++++++-----
3 files changed, 24 insertions(+), 6 deletions(-)

diff --git a/profiles/input/device.c b/profiles/input/device.c
index d6e97c7..e2f97dc 100644
--- a/profiles/input/device.c
+++ b/profiles/input/device.c
@@ -85,12 +85,18 @@ struct input_device {
};

static int idle_timeout = 0;
+static gboolean encryption_enabled = TRUE;

void input_set_idle_timeout(int timeout)
{
idle_timeout = timeout;
}

+void input_enable_encryption(gboolean state)
+{
+ encryption_enabled = state;
+}
+
static void input_device_enter_reconnect_mode(struct input_device *idev);

static void input_device_free(struct input_device *idev)
@@ -424,7 +430,7 @@ static int hidp_add_connection(struct input_device *idev)
strncpy(req->name, idev->name, sizeof(req->name) - 1);

/* Encryption is mandatory for keyboards */
- if (req->subclass & 0x40) {
+ if (encryption_enabled && (req->subclass & 0x40)) {
if (!bt_io_set(idev->intr_io, &gerr,
BT_IO_OPT_SEC_LEVEL, BT_IO_SEC_MEDIUM,
BT_IO_OPT_INVALID)) {
diff --git a/profiles/input/device.h b/profiles/input/device.h
index da2149c..af9f5ea 100644
--- a/profiles/input/device.h
+++ b/profiles/input/device.h
@@ -28,6 +28,7 @@ struct input_device;
struct input_conn;

void input_set_idle_timeout(int timeout);
+void input_enable_encryption(gboolean state);

int input_device_register(struct btd_service *service);
void input_device_unregister(struct btd_service *service);
diff --git a/profiles/input/manager.c b/profiles/input/manager.c
index 6ef83f4..1a70391 100644
--- a/profiles/input/manager.c
+++ b/profiles/input/manager.c
@@ -97,15 +97,26 @@ static int input_init(void)
config = load_config_file(CONFIGDIR "/input.conf");
if (config) {
int idle_timeout;
+ gboolean encryption;

idle_timeout = g_key_file_get_integer(config, "General",
- "IdleTimeout", &err);
- if (err) {
- DBG("input.conf: %s", err->message);
- g_error_free(err);
+ "IdleTimeout", &err);
+ if (!err) {
+ DBG("input.conf: IdleTimeout=%d", idle_timeout);
+ input_set_idle_timeout(idle_timeout * 60);
+ } else {
+ g_clear_error(&err);
}

- input_set_idle_timeout(idle_timeout * 60);
+ encryption = g_key_file_get_boolean(config, "General",
+ "Encryption", &err);
+ if (!err) {
+ DBG("input.conf: Encryption=%s", encryption ?
+ "true" : "false");
+ input_enable_encryption(encryption);
+ } else {
+ g_clear_error(&err);
+ }
}

btd_profile_register(&input_profile);
--
1.9.0.279.gdc9e3eb

2014-03-05 23:09:46

by Marcel Holtmann

Subject: Re: [PATCH] input: Control HID encryption from input.conf

Hi Petri,

> Add support for optional config item "Encryption=<true|false>" in input.conf.
> Allows HID encryption to be disabled for keyboards and keyboard-like devices
> that don't work properly with encryption due to firmware bugs.

really? I assumed the specification and qualification is pretty clear. Encryption is a must requirement for keyboard devices. Do you have traces where this causes issues.

I would also like to hear about samples of these broken keyboards. I have to admit that I have not personally used a Bluetooth keyboard in a while, but when developing the HID support back in the days, the encryption was never an issue. So I am surprised to hear that it has gotten worse and I hope this is not a bug on our side.

Regards

Marcel