Skip to contentSkip to conversation

device: Add ConnectionType property.

PATCH In review Your Name 3 participants 5 messages Sep 7, 2024 – Sep 19, 2024 Bluetooth drivers

lore
Share & tools
lore

Your Name

author
#0 lore raw

PATCH 1/2 device: Add ConnectionType property.

From: Vibhav Pant <[email protected]>

---
 src/device.c | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)
src/device.c +28 -0
index 0f18c8c7f..fa149f1d4 100644
@@ -1003,6 +1003,32 @@ static gboolean dev_property_exists_class(const GDBusPropertyTable *property,
 	return device->class != 0;
 }
 
+static gboolean
+dev_property_get_connection_type(const GDBusPropertyTable *property,
+				 DBusMessageIter *iter, void *data)
+{
+	struct btd_device *device = data;
+	const char *str;
+
+	if (device->le && device->bredr)
+		str = "dual";
+	else if (device->le)
+		str = "le";
+	else
+		str = "bredr";
+	dbus_message_iter_append_basic(iter, DBUS_TYPE_STRING, &str);
+	return TRUE;
+}
+
+static gboolean
+dev_property_exists_connection_type(const GDBusPropertyTable *property,
+				    void *data)
+{
+	struct btd_device *device = data;
+
+	return device->bredr || device->le;
+}
+
 static gboolean dev_property_get_class(const GDBusPropertyTable *property,
 					DBusMessageIter *iter, void *data)
 {
@@ -3234,6 +3260,8 @@ static const GDBusPropertyTable device_properties[] = {
 	{ "Alias", "s", dev_property_get_alias, dev_property_set_alias },
 	{ "Class", "u", dev_property_get_class, NULL,
 					dev_property_exists_class },
+	{ "ConnectionType", "s", dev_property_get_connection_type, NULL,
+					dev_property_exists_connection_type },
 	{ "Appearance", "q", dev_property_get_appearance, NULL,
 					dev_property_exists_appearance },
 	{ "Icon", "s", dev_property_get_icon, NULL,
Signature · 1 line
2.46.0
#1 lore raw

This is automated email and please do not reply to this email!

Dear submitter,

Thank you for submitting the patches to the linux bluetooth mailing list. This is a CI test results with your patch series:
PW Link:https://patchwork.kernel.org/project/bluetooth/list/?series=888046

---Test result---

Test Summary:

CheckPatch                    PASS      0.79 seconds
GitLint                       FAIL      0.75 seconds
BuildEll                      PASS      24.49 seconds
BluezMake                     PASS      1648.94 seconds
MakeCheck                     PASS      13.72 seconds
MakeDistcheck                 PASS      176.37 seconds
CheckValgrind                 PASS      250.72 seconds
CheckSmatch                   PASS      353.18 seconds
bluezmakeextell               PASS      119.16 seconds
IncrementalBuild              PASS      2921.19 seconds
ScanBuild                     PASS      982.22 seconds

Details
##############################
Test: GitLint - FAIL
Desc: Run gitlint
Output:
[BlueZ,1/2] device: Add ConnectionType property.

WARNING: I3 - ignore-body-lines: gitlint will be switching from using Python regex 'match' (match beginning) to 'search' (match anywhere) semantics. Please review your ignore-body-lines.regex option accordingly. To remove this warning, set general.regex-style-search=True. More details: https://jorisroovers.github.io/gitlint/configuration/#regex-style-search
1: T3 Title has trailing punctuation (.): "[BlueZ,1/2] device: Add ConnectionType property."
[BlueZ,2/2] org.bluez.Device: Add documentation for ConnectionType.

WARNING: I3 - ignore-body-lines: gitlint will be switching from using Python regex 'match' (match beginning) to 'search' (match anywhere) semantics. Please review your ignore-body-lines.regex option accordingly. To remove this warning, set general.regex-style-search=True. More details: https://jorisroovers.github.io/gitlint/configuration/#regex-style-search
1: T3 Title has trailing punctuation (.): "[BlueZ,2/2] org.bluez.Device: Add documentation for ConnectionType."

---
Regards,
Linux Bluetooth

#2 lore raw
In reply to Your Name

Hi Vibhav,

Quoted 45 lines · Your Name

From: Vibhav Pant <[email protected]>

---
 src/device.c | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)
src/device.c +26 -0
index 0f18c8c7f..fa149f1d4 100644
@@ -1003,6 +1003,32 @@ static gboolean dev_property_exists_class(const GDBusPropertyTable *property,
        return device->class != 0;
 }
 
+static gboolean
+dev_property_get_connection_type(const GDBusPropertyTable *property,
+                                DBusMessageIter *iter, void *data)
+{
+       struct btd_device *device = data;
+       const char *str;
+
+       if (device->le && device->bredr)
+               str = "dual";
+       else if (device->le)
+               str = "le";
+       else
+               str = "bredr";
+       dbus_message_iter_append_basic(iter, DBUS_TYPE_STRING, &str);
+       return TRUE;
+}
+
+static gboolean
+dev_property_exists_connection_type(const GDBusPropertyTable *property,
+                                   void *data)
+{
+       struct btd_device *device = data;
+
+       return device->bredr || device->le;
+}
+
 static gboolean dev_property_get_class(const GDBusPropertyTable *property,
                                        DBusMessageIter *iter, void *data)
 {

src/device.c static const GDBusPropertyTable device_properties[] = { · line 3264

Quoted 5 lines
src/device.c +2 -0
@@ -3234,6 +3260,8 @@ static const GDBusPropertyTable device_properties[] = {
        { "Alias", "s", dev_property_get_alias, dev_property_set_alias },
        { "Class", "u", dev_property_get_class, NULL,
                                        dev_property_exists_class },
+       { "ConnectionType", "s", dev_property_get_connection_type, NULL,
+                                       dev_property_exists_connection_type },

Don't really like where this is going, if we need bearer specific API like this then we need a whole new interface, besides you didn't really explain the reason why you needed this, is that to determine if there are SDP records to be fetched? We could perhaps implement something like a Bearer property alongside ConnectBearer but many of the properties of Device are sort of making sense only for one bearer.

Quoted 7 lines
src/device.c +0 -1
        { "Appearance", "q", dev_property_get_appearance, NULL,
                                        dev_property_exists_appearance },
        { "Icon", "s", dev_property_get_icon, NULL,
--

2.46.0

Signature · 1 line
Luiz Augusto von Dentz
#3 lore raw
Quoted 7 lines · Luiz Augusto von Dentz

Hi Vibhav,

Don't really like where this is going, if we need bearer specific API like this then we need a whole new interface, besides you didn't really explain the reason why you needed this, is that to determine if there are SDP records to be fetched?

ConnectionType is needed only to differentiate between LE and BR/EDR devices. This is because the NT kernel creates device objects for every discovered Bluetooth device, and BR/EDR and LE devices exist under different paths/hierarchies (I suppose the equivalent would be BR/EDR devices existing under /dev/bluetooth/devices/bredr, and LE devices under /dev/bluetooth/devices/le). Essentially, BR/EDR and LE devices can be differentiated in several significant ways, and I could not find a way to reliably do that. The current code checks if AddressType is random, otherwise tries to see if the device object has org.bluez.Gatt* interfaces to determine if the connection bearer is LE. However, I understand that's not reliable, and doesn't determine if the device is connected using dual-mode.

That being said, Bearer does sound like a better name for the property, and I can modify the patch to that effect. However, I don't foresee the need for a ConnectBearer property for Wine. Ultimately, my goal is to implement Bluetooth functionality with the minimum possible API surface added to BlueZ, hence the small (ish) patches.

Signature · 3 lines
Vibhav Pant
[email protected]
GPG: 7ED1 D48C 513C A024 BE3A  785F E3FB 28CB 6AB5 9598
#4 lore raw

Hi Vibhav,

Quoted 24 lines · Vibhav Pant
Quoted 7 lines · Luiz Augusto von Dentz

Hi Vibhav,

Don't really like where this is going, if we need bearer specific API like this then we need a whole new interface, besides you didn't really explain the reason why you needed this, is that to determine if there are SDP records to be fetched?

ConnectionType is needed only to differentiate between LE and BR/EDR devices. This is because the NT kernel creates device objects for every discovered Bluetooth device, and BR/EDR and LE devices exist under different paths/hierarchies (I suppose the equivalent would be BR/EDR devices existing under /dev/bluetooth/devices/bredr, and LE devices under /dev/bluetooth/devices/le). Essentially, BR/EDR and LE devices can be differentiated in several significant ways, and I could not find a way to reliably do that. The current code checks if AddressType is random, otherwise tries to see if the device object has org.bluez.Gatt* interfaces to determine if the connection bearer is LE. However, I understand that's not reliable, and doesn't determine if the device is connected using dual-mode.

Quoted 5 lines

That being said, Bearer does sound like a better name for the property, and I can modify the patch to that effect. However, I don't foresee the need for a ConnectBearer property for Wine. Ultimately, my goal is to implement Bluetooth functionality with the minimum possible API surface added to BlueZ, hence the small (ish) patches.

Ive left this out for now, we can revisit this if we find out that it is really necessary in the future. Btw, how is the wine support going, perhaps we can put something under doc or in the guthub wiki about the compatibility with windows.

Quoted 4 lines

--
Vibhav Pant
[email protected]
GPG: 7ED1 D48C 513C A024 BE3A 785F E3FB 28CB 6AB5 9598

Signature · 1 line
Luiz Augusto von Dentz