Skip to contentSkip to conversation

Bluetooth: hci_ll: Add DT control for enhanced SCO setup erratum

PATCH In review Stefano Radaelli 3 participants 8 messages Feb 23 – Feb 24

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

cover Bluetooth: hci_ll: Add DT control for enhanced SCO setup erratum

While validating SCO audio on a platform using TI WL183x Bluetooth modules with the hci_ll driver, we observed failures when the HCI Enhanced Setup Synchronous Connection command was used.

Although the controller advertises support for the command, SCO setup fails in certain configurations (e.g. BT_VOICE_TRANSPARENT/mSBC). This matches the scenario described in commit 05abad857277 ("Bluetooth: HCI: Add HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN quirk").

Initially, we considered setting the HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN quirk unconditionally in hci_ll. However, this would affect all TI controllers handled by the driver, including configurations where the enhanced setup command works correctly.

To avoid hardcoding the quirk globally, this series introduces an optional DT property that describes the controller limitation and allows enabling the quirk on affected boards only.

Patch 1 updates the TI Bluetooth DT binding.
Patch 2 adds support for the property in hci_ll and sets the quirk during controller setup when requested.

Tested on VAR-SOM-MX6 using TI WL183x modules where SCO audio was previously failing.

Comments welcome.

Stefano Radaelli (2):
dt-bindings: net: bluetooth: ti: Add property for enhanced SCO setup
erratum
Bluetooth: hci_ll: Add DT property to disable enhanced SCO setup

 .../devicetree/bindings/net/bluetooth/ti,bluetooth.yaml  | 7 +++++++
 drivers/bluetooth/hci_ll.c                               | 9 +++++++++
 2 files changed, 16 insertions(+)

base-commit: 6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f

Signature · 1 line
2.47.3
#1 lore raw

PATCH 1/2 dt-bindings: net: bluetooth: ti: Add property for enhanced SCO setup erratum

From: Stefano Radaelli <[email protected]>

Some Bluetooth controller/firmware combinations advertise support for the HCI Enhanced Setup Synchronous Connection command but do not handle it correctly for SCO setup in certain configurations (e.g. BT_VOICE_TRANSPARENT/mSBC). This results in SCO audio not working unless the enhanced setup path is avoided.

Add an optional boolean property, ti,no-enhanced-setup-sync-conn, to describe this controller limitation.

---
 .../devicetree/bindings/net/bluetooth/ti,bluetooth.yaml    | 7 +++++++
 1 file changed, 7 insertions(+)
Documentation/devicetree/bindings/net/bluetooth/ti,bluetooth.yaml +7 -0
index 290abc22e18a..92ed443cb427 100644
@@ -71,6 +71,13 @@ properties:
     items:
       - const: bd-address
 
+  ti,no-enhanced-setup-sync-conn:
+    type: boolean
+    description:
+      Indicates that the Bluetooth controller/firmware does not correctly
+      handle the HCI Enhanced Setup Synchronous Connection command for SCO
+      setup, despite advertising support for it.
+
 required:
   - compatible
 
Signature · 1 line
2.47.3
#7 lore raw

PATCH 2/2 Bluetooth: hci_ll: Add DT property to disable enhanced SCO setup

In reply to Stefano Radaelli

The HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN quirk was introduced by commit 05abad857277 ("Bluetooth: HCI: Add HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN quirk") to handle controllers that advertise support for the HCI Enhanced Setup Synchronous Connection command but fail to use it properly with some SCO configurations (e.g. BT_VOICE_TRANSPARENT/mSBC).

This behavior has been observed on TI controllers used with the hci_ll driver, where SCO setup can fail unless the enhanced setup path is avoided.

Add support for the optional DT property "ti,no-enhanced-setup-sync-conn" and, when present, set HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN during controller setup, as intended by the quirk documentation.

---
 drivers/bluetooth/hci_ll.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/bluetooth/hci_ll.c +9 -0
index 91acf24f1ef5..5f69f0654c19 100644
@@ -68,6 +68,7 @@ struct ll_device {
 	struct gpio_desc *enable_gpio;
 	struct clk *ext_clk;
 	bdaddr_t bdaddr;
+	bool broken_enhanced_setup;
 };
 
 struct ll_struct {
@@ -656,6 +657,10 @@ static int ll_setup(struct hci_uart *hu)
 			hci_set_quirk(hu->hdev, HCI_QUIRK_INVALID_BDADDR);
 	}
 
+	if (lldev->broken_enhanced_setup)
+		hci_set_quirk(hu->hdev,
+			      HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN);
+
 	/* Operational speed if any */
 	if (hu->oper_speed)
 		speed = hu->oper_speed;
@@ -710,6 +715,10 @@ static int hci_ti_probe(struct serdev_device *serdev)
 	of_property_read_u32(serdev->dev.of_node, "max-speed", &max_speed);
 	hci_uart_set_speeds(hu, 115200, max_speed);
 
+	lldev->broken_enhanced_setup =
+	       of_property_read_bool(serdev->dev.of_node,
+				     "ti,no-enhanced-setup-sync-conn");
+
 	/* optional BD address from nvram */
 	bdaddr_cell = nvmem_cell_get(&serdev->dev, "bd-address");
 	if (IS_ERR(bdaddr_cell)) {
Signature · 1 line
2.47.3
#2 lore raw

Re: [PATCH v1 1/2] dt-bindings: net: bluetooth: ti: Add property for enhanced SCO setup erratum

In reply to Stefano Radaelli
Quoted 3 lines · Stefano Radaelli

From: Stefano Radaelli <[email protected]>

Some Bluetooth controller/firmware combinations advertise support for the

Which ones?

Quoted 18 lines

HCI Enhanced Setup Synchronous Connection command but do not handle it correctly for SCO setup in certain configurations (e.g. BT_VOICE_TRANSPARENT/mSBC). This results in SCO audio not working unless the enhanced setup path is avoided.

Add an optional boolean property, ti,no-enhanced-setup-sync-conn, to describe this controller limitation.

---
 .../devicetree/bindings/net/bluetooth/ti,bluetooth.yaml    | 7 +++++++
 1 file changed, 7 insertions(+)
Documentation/devicetree/bindings/net/bluetooth/ti,bluetooth.yaml +0 -0
index 290abc22e18a..92ed443cb427 100644

Documentation/devicetree/bindings/net/bluetooth/ti,bluetooth.yaml properties: · line 79

Quoted 9 lines
Documentation/devicetree/bindings/net/bluetooth/ti,bluetooth.yaml +6 -0
@@ -71,6 +71,13 @@ properties:
     items:
       - const: bd-address
 
+  ti,no-enhanced-setup-sync-conn:
+    type: boolean
+    description:
+      Indicates that the Bluetooth controller/firmware does not correctly
+      handle the HCI Enhanced Setup Synchronous Connection command for SCO
+      setup, despite advertising support for it.

This looks a lot like controller issue, thus deducible from the compatible. And even if particular firmware has problem, how do you see changing static DTS on linux-firmware update?

Best regards,
Krzysztof

#3 lore raw

Re: [PATCH v1 1/2] dt-bindings: net: bluetooth: ti: Add property for enhanced SCO setup erratum

Quoted 8 lines · Krzysztof Kozlowski
Quoted 3 lines · Stefano Radaelli

From: Stefano Radaelli <[email protected]>

Some Bluetooth controller/firmware combinations advertise support for the

Which ones?

Quoted 6 lines

This looks a lot like controller issue, thus deducible from the compatible. And even if particular firmware has problem, how do you see changing static DTS on linux-firmware update?

Best regards,
Krzysztof

Hi Krzysztof,

thank you for your review!

We have observed this behaviour specifically on TI WL1831 modules.

When using the HCI Enhanced Setup Synchronous Connection command, SCO setup fails even though the controller reports the command as supported. The only working path for this controller is to fall back to the legacy HCI Setup Synchronous Connection (0x0028). The behaviour matches the scenario described in commit 05abad857277, which introduced HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN for that reason.

We are using the latest firmware available from TI for the wl18xx family (from the official TI repository), so this is not related to an outdated firmware version.

The reason for proposing a DT property was to keep the workaround flexible and avoid forcing the quirk on all TI controllers handled by hci_ll without confirmation.

However, since this may be considered deducible from the compatible, would you prefer enabling HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN automatically for "ti,wl1831-st" instead of using a DT property?

Thanks for the feedback.

Best regards,
Stefano

#6 lore raw
In reply to Stefano Radaelli

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=1056461

---Test result---

Test Summary:

CheckPatch                    PENDING   0.30 seconds
GitLint                       PENDING   0.27 seconds
SubjectPrefix                 FAIL      0.54 seconds
BuildKernel                   PASS      25.55 seconds
CheckAllWarning               PASS      28.47 seconds
CheckSparse                   PASS      32.15 seconds
BuildKernel32                 PASS      25.25 seconds
TestRunnerSetup               PASS      558.65 seconds
TestRunner_l2cap-tester       PASS      28.27 seconds
TestRunner_iso-tester         PASS      96.32 seconds
TestRunner_bnep-tester        PASS      6.34 seconds
TestRunner_mgmt-tester        FAIL      114.13 seconds
TestRunner_rfcomm-tester      PASS      9.54 seconds
TestRunner_sco-tester         FAIL      14.50 seconds
TestRunner_ioctl-tester       PASS      10.29 seconds
TestRunner_mesh-tester        FAIL      12.51 seconds
TestRunner_smp-tester         PASS      8.74 seconds
TestRunner_userchan-tester    PASS      6.66 seconds
IncrementalBuild              PENDING   0.75 seconds

Details
##############################
Test: CheckPatch - PENDING
Desc: Run checkpatch.pl script
Output:

##############################
Test: GitLint - PENDING
Desc: Run gitlint
Output:

##############################
Test: SubjectPrefix - FAIL
Desc: Check subject contains "Bluetooth" prefix Output:
"Bluetooth: " prefix is not specified in the subject
##############################
Test: TestRunner_mgmt-tester - FAIL
Desc: Run mgmt-tester with test-runner Output:
Total: 494, Passed: 489 (99.0%), Failed: 1, Not Run: 4

Failed Test Cases
Read Exp Feature - Success Failed 0.107 seconds
##############################
Test: TestRunner_sco-tester - FAIL
Desc: Run sco-tester with test-runner
Output:
WARNING: possible circular locking dependency detected

BUG: sleeping function called from invalid context at net/core/sock.c:3782

Total: 30, Passed: 30 (100.0%), Failed: 0, Not Run: 0
##############################
Test: TestRunner_mesh-tester - FAIL
Desc: Run mesh-tester with test-runner Output:
Total: 10, Passed: 8 (80.0%), Failed: 2, Not Run: 0

Failed Test Cases

Mesh - Send cancel - 1                               Timed out    2.667 seconds
Mesh - Send cancel - 2                               Timed out    1.992 seconds

##############################
Test: IncrementalBuild - PENDING
Desc: Incremental build with the patches in the series Output:

---
Regards,
Linux Bluetooth

#4 lore raw

Re: [PATCH v1 1/2] dt-bindings: net: bluetooth: ti: Add property for enhanced SCO setup erratum

In reply to Stefano Radaelli
Quoted 30 lines · Stefano Radaelli
Quoted 8 lines · Krzysztof Kozlowski
Quoted 3 lines · Stefano Radaelli

From: Stefano Radaelli <[email protected]>

Some Bluetooth controller/firmware combinations advertise support for the

Which ones?

Quoted 6 lines

This looks a lot like controller issue, thus deducible from the compatible. And even if particular firmware has problem, how do you see changing static DTS on linux-firmware update?

Best regards,
Krzysztof

Hi Krzysztof,

thank you for your review!

We have observed this behaviour specifically on TI WL1831 modules.

When using the HCI Enhanced Setup Synchronous Connection command, SCO setup fails even though the controller reports the command as supported. The only working path for this controller is to fall back to the legacy HCI Setup Synchronous Connection (0x0028). The behaviour matches the scenario described in commit 05abad857277, which introduced HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN for that reason.

Quoted 7 lines

We are using the latest firmware available from TI for the wl18xx family (from the official TI repository), so this is not related to an outdated firmware version.

The reason for proposing a DT property was to keep the workaround flexible and avoid forcing the quirk on all TI controllers handled by hci_ll without confirmation.

I did not propose to enable it on ALL TI controllers. I asked to deduce it from given compatible, so to enable it on affected devices.

Quoted 5 lines

However, since this may be considered deducible from the compatible, would you prefer enabling HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN automatically for "ti,wl1831-st" instead of using a DT property?

Yes

Best regards,
Krzysztof

#5 lore raw

Re: [PATCH v1 1/2] dt-bindings: net: bluetooth: ti: Add property for enhanced SCO setup erratum

Hi Krzysztof,

Quoted 8 lines · Krzysztof Kozlowski
Quoted 5 lines

However, since this may be considered deducible from the compatible, would you prefer enabling HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN automatically for "ti,wl1831-st" instead of using a DT property?

Yes

We have now tested the WL183x family and observed the same behaviour across WL1831, WL1835 and WL1837 modules.

Since this affects the WL183x family and is therefore deducible from the compatible, I will respin the patch to enable HCI_QUIRK_BROKEN_ENHANCED_SETUP_SYNC_CONN automatically for:
- "ti,wl1831-st"
- "ti,wl1835-st"
- "ti,wl1837-st"

Thanks for the guidance.

Best regards,
Stefano