Skip to contentSkip to conversation

Bluetooth: hci_sync: fix mesh adv timeout units

PATCH Applied Pauli Virtanen 3 participants 6 messages Oct 5 – Oct 6

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

cover Bluetooth: hci_sync: fix mesh adv timeout units

For mesh advertisements = adv_info::mesh nonzero, the time units are milliseconds instead of seconds like normally.

This was forgotten in hci_schedule_adv_instance_sync() and hci_enable_ext_advertising_sync() which scheduled 1000x too large timeouts.

Fix that. This is the cause for the mesh-tester "Mesh - Send cancel" failures.

Since the timeout units depend on the adv type, make hdev->adv_instance_timeout to just bool indicating if timeout is active or not.

Triaged via LLM. I wrote the patches.

Assisted-by opencode:qwen3.8-flash-next

Pauli Virtanen (2):
Bluetooth: hci_sync: fix mesh adv timeout units
Bluetooth: make hdev->adv_instance_timeout a boolean

 include/net/bluetooth/hci_core.h |  2 +-
 net/bluetooth/hci_core.c         |  6 +++---
 net/bluetooth/hci_sync.c         | 14 ++++++++------
 net/bluetooth/mgmt.c             |  2 +-
 4 files changed, 13 insertions(+), 11 deletions(-)
Signature · 1 line
2.55.0
#1 lore raw

PATCH 1/2 Bluetooth: hci_sync: fix mesh adv timeout units

Mesh adv instance timeout and duration are in milliseconds, see mesh_send_sync, hci_add_adv_instance. hci_schedule_adv_instance_sync() and hci_enable_ext_advertising_sync() ignore this, and scale the timeout wrong.

Fix for mesh instances by using msecs_to_jiffies() in hci_schedule_adv_instance_sync and fixing the scaling in hci_enable_ext_advertising_sync.

Fixes BlueZ "Mesh - Send cancel" test failures.

Fixes b338d91703fa ("Bluetooth: Implement support for Mesh")
Assisted-by opencode:qwen3.8-flash-next sashiko
---
 net/bluetooth/hci_sync.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
net/bluetooth/hci_sync.c +4 -2
index 1fe11d3ef1aa..94dcbbc767e5 100644
@@ -1670,7 +1670,8 @@ int hci_enable_ext_advertising_sync(struct hci_dev *hdev, u8 instance)
 	 * scheduling it.
 	 */
 	if (adv && adv->timeout) {
-		u16 duration = adv->timeout * MSEC_PER_SEC;
+		u16 duration = adv->mesh ? adv->timeout :
+					   adv->timeout * MSEC_PER_SEC;
 
 		/* Time = N * 10 ms */
 		set->duration = cpu_to_le16(duration / 10);
@@ -2101,7 +2102,8 @@ int hci_schedule_adv_instance_sync(struct hci_dev *hdev, u8 instance,
 		hdev->adv_instance_timeout = timeout;
 		queue_delayed_work(hdev->req_workqueue,
 				   &hdev->adv_instance_expire,
-				   secs_to_jiffies(timeout));
+				   adv->mesh ? msecs_to_jiffies(timeout) :
+					       secs_to_jiffies(timeout));
 	}
 
 	/* If we're just re-scheduling the same instance again then do not
Signature · 1 line
2.55.0
#2 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/series/1179822/

---Test result---

Test Summary:

CheckPatch                    PASS      2.14 seconds
VerifyFixes                   PASS      0.13 seconds
VerifySignedoff               PASS      0.13 seconds
GitLint                       PASS      0.69 seconds
SubjectPrefix                 PASS      1.87 seconds
BuildKernel                   PASS      29.74 seconds
CheckAllWarning               PASS      33.57 seconds
CheckSparse                   PASS      36.52 seconds
BuildKernel32                 PASS      29.49 seconds
CheckKernelLLVM               PASS      37.48 seconds
TestRunnerSetup               PASS      781.61 seconds
TestRunner_l2cap-tester       PASS      12.58 seconds
TestRunner_iso-tester         PASS      10.75 seconds
TestRunner_bnep-tester        PASS      3.13 seconds
TestRunner_mgmt-tester        PASS      61.35 seconds
TestRunner_rfcomm-tester      PASS      4.49 seconds
TestRunner_sco-tester         PASS      5.60 seconds
TestRunner_ioctl-tester       PASS      4.66 seconds
TestRunner_mesh-tester        PASS      5.51 seconds
TestRunner_smp-tester         PASS      4.33 seconds
TestRunner_userchan-tester    PASS      3.17 seconds
TestRunner_6lowpan-tester     PASS      15.96 seconds
IncrementalBuild              PASS      60.62 seconds

https://github.com/bluez/bluetooth-next/pull/863

---
Regards,
Linux Bluetooth

#3 lore raw

PATCH 2/2 Bluetooth: make hdev->adv_instance_timeout a boolean

In reply to Pauli Virtanen

Mesh uses millisecond time units for advertisements instead of seconds as normally. These values go to hdev->adv_instance_timeout, which then has mixed units. hdev->adv_instance_timeout is however only used as a true/false flag.

Make hdev->adv_instance_timeout boolean flag, to avoid possible future problems with its mixed-units values.

---
 include/net/bluetooth/hci_core.h | 2 +-
 net/bluetooth/hci_core.c         | 6 +++---
 net/bluetooth/hci_sync.c         | 8 ++++----
 net/bluetooth/mgmt.c             | 2 +-
 4 files changed, 9 insertions(+), 9 deletions(-)
include/net/bluetooth/hci_core.h +1 -1
index f631d9484a0b..aa04dec45153 100644
@@ -601,7 +601,7 @@ struct hci_dev {
 	struct list_head	adv_instances;
 	unsigned int		adv_instance_cnt;
 	__u8			cur_adv_instance;
-	__u16			adv_instance_timeout;
+	bool			adv_instance_timeout;
 	struct delayed_work	adv_instance_expire;
 
 	struct idr		adv_monitors_idr;
net/bluetooth/hci_core.c +3 -3
index 05d1ca25f44f..416803d8d604 100644
@@ -1617,7 +1617,7 @@ int hci_remove_adv_instance(struct hci_dev *hdev, u8 instance)
 	if (hdev->cur_adv_instance == instance) {
 		if (hdev->adv_instance_timeout) {
 			cancel_delayed_work(&hdev->adv_instance_expire);
-			hdev->adv_instance_timeout = 0;
+			hdev->adv_instance_timeout = false;
 		}
 		hdev->cur_adv_instance = 0x00;
 	}
@@ -1647,7 +1647,7 @@ void hci_adv_instances_clear(struct hci_dev *hdev)
 
 	if (hdev->adv_instance_timeout) {
 		disable_delayed_work(&hdev->adv_instance_expire);
-		hdev->adv_instance_timeout = 0;
+		hdev->adv_instance_timeout = false;
 	}
 
 	list_for_each_entry_safe(adv_instance, n, &hdev->adv_instances, list) {
@@ -2437,7 +2437,7 @@ struct hci_dev *hci_alloc_dev_priv(int sizeof_priv)
 	hdev->adv_tx_power = HCI_TX_POWER_INVALID;
 	hdev->adv_instance_cnt = 0;
 	hdev->cur_adv_instance = 0x00;
-	hdev->adv_instance_timeout = 0;
+	hdev->adv_instance_timeout = false;
 
 	hdev->advmon_allowlist_duration = 300;
 	hdev->advmon_no_filter_duration = 500;
net/bluetooth/hci_sync.c +4 -4
index 94dcbbc767e5..7a07aba43262 100644
@@ -468,7 +468,7 @@ static void reenable_adv(struct work_struct *work)
 static void cancel_adv_timeout(struct hci_dev *hdev)
 {
 	if (hdev->adv_instance_timeout) {
-		hdev->adv_instance_timeout = 0;
+		hdev->adv_instance_timeout = false;
 		cancel_delayed_work(&hdev->adv_instance_expire);
 	}
 }
@@ -568,7 +568,7 @@ static void adv_timeout_expire(struct work_struct *work)
 
 	hci_dev_lock(hdev);
 
-	hdev->adv_instance_timeout = 0;
+	hdev->adv_instance_timeout = false;
 
 	if (hdev->cur_adv_instance == 0x00)
 		goto unlock;
@@ -2099,7 +2099,7 @@ int hci_schedule_adv_instance_sync(struct hci_dev *hdev, u8 instance,
 
 	/* Only use work for scheduling instances with legacy advertising */
 	if (!ext_adv_capable(hdev)) {
-		hdev->adv_instance_timeout = timeout;
+		hdev->adv_instance_timeout = true;
 		queue_delayed_work(hdev->req_workqueue,
 				   &hdev->adv_instance_expire,
 				   adv->mesh ? msecs_to_jiffies(timeout) :
@@ -5572,7 +5572,7 @@ int hci_dev_close_sync(struct hci_dev *hdev)
 
 	if (hdev->adv_instance_timeout) {
 		cancel_delayed_work_sync(&hdev->adv_instance_expire);
-		hdev->adv_instance_timeout = 0;
+		hdev->adv_instance_timeout = false;
 	}
 
 	err = hci_dev_shutdown(hdev);
net/bluetooth/mgmt.c +1 -1
index aea3482e3893..2e35b09579d3 100644
@@ -1299,7 +1299,7 @@ void mgmt_advertising_removed(struct sock *sk, struct hci_dev *hdev,
 static void cancel_adv_timeout(struct hci_dev *hdev)
 {
 	if (hdev->adv_instance_timeout) {
-		hdev->adv_instance_timeout = 0;
+		hdev->adv_instance_timeout = false;
 		cancel_delayed_work(&hdev->adv_instance_expire);
 	}
 }
Signature · 1 line
2.55.0
#4 lore raw
In reply to Pauli Virtanen
This series was applied to bluetooth/bluetooth-next.git (master) by Luiz Augusto von Dentz <[email protected]>:

Hello:

This series was applied to bluetooth/bluetooth-next.git (master) by Luiz Augusto von Dentz <[email protected]>:

Quoted 8 lines · On Mon, 5 Oct 2026 21:13:53 +0300 you wrote:

For mesh advertisements = adv_info::mesh nonzero, the time units are milliseconds instead of seconds like normally.

This was forgotten in hci_schedule_adv_instance_sync() and hci_enable_ext_advertising_sync() which scheduled 1000x too large timeouts.

[...]

Here is the summary with links:
- [1/2] Bluetooth: hci_sync: fix mesh adv timeout units

    https://git.kernel.org/bluetooth/bluetooth-next/c/87a49b610708

- [2/2] Bluetooth: make hdev->adv_instance_timeout a boolean (no matching commit)

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
#5 lore raw
In reply to Pauli Virtanen
This series was applied to bluetooth/bluetooth-next.git (master) by Luiz Augusto von Dentz <[email protected]>:

Hello:

This series was applied to bluetooth/bluetooth-next.git (master) by Luiz Augusto von Dentz <[email protected]>:

Quoted 8 lines · On Mon, 5 Oct 2026 21:13:53 +0300 you wrote:

For mesh advertisements = adv_info::mesh nonzero, the time units are milliseconds instead of seconds like normally.

This was forgotten in hci_schedule_adv_instance_sync() and hci_enable_ext_advertising_sync() which scheduled 1000x too large timeouts.

[...]

Here is the summary with links:
- [1/2] Bluetooth: hci_sync: fix mesh adv timeout units (no matching commit)
- [2/2] Bluetooth: make hdev->adv_instance_timeout a boolean
https://git.kernel.org/bluetooth/bluetooth-next/c/dedfbdb9bf06

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html