Commit 0ea9fd001a14 ("Bluetooth: Shutdown controller after workqueues
are flushed or cancelled") introduced a regression that makes mtkbtsdio
driver stops working:
[ 36.593956] Bluetooth: hci0: Firmware already downloaded
[ 46.814613] Bluetooth: hci0: Execution of wmt command timed out
[ 46.814619] Bluetooth: hci0: Failed to send wmt func ctrl (-110)
The shutdown callback depends on the result of hdev->rx_work, so we
should call it before flushing rx_work:
-> btmtksdio_shutdown()
-> mtk_hci_wmt_sync()
-> __hci_cmd_send()
-> wait for BTMTKSDIO_TX_WAIT_VND_EVT gets cleared
-> btmtksdio_recv_event()
-> hci_recv_frame()
-> queue_work(hdev->workqueue, &hdev->rx_work)
-> clears BTMTKSDIO_TX_WAIT_VND_EVT
So move the shutdown callback before flushing TX/RX queue to resolve the
issue.
Reported-and-tested-by: Mattijs Korpershoek <redacted>
Tested-by: Hsin-Yi Wang <redacted>
Cc: Guenter Roeck <linux@roeck-us.net>
Fixes: 0ea9fd001a14 ("Bluetooth: Shutdown controller after workqueues are flushed or cancelled")
Signed-off-by: Kai-Heng Feng <redacted>
---
net/bluetooth/hci_core.c | 16 ++++++++--------
1 file changed, 8 insertions(+), 8 deletions(-)
@@ -1735,6 +1735,14 @@ int hci_dev_do_close(struct hci_dev *hdev)hci_leds_update_powered(hdev,false);+if(!hci_dev_test_flag(hdev,HCI_UNREGISTER)&&+!hci_dev_test_flag(hdev,HCI_USER_CHANNEL)&&+test_bit(HCI_UP,&hdev->flags)){+/* Execute vendor specific shutdown routine */+if(hdev->shutdown)+hdev->shutdown(hdev);+}+/* Flush RX and TX works */flush_work(&hdev->tx_work);flush_work(&hdev->rx_work);
@@ -1798,14 +1806,6 @@ int hci_dev_do_close(struct hci_dev *hdev)clear_bit(HCI_INIT,&hdev->flags);}-if(!hci_dev_test_flag(hdev,HCI_UNREGISTER)&&-!hci_dev_test_flag(hdev,HCI_USER_CHANNEL)&&-test_bit(HCI_UP,&hdev->flags)){-/* Execute vendor specific shutdown routine */-if(hdev->shutdown)-hdev->shutdown(hdev);-}-/* flush cmd work */flush_work(&hdev->cmd_work);
From: Hsin-Yi Wang <hidden> Date: 2021-08-10 04:11:07
On Tue, Aug 10, 2021 at 1:44 AM Kai-Heng Feng
[off-list ref] wrote:
Commit 0ea9fd001a14 ("Bluetooth: Shutdown controller after workqueues
are flushed or cancelled") introduced a regression that makes mtkbtsdio
driver stops working:
[ 36.593956] Bluetooth: hci0: Firmware already downloaded
[ 46.814613] Bluetooth: hci0: Execution of wmt command timed out
[ 46.814619] Bluetooth: hci0: Failed to send wmt func ctrl (-110)
The shutdown callback depends on the result of hdev->rx_work, so we
should call it before flushing rx_work:
-> btmtksdio_shutdown()
-> mtk_hci_wmt_sync()
-> __hci_cmd_send()
-> wait for BTMTKSDIO_TX_WAIT_VND_EVT gets cleared
-> btmtksdio_recv_event()
-> hci_recv_frame()
-> queue_work(hdev->workqueue, &hdev->rx_work)
-> clears BTMTKSDIO_TX_WAIT_VND_EVT
So move the shutdown callback before flushing TX/RX queue to resolve the
issue.
Reported-and-tested-by: Mattijs Korpershoek <redacted>
Tested-by: Hsin-Yi Wang <redacted>
Hello,
Sorry for confusion, but the version I tested is this one:
https://lkml.org/lkml/2021/8/4/486 (shutdown is prior to the
test_and_clear HCI_UP)
I tested this version and still see the error I've seen before.
@@ -1735,6 +1735,14 @@ int hci_dev_do_close(struct hci_dev *hdev)hci_leds_update_powered(hdev,false);+if(!hci_dev_test_flag(hdev,HCI_UNREGISTER)&&+!hci_dev_test_flag(hdev,HCI_USER_CHANNEL)&&+test_bit(HCI_UP,&hdev->flags)){+/* Execute vendor specific shutdown routine */+if(hdev->shutdown)+hdev->shutdown(hdev);+}+/* Flush RX and TX works */flush_work(&hdev->tx_work);flush_work(&hdev->rx_work);
@@ -1798,14 +1806,6 @@ int hci_dev_do_close(struct hci_dev *hdev)clear_bit(HCI_INIT,&hdev->flags);}-if(!hci_dev_test_flag(hdev,HCI_UNREGISTER)&&-!hci_dev_test_flag(hdev,HCI_USER_CHANNEL)&&-test_bit(HCI_UP,&hdev->flags)){-/* Execute vendor specific shutdown routine */-if(hdev->shutdown)-hdev->shutdown(hdev);-}-/* flush cmd work */flush_work(&hdev->cmd_work);--
On Tue, Aug 10, 2021 at 12:10 PM Hsin-Yi Wang [off-list ref] wrote:
On Tue, Aug 10, 2021 at 1:44 AM Kai-Heng Feng
[off-list ref] wrote:
quoted
Commit 0ea9fd001a14 ("Bluetooth: Shutdown controller after workqueues
are flushed or cancelled") introduced a regression that makes mtkbtsdio
driver stops working:
[ 36.593956] Bluetooth: hci0: Firmware already downloaded
[ 46.814613] Bluetooth: hci0: Execution of wmt command timed out
[ 46.814619] Bluetooth: hci0: Failed to send wmt func ctrl (-110)
The shutdown callback depends on the result of hdev->rx_work, so we
should call it before flushing rx_work:
-> btmtksdio_shutdown()
-> mtk_hci_wmt_sync()
-> __hci_cmd_send()
-> wait for BTMTKSDIO_TX_WAIT_VND_EVT gets cleared
-> btmtksdio_recv_event()
-> hci_recv_frame()
-> queue_work(hdev->workqueue, &hdev->rx_work)
-> clears BTMTKSDIO_TX_WAIT_VND_EVT
So move the shutdown callback before flushing TX/RX queue to resolve the
issue.
Reported-and-tested-by: Mattijs Korpershoek <redacted>
Tested-by: Hsin-Yi Wang <redacted>
Hello,
Sorry for confusion, but the version I tested is this one:
https://lkml.org/lkml/2021/8/4/486 (shutdown is prior to the
test_and_clear HCI_UP)
I tested this version and still see the error I've seen before.
Ah, sorry for causing your trouble, I am the one who got confused :(
HCI_UP is obviously required for hci_req_sync() to work.
Let me resend one, and sorry again!
Kai-Heng
@@ -1735,6 +1735,14 @@ int hci_dev_do_close(struct hci_dev *hdev)hci_leds_update_powered(hdev,false);+if(!hci_dev_test_flag(hdev,HCI_UNREGISTER)&&+!hci_dev_test_flag(hdev,HCI_USER_CHANNEL)&&+test_bit(HCI_UP,&hdev->flags)){+/* Execute vendor specific shutdown routine */+if(hdev->shutdown)+hdev->shutdown(hdev);+}+/* Flush RX and TX works */flush_work(&hdev->tx_work);flush_work(&hdev->rx_work);
@@ -1798,14 +1806,6 @@ int hci_dev_do_close(struct hci_dev *hdev)clear_bit(HCI_INIT,&hdev->flags);}-if(!hci_dev_test_flag(hdev,HCI_UNREGISTER)&&-!hci_dev_test_flag(hdev,HCI_USER_CHANNEL)&&-test_bit(HCI_UP,&hdev->flags)){-/* Execute vendor specific shutdown routine */-if(hdev->shutdown)-hdev->shutdown(hdev);-}-/* flush cmd work */flush_work(&hdev->cmd_work);--