Hi Luiz, On 06.02.24 23:22, Luiz Augusto von Dentz wrote:
Hi Jonas, On Tue, Feb 6, 2024 at 6:08 AM Jonas Dreßler <verdre@xxxxxxx> wrote:Since commit 4c67bc74f016 ("[Bluetooth] Support concurrent connect requests"), the kernel supports trying to connect again in case the bluetooth card is busy and fails to connect. The logic that should handle this became a bit spotty over time, and also cards these days appear to fail with more errors than just "Command Disallowed". This series refactores the handling of concurrent connection requests by serializing all "Create Connection" commands for ACL connections similar to how we do it for LE connections. --- v1: https://lore.kernel.org/linux-bluetooth/20240102185933.64179-1-verdre@xxxxxxx/ v2: https://lore.kernel.org/linux-bluetooth/20240108183938.468426-1-verdre@xxxxxxx/ v3: https://lore.kernel.org/linux-bluetooth/20240108224614.56900-1-verdre@xxxxxxx/ v4: - Removed first two commits since they are already applied - Removed a BT_DBG() message in the acl_create_connection() function while moving to hci_sync because it seemed out of place in hci_sync - Added a mention of the test failure in mgmt-tester to commit message Jonas Dreßler (2): Bluetooth: hci_conn: Only do ACL connections sequentially Bluetooth: Remove pending ACL connection attempts include/net/bluetooth/hci.h | 1 + include/net/bluetooth/hci_core.h | 1 - include/net/bluetooth/hci_sync.h | 3 ++ net/bluetooth/hci_conn.c | 83 +++----------------------------- net/bluetooth/hci_event.c | 21 ++------ net/bluetooth/hci_sync.c | 70 +++++++++++++++++++++++++++ 6 files changed, 86 insertions(+), 93 deletions(-) -- 2.43.0Doesn't seem to work with the new test: Sequential connect - setup complete Sequential connect - run Create connection finished Connect failed for Pair Device Create connection finished Sequential connect - test timed out Sequential connect - teardown
Oh you're right, it's because of the increased delay to 5.12 seconds for sending the page timeout. We're actually waiting for connections to fail, so the test will now take at least 5.12s * 3 instead of 2s * 3, which means test timeout needs to be increased from 7 to more like 16 seconds. Sorry for not catching this one, should've run the test before submitting the patch... Cheers, Jonas
