Re: [PATCH 3/3] Bluetooth: Don't leave abandoned SCO links up in the controller

[Luiz Augusto von Dentz]

Hi Hitalo,

On Thu, Oct 8, 2026 at 11:28 AM Hitalo Souza <enghitalo@xxxxxxxxx> wrote:
>
> Hi Luiz,
>
> On Wed, Oct 7, 2026 at 2:15 PM Luiz Augusto von Dentz <luiz.dentz@xxxxxxxxx> wrote:
> > Sashiko flagged quite a few problems:
> >
> > https://sashiko.dev/#/patchset/20261007154353.148223-1-enghitalo%40gmail.com
> >
> > We need to check if the logic of hci_sco_disconnect_orphan couldn't be
> > made more generically, so in case the connection was aborted but we
> > received the connection complete that shall always result in
> > HCI_OP_DISCONNECT so the handle don't stay active in the controller.
>
> Thanks, agreed. v2 does that in two patches:
>
> - when hci_conn_set_handle() refuses the handle because the connection
>   is being aborted, the Connection Complete, Synchronous Connection
>   Complete and LE Connection Complete handlers now disconnect the link,
>   with the abort reason;
> - an ACL link that completes after its connection was already aborted
>   and deleted (an incoming connection, where Create Connection Cancel
>   fails at once) is disconnected too, as SCO links already are.
>
> Both reproduce in an emulator (ACL outgoing and incoming, and LE):
> without them the link stays up. The SCO patch now uses the same helper.

I don't see any new test cases though, so I would appreciate it if you
include them (e.g. l2cap-tester, iso-tester, sco-tester that attempts
to abort the connection in between receiving connection complete
events).

> The other two Sashiko findings are addressed as well: the legacy
> Connection Complete path in the SCO patch, and the air mode
> notification of a SCO link that failed in a separate patch, as it
> predates this series (1f8330ea1692).
>
> v2 is here:
> https://lore.kernel.org/linux-bluetooth/20261008151238.2138665-1-enghitalo@xxxxxxxxx/
>
> Hitalo



-- 
Luiz Augusto von Dentz