This patch is very much incorrect. Will send a v2 later.
On Mon, 12 May 2025 at 15:50, Sanchayan Maity <sanchayan@xxxxxxxxxxxxx> wrote:
>
> Creating the media request before suspend can result in
> a situation where media_request_reply is not called as
> the suspend action for a profile may not depend on the
> callback of the stop operation.
>
> This is the case for ASHA profile as devices are not
> expected to acknowledge the receipt of a stop command.
>
> Moving media_request_create before media_transport_suspend
> should not change the behaviour of release.
> ---
> profiles/audio/transport.c | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/profiles/audio/transport.c b/profiles/audio/transport.c
> index c5ae3a71b..dc4f8fea0 100644
> --- a/profiles/audio/transport.c
> +++ b/profiles/audio/transport.c
> @@ -917,29 +917,29 @@ static DBusMessage *release(DBusConnection *conn, DBusMessage *msg,
> if (g_str_equal(member, "Acquire")) {
> media_request_reply(owner->pending, ECANCELED);
> media_owner_remove(owner);
> } else {
> return btd_error_in_progress(msg);
> }
> }
>
> transport_set_state(transport, TRANSPORT_STATE_SUSPENDING);
>
> + req = media_request_create(msg, id);
> + media_owner_add(owner, req);
> +
> id = media_transport_suspend(transport, owner);
> if (id == 0) {
> media_transport_remove_owner(transport);
> return g_dbus_create_reply(msg, DBUS_TYPE_INVALID);
> }
>
> - req = media_request_create(msg, id);
> - media_owner_add(owner, req);
> -
> return NULL;
> }
>
> static gboolean get_device(const GDBusPropertyTable *property,
> DBusMessageIter *iter, void *data)
> {
> struct media_transport *transport = data;
> const char *path;
>
> if (transport->device)
> --
> 2.49.0
>
