Bluetooth: conn: Don't wait for buf allocation in syswq #71666
Conversation
zephyrbot
added
area: Bluetooth Host
zephyrbot
requested review from
alwa-nordic,
hermabe,
jhedberg,
sjanc,
Thalley and
theob-pro
subsys/bluetooth/host/conn.c
Outdated
| @@ -1471,6 +1471,12 @@ struct net_buf *bt_conn_create_pdu_timeout(struct net_buf_pool *pool, | |||
| */ | |||
| __ASSERT_NO_MSG(!k_is_in_isr()); | |||
|
|
|||
| if (!K_TIMEOUT_EQ(timeout, K_NO_WAIT) && | |||
| k_current_get() == k_work_queue_thread_get(&k_sys_work_q)) { | |||
| LOG_DBG("Blocking the SYSWQ is bad, mmkay?"); | |||
There was a problem hiding this comment.
Mr. Mackey says that we should have proper debug logs
| LOG_DBG("Blocking the SYSWQ is bad, mmkay?"); |
IMO this does not require a log statement at all
There was a problem hiding this comment.
I think it does since it's a "silent" override. I would be confused at least that my supposedly-blocking alloc is non blocking.
I tried to look into dumping the timeout value in the warning below but it's opaque and has no getters in zephyr :/
I'll reword it ofc
There was a problem hiding this comment.
That is a fair point. I think we should add a similar LOG statement in bt_att_req_alloc as well then
We can't just keep blocking the syswq, people be mad. Override the timeout value so we always have K_NO_WAIT. Also print a message to whomever's listening. Signed-off-by: Jonathan Rico <jonathan.rico@nordicsemi.no>
jori-nordic
force-pushed
the
conn-dont-wait-in-syswq
branch
from
f786bfb to
982f4ee
Compare
|
@Thalley @alwa-nordic I'm debugging a deadlock rn caused by an audio sample blocking in the same way in the syswq. edit: A whole lot of audio samples do the badness, so we'll probably have to add error handling there. |
@jori-nordic can you elaborate? |
|
@Thalley they do a |
We can't just keep blocking the syswq, people be mad. Override the timeout value so we always have K_NO_WAIT. Also print a message to whomever's listening.