accel/qaic: Fix NNC message corruption
authorJeffrey Hugo <quic_jhugo@quicinc.com>
Wed, 17 May 2023 19:35:40 +0000 (13:35 -0600)
committerJeffrey Hugo <quic_jhugo@quicinc.com>
Tue, 23 May 2023 15:51:38 +0000 (09:51 -0600)
commite997c218ad736fd6f524d73a987bad9d94128d3d
tree90cc9c9908228f4eae8a0f5444535ae161dfb752
parent75af0a585af93183ba68bb1b45d0d7a61e963712
accel/qaic: Fix NNC message corruption

If msg_xfer() is unable to queue part of a NNC message because the MHI ring
is full, it will attempt to give the QSM some time to drain the queue.
However, if QSM fails to make any room, msg_xfer() will fail and tell the
caller to try again.  This is problematic because part of the message may
have been committed to the ring and there is no mechanism to revoke that
content.  This will cause QSM to receive a corrupt message.

The better way to do this is to check if the ring has enough space for the
entire message before committing any of the message.  Since msg_xfer() is
under the cntl_mutex no one else can come in and consume the space.

Fixes: 129776ac2e38 ("accel/qaic: Add control path")
Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Reviewed-by: Pranjal Ramajor Asha Kanojiya <quic_pkanojiy@quicinc.com>
Reviewed-by: Carl Vanderlip <quic_carlv@quicinc.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230517193540.14323-6-quic_jhugo@quicinc.com
drivers/accel/qaic/qaic_control.c