There is a good chance that while the channel ring gets processed, the STOP or RESET command for the channel might be received from the MHI host. In those cases, the entire channel ring processing needs to be protected by chan->lock to prevent the race where the corresponding channel ring might be reset.
While at it, let's also add a sanity check to make sure that the ring is started before processing it. Because, if the STOP/RESET command gets processed while mhi_ep_ch_ring_worker() waited for chan->lock, the ring would've been reset.
Cc: stable@vger.kernel.org # 5.19 Fixes: 03c0bb8ec983 ("bus: mhi: ep: Add support for processing channel rings") Signed-off-by: Manivannan Sadhasivam manivannan.sadhasivam@linaro.org --- drivers/bus/mhi/ep/main.c | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-)
diff --git a/drivers/bus/mhi/ep/main.c b/drivers/bus/mhi/ep/main.c index 0bce6610ebf1..2362fcc8b32c 100644 --- a/drivers/bus/mhi/ep/main.c +++ b/drivers/bus/mhi/ep/main.c @@ -730,24 +730,37 @@ static void mhi_ep_ch_ring_worker(struct work_struct *work) list_del(&itr->node); ring = itr->ring;
+ chan = &mhi_cntrl->mhi_chan[ring->ch_id]; + mutex_lock(&chan->lock); + + /* + * The ring could've stopped while we waited to grab the (chan->lock), so do + * a sanity check before going further. + */ + if (!ring->started) { + mutex_unlock(&chan->lock); + kfree(itr); + continue; + } + /* Update the write offset for the ring */ ret = mhi_ep_update_wr_offset(ring); if (ret) { dev_err(dev, "Error updating write offset for ring\n"); + mutex_unlock(&chan->lock); kfree(itr); continue; }
/* Sanity check to make sure there are elements in the ring */ if (ring->rd_offset == ring->wr_offset) { + mutex_unlock(&chan->lock); kfree(itr); continue; }
el = &ring->ring_cache[ring->rd_offset]; - chan = &mhi_cntrl->mhi_chan[ring->ch_id];
- mutex_lock(&chan->lock); dev_dbg(dev, "Processing the ring for channel (%u)\n", ring->ch_id); ret = mhi_ep_process_ch_ring(ring, el); if (ret) {