net/mlx5e: Fix handling of compressed CQEs in case of low NAPI budget

[ Upstream commit 9df86bd ]

When CQE compression is enabled, compressed CQEs use the following
structure: a title is followed by one or many blocks, each containing 8
mini CQEs (except the last, which may contain fewer mini CQEs).

Due to NAPI budget restriction, a complete structure is not always
parsed in one NAPI run, and some blocks with mini CQEs may be deferred
to the next NAPI poll call - we have the mlx5e_decompress_cqes_cont call
in the beginning of mlx5e_poll_rx_cq. However, if the budget is
extremely low, some blocks may be left even after that, but the code
that follows the mlx5e_decompress_cqes_cont call doesn't check it and
assumes that a new CQE begins, which may not be the case. In such cases,
random memory corruptions occur.

An extremely low NAPI budget of 8 is used when busy_poll or busy_read is
active.

This commit adds a check to make sure that the previous compressed CQE
has been completely parsed after mlx5e_decompress_cqes_cont, otherwise
it prevents a new CQE from being fetched in the middle of a compressed
CQE.

This commit fixes random crashes in __build_skb, __page_pool_put_page
and other not-related-directly places, that used to happen when both CQE
compression and busy_poll/busy_read were enabled.

Fixes: 7219ab3 ("net/mlx5e: CQE compression")
Signed-off-by: Maxim Mikityanskiy <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>

drivers/net/ethernet/mellanox/mlx5/core/en_rx.c

@@ -1267,8 +1267,11 @@ int mlx5e_poll_rx_cq(struct mlx5e_cq *cq, int budget)
 	if (unlikely(!test_bit(MLX5E_RQ_STATE_ENABLED, &rq->state)))
 		return 0;
 
-	if (cq->decmprs_left)
+	if (cq->decmprs_left) {
 		work_done += mlx5e_decompress_cqes_cont(rq, cq, 0, budget);
+		if (cq->decmprs_left || work_done >= budget)
+			goto out;
+	}
 
 	cqe = mlx5_cqwq_get_cqe(&cq->wq);
 	if (!cqe) {