net/mlx5: HWS, change error flow on matcher disconnect
authorYevgeny Kliteynik <kliteyn@nvidia.com>
Thu, 2 Jan 2025 18:14:05 +0000 (20:14 +0200)
committerJakub Kicinski <kuba@kernel.org>
Tue, 7 Jan 2025 00:33:40 +0000 (16:33 -0800)
commit1ce840c7a659aa53a31ef49f0271b4fd0dc10296
tree80de8c5f754dd098456676f92210403b25c3c580
parentcc611ab6c712eaa1ed3fd4321d91e66cfe3245a3
net/mlx5: HWS, change error flow on matcher disconnect

Currently, when firmware failure occurs during matcher disconnect flow,
the error flow of the function reconnects the matcher back and returns
an error, which continues running the calling function and eventually
frees the matcher that is being disconnected.
This leads to a case where we have a freed matcher on the matchers list,
which in turn leads to use-after-free and eventual crash.

This patch fixes that by not trying to reconnect the matcher back when
some FW command fails during disconnect.

Note that we're dealing here with FW error. We can't overcome this
problem. This might lead to bad steering state (e.g. wrong connection
between matchers), and will also lead to resource leakage, as it is
the case with any other error handling during resource destruction.

However, the goal here is to allow the driver to continue and not crash
the machine with use-after-free error.

Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Itamar Gozlan <igozlan@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://patch.msgid.link/20250102181415.1477316-7-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
drivers/net/ethernet/mellanox/mlx5/core/steering/hws/matcher.c