net/mlx5: Fix fatal error handling during device load

Currently, in case of fatal error during mlx5_load_one(), we cannot
enter error state until mlx5_load_one() is finished, what can take
several minutes until commands will get timeouts, because these commands
can't be processed due to the fatal error.
Fix it by setting dev->state as MLX5_DEVICE_STATE_INTERNAL_ERROR before
requesting the lock.

Fixes: c1d4d2e92a ("net/mlx5: Avoid calling sleeping function by the health poll thread")
Signed-off-by: Shay Drory <shayd@mellanox.com>
Reviewed-by: Moshe Shemesh <moshe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
This commit is contained in:
Shay Drory 2020-05-07 09:32:53 +03:00 committed by Saeed Mahameed
parent 42ea9f1b5c
commit b6e0b6bebe

View file

@ -192,15 +192,23 @@ static bool reset_fw_if_needed(struct mlx5_core_dev *dev)
void mlx5_enter_error_state(struct mlx5_core_dev *dev, bool force)
{
bool err_detected = false;
/* Mark the device as fatal in order to abort FW commands */
if ((check_fatal_sensors(dev) || force) &&
dev->state == MLX5_DEVICE_STATE_UP) {
dev->state = MLX5_DEVICE_STATE_INTERNAL_ERROR;
err_detected = true;
}
mutex_lock(&dev->intf_state_mutex);
if (dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR)
goto unlock;
if (!err_detected && dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR)
goto unlock;/* a previous error is still being handled */
if (dev->state == MLX5_DEVICE_STATE_UNINITIALIZED) {
dev->state = MLX5_DEVICE_STATE_INTERNAL_ERROR;
goto unlock;
}
if (check_fatal_sensors(dev) || force) {
if (check_fatal_sensors(dev) || force) { /* protected state setting */
dev->state = MLX5_DEVICE_STATE_INTERNAL_ERROR;
mlx5_cmd_flush(dev);
}