Commit 2808b37b authored by Roy Novich's avatar Roy Novich Committed by Saeed Mahameed
Browse files

net/mlx5: Allow async trigger completion execution on single CPU systems 



For a single CPU system, the kernel thread executing mlx5_cmd_flush()
never releases the CPU but calls down_trylock(&cmd→sem) in a busy loop.
On a single processor system, this leads to a deadlock as the kernel
thread which executes mlx5_cmd_invoke() never gets scheduled. Fix this,
by adding the cond_resched() call to the loop, allow the command
completion kernel thread to execute.

Fixes: 8e715cd6 ("net/mlx5: Set command entry semaphore up once got index free")
Signed-off-by: default avatarAlexander Schmidt <alexschm@de.ibm.com>
Signed-off-by: default avatarRoy Novich <royno@nvidia.com>
Reviewed-by: default avatarMoshe Shemesh <moshe@nvidia.com>
Signed-off-by: default avatarSaeed Mahameed <saeedm@nvidia.com>
parent 15f8f168

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

+8 −3
Original line number Diff line number Diff line
@@ -1770,12 +1770,17 @@ void mlx5_cmd_flush(struct mlx5_core_dev *dev) 
	struct mlx5_cmd *cmd = &dev->cmd;

	int i;



	for (i = 0; i < cmd->max_reg_cmds; i++)

		while (down_trylock(&cmd->sem))

	for (i = 0; i < cmd->max_reg_cmds; i++) {

		while (down_trylock(&cmd->sem)) {

			mlx5_cmd_trigger_completions(dev);

			cond_resched();

		}

	}



	while (down_trylock(&cmd->pages_sem))

	while (down_trylock(&cmd->pages_sem)) {

		mlx5_cmd_trigger_completions(dev);

		cond_resched();

	}



	/* Unlock cmdif */

	up(&cmd->pages_sem);