[PATCH 4/6] nvmet: don't defer passthrough commands with trivial effects to the workqueue

Sagi Grimberg sagi at grimberg.me
Sun Dec 25 02:14:51 PST 2022


Reviewed-by: Sagi Grimberg <sagi at grimberg.me>


On 12/23/22 09:18, Christoph Hellwig wrote:
> Mask out the "Command Supported" and "Logical Block Content Change" bits
> and only defer execution of commands that have non-trivial effects to
> the workqueue for synchronous execution.  This allows to execture admin

allows to execute admin

> commands asynchronously on controllers that provide a Command Supported
> and Effects log page, and will keep allowing to execute Write commands
> asynchronously once command effects on I/O commands are taken into
> account.
> 
> Fixes: c1fef73f793b ("nvmet: add passthru code to process commands")
> Reviewed-by: Kanchan Joshi <joshi.k at samsung.com>
> Signed-off-by: Christoph Hellwig <hch at lst.de>
> ---
>   drivers/nvme/target/passthru.c | 11 +++++------
>   1 file changed, 5 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/nvme/target/passthru.c b/drivers/nvme/target/passthru.c
> index 79af5140af8bfe..adc0958755d66f 100644
> --- a/drivers/nvme/target/passthru.c
> +++ b/drivers/nvme/target/passthru.c
> @@ -334,14 +334,13 @@ static void nvmet_passthru_execute_cmd(struct nvmet_req *req)
>   	}
>   
>   	/*
> -	 * If there are effects for the command we are about to execute, or
> -	 * an end_req function we need to use nvme_execute_passthru_rq()
> -	 * synchronously in a work item seeing the end_req function and
> -	 * nvme_passthru_end() can't be called in the request done callback
> -	 * which is typically in interrupt context.
> +	 * If a command needs post-execution fixups, or there are any
> +	 * non-trivial effects, make sure to execute the command synchronously
> +	 * in a workqueue so that nvme_passthru_end gets called.
>   	 */
>   	effects = nvme_command_effects(ctrl, ns, req->cmd->common.opcode);
> -	if (req->p.use_workqueue || effects) {
> +	if (req->p.use_workqueue ||
> +	    (effects & ~(NVME_CMD_EFFECTS_CSUPP | NVME_CMD_EFFECTS_LBCC))) {
>   		INIT_WORK(&req->p.work, nvmet_passthru_execute_cmd_work);
>   		req->p.rq = rq;
>   		queue_work(nvmet_wq, &req->p.work);

Other than that,

Reviewed-by: Sagi Grimberg <sagi at grimberg.me>



More information about the Linux-nvme mailing list