[PATCH] nvme-fc: change default devloss_tmo to 30s

[PATCH] nvme-fc: change default devloss_tmo to 30s

[James Smart]

Transport default devloss_tmo is 60s. However, both lldd's use a
driver default devloss_tmo of 30s.  Rather having an inconsistency if
the transport value is used, set the transport default to 30s.

Signed-off-by: James Smart <jsmart2021@gmail.com>
CC: Himanshu Madhani <himanshu.madhani@oracle.com>
CC: Arun Easi <aeasi@marvell.com>
---
 drivers/nvme/host/fc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c
index 1921d2195541..0aa3767cbb1c 100644
--- a/drivers/nvme/host/fc.c
+++ b/drivers/nvme/host/fc.c
@@ -25,7 +25,7 @@ enum nvme_fc_queue_flags {
 	NVME_FC_Q_LIVE,
 };
 
-#define NVME_FC_DEFAULT_DEV_LOSS_TMO	60	/* seconds */
+#define NVME_FC_DEFAULT_DEV_LOSS_TMO	30	/* seconds */
 
 struct nvme_fc_queue {
 	struct nvme_fc_ctrl	*ctrl;
-- 
2.26.1


_______________________________________________
linux-nvme mailing list
linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH] nvme-fc: change default devloss_tmo to 30s

[himanshu.madhani]

On 5/1/20 4:45 PM, James Smart wrote:
> Transport default devloss_tmo is 60s. However, both lldd's use a
> driver default devloss_tmo of 30s.  Rather having an inconsistency if
> the transport value is used, set the transport default to 30s.
> 
> Signed-off-by: James Smart <jsmart2021@gmail.com>
> CC: Himanshu Madhani <himanshu.madhani@oracle.com>
> CC: Arun Easi <aeasi@marvell.com>
> ---
>   drivers/nvme/host/fc.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c
> index 1921d2195541..0aa3767cbb1c 100644
> --- a/drivers/nvme/host/fc.c
> +++ b/drivers/nvme/host/fc.c
> @@ -25,7 +25,7 @@ enum nvme_fc_queue_flags {
>   	NVME_FC_Q_LIVE,
>   };
>   
> -#define NVME_FC_DEFAULT_DEV_LOSS_TMO	60	/* seconds */
> +#define NVME_FC_DEFAULT_DEV_LOSS_TMO	30	/* seconds */
>   
>   struct nvme_fc_queue {
>   	struct nvme_fc_ctrl	*ctrl;
> 

Thanks for making it consistent.

Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>

-- 
Himanshu Madhani
Oracle Linux Engineering

_______________________________________________
linux-nvme mailing list
linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH] nvme-fc: change default devloss_tmo to 30s

[Christoph Hellwig]

On Fri, May 01, 2020 at 02:45:49PM -0700, James Smart wrote:
> Transport default devloss_tmo is 60s. However, both lldd's use a
> driver default devloss_tmo of 30s.  Rather having an inconsistency if
> the transport value is used, set the transport default to 30s.

Can we just have an interface that queries the LLDD instead of
relying on the drivers and the transport having the same value?

_______________________________________________
linux-nvme mailing list
linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH] nvme-fc: change default devloss_tmo to 30s

[James Smart]

On 5/1/2020 11:38 PM, Christoph Hellwig wrote:
> On Fri, May 01, 2020 at 02:45:49PM -0700, James Smart wrote:
>> Transport default devloss_tmo is 60s. However, both lldd's use a
>> driver default devloss_tmo of 30s.  Rather having an inconsistency if
>> the transport value is used, set the transport default to 30s.
> Can we just have an interface that queries the LLDD instead of
> relying on the drivers and the transport having the same value?
>
>

we have that already - not queries, but rather transport can provide it. 
Guess I can hard-enforce the driver sets a value and eliminate the 
transport default.

- james


_______________________________________________
linux-nvme mailing list
linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme