[SRU X] [PATCH 0/1] NVMe polling on timeout

Guilherme G. Piccoli gpiccoli at canonical.com
Fri Dec 21 14:50:16 UTC 2018


On 21/12/2018 11:59, Kleber Souza wrote:
> 
> Hi Guilherme,
> 
> Have we run any I/O stressing test to make sure it didn't introduce at
> least a easily triggered regression?
> 

Sure Kleber, we have exercised this patch in AWS instance for a while -
it does not show any issues, but if an Abort happens, it may prevent
more serious issues (like a RO remount of rootfs). It's a like safe
measure introduced by the nvme maintainer.

Thanks,


Guilherme


> 
> Thanks,
> 
> Kleber
> 
> 



More information about the kernel-team mailing list