[PATCH] nvme-tcp: delay error recovery after link drop

[Christoph Hellwig]

On Wed, Sep 07, 2022 at 10:57:20AM +0300, Sagi Grimberg wrote:
>> I'm not sure the NVMe/TCP host stack obeys this rule.
>>
>> The problem is, when the host fails to follow this rule, some NVMe/TCP 
>> controllers are found to corrupt data during cable pull tests.
>
> The delay needs to come explicitly from the controller, not guessed
> blindly by the host.

No timeout can work properly.  What we need is a proper handshake.