Re: [PATCH v1] mpt3sas: Fix calltrace observed while running IO & host reset

[Bart Van Assche]

On 06/21/18 22:35, Sreekanth Reddy wrote:
No, lock_system_sleep() is not inserted in the interrupt context. we
have inserted it in .resume() call back function just before issuing
the IOC reset.

That's the wrong place to insert a lock_system_sleep() call. Please have 
a look at drivers/scsi/scsi_transport_spi.c for an example of how to use 
that function.

Thanks,

Bart.