On Thu, Nov 12, 2015 at 06:04:27PM +0300, Sergei Shtylyov wrote: > Hello. > > On 11/12/2015 5:31 PM, Wolfram Sang wrote: > > >From: Wolfram Sang > > > >After making sure to reinit the HW and clear interrupts in the timeout > >case, we know that interrupts are always disabled in the sections > >protected by the spinlock. > > What about SMP? Spinlocks are mostly necessary for the SMP support, their > tole isn't limited to disabling interrupts... We have per-adapter locks in the core. Once a transfer is started, you can't access the adapter unless the transfer is finished. But then, all interrupts are off again.