On Fri, May 21, 2021 at 7:56 AM Thomas Gleixner [off-list ref] wrote:

Nitesh,

On Thu, May 20 2021 at 20:03, Nitesh Lal wrote:

quoted

On Thu, May 20, 2021 at 5:57 PM Nitesh Lal [off-list ref] wrote:

quoted

I think here to ensure that we are not breaking any of the drivers we have
to first analyze all the existing drivers and understand how they are using
this API.
AFAIK there are three possible scenarios:

- A driver use this API to spread the IRQs
  + For this case we should be safe considering the spreading is naturally
    done from the IRQ subsystem itself.

Forgot to mention another thing in the above case is to determine whether
it is true for all architectures or not as Thomas mentioned.

Yes.

quoted

quoted

- A driver use this API to actually set the hint
  + These drivers should have no functional impact because of this revert

Correct.

quoted

quoted

- Driver use this API to force a certain affinity mask
  + In this case we have to replace the API with the irq_force_affinity()

irq_set_affinity() or irq_set_affinity_and_hint()

Ah yes! my bad. _force_ doesn't check the mask against the online CPUs.
Hmm, I didn't realize that you also added irq_set_affinity_and_hint()
in your last patchset.

quoted

quoted

I can start looking into the individual drivers, however, testing them will
be a challenge.

The only way to do that is to have the core infrastructure added and

Right.

then send patches changing it in the way you think. The relevant
maintainers/developers should be able to tell you when your analysis
went south. :)

Ack will start looking into this.


--
Thanks
Nitesh