Re: [PATCH] aacraid: reply queue mapping to CPUs based of IRQ affinity

On 13/04/2023 22:52, Sagar.Biradar@microchip.com wrote:

[snip]

Yes we have reserved commands, that originate from within the driver. We rely on the reply_map mechanism (from the original patch) to get interrupt vector for the reserved commands too.

Thanks Sagar

Also, there is this patch which addresses the concerns John Garry raised. https://urldefense.com/v3/__https://lore.kernel.org/lkml/20220929033428.2594...

This patch explains how the coordination happens when a CPU goes offline. IPI can be read InterProcessor Interrupt. The request shall be completed from the CPU where it is running when the original CPU goes offline.

Thanks Sagar

Can you please use standard mailing list practices in your replies, i.e. quote original mail in the reply here? I don't know what was added to the thread in this latest reply.

Thanks, John