I am learning the new RCU based lockless synchronization approach in the linux kernel.
I already have a Kernel module which maintains hash_table (kernel hash api).
Until now I was using spin_lock_irqsave() in order to synchronize process context code trying to modify this hash while it may be getting updated in the softirq context.
I realised that because of this I am curtailing the benifits of softirq as I am serializing the access to resource, when softirq are invented to do just that.
So, now I think RCU is the best suited mechanism for my scenario and I can use hash_add_rcu hash_for_each_rcu to take benefits of RCU.
However will this mechanism work correctly with updaters (writers) in process context concurrently accessing with writers in softirq context?