On Wed, 3 Jul 2024 09:56:45 -0700 Mina Almasry wrote:

quoted

Is this really sufficient in terms of locking? @binding is not
RCU-protected and neither is the reader guaranteed to be in
an RCU critical section. Actually the "reader" tries to take a ref
and use this struct so it's not even a pure reader.

Let's add a lock or use one of the existing locks

Can we just use rtnl_lock() for this synchronization? It seems it's
already locked everywhere that access mp_params.mp_priv, so the
WRITE/READ_ONCE are actually superfluous. Both the dmabuf bind/unbind
already lock rtnl_lock, and the only other place that access
mp_params.mp_priv is page_pool_init(). I think it's reasonable to
assume rtnl_lock is also held during page_pool_init, no? AFAICT it
would be very weird for some code path to be reconfiguring the driver
page_pools without holding rtnl_lock?

What I wanna do here is delete the incorrect comment, remove the
READ/WRITE_ONCE, and maybe add a DEBUG_NET_WARN_ON(!rtnl_is_locked())
in mp_dmabuf_devmem_init() but probably that is too defensive.

The only concern I have is driver error recovery paths. They may be
async and may happen outside of rtnl_lock. Same situation we have
with the queue <> NAPI <> IRQ mapping helpers. queue <> NAPI <> IRQ
helpers require rtnl_lock today, and Intel recently had a number of
fixes because that complicates their error recovery paths.

But I guess any locking here will take non-trivial amount of analysis.
Let's go with rtnl_lock, that's fine.