[openib-general] Re: [PATCH 9 of 20] ipath - char devices for diagnostics and lightweight subnet management

[Roland Dreier]

OK, now I can see the change you made:

 > +atomic_t ipath_sma_alive;
 > +DEFINE_SPINLOCK(ipath_sma_lock);	/* SMA receive */

So why is ipath_sma_alive an atomic_t (and why isn't it static)?
You never modify ipath_sma_alive outside of your spinlock, so I don't
see what having it be atomic buys you.

 - R.