Peter Zijlstra <a.p.zijlstra@xxxxxxxxx> wrote: > + * This is the default implementation, which uses atomic64_t. Which is > + * rather pointless. The whole point behind local64_t is that some processors > + * can perform atomic adds and subtracts in a manner which is atomic wrt IRQs > + * running on this CPU. local64_t allows exploitation of such capabilities. Interesting... What FRV does in atomic64-ops.S should probably be rebranded local64_t, and atomic64_t ops be based on that in non-SMP mode. What I did on FRV was to emulate LL/ST instructions using some of the excessive numbers of conditional bits to do so - but it only works on UP systems. David -- To unsubscribe from this list: send the line "unsubscribe linux-arch" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html