Re: [RFC patch] introduce sys_membarrier(): process-wide memorybarrier (v9)

[Steven Rostedt]

On Thu, 2010-02-25 at 12:51 -0500, Mathieu Desnoyers wrote:

> But... either way we chose, we can extend the system call flags and parameters
> as needed, so I think it really should not be part of this initial
> implementation.

I agree here too.

If you have two different tasks doing lockless RCU or what not on shared
memory, it's best to stick with the mb() on the reader side. Yeah, it
makes the performance go down, but heck, I'm really worried about the
crazy complexity that wound need to go into the kernel to prevent this.

-- Steve


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/