On Fri, Jun 6, 2014 at 10:11 AM, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
>
> +config ARCH_NO_ATOMIC_RMW
> + def_bool y
> + depends on PARISC || SPARC32 || METAG_ATOMICITY_LOCK1 || (TILE && !TILEGX) || (ARC && !ARC_HAS_LLSC)
Ugh. We've had these kinds of things before, and they are broken and
nasty to maintain.
Just make it
config ARCH_SUPPORTS_ATOMIC_RMW
bool
which defaults to no. And then make MUTEX_SPIN_ON_OWNER depend on that.
And then we can add "select ARCH_SUPPORTS_ATOMIC_RMW" to the few
architectures we (a) care about and (b) know work. So start with x86,
arm, powerpc and sparc64, and then the rest can just add their own
oneliners if they care.
Remember, most people really won't ever care about this, simply
because it only matters if you have enough CPU's for the whole
spinning thing to make a noticeable difference. So missing some odd
architecture _really_ doesn't matter.
And we really really *really* shouldn't have these kinds of "random
really odd architecture details" in the generic code, not even if it's
something as specific as kernel/Kconfig.locks.
Linus
Linus
--
To unsubscribe from this list: send the line "unsubscribe linux-parisc" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
