Re: [PATCH 0/3] jump label: updates for 2.6.37

[Andi Kleen <andi@firstfloor.org>]

On Wed, Nov 24, 2010 at 09:29:22AM +0100, Peter Zijlstra wrote:
> On Tue, 2010-11-23 at 18:10 -0500, Steven Rostedt wrote:
> 
> > Anyway, I just tried what you explained with the current kernel, with
> > and without jump labels and, without jump labels, the module has its
> > kmalloc tracepoint traced, but with jump labels it does not. So we can
> > treat this as a regression, which is something that can go into an -rc.
> > 
> > The change log must state that this _is_ a regression, or Linus may not
> > accept it.
> 
> I really dislike the first patch... Preferably I'd simply fully revert
> all the jump-label stuff and try again next round with a saner
> interface.

Agreed.

> 
> There's a really good simple fix for this, simply disable the gcc
> trickery for .37 and use the fallback.
> 
> Then for .38, mandate the key type to be atomic_t * and switch to the
> SWITCH_POINT() interface from hpa.

This would also allow getting rid of the hash tables and use
binary search, which is imho much cleaner and simpler.

I will refresh my older patches to do that.

-Andi

-- 
ak@linux.intel.com -- Speaking for myself only.