Re: [PATCH] Immunize rcu_dereference() against crazy compiler writers

2007-07-17 Thread Paul E. McKenney
On Tue, Jul 17, 2007 at 02:46:00AM -0700, Andrew Morton wrote: > On Wed, 11 Jul 2007 18:00:58 -0700 "Paul E. McKenney" <[EMAIL PROTECTED]> > wrote: > > > Turns out that compiler writers are a bit more aggressive about optimizing > > than one might expect. This patch prevents a number of such opt

Re: [PATCH] Immunize rcu_dereference() against crazy compiler writers

2007-07-17 Thread Andrew Morton
On Wed, 11 Jul 2007 18:00:58 -0700 "Paul E. McKenney" <[EMAIL PROTECTED]> wrote: > Turns out that compiler writers are a bit more aggressive about optimizing > than one might expect. This patch prevents a number of such optimizations > from messing up rcu_deference(). This is not merely a theore

Re: [PATCH] Immunize rcu_dereference() against crazy compiler writers

2007-07-12 Thread Josh Triplett
On Wed, 2007-07-11 at 18:00 -0700, Paul E. McKenney wrote: > Turns out that compiler writers are a bit more aggressive about optimizing > than one might expect. This patch prevents a number of such optimizations > from messing up rcu_deference(). This is not merely a theoretical > problem, as evi

Re: [PATCH] Immunize rcu_dereference() against crazy compiler writers

2007-07-12 Thread Paul E. McKenney
On Thu, Jul 12, 2007 at 04:03:19PM +0200, Andi Kleen wrote: > "Paul E. McKenney" <[EMAIL PROTECTED]> writes: > > > Turns out that compiler writers are a bit more aggressive about optimizing > > than one might expect. This patch prevents a number of such optimizations > > from messing up rcu_defer

Re: [PATCH] Immunize rcu_dereference() against crazy compiler writers

2007-07-12 Thread Andi Kleen
"Paul E. McKenney" <[EMAIL PROTECTED]> writes: > Turns out that compiler writers are a bit more aggressive about optimizing > than one might expect. This patch prevents a number of such optimizations > from messing up rcu_deference(). This is not merely a theoretical > problem, as evidenced by t