lkml.org 
[lkml]   [2011]   [Aug]   [1]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: RCU: how to suppress warnings from rcu_assign_pointer?
On Sat, Jul 30, 2011 at 08:16:09PM -0700, David Miller wrote:
> From: Eric Dumazet <eric.dumazet@gmail.com>
> Date: Sat, 30 Jul 2011 14:17:05 +0200
>
> > Le vendredi 29 juillet 2011 à 10:50 -0700, Stephen Hemminger a écrit :
> >> Gcc now generates warnings from rcu_assign_pointer when passed the
> >> address of something for example:
> >> rcu_assign_pointer(dev_queue->qdisc, &noop_qdisc);
> >> This warning is harmless and should be surpressed but there maybe
> >> other cases where we want that Gcc warning.
> >>
> >> I tried various combinations of in rcu_assign_pointer macro
> >> #pragma GCC diagnostic push
> >> #pragma GCC diagnostic ignored "-Wlogical-op"
> >> ...
> >> #pragma GCC diagnostic pop
> >> but macro's and pragma's don't nest with the correct scope for
> >> this.
> >>
> >> Maybe some one with more Gcc foo and time to waste could take
> >> a crack at it.
> >
> > I would just remove the test from rcu_assign_pointer().
> >
> > We now have RCU_INIT_POINTER() for places we dont want/afford the
> > smp_wmb()
>
> Agreed:
>
> Acked-by: David S. Miller <davem@davemloft.net>

Very good!!! Many thanks to the three of you!

I have queued this, adding Eric's Signed-off-by. Eric, could you please
reply either giving me explicit permission to do this or telling me that
you would rather that I apply under my own Signed-off-by? (I would guess
that the former is OK with you, but I really do need to know for sure.)
Please see below for the current state of this commit in my on-laptop
-rcu tree.

I have also added a second commit upgrading comment headers to call out
exactly when RCU_INIT_POINTER() may be used.

Thanx, Paul

------------------------------------------------------------------------

rcu: Make rcu_assign_pointer() unconditionally insert a memory barrier

Recent changes to gcc give warning messages on rcu_assign_pointers()'s
checks that allow it to determine when it is OK to omit the memory
barrier. Stephen Hemminger tried a number of gcc tricks to silence
this warning, but #pragmas and CPP macros do not work together in the
way that would be required to make this work.

However, we now have RCU_INIT_POINTER(), which already omits this
memory barrier, and which therefore may be used when assigning NULL to
an RCU-protected pointer that is accessible to readers. This commit
therefore makes rcu_assign_pointer() unconditionally emit the memory
barrier.

Reported-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>

diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h
index ea80396..b2e5fe8 100644
--- a/include/linux/rcupdate.h
+++ b/include/linux/rcupdate.h
@@ -443,9 +443,7 @@ extern int rcu_my_thread_group_empty(void);
})
#define __rcu_assign_pointer(p, v, space) \
({ \
- if (!__builtin_constant_p(v) || \
- ((v) != NULL)) \
- smp_wmb(); \
+ smp_wmb(); \
(p) = (typeof(*v) __force space *)(v); \
})

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
\
 
 \ /
  Last update: 2011-08-01 07:27    [W:0.067 / U:0.644 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site