Re: rcu_init allocation failure trace on bootup

["Paul E. McKenney" <paulmck@linux.vnet.ibm.com>]

On Mon, May 06, 2013 at 03:26:53PM -0400, Dave Jones wrote:
> On Mon, May 06, 2013 at 12:05:22PM -0700, Paul E. McKenney wrote:
>  > On Mon, May 06, 2013 at 01:56:47PM -0400, Dave Jones wrote:
>  > > [    0.000000]  RCU debugfs-based tracing is enabled.
>  > > [    0.000000]  RCU lockdep checking is enabled.
>  > > [    0.000000] ------------[ cut here ]------------
>  > > [    0.000000] WARNING: at mm/nobootmem.c:215 ___alloc_bootmem_nopanic+0x3a/0x7d()
>  > > [    0.000000] Modules linked in:
>  > > [    0.000000] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 3.9.0+ #128
>  > > [    0.000000] Hardware name:                  /D975XBX, BIOS BX97510J.86A.1209.2006.0601.1340 06/01/2006
>  > > [    0.000000]  ffffffff819f4246 ffffffff81c01e68 ffffffff816a0db2 ffffffff81c01ea8
>  > > [    0.000000]  ffffffff8104b1a0 ffffffff819e1078 0000000000000000 0000000000000008
>  > > [    0.000000]  ffffffffffffffff 0000000000000040 0000000000000000 ffffffff81c01eb8
>  > > [    0.000000] Call Trace:
>  > > [    0.000000]  [<ffffffff816a0db2>] dump_stack+0x19/0x1b
>  > > [    0.000000]  [<ffffffff8104b1a0>] warn_slowpath_common+0x70/0xa0
>  > > [    0.000000]  [<ffffffff8104b1ea>] warn_slowpath_null+0x1a/0x20
>  > > [    0.000000]  [<ffffffff81ef8914>] ___alloc_bootmem_nopanic+0x3a/0x7d
>  > > [    0.000000]  [<ffffffff81ef8b87>] __alloc_bootmem+0x13/0x39
>  > > [    0.000000]  [<ffffffff81f090ad>] alloc_bootmem_cpumask_var+0x1a/0x21
>  > > [    0.000000]  [<ffffffff81ef1f5a>] rcu_init+0x7f/0x261
>  > > [    0.000000]  [<ffffffff81ef4acb>] ? perf_event_init+0x178/0x181
>  > > [    0.000000]  [<ffffffff81ed6c79>] start_kernel+0x232/0x3d3
>  > > [    0.000000]  [<ffffffff81ed6846>] ? repair_env_string+0x5e/0x5e
>  > > [    0.000000]  [<ffffffff81ed6577>] x86_64_start_reservations+0x2a/0x2c
>  > > [    0.000000]  [<ffffffff81ed6646>] x86_64_start_kernel+0xcd/0xd1
>  > > [    0.000000] ---[ end trace 0acfc82c197165dd ]---
>  > > [    0.000000]  Experimental no-CBs for all CPUs
>  > > [    0.000000]  Experimental no-CBs CPUs: 0-1.
>  > > [    0.000000] NO_HZ: Full dynticks CPUs: 1.
>  > 
>  > Do you have commit fa0f93cd6a (rcu: Don't allocate bootmem from rcu_init())?
>  > If not, could you please try applying it?
> 
> That commit isn't in Linus tree.

You are quite right, I am getting ahead of myself.  Please see below
for the patch.

							Thanx, Paul

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

rcu: Don't allocate bootmem from rcu_init()

When rcu_init() is called we already have slab working, allocating
bootmem at that point results in warnings and an allocation from
slab.  This commit therefore changes alloc_bootmem_cpumask_var() to
alloc_cpumask_var() in rcu_bootup_announce_oddness(), which is called
from rcu_init().

Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>

diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h
index ca6e39c..44b0998 100644
--- a/kernel/rcutree_plugin.h
+++ b/kernel/rcutree_plugin.h
@@ -86,7 +86,7 @@ static void __init rcu_bootup_announce_oddness(void)
 #ifdef CONFIG_RCU_NOCB_CPU
 #ifndef CONFIG_RCU_NOCB_CPU_NONE
 	if (!have_rcu_nocb_mask) {
-		alloc_bootmem_cpumask_var(&rcu_nocb_mask);
+		alloc_cpumask_var(&rcu_nocb_mask, GFP_KERNEL);
 		have_rcu_nocb_mask = true;
 	}
 #ifdef CONFIG_RCU_NOCB_CPU_ZERO