Re: [BUG] set_mempolicy(MPOL_INTERLEAV) cause kernel panic

["KAMEZAWA Hiroyuki" <kamezawa.hiroyu@jp.fujitsu.com>]

KAMEZAWA Hiroyuki wrote:
> Andrew Morton wrote:
>> On Fri, 24 Jul 2009 15:51:51 -0700 (PDT)
>> David Rientjes <rientjes@google.com> wrote:
>
>> afaik we don't have a final patch for this.  I asked Motohiro-san about
>> this and he's proposing that we revert the offending change (which one
>> was it?) if nothing gets fixed soon - the original author is on a
>> lengthy vacation.
>>
>>
>> If we _do_ have a patch then can we start again?  Someone send out the
>> patch
>> and let's take a look at it.
> Hmm, like this ? (cleaned up David's one because we shouldn't have
> extra nodemask_t on stack.)
>
> Problems are
>   - rebind() is maybe broken but no good idea.
>    (but it seems to be broken in old kernels
>   - Who can test this is only a user who has possible node on SRAT.
>

> +		/* should we call is_valid_nodemask() here ?*/
>  		if (pol->flags & MPOL_F_RELATIVE_NODES)
>  			mpol_relative_nodemask(&cpuset_context_nmask, nodes,
> -					       &cpuset_current_mems_allowed);
> +					       &cpuset_context_nmask);
Sorry this part is buggy.

But to fix this, we use extra nodemask here and this patch will
allocate 3 nodemasks on stack!
Then, here is a much easier fix. for trusting cpuset more.

==
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

task->mems_allowed is guaranteed to be only includes valid nodes when
it is set under cpuset. but, at init, all possible nodes are included.
fix it.

And at cpuset-rebind, caluculated result can be a invaild one.
In that case, trust cpuset's one

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
---
 init/main.c    |    4 ++--
 mm/mempolicy.c |    7 ++++++-
 2 files changed, 8 insertions(+), 3 deletions(-)

Index: mmotm-2.6.31-Jul16/init/main.c
===================================================================
--- mmotm-2.6.31-Jul16.orig/init/main.c
+++ mmotm-2.6.31-Jul16/init/main.c
@@ -855,9 +855,9 @@ static int __init kernel_init(void * unu
 	lock_kernel();

 	/*
-	 * init can allocate pages on any node
+	 * init can allocate pages on any online node
 	 */
-	set_mems_allowed(node_possible_map);
+	set_mems_allowed(node_state[N_HIGH_MEMORY]);
 	/*
 	 * init can run on any cpu.
 	 */
Index: mmotm-2.6.31-Jul16/mm/mempolicy.c
===================================================================
--- mmotm-2.6.31-Jul16.orig/mm/mempolicy.c
+++ mmotm-2.6.31-Jul16/mm/mempolicy.c
@@ -290,7 +290,12 @@ static void mpol_rebind_nodemask(struct
 			    *nodes);
 		pol->w.cpuset_mems_allowed = *nodes;
 	}
-
+	/*
+	 * tmp can be invalid ...just use cpuset's one in that case.
+	 */
+	if (nodes_empty(tmp) ||
+	    ((pol->mode == MPOL_BIND) && !is_valid_nodemask(&tmp)))
+		tmp = *nodes;
 	pol->v.nodes = tmp;
 	if (!node_isset(current->il_next, tmp)) {
 		current->il_next = next_node(current->il_next, tmp);





--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>