On Fri, Aug 28, 2015 at 06:06:32PM -0400, Tejun Heo wrote:

On Fri, Aug 28, 2015 at 04:44:32PM -0400, Tejun Heo wrote:

quoted

Ah, cool, so it was a bug from slub.  Punting to return path still has
some niceties but if we can't consistently get rid of stack
consumption it's not that attractive.  Let's revisit it later together
with hard limit reclaim.

So, I can't check right now but I'm pretty sure I was using SLAB on my
test config, so this issue may exist there too.

Yeah, SLAB is broken too. It was accidentally broken by commit
4167e9b2cf10 ("mm: remove GFP_THISNODE"), which among other things made
SLAB filter out __GFP_WAIT from gfp flags when probing a NUMA node. I'll
take a look what we can do with that.

Thanks,
Vladimir

--
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>