Messages in this thread | | | Date | Fri, 12 Jun 2009 14:43:50 +0300 | Subject | Re: [PATCH v2] slab,slub: ignore __GFP_WAIT if we're booting or suspending | From | Pekka Enberg <> |
| |
Hi Ben,
On Fri, Jun 12, 2009 at 2:41 PM, Benjamin Herrenschmidt<benh@kernel.crashing.org> wrote: >> That said, Nick and Ingo seem to think special-casing is questionable >> and I haven't had green light for any of the patches yet. The gfp >> sanitization patch adds some overhead to kmalloc() and page allocator >> paths which is obviously a concern. > > Let's wait and see what Linus thinks...
Yup, lets do that.
On Fri, Jun 12, 2009 at 2:41 PM, Benjamin Herrenschmidt<benh@kernel.crashing.org> wrote: >> So while we continue to discuss this, I'd really like to proceed with >> the patch below. At least it should allow people to boot their kernels >> (although it will produce warnings). I really don't want to keep other >> people waiting for us to reach a resolution on this. Are you OK with >> that? > > I don't care -how- we achieve the result I want as long as we achieve > it, which is to remove the need for callers to care. My approach was one > way to do it, I'm sure there's a better one. That's not the point. I'm > too tried now to properly review your patch and I'll need to test it > tomorrow morning, but it looks ok except for the WARN_ON maybe.
OK, the WARN_ON is there because you will get warnings for might_sleep() et al as well.
Pekka
| |