Re: [RFC] Tagging of vmalloc pages for supporting the pmalloc allocator

From: Michal Hocko
Date: Fri Aug 04 2017 - 04:12:48 EST


On Fri 04-08-17 11:02:46, Igor Stoppa wrote:
>
>
> On 03/08/17 18:15, Michal Hocko wrote:
>
> > I would check the one where we have mapping. It is rather unlikely
> > vmalloc users would touch this one.
>
> That was also the initial recommendation from Jerome Glisse, but it
> seemed unusable, because of the related comment.
>
> I should have asked for clarifications back then :-(
>
> But it's never too late ...
>
>
> struct page {
> /* First double word block */
> unsigned long flags; /* Atomic flags, some possibly
> * updated asynchronously */
> union {
> struct address_space *mapping; /* If low bit clear, points to
> * inode address_space, or NULL.
> * If page mapped as anonymous
> * memory, low bit is set, and
> * it points to anon_vma object:
> * see PAGE_MAPPING_ANON below.
> */
> ...
> }
>
> mapping seems to be used exclusively in 2 ways, based on the value of
> its lower bit.

Not really. The above applies to LRU pages. Please note that Slab pages
use s_mem and huge pages use compound_mapcount. If vmalloc pages are
using none of those already you can add a new field there.

--
Michal Hocko
SUSE Labs