Re: [PATCH V2] mm: Introduce GFP_PGTABLE

On 01/16/2019 06:14 PM, Michal Hocko wrote:

On Wed 16-01-19 04:30:18, Matthew Wilcox wrote:

quoted

On Wed, Jan 16, 2019 at 07:57:03AM +0100, Michal Hocko wrote:

quoted

On Wed 16-01-19 11:51:32, Anshuman Khandual wrote:

quoted

All architectures have been defining their own PGALLOC_GFP as (GFP_KERNEL |
__GFP_ZERO) and using it for allocating page table pages. This causes some
code duplication which can be easily avoided. GFP_KERNEL allocated and
cleared out pages (__GFP_ZERO) are required for page tables on any given
architecture. This creates a new generic GFP flag flag which can be used
for any page table page allocation. Does not cause any functional change.

GFP_PGTABLE is being added into include/asm-generic/pgtable.h which is the
generic page tabe header just to prevent it's potential misuse as a general
allocation flag if included in include/linux/gfp.h.

I haven't reviewed the patch yet but I am wondering whether this is
really worth it without going all the way down to unify the common code
and remove much more code duplication. Or is this not possible for some
reason?

Exactly what I suggested doing in response to v1.

Also, the approach taken here is crazy.  x86 has a feature that no other
architecture has bothered to implement yet -- accounting page tables
to the process.  Yet instead of spreading that goodness to all other
architectures, Anshuman has gone to more effort to avoid doing that.

Yes, I believe the only reason this is x86 only is that each arch would
have to be tweaked separately. So a cleanup in _that_ regard would be
helpful. There is no real reason to have ptes accounted only for x86.
There might be some exceptions but well, our asm-generic allows to opt
in for generic implementation or override it with a special one. The
later should be an exception rather than the rule.

Fair enough. So we seem to have agreement over __GFP_ACCOUNT for user page
tables but not for the kernel. But should we accommodate __GFP_RETRY_MAYFAIL
or drop them altogether (including multi order allocation requests) ?