[RFC,0/7] mm: PG_reserved cleanups and documentation

Message

David Hildenbrand Dec. 5, 2018, 12:28 p.m. UTC

I was recently going over all users of PG_reserved. Short story: it is
difficult and sometimes not really clear if setting/checking for
PG_reserved is only a relict from the past. Easy to break things.

I had way more cleanups in this series inititally,
but some architectures take PG_reserved as a way to apply a different
caching strategy (for MMIO pages). So I decided to only include the most
obvious changes (that are less likely to break something).

So let's see if the documentation update for PG_reserved I crafted
actually covers most cases or if there is plenty more.

Most notably, for device memory we can hopefully soon stop setting
it PG_reserved

I only briefly tested this on s390x.

David Hildenbrand (7):
  agp: efficeon: no need to set PG_reserved on GATT tables
  s390/vdso: don't clear PG_reserved
  powerpc/vdso: don't clear PG_reserved
  riscv/vdso: don't clear PG_reserved
  m68k/mm: use __ClearPageReserved()
  arm64: kexec: no need to ClearPageReserved()
  mm: better document PG_reserved

 arch/arm64/kernel/machine_kexec.c |  1 -
 arch/m68k/mm/memory.c             |  2 +-
 arch/powerpc/kernel/vdso.c        |  2 --
 arch/riscv/kernel/vdso.c          |  1 -
 arch/s390/kernel/vdso.c           |  2 --
 drivers/char/agp/efficeon-agp.c   |  2 --
 include/linux/page-flags.h        | 18 ++++++++++++++++--
 7 files changed, 17 insertions(+), 11 deletions(-)

Comments

Michal Hocko Dec. 5, 2018, 12:56 p.m. UTC | #1

On Wed 05-12-18 13:28:44, David Hildenbrand wrote:
[...]
> Most notably, for device memory we can hopefully soon stop setting
> it PG_reserved

I am busy as hell so I am not likely to look at specific patche this
week. But could you be more specific why we need to get rid of other
PG_reserved users before we can do so for device memory?

I am all for removing relicts because they just confuse people but I
fail to see any relation here.

David Hildenbrand Dec. 5, 2018, 1:04 p.m. UTC | #2

On 05.12.18 13:56, Michal Hocko wrote:
> On Wed 05-12-18 13:28:44, David Hildenbrand wrote:
> [...]
>> Most notably, for device memory we can hopefully soon stop setting
>> it PG_reserved
> 
> I am busy as hell so I am not likely to look at specific patche this
> week. But could you be more specific why we need to get rid of other
> PG_reserved users before we can do so for device memory?
> 

No worries, this has time.

For device memory, nothing should really be needed. I am only collecting
and docuumenting users and this is one user soon to go (eventually) :)

> I am all for removing relicts because they just confuse people but I
> fail to see any relation here.
> 

It's really only "why is this patch set not bigger", nothing related to
device memory actually :)