On Mon, 8 Feb 2021 18:44:22 +0100 Christophe Leroy [off-list ref] wrote:

Le 06/02/2021 à 00:41, Andrew Morton a écrit :

quoted

On Thu, 4 Feb 2021 16:22:39 +0100 Thomas Bogendoerfer [off-list ref] wrote:

quoted

On Thu, Feb 04, 2021 at 09:39:42AM +0800, Huang Pei wrote:

quoted

MIPS page fault path(except huge page) takes 3 exceptions (1 TLB Miss
+ 2 TLB Invalid), butthe second TLB Invalid exception is just
triggered by __update_tlb from do_page_fault writing tlb without
_PAGE_VALID set. With this patch, user space mapping prot is made
young by default (with both _PAGE_VALID and _PAGE_YOUNG set),
and it only take 1 TLB Miss + 1 TLB Invalid exception

Remove pte_sw_mkyoung without polluting MM code and make page fault
delay of MIPS on par with other architecture

Signed-off-by: Huang Pei <redacted>
---
  arch/mips/mm/cache.c    | 30 ++++++++++++++++--------------
  include/linux/pgtable.h |  8 --------
  mm/memory.c             |  3 ---
  3 files changed, 16 insertions(+), 25 deletions(-)

Acked-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>

Andrew, can you take this patch through your tree ?

Sure.  I'll drop Christophe's "mm/memory.c: remove pte_sw_mkyoung()"
(https://lkml.kernel.org/r/f302ef92c48d1f08a0459aaee1c568ca11213814.1612345700.git.christophe.leroy@csgroup.eu)
in favour of this one.

Pitty. My patch was improving page faults on powerpc/32.

How does it do that?  By running pte_mkyoung() for powerpc32?  Such a
change is still valid, isn't it?

That one is only addressing MIPS.

It cleans up core code nicely, by removing a MIPS wart.  We can still
add a ppc32 wart?

Any plan to take the series from Nick 
https://patchwork.kernel.org/project/linux-mm/list/?series=404539 ?

I expect so.  After -rc1, if the churn is settling down and reviewers
are happy enough.