2019-12-10 13:50:19

by chengkaitao

[permalink] [raw]
Subject: [PATCH] mm: cleanup some useless code

From: Kaitao Cheng <[email protected]>

There is a duplicate PageTransHuge, because hpage_nr_pages have
the same one.

Signed-off-by: Kaitao Cheng <[email protected]>
---
mm/memcontrol.c | 7 +++----
1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index bc01423277c5..870284d3ee9d 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -6678,7 +6678,6 @@ void mem_cgroup_migrate(struct page *oldpage, struct page *newpage)
{
struct mem_cgroup *memcg;
unsigned int nr_pages;
- bool compound;
unsigned long flags;

VM_BUG_ON_PAGE(!PageLocked(oldpage), oldpage);
@@ -6700,8 +6699,7 @@ void mem_cgroup_migrate(struct page *oldpage, struct page *newpage)
return;

/* Force-charge the new page. The old one will be freed soon */
- compound = PageTransHuge(newpage);
- nr_pages = compound ? hpage_nr_pages(newpage) : 1;
+ nr_pages = hpage_nr_pages(newpage);

page_counter_charge(&memcg->memory, nr_pages);
if (do_memsw_account())
@@ -6711,7 +6709,8 @@ void mem_cgroup_migrate(struct page *oldpage, struct page *newpage)
commit_charge(newpage, memcg, false);

local_irq_save(flags);
- mem_cgroup_charge_statistics(memcg, newpage, compound, nr_pages);
+ mem_cgroup_charge_statistics(memcg, newpage, PageTransHuge(newpage),
+ nr_pages);
memcg_check_events(memcg, newpage);
local_irq_restore(flags);
}
--
2.20.1


2019-12-10 14:08:51

by Michal Hocko

[permalink] [raw]
Subject: Re: [PATCH] mm: cleanup some useless code

On Tue 10-12-19 05:49:10, chengkaitao wrote:
> From: Kaitao Cheng <[email protected]>
>
> There is a duplicate PageTransHuge, because hpage_nr_pages have
> the same one.

I am not a great fan of changes like these - a minor code churn without
a clear benefit - but in this particular case I do agree that there is
an improvement because we have two states to represent the same thing.
I have hard time believing this makes a more optimal code as
PageTransHuge is a trivial check so it doesn't make much sense to
replicate the state locally.

That being said the changelog could be improved a bit. What about
something like the following?
"
Compound pages handling in mem_cgroup_migrate is more convoluted than
necessary. The state is duplicated in compound variable and the same
could be achieved by PageTransHuge check which is trivial and
hpage_nr_pages is already PageTransHuge aware.

It is much simpler to just use hpage_nr_pages for nr_pages and replace
the local variable by PageTransHuge check directly
"

> Signed-off-by: Kaitao Cheng <[email protected]>

To the change
Acked-by: Michal Hocko <[email protected]>

Thanks!
> ---
> mm/memcontrol.c | 7 +++----
> 1 file changed, 3 insertions(+), 4 deletions(-)
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index bc01423277c5..870284d3ee9d 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -6678,7 +6678,6 @@ void mem_cgroup_migrate(struct page *oldpage, struct page *newpage)
> {
> struct mem_cgroup *memcg;
> unsigned int nr_pages;
> - bool compound;
> unsigned long flags;
>
> VM_BUG_ON_PAGE(!PageLocked(oldpage), oldpage);
> @@ -6700,8 +6699,7 @@ void mem_cgroup_migrate(struct page *oldpage, struct page *newpage)
> return;
>
> /* Force-charge the new page. The old one will be freed soon */
> - compound = PageTransHuge(newpage);
> - nr_pages = compound ? hpage_nr_pages(newpage) : 1;
> + nr_pages = hpage_nr_pages(newpage);
>
> page_counter_charge(&memcg->memory, nr_pages);
> if (do_memsw_account())
> @@ -6711,7 +6709,8 @@ void mem_cgroup_migrate(struct page *oldpage, struct page *newpage)
> commit_charge(newpage, memcg, false);
>
> local_irq_save(flags);
> - mem_cgroup_charge_statistics(memcg, newpage, compound, nr_pages);
> + mem_cgroup_charge_statistics(memcg, newpage, PageTransHuge(newpage),
> + nr_pages);
> memcg_check_events(memcg, newpage);
> local_irq_restore(flags);
> }
> --
> 2.20.1

--
Michal Hocko
SUSE Labs