diff mbox series

[v7,1/3] memcontrol: define root_mem_cgroup for CONFIG_MEMCG=n cases

Message ID 2050f8a1bc181a9aaf01e0866e230e23216000f4.1721384771.git.wqu@suse.com (mailing list archive)
State New, archived
Headers show
Series btrfs: try to allocate larger folios for metadata | expand

Commit Message

Qu Wenruo July 19, 2024, 10:28 a.m. UTC
There is an incoming btrfs patchset, which will use @root_mem_cgroup as
the active cgroup to attach metadata folios to its internal btree
inode, so that btrfs can skip the possibly costly charge for the
internal inode which is only accessible by btrfs itself.

However @root_mem_cgroup is not always defined (not defined for
CONFIG_MEMCG=n case), thus all such callers need to do the extra
handling for different CONFIG_MEMCG settings.

So here we add a special macro definition of root_mem_cgroup, making it
to always be NULL.

The advantage of this, other than pulling the pointer definition out,
is that we will avoid wasting global data section space for such
pointer.

Signed-off-by: Qu Wenruo <wqu@suse.com>
---
 include/linux/memcontrol.h | 6 ++++++
 1 file changed, 6 insertions(+)

Comments

Michal Hocko July 19, 2024, 11:13 a.m. UTC | #1
On Fri 19-07-24 19:58:39, Qu Wenruo wrote:
> There is an incoming btrfs patchset, which will use @root_mem_cgroup as
> the active cgroup to attach metadata folios to its internal btree
> inode, so that btrfs can skip the possibly costly charge for the
> internal inode which is only accessible by btrfs itself.
> 
> However @root_mem_cgroup is not always defined (not defined for
> CONFIG_MEMCG=n case), thus all such callers need to do the extra
> handling for different CONFIG_MEMCG settings.
> 
> So here we add a special macro definition of root_mem_cgroup, making it
> to always be NULL.

Isn't just a declaration sufficient? Nothing should really dereference
the pointer anyway.
Qu Wenruo July 19, 2024, 9:58 p.m. UTC | #2
在 2024/7/19 20:43, Michal Hocko 写道:
> On Fri 19-07-24 19:58:39, Qu Wenruo wrote:
>> There is an incoming btrfs patchset, which will use @root_mem_cgroup as
>> the active cgroup to attach metadata folios to its internal btree
>> inode, so that btrfs can skip the possibly costly charge for the
>> internal inode which is only accessible by btrfs itself.
>>
>> However @root_mem_cgroup is not always defined (not defined for
>> CONFIG_MEMCG=n case), thus all such callers need to do the extra
>> handling for different CONFIG_MEMCG settings.
>>
>> So here we add a special macro definition of root_mem_cgroup, making it
>> to always be NULL.
>
> Isn't just a declaration sufficient? Nothing should really dereference
> the pointer anyway.
>
That can pass the compile, but waste the extra bytes for the pointer in
the data section, even if no one is utilizing that pointer.

Thanks,
Qu
Michal Hocko July 22, 2024, 7:32 a.m. UTC | #3
On Sat 20-07-24 07:28:45, Qu Wenruo wrote:
> 
> 
> 在 2024/7/19 20:43, Michal Hocko 写道:
> > On Fri 19-07-24 19:58:39, Qu Wenruo wrote:
> > > There is an incoming btrfs patchset, which will use @root_mem_cgroup as
> > > the active cgroup to attach metadata folios to its internal btree
> > > inode, so that btrfs can skip the possibly costly charge for the
> > > internal inode which is only accessible by btrfs itself.
> > > 
> > > However @root_mem_cgroup is not always defined (not defined for
> > > CONFIG_MEMCG=n case), thus all such callers need to do the extra
> > > handling for different CONFIG_MEMCG settings.
> > > 
> > > So here we add a special macro definition of root_mem_cgroup, making it
> > > to always be NULL.
> > 
> > Isn't just a declaration sufficient? Nothing should really dereference
> > the pointer anyway.
> > 
> That can pass the compile, but waste the extra bytes for the pointer in
> the data section, even if no one is utilizing that pointer.

Are you sure that the mere declaration will be defined in the data section?
diff mbox series

Patch

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 030d34e9d117..ae5c78719454 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -346,6 +346,12 @@  enum page_memcg_data_flags {
 
 #define __FIRST_OBJEXT_FLAG	(1UL << 0)
 
+/*
+ * For CONFIG_MEMCG=n case, still define a root_mem_cgroup, but that will
+ * always be NULL and not taking any global data section space.
+ */
+#define root_mem_cgroup		(NULL)
+
 #endif /* CONFIG_MEMCG */
 
 enum objext_flags {