[v6,0/1] mm: report per-page metadata information

  Changelog:
v6:
	- Interface changes
	  	- Added per-node nr_page_metadata and
		  nr_page_metadata_boot fields that are exported
		  in /sys/devices/system/node/nodeN/vmstat
		- nr_page_metadata exclusively tracks buddy
		  allocations while nr_page_metadata_boot
		  exclusively tracks memblock allocations.
		- Modified PageMetadata in /proc/meminfo to only
		  include buddy allocations so that it is
		  comparable against MemTotal in /proc/meminfo
		- Removed the PageMetadata field added in
		  /sys/devices/system/node/nodeN/meminfo
	- Addressed bugs reported by the kernel test bot.
	  	- All occurences of __mod_node_page_state have
		  been replaced by mod_node_page_state.
	- Addressed comments from Muchun Song.
	  	- Removed page meta data accouting from
		  mm/hugetlb.c. When CONFIG_SPARSEMEM_VMEMMAP
		  is disabled struct pages should not be returned
		  to buddy.
	- Modified the cover letter with the results and analysis
		- From when memory_hotplug.memmap_on_memory is
		  alternated between 0 and 1.
		- To account for the new interface changes.

v5:
	- Addressed comments from Muchun Song.
		- Fixed errors introduced in v4 when
		  CONFIG_SPARSEMEM_VMEMMAP is disabled by testing
		  against FLATMEM and SPARSEMEM memory models.
		- Handled the condition wherein the allocation of
		  walk.reuse_page fails, by moving NR_PAGE_METADATA
		  update into the clause if( walk.reuse_page ).
		- Removed the usage of DIV_ROUND_UP from
		  alloc_vmemmap_page_list since "end - start" is
		  always a multiple of PAGE_SIZE.
		- Modified alloc_vmemmap_page_list to update
		  NR_PAGE_METADATA once instead of every loop.
v4:
	- Addressed comment from Matthew Wilcox.
		- Used __node_stat_sub_folio and __node_stat_add_folio
		  instead of __mod_node_page_state in mm/hugetlb.c.
		- Used page_pgdat wherever possible in the entire patch.
		- Used DIV_ROUND_UP() wherever possible in the entire
		  patch.
v3:
	- Addressed one comment from Matthew Wilcox.
	  	- In free_page_ext, page_pgdat() is now extracted
		  prior to freeing the memory.
v2:
	- Fixed the three bugs reported by kernel test robot.
	- Enhanced the commit message as recommended by David Hildenbrand.
	- Addressed comments from Matthew Wilcox:
	  	- Simplified alloc_vmemmap_page_list() and
		  free_page_ext() as recommended.
		- Used the appropriate comment style in mm/vmstat.c.
		- Replaced writeout_early_perpage_metadata() with
		  store_early_perpage_metadata() to reduce ambiguity
		  with what swap does.
	- Addressed comments from Mike Rapoport:
	  	- Simplified the loop in alloc_vmemmap_page_list().
		- Could NOT address a comment to move
		  store_early_perpage_metadata() near where nodes
		  and page allocator are initialized.
		- Included the vmalloc()ed page_ext in accounting
		  within free_page_ext().
		- Made early_perpage_metadata[MAX_NUMNODES] static.


Previous approaches and discussions
-----------------------------------
v5:
https://lore.kernel.org/all/20231101230816.1459373-1-souravpanda@google.com
v4:
https://lore.kernel.org/all/20231031223846.827173-1-souravpanda@google.com
v3:
https://lore.kernel.org/all/20231031174459.459480-1-souravpanda@google.com
v2:
https://lore.kernel.org/all/20231018005548.3505662-1-souravpanda@google.com
v1:
https://lore.kernel.org/r/20230913173000.4016218-2-souravpanda@google.com

Hi!

This patch adds two new per-node fields, namely nr_page_metadata and
nr_page_metadata_boot to /sys/devices/system/node/nodeN/vmstat and a
global PageMetadata field to /proc/meminfo. This information can be
used by users to see how much memory is being used by per-page
metadata, which can vary depending on build configuration, machine
architecture, and system use.

Per-page metadata is the amount of memory that Linux needs in order to
manage memory at the page granularity. The majority of such memory is
used by "struct page" and "page_ext" data structures.


Background
----------

Kernel overhead observability is missing some of the largest
allocations during runtime, including vmemmap (struct pages) and
page_ext. This patch aims to address this problem by exporting a
new metric PageMetadata.

On the contrary, the kernel does provide observibility for boot memory
allocations. For example, the metric reserved_pages depicts the pages
allocated by the bootmem allocator. This can be simply calculated as
present_pages - managed_pages, which are both exported in /proc/zoneinfo.
The metric reserved_pages is primarily composed of struct pages and
page_ext.

What about the struct pages (allocated by bootmem allocator) that are
free'd during hugetlbfs allocations and then allocated by buddy-allocator
once hugtlbfs pages are free'd?

/proc/meminfo MemTotal changes: MemTotal does not include memblock
allocations but includes buddy allocations. However, during runtime
memblock allocations can be shifted into buddy allocations, and therefore
become part of MemTotal.

Once the struct pages get allocated by buddy allocator, we lose track of
these struct page allocations overhead accounting. Therefore, we must
export new metrics. nr_page_metadata and nr_page_metadata_boot
exclusively track the struct page and page ext allocations made by the
buddy allocator and memblock allocator, respectively for each node.
PageMetadata in /proc/meminfo would report the struct page and page ext
allocations made by the buddy allocator.

Results and analysis
--------------------

Memory model: Sparsemem-vmemmap
$ echo 1 > /proc/sys/vm/hugetlb_optimize_vmemmap

$ cat /proc/meminfo | grep MemTotal
	MemTotal:       32919124 kB
$ cat /proc/meminfo | grep Meta
	PageMetadata:      65536 kB
$ cat /sys/devices/system/node/node0/vmstat | grep "nr_page_metadata"
	nr_page_metadata 8192
	nr_page_metadata_boot 65536
$ cat /sys/devices/system/node/node1/vmstat | grep "nr_page_metadata"
        nr_page_metadata 8192
	nr_page_metadata_boot 65536

AFTER HUGTLBFS RESERVATION
$ echo 512 > /proc/sys/vm/nr_hugepages

$ cat /proc/meminfo | grep MemTotal
	MemTotal:       32935508 kB
$ cat /proc/meminfo | grep Meta
	PageMetadata:      67584 kB
$ cat /sys/devices/system/node/node0/vmstat | grep "nr_page_metadata"
	nr_page_metadata 8448
	nr_page_metadata_boot 63488
$ cat /sys/devices/system/node/node1/vmstat | grep "nr_page_metadata"
	nr_page_metadata 8448
	nr_page_metadata_booy 63488


AFTER FREEING HUGTLBFS RESERVATION
$ echo 0 > /proc/sys/vm/nr_hugepages

$ cat /proc/meminfo | grep MemTotal
        MemTotal:       32935508 kB
$ cat /proc/meminfo | grep Meta
        PageMetadata:      81920 kB
$ cat /sys/devices/system/node/node0/vmstat | grep "nr_page_metadata"
        nr_page_metadata 10240
	nr_page_metadata_boot 63488
$ cat /sys/devices/system/node/node1/vmstat | grep "nr_page_metadata"
        nr_page_metadata 10240
	nr_page_metadata_boot 63488

------------------------

Memory Hotplug with memory_hotplug.memmap_on_memory=1

BEFORE HOTADD
$ cat /proc/meminfo | grep MemTotal
	MemTotal:       32919124 kB
$ cat /proc/meminfo | grep PageMeta
	PageMetadata:      65536 kB
$ cat /sys/devices/system/node/node0/vmstat | grep "nr_page_metadata"
	nr_page_metadata 8192
	nr_page_metadata_boot 65536
$ cat /sys/devices/system/node/node1/vmstat | grep "nr_page_metadata"
	nr_page_metadata 8192
	nr_page_metadata_boot 65536

AFTER HOTADDING 1GB
$ cat /proc/meminfo | grep MemTotal
	MemTotal:       33951316 kB
$ cat /proc/meminfo | grep PageMeta
	PageMetadata:      83968 kB
$ cat /sys/devices/system/node/node0/vmstat | grep "nr_page_metadata"
	nr_page_metadata 12800
	nr_page_metadata_boot 65536
$ cat /sys/devices/system/node/node1/vmstat | grep "nr_page_metadata"
	nr_page_metadata 8192
	nr_page_metadata_boot 65536

--------------------------

Memory Hotplug with memory_hotplug.memmap_on_memory=0

BEFORE HOTADD
$ cat /proc/meminfo | grep MemTotal
	MemTotal:       32919124 kB
$ cat /proc/meminfo | grep PageMeta
	PageMetadata:      65536 kB
$ cat /sys/devices/system/node/node0/vmstat | grep "nr_page_metadata"
	nr_page_metadata 8192
	nr_page_metadata_boot 65536
$ cat /sys/devices/system/node/node1/vmstat | grep "nr_page_metadata"
	nr_page_metadata 8192
	nr_page_metadata_boot 65536

AFTER HOTADDING 1GB
$ cat /proc/meminfo | grep MemTotal
	MemTotal:       33967700 kB
$ cat /proc/meminfo | grep PageMeta
	PageMetadata:      83968 kB
$ cat /sys/devices/system/node/node0/vmstat | grep "nr_page_metadata"
	nr_page_metadata 12800
	nr_page_metadata_boot 65536
$ cat /sys/devices/system/node/node1/vmstat | grep "nr_page_metadata"
	nr_page_metadata 8192
	nr_page_metadata_boot 65536

Sourav Panda (1):
  mm: report per-page metadata information

 Documentation/filesystems/proc.rst |  3 +++
 fs/proc/meminfo.c                  |  7 +++++++
 include/linux/mmzone.h             |  4 ++++
 include/linux/vmstat.h             |  4 ++++
 mm/hugetlb_vmemmap.c               | 19 ++++++++++++++----
 mm/mm_init.c                       |  3 +++
 mm/page_alloc.c                    |  1 +
 mm/page_ext.c                      | 32 +++++++++++++++++++++---------
 mm/sparse-vmemmap.c                |  8 ++++++++
 mm/sparse.c                        |  7 ++++++-
 mm/vmstat.c                        | 26 +++++++++++++++++++++++-
 11 files changed, 99 insertions(+), 15 deletions(-)
  

On Tue, Dec 05, 2023 at 02:31:17PM -0800, Sourav Panda wrote:
> Changelog:
> v6:
> 	- Interface changes
> 	  	- Added per-node nr_page_metadata and
> 		  nr_page_metadata_boot fields that are exported
> 		  in /sys/devices/system/node/nodeN/vmstat

Again, please do not add any new fields to existing sysfs files, that's
not going to work well.  You can add a new sysfs file, that's file, but
do not continue the abuse of the sysfs api in this file.

thanks,

greg k-h
  

Hi Greg,

Sourav removed the new field from sys/device../nodeN/meminfo as you
requested; however, in nodeN/vmstat fields still get appended, as
there is code that displays every item in zone_stat_item,
node_stat_item without option to opt-out. I mentioned it to you at
LPC.

In my IOMMU [1] series, there are also fields that are added to
node_stat_item that as result are appended to nodeN/vmstat.

Pasha

[1] https://lore.kernel.org/all/20231130201504.2322355-1-pasha.tatashin@soleen.com

On Tue, Dec 5, 2023 at 9:36 PM Greg KH <gregkh@linuxfoundation.org> wrote:
>
> On Tue, Dec 05, 2023 at 02:31:17PM -0800, Sourav Panda wrote:
> > Changelog:
> > v6:
> >       - Interface changes
> >               - Added per-node nr_page_metadata and
> >                 nr_page_metadata_boot fields that are exported
> >                 in /sys/devices/system/node/nodeN/vmstat
>
> Again, please do not add any new fields to existing sysfs files, that's
> not going to work well.  You can add a new sysfs file, that's file, but
> do not continue the abuse of the sysfs api in this file.
>
> thanks,
>
> greg k-h
  

On Tue, Dec 05, 2023 at 09:57:36PM -0500, Pasha Tatashin wrote:
> Hi Greg,
> 
> Sourav removed the new field from sys/device../nodeN/meminfo as you
> requested; however, in nodeN/vmstat fields still get appended, as
> there is code that displays every item in zone_stat_item,
> node_stat_item without option to opt-out. I mentioned it to you at
> LPC.

Sorry, I thought that was a proc file, not a sysfs file.  Don't grow
that file please, it should not be that way and adding to it will just
cause others to also want to add to it and we end up with the whole proc
file mess...

> In my IOMMU [1] series, there are also fields that are added to
> node_stat_item that as result are appended to nodeN/vmstat.

I missed that, that too shouldn't be done please.

Again, sysfs is "one value per file" for a reason, don't abuse the fact
that we missed this abuse of the rules for that file by adding more
things to it.

thanks,

greg k-h
  

On Wed, 6 Dec 2023 12:12:10 +0900 Greg KH <gregkh@linuxfoundation.org> wrote:

> On Tue, Dec 05, 2023 at 09:57:36PM -0500, Pasha Tatashin wrote:
> > Hi Greg,
> > 
> > Sourav removed the new field from sys/device../nodeN/meminfo as you
> > requested; however, in nodeN/vmstat fields still get appended, as
> > there is code that displays every item in zone_stat_item,
> > node_stat_item without option to opt-out. I mentioned it to you at
> > LPC.
> 
> Sorry, I thought that was a proc file, not a sysfs file.  Don't grow
> that file please, it should not be that way and adding to it will just
> cause others to also want to add to it and we end up with the whole proc
> file mess...
> 
> > In my IOMMU [1] series, there are also fields that are added to
> > node_stat_item that as result are appended to nodeN/vmstat.
> 
> I missed that, that too shouldn't be done please.
> 
> Again, sysfs is "one value per file" for a reason, don't abuse the fact
> that we missed this abuse of the rules for that file by adding more
> things to it.

I'm afraid that horse has bolted.

hp2:/usr/src/25> wc /sys/devices/system/node/node0/vmstat 
  61  122 1309 /sys/devices/system/node/node0/vmstat

We're never going to unpick this into 61 separate files so adding new
files at this stage is pointless.
  

On Wed, Dec 06, 2023 at 07:59:13AM -0800, Andrew Morton wrote:
> On Wed, 6 Dec 2023 12:12:10 +0900 Greg KH <gregkh@linuxfoundation.org> wrote:
> 
> > On Tue, Dec 05, 2023 at 09:57:36PM -0500, Pasha Tatashin wrote:
> > > Hi Greg,
> > > 
> > > Sourav removed the new field from sys/device../nodeN/meminfo as you
> > > requested; however, in nodeN/vmstat fields still get appended, as
> > > there is code that displays every item in zone_stat_item,
> > > node_stat_item without option to opt-out. I mentioned it to you at
> > > LPC.
> > 
> > Sorry, I thought that was a proc file, not a sysfs file.  Don't grow
> > that file please, it should not be that way and adding to it will just
> > cause others to also want to add to it and we end up with the whole proc
> > file mess...
> > 
> > > In my IOMMU [1] series, there are also fields that are added to
> > > node_stat_item that as result are appended to nodeN/vmstat.
> > 
> > I missed that, that too shouldn't be done please.
> > 
> > Again, sysfs is "one value per file" for a reason, don't abuse the fact
> > that we missed this abuse of the rules for that file by adding more
> > things to it.
> 
> I'm afraid that horse has bolted.
> 
> hp2:/usr/src/25> wc /sys/devices/system/node/node0/vmstat 
>   61  122 1309 /sys/devices/system/node/node0/vmstat
> 
> We're never going to unpick this into 61 separate files so adding new
> files at this stage is pointless.

But if it keeps growing, it will eventually overflow and start crashing
the kernel and you will then have to do the horrid thing of turning it
into a binary sysfs file.

So I can please ask that no new entries be added to the file please,
let's not keep making things worse.  For new items, just add new files,
don't add to the existing mess.

thanks,

greg k-h