[PATCH 1/3] mm/filemap: add filemap_add_folio_nocharge()

[Boris Burkov <boris@bur.io>]

Btrfs currently tracks its metadata pages in the page cache, using a
fake inode (fs_info->btree_inode) with offsets corresponding to where
the metadata is stored in the filesystem's full logical address space.

A consequence of this is that when btrfs uses filemap_add_folio(), this
usage is charged to the cgroup of whichever task happens to be running
at the time. These folios don't belong to any particular user cgroup, so
I don't think it makes much sense for them to be charged in that way.
Some negative consequences as a result:
- A task can be holding some important btrfs locks, then need to lookup
  some metadata and go into reclaim, extending the duration it holds
  that lock for, and unfairly pushing its own reclaim pain onto other
  cgroups.
- If that cgroup goes into reclaim, it might reclaim these folios a
  different non-reclaiming cgroup might need soon. This is naturally
  offset by LRU reclaim, but still.

A very similar proposal to use the root cgroup was previously made by
Qu, but he was advised by Christoph to instead introduce a _nocharge
variant of filemap_add_folio.

Link: https://lore.kernel.org/linux-mm/b5fef5372ae454a7b6da4f2f75c427aeab6a07d6.1727498749.git.wqu@suse.com/
Suggested-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Boris Burkov <boris@bur.io>
---
 include/linux/pagemap.h |  2 ++
 mm/filemap.c            | 23 +++++++++++++++++------
 2 files changed, 19 insertions(+), 6 deletions(-)

diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
index e63fbfbd5b0f..acc8d390ecbb 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -1237,6 +1237,8 @@ size_t fault_in_readable(const char __user *uaddr, size_t size);
 
 int add_to_page_cache_lru(struct page *page, struct address_space *mapping,
 		pgoff_t index, gfp_t gfp);
+int filemap_add_folio_nocharge(struct address_space *mapping,
+			       struct folio *folio, pgoff_t index, gfp_t gfp);
 int filemap_add_folio(struct address_space *mapping, struct folio *folio,
 		pgoff_t index, gfp_t gfp);
 void filemap_remove_folio(struct folio *folio);
diff --git a/mm/filemap.c b/mm/filemap.c
index bada249b9fb7..ccc9cfb4d418 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -955,20 +955,15 @@ noinline int __filemap_add_folio(struct address_space *mapping,
 }
 ALLOW_ERROR_INJECTION(__filemap_add_folio, ERRNO);
 
-int filemap_add_folio(struct address_space *mapping, struct folio *folio,
+int filemap_add_folio_nocharge(struct address_space *mapping, struct folio *folio,
 				pgoff_t index, gfp_t gfp)
 {
 	void *shadow = NULL;
 	int ret;
 
-	ret = mem_cgroup_charge(folio, NULL, gfp);
-	if (ret)
-		return ret;
-
 	__folio_set_locked(folio);
 	ret = __filemap_add_folio(mapping, folio, index, gfp, &shadow);
 	if (unlikely(ret)) {
-		mem_cgroup_uncharge(folio);
 		__folio_clear_locked(folio);
 	} else {
 		/*
@@ -986,6 +981,22 @@ int filemap_add_folio(struct address_space *mapping, struct folio *folio,
 	}
 	return ret;
 }
+EXPORT_SYMBOL_GPL(filemap_add_folio_nocharge);
+
+int filemap_add_folio(struct address_space *mapping, struct folio *folio,
+				pgoff_t index, gfp_t gfp)
+{
+	int ret;
+
+	ret = mem_cgroup_charge(folio, NULL, gfp);
+	if (ret)
+		return ret;
+
+	ret = filemap_add_folio_nocharge(mapping, folio, index, gfp);
+	if (ret)
+		mem_cgroup_uncharge(folio);
+	return ret;
+}
 EXPORT_SYMBOL_GPL(filemap_add_folio);
 
 #ifdef CONFIG_NUMA
-- 
2.50.1