Received-SPF: pass (google.com: best guess record for domain of linux-nfs-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   NeilBrown <neilb@suse.de>
To:     Trond Myklebust <trondmy@hammerspace.com>,
        "Anna.Schumaker\@Netapp.com" <Anna.Schumaker@Netapp.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Jan Kara <jack@suse.cz>, Michal Hocko <mhocko@kernel.org>
Date:   Tue, 07 Apr 2020 09:44:25 +1000
Cc:     linux-mm@kvack.org, linux-nfs@vger.kernel.org,
        LKML <linux-kernel@vger.kernel.org>
Subject: [PATCH 2/2] MM: Discard NR_UNSTABLE_NFS, use NR_WRITEBACK instead.
In-Reply-To: <draft-87d08kw57p.fsf@notabene.neil.brown.name>
References: <draft-87d08kw57p.fsf@notabene.neil.brown.name>
Message-ID: <878sj8w55y.fsf@notabene.neil.brown.name>
MIME-Version: 1.0
Content-Type: multipart/signed; boundary="=-=-=";
        micalg=pgp-sha256; protocol="application/pgp-signature"
Sender: linux-nfs-owner@vger.kernel.org
Precedence: bulk

--=-=-=
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable


After an NFS page has been written it is considered "unstable" until a
COMMIT request succeeds.  If the COMMIT fails, the page will be
re-written.

These "unstable" pages are currently accounted as "reclaimable", either
in WB_RECLAIMABLE, or in NR_UNSTABLE_NFS which is included in a
'reclaimable' count.  This might have made sense when sending the COMMIT
required a separate action by the VFS/MM (e.g.  releasepage() used to
send a COMMIT).  However now that all writes generated by ->writepages()
will automatically be followed by a COMMIT (since commit 919e3bd9a875
("NFS: Ensure we commit after writeback is complete")) it makes more
sense to treat them as writeback pages.

So this patch removes NR_UNSTABLE_NFS and accounts unstable pages in
NR_WRITEBACK and WB_WRITEBACK.

A particular effect of this change is that when
wb_check_background_flush() calls wb_over_bg_threshold(), the latter
will report 'true' a lot less often as the 'unstable' pages are no
longer considered 'dirty' (as there is nothing that writeback can do
about them anyway).

Currently wb_check_background_flush() will trigger writeback to NFS even
when there are relatively few dirty pages (if there are lots of unstable
pages), this can result in small writes going to the server (10s of
Kilobytes rather than a Megabyte) which hurts throughput.
With this patch, there are fewer writes which are each larger on average.

Where the NR_UNSTABLE_NFS count was included in statistics
virtual-files, the entry is retained, but the value is hard-coded as
zero.  static trace points which record this counter no longer report
it.

Signed-off-by: NeilBrown <neilb@suse.de>
=2D--
 Documentation/filesystems/proc.rst |  4 ++--
 drivers/base/node.c                |  2 +-
 fs/fs-writeback.c                  |  1 -
 fs/nfs/internal.h                  | 10 +++++++---
 fs/nfs/write.c                     |  4 ++--
 fs/proc/meminfo.c                  |  3 +--
 include/linux/mmzone.h             |  1 -
 include/trace/events/writeback.h   |  5 +----
 mm/memcontrol.c                    |  1 -
 mm/page-writeback.c                | 17 ++++-------------
 mm/page_alloc.c                    |  6 ++----
 mm/vmstat.c                        | 13 ++++++++++---
 12 files changed, 30 insertions(+), 37 deletions(-)

diff --git a/Documentation/filesystems/proc.rst b/Documentation/filesystems=
/proc.rst
index 38b606991065..092b7b44d158 100644
=2D-- a/Documentation/filesystems/proc.rst
+++ b/Documentation/filesystems/proc.rst
@@ -1042,8 +1042,8 @@ PageTables
               amount of memory dedicated to the lowest level of page
               tables.
 NFS_Unstable
=2D              NFS pages sent to the server, but not yet committed to sta=
ble
=2D	      storage
+              Always zero. Previous counted pages which had been written to
+              the server, but has not been committed to stable storage.
 Bounce
               Memory used for block device "bounce buffers"
 WritebackTmp
diff --git a/drivers/base/node.c b/drivers/base/node.c
index 10d7e818e118..15f5ed6a8830 100644
=2D-- a/drivers/base/node.c
+++ b/drivers/base/node.c
@@ -439,7 +439,7 @@ static ssize_t node_read_meminfo(struct device *dev,
 		       nid, K(i.sharedram),
 		       nid, sum_zone_node_page_state(nid, NR_KERNEL_STACK_KB),
 		       nid, K(sum_zone_node_page_state(nid, NR_PAGETABLE)),
=2D		       nid, K(node_page_state(pgdat, NR_UNSTABLE_NFS)),
+		       nid, 0,
 		       nid, K(sum_zone_node_page_state(nid, NR_BOUNCE)),
 		       nid, K(node_page_state(pgdat, NR_WRITEBACK_TEMP)),
 		       nid, K(sreclaimable +
diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 76ac9c7d32ec..c5bdf46e3b4b 100644
=2D-- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -1070,7 +1070,6 @@ static void bdi_split_work_to_wbs(struct backing_dev_=
info *bdi,
 static unsigned long get_nr_dirty_pages(void)
 {
 	return global_node_page_state(NR_FILE_DIRTY) +
=2D		global_node_page_state(NR_UNSTABLE_NFS) +
 		get_nr_dirty_inodes();
 }
=20
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index f80c47d5ff27..749da02b547a 100644
=2D-- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -652,7 +652,8 @@ void nfs_super_set_maxbytes(struct super_block *sb, __u=
64 maxfilesize)
 }
=20
 /*
=2D * Record the page as unstable and mark its inode as dirty.
+ * Record the page as unstable (an extra writeback period) and mark its
+ * inode as dirty.
  */
 static inline
 void nfs_mark_page_unstable(struct page *page, struct nfs_commit_info *cin=
fo)
@@ -660,8 +661,11 @@ void nfs_mark_page_unstable(struct page *page, struct =
nfs_commit_info *cinfo)
 	if (!cinfo->dreq) {
 		struct inode *inode =3D page_file_mapping(page)->host;
=20
=2D		inc_node_page_state(page, NR_UNSTABLE_NFS);
=2D		inc_wb_stat(&inode_to_bdi(inode)->wb, WB_RECLAIMABLE);
+		/* This page is really still in write-back - just that the
+		 * writeback is happening on the server now.
+		 */
+		inc_node_page_state(page, NR_WRITEBACK);
+		inc_wb_stat(&inode_to_bdi(inode)->wb, WB_WRITEBACK);
 		__mark_inode_dirty(inode, I_DIRTY_DATASYNC);
 	}
 }
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index c478b772cc49..2e15a56620b3 100644
=2D-- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -958,9 +958,9 @@ nfs_mark_request_commit(struct nfs_page *req, struct pn=
fs_layout_segment *lseg,
 static void
 nfs_clear_page_commit(struct page *page)
 {
=2D	dec_node_page_state(page, NR_UNSTABLE_NFS);
+	dec_node_page_state(page, NR_WRITEBACK);
 	dec_wb_stat(&inode_to_bdi(page_file_mapping(page)->host)->wb,
=2D		    WB_RECLAIMABLE);
+		    WB_WRITEBACK);
 }
=20
 /* Called holding the request lock on @req */
diff --git a/fs/proc/meminfo.c b/fs/proc/meminfo.c
index 8c1f1bb1a5ce..9bd94b5a9658 100644
=2D-- a/fs/proc/meminfo.c
+++ b/fs/proc/meminfo.c
@@ -106,8 +106,7 @@ static int meminfo_proc_show(struct seq_file *m, void *=
v)
 	show_val_kb(m, "PageTables:     ",
 		    global_zone_page_state(NR_PAGETABLE));
=20
=2D	show_val_kb(m, "NFS_Unstable:   ",
=2D		    global_node_page_state(NR_UNSTABLE_NFS));
+	show_val_kb(m, "NFS_Unstable:   ", 0);
 	show_val_kb(m, "Bounce:         ",
 		    global_zone_page_state(NR_BOUNCE));
 	show_val_kb(m, "WritebackTmp:   ",
diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index e84d448988b6..3937f2be27d8 100644
=2D-- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -237,7 +237,6 @@ enum node_stat_item {
 	NR_FILE_THPS,
 	NR_FILE_PMDMAPPED,
 	NR_ANON_THPS,
=2D	NR_UNSTABLE_NFS,	/* NFS unstable pages */
 	NR_VMSCAN_WRITE,
 	NR_VMSCAN_IMMEDIATE,	/* Prioritise for reclaim when writeback ends */
 	NR_DIRTIED,		/* page dirtyings since bootup */
diff --git a/include/trace/events/writeback.h b/include/trace/events/writeb=
ack.h
index d94def25e4dc..45b5fbdb1f62 100644
=2D-- a/include/trace/events/writeback.h
+++ b/include/trace/events/writeback.h
@@ -542,7 +542,6 @@ TRACE_EVENT(global_dirty_state,
 	TP_STRUCT__entry(
 		__field(unsigned long,	nr_dirty)
 		__field(unsigned long,	nr_writeback)
=2D		__field(unsigned long,	nr_unstable)
 		__field(unsigned long,	background_thresh)
 		__field(unsigned long,	dirty_thresh)
 		__field(unsigned long,	dirty_limit)
@@ -553,7 +552,6 @@ TRACE_EVENT(global_dirty_state,
 	TP_fast_assign(
 		__entry->nr_dirty	=3D global_node_page_state(NR_FILE_DIRTY);
 		__entry->nr_writeback	=3D global_node_page_state(NR_WRITEBACK);
=2D		__entry->nr_unstable	=3D global_node_page_state(NR_UNSTABLE_NFS);
 		__entry->nr_dirtied	=3D global_node_page_state(NR_DIRTIED);
 		__entry->nr_written	=3D global_node_page_state(NR_WRITTEN);
 		__entry->background_thresh =3D background_thresh;
@@ -561,12 +559,11 @@ TRACE_EVENT(global_dirty_state,
 		__entry->dirty_limit	=3D global_wb_domain.dirty_limit;
 	),
=20
=2D	TP_printk("dirty=3D%lu writeback=3D%lu unstable=3D%lu "
+	TP_printk("dirty=3D%lu writeback=3D%lu "
 		  "bg_thresh=3D%lu thresh=3D%lu limit=3D%lu "
 		  "dirtied=3D%lu written=3D%lu",
 		  __entry->nr_dirty,
 		  __entry->nr_writeback,
=2D		  __entry->nr_unstable,
 		  __entry->background_thresh,
 		  __entry->dirty_thresh,
 		  __entry->dirty_limit,
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index ca194864d802..41b450b0ca29 100644
=2D-- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -4326,7 +4326,6 @@ void mem_cgroup_wb_stats(struct bdi_writeback *wb, un=
signed long *pfilepages,
=20
 	*pdirty =3D memcg_exact_page_state(memcg, NR_FILE_DIRTY);
=20
=2D	/* this should eventually include NR_UNSTABLE_NFS */
 	*pwriteback =3D memcg_exact_page_state(memcg, NR_WRITEBACK);
 	*pfilepages =3D memcg_exact_page_state(memcg, NR_INACTIVE_FILE) +
 			memcg_exact_page_state(memcg, NR_ACTIVE_FILE);
diff --git a/mm/page-writeback.c b/mm/page-writeback.c
index 4c9875971de5..d16f6a59bce4 100644
=2D-- a/mm/page-writeback.c
+++ b/mm/page-writeback.c
@@ -504,7 +504,6 @@ bool node_dirty_ok(struct pglist_data *pgdat)
 	unsigned long nr_pages =3D 0;
=20
 	nr_pages +=3D node_page_state(pgdat, NR_FILE_DIRTY);
=2D	nr_pages +=3D node_page_state(pgdat, NR_UNSTABLE_NFS);
 	nr_pages +=3D node_page_state(pgdat, NR_WRITEBACK);
=20
 	return nr_pages <=3D limit;
@@ -758,7 +757,7 @@ static void mdtc_calc_avail(struct dirty_throttle_contr=
ol *mdtc,
  * bounded by the bdi->min_ratio and/or bdi->max_ratio parameters, if set.
  *
  * Return: @wb's dirty limit in pages. The term "dirty" in the context of
=2D * dirty balancing includes all PG_dirty, PG_writeback and NFS unstable =
pages.
+ * dirty balancing includes all PG_dirty and PG_writeback pages.
  */
 static unsigned long __wb_calc_thresh(struct dirty_throttle_control *dtc)
 {
@@ -1566,7 +1565,7 @@ static void balance_dirty_pages(struct bdi_writeback =
*wb,
 	struct dirty_throttle_control * const mdtc =3D mdtc_valid(&mdtc_stor) ?
 						     &mdtc_stor : NULL;
 	struct dirty_throttle_control *sdtc;
=2D	unsigned long nr_reclaimable;	/* =3D file_dirty + unstable_nfs */
+	unsigned long nr_reclaimable;	/* =3D file_dirty */
 	long period;
 	long pause;
 	long max_pause;
@@ -1589,14 +1588,7 @@ static void balance_dirty_pages(struct bdi_writeback=
 *wb,
 		unsigned long m_thresh =3D 0;
 		unsigned long m_bg_thresh =3D 0;
=20
=2D		/*
=2D		 * Unstable writes are a feature of certain networked
=2D		 * filesystems (i.e. NFS) in which data may have been
=2D		 * written to the server's write cache, but has not yet
=2D		 * been flushed to permanent storage.
=2D		 */
=2D		nr_reclaimable =3D global_node_page_state(NR_FILE_DIRTY) +
=2D					global_node_page_state(NR_UNSTABLE_NFS);
+		nr_reclaimable =3D global_node_page_state(NR_FILE_DIRTY);
 		gdtc->avail =3D global_dirtyable_memory();
 		gdtc->dirty =3D nr_reclaimable + global_node_page_state(NR_WRITEBACK);
=20
@@ -1940,8 +1932,7 @@ bool wb_over_bg_thresh(struct bdi_writeback *wb)
 	 * as we're trying to decide whether to put more under writeback.
 	 */
 	gdtc->avail =3D global_dirtyable_memory();
=2D	gdtc->dirty =3D global_node_page_state(NR_FILE_DIRTY) +
=2D		      global_node_page_state(NR_UNSTABLE_NFS);
+	gdtc->dirty =3D global_node_page_state(NR_FILE_DIRTY);
 	domain_dirty_limits(gdtc);
=20
 	if (gdtc->dirty > gdtc->bg_thresh)
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index e5f76da8cd4e..24678d6e308d 100644
=2D-- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -5237,7 +5237,7 @@ void show_free_areas(unsigned int filter, nodemask_t =
*nodemask)
=20
 	printk("active_anon:%lu inactive_anon:%lu isolated_anon:%lu\n"
 		" active_file:%lu inactive_file:%lu isolated_file:%lu\n"
=2D		" unevictable:%lu dirty:%lu writeback:%lu unstable:%lu\n"
+		" unevictable:%lu dirty:%lu writeback:%lu unstable:0\n"
 		" slab_reclaimable:%lu slab_unreclaimable:%lu\n"
 		" mapped:%lu shmem:%lu pagetables:%lu bounce:%lu\n"
 		" free:%lu free_pcp:%lu free_cma:%lu\n",
@@ -5250,7 +5250,6 @@ void show_free_areas(unsigned int filter, nodemask_t =
*nodemask)
 		global_node_page_state(NR_UNEVICTABLE),
 		global_node_page_state(NR_FILE_DIRTY),
 		global_node_page_state(NR_WRITEBACK),
=2D		global_node_page_state(NR_UNSTABLE_NFS),
 		global_node_page_state(NR_SLAB_RECLAIMABLE),
 		global_node_page_state(NR_SLAB_UNRECLAIMABLE),
 		global_node_page_state(NR_FILE_MAPPED),
@@ -5283,7 +5282,7 @@ void show_free_areas(unsigned int filter, nodemask_t =
*nodemask)
 			" anon_thp: %lukB"
 #endif
 			" writeback_tmp:%lukB"
=2D			" unstable:%lukB"
+			" unstable:0kB"
 			" all_unreclaimable? %s"
 			"\n",
 			pgdat->node_id,
@@ -5305,7 +5304,6 @@ void show_free_areas(unsigned int filter, nodemask_t =
*nodemask)
 			K(node_page_state(pgdat, NR_ANON_THPS) * HPAGE_PMD_NR),
 #endif
 			K(node_page_state(pgdat, NR_WRITEBACK_TEMP)),
=2D			K(node_page_state(pgdat, NR_UNSTABLE_NFS)),
 			pgdat->kswapd_failures >=3D MAX_RECLAIM_RETRIES ?
 				"yes" : "no");
 	}
diff --git a/mm/vmstat.c b/mm/vmstat.c
index c9c0d71f917f..228d9f6e1c5c 100644
=2D-- a/mm/vmstat.c
+++ b/mm/vmstat.c
@@ -1108,7 +1108,7 @@ int fragmentation_index(struct zone *zone, unsigned i=
nt order)
 					TEXT_FOR_HIGHMEM(xx) xx "_movable",
=20
 const char * const vmstat_text[] =3D {
=2D	/* enum zone_stat_item countes */
+	/* enum zone_stat_item counters */
 	"nr_free_pages",
 	"nr_zone_inactive_anon",
 	"nr_zone_active_anon",
@@ -1162,7 +1162,6 @@ const char * const vmstat_text[] =3D {
 	"nr_file_hugepages",
 	"nr_file_pmdmapped",
 	"nr_anon_transparent_hugepages",
=2D	"nr_unstable",
 	"nr_vmscan_write",
 	"nr_vmscan_immediate_reclaim",
 	"nr_dirtied",
@@ -1707,8 +1706,16 @@ static void *vmstat_start(struct seq_file *m, loff_t=
 *pos)
 static void *vmstat_next(struct seq_file *m, void *arg, loff_t *pos)
 {
 	(*pos)++;
=2D	if (*pos >=3D NR_VMSTAT_ITEMS)
+	if (*pos >=3D NR_VMSTAT_ITEMS) {
+		/*
+		 * Deprecated counters which are no longer represented
+		 * in vmstat arrays. We just lie about them to be always
+		 * 0 to not break userspace which might expect them in
+		 * the output.
+		 */
+		seq_puts(m, "nr_unstable 0");
 		return NULL;
+	}
 	return (unsigned long *)m->private + *pos;
 }
=20
=2D-=20
2.26.0


--=-=-=
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----

iQIzBAEBCAAdFiEEG8Yp69OQ2HB7X0l6Oeye3VZigbkFAl6LvtkACgkQOeye3VZi
gbn3+hAAl4JHaokEZf0M4UPhW7wrfReHkOp23r6mHwZGhMb5nSNvujzMpt+iuFrX
7i8GkILPaEkLFTYJ71sbJPchyZz/1NI65imEIGMnD5gNVuWHeeOULGi7Ydxas9kn
a+ZTn7mURcGf3j2R0g9+QNPEGo6hL3KAYDbEvDo66qLYfO9abHb9L1khNkFpAGkn
WAaxPqPkG2nW71+g/mwo/WOblMfEDCeoHPfI0T3Yp0JJbnBhQm5Q/wK7/8FrL9id
M7bpmvbw8Z+RjW5NWv6VPINPFBGGyu52/wnjLPWzH346+h9DiOxfuhh7kMZs/t53
auDOTR9GTNbL/0nq5B/AqnlWS9fTDbdzu1WJn7UgmRM75m06MYApaBGdLpbTxmrx
huCt6UEnNcYErwXrEav9rDcyHYn9kcEpeWPCsUPEQWk9rDYTvTzaR2Pzsz41rNIS
+fVhK16xp9I1aPsnRbqMhATfewVW81RBU+SXxrLCwLF132H2jpD3W3T8pFfHFEYG
pSkQBR4eE5HML+l+vcY27YwSaBCx+98J9bz7fKA2/Cv+H4ONluYAerTU/Qwwuiwk
ItY74XgswpeIRf5/hYkUVgHcDhcwtwa7aiJSOVO3VvSsm/gjm99/kcuPn7S1xT2O
XGJGLTezl1iM3Qz6Uqbw/N/h0CAR4KoF+UHMOzXpFMQo4VAC9zo=
=6+Sd
-----END PGP SIGNATURE-----
--=-=-=--