The patch titled
Subject: cephfs: don't set/clear bdi_congestion
has been added to the -mm tree. Its filename is
cephfs-dont-set-clear-bdi_congestion.patch
This patch should soon appear at
https://ozlabs.org/~akpm/mmots/broken-out/cephfs-dont-set-clear-bdi_congestion.patch
and later at
https://ozlabs.org/~akpm/mmotm/broken-out/cephfs-dont-set-clear-bdi_congestion.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: NeilBrown <neilb@xxxxxxx>
Subject: cephfs: don't set/clear bdi_congestion
The bdi congestion framework is no-longer used - writeback uses other
mechanisms to manage throughput.
So remove calls to set_bdi_congested() and clear_bdi_congested(), and
remove the writeback_count which is used only to guide the setting and
clearing.
The congestion_kb mount option is no longer meaningful, but as it is
visible to user-space, removing it needs more consideration.
Link: https://lkml.kernel.org/r/164325158958.29787.8840004338500709466.stgit@noble.brown
Signed-off-by: NeilBrown <neilb@xxxxxxx>
Cc: Anna Schumaker <Anna.Schumaker@xxxxxxxxxx>
Cc: Chao Yu <chao@xxxxxxxxxx>
Cc: Christoph Hellwig <hch@xxxxxx>
Cc: Darrick J. Wong <djwong@xxxxxxxxxx>
Cc: Dave Chinner <david@xxxxxxxxxxxxx>
Cc: Ilya Dryomov <idryomov@xxxxxxxxx>
Cc: Jaegeuk Kim <jaegeuk@xxxxxxxxxx>
Cc: Jeff Layton <jlayton@xxxxxxxxxx>
Cc: Jens Axboe <axboe@xxxxxxxxx>
Cc: Lars Ellenberg <lars.ellenberg@xxxxxxxxxx>
Cc: Miklos Szeredi <miklos@xxxxxxxxxx>
Cc: Paolo Valente <paolo.valente@xxxxxxxxxx>
Cc: Philipp Reisner <philipp.reisner@xxxxxxxxxx>
Cc: Ryusuke Konishi <konishi.ryusuke@xxxxxxxxx>
Cc: Trond Myklebust <trond.myklebust@xxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
fs/ceph/addr.c | 27 ---------------------------
fs/ceph/super.c | 2 --
fs/ceph/super.h | 2 --
3 files changed, 31 deletions(-)
--- a/fs/ceph/addr.c~cephfs-dont-set-clear-bdi_congestion
+++ a/fs/ceph/addr.c
@@ -57,11 +57,6 @@
* accounting is preserved.
*/
-#define CONGESTION_ON_THRESH(congestion_kb) (congestion_kb >> (PAGE_SHIFT-10))
-#define CONGESTION_OFF_THRESH(congestion_kb) \
- (CONGESTION_ON_THRESH(congestion_kb) - \
- (CONGESTION_ON_THRESH(congestion_kb) >> 2))
-
static int ceph_netfs_check_write_begin(struct file *file, loff_t pos, unsigned int len,
struct folio *folio, void **_fsdata);
@@ -561,10 +556,6 @@ static int writepage_nounlock(struct pag
dout("writepage %p page %p index %lu on %llu~%llu snapc %p seq %lld\n",
inode, page, page->index, page_off, len, snapc, snapc->seq);
- if (atomic_long_inc_return(&fsc->writeback_count) >
- CONGESTION_ON_THRESH(fsc->mount_options->congestion_kb))
- set_bdi_congested(inode_to_bdi(inode), BLK_RW_ASYNC);
-
req = ceph_osdc_new_request(osdc, &ci->i_layout, ceph_vino(inode), page_off, &len, 0, 1,
CEPH_OSD_OP_WRITE, CEPH_OSD_FLAG_WRITE, snapc,
ceph_wbc.truncate_seq, ceph_wbc.truncate_size,
@@ -621,10 +612,6 @@ static int writepage_nounlock(struct pag
ceph_put_wrbuffer_cap_refs(ci, 1, snapc);
ceph_put_snap_context(snapc); /* page's reference */
- if (atomic_long_dec_return(&fsc->writeback_count) <
- CONGESTION_OFF_THRESH(fsc->mount_options->congestion_kb))
- clear_bdi_congested(inode_to_bdi(inode), BLK_RW_ASYNC);
-
return err;
}
@@ -704,12 +691,6 @@ static void writepages_finish(struct cep
BUG_ON(!page);
WARN_ON(!PageUptodate(page));
- if (atomic_long_dec_return(&fsc->writeback_count) <
- CONGESTION_OFF_THRESH(
- fsc->mount_options->congestion_kb))
- clear_bdi_congested(inode_to_bdi(inode),
- BLK_RW_ASYNC);
-
ceph_put_snap_context(detach_page_private(page));
end_page_writeback(page);
dout("unlocking %p\n", page);
@@ -952,14 +933,6 @@ get_more_pages:
dout("%p will write page %p idx %lu\n",
inode, page, page->index);
- if (atomic_long_inc_return(&fsc->writeback_count) >
- CONGESTION_ON_THRESH(
- fsc->mount_options->congestion_kb)) {
- set_bdi_congested(inode_to_bdi(inode),
- BLK_RW_ASYNC);
- }
-
-
pages[locked_pages++] = page;
pvec.pages[i] = NULL;
--- a/fs/ceph/super.c~cephfs-dont-set-clear-bdi_congestion
+++ a/fs/ceph/super.c
@@ -801,8 +801,6 @@ static struct ceph_fs_client *create_fs_
fsc->filp_gen = 1;
fsc->have_copy_from2 = true;
- atomic_long_set(&fsc->writeback_count, 0);
-
err = -ENOMEM;
/*
* The number of concurrent works can be high but they don't need
--- a/fs/ceph/super.h~cephfs-dont-set-clear-bdi_congestion
+++ a/fs/ceph/super.h
@@ -120,8 +120,6 @@ struct ceph_fs_client {
struct ceph_mds_client *mdsc;
- atomic_long_t writeback_count;
-
struct workqueue_struct *inode_wq;
struct workqueue_struct *cap_wq;
_
Patches currently in -mm which might be from neilb@xxxxxxx are
remove-inode_congested.patch
remove-bdi_congested-and-wb_congested-and-related-functions.patch
f2fs-change-retry-waiting-for-f2fs_write_single_data_page.patch
f2f2-replace-some-congestion_wait-calls-with-io_schedule_timeout.patch
cephfs-dont-set-clear-bdi_congestion.patch
fuse-dont-set-clear-bdi_congested.patch
nfs-remove-congestion-control.patch
block-bfq-ioschedc-use-false-rather-than-blk_rw_async.patch
remove-congestion-tracking-framework.patch
mm-discard-__gfp_atomic.patch
