This framework is no longer used - so discard it.
Signed-off-by: NeilBrown
---
include/linux/backing-dev-defs.h |8 -
include/linux/backing-dev.h |2 -
include/trace/events/writeback.h | 28 ---
mm/backing-dev.c | 57 --
bfq_get_queue() expects a "bool" for the third arg, so pass "false"
rather than "BLK_RW_ASYNC" which will soon be removed.
Signed-off-by: NeilBrown
---
block/bfq-iosched.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c
index 0c612
Linux no longer uses the bdi congestion tracking framework.
So remove code from bdi which tries to support it.
Also remove the "nfs_congestion_kb" sysctl. This is a user-visible
change, but unlikely to be a problematic one.
Signed-off-by: NeilBrown
---
fs/nfs/sysctl.c |7 --
The bid congestion framework is no longer used to manage writeout etc,
so drop updating it in fuse.
Signed-off-by: NeilBrown
---
fs/fuse/control.c | 17 -
fs/fuse/dev.c |8
2 files changed, 25 deletions(-)
diff --git a/fs/fuse/control.c b/fs/fuse/control.c
ind
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
As congestion is no longer tracked, contestion_wait() is effectively
equivalent to io_schedule_timeout().
It isn't clear to me what these contestion_wait() calls are waiting
for, so I cannot change them to wait for some particular event.
So simply change them to io_schedule_timeout(), which will ha
f2fs_write_single_data_page() can return -EAGAIN if it cannot get
the cp_rwsem lock - it holds a page lock and so cannot wait for it.
Some code which calls f2fs_write_single_data_page() use
congestion_wait() and then tries again. congestion_wait() doesn't do
anything useful as congestion is no lo
These functions are no longer useful as the only bdis that report
congestion are in ceph, fuse, and nfs. None of those bdis can be the
target of the calls in drbd, ext2, nilfs2, or xfs.
Removing the test on bdi_write_contested() in current_may_throttle()
could cause a small change in behaviour, b
Congestion hasn't been reliably tracked for quite some time.
Most MM uses of it for guiding writeback decisions were removed in 5.16.
Some other uses were removed in 17-rc1.
This series removes the remaining places that test for congestion, and
the few places which still set it.
The second patch
inode_congested() reports if the backing-device for the inode is
congested. Few bdi report congestion any more, only ceph, fuse, and
nfs. Having support just for those is unlikely to be useful.
The places which test inode_congested() or it variants like
inode_write_congested(), avoid initiating
10 matches
Mail list logo
