On 2021/5/18 4:52 AM, Theodore Y. Ts'o wrote:

On Mon, May 17, 2021 at 11:57:09AM +0800, Wang Jianchao wrote:

quoted

Right now, discard is issued and waited to be completed in jbd2
commit kthread context after the logs are committed. When large
amount of files are deleted and discard is flooding, jbd2 commit
kthread can be blocked for long time. Then all of the metadata
operations can be blocked to wait the log space.

One case is the page fault path with read mm->mmap_sem held, which
wants to update the file time but has to wait for the log space.
When other threads in the task wants to do mmap, then write mmap_sem
is blocked. Finally all of the following read mmap_sem requirements
are blocked, even the ps command which need to read the /proc/pid/
-cmdline. Our monitor service which needs to read /proc/pid/cmdline
used to be blocked for 5 mins.

This patch moves the discard out of jbd2 kthread context and do it
in kworker. And drain the kwork when cannot get space in mb buddy.
This is done out of jbd2 handle and won't block the commit process.
After that, we could use blk-wbt or other method to throttle the
discard and needn't to worry it block the jbd2 commit kthread any
more.

Wouldn't be much simpler to do something like this?

		if (discard_bio) {
-			submit_bio_wait(discard_bio);
-			bio_put(discard_bio);
+			submit_bio(discard_bio);
		}


We're throwing away the return value from submit_bio_wait(), so
there's no real need to wait for I/O to complete so we can fetch the
I/O status.

That way we don't need to move all of this to a kworker context.

The submit_bio also needs to be out of jbd2 commit kthread as it may be
blocked due to blk-wbt or no enough request tag. ;)

Best Regards
Jianchao

     	    	       	       	   - Ted