On 2023/4/27 14:36, Baokun Li wrote:
On 2023/4/27 12:50, Ming Lei wrote:
Hello Matthew,
On Thu, Apr 27, 2023 at 04:58:36AM +0100, Matthew Wilcox wrote:
On Thu, Apr 27, 2023 at 10:20:28AM +0800, Ming Lei wrote:
Hello Guys,
I got one report in which buffered write IO hangs in
balance_dirty_pages,
after one nvme block device is unplugged physically, then umount can't
succeed.
That's a feature, not a bug ... the dd should continue indefinitely?
Can you explain what the feature is? And not see such 'issue' or
'feature'
on xfs.
The device has been gone, so IMO it is reasonable to see FS buffered
write IO
failed. Actually dmesg has shown that 'EXT4-fs (nvme0n1): Remounting
filesystem read-only'. Seems these things may confuse user.
The reason for this difference is that ext4 and xfs handle errors
differently.
ext4 remounts the filesystem as read-only or even just continues,
vfs_write does not check for these.
xfs shuts down the filesystem, so it returns a failure at
xfs_file_write_iter when it finds an error.
``` ext4
ksys_write
vfs_write
ext4_file_write_iter
ext4_buffered_write_iter
ext4_write_checks
file_modified
file_modified_flags
__file_update_time
inode_update_time
generic_update_time
__mark_inode_dirty
ext4_dirty_inode ---> 2. void func, No propagating errors out
__ext4_journal_start_sb
ext4_journal_check_start ---> 1. Error found, remount-ro
generic_perform_write ---> 3. No error sensed, continue
balance_dirty_pages_ratelimited
balance_dirty_pages_ratelimited_flags
balance_dirty_pages
// 4. Sleeping waiting for dirty pages to be freed
__set_current_state(TASK_KILLABLE)
io_schedule_timeout(pause);
```
``` xfs
ksys_write
vfs_write
xfs_file_write_iter
if (xfs_is_shutdown(ip->i_mount))
return -EIO; ---> dd fail
```
balance_dirty_pages() is sleeping in KILLABLE state, so kill -9 of
the dd process should succeed.
Yeah, dd can be killed, however it may be any application(s), :-)
Fortunately it won't cause trouble during reboot/power off, given
userspace will be killed at that time.
Thanks,
Ming
Don't worry about that, we always set the current thread to TASK_KILLABLE
while waiting in balance_dirty_pages().
On second thought, we can determine if the file system has become read-only
when the ext4_file_write_iter() is called on a write file, even though
the fs was
not read-only when the file was opened.
This would end the write process early and free up resources like xfs does.
The patch is below, does anyone have any other thoughts?
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index d101b3b0c7da..d2966268ee41 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -699,6 +699,8 @@ ext4_file_write_iter(struct kiocb *iocb, struct
iov_iter *from)
if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
return -EIO;
+ if (unlikely(sb_rdonly(inode->i_sb)))
+ return -EROFS;
#ifdef CONFIG_FS_DAX
if (IS_DAX(inode))
--
With Best Regards,
Baokun Li
.
