filters: add a filter that accepts EIO instead of other errors

filters: add a filter that accepts EIO instead of other errors

Running a dm-flakey or dm-error test that loads a table that returns
EIO to all IO and then running a command that is expected to fail
with a specific error is racy.

If there is memory pressure at the same time that the table is
loaded, cached inodes can be turfed from memory and the command then
needs to read the inode it is about to act on from disk again. This
results in the inode read getting EIO and failing (e.g. xfs_io will
return a stat() error) rather than having the desired operation
fail.

This results in spurious test failures that look like this:

generic/331       - output mismatch (see /mnt/xfs/runner-41/results-2024-11-20-10:57:31/xfs/generic/331.out.bad)
    --- tests/generic/331.out   2022-12-21 15:53:25.487044098 +1100
    +++ /mnt/xfs/runner-41/results-2024-11-20-10:57:31/xfs/generic/331.out.bad  2024-11-20 11:02:12.123572607 +1100
    @@ -5,7 +5,8 @@
     1886e67cf8783e89ce6ddc5bb09a3944  SCRATCH_MNT/test-331/file1
     1886e67cf8783e89ce6ddc5bb09a3944  SCRATCH_MNT/test-331/file2
     CoW and unmount
    -fdatasync: Input/output error
    +/mnt/xfs/runner-41/scratch/test-331/file2: Input/output error
    +stat: Input/output error
     Compare files
    ...

Add a new "flakey EIO filter" that will catch -any- EIO error from
the command and change it to the error we expected to see.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Zorro lang <zlang@redhat.com>
Signed-off-by: Zorro Lang <zlang@kernel.org>