Suppose a thread sharing the table started a resize, while
sysctl_nr_open got lowered to a value which prohibits it. This is still
going to go through with and without the patch, which is fine.
Further suppose another thread shows up to do a matching expansion while
resize_in_progress == true. It is going to error out since it performs
the sysctl_nr_open check *before* finding out if there is an expansion
in progress. But the aformentioned thread is going to succeded, so the
error is spurious (and it would not happen if the thread showed up a
little bit later).
Checking the sysctl *after* we know there are no pending updates sorts
it out.
While here annotate the thing as unlikely.
Signed-off-by: Mateusz Guzik <mjguzik@xxxxxxxxx>
---
This is a random tidbit I found while looking at the code, I don't think
this is a particularly impactful problem but definitely worth sorting
out in master.
I doubt it warrants backports to stable so I'm not cc-ing it.
fs/file.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/fs/file.c b/fs/file.c
index fb1011cf6b4a..019fb9acf91b 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -278,10 +278,6 @@ static int expand_files(struct files_struct *files, unsigned int nr)
if (nr < fdt->max_fds)
return 0;
- /* Can we expand? */
- if (nr >= sysctl_nr_open)
- return -EMFILE;
-
if (unlikely(files->resize_in_progress)) {
spin_unlock(&files->file_lock);
wait_event(files->resize_wait, !files->resize_in_progress);
@@ -289,6 +285,10 @@ static int expand_files(struct files_struct *files, unsigned int nr)
goto repeat;
}
+ /* Can we expand? */
+ if (unlikely(nr >= sysctl_nr_open))
+ return -EMFILE;
+
/* All good, so we try */
files->resize_in_progress = true;
error = expand_fdtable(files, nr);
--
2.43.0
