[PATCH 4.4 34/47] userns: move user access out of the mutex
From: Greg Kroah-Hartman
Date: Fri Sep 07 2018 - 17:41:14 EST
4.4-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jann Horn <jannh@xxxxxxxxxx>
commit 5820f140edef111a9ea2ef414ab2428b8cb805b1 upstream.
The old code would hold the userns_state_mutex indefinitely if
memdup_user_nul stalled due to e.g. a userfault region. Prevent that by
moving the memdup_user_nul in front of the mutex_lock().
Note: This changes the error precedence of invalid buf/count/*ppos vs
map already written / capabilities missing.
Fixes: 22d917d80e84 ("userns: Rework the user_namespace adding uid/gid...")
Cc: stable@xxxxxxxxxxxxxxx
Signed-off-by: Jann Horn <jannh@xxxxxxxxxx>
Acked-by: Christian Brauner <christian@xxxxxxxxxx>
Acked-by: Serge Hallyn <serge@xxxxxxxxxx>
Signed-off-by: Eric W. Biederman <ebiederm@xxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
kernel/user_namespace.c | 22 ++++++++++------------
1 file changed, 10 insertions(+), 12 deletions(-)
--- a/kernel/user_namespace.c
+++ b/kernel/user_namespace.c
@@ -604,7 +604,16 @@ static ssize_t map_write(struct file *fi
struct uid_gid_extent *extent = NULL;
unsigned long page = 0;
char *kbuf, *pos, *next_line;
- ssize_t ret = -EINVAL;
+ ssize_t ret;
+
+ /* Only allow < page size writes at the beginning of the file */
+ if ((*ppos != 0) || (count >= PAGE_SIZE))
+ return -EINVAL;
+
+ /* Slurp in the user data */
+ if (copy_from_user(kbuf, buf, count))
+ return -EFAULT;
+ kbuf[count] = '\0';
/*
* The userns_state_mutex serializes all writes to any given map.
@@ -645,17 +654,6 @@ static ssize_t map_write(struct file *fi
if (!page)
goto out;
- /* Only allow < page size writes at the beginning of the file */
- ret = -EINVAL;
- if ((*ppos != 0) || (count >= PAGE_SIZE))
- goto out;
-
- /* Slurp in the user data */
- ret = -EFAULT;
- if (copy_from_user(kbuf, buf, count))
- goto out;
- kbuf[count] = '\0';
-
/* Parse the user data */
ret = -EINVAL;
pos = kbuf;