From: Julia Lawall <[email protected]>
list_for_each_entry uses its first argument to move from one element to the
next, so modifying it can break the iteration. The variable re1 is already
used within the loop as a temporary variable, and is not live here.
The semantic match that finds this problem is as follows:
(http://coccinelle.lip6.fr/)
// <smpl>
@r@
iterator name list_for_each_entry;
expression x,E;
position p1,p2;
@@
list_for_each_entry@p1(x,...) { <... x =@p2 E ...> }
@@
expression x,E;
position r.p1,r.p2;
statement S;
@@
*x =@p2 E
...
list_for_each_entry@p1(x,...) S
// </smpl>
Signed-off-by: Julia Lawall <[email protected]>
---
drivers/mtd/ubi/cdev.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/drivers/mtd/ubi/cdev.c b/drivers/mtd/ubi/cdev.c
index 4dfa6b9..3d2d1a6 100644
--- a/drivers/mtd/ubi/cdev.c
+++ b/drivers/mtd/ubi/cdev.c
@@ -798,18 +798,18 @@ static int rename_volumes(struct ubi_device *ubi,
goto out_free;
}
- re = kzalloc(sizeof(struct ubi_rename_entry), GFP_KERNEL);
- if (!re) {
+ re1 = kzalloc(sizeof(struct ubi_rename_entry), GFP_KERNEL);
+ if (!re1) {
err = -ENOMEM;
ubi_close_volume(desc);
goto out_free;
}
- re->remove = 1;
- re->desc = desc;
- list_add(&re->list, &rename_list);
+ re1->remove = 1;
+ re1->desc = desc;
+ list_add(&re1->list, &rename_list);
dbg_msg("will remove volume %d, name \"%s\"",
- re->desc->vol->vol_id, re->desc->vol->name);
+ re1->desc->vol->vol_id, re1->desc->vol->name);
}
mutex_lock(&ubi->device_mutex);
On Sat, 2010-08-07 at 11:09 +0200, Julia Lawall wrote:
> From: Julia Lawall <[email protected]>
>
> list_for_each_entry uses its first argument to move from one element to the
> next, so modifying it can break the iteration. The variable re1 is already
> used within the loop as a temporary variable, and is not live here.
Looks good, thanks, I'll push it to the UBI tree a bit later.
--
Best Regards,
Artem Bityutskiy (Артём Битюцкий)
On Sat, 2010-08-07 at 11:09 +0200, Julia Lawall wrote:
> From: Julia Lawall <[email protected]>
>
> list_for_each_entry uses its first argument to move from one element to the
> next, so modifying it can break the iteration. The variable re1 is already
> used within the loop as a temporary variable, and is not live here.
>
> The semantic match that finds this problem is as follows:
> (http://coccinelle.lip6.fr/)
>
> // <smpl>
> @r@
> iterator name list_for_each_entry;
> expression x,E;
> position p1,p2;
> @@
>
> list_for_each_entry@p1(x,...) { <... x =@p2 E ...> }
>
> @@
> expression x,E;
> position r.p1,r.p2;
> statement S;
> @@
>
> *x =@p2 E
> ...
> list_for_each_entry@p1(x,...) S
> // </smpl>
>
> Signed-off-by: Julia Lawall <[email protected]>
Pushed to ubi-2.6.git tree. This won't be seen in linux-next for some
time though, but I'll make it visible in linux-next at some point soon.
Thanks!
--
Best Regards,
Artem Bityutskiy (Артём Битюцкий)