proc: proc_get_inode should de_put when inode already initialized
PATCH 1/1 proc: proc_get_inode should de_put when inode already initialized
de_get is called before every proc_get_inode, but corresponding de_put is called only when dropping last reference to an inode. This might cause something likeremove_proc_entry: /proc/stats busy, count=14496
to be printed to the syslog.
The fix is to call de_put in case of an already initialized inode in proc_get_inode.
---
fs/proc/inode.c +3 -1
@@ -485,8 +485,10 @@ struct inode *proc_get_inode(struct supe } }unlock_new_inode (inode); - }else + }else {module_put (de->owner); +de_put (de); + }return inode; out_ino:
Quoted 16 lines · On Mon, 23 Feb 2009 22:21:55 +0100 Krzysztof Sachanowicz <[email protected]>…
de_getis called before everyproc_get_inode, but correspondingde_putis called only when dropping last reference to an inode. This might cause something likeremove_proc_entry:/proc/statsbusy, count=14496
to be printed to the syslog.The fix is to call
de_putin case of an already initialized inode inproc_get_inode.Signed-off-by Krzysztof Sachanowicz <[email protected]>Tested-by Marcin Pilipczuk <[email protected]>---
fs/proc/inode.c +0
-0@@ -485,8 +485,10 @@ struct inode *proc_get_inode(struct supe }
fs/proc/inode.c struct inode *proc_get_inode(struct supe · line 494
Quoted 10 lines
fs/proc/inode.c +3
-1}unlock_new_inode (inode); - }else + }else {module_put (de->owner); +de_put (de); + }return inode; out_ino:
This code area looks quite different in linux-next, although the changes there are removing proc_dir_entry.owner altogether and aren't obviously targetted at fixing this bug.
Also...
It's unpleasing to have the de_get() inside the caller and the de_put() inside the callee - it is better to have them both happening at the same level. If it is the case that "de_get is called before every proc_get_inode", then perhaps that operation should simply be moved into proc_get_inode().
Tuesday 24 February 2009 00:25:55 Andrew Morton napisał(a):
Quoted 34 lines
On Mon, 23 Feb 2009 22:21:55 +0100
Quoted 16 lines · Krzysztof Sachanowicz
de_getis called before everyproc_get_inode, but correspondingde_putis called only when dropping last reference to an inode. This might cause something likeremove_proc_entry:/proc/statsbusy, count=14496
to be printed to the syslog.The fix is to call
de_putin case of an already initialized inode inproc_get_inode.Signed-off-by Krzysztof Sachanowicz <[email protected]>Tested-by Marcin Pilipczuk <[email protected]>---
---linux-2.6.29-rc6.orig/fs/proc/inode.c2009-02-23 20:43:32.000000000
+0100 +++linux-2.6.29-rc6/fs/proc/inode.c2009-02-23 20:46:37.000000000
+0100 @@ -485,8 +485,10 @@ struct inode *proc_get_inode(struct supefs/proc/inode.c +0
-0}fs/proc/inode.c struct inode *proc_get_inode(struct supe · line 494
Quoted 10 lines
fs/proc/inode.c +3
-1}unlock_new_inode (inode); - }else + }else {module_put (de->owner); +de_put (de); + }return inode; out_ino:This code area looks quite different in linux-next, although the changes there are removing proc_dir_entry.owner altogether and aren't obviously targetted at fixing this bug.
Quoted 7 lines
Also...
It's unpleasing to have the
de_get()inside the caller and thede_put()inside the callee - it is better to have them both happening at the same level. If it is the case that "de_getis called before everyproc_get_inode", then perhaps that operation should simply be moved intoproc_get_inode().
Yes, but unfortunately in proc_lookup_de() (fs/proc/generic.c) we have:
391 de_get(de); 392 spin_unlock(&proc_subdir_lock); 393 error = -EINVAL; 394 inode = proc_get_inode(dir->i_sb, ino, de);
So if we move de_get() into proc_get_inode(), we will also have to move spin_unlock there. Then we will have spin_lock in proc_lookup_de but spin_unlock in proc_get_inode...
Maybe my solution is not that bad, because usually de_put is called from proc_delete_inode(). Only if iget_locked() returns an already initialized inode we want de_put to be called in proc_get_inode. So the callee need not care about who will eventually call de_put.
Quoted 28 lines · Andrew Morton
Quoted 25 lines · On Mon, 23 Feb 2009 22:21:55 +0100 Krzysztof Sachanowicz <[email protected]>…
de_getis called before everyproc_get_inode, but correspondingde_putis called only when dropping last reference to an inode. This might cause something likeremove_proc_entry:/proc/statsbusy, count=14496
to be printed to the syslog.The fix is to call
de_putin case of an already initialized inode inproc_get_inode.Signed-off-by Krzysztof Sachanowicz <[email protected]>Tested-by Marcin Pilipczuk <[email protected]>---
fs/proc/inode.c +3
-1@@ -485,8 +485,10 @@ struct inode *proc_get_inode(struct supe } }unlock_new_inode (inode); - }else + }else {module_put (de->owner); +de_put (de); + }return inode;
Quoted 5 lines
fs/proc/inode.c struct inode *proc_get_inode(struct supe · line 494
Quoted 1 line
fs/proc/inode.c +0
-0out_ino:This code area looks quite different in linux-next, although the changes there are removing proc_dir_entry.owner altogether and aren't obviously targetted at fixing this bug.
->owner issue is independent of this leak, I'll rebase/edit patches as needed.
Quoted 7 lines
Also...
It's unpleasing to have the
de_get()inside the caller and thede_put()inside the callee - it is better to have them both happening at the same level. If it is the case that "de_getis called before everyproc_get_inode", then perhaps that operation should simply be moved intoproc_get_inode().
Well, yes, this sucks. But unlock_new_inode() will clear I_NEW state, so after function returns, there is no way to distinguish.