Re: [PATCH] fs: xfs: Fix a lock-twice and sleep-in-atomic bug in xfs_iget

[Jia-Ju Bai]

On 06/05/2017 04:32 PM, Shan Hai wrote:



On 2017年06月05日 16:17, Jia-Ju Bai wrote:

The driver may sleep under a read rcu lock, and function call path is:
xfs_iget (acquire the lock by rcu_read_lock)
   "goto out_error_or_again" after xfs_iget_cache_hit
 delay
   schedule_timeout_uninterruptible --> may sleep
Meanwhile, the rcu_read_lock will be called twice in this situation.

To fix it, the lock is released before "goto".

Signed-off-by: Jia-Ju Bai 
---
  fs/xfs/xfs_icache.c |4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index f61c84f8..c2a4722 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -600,8 +600,10 @@ struct xfs_inode *
if (ip) {
  error = xfs_iget_cache_hit(pag, ip, ino, flags, lock_flags);
-if (error)
+if (error) {
+rcu_read_unlock();


Seems you are going to double unlock by doing this, since it is 
unlocked in the xfs_iget_cache_hit.


Thanks
Shan Hai


  goto out_error_or_again;
+}
  } else {
  rcu_read_unlock();
  XFS_STATS_INC(mp, xs_ig_missed);



I think you are right. Please ignore my patch.

Thanks,
Jia-Ju Bai

Re: [PATCH] fs: xfs: Fix a lock-twice and sleep-in-atomic bug in xfs_iget

[Shan Hai]

On 2017年06月05日 16:17, Jia-Ju Bai wrote:

The driver may sleep under a read rcu lock, and function call path is:
xfs_iget (acquire the lock by rcu_read_lock)
   "goto out_error_or_again" after xfs_iget_cache_hit
 delay
   schedule_timeout_uninterruptible --> may sleep
Meanwhile, the rcu_read_lock will be called twice in this situation.

To fix it, the lock is released before "goto".

Signed-off-by: Jia-Ju Bai 
---
  fs/xfs/xfs_icache.c |4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index f61c84f8..c2a4722 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -600,8 +600,10 @@ struct xfs_inode *
  
  	if (ip) {

error = xfs_iget_cache_hit(pag, ip, ino, flags, lock_flags);
-   if (error)
+   if (error) {
+   rcu_read_unlock();


Seems you are going to double unlock by doing this, since it is unlocked 
in the xfs_iget_cache_hit.


Thanks
Shan Hai


goto out_error_or_again;
+   }
} else {
rcu_read_unlock();
XFS_STATS_INC(mp, xs_ig_missed);

[PATCH] fs: xfs: Fix a lock-twice and sleep-in-atomic bug in xfs_iget

[Jia-Ju Bai]

The driver may sleep under a read rcu lock, and function call path is:
xfs_iget (acquire the lock by rcu_read_lock)
  "goto out_error_or_again" after xfs_iget_cache_hit
delay
  schedule_timeout_uninterruptible --> may sleep
Meanwhile, the rcu_read_lock will be called twice in this situation.

To fix it, the lock is released before "goto".

Signed-off-by: Jia-Ju Bai 
---
 fs/xfs/xfs_icache.c |4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index f61c84f8..c2a4722 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -600,8 +600,10 @@ struct xfs_inode *
 
if (ip) {
error = xfs_iget_cache_hit(pag, ip, ino, flags, lock_flags);
-   if (error)
+   if (error) {
+   rcu_read_unlock();
goto out_error_or_again;
+   }
} else {
rcu_read_unlock();
XFS_STATS_INC(mp, xs_ig_missed);
-- 
1.7.9.5