luck/tmp_suning_uos_patched
1
0
Fork 1
Code Issues Pull Requests Packages Projects Releases Wiki Activity

cifs: stuff the fl_owner into "pid" field in the lock request

Browse Source 
Right now, we send the tgid cross the wire. What we really want to send
though is a hashed fl_owner_t since samba treats this field as a generic
lockowner.

It turns out that because we enforce and release locks locally before
they are ever sent to the server, this patch makes no difference in
behavior. Still, setting OFD locks on the server using the process
pid seems wrong, so I think this patch still makes sense.

Signed-off-by: Jeff Layton <jlayton@poochiereds.net>
Signed-off-by: Steve French <smfrench@gmail.com>
Acked-by: Pavel Shilovsky <pshilovsky@samba.org>
Acked-by: Sachin Prabhu <sprabhu@redhat.com>
This commit is contained in:
Jeff Layton 2016-05-24 06:27:44 -04:00 committed by Steve French
parent da01e18a37
commit 3d22462ae9
3 changed files with 15 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
fs/cifs/cifsfs.c
View File

@ -87,6 +87,7 @@ extern mempool_t *cifs_req_poolp;
extern mempool_t *cifs_mid_poolp;
struct workqueue_struct *cifsiod_wq;
__u32 cifs_lock_secret;
/*
* Bumps refcount for cifs super block.
@ -1266,6 +1267,8 @@ init_cifs(void)
spin_lock_init(&cifs_file_list_lock);
spin_lock_init(&GlobalMid_Lock);
get_random_bytes(&cifs_lock_secret, sizeof(cifs_lock_secret));
if (cifs_max_pending < 2) {
cifs_max_pending = 2;
cifs_dbg(FYI, "cifs_max_pending set to min of 2\n");

1
fs/cifs/cifsglob.h
View File

@ -1619,6 +1619,7 @@ void cifs_oplock_break(struct work_struct *work);
extern const struct slow_work_ops cifs_oplock_break_ops;
extern struct workqueue_struct *cifsiod_wq;
extern __u32 cifs_lock_secret;
extern mempool_t *cifs_mid_poolp;

14
fs/cifs/file.c
View File

@ -1112,6 +1112,12 @@ cifs_push_mandatory_locks(struct cifsFileInfo *cfile)
return rc;
}
static __u32
hash_lockowner(fl_owner_t owner)
{
return cifs_lock_secret ^ hash32_ptr((const void *)owner);
}
struct lock_to_push {
struct list_head llist;
__u64 offset;
@ -1178,7 +1184,7 @@ cifs_push_posix_locks(struct cifsFileInfo *cfile)
else
type = CIFS_WRLCK;
lck = list_entry(el, struct lock_to_push, llist);
lck->pid = flock->fl_pid;
lck->pid = hash_lockowner(flock->fl_owner);
lck->netfid = cfile->fid.netfid;
lck->length = length;
lck->type = type;
@ -1305,7 +1311,8 @@ cifs_getlk(struct file *file, struct file_lock *flock, __u32 type,
posix_lock_type = CIFS_RDLCK;
else
posix_lock_type = CIFS_WRLCK;
rc = CIFSSMBPosixLock(xid, tcon, netfid, current->tgid,
rc = CIFSSMBPosixLock(xid, tcon, netfid,
hash_lockowner(flock->fl_owner),
flock->fl_start, length, flock,
posix_lock_type, wait_flag);
return rc;
@ -1505,7 +1512,8 @@ cifs_setlk(struct file *file, struct file_lock *flock, __u32 type,
posix_lock_type = CIFS_UNLCK;
rc = CIFSSMBPosixLock(xid, tcon, cfile->fid.netfid,
current->tgid, flock->fl_start, length,
hash_lockowner(flock->fl_owner),
flock->fl_start, length,
NULL, posix_lock_type, wait_flag);
goto out;
}