[linux-cifs-client] Re: [PATCH 3/3] [CIFS] Enable dfs submounts to handle remote referrals.

[Igor Mammedov]

Steve French wrote:
> I merged your updated patch 2 (patch 1 was merged earlier), but was
> not clear in the following patch (patch 3) why we no longer need to
> check flags for DFSREF_STORAGE_SERVER, are other types (both
> DFSREF_REFERRAL_SERVER and DFSREF_TARGET_FAILBACK) now ok and do not
> require special handling (I wasn't sure about what TARGET_FAILBACK
> does).

Now cifs supports only V3 style referrals only and by ms docs
TARGET_FAILBACK should be ignored by a client except for V4 referrals.

So we could ignore it for a while till we decide to add support for
V4 referrals.


> 
> On Fri, Mar 27, 2009 at 11:43 AM, Jeff Layton <jlayton at redhat.com> wrote:
>> On Tue, 17 Mar 2009 19:44:32 +0300
>> Igor Mammedov <niallain at gmail.com> wrote:
>>
>>> From 73352e0f001d32812f9b18ae85aec35026b7a194 Mon Sep 17 00:00:00 2001
>>> From: Igor Mammedov <niallain at gmail.com>
>>> Date: Tue, 17 Mar 2009 19:00:30 +0300
>>> Subject: [PATCH 3/3] [CIFS] Enable dfs submounts to handle remote referrals.
>>>
>>> Having remote dfs root support in cifs_mount, we can
>>> afford to pass into it UNC that is remote.
>>>
>>> Signed-off-by: Igor Mammedov <niallain at gmail.com>
>>> ---
>>>  fs/cifs/cifs_dfs_ref.c |   32 ++++++++++++++------------------
>>>  1 files changed, 14 insertions(+), 18 deletions(-)
>>>
>>> diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c
>>> index 5fdbf8a..896b08f 100644
>>> --- a/fs/cifs/cifs_dfs_ref.c
>>> +++ b/fs/cifs/cifs_dfs_ref.c
>>> @@ -341,27 +341,23 @@ cifs_dfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd)
>>>
>>>       for (i = 0; i < num_referrals; i++) {
>>>               dump_referral(referrals+i);
>>> -             /* connect to a storage node */
>>> -             if (referrals[i].flags & DFSREF_STORAGE_SERVER) {
>>> -                     int len;
>>> -                     len = strlen(referrals[i].node_name);
>>> -                     if (len < 2) {
>>> -                             cERROR(1, ("%s: Net Address path too short: %s",
>>> +             /* connect to a node */
>>> +             int len;
>>> +             len = strlen(referrals[i].node_name);
>>> +             if (len < 2) {
>>> +                     cERROR(1, ("%s: Net Address path too short: %s",
>>>                                       __func__, referrals[i].node_name));
>>> -                             rc = -EINVAL;
>>> -                             goto out_err;
>>> -                     }
>>> -                     mnt = cifs_dfs_do_refmount(nd->path.mnt,
>>> -                                             nd->path.dentry,
>>> -                                             referrals + i);
>>> -                     cFYI(1, ("%s: cifs_dfs_do_refmount:%s , mnt:%p",
>>> -                                      __func__,
>>> +                     rc = -EINVAL;
>>> +                     goto out_err;
>>> +             }
>>> +             mnt = cifs_dfs_do_refmount(nd->path.mnt,
>>> +                             nd->path.dentry, referrals + i);
>>> +             cFYI(1, ("%s: cifs_dfs_do_refmount:%s , mnt:%p", __func__,
>>>                                       referrals[i].node_name, mnt));
>>>
>>> -                     /* complete mount procedure if we accured submount */
>>> -                     if (!IS_ERR(mnt))
>>> -                             break;
>>> -             }
>>> +             /* complete mount procedure if we accured submount */
>>> +             if (!IS_ERR(mnt))
>>> +                     break;
>>>       }
>>>
>>>       /* we need it cause for() above could exit without valid submount */
>>
>> Looks fine, though obviously patch #2 needs to go in first.
>>
>> --
>> Jeff Layton <jlayton at redhat.com>
>>
> 
> 
> 

-- 

Best regards,

-------------------------
Igor Mammedov,
niallain "at" gmail.com