On Tue, Nov 16, 2010 at 3:48 PM, Steve French <smfrench@xxxxxxxxx> wrote:
> On Sat, Nov 13, 2010 at 10:02 AM, <shirishpargaonkar@xxxxxxxxx> wrote:
>> From: Shirish Pargaonkar <shirishpargaonkar@xxxxxxxxx>
>>
>> Change the name of function mode_to_acl to mode_to_cifs_acl.
>>
>> Handle return code in functions mode_to_cifs_acl and
>> cifs_acl_to_fattr.
>>
>> Pass fid to inode helper function which passes it on to
>> get inode info function so we do not end up opening the
>> opened file again thus breaking oplocks.
> <snip>
>
>> @@ -2113,11 +2118,15 @@ cifs_setattr_nounix(struct dentry *direntry, struct iattr *attrs)
>> }
>>
>> if (attrs->ia_valid & ATTR_MODE) {
>> - rc = 0;
>> #ifdef CONFIG_CIFS_EXPERIMENTAL
>> - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL)
>> - rc = mode_to_acl(inode, full_path, mode);
>> - else
>> + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) {
>> + rc = mode_to_cifs_acl(inode, full_path, mode);
>> + if (rc) {
>> + cFYI(1, "%s: Setting ACL failed with error: %d",
>> + __func__, rc);
>> + goto cifs_setattr_exit;
>> + }
>> + }
>> #endif
>> if (((mode & S_IWUGO) == 0) &&
>> (cifsInode->cifsAttrs & ATTR_READONLY) == 0) {
>
> This changes behavior. When you mount with CIFS_ACL then you are
> still falling through to set the mode bits too (it used to avoid
> setting the mode bits when we can use ACLs to more accurately
> represent the mode). You got rid of the else ... Did you mean to do
> that?
Steve, I was thinking dynperm (mode) should take precedence if it is chosen
as a mount option. Would that be right thing to do?
--
To unsubscribe from this list: send the line "unsubscribe linux-cifs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
