Re: [PATCH 13/16] ceph: implement i_op->atomic_open() andi_op->atomic_create()

[Sage Weil]

Hi Miklos,

I was pleasantly surprised to find that this works as-is.  I simplified 
this to call ceph_lookup_open for the atomic O_CREAT case as well, and it 
passes my basic battery of tests.  An incremental patch is below.

I haven't looked closely at the vfs bits, but from an fs perspective this 
series looks good to me!

Thanks-
sage




diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index 75df600..9d679dd 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -642,37 +642,17 @@ struct file *ceph_atomic_open(struct inode *dir, struct dentry *dentry,
        struct dentry *res = NULL;
        struct file *filp;
 
-       if (!(flags & O_CREAT)) {
-               if (dentry->d_name.len > NAME_MAX)
-                       return ERR_PTR(-ENAMETOOLONG);
-
-               err = ceph_init_dentry(dentry);
-               if (err < 0)
-                       return ERR_PTR(err);
-
-               return ceph_lookup_open(dir, dentry, od, flags, mode);
-       }
-
-       if (d_unhashed(dentry)) {
-               res = ceph_lookup(dir, dentry, NULL);
-               if (IS_ERR(res))
-                       return ERR_CAST(res);
-
-               if (res)
-                       dentry = res;
-       }
+       dout("atomic_open %p dentry %p '%.*s'\n",
+            dir, dentry, dentry->d_name.len, dentry->d_name.name);
 
-       /* We don't deal with positive dentries here */
-       if (dentry->d_inode) {
-               finish_no_open(od, res);
-               return NULL;
-       }
+       if (dentry->d_name.len > NAME_MAX)
+               return ERR_PTR(-ENAMETOOLONG);
 
-       *created = true;
-       filp = ceph_lookup_open(dir, dentry, od, flags, mode);
-       dput(res);
+       err = ceph_init_dentry(dentry);
+       if (err < 0)
+               return ERR_PTR(err);
 
-       return filp;
+       return ceph_lookup_open(dir, dentry, od, flags, mode);
 }
 
 /*
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/