Messages in this thread | | | Date | Thu, 3 Oct 2013 03:50:07 +0100 | From | Al Viro <> | Subject | Re: [rfc] rework aio migrate pages to use aio fs |
| |
On Thu, Oct 03, 2013 at 03:22:51AM +0100, Al Viro wrote: > > + /* > > + * Link the inode to a directory entry by creating a unique name > > + * using the inode sequence number. > > + */ > > + file = ERR_PTR(-ENOMEM); > > + this.name = name; > > + this.len = strlen(name); > > + this.hash = 0; > > Umm... ITYM > struct qstr this = QSTR_INIT("[aio]", 5); > , if not > path.dentry = d_alloc_pseudo(aio_mnt->mnt_sb, > &(struct qstr)QSTR_INIT("[aio]", 5)); > > + if (!path.dentry) > > + goto err_module;
and you've leaked that inode, BTW.
Another thing: I'd rather pull everything about setting the inode up (aops, size, etc.) in there.
Anyway, could you take a look at the last couple of commits in vfs.git#for-bcrl? Commit message on the last one is an atrocity, of course...
| |