From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tao Ma Date: Mon, 30 Jun 2008 10:53:34 +0800 Subject: [Ocfs2-devel] [PATCH 7/8] ocfs2: Add extended attributes support. v1 In-Reply-To: <20080627181034.GB28100@wotan.suse.de> References: <484792D1.3080802@oracle.com> <1212650694-9603-1-git-send-email-tiger.yang@oracle.com> <20080617233221.GD28100@wotan.suse.de> <48649B99.3090205@oracle.com> <20080627181034.GB28100@wotan.suse.de> Message-ID: <48684AAE.3030104@oracle.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ocfs2-devel@oss.oracle.com Hi Mark, Yes, this is also a problem in xattr bucket. Tiger and I both inserted a zeroed tree root as xattr's value and then extend the root. The [PATCH 13/15] also has the same problem. I will revise it in the next round. Regards, Tao Mark Fasheh wrote: > On Fri, Jun 27, 2008 at 03:49:45PM +0800, Tiger Yang wrote: >> I split setting long value operation into two parts. First insert tree >> root in local, if succeed then insert long value into B tree based on >> that tree root. If hit problem in second step, we could only get that >> junk tree root. This is harmless for system. but is it acceptable? > > If the junk root is ignored, and eventually overwritten as more ea's get > inserted then I don't think you need to do any more. > > If the junk root stays in the block and occupies space that would otherwise > be used, then we probably need to handle this.. I imagine a clever user > could DOS the xattr space by making large ones when the system is low on > free space... > --Mark > > -- > Mark Fasheh > > _______________________________________________ > Ocfs2-devel mailing list > Ocfs2-devel at oss.oracle.com > http://oss.oracle.com/mailman/listinfo/ocfs2-devel