From: Wengang Wang Date: Tue, 10 Nov 2020 16:55:57 -0800 Subject: [Ocfs2-devel] [PATCH V2] ocfs2: initialize ip_next_orphan In-Reply-To: <20201110130433.a891a2c79e96c0815079a199@linux-foundation.org> References: <20201109171746.27884-1-wen.gang.wang@oracle.com> <20201110130433.a891a2c79e96c0815079a199@linux-foundation.org> Message-ID: List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ocfs2-devel@oss.oracle.com On 11/10/20 1:04 PM, Andrew Morton wrote: > On Mon, 9 Nov 2020 17:56:41 -0800 Wengang Wang wrote: > >>>> Looking at more on ip_next_orphan, I found it's not initialized when >>>> allocating a new ocfs2_inode_info structure. >>>> >>>> Fix: >>>> ????initialize ip_next_orphan as NULL. >>>> Signed-off-by: Wengang Wang >>> Reviewed-by: Joseph Qi >> Thank you Joseph! >> >> AKPM, could you please pull this patch to your tree? > I'm trying to figure out whether we should add cc:stable to this. But > to know that, I need to know the end-user visible effects of the bug, > and that isn't described in this changelog (it should have been!). > > Please describe this? The observed end-user visible effects is that: The reflink operations from some nodes hang for very long time waiting for the cluster lock on the orphan directory. So yes, we should add cc:stable. Sorry, I missed that part. Thanks, Wengang