Re: [Ocfs2-devel] What's the need of OCFS2_INODE_MAYBE_ORPHANED?
On Thu, Jan 09, 2014 at 07:35:05AM -0600, Goldwyn Rodrigues wrote: On 01/09/2014 04:23 AM, Joel Becker wrote: Unlink can happen from anywhere, but only the last closing node can actually remove the file. MAYBE_ORPHANED tells the node to try for removal at close time. It is absolutely necessary. The reason I asked the query is that OCFS2_INODE_MAYBE_ORPHANED is being set at every dentry downconvert. Is this really necessary because every dentry downconvert does not turn into unlink? (I know it says maybe :/ ) Is it okay to set it when the open_lock fails or is it too late in the process? If another node has performed an unlink, it would need to get the open lock before it performs the inode wipe. So we should be safe that way? Is there anything incorrect in this design? It's not safe. Srini has already answered this on the other part of the thread. I'll address your other comments there. Joel -- Goldwyn -- You look in her eyes, the music begins to play. Hopeless romantics, here we go again. http://www.jlbec.org/ jl...@evilplan.org ___ Ocfs2-devel mailing list Ocfs2-devel@oss.oracle.com https://oss.oracle.com/mailman/listinfo/ocfs2-devel
Re: [Ocfs2-devel] What's the need of OCFS2_INODE_MAYBE_ORPHANED?
On Thu, Jan 09, 2014 at 11:27:15AM -0600, Goldwyn Rodrigues wrote: Yes, I did not consider that. How about using open locks ro_holders count to identify this? That may just work. Thanks! One problem I see in using open lock for this is it could be late. Consider the scenario where node A removes the dentry and then the node crashes before trying the try_open_lock. Node B does the file close later but it doesn't know that the file was unlinked and doesn't do the clean up. To me it appears OCFS2_INODE_MAYBE_ORPHANED is necessary. Any delay it is causing must be addressed differently. No, I don't mean to remove the OCFS2_INODE_MAYBE_ORPHANED flag, but set it conditionally in ocfs2_dentry_convert_worker() based on the value of the open locks held. I'm confused by what you are attempting here. We hold the dentry lock until the final dpu() (see the comment in fs/ocfs2/dcache.c). We should never have ro_holders==0 unless we're flushing the entry from the dcache. Do you mean something else? Joel -- Now Someone's on the telephone, desperate in his pain. Someone's on the bathroom floor doing her cocaine. Someone's got his finger on the button in some room. No one can convince me we aren't gluttons for our doom. http://www.jlbec.org/ jl...@evilplan.org ___ Ocfs2-devel mailing list Ocfs2-devel@oss.oracle.com https://oss.oracle.com/mailman/listinfo/ocfs2-devel
Re: [Ocfs2-devel] [PATCH 1/1] o2dlm: fix NULL pointer dereference in o2dlm_blocking_ast_wrapper
On Fri, Jan 10, 2014 at 05:19:13PM -0800, Srinivas Eeda wrote: From: Srinivas Eeda seeda@srini.(none) A tiny race between BAST and unlock message causes the NULL dereference. A node sends an unlock request to master and receives a response. Before processing the response it receives a BAST from the master. Since both requests are processed by different threads it creates a race. While the BAST is being processed, lock can get freed by unlock code. This patch makes bast to return immediately if lock is found but unlock is pending. The code should handle this race. We also have to fix master node to skip sending BAST after receiving unlock message. Did the master send the BAST after the unlock, or does that race too? Does the master know the unlock has succeeded, or does it just think so? @@ -385,8 +385,13 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data, head = res-granted; list_for_each_entry(lock, head, list) { - if (lock-ml.cookie == cookie) - goto do_ast; + /* if lock is found but unlock is pending ignore the bast */ + if (lock-ml.cookie == cookie) { + if (lock-unlock_pending) + break; + else + goto do_ast; + } This breaks out for asts as well as basts. Can't that cause problems with the unlock ast expected by the caller? Joel -- Not being known doesn't stop the truth from being true. - Richard Bach http://www.jlbec.org/ jl...@evilplan.org ___ Ocfs2-devel mailing list Ocfs2-devel@oss.oracle.com https://oss.oracle.com/mailman/listinfo/ocfs2-devel
Re: [Ocfs2-devel] [PATCH 1/1] o2dlm: fix NULL pointer dereference in o2dlm_blocking_ast_wrapper
On 2014/1/11 9:19, Srinivas Eeda wrote: From: Srinivas Eeda seeda@srini.(none) A tiny race between BAST and unlock message causes the NULL dereference. A node sends an unlock request to master and receives a response. Before processing the response it receives a BAST from the master. Since both requests are processed by different threads it creates a race. While the BAST is being processed, lock can get freed by unlock code. This patch makes bast to return immediately if lock is found but unlock is pending. The code should handle this race. We also have to fix master node to skip sending BAST after receiving unlock message. Below is the crash stack BUG: unable to handle kernel NULL pointer dereference at 0048 IP: [a015e023] o2dlm_blocking_ast_wrapper+0xd/0x16 [a034e3db] dlm_do_local_bast+0x8e/0x97 [ocfs2_dlm] [a034f366] dlm_proxy_ast_handler+0x838/0x87e [ocfs2_dlm] [a0308abe] o2net_process_message+0x395/0x5b8 [ocfs2_nodemanager] [a030aac8] o2net_rx_until_empty+0x762/0x90d [ocfs2_nodemanager] [81071802] worker_thread+0x14d/0x1ed Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com --- fs/ocfs2/dlm/dlmast.c |9 +++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/fs/ocfs2/dlm/dlmast.c b/fs/ocfs2/dlm/dlmast.c index b46278f..dbc6cee 100644 --- a/fs/ocfs2/dlm/dlmast.c +++ b/fs/ocfs2/dlm/dlmast.c @@ -385,8 +385,13 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data, head = res-granted; list_for_each_entry(lock, head, list) { - if (lock-ml.cookie == cookie) - goto do_ast; + /* if lock is found but unlock is pending ignore the bast */ + if (lock-ml.cookie == cookie) { + if (lock-unlock_pending) + break; + else + goto do_ast; + } } mlog(0, Got %sast for unknown lock! cookie=%u:%llu, name=%.*s, I found you sent a version on Jan 30, 2012. https://oss.oracle.com/pipermail/ocfs2-devel/2012-January/008469.html Compared with the old version, this version only saves a little bit CPU, am I right? ___ Ocfs2-devel mailing list Ocfs2-devel@oss.oracle.com https://oss.oracle.com/mailman/listinfo/ocfs2-devel
Re: [Ocfs2-devel] [PATCH 1/1] o2dlm: fix NULL pointer dereference in o2dlm_blocking_ast_wrapper
On 01/13/2014 08:06 PM, Joseph Qi wrote: On 2014/1/11 9:19, Srinivas Eeda wrote: From: Srinivas Eeda seeda@srini.(none) A tiny race between BAST and unlock message causes the NULL dereference. A node sends an unlock request to master and receives a response. Before processing the response it receives a BAST from the master. Since both requests are processed by different threads it creates a race. While the BAST is being processed, lock can get freed by unlock code. This patch makes bast to return immediately if lock is found but unlock is pending. The code should handle this race. We also have to fix master node to skip sending BAST after receiving unlock message. Below is the crash stack BUG: unable to handle kernel NULL pointer dereference at 0048 IP: [a015e023] o2dlm_blocking_ast_wrapper+0xd/0x16 [a034e3db] dlm_do_local_bast+0x8e/0x97 [ocfs2_dlm] [a034f366] dlm_proxy_ast_handler+0x838/0x87e [ocfs2_dlm] [a0308abe] o2net_process_message+0x395/0x5b8 [ocfs2_nodemanager] [a030aac8] o2net_rx_until_empty+0x762/0x90d [ocfs2_nodemanager] [81071802] worker_thread+0x14d/0x1ed Signed-off-by: Srinivas Eeda srinivas.e...@oracle.com --- fs/ocfs2/dlm/dlmast.c |9 +++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/fs/ocfs2/dlm/dlmast.c b/fs/ocfs2/dlm/dlmast.c index b46278f..dbc6cee 100644 --- a/fs/ocfs2/dlm/dlmast.c +++ b/fs/ocfs2/dlm/dlmast.c @@ -385,8 +385,13 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data, head = res-granted; list_for_each_entry(lock, head, list) { -if (lock-ml.cookie == cookie) -goto do_ast; +/* if lock is found but unlock is pending ignore the bast */ +if (lock-ml.cookie == cookie) { +if (lock-unlock_pending) +break; +else +goto do_ast; +} } mlog(0, Got %sast for unknown lock! cookie=%u:%llu, name=%.*s, I found you sent a version on Jan 30, 2012. https://oss.oracle.com/pipermail/ocfs2-devel/2012-January/008469.html Compared with the old version, this version only saves a little bit CPU, am I right? Yes you are right. I made the change as Goldwyn suggested which is a good thing to have :) ___ Ocfs2-devel mailing list Ocfs2-devel@oss.oracle.com https://oss.oracle.com/mailman/listinfo/ocfs2-devel
Re: [Ocfs2-devel] [PATCH 1/1] o2dlm: fix NULL pointer dereference in o2dlm_blocking_ast_wrapper
On 01/13/2014 07:37 AM, Joel Becker wrote: On Fri, Jan 10, 2014 at 05:19:13PM -0800, Srinivas Eeda wrote: From: Srinivas Eeda seeda@srini.(none) A tiny race between BAST and unlock message causes the NULL dereference. A node sends an unlock request to master and receives a response. Before processing the response it receives a BAST from the master. Since both requests are processed by different threads it creates a race. While the BAST is being processed, lock can get freed by unlock code. This patch makes bast to return immediately if lock is found but unlock is pending. The code should handle this race. We also have to fix master node to skip sending BAST after receiving unlock message. Did the master send the BAST after the unlock, or does that race too? Does the master know the unlock has succeeded, or does it just think so? I think it's due to a race but I haven't debugged the master. My guess is unlock request sneaked in before the dlm_flush_asts was called. However non master node should handle this race as well, so just did that part which fixed a bug we were seeing. @@ -385,8 +385,13 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data, head = res-granted; list_for_each_entry(lock, head, list) { -if (lock-ml.cookie == cookie) -goto do_ast; +/* if lock is found but unlock is pending ignore the bast */ +if (lock-ml.cookie == cookie) { +if (lock-unlock_pending) +break; +else +goto do_ast; +} This breaks out for asts as well as basts. Can't that cause problems with the unlock ast expected by the caller? if unlock_pending is set, then the node is trying to unlock an existing lock and shouldn't receive any asts ? Joel ___ Ocfs2-devel mailing list Ocfs2-devel@oss.oracle.com https://oss.oracle.com/mailman/listinfo/ocfs2-devel