[LU-201] unregistered bulk after reply unregistered causes "ptlrpc_set_wait" to hang

• Status: Resolved
• Project: Lustre
• Affects Version/s: Lustre 2.0.0
• Fix Version/s: Lustre 2.1.0
• Type: Bug
• Priority: Minor
• Reporter: Hongchao Zhang
• Assignee: Hongchao Zhang
• Resolution: Fixed
• Severity: 3
• Bugzilla ID: 21,760

Description

in ptlrpc_set_wait, the bulk is not unregistered if the reply was unregistered asynchronously, and the phase of the request was also changed into
"RQ_PHASE_UNREGISTERING", which causes the "ptlrpc_set_wait" to hang for the bulk's state remains to be active (bulk->bd_network_rw == 1)

Comments

[Build Master (Inactive)]

Integrated in lustre-reviews » server,el6-i686 #122
LU-201 unregister bulk after unregistering reply

hongchao.zhang : ebadcfeb189e1b17fbd99eda1544aee243866786
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-reviews » client,el5-x86_64 #122
LU-201 unregister bulk after unregistering reply

hongchao.zhang : ebadcfeb189e1b17fbd99eda1544aee243866786
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-reviews » client,el6-x86_64 #122
LU-201 unregister bulk after unregistering reply

hongchao.zhang : ebadcfeb189e1b17fbd99eda1544aee243866786
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-reviews » client,ubuntu-x86_64 #122
LU-201 unregister bulk after unregistering reply

hongchao.zhang : ebadcfeb189e1b17fbd99eda1544aee243866786
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-reviews » client,el6-i686 #122
LU-201 unregister bulk after unregistering reply

hongchao.zhang : ebadcfeb189e1b17fbd99eda1544aee243866786
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-reviews » server,el5-i686 #122
LU-201 unregister bulk after unregistering reply

hongchao.zhang : ebadcfeb189e1b17fbd99eda1544aee243866786
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-reviews » server,el5-x86_64 #122
LU-201 unregister bulk after unregistering reply

hongchao.zhang : ebadcfeb189e1b17fbd99eda1544aee243866786
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-reviews » server,el6-x86_64 #122
LU-201 unregister bulk after unregistering reply

hongchao.zhang : ebadcfeb189e1b17fbd99eda1544aee243866786
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-reviews » client,el5-i686 #122
LU-201 unregister bulk after unregistering reply

hongchao.zhang : ebadcfeb189e1b17fbd99eda1544aee243866786
Files :

• lustre/ptlrpc/client.c

[Cory Spitz]

Are there other end cases to address other than the ones Oleg mentions in Lustre bugzilla 21760 comment #73?

[Build Master (Inactive)]

Integrated in lustre-master » client,el5-x86_64 #27
LU-201 unregister bulk after unregistering reply

Oleg Drokin : f51ba918d4cb214d379f8b60f383b4e3102e08de
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-master-centos5 #199
LU-201 unregister bulk after unregistering reply

Oleg Drokin : f51ba918d4cb214d379f8b60f383b4e3102e08de
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-master » client,el5-i686 #27
LU-201 unregister bulk after unregistering reply

Oleg Drokin : f51ba918d4cb214d379f8b60f383b4e3102e08de
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-master » client,ubuntu-x86_64 #27
LU-201 unregister bulk after unregistering reply

Oleg Drokin : f51ba918d4cb214d379f8b60f383b4e3102e08de
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-master » server,el6-x86_64 #27
LU-201 unregister bulk after unregistering reply

Oleg Drokin : f51ba918d4cb214d379f8b60f383b4e3102e08de
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-master » client,el6-i686 #27
LU-201 unregister bulk after unregistering reply

Oleg Drokin : f51ba918d4cb214d379f8b60f383b4e3102e08de
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-master » server,el5-x86_64 #27
LU-201 unregister bulk after unregistering reply

Oleg Drokin : f51ba918d4cb214d379f8b60f383b4e3102e08de
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-master » client,el6-x86_64 #27
LU-201 unregister bulk after unregistering reply

Oleg Drokin : f51ba918d4cb214d379f8b60f383b4e3102e08de
Files :

• lustre/ptlrpc/client.c

[Build Master (Inactive)]

Integrated in lustre-master » server,el5-i686 #27
LU-201 unregister bulk after unregistering reply

Oleg Drokin : f51ba918d4cb214d379f8b60f383b4e3102e08de
Files :

• lustre/ptlrpc/client.c

[Hongchao Zhang]

patch has been pushed into master, close the issue, please reopen it if needed

[Cory Spitz]

Since we haven't had to revisit this, it might not really be a problem (or our networks are have integrity), but what about http://jira.whamcloud.com/browse/LU-201?focusedCommentId=12839&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-12839 ?