Hello,

 

I've got yet another byte range locking question.

 

MS-CIFS (v0.1) states in section 2.2.4.32.1:

 

"If the Locks vector contains one and only one entry
(NumberOfRequestedLocks ==

1) and TypeOfLock has the CANCEL_LOCK bit set, the client is requesting
that the

server cancel a previously requested but unacknowledged lock."

 

My newly pushed smbtorture4 test, part of RAW-LOCK-ASYNC, shows a W2K8R2
server accepting a request with the CANCEL_LOCK bit set that contains
multiple entries in the lock array, ie NumberOfRequestedLocks == 2.  The
server then seems to cancel the first outstanding lock and return
SUCCESS.

 

>From the description in the spec I would expect the server to return an
error, something like STATUS_INVALID_PARAMETER in this case.  Is this an
implementation specific issue to Windows or is the MS-CIFS spec
incorrect?

 

Attached is a pcap showing this behavior.  The new test was added to
smbtorture in:

 

http://gitweb.samba.org/?p=samba.git;a=commit;h=48358b3eaa425d8fbfec7bfd
8ccf56860b5a1ba0

 

Thanks,

Steven Danneman | Software Development Engineer
Isilon Systems    P +1-206-315-7500     F  +1-206-315-7501
www.isilon.com        

    How breakthroughs begin. (tm)

<<image001.gif>>

Attachment: multiple_cancel.pcap
Description: multiple_cancel.pcap

_______________________________________________
cifs-protocol mailing list
[email protected]
https://lists.samba.org/mailman/listinfo/cifs-protocol

Reply via email to