Age | Commit message (Collapse) | Author | Files | Lines |
|
Allow for large files in test_setup_copy_chunk():
Write test data in 1M IOs, rather than attempting to do the whole
thing in one go.
Add copychunk bad resume key test:
Send a copy chunk request with an intentionally bogus resume key
(source key handle).
Add copychunk src=dest test:
Test copychunk requests where the source and destination handles refer
to the same file.
Add copychunk src=dest overlap test.
Add desired access args to test_setup_copy_chunk().
Add copychunk_bad_access test:
Open the copychunk source and destination files with differing
desired_access values. Confirm copychunk response matches 2k8 and 2k12
behaviour.
Add copy_chunk_src_exceed test:
Attempts to copy more data than is present in the copychunk source
file.
Add copy_chunk_src_exceed_multi test:
Test whether the first chunk in a multi-chunk copychunk request is
written to disk, where the second chunk is invalid due to src file
overrun.
Add copy_chunk_sparse_dest test:
Issue a request where the target offset exceeds the file size, resulting
in a sparse region.
Add copy_chunk_max_output_sz test.
Reviewed by: Jeremy Allison <jra@samba.org>
|
|
Also change test_ioctl_get_shadow_copy() to use torture_skip(), and
clean up test output.
Reviewed by: Jeremy Allison <jra@samba.org>
|
|
Reviewed by: Jeremy Allison <jra@samba.org>
|
|
If FSCTL_SRV_ENUM_SNAPS fails with NT_STATUS_NOT_SUPPORTED then skip the
test, this means we can run the full ioctl test suite as part of
autobuild.
Reviewed by: Jeremy Allison <jra@samba.org>
|
|
MS-SMB2 3.3.4.4 documents cases where a ntstatus indicating an error
should not be considered a failure. In such a case the output data
buffer should be sent to the client rather than an error response
packet.
Add a new fsctl copy_chunk test to confirm field limits are sent back
in response to an oversize chunk request.
Reviewed by: Jeremy Allison <jra@samba.org>
|
|
Found by Coverity.
|
|
Reviewed-by: David Disseldorp <ddiss@samba.org>
Autobuild-User(master): David Disseldorp <ddiss@samba.org>
Autobuild-Date(master): Fri Nov 9 14:53:27 CET 2012 on sn-devel-104
|
|
metze
|
|
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Sun Sep 23 05:47:30 CEST 2012 on sn-devel-104
|
|
|
|
|
|
|
|
Test that when turning the related flag back off for the
last compound request, the return code changes from
invalid parameter to file closed.
Pair-Programmed-With: Michael Adam <obnox@samba.org>
|
|
|
|
|
|
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
|
|
metze
|
|
(Error was set to an unused variable)
Autobuild-User(master): Michael Adam <obnox@samba.org>
Autobuild-Date(master): Tue Sep 11 22:09:18 CEST 2012 on sn-devel-104
|
|
(Error was set to an unused variable)
|
|
(Error was set to an unused variable)
|
|
|
|
|
|
This test verifies that the delete on close flag is kept during
a disconnect and is still active on the reconnected handle.
When the reconnected handle is closed, the file is deleted.
|
|
This test used to flap because there was an invalid check of
the file-id. This might or might not be the same after a
new open. Hence the flapping.
The new version simply opens a file as durable handle with
delete_on_close set, writes a byte and closes the file.
Then on a new connect, the file is opened again, and it is checked
that the file has been created (again) and that it is empty.
|
|
I/O and one handle
Signed-off-by: Stefan Metzmacher <metze@samba.org>
|
|
connection only
And do a tcp disconnect followed by a session reconnect instead
of immediately using the already opened second connection.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
|
|
The DH2Q response blob is not sent upon successful durable handle v2 reconnect.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
|
|
struct)
Signed-off-by: Stefan Metzmacher <metze@samba.org>
|
|
only use one smb2_create i/o struct and store the create GUID centrally
Signed-off-by: Stefan Metzmacher <metze@samba.org>
|
|
don't copy the old request, but only set the necessary fields
Signed-off-by: Stefan Metzmacher <metze@samba.org>
|
|
Signed-off-by: Stefan Metzmacher <metze@samba.org>
|
|
Signed-off-by: Stefan Metzmacher <metze@samba.org>
|
|
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
|
|
They show that only the content of the DH2C request matches and
all other values in the SMB2 Create request are ignored.
metze
|
|
metze
|
|
This is fixed in Windows 2012 and matches the [MS-SMB2] documentation
now.
metze
|
|
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Mon Aug 27 17:43:09 CEST 2012 on sn-devel-104
|
|
metze
|
|
metze
|
|
metze
|
|
changing them
metze
|
|
metze
|
|
metze
|
|
This way we can give anonymous full access to the directory.
metze
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Fri Jun 22 11:30:06 CEST 2012 on sn-devel-104
|
|
metze
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Fri Jun 22 03:37:54 CEST 2012 on sn-devel-104
|
|
metze
|
|
metze
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Thu Jun 21 10:19:00 CEST 2012 on sn-devel-104
|
|
|
|
to make smbtorture report the error instead of complaining about missing torture_ call
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User: Stefan Metzmacher <metze@samba.org>
Autobuild-Date: Tue May 29 13:57:42 CEST 2012 on sn-devel-104
|
|
metze
Autobuild-User: Stefan Metzmacher <metze@samba.org>
Autobuild-Date: Thu May 17 21:53:11 CEST 2012 on sn-devel-104
|