IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
If the "socket address" parameter is a null string that is an invalid value for Samba 3.2 but valid for Samba 3.0.
Jeremy.
(This used to be commit c65726d418601cdc86202c0371615e4f7f3d843c)
xp/2003 explorer freezes browsing shares on samba ipv6 hosts. Caused by missing
reply packet to SMB printclose packet.
Jeremy
(This used to be commit ecf2b906f476e4a764d0e53eed84b9b75a2062c0)
map_nt_error_from_unix() now assumes that it is called in
an error path and returns an error even for a given errno == 0.
The original behaviour of unix_convert() used the mapping
of errno == 0 ==> NT_STATUS_OK to return success through
an error path.
I think this must have been an oversight, and unix_convert() worked
only by coincidence (or because explicitly using the knowledge
of the conceptually wrong working of map_nt_error_from_unix().
This patch puts this straight by not interpreting errno == 0
as an error condition and proceeding in that case.
Jeremy - please check!
Michael
(This used to be commit ec5956ab0df1b3f567470b2481b73da9c3c67371)
one of our virtualised functions, such as db_open(), but error is only
set when a system call fails, and it is not uncommon for us to fail a
function internally without ever making a system call. That led to us
passing back success when a function had in fact failed.
I found two places where we relied on map_nt_error_from_unix()
returning success when errno==0, but lots and lots of places where we
relied on the reverse, so I fixed those two places.
map_nt_error_from_unix() will now always return an error, returning
NT_STATUS_UNSUCCESSFUL if errno is 0
(cherry picked from commit 69d40ca4c1af925d4b0e59ddc69ef8c26e6501d1)
(This used to be commit 834684a524a24bb4eb46b4af583d39947dc87d95)
I think chain_reply() is one of the most tricky parts of Samba. This recursion
needs to go away, we need to sequentially walk the chain list.
(This used to be commit af2b01d85188d2301580643f7e862e3e3988aadc)
Untested code is broken code.... Test follows later, it's quite an intrusive
change to libsmb/
(This used to be commit 0ff16e8573f3c312f10fc723648319fa1f514ac0)
case files. Reported by Daniel Johnson <Progman2000@usa.net>.
The smb_set_file_time() call to set the filetimes is failing
because it's using the unmodified name passed in by the
client, not the modified name (matching case on the
disk) that comes out from create_file().
Jeremy.
(This used to be commit 1706a33e78347d14a8b09fd21b87d57bad543bcd)
then don't allow create_file() to call down to
create_file_unixpath() with a stream name.
Jeremy.
(This used to be commit 167d611b525db6d103a7f83fd5a19792be4e7745)
This fixes various build warnings on our platform. I'm sure I haven't
caught them all, but it's a start.
(This used to be commit 6b73f259cb67d9dda9127907d706f9244a871fa3)
This scans the list of pipes registered via rpc_pipe_register_commands instead
of using static tables.
(This used to be commit 283e6039989adea1c8921b3600b410cb67b6492a)
leaving panic or more appropriate action to callers.
This can for instance happen in a clustered setup, when
ctdb is not running. This allows for a more defined behaviour,
reducing chicken-egg-problems.
Michael
(This used to be commit 3b6541ff483d5fea4fd30bb6a01a560c05028d4e)
Jeremy, setting "exists" to True in _dfs_Add prevented the initial creation of
a new symlink for me, because the SMB_VFS_UNLINK failed. This also exists in
3.2. I only check it into 3.3 as I would like you to look at it first.
Thanks,
Volker
(This used to be commit f58b098a4172949018e84c4d0c722d6eb9bba514)
At least form_junctions() does not initialize refcount, and I don't see it in
get_referred_path(). For the latters, the callers might initialize it. But even
if they did, I think parse_msdfs_symlink() should unconditionally return the
number of referrals it found. I don't think it makes sense to count them up
from somewhere else.
(This used to be commit 73172113484a2c6c9af014f59e84707716c61a7d)
fsp_belongs_conn only used the vuid struct member anyway, and this is available
in the smb_request structure as well.
(This used to be commit 64e9372ab997739d46669c0cc4a4c6edb11d5e64)
check_fsp only used the vuid struct member anyway, and this is available in the
smb_request structure as well.
(This used to be commit 8d364c4c3311b406847158fc37e9208d298cf8ba)
Without this the changed checks in can_delete_file_in_directory give DELETE
access where there is none. So we can end up granting the ntcreate&x preparing
the unlink where we should not, which leads to a NT_STATUS_ACCESS_DENIED at
close time later, which in turn does *not* give the access denied error message
in the Windows GUI.
can_delete_file_in_directory will grant access now by looking at the directory
permissions.
(This used to be commit 51b5364c2afb3a18df4bec2bc1624760ccc01676)
With at least NFSv4 ACLs around the write permission for the owner is a bogus
check if we can delete a file in a directory. Like in Windows, there are two
ways which can grant us such: First, the DELETE permission on the file itself,
or if that does not help, the DELETE_CHILD permission on the directory. It
might be a bit more code that runs, but essentially we should end up with the
same set of syscalls in the non-acl case.
(This used to be commit daa9b056645a45edfb3a70e3536011ebe5678970)
Did not measure it, but I think a single write is better than a read and a
conditional branch
(This used to be commit abe1bed665ad8d1dbf9177dcbb9344b25df9594c)
