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.
nameelect.c: Tidied up settings of work->ServerType when unbecoming things.
nmbd.c: Fixed pidFile warning.
server.c: Fixed pidFile warning.
Jeremy.
(This used to be commit 94d53dcac5)
reply.c: Changed reply_open_and_X to split out the oplock
request bits from core and extended and if an oplock was granted only set
the corresponding bit on reply.
server.c: Added code to dynamically allocate i/o buffers in oplock_break
(prevents recursion problems) , also made reset of sent_oplock_break
explicit.
Jeremy.
(This used to be commit 16e55ee2b8)
can't sent a oplock break twice on the same file.
changed some debug levels in the oplock code to level 0 so we can
track down a bug
zero the returned Files[] entry in find_free_file()
don't try to overcome client bugs in the handling of non-encrypted
passwords if in server level security mode
added paranoid null termination of password buffers
slight change to my ajt_panic() routine
(This used to be commit e360c79c9c)
- fixed process_logon_packet() to always send the reply to/from the
correct netbios name (it should match the query)
- only do a NT style QUERYFORPDC if the query came in on a
NT_LOGON_MAILSLOT
Luke, can you check to see if these changes work in your branch with
NT logons?
(This used to be commit e89611985b)
files that aren't open if the file happened to close while the oplock
was in transit. We would end up sending a oplock break request on
another random file (actually the open file that happened to have the
highest fnum). Then we wouldn't get a response, so smbd would keep
sending and would actually flood the net with an infinite number of
oplock break requests!
(This used to be commit 0da20e0aec)
[%U]
path = %H
to get a new homes area. I wonder when it broke?
Anyway, I've fixed it. We can all rest easy now, knowing that you
really can produce an incomprehensible smb.conf and get it to work.
(This used to be commit 2245da4ad6)
added test in smbtorture for the server updating the directory modify
time when a file is added to a directory
cleanup in smbtorture so no garbage files are left on the server
(This used to be commit 3a5e07f1e9)
added support for choosing the protocol level in smbtorture (-m option)
use -1 for null date in cli_close()
get the attributes right in cli_open()
(This used to be commit d64d40a6ec)
"interpret_security" and similar functions. This also means that
testparm produces string values for these enumerated types rather than
an integer. This was done to make the web interface much cleaner.
While I was doing this I found a couple of very nasty bugs in the
parm_table code. Someone had added a handle_character_set() function
that treated the char** pointer for a P_STRING as an integer! This
would have caused memory corruption for anyone using the "character
set" option. The loadparm code is delicate - please be careful about
using it's more esoteric features!
I've also removed the "coding system" global and made it a
P_STRING. The integer value was never used anywhere in Samba.
(This used to be commit bac6b0daa2)
master browsers. Once we have become a local master, we would
*NEVER* stop announcing ourselves as such. This has serious
implications for the new nmbd code.
Jeremy.
(This used to be commit 8e8d239d9e)
with local message processing.
reply.c: Added check to reply_lockingX for chain after oplock break.
server.c: Added receive_next_smb().
trans2.c: Changed reply_trans2 to use receive_next_smb() to cope
with local message processing.
(This used to be commit f4ae644e13)
is 0x1c. This allows remote logon servers when using a Samba WINS server.
I did some experiments with a real NT WINS server and this behaviour
is much closer to being correct. It will only be totally correct when
we add proper group name support.
(This used to be commit c43b7e29ba)
typecasting for the debug messages of the data dump. hm.
#ifdef NTDOMAIN
ntclient.c rpc_pipes/lsaparse.c rpc_pipes/ntclientnet.c rpc_pipes/smbparse.c :
adding "Net Server Password Set" functionality to smbclient.
#endif
(This used to be commit ad43183d2d)
separated out smb server-mode password validation into a separate file.
added called and calling netbios names to client gen state: referenced
section in rfc1002.txt.
created workstation trust account checking code in ntclient.c
there might be a bug in reply_session_setup_andX. i indented and added { }
around single-line if statements: the lm password checking code now doesn't
look right (around the GUEST_SESSSETUP bits). *no code semantics have been
changed by the indentation process*.
(This used to be commit f27966957f)
understand the full range of byte offsets in locking requests. Samba
doesn't (due to the 31 bit limitation in fcntl locking)
(This used to be commit af1f408a05)
are ones that I know Samba fails. They are:
1) correct support for retaining locks over a close (ie. the server
must not use posix semantics)
2) support for lock timeouts
3) the server supports multiple locking contexts on the one SMB
connection, distinguished by PID.
4) the server correctly fails overlapping locks made by the same PID (this
goes against POSIX behaviour, which is why it is tricky to implement)
5) the server denies unlock requests by an incorrect client PID
I've been discussing with Jeremy ways that we can re-implement the
locking code to handle these correctly. This test code will be useful
to see that we have got it right.
(This used to be commit 097781e299)
chris' computer (probably about four minutes) meant that chris' commits
didn't get downloaded when i did a cvs update.
therefore, i did a cvs update; make proto; cvs commit, and still got problems.
so, five minutes later, i do a cvs update, and _then_ chris' server.c and
mangle.c mods get downloaded, and i have to do another make proto.
this particular problem has hit us *really* badly in the past, because i
was taking files home (onto a portable that i hadn't reset the time on
since i went to the states) and then recopying them back onto the computer
here. if i do this, i do a touch *.c *.h; cvs -t update. *then* a cvs commit.
(This used to be commit edb9c52948)
