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.
It is changing the global variables "myname" and "myworkgroup"
to "global_myname" and "global_myworkgroup" respectively.
This is to make it very explicit when we are messing
with a global (don't ask - it makes the domain client
code much clearer :-).
Jeremy.
loadparm.c: Started add of 'security=domain' code.
password.c: Fix for security=server NT bugs.
reply.c: Started add of 'security=domain' code.
server.c: Started add of 'security=domain' code.
smb.h: Started add of 'security=domain' code.
Jeremy.
clientgen.c: Changed to fill change password buffer with random stuff.
password.c: Changed to get challenge from genrand.c
server.c: Added #ifdef around O_SYNC.
version.h: Changed to 1.9.19prealpha.
genrand.c:
New code to generate (hopefully) good random numbers for
use in crypto challenges/session keys etc.
PLEASE REVIEW THIS CODE AND SUGGEST IMPROVEMENTS !!!!!!
Jeremy.
proto.h: The unusual. ;)
reply.c: I changes some function names, and updated reply.c to match.
See mangle.c below for more.
server.c: Changed function names and parameters in file mangle.c, so
changed server.c calls to match. See mangle.c below for more.
mangle.c:
I replaced the caching mechanism used for caching reverse mangled name
maps. The old method was a large array of 256-byte strings. Movement
in the stack (including push and pop) was done by memcpy()ing whole
chunks of memory around.
The new system uses the ubi_Cache module which, in turn, uses a splay
tree. Entries are dynamically allocated using a minimum amount of
memory. Searches are non-linear, which should speed things up a bit,
too. Overall, this should save memory and be faster.
Other changes: I streamlined the is_mangled() test and made other
speed enhancements including replacing some static functions with
macros. Added comments, etc.
Note: Per an E'mail conversation with Andrew, the 'mangled stack'
parameter in smb.conf doesn't do anything anymore. The cache is
now set for 16K bytes maximum memory usage. The mangle stack
parameter is silently ignored. This can easily be changed, but
I'd rather introduce a 'mangled cache memory' parameter and
remove 'mangled stack'.
Remaining problems:
While testing the module, I noticed that something is calling
name_map_mangle() twice. The result is that names which contain
illegal characters are getting mangled twice.
Also, the entire module works by overwriting the input string.
This has a variety of nasty side effects.
Summary:
There's a lot still to be done, but the changes I have in place *should*
work in exactly the same way (except for the mangle stack parameter).
The rest of the bugs and other issues are separate.
Chris -)-----
load time (patch from "Marty Leisner" <leisner@sdsp.mc.xerox.com>.
server.c: Patch from Josef Hinteregger <joehtg@joehtg.co.at> - string
could be overwritten when find_service() called recursively.
Jeremy.
was so large that when used with -DUSE_MMAP it caused the unsigned
subtraction to wrap aound and become positive - thus causing
a silly memcpy offset. Thanks to "Michael St. Laurent" <rowl@earthlink.net>
for giving me the core dump that allowed me to track this one
down.
Jeremy.
I needed the client_name() and client_addr() functions in swat so I
could tell who was connecting from where. The problem was that these
functions didn't take a file descriptor parameter they just used
the global "Client".
So I needed to change all calls to pass a parameter ... lots of files.
the status page)
- split the claim_connection() code into its own file
- fixed the claim_connection() code to lock the file when manipulating
it
- always claim a null connection at startup
- fixed a bug in the pidfile code
I needed to modify the way the pidfile is handled in nmbd and smbd to
do this. Jeremy, you may wish to look at what I've done as it probably
breaks the Whistle use of pidfiles. In particular I've removed the -f
option and instead smbd and nmbd always create a pidfile in the lock
directory.
testparm.
In particular I added:
- ability to optionally save default values of all parameters when
calling lp_load(). This can then be used to save only non-default
parameters in lp_dump(). This makes the saved smb.conf (and viewed
parameters in testparm) much shorter
- ability to not load ipc share in lp_load()
- separators in parm_table[] so parameters can be grouped logically.
- flag to mark parameters that are local but which should be also
viewed as global as far as parameters editing is concerned
all I saw" - the book of Jeremy, chapter 1 :-).
So here is the mega-merge of the NTDOM branch server code.
It doesn't include the new client side pieces, we'll look
at that later.
This should give the same functionality, server wise, as
the NTDOM branch does, only merged into the main branch.
Any fixes to domain controler functionality should be
added to the main branch, not the NTDOM branch.
This code compiles without warnings on gcc2.8, but will
need further testing before we are sure all the working
functionality of the NTDOM server branch has been
correctly carried over.
I hereby declare the server side of the NTDOM branch
dead (and all who sail in her :-).
Jeremy.
algorithm was correct).
Finally (I think) fixed the mangled directory stack issue in
scan_directory() correctly. Mangled & non-mangled names are
now being checked correctly. Hurrah to Ulrik Dickow <ukd@kampsax.dk>
who helped isolate this one.
Jeremy.
to fix the fact that retransmit_or_expire_response_records() wasn't looking
at the WINS subnet.
server.c: Patch from jkf@soton.ac.uk to add %p (NIS server path) substitution.
smbpass.c: Fix to stop parsing failing on non-valid lines.
trans2.c: Fix for volume serial number code.
util.c: Patch from jkf@soton.ac.uk to add %p (NIS server path) substitution.
Fix for warnings under RH5. gcc 2.8.
Jeremy.
starts with log.smb and nmbd starts with log.nmb. It also gets rid of
the "log." when using the log.%m construct as %m expands to smb before
a client connects.
loadparm.c: Added "win95 bug compatibility" parameter.
local.h: Replaced MAX_OPEN_FILES back to 100 from 10 (oops).
reply.c: Fixed ulogoff check against uid - changed to vuid.
server.c: Changed file struct save of uid - changed to vuid.
smb.h: Changed id in struct current_user to vuid.
Changed file struct uid to vuid.
time.c: Added "win95 bug compatibility" atime -> mtime return.
trans2.c: Added "win95 bug compatibility" fixes.
uid.c: Changed id in struct current_user to vuid - added checks
to set/reset it.
util.c: Added code to expand environment variables.
version.h : still at 1.9.18 (head branch doesn't matter too much at present).
Jeremy.
chgpasswd.c: Fixed typo in debug message.
includes.h: Fix include for aix.
kanji.c: Added cap_to_sj as inverse of sj_to_cap.
loadparm.c:
local.h:
password.c: Added code for "networkstation user login" parameter.
- patch from Rob Nielsen <ran@adc.com>.
printing.c: Added further aix printing fixes.
reply.c: Changed access time fetch to a function.
trans2.c: Changed access time fetch to a function.
time.c: Changed access time fetch to a function.
server.c: Made NT redirector workaround final.
util.c: Added debug for write_socket failing.
Jeremy.
used, smbd tries to break an oplock to make room for another
file entry. This works well with Windows 95 that seems to keep
batch oplocks around for an arbitrarily long time.
Also changed rlimit code to ask for MAX_OPEN_FILES + 10 (if allowed) as
many systems use file descriptors for directory handles also.
Jeremy.
from Max Khon <max@iclub.nsu.ru>.
chgpasswd.c: Allow old RAP change password to work with encrypted
passwords. Samba can now allow Windows 95/NT clients to securely
change the Lanman password ! (But not the NT hash - that gets lost).
ipc.c:
smbdes.c:
smbpass.c: Support for the above.
server.c: #ifdef'ed out fix for NT redirector bug.
util.c: Fix NIS bug with server name.
Jeremy.
with smb_mode (smb_vwv[3]) of 0x20 == DENY_WRITE + read-only-open.
and smb_ofun (smb_vwv[8]) of 0x11 == Create if not exist plus append
on a read-only share.
This was mapped into the strange unix flags of (O_RDONLY|O_CREAT)
- essentially O_CREAT as O_RDONLY == 0.
We were checking the unix flags directly against O_RDONLY instead
of masking off the open mode flag bits before doing the comparison,
so this open was being refused even though it was valid on a
read-only share.
Also ensured that the O_CREAT bit was masked out of the flags
bit if the open was done on a read-only share (as doing a unix
open( filename, O_RDONLY|O_CREAT, xxx) will create a zero length
file if the user had permission to write into the directory - which
should be denied on a read-only share.
Thanks to Mark Peek @ Whistle for giving me this test case.
Jeremy.
server.c, util.c: Added fix for oplock break requests blocking due to server being
blocked in read call. Bug found by Charles Hoch <hoch@hplcgh.hpl.hp.com>.
Jeremy.
common bug in MS clients where they try to reuse a file descriptor
from an earlier smb connection. This code increases the chance that
the errant client will get an error rather than causing corruption
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.
