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.
with PCL drivers. The problem was we were updating the changeid on every
SETPRINTERDATA/DELETEPRINTERDATA call. We should not do this, we should
just update the 'setprinter' called count. We update the changeid on calls
to SETPRINTER/ADDPRINTER/ADDPRINTEREX etc. Also fixed the correct returning
of the create time on printers.
Jeremy.
the list of groups a user is in succeeds via winbind, we must allow the
lookup of the group name -> gid we are checking if the user is a member
of to go via winbind or /etc/group - as it may be a group on the local
box we are checking against. This is a subtle one.....
Jeremy.
instead of strtok - this fixes a bug with NT users with spaces in their
names when using winbindd. Needs to be added to the other parse_lpXX functions
(currently only added to lprng parsing code).
Jeremy.
CVS commit access :-) has written a simple routine that peeks inside the
MS PE printer driver file format and can tell if a driver is W2K or NT4.x.
So we can now correctly return the driver version number. Hurrah !
JF - this is the code you always wanted ..... :-) :-).
Jeremy.
I hope not). If you encounter strange file-serving behavior after this
patch then back it out. I analysed our stat() usage and realised we
were doing approx. 3 stat calls per open, and 2 per getattr/setattr.
This patch should fix all that. It causes the stat struct returned
from unix_convert() (which now *must* be passed a valid SMB_STRUCT_STAT
pointer) to be passed through into the open code. This should prevent
the multiple stats that were being done so as not to violate layer
encapsulation in the API's.
Herb - if you could run a NetBench test with this code and do a
padc/par syscall test and also run with the current 2.2.0 code
and test the padc/par syscalls I'd appreciate it - you should
find the number of stat calls reduced - not sure by how much.
The patch depends on unix_convert() actually finding the file
and returning a stat struct, or returning a zero'd out stat
struct if the file didn't exist. I believe we can guarentee this
to be the case - I just wasn't confident enough to make this
an assertion before.
Ok ok - I did write this whilst at the Miami conference.....
sometimes you get a little free time at these things :-).
Jeremy.
negative values from the mygetc() function. I've modified the return
line so that it should return values in the 0..255 range for legitimate
characters.
This change should probably be copied into SAMBA_2_2 but I haven't checked
that tree out yet.
Chris -)-----
the trust domain list reply on netlogon pipe was wrong, interim hack until
we have full trust relationships.
changed some unistr2 to parse the ending NULL char.
added a prs_align_needed() function. much like a prs_align but with a
condition. needed for the unistr2 parsing.
J.F.
falling back to the UNIX calls on error. This should fix all problems with
smbd enumerating all users in all groups in all trusted domains via winbindd.
Also changed GETDC to query 1C name rather than 1b name as only the PDC
registers 1b.
Jeremy.
falling back to the UNIX calls on error. This should fix all problems with
smbd enumerating all users in all groups in all trusted domains via winbindd.
Jeremy.
to open_socket_in() from within the open_sockets() function. This has some
effect on the way port numbers are assigned. Basically, if we use the -r
switch we are saying 'use port 137'. If we can't do that, there should
be an error message and a failure. If we don't use -r then we simply select
the first available port. The way it was working, if we selected -r and
port 137 was in use, wierd things would happen as open_socket_in() tried to
find another port.
when looking up the WINS server address.
Please *don't* use lp_wins_server() any more!
The wins_srv_ip() function has the following features:
- If the WINS server was entered as a DNS name then the translation to an
IP address will already have taken place. We used to do this every time
a call to the WINS server was made. Ick.
- The return value of wins_srv_ip() is a struct in_addr. Ready to go.
- When WINS failover is fully implemented, you'll be able to enter a colon-
separated list of WINS servers via the WINS SERVER parameter. Using
lp_wins_server() directly will fail if this syntax is used.
Chris -)-----
.cvsignore remove config.h - not in this directory
include/profile.h profile changes
lib/messages.c added message to return debug level
libsmb/clierror.c cast to get rid of compiler warning
libsmb/smbencrypt.c cast to get rid of compiler warning
profile/profile.c add flush profile stats changes for profile struct
rpc_parse/parse_samr.c fix for compiler warning
rpc_server/srv_samr.c cast to get rid of compiler warning
smbd/ipc.c profile stats
message.c profile stats
smbd/negprot.c profile stats
smbd/nttrans.c profile stats
smbd/trans2.c profile stats
utils/smbcontrol.c new flush stats command
on exit. Needed to fix printing.tdb from groving to 300Mb+ if being
driven by smbclient clients that never ask for status... (effective
DOS attack :-).
Jeremy.
