headers
elrond | 3 Sep 2007 22:28

cvs commit: tng/source/lib util_tng.c tng/source/passdb smbpass.c

elrond      2007/09/03 22:28:08 CEST

Modified files:
   source/lib             util_tng.c 
   source/passdb          smbpass.c 
Log:
pwdb_gethexpwd: Use uchar instead of char

Starting to clean up some useless signedness warnings of
some compilers

Internal-Name: sn-gethexpwd.diff

Revision  Changes                              Path
1.31      +2 -2;  commitid: RKnXoMxY65qIwkws;  tng/source/lib/util_tng.c
1.25      +2 -2;  commitid: RKnXoMxY65qIwkws;  tng/source/passdb/smbpass.c
Permalink | Reply | 0 comments |
headers
elrond | 3 Sep 2007 22:52

cvs commit: tng/source/include vfs.h tng/source/lib genrand.c util.c tng/source/smbd chgpasswd.c vfs-wrap.c vfs.c

elrond      2007/09/03 22:52:56 CEST

Modified files:
   source/include         vfs.h 
   source/lib             genrand.c util.c 
   source/smbd            chgpasswd.c vfs-wrap.c vfs.c 
Log:
Use SMB_STRUCT_DIRENT and sys_readdir more around

This stops some compilers complaining about dirent64 stuff.

Oh, and some const, while there.

Internal-Name: more-dirent64.diff

Revision  Changes                              Path
1.25      +1 -1;  commitid: DPDdJl7p0UedFkws;  tng/source/include/vfs.h
1.12      +1 -1;  commitid: DPDdJl7p0UedFkws;  tng/source/lib/genrand.c
1.66      +2 -2;  commitid: DPDdJl7p0UedFkws;  tng/source/lib/util.c
1.25      +1 -1;  commitid: DPDdJl7p0UedFkws;  tng/source/smbd/chgpasswd.c
1.23      +3 -3;  commitid: DPDdJl7p0UedFkws;  tng/source/smbd/vfs-wrap.c
1.44      +2 -2;  commitid: DPDdJl7p0UedFkws;  tng/source/smbd/vfs.c
Permalink | Reply | 0 comments |
headers
elrond | 4 Sep 2007 23:08

cvs commit: tng/source configure.in tng/source/smbwrapper config.m4

elrond      2007/09/04 23:08:38 CEST

Modified files:
   source                 configure.in 
Added files:
   source/smbwrapper      config.m4 
Log:
configure cleanup: Move smbwrapper tests out

We don't support smbwrapper currently, so I've decided to
move a big bunch of the configure tests into a new
smbwrapper/config.m4.
If we decide to support it later on, we can just include
that file from the main configure.in.

This saves about 10% of the time for configure. :)

Internal-Name: conf-out-smbw.diff

Revision  Changes                               Path
1.105     +3 -21;  commitid: dwck1Wu2rf29Isws;  tng/source/configure.in
1.1       +17 -0                                tng/source/smbwrapper/config.m4 (new)
Permalink | Reply | 0 comments |
headers
elrond | 16 Sep 2007 16:28

cvs commit: tng/source Makefile.in

elrond      2007/09/16 16:28:34 CEST

Modified files:
   source                 Makefile.in 
Log:
Consolidate make output messages

Create some make-vars for the output messages, makes things
nicer to look at, really.

Oh, and as I sometimes love to look at the generated
assembler code, I've added a target for it.

Revision  Changes                                Path
1.152     +71 -64;  commitid: 5LnsoXtaoZ0d7Yxs;  tng/source/Makefile.in
Permalink | Reply | 0 comments |
headers
elrond | 22 Sep 2007 19:56

cvs commit: tng/source/include doserr.h tng/source/smbd reply.c

elrond      2007/09/22 19:56:31 CEST

Modified files:
   source/include         doserr.h 
   source/smbd            reply.c 
Log:
Looks like LOCKING_ANDX_CHANGE_LOCKTYPE is unsupported

It seems, that even windows doesn't like
LOCKING_ANDX_CHANGE_LOCKTYPE. Or at least, it's not yet
understood, how it has to be called to actually do
something useful and not evoke an error.

So instead of doing something completely strange, we
currently error out. Better that way.

Thanks go to the testing team, as usual.

Internal-Name: lock-changetype-0.diff

Revision  Changes                               Path
1.6       +1 -0;  commitid: AHd4ljOggq3r3Lys;   tng/source/include/doserr.h
1.133     +11 -1;  commitid: AHd4ljOggq3r3Lys;  tng/source/smbd/reply.c
Permalink | Reply | 0 comments |
headers
elrond | 22 Sep 2007 20:05

cvs commit: tng/source/include doserr.h tng/source/smbd blocking.c reply.c

elrond      2007/09/22 20:05:29 CEST

Modified files:
   source/include         doserr.h 
   source/smbd            blocking.c reply.c 
Log:
Fully support LOCKING_ANDX_CANCEL_LOCK

lockingandx can cancel pending locks directly, by giving
them exactly again with the request to cancel them.
So there are many ways to cancel pending locks, fun.

Internal-Name: lx-cancel-2.diff

Revision  Changes                               Path
1.7       +1 -0;  commitid: UgR7Vi9uaqzD7Lys;   tng/source/include/doserr.h
1.23      +66 -0;  commitid: UgR7Vi9uaqzD7Lys;  tng/source/smbd/blocking.c
1.134     +17 -0;  commitid: UgR7Vi9uaqzD7Lys;  tng/source/smbd/reply.c
Permalink | Reply | 0 comments |
headers
elrond | 22 Sep 2007 20:33

cvs commit: tng/source/smbd nttrans.c reply.c trans2.c vfs.c

elrond      2007/09/22 20:33:27 CEST

Modified files:
   source/smbd            nttrans.c reply.c trans2.c vfs.c 
Log:
Create fsp_fill_ntinfo and use it around

fsp_fill_ntinfo is mostly a wrapper around vfs_fill_ntinfo,
just for "fsp"s.

Makes a lot of code shorter and just more portable.
Oh, and the binary gets smaller. ;)

Internal-Name: fsp_fill_ntinfo.diff

Revision  Changes                                Path
1.72      +20 -32;  commitid: c4tppVudAYgEfLys;  tng/source/smbd/nttrans.c
1.135     +19 -34;  commitid: c4tppVudAYgEfLys;  tng/source/smbd/reply.c
1.99      +10 -8;  commitid: c4tppVudAYgEfLys;   tng/source/smbd/trans2.c
1.45      +9 -0;  commitid: c4tppVudAYgEfLys;    tng/source/smbd/vfs.c
Permalink | Reply | 0 comments |
headers
elrond | 22 Sep 2007 20:49

cvs commit: tng/source/smbd filename.c open.c

elrond      2007/09/22 20:49:01 CEST

Modified files:
   source/smbd            filename.c open.c 
Log:
Create smbd_recheck_name and use it a little

check_name currently still modifies the passed in name.
We call it more often than needed, really. The second calls
do not modify the name, at least they shouldn't. So

smbd_recheck_name is for the second calls. It also errors
out, if the name would be changed. It allows to structure
the code better and use const a little more. :)

Internal-Name: smbd_recheck_name.diff

Revision  Changes                               Path
1.37      +24 -0;  commitid: CE8CYIEXgKSTlLys;  tng/source/smbd/filename.c
1.59      +4 -2;  commitid: CE8CYIEXgKSTlLys;   tng/source/smbd/open.c
Permalink | Reply | 0 comments |
headers
elrond | 23 Sep 2007 21:09

cvs commit: tng/source/smbd notify.c

elrond      2007/09/23 21:09:28 CEST

Modified files:
   source/smbd            notify.c 
Log:
Notify: Detect renames somewhat better

With "change notify timeout = 30" we now do some more
tracking of changes in directories, etc. This time mostly
for renames.

I'm not quite happy with this stuff anyway, but it's making
things somewhat better anyway.

Let's see, if I get some nicer ideas on how to handle all
this more internally.

Internal-Name: notify-3.diff

Revision  Changes                               Path
1.28      +25 -2;  commitid: rsPFGVw4OQRorTys;  tng/source/smbd/notify.c
Permalink | Reply | 0 comments |
headers
elrond | 23 Sep 2007 21:13

cvs commit: tng/source/smbd notify.c

elrond      2007/09/23 21:13:07 CEST

Modified files:
   source/smbd            notify.c 
Log:
Notify: Move Entry cleanup inside reply-packet function

Move list-removal and free() inside
change_notify_reply_packet(). This makes things nicer to
look at and easier to maintain.

Internal-Name: notify-4.diff

Revision  Changes                                Path
1.29      +23 -24;  commitid: gVDgM7US5ylUsTys;  tng/source/smbd/notify.c
Permalink | Reply | 0 comments |

Gmane