Stefan Metzmacher [Fri, 2 Jan 2009 15:41:21 +0000 (16:41 +0100)]
s4:lib/events: remove unused events_internal.h
external users need to use tevent_internal.h anyway
metze
Stefan Metzmacher [Fri, 2 Jan 2009 15:27:48 +0000 (16:27 +0100)]
s4:selftest: report tevent tests as 'tevent.python'
metze
Stefan Metzmacher [Fri, 2 Jan 2009 14:34:37 +0000 (15:34 +0100)]
s4:lib/events: convert to use tevent_* functions instead of compat macros
metze
Stefan Metzmacher [Fri, 2 Jan 2009 15:28:38 +0000 (16:28 +0100)]
tevent: only provide compat macros if the caller wants them
metze
Stefan Metzmacher [Fri, 2 Jan 2009 15:27:11 +0000 (16:27 +0100)]
tevent: change pytevent to tevent_*
metze
Stefan Metzmacher [Fri, 2 Jan 2009 15:25:29 +0000 (16:25 +0100)]
tevent: use TEVENT_FD_* instead of EVENT_FD_*
metze
Stefan Metzmacher [Fri, 2 Jan 2009 12:39:56 +0000 (13:39 +0100)]
tevent: rename ev_debug_ops => tevent_debug_ops
And related changes, but we keep compat macros
metze
Stefan Metzmacher [Fri, 2 Jan 2009 13:20:37 +0000 (14:20 +0100)]
tevent: rename events => tevent in tevent.pc
metze
Stefan Metzmacher [Fri, 2 Jan 2009 12:26:32 +0000 (13:26 +0100)]
tevent: pass down handler_name and location to the backend layer
metze
Stefan Metzmacher [Fri, 2 Jan 2009 10:44:17 +0000 (11:44 +0100)]
tevent: make tevent_add_*() macros arround _tevent_add_*()
This passes some additional debug information into
the events system to make it easier to find bugs.
metze
Stefan Metzmacher [Fri, 2 Jan 2009 12:39:26 +0000 (13:39 +0100)]
tevent: use HAVE_EPOLL instead of HAVE_EVENTS_EPOLL
metze
Stefan Metzmacher [Fri, 2 Jan 2009 12:35:32 +0000 (13:35 +0100)]
tevent: rename event_register_backend() => tevent_register_backend()
And fix the callers and there init functions.
metze
Stefan Metzmacher [Fri, 2 Jan 2009 10:30:00 +0000 (11:30 +0100)]
tevent: rename event_context_init* => tevent_context_init*()
metze
Stefan Metzmacher [Fri, 2 Jan 2009 10:36:52 +0000 (11:36 +0100)]
tevent: rename event_[s|g]et_fd_flags => tevent_fd_[s|g]et_flags
metze
Stefan Metzmacher [Fri, 2 Jan 2009 10:32:29 +0000 (11:32 +0100)]
tevent: rename event_loop_* => tevent_loop_*
metze
Stefan Metzmacher [Fri, 2 Jan 2009 10:17:00 +0000 (11:17 +0100)]
tevent: rename event_set_default_backend() => tevent_set_default_backend()
metze
Stefan Metzmacher [Fri, 2 Jan 2009 11:53:05 +0000 (12:53 +0100)]
tevent: use libreplace headers instead of system headers
metze
Stefan Metzmacher [Fri, 2 Jan 2009 14:33:23 +0000 (15:33 +0100)]
s4:lib/events: deprecate event_context_find()
metze
Stefan Metzmacher [Fri, 2 Jan 2009 11:06:39 +0000 (12:06 +0100)]
s4:tevent: move event_context_find() to tevent_s4.c
metze
Stefan Metzmacher [Fri, 2 Jan 2009 15:06:31 +0000 (16:06 +0100)]
s4:pyregistry: use tevent_context_init()
metze
Stefan Metzmacher [Fri, 2 Jan 2009 17:10:26 +0000 (18:10 +0100)]
Revert "s4 winbind: Copy SID to avoid "discard const" warning"
This reverts commit
102028ec722d942d7f91eb92e8da4f1480d140d1.
state->ctx is the composite_context, which is a temporary context!
metze
Stefan Metzmacher [Fri, 2 Jan 2009 16:52:14 +0000 (17:52 +0100)]
Revert "s4 winbind: Avoid a "discards const" compiler warning."
This reverts commit
dd8d4b6fd178df6847052dac207a4aafc0697d71.
s3call->response.extra_data.data is a pointer and no fstring!
metze
Stefan Metzmacher [Fri, 2 Jan 2009 15:26:00 +0000 (16:26 +0100)]
librpc/gen_ndr: rerun make idl_full after pidl changes
metze
Stefan Metzmacher [Fri, 2 Jan 2009 15:22:18 +0000 (16:22 +0100)]
pidl: add dummy element for empty unions
This hopefully fixes the build on Solaris and IRIX.
metze
Volker Lendecke [Fri, 2 Jan 2009 13:53:26 +0000 (14:53 +0100)]
Alternative fix for
45db33e73 and
0d443ae7931
Simo is right, we need to ask passdb first. At least this fixes a nasty to find
NT_STATUS_ACCESS_DENIED problem in the build farm for the test run I just did
on host "opi".
Michael, can you re-check if this also fixes the error you found, leading to
the two fixes?
Thanks,
Volker
Volker Lendecke [Fri, 2 Jan 2009 13:51:23 +0000 (14:51 +0100)]
Revert "s:Fix uid_to_sid mapping when the idmap cache is empty."
This reverts commit
9a9b64dbdfce4414ada22d4f882c8c757b5813e1.
Volker Lendecke [Fri, 2 Jan 2009 13:51:06 +0000 (14:51 +0100)]
Revert "s3:idmap: Remove passd check from idmap_backends_unixid_to_sid()."
This reverts commit
45db33e73262d8e195a46fb96405dfb3dc43d6bc.
Michael Adam [Wed, 17 Dec 2008 21:27:39 +0000 (22:27 +0100)]
packaging(RHEL): fix typo in [u]mount.cifs installation
Michael
(cherry picked from commit
835108bcf84a5f47f46d237c7a93c572348125a1)
Signed-off-by: Michael Adam <obnox@samba.org>
Michael Adam [Wed, 17 Dec 2008 17:56:34 +0000 (18:56 +0100)]
packaging(RHEL): fix build of [u]mount.cifs
mount.cifs is now built by make everything.
And it needs extra objects anyways (mtab.o),
which is why the build of mount.cifs failed.
Michael
(cherry picked from commit
96c811cf222be5463a6705feae73d6215685c5f1)
Signed-off-by: Michael Adam <obnox@samba.org>
Stefan Metzmacher [Fri, 2 Jan 2009 08:44:53 +0000 (09:44 +0100)]
Happy New Year!
metze
Volker Lendecke [Thu, 1 Jan 2009 21:05:03 +0000 (22:05 +0100)]
Fix bug 5913.
Never seen this to be a problem, but it doesn't hurt either :-)
Volker Lendecke [Thu, 1 Jan 2009 18:55:20 +0000 (19:55 +0100)]
Remove a global variable
Volker Lendecke [Thu, 1 Jan 2009 18:47:40 +0000 (19:47 +0100)]
fix some nonempty blank lines
Volker Lendecke [Thu, 1 Jan 2009 17:16:42 +0000 (18:16 +0100)]
Remove two pointless globals
Volker Lendecke [Thu, 1 Jan 2009 12:28:38 +0000 (13:28 +0100)]
The Tru64 shell doesn't like (*) in a case statement
Tim Prouty [Thu, 1 Jan 2009 05:24:25 +0000 (21:24 -0800)]
s3: Fix caller of print_fsp_open
Jeremy Allison [Thu, 1 Jan 2009 02:34:52 +0000 (18:34 -0800)]
Fix typo found by Volker. Thanks for the review.
Jeremy.
Jeremy Allison [Thu, 1 Jan 2009 02:06:57 +0000 (18:06 -0800)]
Fix all warnings in source3 with gcc4.3.
Jeremy.
Jeremy Allison [Thu, 1 Jan 2009 00:30:11 +0000 (16:30 -0800)]
Fix more asprintf and "ignoring return code" warnings from gcc 4.3.
Jeremy.
Volker Lendecke [Thu, 25 Dec 2008 13:37:33 +0000 (14:37 +0100)]
Replace a static variable and alarm() calls by using sys_select()
Günther, please check!
Volker Lendecke [Thu, 25 Dec 2008 13:29:39 +0000 (14:29 +0100)]
Fix some nonempty blank lines
Volker Lendecke [Thu, 25 Dec 2008 13:26:17 +0000 (14:26 +0100)]
Make a CONST_DISCARD unnecessary by applying some const
Derrell, please check!
Volker Lendecke [Sat, 20 Dec 2008 09:42:02 +0000 (10:42 +0100)]
Replace "goto again" by "return NT_STATUS_RETRY" in receive_message_or_smb
This gives lower-level routines the chance to indicate a retry condition
Volker Lendecke [Thu, 25 Dec 2008 12:44:11 +0000 (13:44 +0100)]
Disable the use of splice(2)
On a debian lenny installation with a kernel that calls itself 2.6.28, I could
repeatably block smbd using splice. According to the output of netstat, the
recvq of the TCP connection was full and at the same time according to strace
smbd was stuck in a splice syscall trying to read from that TCP connection. We
can not protect ourselves from this condition, as nobody will ever notify us
that something is broken. So sadly we can't use splice yet at all. Maybe in
2.6.29...
Volker Lendecke [Tue, 23 Dec 2008 21:45:03 +0000 (22:45 +0100)]
Try to fix recvfile using splice on Linux
According to the splice(2) manpage, one of the file descriptors *must* be a
pipe. With this patch I could successfully run splice(2) on a debian lenny
installation.
Jeremy, please check!
Volker Lendecke [Wed, 24 Dec 2008 12:52:20 +0000 (13:52 +0100)]
Fix a "nested extern" warning
Volker Lendecke [Wed, 24 Dec 2008 12:51:47 +0000 (13:51 +0100)]
Rename parent_dirname_talloc() to parent_dirname()
Volker Lendecke [Wed, 24 Dec 2008 12:47:41 +0000 (13:47 +0100)]
Remove unused function parent_dirname
Volker Lendecke [Wed, 24 Dec 2008 12:46:22 +0000 (13:46 +0100)]
Use parent_dirname_talloc instead of parent_dirname in smb_unix_mknod
Volker Lendecke [Wed, 24 Dec 2008 12:44:44 +0000 (13:44 +0100)]
Use parent_dirname_talloc instead of parent_dirname in copy_internals
Volker Lendecke [Wed, 24 Dec 2008 12:41:42 +0000 (13:41 +0100)]
Use parent_dirname_talloc instead of parent_dirname in vfswrap_mkdir
Volker Lendecke [Wed, 24 Dec 2008 12:18:35 +0000 (13:18 +0100)]
Make use of TALLOC_MEMDUP in parent_dirname_talloc()
Volker Lendecke [Wed, 24 Dec 2008 12:05:25 +0000 (13:05 +0100)]
Remove unused function get_numlist()
Volker Lendecke [Wed, 31 Dec 2008 17:37:39 +0000 (18:37 +0100)]
recvfile can't be used for printing so far
Stefan Metzmacher [Mon, 29 Dec 2008 15:31:57 +0000 (16:31 +0100)]
s3:winbindd: we don't need to call messaging_reinit() twice
reinit_after_fork() already calls messaging_reinit()
metze
Volker Lendecke [Wed, 31 Dec 2008 09:49:57 +0000 (10:49 +0100)]
Enable individual smbtorture_s3 subtests in make test
Volker Lendecke [Tue, 30 Dec 2008 21:49:01 +0000 (22:49 +0100)]
Fix an ancient uninitialized variable read
The callers of open_file_ntcreate expect *psbuf to be filled correctly
Tim Prouty [Wed, 31 Dec 2008 02:46:53 +0000 (18:46 -0800)]
s3 build: Revert previous patch to fix the build
888524623d179ceb4d540f5ac0f3b013d3ce5735 broke the build. I'm
reverting it for the short term, while I work on a better fix.
Jeremy Allison [Wed, 31 Dec 2008 02:24:39 +0000 (18:24 -0800)]
Fix more "ignore return value" warnings from gcc 4.3.
Jeremy
Tim Prouty [Wed, 31 Dec 2008 00:02:09 +0000 (16:02 -0800)]
librpc: Re-run make idl-full from source3
Tim Prouty [Tue, 30 Dec 2008 21:47:58 +0000 (13:47 -0800)]
pidl: Fix a const build warning in the samba3 merged build
The warning:
../librpc/gen_ndr/ndr_spoolss.c: In function ‘ndr_pull_spoolss_EnumPrinterData’:
../librpc/gen_ndr/ndr_spoolss.c:14792: warning: passing argument 1 of ‘memset’ discards qualifiers from pointer target type
In some case the pidl generated code needs to memset a const struct
field. This causes the above warning. This patch discardeds the const
in this case. I'm open to a more elegant fix, but this eliminates the
warning and should be relatively safe.
Tim Prouty [Tue, 30 Dec 2008 21:41:18 +0000 (13:41 -0800)]
s3/s4: Move CONST_DISCARD into lib/util so it's available from s3 and s4
Tim Prouty [Tue, 30 Dec 2008 21:45:29 +0000 (13:45 -0800)]
s3 build: Remove redundant #defines that are in idl generated headers
This fixes a series of samba3 build warnings in the merged build:
'error: "X" redefined', where X is one of the redundant #defines
Tim Prouty [Tue, 30 Dec 2008 22:40:33 +0000 (14:40 -0800)]
s4: simplify some overly-complicated logic
Tim Prouty [Tue, 30 Dec 2008 20:53:34 +0000 (12:53 -0800)]
s3 build: Fix warning
Stefan Metzmacher [Thu, 18 Dec 2008 14:02:42 +0000 (15:02 +0100)]
s3:cluster: register the client connection via CTDB_CONTROL_TCP_ADD
This means we correctly pass IPv6 addresses too.
(CTDB_CONTROL_TCP_CLIENT was IPv4 only)
metze
Volker Lendecke [Tue, 30 Dec 2008 14:47:06 +0000 (15:47 +0100)]
Fix a memory leak in cups_pull_comment_location
We allocated "request" with ippNew, so we also should ippDelete it.
Volker Lendecke [Tue, 30 Dec 2008 13:05:26 +0000 (14:05 +0100)]
Fix a spinning smbd when printing
Without this, we end up adding more than one timed event. In the event handler
print_notify_event_send_messages() only one event will be deleted, all others
will fire indefinitely.
Volker Lendecke [Mon, 29 Dec 2008 21:07:56 +0000 (22:07 +0100)]
Print the vuid in BENCH-SESSSETUP
Volker Lendecke [Mon, 29 Dec 2008 21:06:08 +0000 (22:06 +0100)]
Second part of the bugfix for #5933
Incrementing the next vuid did not correctly overflow
Now we survive BENCH-SESSSETUP with -o 100000. Takes a while though :-)
Thanks a lot to Ofer Tal <otsmb@shmoop.org> for reporting #5933
Volker Lendecke [Mon, 29 Dec 2008 21:01:57 +0000 (22:01 +0100)]
First part of bugfix for #5933
Ofer Tal <otsmb@shmoop.org> fully correctly noted that we're incrementing
num_validated_vuids twice per session setup, but decrement it only once.
Looking at sesssetup.c we always call register_initial_vuid() before
register_existing_vuid(), so there's no point in incrementing it in
register_existing_vuid().
Jeremy, please check!
Volker Lendecke [Mon, 29 Dec 2008 20:07:41 +0000 (21:07 +0100)]
Move a comment to its place
Volker Lendecke [Mon, 29 Dec 2008 20:03:15 +0000 (21:03 +0100)]
Simplify invalidate_vuid slightly
get_valid_user_struct_internal() checks for UID_FIELD_INVALID itself
Volker Lendecke [Mon, 29 Dec 2008 20:01:13 +0000 (21:01 +0100)]
Simplify is_partial_auth_vuid slightly
Volker Lendecke [Mon, 29 Dec 2008 18:50:39 +0000 (19:50 +0100)]
Simplify invalidate_all_vuids() slightly
invalidate_vuid takes care of removing the user_struct from validated_users
Stefan Metzmacher [Tue, 30 Dec 2008 09:09:59 +0000 (10:09 +0100)]
Fix the merged build
metze
Stefan Metzmacher [Tue, 30 Dec 2008 07:58:52 +0000 (08:58 +0100)]
lib/tevent: fix standalone make distclean realdistclean
Bug report by Brad Hards <bradh@frogmouth.net>.
metze
Stefan Metzmacher [Mon, 29 Dec 2008 20:07:11 +0000 (21:07 +0100)]
pidl: don't generate invalid C code
When we don't know how to handle a type for the python bindings,
we should not generate invalid C code.
Jelmer: please fix do the full fix for this.
metze
Stefan Metzmacher [Mon, 15 Dec 2008 18:28:27 +0000 (19:28 +0100)]
s3:pdb_tdb: store the next rid counter in passdb.tdb instead of winbind_idmap.tdb
This upgrades the TDBSAM_VERSION to 4 and SAMU_BUFFER_V4.
metze
Stefan Metzmacher [Tue, 16 Dec 2008 10:51:14 +0000 (11:51 +0100)]
s3:passdb: add SAMU_BUFFER_V4 with no changes
The upgrade is required because of the followin TDBSAM_VERSION
upgrade.
metze
Stefan Metzmacher [Mon, 15 Dec 2008 17:46:37 +0000 (18:46 +0100)]
s3:passdb: make marshalling struct samu from and to a buffer more generic
This marshalling isn't specific to tdbsam and it's
ugly to have the related functions in two different files.
metze
Stefan Metzmacher [Mon, 29 Dec 2008 19:55:02 +0000 (20:55 +0100)]
s4:cluster: fix dependency to TDB_WRAP
metze
Stefan Metzmacher [Mon, 29 Dec 2008 19:24:57 +0000 (20:24 +0100)]
s4:lib/tevent: rename structs
list=""
list="$list event_context:tevent_context"
list="$list fd_event:tevent_fd"
list="$list timed_event:tevent_timer"
for s in $list; do
o=`echo $s | cut -d ':' -f1`
n=`echo $s | cut -d ':' -f2`
r=`git grep "struct $o" |cut -d ':' -f1 |sort -u`
files=`echo "$r" | grep -v source3 | grep -v nsswitch | grep -v packaging4`
for f in $files; do
cat $f | sed -e "s/struct $o/struct $n/g" > $f.tmp
mv $f.tmp $f
done
done
metze
Stefan Metzmacher [Mon, 29 Dec 2008 16:48:35 +0000 (17:48 +0100)]
s4:pvfs_aio: use tevent
metze
Stefan Metzmacher [Mon, 29 Dec 2008 16:33:51 +0000 (17:33 +0100)]
s4:sys_lease_linux: use tevent
metze
Volker Lendecke [Mon, 29 Dec 2008 16:08:38 +0000 (17:08 +0100)]
Fix SESSSETUP_BENCH torture test for long runtimes
deadtime hits without any files open. To enable long runtimes, open a file with
delete_on_close set.
Stefan Metzmacher [Mon, 29 Dec 2008 10:02:32 +0000 (11:02 +0100)]
s4:cluster: remove old/broken ctdb code
We don't need a full copy of ctdb...
If we want to readd cluster support,
we should just use a ctdb client library.
metze
Stefan Metzmacher [Mon, 29 Dec 2008 10:24:05 +0000 (11:24 +0100)]
s4: remove unused dbwrap code
If we readd this later, we should share the code with
samba3 and make it async/non-blocking.
metze
Stefan Metzmacher [Mon, 29 Dec 2008 10:52:53 +0000 (11:52 +0100)]
s4: Revert "r24284: change brlock_tdb.c to use the dbwrap API. This actually makes the"
This reverts commit
4792285282e2ec78ccb0ebc3cac6465d1d26ef10.
Conflicts:
source4/ntvfs/common/brlock_tdb.c
source4/ntvfs/common/config.mk
Volker Lendecke [Mon, 22 Dec 2008 21:42:44 +0000 (22:42 +0100)]
Add code to test write_data_iov a bit
Volker Lendecke [Mon, 22 Dec 2008 21:32:12 +0000 (22:32 +0100)]
Attempt to fix bug 5953: Make cli_send_smb_direct_writeX use writev
It seems there are SMB servers around which can't cope with the write header
being sent in a packet of its own. With writev we keep the advantage of direct
writes, giving the kernel the chance to coalesce the write calls.
Volker Lendecke [Mon, 22 Dec 2008 21:17:59 +0000 (22:17 +0100)]
Make write_data use write_data_iov
Volker Lendecke [Mon, 22 Dec 2008 21:17:28 +0000 (22:17 +0100)]
Add write_data_iov
Volker Lendecke [Sun, 21 Dec 2008 22:22:30 +0000 (23:22 +0100)]
Add sys_writev
Kai Blin [Sun, 28 Dec 2008 11:04:40 +0000 (12:04 +0100)]
s4 winbind: Copy SID to avoid "discard const" warning
Kai Blin [Fri, 26 Dec 2008 11:45:50 +0000 (12:45 +0100)]
s4 winbind: Add support for WINBINDD_GETGRGID call
Kai Blin [Fri, 26 Dec 2008 10:32:09 +0000 (11:32 +0100)]
s4 libnet: Add support for groupinfo by sid lookup
Kai Blin [Fri, 26 Dec 2008 10:08:13 +0000 (11:08 +0100)]
s4 winbind: Fix typos in debug output
Kai Blin [Fri, 26 Dec 2008 10:04:01 +0000 (11:04 +0100)]
wbinfo4: Add --gid-info option
Kai Blin [Fri, 26 Dec 2008 10:02:02 +0000 (11:02 +0100)]
wbinfo: Add --gid-info option.