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.
* Finally fix parsing idmap uid/gid ranges not to break with spaces
surrounding the '-'
* Allow local groups to renamed by adding info level 2 to
_samr_set_aliasinfo()
* Fix parsing bug in _samr_del_dom_alias() reply
* Prevent root from being deleted via Samba
* Prevent builting groups from being renamed or deleted
* Fix bug in pdb_tdb that broke renaming user accounts
* Make sure winbindd is running when trying to create the Administrators
and Users BUILTIN groups automatically from smbd (and not just check the
winbind nexted groups parameter value).
* Have the top level rid allocator verify that the RID it is about to
grant is not already assigned in our own SAM (retries up to 250 times).
This fixes passdb with existing SIDs assigned to users from the RID algorithm
but not monotonically allocating the RIDs from passdb.
(This used to be commit db1162241f)
Now that I know what all the requirements for this group are
I can generalize the code some more and make it cleaner.
But at least this is working with lusrmgr.msc on XP and 2k now.
(This used to be commit d2c1842978)
group IFF sid_to_gid(S-1-5-32-544) fails and 'winbind nested groups = yes'
* Add a SID domain to the group mapping enumeration passdb call
to fix the checks for local and builtin groups. The SID can be
NULL if you want the old semantics for internal maintenance.
I only updated the tdb group mapping code.
* remove any group mapping from the tdb that have a
gid of -1 for better consistency with pdb_ldap.c.
The fixes the problem with calling add_group_map() in
the tdb code for unmapped groups which might have had
a record present.
* Ensure that we distinguish between groups in the
BUILTIN and local machine domains via getgrnam()
Other wise BUILTIN\Administrators & SERVER\Administrators
would resolve to the same gid.
* Doesn't strip the global_sam_name() from groups in the
local machine's domain (this is required to work with
'winbind default domain' code)
Still todo.
* Fix fallback Administrators membership for root and domain Admins
if nested groups = no or winbindd is not running
* issues with "su - user -c 'groups'" command
* There are a few outstanding issues with BUILTIN\Users that
Windows apparently tends to assume. I worked around this
presently with a manual group mapping but I do not think
this is a good solution. So I'll probably add some similar
as I did for Administrators.
(This used to be commit 612979476a)
Not a bug in the strictest sense, more a clarification. This whole routine
assumes new_gid != NULL anyway, so there's no point in checking.
Volker
(This used to be commit dfbf09c772)
realloc can return NULL in one of two cases - (1) the realloc failed,
(2) realloc succeeded but the new size requested was zero, in which
case this is identical to a free() call.
The error paths dealing with these two cases should be different,
but mostly weren't. Secondly the standard idiom for dealing with
realloc when you know the new size is non-zero is the following :
tmp = realloc(p, size);
if (!tmp) {
SAFE_FREE(p);
return error;
} else {
p = tmp;
}
However, there were *many* *many* places in Samba where we were
using the old (broken) idiom of :
p = realloc(p, size)
if (!p) {
return error;
}
which will leak the memory pointed to by p on realloc fail.
This commit (hopefully) fixes all these cases by moving to
a standard idiom of :
p = SMB_REALLOC(p, size)
if (!p) {
return error;
}
Where if the realloc returns null due to the realloc failing
or size == 0 we *guarentee* that the storage pointed to by p
has been freed. This allows me to remove a lot of code that
was dealing with the standard (more verbose) method that required
a tmp pointer. This is almost always what you want. When a
realloc fails you never usually want the old memory, you
want to free it and get into your error processing asap.
For the 11 remaining cases where we really do need to keep the
old pointer I have invented the new macro SMB_REALLOC_KEEP_OLD_ON_ERROR,
which can be used as follows :
tmp = SMB_REALLOC_KEEP_OLD_ON_ERROR(p, size);
if (!tmp) {
SAFE_FREE(p);
return error;
} else {
p = tmp;
}
SMB_REALLOC_KEEP_OLD_ON_ERROR guarentees never to free the
pointer p, even on size == 0 or realloc fail. All this is
done by a hidden extra argument to Realloc(), BOOL free_old_on_error
which is set appropriately by the SMB_REALLOC and SMB_REALLOC_KEEP_OLD_ON_ERROR
macros (and their array counterparts).
It remains to be seen what this will do to our Coverity bug count :-).
Jeremy.
(This used to be commit 1d710d06a2)
* \PIPE\unixinfo
* winbindd's {group,alias}membership new functions
* winbindd's lookupsids() functionality
* swat (trunk changes to be reverted as per discussion with Deryck)
(This used to be commit 939c3cb5d7)
can be taken out of it, so I decided to commit this in one lump. It changes
the passdb enumerating functions to use ldap paged results where possible. In
particular the samr calls querydispinfo, enumdomusers and friends have
undergone significant internal changes. I have tested this extensively with
rpcclient and a bit with usrmgr.exe. More tests and the merge to trunk will
follow later.
The code is based on a first implementation by Günther Deschner, but has
evolved quite a bit since then.
Volker
(This used to be commit f0bb44ac58)
(based on Simo's code in trunk). Rewritten with the
following changes:
* privilege set is based on a 32-bit mask instead of strings
(plans are to extend this to a 64 or 128-bit mask before
the next 3.0.11preX release).
* Remove the privilege code from the passdb API
(replication to come later)
* Only support the minimum amount of privileges that make
sense.
* Rewrite the domain join checks to use the SeMachineAccountPrivilege
instead of the 'is a member of "Domain Admins"?' check that started
all this.
Still todo:
* Utilize the SePrintOperatorPrivilege in addition to the 'printer admin'
parameter
* Utilize the SeAddUserPrivilege for adding users and groups
* Fix some of the hard coded _lsa_*() calls
* Start work on enough of SAM replication to get privileges from one
Samba DC to another.
* Come up with some management tool for manipultaing privileges
instead of user manager since it is buggy when run on a 2k client
(haven't tried xp). Works ok on NT4.
(This used to be commit 77c10ff9aa)
functions so we can funnel through some well known functions. Should help greatly with
malloc checking.
HEAD patch to follow.
Jeremy.
(This used to be commit 620f2e608f)
not exactly match what you would expect.
XP workstations during login actually do this, so we should better become a
bit more correct. The LDAP query issued is not really fully optimal, but it is
a lot faster and more correct than what was there before. The change in
passdb.h makes it possible that queryuseraliases is done with a single ldap
query.
Volker
(This used to be commit 2508d4ed1e)
correct anymore. If we actually open the tdb before the fork, we end up
opening the tdb twice. Jerry, jra, this also happens in the locking and
printing subsystems. You might want to check it there (not that it actually
happens right now, but this gave me some confusion lately...).
Volker
(This used to be commit 40cad9dcc1)
_samr_query_useraliases shows up with all kind of very weird memberships
(global-groups, machine-accounts, etc.). Sometimes even if there is no
alias-membership at all.
One of the biggest mistakes is to convert any unix-group the user is a
member of, into an alias by default in get_group_from_gid.
get_alias_user_groups should be rewritten to use
pdb_enum_alias_memberships.
Guenther
(This used to be commit 73ab2d2a74)
his book.
This prompted me to look at the code that reads the unix group list. This
code did a lot of name -> uid -> name -> sid translations, which caused
problems. Instead, we now do just name->sid
I also cleaned up some interfaces, and client tools.
Andrew Bartlett
(This used to be commit f9e59f8bc0)
* don't fall back to unmapped UNIX group for
get_local_group_from_sid()
* remove an extra become/unbecome_root() pair
from group enumeration
(This used to be commit da12bbdb0d)
displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings
on some of the 64-bit build farm machines as well as help us out when 64-bit
uid/gid/pid values come along.
(This used to be commit f93528ba00)
and migrate an NT4 domain and still logon from domain members
(tested logon scripts, system policies, profiles, & home directories)
(passdb backend = tdbsam)
removed call to idmap_init_wellknown_sids() from winbindd.c
since the local domain should be handled by the guest passdb backend
(and you don't really always want the Administrator account to be root)
...and we didn't pay attention to this anyways now.
(This used to be commit 837d7c54d3)
Jerry, this is assigned to you. Do you want to answer it?
However, we have to decide what to do if a mapping is to be done for a
unix group not in LDAP....
Volker
(This used to be commit bf449d467c)
* move rid allocation into IDMAP. See comments in _api_samr_create_user()
* add winbind delete user/group functions
I'm checking this in to sync up with everyone. But I'm going to split
the add a separate winbindd_allocate_rid() function for systems
that have an 'add user script' but need idmap to give them a RID.
Life would be so much simplier without 'enable rid algorithm'.
The current RID allocation is horrible due to this one fact.
Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow.
Nothing has changed in the way a samba domain is represented, stored,
or search in the directory so things should be ok with previous installations.
going to bed now.
(This used to be commit 0463045cc7)
The code was nice, but put in the wrong place (group mapping) and not
supported by most of the code, thus useless.
We will put back most of the code when our infrastructure will be changed
so that privileges actually really make sense to be set.
This is a first patch of a set to enhance all our mapping code cleaness and
stability towards a sane next beta for 3.0 code base
Simo.
(This used to be commit e341e7c49f)
We really need idmap_ldap to have a good solution with ldapsam, porting
it from the prvious code is beeing made, the code is really simple to do
so I am confident it is not a problem to commit this code in.
Not committing it would have been worst.
I really would have been able to finish also the group code, maybe we can
put it into a followin release after 3.0.0 even if it may be an upgrade
problem.
The code has been tested and seem to work right, more testing is needed for
corner cases.
Currently winbind pdc (working only for users and not for groups) is
disabled as I was not able to make a complete group code replacement that
works somewhat in a week (I have a complete patch, but there are bugs)
Simo.
(This used to be commit 0e58085978)
The intention is to remove the muliple passdb backends, but we need the
'guest' account to always be there. If the admin adds the guest account to
(say) LDAP, there will only be one backend required for operation.
This helps remove some nasty behaviours with adding accounts to the system
for both the RPC 'create user' and the SAMSYNC code. Users 'added' with
an 'add user/machine' script won't magicly appear, and machine accounts
'pre-added' to unix, but not the smbpasswd file will not cause mayhem.
This commit also implements somthing tridge discussed with me, the concept
of 'default' passdb operation pointers - so that each backend does not
need it's own stub funcitons wrapping the default tdb privilages/group
mapping code.
This also removes an implicit 'sid->name' and 'name->sid' mapping from our
own local SID space, to winbind usernames. When adding mapping for NIS/LDAP
non-sam users in future, we need to be careful.
Andrew Bartlett
(This used to be commit 6f32fa2349)
This patch makes Samba compile cleanly with -Wwrite-strings.
- That is, all string literals are marked as 'const'. These strings are
always read only, this just marks them as such for passing to other functions.
What is most supprising is that I didn't need to change more than a few lines of code (all
in 'net', which got a small cleanup of net.h and extern variables). The rest
is just adding a lot of 'const'.
As far as I can tell, I have not added any new warnings - apart from making all
of tdbutil.c's function const (so they warn for adding that const string to
struct).
Andrew Bartlett
(This used to be commit 92a777d0ea)
Currently this calls back to mapping.c, but we have the framework
to get the information into LDAP and the passdb.tdb (should we? I
think so..).
This has received moderate testing with net rpc vampire and
usrmgr. I found the add_groupmem segfault in add_aliasmem as
well, but that will be another checkin.
Volker
(This used to be commit f30095852f)
When creating a group you have to take care of the fact that the
underlying unix might not like the group name. This change gets around
that problem by giving the add group script the chance to invent a
group name. It then must only return the newly created numerical gid.
Volker
(This used to be commit b959419ed3)
might be ugly, etc - please don't blame me for anything but instead try to fix
the code :-). Compiling of the new sam system can be enabled with the
configure option --with-sam
Removing passdb/passgrp.c as it's unused
fix typo in utils/testparm.c
(This used to be commit 4b7de5ee23)
initialising function. This patch thanks to the work of
"Stefan (metze) Metzmacher" <metze@metzemix.de>
This is partly to enable the transition to SIDs in the the passdb.
Andrew Bartlett
(This used to be commit 96afea638e)
init_group_mapping() a static function and don't call it from any client
programs.
Not sure whether I've made a bigger mess here or not...
(This used to be commit 3c887d9021)
Samba (ab)uses the returns from getpwnam() a lot - in particular it keeps
them around for a long time - often past the next call...
This adds a getpwnam_alloc and a getpwuid_alloc to the collection.
These function as expected, returning a malloced structure that can be
free()ed with passwd_free(&passwd).
This patch also cuts down on the number of calls to getpwnam - mostly by
taking advantage of the fact that the passdb interface is already
case-insensiteve.
With this patch most of the recursive cases have been removed (that I know
of) and the problems are reduced further by not using the sys_ interface
in the new code. This means that pointers to the cache won't be affected.
(This is a tempoary HACK, I intend to kill the password cache entirly).
The only change I'm a little worried about is the change to
rpc_server/srv_samr_nt.c for private groups. In this case we are getting
groups from the new group mapping DB. Do we still need to check for private
groups? I've toned down the check to a case sensitve match with the new code,
but we might be able to kill it entirly.
I've also added a make_modifyable_passwd() function, that copies a passwd
struct into the form that the old sys_getpw* code provided. As far as I can
tell this is only actually used in the pass_check.c crazies, where I moved
the final 'special case' for shadow passwords (out of _Get_Pwnam()).
The matching case for getpwent() is dealt with already, in lib/util_getent.c
Also included in here is a small change to register the [homes] share at vuid
creation rather than just in one varient of the session setup. (This picks
up the SPNEGO cases). The home directory is now stored on the vuid, and I
am hoping this might provide a saner way to do %H substitions.
TODO: Kill off remaining Get_Pwnam_Modify calls (they are not needed), change
the remaining sys_getpwnam() callers to use getpwnam_alloc() and move
Get_Pwnam to return an allocated struct.
Andrew Bartlett
(This used to be commit 1d86c7f942)
Thou shalt not reference SAM_ACCOUNT members directly - always use
pdb_get/pdb_set.
This is achived by making the whole of SAM_ACCOUNT have a .private member,
where the real members live. This caught a pile of examples, and these have
beeen fixed.
The pdb_get..() functions are 'const' (have been for some time) and this
required a few small changes to constify other functions.
I've also added some debugs to the pdb get and set, they can be removed if
requested.
I've rewritten the copy_id2x_to_sam_pass() functions to use the new passdb
interface, but I need the flags info to do it properly.
The pdb_free_sam() funciton now blanks out the LM and NT hashes, and as such
I have removed many extra 'samr_clear_sam_passwd(smbpass)' calls as a result.
Finally, any and all testing is always appriciated - but the basics seem to
work.
Andrew Bartlett
(This used to be commit d3dd28f6c4)
not the privileges. Usually we don't need them, so the memory is free
early.
lib/util_sid.c: added some helper functions to check an SID.
passdb/passdb.c: renamed local_lookup_rid() to local_lookup_sid() and pass
an RID all the way. If the group doesn't exist on the domain SID,
don't return a faked one as it can collide with a builtin one. Some rpc
structures have been badly designed, they return only rids and force the
client to do subsequent lsa_lookup_sid() on the domain sid and the builtin
sid !
rpc_server/srv_util.c: wrote a new version of get_domain_user_groups().
Only the samr code uses it atm. It uses the group mapping code instead of
a bloody hard coded crap. The netlogon code will use it too, but I have to
do some test first.
J.F.
(This used to be commit 6c87e96149)
You can change them with either usermanager->policies->account
or from a command prompt on NT/W2K: net accounts /domain
we can add a rpc accounts to the net command. As the net_rpc.c is still
empty, I did not start. How should I add command to it ? Should I take the
rpcclient/cmd_xxx functions and call them from there ?
alse changed the SAM_UNK_INFO_3 parser, it's an NTTIME. This one is more
for jeremy ;-)
J.F.
(This used to be commit bc28a8eebd)
This time it's a PRIVILEGE_SET struct instead of a simple uint32 array. It
makes much more sense. Also added a uint32 systemaccount to the GROUP_MAP
struct as some privilege showing in USRMGR.EXE are not real privs but a
bitmask flag. I guess it's an heritage from NT 3.0 ! I could setup an NT
3.1 box to verify, but I'm too lazy (yes I still have my CDs).
Added 3 more LSA calls: SetSystemAccount, AddPrivileges and
RemovePrivileges, we can manage all this privilege from UserManager.
Time to change the NT_USER_TOKEN struct and add checks in all the rpc
functions. Fun, fun, fun.
J.F.
(This used to be commit 3f0a9ef2b8)
an array of uint32. That's not perfect but that's better.
Added more privileges too.
Changed the local_lookup_rid/name functions in passdb.c to check if the
group is mapped. Makes the LSA rpc calls return correct groups
Corrected the return code in the LSA server code enum_sids.
Only enumerate well known aliases if they are mapped to real unix groups.
Won't confuse user seeing groups not available.
Added a short/long view to smbgroupedit.
now decoding rpc calls to add/remove privileges to sid.
J.F.
(This used to be commit f29774e589)
include/proto.h: Fix missing (void) in proto.
rpc_server/srv_samr_nt.c: Fix user private group problem by filtering out groups that
clash with users.
smbd/posix_acls.c: Ensure default ACE's are sensible.
utils/pdbedit.c: Fix from Simo Sorce.
Jeremy.
(This used to be commit 29414fe0d6)
as they are generic "file line-by-line" reading routines. lines with
"#" at the front are ignored (as comments). this code started out as
the password file reading code.
(This used to be commit ef6df590fd)
* Added new APIs for modifying groups.
* RIDs are allocated similarly to NT, starting from 1000 and incrementing by 1
for each new user/group.
* RIDs are now consistently in hex
* Fixed bugs reported by Allan Bjorklund <allan@umich.edu>:
- ldap_close_connection is exported by OpenLDAP - changed to ldap_disconnect
- Missing ldap_connect() in getusergroups functions
- ldap_next_entry was being called too early while retrieving a sam_struct
- LDAP globals should be extern in sampassldap.c
* Fixed bugs reported by Martin Hofbauer <mh@bacher.at>
- Newly added workstation trust accounts had attributes DU rather than W.
- User dn's were forced to start with "uid=XX" rather than using the existing
dn.
(This used to be commit 91c77f5432)
attempt at taking lib/uid.c and getting a unix security context
change module that is independent of "cnums" and "snums".
a security context is needed for pipes, not just IPC$ or other
services.
- group database API
added add_group/alias_member, del_group/alias_member,
del_group/alias_entry functions. del_builtin_entry() is
deliberately set to NULL to cause an exception, you cannot
delete builtin aliases.
- parse_lsa.c srv_lsa.c
fixed lookup_names code, it was a load of trash and didn't do
anything.
- cmd_samr.c rpcclient.c srv_samr.c
added "deletegroup", "deletealias", "delaliasmem", "delgroupmem",
"addgroupmem", "addaliasmem", "createalias", "creategroup", to
both client and server code.
server code calls into unix stubs right now, which don't actually
do anything. the only instance where they are expected to do
anything is in appliance mode NOT even in the ldap code or anything.
client code modified to call samr_lookup_names() for group code
(because we can) and lsa_lookup_names() for alias code (because
we have to).
- srv_lookup.c
oops, lookup on unsplit name, we got lookup on DOMAIN, DOMAIN\name
instead of DOMAIN, name.
(This used to be commit b8175702ef)
query.
domain groups now work, hurrah! only thing is that the description is
one character long, don't know why (which is wierd in itself).
(This used to be commit 78a86c6496)
need to check, when looking up group members, that a group member is
a unix user [being mapped to an nt user] FIRST then if that fails
check that a group member is a unix group [being mapped to an nt group].
why? because you can have group names in a unix /etc/group file with
the same name as users.
this _might_ be a problem...
(This used to be commit 585d47644d)
- split sam_passwd and smb_passwd into separate higher-order function tables
- renamed struct smb_passwd's "smb_user" to "unix_user". added "nt_user"
plus user_rid, and added a "wrap" function in both sam_passwd and smb_passwd
password databases to fill in the blank entries that are not obtained
from whatever password database API instance is being used.
NOTE: whenever a struct smb_passwd or struct sam_passwd is used, it MUST
be initialised with pwdb_sam_init() or pwd_smb_init(), see chgpasswd.c
for the only example outside of the password database APIs i could find.
- added query_useraliases code to rpcclient.
- dealt with some nasty interdependencies involving non-smbd programs
and the password database API. this is still not satisfactorily
resolved completelely, but it's the best i can do for now.
- #ifdef'd out some password database options so that people don't
mistakenly set them unless they recompile to _use_ those options.
lots of debugging done, it's still not finished. the unix/NT uid/gid
and user-rid/group-rid issues are better, but not perfect. the "BUILTIN"
domain is still missing: users cannot be added to "BUILTIN" groups yet,
as we only have an "alias" db API and a "group" db API but not "builtin-alias"
db API...
(This used to be commit 5d5d7e4de7)
would only be a domain group rid. it can also be a local group rid,
which causes us problems in attempting to turn a unix gid into the
correct rid (domain group or local group).
sooo.... the fix is _in_ there, we just can't use it because it causes
link / knock-on problems in nmbd.
(This used to be commit e4ee653870)
now need search capability on S-1-5-20, which will need argh, a
"group database API" on S-1-5-20, and the ability to add BUILTIN\Admins
etc to "local group map" argh.
(This used to be commit a24f6eb00b)
unix groups are not explicitly mapped.
i.e as a PDC or BDC you can have domain groups, as a member of a domain
you cannot.
as a member of a domain, unmapped unix groups are assumed to be aliases,
and as a PDC or BDC, unmapped unix groups are assumed to be unix groups.
there is _one_ other check needed with aliases to be added: unmapped unix
groups that have the same name as an NT group on the PDC (for which i will
need to write an LsaLookupNames call) should be assumed to be domain groups
on the PDC.
(This used to be commit 53b49b44e1)
Tidied up some of the mess (no other word for it). Still doesn't
compile cleanly. There are calls with incorrect parameters that
don't seem to be doing the right thing.
This code still needs surgery :-(.
Jeremy.
(This used to be commit 18ff93a9ab)
the pre-alpha "domain group" etc parameters have disappeared.
- interactive debug detection
- re-added mem_man (andrew's memory management, detects memory corruption)
- american spellings of "initialise" replaced with english spelling of
"initialise".
- started on "lookup_name()" and "lookup_sid()" functions. proper ones.
- moved lots of functions around. created some modules of commonly used
code. e.g the password file locking code, which is used in groupfile.c
and aliasfile.c and smbpass.c
- moved RID_TYPE_MASK up another bit. this is really unfortunate, but
there is no other "fast" way to identify users from groups from aliases.
i do not believe that this code saves us anything (the multipliers)
and puts us at a disadvantage (reduces the useable rid space).
the designers of NT aren't silly: if they can get away with a user-
interface-speed LsaLookupNames / LsaLookupSids, then so can we. i
spoke with isaac at the cifs conference, the only time for example that
they do a security context check is on file create. certainly not on
individual file reads / writes, which would drastically hit their
performance and ours, too.
- renamed myworkgroup to global_sam_name, amongst other things, when used
in the rpc code. there is also a global_member_name, as we are always
responsible for a SAM database, the scope of which is limited by the role
of the machine (e.g if a member of a workgroup, your SAM is for _local_
logins only, and its name is the name of your server. you even still
have a SID. see LsaQueryInfoPolicy, levels 3 and 5).
- updated functionality of groupname.c to be able to cope with names
like DOMAIN\group and SERVER\alias. used this code to be able to
do aliases as well as groups. this code may actually be better
off being used in username mapping, too.
- created a connect to serverlist function in clientgen.c and used it
in password.c
- initialisation in server.c depends on the role of the server. well,
it does now.
- rpctorture. smbtorture. EXERCISE EXTREME CAUTION.
(This used to be commit 0d21e1e609)
groupdb/aliasdb.c groupdb/aliasfile.c groupdb/groupfile.c: Don't use snprinf, use slprintf.
include/includes.h: Fix YP problem.
include/smb.h: Fix ZERO_STRUCTP.
lib/util_sock.c: Added strerror() in debugs.
passdb/ldap.c: Don't use snprinf, use slprintf.
rpc_client/cli_lsarpc.c rpc_client/cli_pipe.c rpc_parse/parse_sec.c rpc_server/srv_pipe.c: Don't use snprinf, use slprintf.
script/installman.sh: DGUX changes.
smbd/open.c smbd/oplock.c: Fixed gcc warnings.
web/swat.c: Changes USER to SWAT_USER.
(This used to be commit 4c2b5a0098)
