| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
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.
(This used to be commit e6bda112ebe0d41f54c4249b5c2e1f24011347e1)
|
|
internals after Andrews' code had memset it to zero (this was causing core
dumps).
charcnv.c: Fixes for ISO8859-2 from Petr Hubeny <psh@capitol.cz>.
Jeremy.
(This used to be commit df8783ca76d543d200c743f515a185cfea2880df)
|
|
get_smbpwd_entry (now an internal function to smbpass.c)
to a more UNIX-like :
getsmbpwnam() - get entry by name.
getsmbpwuid() - get entry by uid.
Changed the type returned by the smbpasswd enumeration
functions to be a void * so that people don't come to
depend on it being a FILE *.
These abstractions should make it much easier to
replace the smbpasswd file with a better backend
in future.
Other files changed are to match the above changes.
Jeremy.
(This used to be commit 1161cfb7f2b0d5a6d3e2b524a14a6f325ce70efb)
|
|
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 -)-----
(This used to be commit 8759bec11ba483b2292b0e513b85c98ed5e3e2d4)
|
|
with gcc. (Not a big change although it looks like it :-).
Jeremy.
(This used to be commit cd2613c57261456485fe4eeecfda209ada70de8e)
|
|
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.
(This used to be commit a776058900a727591bd7b69debdaa25c0e31d693)
|
|
"name resolve order" parameter.
source/Makefile: Re-ordered link for name resolve order code.
source/clientgen.c:
source/clientutil.c: Added calls to resolve_name().
source/includes.h: Added HPUX zombie fix.
source/loadparm.c: Added new name resolve order parameter.
source/namequery.c: Re-wrote to include parsing of lmhosts file, new resolve_name()
function requested by John.
source/nmbd.c: Tell resolve_name not to do WINS lookups if we are the WINS server.
source/nmbd_lmhosts.c: Call lmhosts parsing functions in namequery.c
source/password.c: Call resolve_name() to lookup security=server name.
source/reply.c:
source/time.c:
source/trans2.c: "fake directory create times" fix from Jim Hague - hague@research.canon.com.au.
source/util.c: Removed isalnum() test in Get_Hostname() that seems to cause
problems on many systems.
Jeremy.
(This used to be commit 7f118970da7c43eaddcf92dc056d3e849f1e7d5c)
|
|
have the netbios name
- fix another kill connection bug
(This used to be commit c634b799874795d42dae28fb4440ea452dc89b1b)
|
|
(This used to be commit df62c80e1d04059905b8a3c5bf9073ba91331e99)
|
|
Allows long share lists to be browsed.
Browsing *into* a long share name still fails, though.
(Luke - you may need to look into this).
Jeremy.
(This used to be commit 5299d1b49f2bfd6cf84a687548904206f4a18a41)
|
|
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.
(This used to be commit 118ba4d77a33248e762a2cf843fb7cbc906ee6e7)
|
|
This is the checkin that fixes the infamous Visual C++ 'file
has changed' bug.
I feel *SO* good about that :-).
charset.c: Added (void) to fix Herb's fussy compiler.
loadparm.c: Removed "win95 bug compatibility" (didn't
like it much anyway :-). Added "dos filetime resolution"
instead.
reply.c: Added the 2 second timestamp resolution fix that the song above is about.
time.c: Removed unneeded get_access_time() function.
trans2.c : Removed unneeded "win95 bug compatibility" code.
Jeremy.
(This used to be commit 10d628e4aeaecc573de27e251fec7b91844cba40)
|
|
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.
(This used to be commit adc903bcf59ad1664babd7f1d43675d3a75bfbc9)
|
|
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.
(This used to be commit a031404623c22d62f8de035be2239f609af08112)
|
|
This gets the minimum timestamp associated with a file.
reply.c and trans2.c then return this as the create
time. Designed to fix problems with VC++ and others.
Jeremy.
(This used to be commit e3d5f6196d6eff707c78941696a368216e2a7410)
|
|
This is merely updating the Copyright statements from 1997 to 1998.
It's a once a year thing :-).
NO OTHER CHANGES WERE MADE.
Jeremy.
(This used to be commit b9c16977231efb274e08856f7f3f4408dad6d96c)
|
|
server.c: Test fix for NT worstation SMBmv oplock bug.
smbdes.c: Addition of 'forward' parameter in preparation of allowing
password change.
Jeremy.
(This used to be commit 0b0b1fb122a52e67a8fdc77d013ad0b3bbb90d19)
|
|
server.c: Fixed security=share problem where the vuid was still being
looked at.
Jeremy.
(This used to be commit ab8d615fe2004c3ca93dd2978ba988ea89d7fd74)
|
|
(This used to be commit 7b031586ca33a381eb0e27f3557f43c2550df5f8)
|
|
spaces
(This used to be commit 93f0619e049d1598db0c3022aeccf33910b0550f)
|
|
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.
(This used to be commit 16e55ee2b8be9a4210d8cf87691cdf42373759d2)
|
|
can't sent a oplock break twice on the same file.
changed some debug levels in the oplock code to level 0 so we can
track down a bug
zero the returned Files[] entry in find_free_file()
don't try to overcome client bugs in the handling of non-encrypted
passwords if in server level security mode
added paranoid null termination of password buffers
slight change to my ajt_panic() routine
(This used to be commit e360c79c9cec681c4609783019749773d3e79386)
|
|
(This used to be commit ee3042eefb47bbdbefc83ab3f0f407c7dea4d8a0)
|
|
with local message processing.
reply.c: Added check to reply_lockingX for chain after oplock break.
server.c: Added receive_next_smb().
trans2.c: Changed reply_trans2 to use receive_next_smb() to cope
with local message processing.
(This used to be commit f4ae644e13f2c4479dfc94c77c0a8295edf54086)
|
|
remove problems.
Jeremy
(This used to be commit 4a36ac236c2ad634f05efcd0179875d09988614a)
|
|
separated out smb server-mode password validation into a separate file.
added called and calling netbios names to client gen state: referenced
section in rfc1002.txt.
created workstation trust account checking code in ntclient.c
there might be a bug in reply_session_setup_andX. i indented and added { }
around single-line if statements: the lm password checking code now doesn't
look right (around the GUEST_SESSSETUP bits). *no code semantics have been
changed by the indentation process*.
(This used to be commit f27966957fa7f16d337a4a58719239d036deab4c)
|
|
added a #define around the alignment thing: it's a way to stop
NetMonitor from decoding your packets!!!!
proto.h :
usual.
reply.c :
added what i believe to be the correct error messages for getting
correct domain joining.
smb.h :
some guesses at good names of the SAMR_XXXX functions. sorting
out the SAMR_LOOKUP_RIDS function. this is *not* the same as
the LSA_LOOKUP_RIDS function, unless paul accidentally put it
on the ntlsa pipe by mistake, instead of the samr pipe :-)
rpc_pipes/lsa_hnd.c rpc_pipes/pipe_hnd.c :
moved creation and allocation of unique policy handles into this module.
rpc_pipes/pipesamr.c rpc_pipes/samrparse.c rpc_pipes/smbparse.c :
SAMR_LOOKUP_RIDS is beginning to look _suspiciously_ like the
LSA_LOOKUP_RIDS function. but i know that there are subtle
discrepancies.
(This used to be commit 6bc07b0b4193e28b13a675fece8d9d6b365a7eb0)
|
|
Modified Files:
---------------
Makefile:
adding extra files
ipc.c :
send_trans_reply() - alignment issue. this makes the alignment
the same as that in NT. this should be looked at by people who
understand the SMB stuff better than i.
api_fd_commands[] - added samr and wkssvc pipes.
loadparm.c :
lp_domain_controller() changed to mean "samba is a domain controller".
it's a "yes/no" parameter, now. no, it isn't used _anywhere_.
namedbwork.c nameelect.c :
if "domain controller = yes" then add SV_TYPE_DOMAIN_CTRL to the
host _and_ workgroup announcements. yes, you must do both: nt does.
namelogon.c :
important NETLOGON bug in SAMLOGON request parsing, which may be
the source of some people's problems with logging on to the Samba PDC.
password.c :
get_smbpwnam() renamed to get_smbpwd_entry().
pipes.c :
added samr and wkssvc pipes.
proto.h :
usual. can we actually _remove_ proto.h from the cvs tree, and
have it as one of the Makefile dependencies, or something?
reply.c :
get_smbpwnam() renamed to get_smbpwd_entry() - also changed response
error code when logging in from a WORKSTATION$ account. yes, paul
is right: we need to know when to return the right error code, and why.
server.c :
added call to reset_chain_pnum().
#ifdef NTDOMAIN added call to init_lsa_policy_hnd() #endif. jeremy,
you'd be proud: i did a compile without NTDOMAIN, and caught a link
error for this function.
smb.h :
defines and structures for samr and wkssvc pipes.
smbpass.c :
modified get_smbpwnam() to get_smbpwd_entry() and it now takes
two arguments. one for the name; if this is null, it looks up
by smb_userid instead.
oh, by the way, smb_userids are actually domain relative ids
(RIDs). concatenate a RID with the domain SID, and you have
an internet globally unique way of identifying a user.
we're using RIDs in the wrong way....
added mod_smbpwnam() function. this was based on code in smbpasswd.c
rpc_pipes/lsaparse.c :
added enum trusted domain parsing. this is incomplete: i need
a packet trace to write it properly.
rpc_pipes/pipe_hnd.c :
added reset_chain_pnum() function.
rpc_pipes/pipenetlog.c :
get_smbpwnam() function renamed to get_smbpwd_entry().
arcfour() issues.
removed capability of get_md4pw() function to automatically add
workstation accounts. this should either be done using
smbpasswd -add MACHINE$, or by using \PIPE\samr.
rpc_pipes/pipe_util.c :
create_pol_hnd() - creates a unique LSA Policy Handle. overkill
function: uses a 64 bit sequence number; current unix time and
the smbd pid.
rpc_pipes/smbparse.c :
arcfour() issues.
smb_io_unistr2() should advance by uni_str_len not uni_max_len.
smb_io_smb_hdr_rb() - request bind uses uint16 for the context
id, and uint8 for the num_syntaxes. oops, i put these both as
uint32s.
Added Files:
------------
rpc_pipes/lsa_hnd.c :
on the samr pipe, allocate and associate an LSA Policy Handle
with a SID. you receive queries with the LSA Policy Handle,
and have to turn this back into a SID in order to answer the
query...
rpc_pipes/pipesamr.c rpc_pipes/samrparse.c
\PIPE\samr processing. samr i presume is the SAM Replication pipe.
rpc_pipes/pipewkssvc.c rpc_pipes/wksparse.c
\PIPE\wkssvc processing. the Workstation Service pipe?
holy cow.
(This used to be commit 1bd084b3e690eb26a1006d616075e53d711ecd2f)
|
|
loadparm.c : to equivalent to version 1.67
reply.c : to equivalent to version 1.69
server.c : to equivalent to version 1.122
util.c : to equivalent to version 1.98
to remove the incorrect changes.
proto.h: The usual.
rpc_pipes/smbparse.c : Backeting stuff that SHOULD NOT BE IN THE
none-NTDOMAIN build !
Jeremy.
(This used to be commit 6064c9d80fd9fcc3ceec528494ba5e2591610098)
|
|
'<' and '|' characters indicate read file and execute command respectively,
and feed the output into the parameter (!!!).
'<$' and '|$' means run standard_sub_basic() on them.
this is going to be fun to document in smb.conf.5....
also, Christian created a new "online" service parameter. services can
be taken "off-line"....
(This used to be commit 15f44d28916cdc1432bffdbb999c7cf7efd8fb86)
|
|
which are pipes on the IPC$ connection.
created mechanism to record pipe names in a separate pipes_struct. it
is planned to expand this, to return sensible things like interface
structures, and policy handles (RPC_IFACE and LSA_POL_HND). and the like.
(This used to be commit 33cce5fac0e2f818a19a6c4e6a797ef44f3b5c75)
|
|
(This used to be commit c63fee2b282c8b53f87e63995384602b66a805a6)
|
|
(This used to be commit 0f15558efb26b7215540a588dfe8733e9346d407)
|
|
1) the oplock macros in smb.h used | where they should have used
&. This means that smbd thought that all clients were always
requesting oplocks. This would have _really_ confused smbclient
and smbfs when they started receiving async oplock break requests when
they don't even know what an oplock is!
2) an oplock break request from a client can be embedded in a normal
lockingX request, and will be if the client has batched any lock
requests internally. The smbd code assumed that all oplock break
requests had num_locks==num_ulocks==0 which is not true. The only
thing special about a oplock break request with
num_locks==num_ulocks==0 is that no reply is sent. Otherwise it is
processed as a normal locking request in addition to the oplock break
processing.
These two fixes get the MS mail system in Win98 working on a Samba
1.9.18 network drive.
Andrew
(This used to be commit ed71534df56d0296280dbde1859597fb42002088)
|
|
adding bits for new nt domain code
byteorder.h :
trying to get macros right, and not to crash on SUNOS5...
client.c :
added #ifdef NTDOMAIN, and created do_nt_login() function. don't
want to have to recompile client.c unless absolutely necessary.
credentials.c :
moved deal_with_creds() [possibly inappropriately] into credentials.c
ipc.c reply.c server.c uid.c :
attempting to make (un)become_root() functions calleable from smbclient.
this is a little tricky: smbclient might have to be another setuid
root program, immediately setuid'ing to non-root, so that we can
reset-uid to root to get at the smbpasswd file. or, have a secure
pipe mechanism to smbd to grab smbpasswd entries. or the like.
smbdes.c smbencrypt.c :
created a function to generate lm and nt owf hashes.
lsaparse.c ntclient.c smbparse.c :
added nt client LSA_AUTH2 code. it works, too!
pipenetlog.c pipentlsa.c pipesrvsvc.c :
simplification. code-shuffling. getting that damn offset right
for the opcode in RPC_HDR.
smb.h :
changed dcinfo xxx_creds to DOM_CRED structures instead of DOM_CHAL.
we might need to store the server times as well.
proto.h :
the usual.
(This used to be commit 82436a3d99d4bdce249ce9ff27fd2ca4b2447e07)
|
|
interface
The new code uses a source netbios name equal to the Samba servers
name, not the client name. It also uses NetWkstaUserLogon to do a full
network logon. This means it will honour the servers logon
restrictions (such as login times etc).
(This used to be commit 11de90f972f6d83974425e80014f54e15d495413)
|
|
'dos filetimes' parameter for UTIME fix.
locking_shm.c: Fixed typo (sorry Andrew :-).
namepacket.c: Changed lp_interfaces_only() to lp_bind_interfaces_only().
proto.h: The usual.
reply.c: Made filetime calls use new file_utime call (wrapper for sys_utime).
server.c: Made filetime calls use new file_utime call (wrapper for sys_utime).
system.c: Added Andrew's sanity checks to times in sys_utime().
time.c: Moved set_filetime() to server.c. Made null_mtime() global.
trans2.c: Made filetime calls use new file_utime call (wrapper for sys_utime).
Jeremy (jallison@whistle.com)
(This used to be commit 41a1d81c112a82ad2ae1b3c4ee81051f133ce1ed)
|
|
The shares modes code is now split into separate files. The shared
memory implementation is in locking_shm.c. The slow implementation is
in locking_slow.c
It is all controlled by a struct share_ops structure that has function
pointers to the implementation of all the functions needed by a share
modes implementation. An initialisation function sets up this
structure. This will make adding new implementations easy and clean.
This also allowed me to get rid of the ugly code in smbstatus. Now
status.c links to the locking code and calls methods in share_ops.
I also renamed some things and generally organised things in a much
cleaner fashion. Defines and structures specific to each
implementation have been moved to the appropriate file and out of
smb.h.
(This used to be commit 65ab9adaa0d356b8041ed8a507ea52117f2a284e)
|
|
whatever that means.
(This used to be commit ce54f3a7a965b0a9f888dc45d6f4633609d5ca08)
|
|
locking.c: Adding Andrews become_root code to the main branch.
pipes.c: Fixing the close_file issue.
proto.h: The usual.
reply.c: Move smb_pass into NTDOMAIN defined code. Fixing the close_file issue.
server.c: Fixing the close_file issue.
trans2.c: Fixing the close_file issue.
uid.c: Adding Andrews become_root code to the main branch.
Jeremy (jallison@whistle.com)
(This used to be commit 16fd4337f79ce33f91050c96c4a566221c5d9126)
|
|
whoops, the SAM Logon structure was wrong. updated this, and
cifsntdomain.txt. more debug info in pipenetlog.c. the crash
is somewhere around deal_with_credentials().
byteorder.h :
put in uint8, uint16 and uint32 typecasts around debug info, because
sign extending was resulting in ffffffe8 being displayed instead of e8.
credentials.c :
some debugging info, because i'm tracking a coredump. without gdb.
nothing like making things difficult.
reply.c :
whoops, missed this (important) bit from paul's code, which tells
the NT workstation that the MACHINE$ entry doesn't already exist,
and we're going to create a default entry with a password "machine"
right now.
proto.h:
the usual.
(This used to be commit ed606bc7d4e6fb1091e527ea70a3e950d50a1db4)
|
|
namedbsubnet.c: Stop registering 1x name unless we can be a local master.
reply.c: Remove ERRbaddirectory code.
server.c: Remove abort() - use exit_server() instead.
trans2.c: Remove ERRbaddirectory code.
Jeremy (jallison@whistle.com)
(This used to be commit 76247228896d39312ba896fa229076be3271e2e4)
|
|
change MAX_PASSWORD_LENGTH to MAX_PASS_LEN to prevent conflict on some
systems
add #ifdef around soft link dependent code (for systems that don't
have soft links)
(This used to be commit e10ba4b97a219e87b62d32834bf15ed2e323ed2e)
|
|
locking.c: Removed USE_OPLOCKS - now the default.
params.c: Removed unused variable.
proto.h: Updated.
reply.c: Removed USE_OPLOCKS - now the default.
server.c: Removed USE_OPLOCKS - now the default.
smb.h: Removed USE_OPLOCKS - now the default.
smbparse.c: Changed shadowed variable.
status.c: Removed USE_OPLOCKS - now the default.
util.c: Removed USE_OPLOCKS - now the default.
Jeremy (jallison@whistle.com)
(This used to be commit b93509846d6291771787af457500eec8984ee6bd)
|
|
solaris crashes.
reply.c: Added NT specific error code. Put oplock break code in correct place
in reply_lockingX.
server.c: Removed unneeded error mapping stuff.
Fixed race condition in oplock code.
trans2.c: Added NT specific error code.
util.c: Added paranoia check in interpret_addr. Some core dumps
reported here. Upped fcntl debug levels.
Andrew. Please check the NT specific error code handling (search
for the string "/* Ugly - NT specific hack - but needed (JRA) */",
this makes NT and 95 clients behave correctly here - please check
your Visual Basic apps with this code.
Jeremy (jallison@whistle.com).
(This used to be commit 97ee4a5f69bd9cfbbc8710a1a04d80db0ee40104)
|
|
change from ERRbaddirectory to ERRbadpath for ENOTDIR errors. This
reverts to the old Samba code. I've done quite a bit of testing
against NT4 and have yet to get it to produce the ERRbaddirectory
error code. Producing ERRbaddirectory made a visual basic application
that was sent to me not run. This might explain some of the "it
doesn't work any more" complaints we've got about 1.9.17.
Jeremy, can you remember how you got NT to produce ERRbaddirectory?
There might be some specific circumstances we need to cover.
(This used to be commit 1ed901ddffef48cdce87017bab0fd613e6ca8637)
|
|
Yipeee. At least as far as I can check in a short time :-).
local.h: Changed OPLOCK_BREAK_TIMEOUT to 30 seconds.
locking.c: Big changes to delete oplocks on a share mode entry.
proto.h: updated.
reply.c: Added oplock break code in lockingX reply & readbraw reply.
server.c: Add batch oplock code. Force server shutdown if client fails
to respond to oplock break.
smb.h: Fix silly slow share mode oplock define bug.
status.c: Add oplock status info.
Jeremy (jallison@whistle.com)
(This used to be commit 4c83d37239f15f855fc10f01d7b4bf4217fb9eda)
|
|
local.h: Add OPLOCK_BREAK_TIMEOUT.
password.c: Fix for paranoia password server security bug.
proto.h: Updated.
reply.c: Oplock changes.
server.c: Massive oplock changes - nearly there....
smb.h: oplock definitions.
util.c: Add local message processing queues for oplocks.
Jeremy (jallison@whistle.com)
(This used to be commit 92f1553db2cdf6f32881eb984a87050cf3e4760b)
|
|
Jeremy (jallison@whistle.com).
(This used to be commit b11833f1bdecf1c3c3e5bc7494741605cba92a1e)
|
|
so should have no effect on other work.
Jeremy (jallison@whistle.com)
(This used to be commit 7e3d4c8b21f63a06d32605d230129e36883ad08c)
|
