Age | Commit message (Collapse) | Author | Files | Lines |
|
<ying@almaden.ibm.com>.
Jeremy.
(This used to be commit 08c168242364bf4d415f49d134e507a7e234611b)
|
|
that libsmb/ creates a local tcp socket then launches smbd as a subprocess
attached to that socket. smbd thinks it is being launched from inetd.
to use it do the following:
- compile with -DSMB_REGRESSION_TEST
- run like this (also works with smbtorture etc)
export SMBD_TEST=1
export LIBSMB_PROG=bin/smbd
smbclient //server/share -Uuser%pass
obviously you need to setup a smb.conf etc. Using --prefix to configure
is useful.
The aim of all this stuff is to add a decent set of regression tests
to the build farm, so we know if smbd actually runs correctly on all the
platforms, not just builds. We can run smbtorture, masktest, locktest etc,
plus a bunch of smbclient scripts and any new tests we write.
This doesn't help much with nmbd (at least not yet) but its a good start.
(This used to be commit 7e8e6ae9a88c4d2587eb4e7f0501cd71bd36ebb2)
|
|
years old) bug when chainging a sessionsetup_and_X and tcon together.
The wrong username was being entered into the tdb, even though the
correct user was used for accessing files. This is related to the fact
that authorise_login() is not used for sessionsetup, but only for tcon
auths.
Jeremy.
(This used to be commit 0187cd6aef7586d7ad4bdc70c50f3f2e7c69519c)
|
|
but the code suffered from bitrot and is not now reentrant. That means
we can get bizarre behaviour
i've fixed this by making next_token() reentrant and creating a
next_token_nr() that is a small non-reentrant wrapper for those lumps
of code (mostly smbclient) that have come to rely on the non-reentrant
behaviour
(This used to be commit 674ee2f1d12b0afc164a9e9072758fd1c5e54df7)
|
|
send unaligned unicode strings sometimes!
Fixed our handling of the workgroup name tacked on the end of the
NT1 negprot response (a unaligned unicode)
fixed a couple of places where we should be using the message_end fns instead
of pre-calculated buffer lengths
(This used to be commit 86613493a9b2e56523153486931d0bf8d39beb7a)
|
|
Samba Users Group Japan, ensure that we don't use dos_to_unix(xx,True),
but always use dos_to_unix(xx,False) to prevent overwriting.
Jeremy.
(This used to be commit 244aec8ea623fec828add3ab09c5003bf32bd5c7)
|
|
Jermey.
(This used to be commit 05a2911403a0710d994a618e72743205a3b0b87a)
|
|
(to allow unmount)
(This used to be commit 15b17a80db605a55f667c95fb7e316877a441887)
|
|
Jeremy.
(This used to be commit 2c2fc8513699eb39721ac1d65fa1fdaecde526a8)
|
|
Jeremy.
(This used to be commit e033f758914f6631545af17041b4e49c1ff885d9)
|
|
An-embrace-and-extending we will go...
Jeremy.
(This used to be commit 43bf9b2c7aad15d65bac3a9c4f799be050a074cd)
|
|
(This used to be commit 53803b008717e8606b347ed8baab78121c6e9657)
|
|
Spotted by Joe Doran <joed@interlude.eu.org>
(This used to be commit 3e3b9bc5380652d882c02e7286258f0aabcaf395)
|
|
(This used to be commit a542f4513ab792363fd5772582c6d317aa913257)
|
|
smbd/posix_acls.c: Attempt to fix the "lose default acl" problem in Solaris.
Needs testing.
lib/sysacls.c: Typo fix.
Jeremy.
(This used to be commit d989f8bd3e1524183a24fb67be1af05b3289f648)
|
|
cleaned on clients abending connections. Thanks Andrew !
Jeremy.
(This used to be commit 1b3977c5367a0b713b194f369abd9872ae01ac2a)
|
|
Jeremy.
(This used to be commit f004471aac3a1a0f2e40c05d1aadfe68456bc103)
|
|
Jeremy.
(This used to be commit 4338ee78c3d7bcf4b9fac383ff2f572d882ab97c)
|
|
Jeremy.
(This used to be commit 30cffea8fa813a7448648b496a19e751441e436e)
|
|
Also added smbtorture test for this.
Jeremy.
(This used to be commit 6d65556ae8bea45a203defaded8436cbb56965e1)
|
|
We still need to set the large write reply correctly.
Jeremy.
(This used to be commit 810dae29b8b45e91c0c35a4d96202c08b13d4e82)
|
|
(This used to be commit 8ec9c87b5d1a7dae17d5b1a30f58effaf5e69e4b)
|
|
(This used to be commit 870e9e4e89dcf910c595cabb28f69ad070942072)
|
|
Jeremy.
(This used to be commit 18e652a5e0d30d033be70e512cd94bf867507f64)
|
|
(This used to be commit 68c0dcec3d3c17aae76ce093a601c9f452af5a0c)
|
|
(This used to be commit d6dcca7cc9bc5a1948c1b6126ca4f6cb9ccacc52)
|
|
otherwise security scanners may think we are vulnerable!
(This used to be commit ee8cb88682421464016d56209eecea764bddc032)
|
|
Jeremy.
(This used to be commit cf5015f15935605cf69078bc15251db61ddc48c7)
|
|
we already have space for this we just need to understand the length correctly).
Jeremy.
(This used to be commit 19145bae720bbcc32dcab380c62a33d1f0e3eef0)
|
|
Jeremy.
(This used to be commit ec64ddac5fa4688b18484d6599b17b8c4bcd34c0)
|
|
CAP_LARGE_READX|CAP_LARGE_WRITEX bits on negprot and out W2K
performance goes through the roof......
And as we *always* offer 64 buffers we can do this with this
simple change.....
Jeremy.
(This used to be commit c328dda0fa081e79049d7a9ddac06e80cc8b331b)
|
|
Jeremy.
(This used to be commit d7aa42e4593b02ee6e487f7a4633bd7e7620ef2f)
|
|
(This used to be commit 083b74c743f0026693fa0fbe665ed08a3ac706b8)
|
|
smbd/connection.c: Sync up with code in 2.2
Jeremy.
(This used to be commit 87025c223dd33f2e02060c2a5cd45502946c87c6)
|
|
We will need this when we use finer grained locking for max connections.
Jeremy.
(This used to be commit c6cd42a6791e26174eb795fd08ddbbd797e5a9cf)
|
|
- it didn't honour the clear flag
- it locked the database (not necessary)
the previous code would do 10^6 kill() operations for 1000 people
logging in - and all with the database locked. Not very scalable.
Still haven't added the counter, that would make it really efficient
(This used to be commit 6bdbeb4a32e52fb05328dcf38dbb90780e1761ef)
|
|
Now tested max connections and it works again. Hurrah !
Jeremy.
(This used to be commit 6565b8c9781f571735068aba5044a913b5fe31fe)
|
|
This needs testing !
Tidied up tabs in tdb.c.
Jeremy.
(This used to be commit 0852465053d0c1a23b6cfe6097267291b0595ef8)
|
|
pointed out by Elrond
(This used to be commit 572a37534b7ee010aafd7e8bcb9da52957306f97)
|
|
debug 0 warning.
Jeremy.
(This used to be commit be7e1d0f2c078bd07c7087e1e36636dcd6d0a5d8)
|
|
bits before checking if we should change them on non-acl systems.
Jeremy.
(This used to be commit aba243ca0867a0787f9f7c7b2cda6143bcc53087)
|
|
Removed "restrict acl with mask" - redundent.
Jeremy.
(This used to be commit 0db8a61d71f25ffa0e5c585e02e2fce973867156)
|
|
(This used to be commit 88a7b2c3f80b17ca8ab2112e9a98f55f1a11d88c)
|
|
%I getpeername() failure.
Jeremy.
(This used to be commit 15584d30cd51650752fcbdd4f0dfc193fed18e5a)
|
|
Jeremy.
(This used to be commit f100e091abc57a9ba983e7c3cf84bfda2dbc2e18)
|
|
o sed 's/pdb_clear_sam/pdb_free_sam/g'
o add pdb_reset_sam()
o password changing should be ok now as well.
(This used to be commit 96d0e7c3301ad990f6c83b9c216720cb32661fb5)
|
|
about it to stop account lockouts with "security=server" mode. Sorry for
the delay David.
Jeremy.
(This used to be commit e8819715038ed49b07ed5639b1b88ad12e994b53)
|
|
I did some basic tests but I have probably broken something. Notably the
password changing. So don't cry ;-)
J.F.
(This used to be commit a4a4c02b12f030a3b9e6225b999c90689dfc4719)
|
|
return code from smb_pam_accountcheck()
(This used to be commit d5d6f01aaf8d344bb44dbe047c2f760ca220529e)
|
|
and using autoconf tests.
Added "restrict acl with mask" parameter.
Jeremy.
(This used to be commit 7792e32ba7fd734cc68b354f31c382ac11521fe8)
|