1
0
mirror of https://github.com/samba-team/samba.git synced 2025-01-28 17:47:29 +03:00

44 Commits

Author SHA1 Message Date
Andrew Tridgell
64c0a5c4a0 handle clock skew in getatr test -
Andrew Tridgell
f9e3b91f58 fixed -c option to NBENCH test -
Andrew Tridgell
8095375623 added a "XCOPY" test that simulates the open calls made by xcopy /O
we currently fail this test
-
Andrew Tridgell
81fb8845e5 added a cli_qfilename() test to the trans2 tests -
Andrew Tridgell
22be40b6b8 got rid of a silly '*' in printout -
Andrew Tridgell
10ee1d22e6 - only show 1 cleanup msg per client
- show client count in progress
-
Andrew Tridgell
36f816a5e1 rewrote smbtorture to use the new dbench 2 format and methods -
Tim Potter
6a58c9bd06 Removed version number from file header.
Changed "SMB/Netbios" to "SMB/CIFS" in file header.
-
Tim Potter
1906903f06 Quieten some pointer cast warnings. -
Andrew Bartlett
a855dfb2e0 Update the NT_STATUS -> DOS error table.
This new table is rather different to the old one (see diff posted to the
list for a sorted list of differences) and needs a *lot* of testing.

It does however seem to line up much better with what NT is using, as
exampled by the change to the OBJECT_NAME_COLLISION DOS error, it now matches
win2k where it didn't before.

I can't see any critical errors we now get wrong, and I know that the auth
errors are correct as per my on-the-wire observations.

This table was produced (and I hope to comment this better later) by
using the ERRMAPEXTRACT smbtorture tool, a Win2k domain member and the
'name_to_ntstatus' auth module on the HEAD PDC.  This module returned
the username as the error, and the NT box was forced to give me a dos
error becouse thats all I negotiated on that connection.  Hence the map.

Andrew Bartlett
-
Andrew Bartlett
8f343beae7 Make a couple more of the warnings also be C comments -
Andrew Bartlett
d879d2ffff A few changes to always output the erorr mapping even when the error was
'squashed' or the session setup succeeded.  (How the latter occurs I don't know).
-
Andrew Bartlett
d47e61eb46 Workaround some 'smarts' in Win2k.
Win2k kills off connections that have issued a negprot but havn't yet issued
a session setup when a second connection appears from the same host.
-
Andrew Bartlett
da9444d98f Re-add bail on failure. -
Andrew Tridgell
27143900e7 fixed ERRMAPEXTRACT torture to work with win2k -
Andrew Bartlett
95abb2473b Minor update to make the output 'real C'. (The output is intended to be a C
format table for inclusion back into Samba).

Andrew Bartlett
-
Tim Potter
778f5f77a6 Got medieval on another pointless extern. Removed extern struct ipzero
and replaced with two functions:

	void zero_ip(struct in_adder *ip);
	BOOL is_zero_ip(struct in_addr ip);
-
Andrew Bartlett
adc7a6048c Add a new torture test to extract a NT->DOS error map from an NT member of a
samba domain.

The PDC must be running a special authenticaion module that spits out NT errors
based on username.

Andrew Bartlett
-
Andrew Tridgell
e4fb073af0 fix maxfd test to start deleting at right file -
Martin Pool
79ec88f0da Store some path names in global variables initialized to configure
default, rather than in preprocessor macros.
-
Andrew Tridgell
441f9415b3 the CASETABLE torture target now generates the complete unicode
equivalence table for a server. This was inspired by the chargen win32
code from monyo.

This takes a *long* time to run against a Samba server due to the case
insensitive comparisons in the filesystem. That makes it a N^2
operation, and N is 64k. It is linear on NT.
-
Andrew Tridgell
eb668b54af added a little smbtorture test for dumping the unicode table of a
server. This is just a framework right now - I want this to eventually
replace the win32 test code from monyo

The interesting this about this test is that it shows up a really
horrible performance bug in our stat cache code. I'll see if I can fix
it.
-
Andrew Tridgell
2d507ec669 fixed ctemp in server and client. It turns out that ctemp on NT is completely broken, and it's pointless to emulate their brokenness completely in this case, but at least this makes us use approximately the same packet format. The spec is complelet wrong in this case -
Andrew Bartlett
0ff0a91a6b Fix (I hope) for a number of little compile warnings found by the IRIX
compilers on the build farm.

Andrew Bartlett
-
Andrew Tridgell
f41c3bb80f declare dbf in one spot -
Andrew Tridgell
1af8bf34f1 replaced stdio in many parts of samba with a XFILE. XFILE is a cut-down
replacemnt of stdio that doesn't suffer from the 8-bit filedescriptor
limit that we hit with nasty consequences on some systems

I would eventually prefer us to have a configure test to see if we need
to replace stdio, but for now this code needs to be tested widely so
I'm enabling it by default.
-
Andrew Tridgell
9341e5534d - fixed proto.h build on systems using a parallel make
- changed DENY1 and DENY2 tests to only report errors
-
Andrew Tridgell
a583af3c2f it turns out that XP agrees with the samba head branch over the right
error codes for locking, so fix the test rather than fixing samba
-
Andrew Tridgell
aaf538a830 fixed compilation of torture -
Andrew Tridgell
1d36250e33 converted another bunch of stuff to NTSTATUS -
Andrew Tridgell
83d9896c1e converted smbd to use NTSTATUS by default
major changes include:

- added NSTATUS type
- added automatic mapping between dos and nt error codes
- changed all ERROR() calls to ERROR_DOS() and many to ERROR_NT()
  these calls auto-translate to the client error code system
- got rid of the cached error code and the writebmpx code

We eventually will need to also:
- get rid of BOOL, so we don't lose error info
- replace all ERROR_DOS() calls with ERROR_NT() calls

but that is too much for one night
-
Andrew Tridgell
6513b94d19 more minor torture updates -
Andrew Tridgell
9c483987a5 the beginnings of a TRANS2 scanner -
Andrew Tridgell
7724243fe7 minor bug fixes to smbtorture -
Andrew Tridgell
3eba9606f7 a bunch of fixes from the sflight to seattle
in particular:
 - fixed NT status code for a bunch of ops
 - fixed handling of protocol levels in ms_fnmatch
-
Tim Potter
577f6065f6 Converted to new error API.
Now that we support NT errors from the client, appropriate values will need
to be filled in for the various invocations of check_error() within the
torture code.
-
Andrew Bartlett
dc588ee87a Update smbtorture in line with SAMBA_2_2
Andrew Bartlett
-
Andrew Tridgell
c41fc06376 strchr and strrchr are macros when compiling with optimisation in gcc, so we can't redefine them. damn. -
Andrew Tridgell
debb471267 The big character set handling changeover!
This commit gets rid of all our old codepage handling and replaces it with
iconv. All internal strings in Samba are now in "unix" charset, which may
be multi-byte. See internals.doc and my posting to samba-technical for
a more complete explanation.
-
Andrew Bartlett
aa5708de69 Not all OSs have setbuffer, so we better check for it.
Andrew Bartlett
-
Andrew Bartlett
a73aefef09 Make smbtorture return status values for use in the build_farm tests.
This should not have changed behaviour, and it has been tested.

Andrew Bartlett
-
Jeremy Allison
015a077acc Ensure we return correct error on trying to unlock a region not locked.
Jeremy.
-
Jeremy Allison
2c2fc85136 Extra debug in open.c, fix for bad debug message in reply.c
Jeremy.
-
Andrew Tridgell
dd12c47645 moved all our torture code to a separate directory -