1
0
mirror of https://github.com/samba-team/samba.git synced 2025-07-22 16:59:09 +03:00
Commit Graph

2526 Commits

Author SHA1 Message Date
d3379fe61b r6172: Tidy up error processing significantly. Remove unix_ERR_XXX global nastyness.
Jeremy.
2007-10-10 10:56:27 -05:00
065ab9182d r6169: Fix bug #2563. Infinite loop on non-existant file with findnext.
Jeremy
2007-10-10 10:56:26 -05:00
75a0c11de4 r6163: Well I never, all this time the t2open code was incorrectly decting
oplock requests and no one noticed. Hmmm. Don't think this is a commonly
used codepath :-).
Jeremy.
2007-10-10 10:56:26 -05:00
fb05ac4c03 r6160: Ensure allocation size is correctly returned for OpenX. Only set allocation
on create/truncate for nttrans.
Jeremy.
2007-10-10 10:56:26 -05:00
326124a7b3 r6152: Correctly check OpenX open modes.
Jeremy.
2007-10-10 10:56:25 -05:00
994694f7f2 r6149: Fixes bugs #2498 and 2484.
1. using smbc_getxattr() et al, one may now request all access control
   entities in the ACL without getting all other NT attributes.
2. added the ability to exclude specified attributes from the result set
   provided by smbc_getxattr() et al, when requesting all attributes,
   all NT attributes, or all DOS attributes.
3. eliminated all compiler warnings, including when --enable-developer
   compiler flags are in use.  removed -Wcast-qual flag from list, as that
   is specifically to force warnings in the case of casting away qualifiers.

Note: In the process of eliminating compiler warnings, a few nasties were
      discovered.  In the file libads/sasl.c, PRIVATE kerberos interfaces
      are being used; and in libsmb/clikrb5.c, both PRIAVE and DEPRECATED
      kerberos interfaces are being used.  Someone who knows kerberos
      should look at these and determine if there is an alternate method
      of accomplishing the task.
2007-10-10 10:56:24 -05:00
989fcb4c08 r6146: Added OS/2 EA support in trans2_open and trans2_mkdir. Fixed in nttrans_create.
Jeremy.
2007-10-10 10:56:24 -05:00
106f91ff65 r6141: Fix OS/2 EA's for NTcreate. OpenX and mkdir to follow.
Jeremy.
2007-10-10 10:56:24 -05:00
7c4262d6ba r6138: Refactor length reading out of EA code so we can read length
external to buffer. Needed to add EA code to create calls.
Jeremy.
2007-10-10 10:56:24 -05:00
b42eaf424e r6124: Fix for bug #2242 from Tom Lackemann <cessnatomny@yahoo.com> - don't
set SD on an NTtransact create unless we created the file.
Jeremy.
2007-10-10 10:56:23 -05:00
14eab937a5 r6122: Added "hack" fix for broken SMB_INFO_VOLUME level - used by OS/2
I believe. Reported as fixed by Marcel Müller <mueller@maazl.de>.
Jeremy.
2007-10-10 10:56:23 -05:00
8c64cd368f r6106: Fix bug #2551. It turns out that the incoming flags2 flag FLAGS2_LONG_PATH_COMPONENTS
determines if a reply is uppercased on a SMBsearch request, not the protocol level.
This could clear up quite a few hacks going forward I think.
Jeremy.
2007-10-10 10:56:22 -05:00
057e4bf641 r6082: Fix for bugid #2541. Fixes from Guenter Kukkukk <guenter.kukkukk@kukkukk.com>
Jeremy.
2007-10-10 10:56:21 -05:00
d6ad1f8df0 r6069: Fix for bug #2541. Ensure we recognise LANMAN2.1 as OS/2 and select LANMAN2
protocol, ensure the EA size is always correctly set on a query for a file with no EA's.
Jeremy.
2007-10-10 10:56:20 -05:00
e15c36e949 r6067: Patch from Marcel Müller <mueller@maazl.de> - fix OS/2 EA_FROM_LIST
info level.
Jeremy.
2007-10-10 10:56:20 -05:00
56cc756b9d r6063: Fix for bug #2533 - ensure SMBsearch unsets UNICODE flags2 bit as this SMB is
DOS codepage only.
Jeremy.
2007-10-10 10:56:20 -05:00
e3c7d08bb6 r6060: It's not quite accurate to say not having write access causes a group
entry never to match - it matches but if doesn't grant access is recorded
so the "other" entry isn't subsequently checked.
Fix the algorithm.
Jeremy.
2007-10-10 10:56:19 -05:00
6609b209f5 r6057: Don't put the assert in the wrong place :-).
Jeremy.
2007-10-10 10:56:19 -05:00
81fb337286 r6055: Fix algorithm. If any of the primary or supplementary group ids match
a "allow" entry of GROUP or GROUP_OBJ, then access is allowed. It doesn't
terminate on the first match. Added debug to show where the match occured
(or didn't).
Jeremy.
2007-10-10 10:56:19 -05:00
e831cef618 r6053: Fixup dfs path with the new wildcard parser code split out.
Jeremy.
2007-10-10 10:56:19 -05:00
80e788143a r6049: Ensure "dos filetime" checks file ACLs correctly. May fix Excel "read-only"
issue.
Jeremy.
2007-10-10 10:56:18 -05:00
f8b67159fc r6048: Split out the check_path_syntax into a findfirst/next/wildcard version.
The semantics are different with wildcards.
Jeremy.
2007-10-10 10:56:18 -05:00
67f6473f50 r6044: Ensure the old search calls always ask mask_match to translate
patterns like ????????.??? - even if using an NT1 protocol.
Matches W2K3 behavior.
Jeremy.
2007-10-10 10:56:18 -05:00
cf8949f684 r6022: Fix for bug #2533. Incorrect dir listings from OS/2 clients.
Jeremy.
2007-10-10 10:56:16 -05:00
4868e42027 r6020: Never do NT status codes with protocols before NT1 as we don't get client caps.
Jeremy.
2007-10-10 10:56:16 -05:00
4e0ac63c36 r6014: rather large change set....
pulling back all recent rpc changes from trunk into
3.0.  I've tested a compile and so don't think I've missed
any files.  But if so, just mail me and I'll clean backup
in a couple of hours.

Changes include \winreg, \eventlog, \svcctl, and
general parse_misc.c updates.

I am planning on bracketing the event code with an
#ifdef ENABLE_EVENTLOG until I finish merging Marcin's
changes (very soon).
2007-10-10 10:56:15 -05:00
21e3cf2f8f r6001: Oops. Checing the wrong tagtype - should have been SMB_ACL_GROUP, not SMB_ACL_MASK.
Fix bug #2521.
Jeremy.
2007-10-10 10:56:15 -05:00
9e77da9320 r5961: final round of compiler warning fixes based on feedback from Jason Mader 2007-10-10 10:56:13 -05:00
e61f7a6b72 r5960: Integer wrap paranoia when reading EA buffers.
Jeremy.
2007-10-10 10:56:12 -05:00
aeb3cea54e r5958: BUG 1843: patch from james peach to fix quotas (with no soft limits) on IRIX 2007-10-10 10:56:12 -05:00
19a639ac46 r5952: BUG 2469: patch from Jason Mader to cleanup compiler warning when not using krb5 2007-10-10 10:56:11 -05:00
8aa2b2617b r5936: Fix bugs in earlier work in progress commit. Closer to properly supporting EA's now.
Jeremy.
2007-10-10 10:56:09 -05:00
22fca74657 r5933: We were handling setting of EA's incorrectly - we should be able to set
a list. Also not converting names from DOS CP to UNIX CP correctly. This
code doesn't quite work yet but it's a work in progress to be fixed
tomorrow (don't want to lose it).
Jeremy.
2007-10-10 10:56:09 -05:00
e32a58742e r5922: Fix for NASTY NASTY bug #2501. All my fault :-(. Brown paper bag time.
Stops Win98 from looping doing findnext on a singleton directory. More
testing very welcome.
Jeremy.
2007-10-10 10:56:09 -05:00
f7218d1c66 r5905: Fix two warnings found by AIX. They might actually be bugs on 64-bit
platforms.

Volker
2007-10-10 10:56:07 -05:00
8b2b652c0a r5893: Get us to pass some of the Samba4 EA tests. EA_LIST seems to be
working - need to valgrind it to be sure.
Jeremy.
2007-10-10 10:56:07 -05:00
b033a302b5 r5874: Work in progress. EA list infolevel.
Jeremy
2007-10-10 10:56:06 -05:00
0e399422f7 r5872: Restructure ea list code to start to fix the EA_LIST calls. This does not
work yet but it's a work in progress.
Jeremy.
2007-10-10 10:56:06 -05:00
a417feaa41 r5864: Stop using 'INFO' names on find calls. Getting ready to fixup OS/2 EA
support.
Jeremy.
2007-10-10 10:56:06 -05:00
c137362871 r5854: Fix coredump when OS/2 checks for long file name support (with .+,;=[].).
Bugzilla 2338.

Thanks to Guenter Kukkukk <guenter.kukkukk@kukkukk.com>
2007-10-10 10:56:05 -05:00
c0924a497b r5848: Fix inherit owner to chown directories in a porable and secure way (from
just secure on Linux). Idea from tridge.
Jeremy.
2007-10-10 10:56:05 -05:00
a4b4766e42 r5844: Fix typo. Add cast for bug #2464 reported by jason@ncac.gwu.edu.
Jeremy.
2007-10-10 10:56:05 -05:00
1f988333ec r5843: Removed unused variable - pointed out by jason@ncac.gwu.edu in bugid #2460.
Jeremy.
2007-10-10 10:56:05 -05:00
712ef86ac0 r5836: Correct fix for OS/2 bug #2335. The error return code in this case is protocol
level dependent.
Jeremy.
2007-10-10 10:56:04 -05:00
a7a7af6815 r5833: Revert change for bug #2335. Tested with a Win9x client and the original
error code is correct. I have some theories as to why this is different,
I'll add some code to the torture tester to confirm.
Jeremy.
2007-10-10 10:56:04 -05:00
2312ac9d42 r5831: Along with jra's recent checkins, fix dir/xcopy of empty dirs on OS/2. Bugs
2335, 2337.
2007-10-10 10:56:04 -05:00
19e017a78a r5825: Fix one more DISK_FREE call - spotted by Ying Li <ying.li2@hp.com>.
Jeremy.
2007-10-10 10:56:04 -05:00
b5d31b2caf r5822: Actually return an error message if disk_free fails ! Pointed out by Ying Li <ying.li2@hp.com>.
Jeremy.
2007-10-10 10:56:03 -05:00
479bb6ea95 r5816: Catch the "." resume name case also.
Jeremy.
2007-10-10 10:56:03 -05:00
c03524363f r5813: Fix bug found by Jim McDonough <jmcd@us.ibm.com>
Win9x can send a resume name of "..". This will cause the parser to
complain (it thinks we're asking for the directory above the shared
path). Catch this as the resume name is only compared, never used in
a file access and replace it with "..".
Jeremy.
2007-10-10 10:56:03 -05:00