summaryrefslogtreecommitdiff
path: root/source3/smbd
AgeCommit message (Collapse)AuthorFilesLines
2012-09-13Sigh :-(. Removing optimization prematurely is the root of all evil :-(.Jeremy Allison3-0/+15
Sorry for the mistake, but the LastDir singleton cache in vfs_ChDir() actually plays an important role. When we're processing a stream of SMB1/SMB2/SMB3 requests we don't want to add a chdir()/getcwd() system call pair on every request if they're all on the same connection and dealing with the same base path. I did some testing with a program that times 1,000,000 chdir() requests vs. 1,000,000 strcmp requests and it's a penalty of 10x doing the system calls. Just because it's old code, doesn't mean it's bad :-(. Autobuild-User(master): Jeremy Allison <jra@samba.org> Autobuild-Date(master): Thu Sep 13 21:31:42 CEST 2012 on sn-devel-104
2012-09-12Move cached cwd onto conn struct.Jeremy Allison2-1/+15
This enables us to make VFS modules safe for use in root called code when we've changed directory under conn->connectpath.
2012-09-12Remove ancient "optimization" global LastDir.Jeremy Allison3-18/+3
2012-09-12s3: remove unneeded noquota.c fileBjörn Jacke1-38/+0
Autobuild-User(master): Björn Jacke <bj@sernet.de> Autobuild-Date(master): Wed Sep 12 16:50:50 CEST 2012 on sn-devel-104
2012-09-12s3-rap: Open printers with the right access mask.Andreas Schneider1-3/+3
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org> Autobuild-Date(master): Wed Sep 12 14:05:42 CEST 2012 on sn-devel-104
2012-09-12smbd: Add extra VFS hooks to get the posix ACL as a blobAndrew Bartlett1-0/+21
This will allow us to hash this, rather than the NT ACL it maps to. This will in turn allow us to know if the NT ACL is valid even if we have to change the mapping code. Andrew Bartlett Autobuild-User(master): Andrew Bartlett <abartlet@samba.org> Autobuild-Date(master): Wed Sep 12 07:06:01 CEST 2012 on sn-devel-104
2012-09-12smbd: Remove pre-allocation of ACL array in sys_acl_init()Andrew Bartlett2-4/+4
Instead, this is just handled with realloc in sys_acl_create_entry() This allows us to remove the size element from the SMB_ACL_T. Andrew Bartlett
2012-09-12Make metze happy and the code clearer :-).Jeremy Allison1-0/+6
Ensure we know after the destructor fires we're never going to look at this again. Autobuild-User(master): Jeremy Allison <jra@samba.org> Autobuild-Date(master): Wed Sep 12 03:00:21 CEST 2012 on sn-devel-104
2012-09-11quota: move function again to its belonging ifdef blockBjörn Jacke1-11/+9
Autobuild-User(master): Björn Jacke <bj@sernet.de> Autobuild-Date(master): Tue Sep 11 04:44:28 CEST 2012 on sn-devel-104
2012-09-11s3:proto.h: remove duplicate disk_quotas prototypesBjörn Jacke1-10/+0
2012-09-10s3: delete requests are not specialVolker Lendecke1-16/+0
The only difference between batch and exclusive oplocks is the time of the check: Batch is checked before the share mode check, exclusive after. Signed-off-by: Jeremy Allison <jra@samba.org>
2012-09-10Fix talloc memory heirarchy bug. If there's an SMB2 sessionsetup in flight ↵Jeremy Allison1-5/+36
when we're shut down, we end up freeing the struct smbXsrv_session *session pointer twice. Autobuild-User(master): Jeremy Allison <jra@samba.org> Autobuild-Date(master): Mon Sep 10 23:34:06 CEST 2012 on sn-devel-104
2012-09-10uid_t is a synonym for int on osXMatthieu Patou1-1/+1
This function is clearly osX oriented so we adapt the modifier to the target
2012-09-08s3:vfs_default: add basic support for durable handle request and reconnectMichael Adam2-0/+586
We only grant durable handles for CIFS/SMB2 only access, that means "kernel oplocks", "kernel share modes" and "posix locking" need to be set to "no". For now we also don't grant durable handles if delete on close is active on the handle. Pair-Programmed-With: Stefan Metzmacher <metze@samba.org> Pair-Programmed-With: Volker Lendecke <vl@samba.org>
2012-09-08s3:smbd: make fd_open() publicMichael Adam2-4/+6
2012-09-08s3:smbd: don't set kernel flock if "kernel share modes = no"Michael Adam1-1/+1
Signed-off-by: Stefan Metzmacher <metze@samba.org> Autobuild-User(master): Stefan Metzmacher <metze@samba.org> Autobuild-Date(master): Sat Sep 8 13:05:05 CEST 2012 on sn-devel-104
2012-09-08s3:smbd: add basic support for durable handle v2 request and reconnectMichael Adam1-2/+129
This does not yet cover persistent handle support which is also negotiated through these create request blobs. Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
2012-09-08s3:smbd: add basic support for durable handle request and reconnectMichael Adam1-46/+203
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org> Pair-Programmed-With: Volker Lendecke <vl@samba.org>
2012-09-08s3:smbd: initial durable handle support: special treatment of durable ↵Michael Adam1-0/+44
handles in close Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
2012-09-08s3:vfs: add durable VFS operationsMichael Adam1-0/+34
This allows a VFS module to implement durable handles in different ways. Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
2012-09-08s3:smbd: add disconnected checks to the open code.Stefan Metzmacher1-0/+31
(delay_for_batch_oplocks, open_mode_check, and delay_for_exclusive_oplocks) Pair-Programmed-With: Michael Adam <obnox@samba.org>
2012-09-08s3:smbd: also close durable file handles in a tdisMichael Adam1-2/+9
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
2012-09-08s3:smbXsrv_open: add smb2srv_open_recreate() to support durable handlesStefan Metzmacher2-1/+210
metze
2012-09-08s3: Fix a panic when shutting downVolker Lendecke1-10/+32
When a client disconnects while we have aio open, there is no close request that cleans up. We can't send out the replies anymore, so just drop the aio requests that are pending. Found using the new python lib writing multiple files simultaneously TODO: check tdis and logoff Signed-off-by: Stefan Metzmacher <metze@samba.org> Autobuild-User(master): Stefan Metzmacher <metze@samba.org> Autobuild-Date(master): Sat Sep 8 01:27:34 CEST 2012 on sn-devel-104
2012-09-07s3-smbd: Remove OSF1 support from old (pre-sysquotas) codeBjörn Jacke1-65/+0
"recent" Tru64 releases had QUOTACTL_4B support
2012-09-06s3:quota: don't force the block size to 512Björn Jacke1-3/+0
there is no point in forcing the block size to 512 when curblocks is 1. This will only lead to false quota reporting. See bug #3272
2012-09-06s3-smbd: Remove remaining references to removed OS support in old-style ↵Andrew Bartlett1-6/+6
quota code Signed-off-by: Björn Jacke <bj@sernet.de> Autobuild-User(master): Björn Jacke <bj@sernet.de> Autobuild-Date(master): Thu Sep 6 10:47:18 CEST 2012 on sn-devel-104
2012-09-06s3-smbd: Remove IRIX support from old (pre-sysquotas) codeAndrew Bartlett1-129/+0
This should be essentially dead code, as while it is possible to disable sysquotas, the sysquotas has since 2003 supported the 4A varient used on IRIX. Andrew Bartlett Signed-off-by: Björn Jacke <bj@sernet.de>
2012-09-06s3-smbd: Remove HPUX support from old (pre-sysquotas) codeAndrew Bartlett1-19/+0
This should be essentially dead code, as while it is possible to disable sysquotas, the sysquotas has since 2003 supported the 4A varient used on HPUX. Andrew Bartlett Signed-off-by: Björn Jacke <bj@sernet.de>
2012-09-06s3-smbd: Remove BSD support from old (pre-sysquotas) codeAndrew Bartlett1-226/+3
This is essentially dead code, as while it is possible to disable sysquotas, the default for BSD is now to use sysquotas with the addition of 4B support via the recent Darwin patch by James Peach. NFS support has been in sysquotas for quite some time now. Andrew Bartlett Signed-off-by: Björn Jacke <bj@sernet.de>
2012-09-06s3-smbd: Remove Linux support from old (pre-sysquotas) codeAndrew Bartlett1-438/+1
This is essentially dead code, as while it is possible to disable sysquotas, the default for Linux has been to use sysquotas for quite some time now. Andrew Bartlett Signed-off-by: Björn Jacke <bj@sernet.de>
2012-09-06build: Remove Unicos support (quota in particular)Andrew Bartlett1-89/+0
Unicos machines are long gone now (Cray now make Linux compute nodes), so remove the quota support. Andrew Bartlett Signed-off-by: Björn Jacke <bj@sernet.de>
2012-09-05Remove an unncessary FTRUNCATE call in the open path. We can rely on O_TRUC.Jeremy Allison1-29/+1
Suggested by Volker.
2012-09-05s3:smb2_create: check for SMB2_CREATE_TAG_DHNC firstStefan Metzmacher1-43/+14
metze Autobuild-User(master): Stefan Metzmacher <metze@samba.org> Autobuild-Date(master): Wed Sep 5 12:27:25 CEST 2012 on sn-devel-104
2012-09-05s3: Slightly simplify fd_open_atomicVolker Lendecke1-8/+7
Replace an if-statement by a direct assignment Signed-off-by: Jeremy Allison <jra@samba.org> Autobuild-User(master): Jeremy Allison <jra@samba.org> Autobuild-Date(master): Wed Sep 5 01:56:46 CEST 2012 on sn-devel-104
2012-09-04s3: Put a comment into the right placeVolker Lendecke1-4/+5
Signed-off-by: Jeremy Allison <jra@samba.org>
2012-09-04s3: Factor out disposition_to_open_flagsVolker Lendecke1-24/+57
Signed-off-by: Jeremy Allison <jra@samba.org>
2012-09-04s3: Factor out calculation of clear_adsVolker Lendecke1-5/+17
Signed-off-by: Jeremy Allison <jra@samba.org>
2012-09-04s3: Slightly simplify open_file_ntcreateVolker Lendecke1-5/+5
We have not set flags2 before, so do direct assignment and not |= Signed-off-by: Jeremy Allison <jra@samba.org>
2012-09-04s3: Make "open_match_attributes" staticVolker Lendecke2-12/+6
Signed-off-by: Jeremy Allison <jra@samba.org>
2012-09-04s3:smbd: in sys_disk_free(), improve a debug messageMichael Adam1-1/+1
2012-09-04s3:smbd: in sys_disk_free(), improve a debug messageMichael Adam1-1/+1
2012-09-04s3:smbd: in sys_disk_free(), fix line length and indentation of debug statementMichael Adam1-2/+3
2012-09-04s3:smbd: in sys_disk_free(), fix a debug messageMichael Adam1-1/+1
2012-09-02s3: remove some duplicate quota codeBjörn Jacke1-27/+12
2012-09-01s3: Make an if statement a bit easier to readVolker Lendecke1-2/+3
Fix indentation a bit Signed-off-by: Jeremy Allison <jra@samba.org> Autobuild-User(master): Jeremy Allison <jra@samba.org> Autobuild-Date(master): Sat Sep 1 07:07:12 CEST 2012 on sn-devel-104
2012-08-31Change the S3 fileserver over to se_file_access_check().Jeremy Allison1-6/+9
Don't set the priv_open_requested yet until the open-for-backup request is correctly passed in.
2012-08-30Windows does canonicalization of inheritance bits. Do the same.Jeremy Allison1-0/+35
We need to filter out the SEC_DESC_DACL_AUTO_INHERITED|SEC_DESC_DACL_AUTO_INHERIT_REQ bits. If both are set we store SEC_DESC_DACL_AUTO_INHERITED as this alters whether SEC_ACE_FLAG_INHERITED_ACE is set when an ACE is inherited. Otherwise we zero these bits out. See: http://social.msdn.microsoft.com/Forums/eu/os_fileservices/thread/11f77b68-731e-407d-b1b3-064750716531 for details.
2012-08-30Change the other two places where we set a security descriptor given by the ↵Jeremy Allison1-5/+1
client to got through set_sd(), the canonicalize sd function.
2012-08-30Re-add set_sd(), called from set_sd_blob(). Allows us to centralize all ACL ↵Jeremy Allison2-14/+28
canonicalization.