summaryrefslogtreecommitdiff
path: root/source4/ntvfs/posix
AgeCommit message (Collapse)AuthorFilesLines
2007-10-10r3064: - use UINT8_MAX and UINT16_MAX instead of hex values for ↵Andrew Tridgell2-3/+3
idr_get_new() limits - change idr_get_new() to use > instead of >= in the limit check (This used to be commit 834b09929bcb8aabdd151b7c2306001497cabdb4)
2007-10-10r3057: - moved the idtree.c code into lib/Andrew Tridgell1-1/+0
- converted the tid handling to use a idtree instead of bitmaps (This used to be commit 4220914179d10132057216650b65ed7f7679717e)
2007-10-10r3056: added a id -> pointer data structure (a type of radix tree). This isAndrew Tridgell6-130/+134
an extremely efficient way of mapping from an integer handle (such as an open file handle) to a pointer (such as the structure containing the open file information). The code is taken from lib/idr.c in the 2.6 Linux kernel, and is very fast and space efficient. By using talloc it even has auto cleanup. This commit converts the handling of open file handles and open directory search handles to use the idtree routines. In combination with talloc destructors, this simplifies the structure handling in the pvfs backend a lot. For example, we no longer need to keep a linked list of open directory searches at all, and we no longer need to do linear scans of the list of open files on most operations. The end result is that the pvfs code is now extremely scalable. You can have 10s of thousands of open files and open searches and the code still runs very fast. I have also added a small optimisation into the file close path, to avoid looking in the byte range locking database if we know that there are no locks outstanding. (This used to be commit 16835a0ef91a16fa01145b773aad8d43da215dbf)
2007-10-10r3039: This solves the problem of async handlers in ntvfs backends not beingAndrew Tridgell2-3/+34
in the right state when called. For example, when we use the unixuid handler in the chain of handlers, and a backend decides to continue a call asynchronously then we need to ensure that the continuation happens with the right security context. The solution is to add a new ntvfs operation ntvfs_async_setup(), which calls all the way down through the layers, setting up anything that is required, and takes a private pointer. The backend wanting to make a async calls can use ntvfs_async_setup() to ensure that the modules above it are called when doing async processing. (This used to be commit a256e71029727fa1659ade6257085df537308c7d)
2007-10-10r3035: if the ntvfs layers prior to us have said that we can't perform anAndrew Tridgell1-1/+2
operation asynchronously (such as the nbench module), then ignore lock timeouts, as they would make no sense (This used to be commit 2894dd0ac0ddd0ae5b4d536d5cff0690bbfab1a0)
2007-10-10r3034: - fixed a bug in message dispatch, when the dispatch function called ↵Andrew Tridgell3-14/+38
messaging_deregister() - added a pvfs_lock_close_pending() hook to remove pending locks on file close - fixed the private ptr argument to messaging_deregister() in pvfs_wait - fixed a bug in continuing lock requests after a lock that is blocking a pending lock is removed - removed bogus brl_unlock() call in lock continue - corrected error code for LOCKING_ANDX_CHANGE_LOCKTYPE - expanded the lock cancel test suite to test lock cancel by unlock and by close - added a testsuite for LOCKING_ANDX_CHANGE_LOCKTYPE (This used to be commit 5ef80f034d4aa4dd6810532c63ad041bfc019cb8)
2007-10-10r3031: added support for lock cancelation, which effectively just triggers ↵Andrew Tridgell3-6/+58
an early lock timeout added support for more of the bizarre special lock offset semantics of w2k3 (This used to be commit d5bfc910b1200fb283e26572dc57fcf93652fd32)
2007-10-10r3029: implemented byte range lock timeouts.Andrew Tridgell5-7/+317
This adds a pvfs_wait_message() routine which uses the new messaging system, event timers and talloc destructors to give a nice generic async event handling system with a easy to use interface. The extensions to pvfs_lock.c are based on calls to pvfs_wait_message() routines. We now pass all of our smbtorture locking tests, although while writing this code I have thought of some additonal tests that should be added, particularly for lock cancel operations. I'll work on that soon. This commit also extends the smbtorture lock tests to test the rather weird 0xEEFFFFFF locking semantics that I have discovered in win2003. Win2003 treats the 0xEEFFFFFF boundary as special, and will give different error codes on either side of it. Locks on both sides are allowed, the only difference is which error code is given when a lock is denied. Anyone like to hazard a guess as to why? It has me stumped. (This used to be commit 4395c0557ab175d6a8dd99df03c266325949ffa5)
2007-10-10r3012: added initial support for byte range locking in the posix vfs. This isAndrew Tridgell7-22/+227
enough for us to pass locktest, but does not yet support lock timeouts and some of the other esoteric features. (This used to be commit 58a92abd88f190bc60894a68e0528e95ae33fe39)
2007-10-10r2984: fixed the error code for a non-terminal component of a path name not ↵Andrew Tridgell1-2/+1
existing (This used to be commit c6a922dbd844c69bbba83c259169655d17e9f87f)
2007-10-10r2941: added pvfs_flush() implementation to the posix backendAndrew Tridgell4-10/+67
(This used to be commit dac00ef3b0d0f8467367d7b6ff77db7b63a042cc)
2007-10-10r2931: use next_codepoint() to ensure we properly handle multi-byte ↵Andrew Tridgell1-2/+6
characters in pvfs_unix_path() (This used to be commit 0acf95d84c51d564f7215d34364c37ca74a2a150)
2007-10-10r2928: - fixed the handling of reserved names (rejecting them with ↵Andrew Tridgell3-38/+23
ACCESS_DENIED) - don't check for '.' specially in checking for legal names. Longhorn doesn't do this any more, and its a real pain. Longhorn allows for filenames ending in '.', and with as many '.' elements as you like. (This used to be commit 0a475175c53016bfa5b8246819676ddcd8b66feb)
2007-10-10r2927: imported the hash2 name mangling code from Samba3 into Samba4, butAndrew Tridgell5-39/+693
heavily modified to suit the Samba4 architecture. Samba4 with posix backend now passes the BASE-MANGLE test (This used to be commit ed52d69e8a065b6a8df2fb73c89be67acfdbca65)
2007-10-10r2926: name->dos.attrib is not valid unless name->exists is trueAndrew Tridgell1-2/+4
(This used to be commit f85a0b725d3bf67531e6874d4e8558fb050bb66f)
2007-10-10r2916: longhorn client doesn't bother setting the directory bit in ntcreatexAndrew Tridgell1-1/+6
when opening directories. This change allows samba4 server to work with longhorn client. (This used to be commit 4a8b0302eff37ccfcb66b91259ee5da20879c652)
2007-10-10r2904: - fixed the old style SMBsearch to return the pvfs shortname, not a ↵Andrew Tridgell2-3/+6
truncated long name. - short name can be up to 12 bytes, not 11 (This used to be commit 657103ec6264bf4c2986fedd3fc6577746395d49)
2007-10-10r2893: added very primitive name mangling support to pvfsAndrew Tridgell2-4/+58
(This used to be commit 749fa73544201f521d9cd3fa972b89cae99bc2c0)
2007-10-10r2800: removed the warning about using the posix ntvfs handler, as it is nowAndrew Tridgell1-2/+0
considerably more complete than the simple handler (This used to be commit e6ecd31be8e91f552d891c41b2f8737519a490a1)
2007-10-10r2794: a very simple version of the unixuid NTVFS pass-thru module. InAndrew Tridgell1-55/+0
conjunction with the posix backend this gives us a way to correctly setup the unix security context in Samba4. I chose the following method to determine the unix uid's and gid's to use given the list of SIDs from the login process - look for a "UnixID" field in the sam record. If present, then use it (check if the record is of the right type as well) - if UnixID is not present, then look for the "UnixName" sam field. If it is present then use getpwnam() or getgrnam() to find the unix id. - if UnixID and UnixName are not present, then look for a unix account of the right type called by the same name as the sAMAccountName field. - if none of the above work then fail the operation with NT_STATUS_ACCESS_DENIED obviously these steps only work well with a local SAM. It will need to be more sophisticated in future. I did not put any cache in place at all. That will need to be added for decent performance. (This used to be commit 78b67d19b9766131f0270e451089ee5bb1aa8bd9)
2007-10-10r2789: fix compileSimo Sorce1-1/+1
(This used to be commit a34ba8dafebb65e59616a0e68cc3d6fe3a349801)
2007-10-10r2788: prevent a memory leak in the pvfs search backendAndrew Tridgell2-4/+6
(This used to be commit 1de22070610231e60d329f56997bbec2cc674a4e)
2007-10-10r2786: - match on both long and short name for search posix backendAndrew Tridgell3-4/+29
- a final name component of . is illegal (This used to be commit 11c852170b83e5adbdb58407e1c7d3aeb4ab5bb8)
2007-10-10r2751: this is a new ntvfs design which tries to solve:Stefan Metzmacher12-69/+95
- the stacking of modules - finding the modules private data - hide the ntvfs details from the calling layer - I set NTVFS_INTERFACE_VERSION 0 till we are closer to release (because we need to solve some async problems with the module stacking) metze (This used to be commit 3ff03b5cb21bb79afdd3b1609be9635f6688a539)
2007-10-10r2679: fixed an uninitialised variable found with valgrindAndrew Tridgell1-1/+2
(This used to be commit 9087fab0adcf1791caeb795509ca9f14f5f47e82)
2007-10-10r2671: we're getting too many errors caused by the talloc_realloc() API notAndrew Tridgell1-6/+1
taking a context (so when you pass a NULL pointer you end up with memory in a top level context). Fixed it by changing the API to take a context. The context is only used if the pointer you are reallocing is NULL. (This used to be commit 8dc23821c9f54b2f13049b5e608a0cafb81aa540)
2007-10-10r2660: - converted the libcli/raw/ library to use talloc_increase_ref_count()Andrew Tridgell7-16/+66
rather than manual reference counts - properly support SMBexit in the cifs and posix backends - added a logoff method to all backends With these changes the RAW-CONTEXT test now passes against the posix backend (This used to be commit c315d6ac1cc40546fde1474702a6d66d07ee13c8)
2007-10-10r2656: moved the seteuid configure tests into the posix backend (these testsAndrew Tridgell1-0/+57
don't actually work yet, that will come later) (This used to be commit 46b790c19da25ba88d29f555f828688bb05e531d)
2007-10-10r2654: fixed some more server memory leaks. We are now down to a single leakAndrew Tridgell1-1/+1
of 16 bytes, caused by the 16 byte data_blob in the smb_signing code. (This used to be commit 2f1b788e09686e065d22f621f5c0c585192c6740)
2007-10-10r2613: use a talloc destructor to ensure that file descriptors are not leakedAndrew Tridgell1-2/+26
on abnormal termination of a connection. As long as the top level connection structure is freed then that should cascade down to the file structure, and call this destructor which will close the open file descriptor. In general I'd like to use this technique in any place in Samba4 where we hold operating system resources that we need to make sure are released on abnormal termination. (This used to be commit ed87b7fcbd9fedc155528ce6dd8ab5d5fce637b2)
2007-10-10r2573: - added a configure test for nanosecond time resolution in struct statAndrew Tridgell3-0/+52
(recently Linux systems support this, allowing us to support the full resolution in NTTIME) - use nanosecond resolution in the posix backend if available - moved the configure tests and list of object files for the posix backend into ntvfs/posix/ to keep them more neatlly separated. (This used to be commit d92ad9f307fe16a3b253a0555b437f14c94b4dd7)
2007-10-10r2561: completely redid the ntvfs module chaining code, You can now do ↵Andrew Tridgell13-26/+42
something like: ntvfs handler = nbench posix and the nbench pass-thru module will be called before the posix module. The chaining logic is now much saner, and less racy, with each level in the chain getting its own private pointer rather than relying on save/restore logic in the pass-thru module. The only pass-thru module we have at the moment is the nbench one (which records all traffic in a nbench compatibe format), but I plan on soon writing a "unixuid" pass-thru module that will implement the setegid()/setgroups()/seteuid() logic for standard posix uid handling. This separation of the posix backend from the uid handling should simplify the code, and make development easier. I also modified the nbench module so it can do multiple chaining, so if you want to you can do: ntvfs module = nbench nbench posix and it will save 2 copies of the log file in /tmp. This is really only useful for testing at the moment until we have more than one pass-thru module. (This used to be commit f84c0af35cb54c8fdc4933afefc18fa4c062aae4)
2007-10-10r2524: a simple pvfs rename implementation to make testing easierAndrew Tridgell2-10/+69
(This used to be commit 98c1c75076fdb0df3a7c616f8c2e1ed138a6ff9a)
2007-10-10r2520: - finished implementing the server side of the old style search requestsAndrew Tridgell1-81/+200
(This used to be commit 4e4859c06b9de5fe60ebd17cfb09eed480b79ec1)
2007-10-10r2503: the RAW-SEARCH test now mostly passes against the posix backendAndrew Tridgell4-103/+264
(This used to be commit 9710f24b1fd103d5656c9585cdfed96449cf9f97)
2007-10-10r2469: complete overhaul of the old-style RAW_SEARCH_ calls (the OS/2 andAndrew Tridgell2-21/+91
original core level calls). The old code was completely wrong in many respects. also fixed the EA_SIZE level in the server extended the RAW-SEARCH test suite to test the new code properly (This used to be commit 71480271ad84b57fcdde264a54bb2408cf783255)
2007-10-10r2438: compile on systems without O_DIRECTORY (probably won't work, but I'll ↵Andrew Tridgell1-0/+5
get to that later) (This used to be commit 5e2027f8ec7248e48d1dfb94c3688c78c64a85c5)
2007-10-10r2436: the second big lump of posix vfs code.Andrew Tridgell14-186/+817
this is still just a skeleton, and many of the functions are just based on the simple vfs backend, they are there to allow me to run smbtorture tests against the real parts of the posix backend. (This used to be commit f2fa7fe565e89360dba3bb5434d3a6a36f398348)
2007-10-10r2404: the first large lump of posix vfs stuff.Andrew Tridgell9-37/+1137
this is still very much a skeleton (with many limbs missing too!). I am committing this early to get some feedback on the approach taken. (This used to be commit 40d5cae5ebbfe328e193eadb685df6a370730299)
2007-10-10r2252: don't register the same name twiceAndrew Tridgell1-2/+2
(This used to be commit fdb675bbad1322ddd94c646f67803b9678468a64)
2007-10-10r2251: forgot to add vfs_posix.h in my last commitAndrew Tridgell2-2/+261
note that this is just a skeleton so far. More to come soon. (This used to be commit efc8850b9aa9348f5f7c4b342aa76dab1635e7d4)
2007-10-10r2249: got rid of some more mem_ctx elements in structuresAndrew Tridgell1-12/+30
(This used to be commit 21ef338cbbe96acc8594ffc550ef60c6a40fb951)
2007-10-10r1916: return a more clear error, we are miss configure in this caseStefan Metzmacher1-1/+1
metze (This used to be commit a908f831cb9dd265f9b183512053f9e608feec3d)
2007-10-10r1286: rename struct tcon_context to smbsrv_tconStefan Metzmacher1-1/+1
metze (This used to be commit a6c0ca9de52b2395b092cb245bb94cbd55dfdd46)
2007-10-10r1280: rename struct request_context to smbsrv_requestStefan Metzmacher1-1/+1
metze (This used to be commit a85d2db5826a84b812ea5162a11f54edd25f74e3)
2007-10-10r312: let ntvfs posix backend return NT_STATUS_ACCESS_DENIED in the connect hookStefan Metzmacher1-85/+14
and print out an error message to the debug log which say: use 'cifs' or 'simple' as ntvfs handler this also warns about 'root' fileaccess in the 'simple' module the 'default' ntvfs handler is now registered by the posix backend metze (This used to be commit 84b3589daa60cfdd2c868d9468192b0a6e1eebae)
2003-11-25CVS: ----------------------------------------------------------------------Jelmer Vernooij1-7/+9
CVS: Enter Log. Lines beginning with `CVS:' are removed automatically CVS: CVS: Committing in . CVS: CVS: Modified Files: CVS: Makefile.in configure.in include/includes.h include/ntvfs.h CVS: include/smb.h lib/iconv.c lib/module.c ntvfs/ntvfs_base.c CVS: ntvfs/cifs/vfs_cifs.c ntvfs/ipc/vfs_ipc.c CVS: ntvfs/posix/vfs_posix.c ntvfs/print/vfs_print.c CVS: ntvfs/reference/vfs_ref.c ntvfs/simple/vfs_simple.c CVS: passdb/pdb_interface.c CVS: Added Files: CVS: include/module.h CVS: ---------------------------------------------------------------------- Update to the modules system. Fixed: - get rid of smb_probe_module - merge older updates from 3.0 - introduced register_subsystem() and register_backend() functions - adapt ntvfs and charset to use new register functions - made smb_load_modules() work recursively (e.g. 'preload modules = /usr/lib/samba') - got rid of some old remains Things that still need work: - Did I break tankFS? I don't think so, but I can't test it here :-( - Add 'postload modules = ' (for modules that need to be loaded after fork() in smbd, if applicable) - Convert RPC, auth, passdb, etc to use new register_{subsystem,backend}() functions - Accept wildcards in 'preload modules' option, instead of loading recursively (This used to be commit 7512b9ab1a8b3103f7a6c13f736353c46a26b668)
2003-08-13first public release of samba4 codeAndrew Tridgell1-0/+151
(This used to be commit b0510b5428b3461aeb9bbe3cc95f62fc73e2b97f)
generated by cgit (git 2.34.1) at 2024-12-22 12:24:28 +0000