summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* modified cli_read() and cli_write() to issue multiple outstandingAndrew Tridgell1998-10-045-58/+139
| | | | | | | | | read/write requests for large reads. up to max_mux requests may be outstanding. This gives _much_ better throughput and should allow smbsh to saturate just about any network. this is an implementation of the "fast SMB" method I described on the CIFS list a couple of months back.
* use dummy file descriptors opened on /dev/null to ensure that the smbwAndrew Tridgell1998-10-045-12/+23
| | | | file descriptor allocation order is identical to the kernels.
* more smbw cleanups.Andrew Tridgell1998-10-049-167/+288
| | | | | | - cleaned up prototyping. Unfortunately we can't auto-prototype wrapped.c because it replaces system functions. - split stat functions into smbw_stat.c
* - split smbw directory code into smbw_dir.cAndrew Tridgell1998-10-043-619/+658
| | | | | - fixed a bug in the handling of multiple connections to different servers
* support a few more function types (like ino_t and off_t)Andrew Tridgell1998-10-041-1/+1
|
* use const charAndrew Tridgell1998-10-041-2/+2
|
* simplied the layout of the smbwrapper code. All those 3 line filesAndrew Tridgell1998-10-0432-1076/+556
| | | | were getting unwieldy
* we need to do a load_interfaces() at startup to support broadcastAndrew Tridgell1998-10-041-0/+2
| | | | resolution of names.
* use __XXXdir() instead of __libc_XXXdir()Andrew Tridgell1998-10-036-26/+5
|
* drat.Andrew Tridgell1998-10-038-6/+263
| | | | | | | on Linux 2.0 libc detects the lack of getdents in the kernel and used opendir() etc. so we need to implement those ... it would be needed for a port to others OSes anyway I suppose.
* fixed a authentication problem with non-encrypting serversAndrew Tridgell1998-10-031-2/+3
|
* fixed initialisation bug in rpcclient (stdout is not a constant)Andrew Tridgell1998-10-031-1/+3
|
* don't call functions that aren't there yet. (Luke had the codeAndrew Tridgell1998-10-031-0/+2
| | | | commented out in the makefile)
* added fchdir() supportAndrew Tridgell1998-10-035-5/+67
|
* fixed vi on smbwrappper (it was a problem in cli_read())Andrew Tridgell1998-10-034-8/+6
|
* fixed wrapper for access(). This gets xedit working.Andrew Tridgell1998-10-031-1/+1
|
* support getcwd() in smbwrapperAndrew Tridgell1998-10-034-2/+66
|
* added mkdir() and rmdir() supportAndrew Tridgell1998-10-035-0/+152
|
* added simple device/inode number support based on a checksum of theAndrew Tridgell1998-10-034-5/+21
| | | | filename
* added a bit more to the docsAndrew Tridgell1998-10-033-14/+25
|
* added lseek() support for directoriesAndrew Tridgell1998-10-035-6/+85
|
* fixed a bug in name_len() (thanks to kooros@kooros.netrack.net)Andrew Tridgell1998-10-031-7/+7
|
* added lseek() to smbwrapperAndrew Tridgell1998-10-032-7/+54
|
* use O_ACCMODEAndrew Tridgell1998-10-031-1/+1
|
* - always open for reading (otherwise getattrE won't work).Andrew Tridgell1998-10-031-4/+11
| | | | - added somemore NT error codes
* updated prototypesAndrew Tridgell1998-10-031-2/+2
|
* define O_ACCMODE if not definedAndrew Tridgell1998-10-031-0/+4
|
* added lseekAndrew Tridgell1998-10-031-1/+1
|
* added basic chmod(), chown() and utime() support (not fullyAndrew Tridgell1998-10-037-4/+257
| | | | | | | implemented). this is enough to be able to edit files using emacs on the smbwrapper filesystem
* added unlink() and rename() support to smbwrapperAndrew Tridgell1998-10-0313-67/+242
|
* ignore *.p filesAndrew Tridgell1998-10-031-0/+1
|
* lots of improvements to smbwrapper. It now works with Samba, Win95 andAndrew Tridgell1998-10-038-66/+271
| | | | WinNT servers, and correctly lists share names in /smb/SERVER/
* - ignore *.p filesAndrew Tridgell1998-10-031-0/+1
|
* - ignore *.p filesAndrew Tridgell1998-10-032-87/+222
| | | | | | | | | | | | - enable cli_RNetSharEnum - fix password handling in sesssetup for NT1 protocol - handle partial reads and writes in cli_{read,write} - added cli_getattrE - modify cli_qpathinfo() to swap byte order for win95 servers - handle temporary errors from FINDFIRST/FINDNEXT from win95 servers, when we get a error we sleep for a bit and retry - return approx unix errno from cli_error(). Need to add a lot more cases to this.
* - ignore *.p filesAndrew Tridgell1998-10-032-1/+2
| | | | - make msleep() non-static
* updated prototypesAndrew Tridgell1998-10-031-3/+7
|
* - remove .p files in make cleanAndrew Tridgell1998-10-032-3/+4
| | | | - added readlink support to smbwrapper
* added capabilities ab win95 fields to client structure. Used forAndrew Tridgell1998-10-031-0/+2
| | | | detecting win95 server for bug handling
* Makefile.in :Luke Leighton1998-10-029-582/+780
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - added srvsvc client files clientgen.c : - replaced cli_error(cli, int *cls, int *err) with cli_error(cli, uint8 cls, uint32 *err). this version detects 32 bit status messages. the DOS error "MORE_DATA", the equivalent of the 32 bit *warning* 0x8000 0005 (STATUS_BUFFER_OVERFLOW), was being processed as an error, terminating the cli_receive_trans() call. cli_pipe.c : - replaced calls that had been incorrectly modified from 32 bit warnings (0x8000 0005 - STATUS_BUFFER_OVERFLOW) to 8 bit DOS errors (0x01 0xEA - MORE_DATA). the use of the old version of cli_error (DOS only) instead of the new one (DOS and 32 bit) caused the dce/rpc client code to fail. - replaced 2 space indentation with tab indentation in all functions. cli_srvsvc.c : cmd_srvsvc.c : - added these files back in, fixing them up to use jeremy's modified versions of the dce/rpc client functions. parse_srv.c : - added back in some "unused" functions required by dce/rpc client-side code. it would be helpful if all such "unused" functions could be added back in. rpcclient.c : - added "session", "file", "share", "connection" enumeration functions back in. these are equivalent to nt's "NetXXXXXEnum" Win32 (MSDN) functions. - added "srvinfo" function back in. this is equivalent to nt's NetServerGetInfo Win32 (MSDN) function.
* - static function "create_new_hashes" was identical to "nt_lm_owf_gen".Luke Leighton1998-10-024-36/+41
| | | | | | | | | | create_new_hashes didn't zero the buffer for the md4hash: nt_lm_owf_gen did, because jeremy sorted this out a couple of days ago. call nt_lm_owf_gen instead. - call SMBOWFencrypt from SMBencrypt and SMBNTencrypt. - added #ifdef DEBUG_PASSWORD debug password calls.
* microsoft is slowly fixing some buffer overflow errors in dce/rpc code (SP4).Luke Leighton1998-10-021-2/+2
| | | | | | | | therefore, they are being more strict, first in the server-side code, and now in the client-side code. this fixes a bind-request that was too short by 16 bytes, and an rpc-request that was too long by 24 bytes.
* some changes in smbtorture as a result of clientgen interface changesAndrew Tridgell1998-10-021-5/+5
| | | | | | | for smbwrapper note that Lukes client changes broke smbtorture. This doesn't fix it, but at least I didn't break it more than it was.
* the guts of the smbwrapper code. I may change the layout of this atAndrew Tridgell1998-10-0219-0/+1700
| | | | some stage.
* several clientgen mods to support smbwrapper. In particular addedAndrew Tridgell1998-10-021-3/+272
| | | | | cli_list() for directory listing and expended some other functions a bit.
* added sys_getwd()Andrew Tridgell1998-10-021-3/+13
|
* use sys_stat() not file_exist() for codepages. (we don't want dosAndrew Tridgell1998-10-021-2/+2
| | | | translation!)
* a couple of mode for smbwrapperAndrew Tridgell1998-10-022-3/+7
|
* makefile support for smbwrapperAndrew Tridgell1998-10-021-4/+24
| | | | dosn't compile by default. You need to do "make smbwrapper/smbwrapper.so"
* got rid of USE_FILES_ARRAY code (it was unused)Andrew Tridgell1998-10-011-106/+4
|
* Fixed *nasty* bug in nt_lm_owf_gen() - this function wasJeremy Allison1998-10-011-1/+3
| | | | | | not ensuring a zero filled buffer before doing crypto stuff. See PR#10121. Beware of this ! Jeremy.
generated by cgit (git 2.34.1) at 2025-10-01 03:50:09 +0000