GitLab

Retagged.

These two displays take different code paths, and *INFO goes via a check to workaround a bug in Win98 servers. Unfortunately, the check no longer works because Microsoft don't return the '.' and '..' entries in SMB Trans2 FIND_FIRST, and the logic previously always set the "need to frig the dates" flag. Now, the flag is only set if something comes back and after doing a test swap the results are also correct.
Turned on the 'auto close Trans2 FIND_FIRST' flag in the request block to save having to do a second transaction to close it.
Swapped the LDMFD/STMFD for Push and Pull in the assembler.
Tested against WinXP SP3, dates are no longer bonkers.

Version 2.35. Tagged as 'LanManFS-2_35'

Detail:
 Added an extra check after the Xlate path resolution code so that
 the case of the path given to resolve is honoured. This now means that
 the strcmp() done in SMB_Rename fails and the rename is actually sent to
 the server. Before "*rename test Test" would have the path resolved and
 since 'test' already existed both names would be the same, in this example
 something like "*rename A:\test A:\test" would result.
Admin:
 Made the non LONGNAMES version compile again (not tested).
 Removed local definitions of FSControl reason codes, uses Interface header
 now.
 Made the case of #includes match CVS to allow future compilation on a case
 sensitive system.
 Deleted unused "Version" header.

Version 2.34. Tagged as 'LanManFS-2_34'

Detail:
  When the underlying FS is NTFS, eg. WinXP Pro, the requested
  resume keys for a directory search are always zero - this is passed
  back via OSGBPB and on the next call it triggers a new search. So
  we get stuck in an infinite loop.
  Now checks for a resume key stuck at zero, and makes a fake one static
  to the search context then sets the continuation bit in the SMB_FIND_NEXT
  call because the server returning the duff key remembers the position.
  Dir_CallbackFn could return with "taken" undefined. Move a few lines
  higher.
  Missing "#else" added to CHECK_ARMBOOT_EXISTS so if this is disabled
  no further attempt is made to find !ArmBoot.
  Couple of typos corrected.
  Spelling of "disk" changed to "disc" in an error message.
Admin:
  Tested with XP Pro SP3 with NTFS, directories which previously hung
  the filer now open correctly. Checked with XP Home SP3 with FAT32 to
  verify behaviour unchanged from 2.32.
  Note, with large directories the initial path translation triggers a
  dir search too which takes ages compared with the actual number of
  entries returned. This could be cached to make counting far faster.

Version 2.33. Tagged as 'LanManFS-2_33'

Detail:
  Too many to list here: see doc/RPSChanges.
Admin:
  Received from Rob Sprowson approx 2009-02-08. Builds, but not tested by ROOL.

Version 2.32. Tagged as 'LanManFS-2_32'

Detail:
  Security fix MS09-001 (detailed in Microsoft Knowledge base article KB958687)
  fixes some vulnerabilities in the SMB protocol which LanManFS uses. This
  highlighted an error in SMB_WriteRaw which reported zero data bytes but then
  supplied an offset to the non existant bytes: this is now faulted by Windows.
  As there are no data bytes, the offset should also be zero.
Admin:
  Fix received from Rob Sprowson 2009-01-19.

Version 2.31. Tagged as 'LanManFS-2_31'

Version 2.30. Tagged as 'LanManFS-2_30'

Detail:
  > Default filetype now text
  >
  > Default type given to any file that has neither a dos extent, nor a
  > ',xxx' riscos extent at the server
  >
  > Ensure that files given the default type never have a ',xxx' riscos
  > extent appended.
  >
  > Add environment variable 'LMDefaultType' which checked for a default
  > type on module initialisation .. thus setting this to (e.g.) dos, then
  > rmreinitting LanManFS will change the default type to (e.g.) dos.
Admin:
  Not tested here yet.

Version 2.29. Tagged as 'LanManFS-2_29'

> Further mods to make sure that directory naming/renaming cannot attach
> the ,xxx filetyping.

Version 2.28. Tagged as 'LanManFS-2_28'

Update from jwb - no longer appends ,xxx to filenames on server if the filetype can already be determined from the file extension.

Version 2.27. Tagged as 'LanManFS-2_27'

Tightened up checking of when sockets get opened and closed in an attempt to work out why Freeway gets upset when LanManFS is configured in IP mode.

Changed app date that OmniClient shows to be dd-mmm-yy
Added ability to include slimmer ROMSPRITES and deleted the resources
phase from the makefile: the sprites were being included once in
ResourceFS and again embedded in the module.
If the FS is unset it says so in *status rather than a blank line.

Version 2.26. Tagged as 'LanManFS-2_26'

Directory rename fixed - the mask being passed by the rename routine didn't have the ATTR_DIR bit set.Discovered this after reading lots of the spec which is also now included here in /doc.

Copying files out of LanManFS filer "forgot" their filetype even though
they appeared right in the filer,fixed.
A stray debugging printf removed,along with one compiler warning.
New sprites (yuck) to match !Omni.
Reordered the shutdown in two places,first Omni_Shutdown bins the mounts
lists which SMB_Shutdown uses.Second in NetBEUI mode the announcement that
a protocol has terminated goes *after* the shutdown - otherwise you sit
around for about 20s wondering where the link has gone.

Version 2.25. Tagged as 'LanManFS-2_25'

On saving a file if the type was text no filetype information was
saved and instead it relied on the default for unknown filetypes being
text.
The default unknown filetype is now &FE4,plus the above logic would
have conflicted with the MimeMap code if for example a file of type
text was saved but which had an extension which was in the mimemap of
type something other than text.

Version 2.24. Tagged as 'LanManFS-2_24'

Change lmls command to ListFS in line with NFS
Supress LF when ListFS is issued but there are no servers present.
Cast a few things to shut some compiler warnings up.
Addition of MimeMap lookups where the file has no ",ttt" style type
appended.Previously any such files were all reported as text files so
viewing a machine would typically show whole directories full of
identical filetypes.

Version 2.23. Tagged as 'LanManFS-2_23'

Detail:
  Now happens every 4500 cs rather than every 4501. Big deal.
Admin:
  Untested - will work.
  I'm bored.

Version 2.22. Tagged as 'LanManFS-2_22'

Admin:
  Tested.


Version 2.21. Tagged as 'LanManFS-2_21'

Detail:
  Due to re-ordering of object files, the CMHG header wasn't being
    generated by the time it was first required for inclusion.  Static
    dependencies have been added to fix this.  This is going to affect
    a lot of components which rely on amu's odd order to processing
    object files (back to front compared to other makes).
Admin:
  Fixes Bugzilla bug #4064
  Tested manually.

Version 2.20. Tagged as 'LanManFS-2_20'

  Removed BBE specific rules due to above change obsoleting it.
Admin:
  Built.
  Requires Library 0.70 or later (RiscOS/Library; Library-0_70)

Version 2.19. Tagged as 'LanManFS-2_19'

Detail:
  This module was ignoring the possibility of the Internet module being
    re-initialised underneath it - something which can happen if there
    is an Internet module in flash as well as main ROM.  This is an
    experimental fix for this problem.
Admin:
  Not tested.

Version 2.18. Tagged as 'LanManFS-2_18'

Detail:
  The wrong bit was being checked to see if we were using the extended
    security model (CAP_EXTENDED_SECURITY).  Thus the wrong format was
    being used in SMBsesssetup, which meant that we couldn't connect
    to Windows 2000.
  Much better debugging added to the name resolution code - this helped
    diagnose the faulty photocopier problem and is left in case anything
    like that happens again!
Admin:
  Tested - I can now connect to Win 2000 shares.

Version 2.17. Tagged as 'LanManFS-2_17'

Detail:
  Private bbe rule to deal with Ant putting files in the wrong place :-(
  Resource files should be in Resources/... and not in the component root.
Admin:
  Tested on Fun8JP Source & BBE builds.

Version 2.16. Tagged as 'LanManFS-2_16'

Detail:
  Some symbol names were wrong which only showed up in ROM builds.
Admin:
  Tested in Ursula ROM build, and tested RAM build.

Version 2.15. Tagged as 'LanManFS-2_15'

Detail:
  Pass OPTIONS=-DNO_NETBEUI from the Components file if you do not need
    NetBEUI support, and LanManFS will be 12K smaller.
  In NetBEUI mode, LanManFS sends Service_DCIProtocolStatus as required.
  In IP mode, LanManFS notices Service_DCIProtocolStatus from the Internet
    module when it dies and marks all its sockets as needing reconnection.
  Booting behaviour changed to look for !ArmBoot again and not attempt
    the boot if it is missing.  Removed lots of grotty backdoor hacks
    that used to be used to determine this.
Admin:
  Not tested at all.  This will need very thorough testing before it can
    be considered for release.

Version 2.14. Tagged as 'LanManFS-2_14'

Detail:
  NB_Startup was failing to stop the LLC timers if the function
    was going to fail after hooking onto TickerV.  This meant that
    if LanManFS never found an appropriate driver for NetBIOS use,
    then killing the module resulted in repeated aborts when the
    TickerV handler has invoked.
Admin:
  Untested.
  Fixes Bugzilla bug #26

Version 2.13. Tagged as 'LanManFS-2_13'

Detail:
  SVC stack overflow occurred due to the recursive method used to
    discard the idle-out prevention responses.  This no longer
    happens.
  Fixed a bit of debugging which caused data aborts!
  Optimised a select() call to pass s + 1 as the first parameter to
    save Internet time in processing the fd_set.
Admin:
  Tested on desktop build, with the idle timers set to send idle outs
    every second (so we don't have to wait 100 hours for the problem
    to occur).  No longer get problems with the machine stiffing.


Version 2.12. Tagged as 'LanManFS-2_12'

  Now doesn't require that the !ArmBoot object is a directory
Detail:
  The code in Omni.c was carefully checking that !ARMBOOT existed
    before attempting to run the boot file.  Unfortunately, it was
    using a method which bypassed the filename resolution (that does
    the ,xxx filetype name mapping), so it did not find the new Obey
    file in the 400 series baseline.  The check has been removed.
  The code in SMB.c was being caught out on an uninitialised variable
    usage when the attribute cache already held details of the object
    being sought.  The variable would have been initialised on a non-
    cached lookup, but the special case of booting a machine via LanManFS
    manages to get a cached lookup without having run through the routine
    before, resulting in a strcpy() with a destination of 0.
Admin:
  Tested on Risc PC.
  Fixes fault 1511 (STB-400 Generic)

Version 2.11. Tagged as 'LanManFS-2_11'

Detail:
  LanManFS does not like it if you create (independently, using a PC or
    otherwise) files with names like "myfile,fff" which you intend to
    be displayed as files with type &FFF (ie. Text) on a RISC OS machine.
    If you tried to access the file for reading it, such as loading it
    into an editor, that worked due to the name matching resolution.
    However, any attempt to update the file caused LanManFS to attempt
    to write the file without the extension and not notice that a file
    with a ,fff extension already existed (Text files are special cased
    in the current implementation of name translation - see LanManFS
    Functional Specification for details and rationale).  Attempts to
    save typically succeed (giving you two files: myfile and myfile,fff)
    but generate "Operation not permitted" or such like.  This stemmed
    from the attempt being made by LanManFS to rename a file to the same
    name.  SMB_Rename now traps that and tells its caller that the rename
    succeeded.  The Xlate.c code now indicates correctly that the filename
    extension has been altered when a ,fff extension has to be chopped off.
Admin:
  Tested briefly on desktop build against art-work's SAMBA server.
  Bug reported by TopCat engineers.

Version 2.10. Tagged as 'LanManFS-2_10'

  Removed obsolete header files.
  Removed obsolete assembler.
  Fixed flag preservation in remaining assembler.
Detail:
  See above.
Admin:
  Built RAM build.

Version 2.09. Tagged as 'LanManFS-2_09'

Detail:
  Added Install phase to Makefile.
  Added !MkInstall.
Admin:
  Tested on FastNC Ethernet card build.

Version 2.08. Tagged as 'LanManFS-2_08'

  Fixed search handle haemorrhaging.
Detail:
  The directory lookup routines cached directory search handles to avoid
    hammering the server with repeated requests to open directories.
    Unfortunately, it was possible for these to be forgotten without
    being freed up leading to resource loss at the server end.  Eventually,
    the server gives up servicing you and reports errors.  This will only
    occur once a significant number of accesses have been performed (the
    number depends entirely on server configuration).  This leak is now
    fixed.
Admin:
  Tested against Cerium - debug reports that no handles are being leaked.

Version 2.07. Tagged as 'LanManFS-2_07'

  Name mapping table modified to prefer mapping hard space to space.
Detail:
  Fixes hard space->space mismapping.
  Doesn't leak sockets during startup if driver isn't ready.
Admin:
  Check-in of previously uncommitted bug fixes to get final version number


Version 2.06. Tagged as 'LanManFS-2_06'

Detail:
  The module used to have specific knowledge of the driver to which it
    was supposed to bind if it was unable to find any active drivers at
    the time that LanManFS was initialised.  This meant that if the
    LanManFS module was placed in ROM (eg. in STB3) and the drivers
    hadn't initialised by the time that LanManFS was initialised, it
    would sit and wait for EtherH to arrive - which doesn't happen
    for ATM solutions, and doesn't happen for Ethernet in STB3 because
    we no longer use EtherH!
Admin:
  Tested in STB22 expansion cards, and in STB3 ROM builds for both
    Ethernet and ATM solutions.

Version 2.05. Tagged as 'LanManFS-2_05'

  Introduced new error message for re-entrancy prevention trap to use.
Detail:
  When 'pinging' an SMB server, LanManFS does not wait for any response
    but the response reading routine knows to just throw away any old
    SMBchkpth responses that it gets and try reading again.
  Re-entrancy trap now has its own error "LanManFS in use" &1663E, which
    means you no longer see "!Armboot files nested too deeply" which is
    confusing.  The only way you can provoke this message is if you use
    Alt-Break whilst the NetBIOS/IP code is executing.  The error plays
    the same part as "FileCore in use" does for FileCore.
Admin:
  Verified module still works and the anti idle-out features still work.

Version 2.04. Tagged as 'LanManFS-2_04'

  Fix for ./.. files in short filename shares.
  Long multiply now provided in assembler.
Detail:
  *LMLS is a new filesystem-specific command which performs the same
    task as "smbclient -L" does under UNIX - ie. listing all the shares
    available on the specified host.
  When enumerating the shares on a remote server, LanManFS now remembers
    all the printers, IPC, device and disc shares (for *LMLS to output).
  The 32x32->64 multiply routine is now provided in assembler instead of
    the previous huge amount of obscure C taken from NSPRLib.  This is
    more understandable (and almost certainly far smaller and quicker
    and doesn't need a hack to stop the compiler mis-optimising it!)
  Old short filename shares still had problems with the . and .. file
    filtering.  Programmer fault there - should've let the compiler do
    the optimisations :-)
  Server info string fixed to byte-reverse the port numbers when printing
    the string into the info buffer.  Now claims to be talking to port 139
    as it should.
Admin:
  Tested in general desktop use for several days, plus experimental versions
    tested in other boot ROMs.

Version 2.03. Tagged as 'LanManFS-2_03'

Detail:
  Long filename flag tarnsferred to be a server property rather than
    a share property, otherwise subsequent shares to the same server
    do not get long filenames (because the subsequent shares don't
    have to go through the connection negotiation phase)
Admin:
  Tested on by mounting lots of drives through desktop Omni frontend.

Version 2.02. Tagged as 'LanManFS-2_02'

Detail:
  Any filename starting with . was being hidden.
Admin:
  Tested.  Confirmed . and .. are not there but files such as .rhosts
    are visible.

Version 2.01. Tagged as 'LanManFS-2_01'

  Support for spaces in machine names.
  Merge of sbrodie_LanManFS_dev branch to trunk.
Detail:
  LanManFS 2.00 supports the "NT LM 0.12" protocol, enabling it to
    use long filenames on mounted shares.
Admin:
  Supporting documentation:
    1215,256/FS: LanManFS Software Functional Specification.
  Same as LanManFS-1_87_1_1_1_1_2_13.

Version 2.00. Tagged as 'LanManFS-2_00'