Net user command:
- Fix typos in the /ACTIVE option.
- Implement the /COUNTRYCODE, /PASSWORDCHG and /PASSWORDREQ options.
svn path=/trunk/; revision=65179
Whatever happened during format, always unmount the volume.
This allows switching FS in usetup after formating an ext2 volume.
As a record, before (read, with the IopParseDevice) hack, this wasn't possible and if formating with ext2 was possible, kernel would have written with either RawFS or FastFAT driver. Which is not exactly what we want ;-).
But no joy guys... The Ext2Fsd needs to work in ReactOS first before we can effectively install ReactOS on ext2 volumes.
In any case, that's a clear step forward :-)
svn path=/trunk/; revision=65174
Implement some kind of volume dismount in FastFAT (ie, implement VfatDismountVolume())
This is not fully perfect situation, but it does most of the job (+ leaking a few things). So far, this is not dramatic as its main use is for 1st stage. This will have to be improved later on.
CORE-8732 #comment Can you retest please?
svn path=/trunk/; revision=65173
- Correctly set the VdmPower flag when the process being started is a NT VDM
- Return a correct status code if VDM access is forbidden.
svn path=/trunk/; revision=65170
- Mark almost all of the exported console APIs hotpatchable (normally on windows 2k3, all the exports of kernel32 are hotpatchable...)
- Version check mismatch is not a ReactOS error by itself so adjust the DPRINT a bit.
svn path=/trunk/; revision=65169
Simplify buffer size management in GetNtfsFileRecord(). As tests shown, Windows doesn't care about the real size of the file record and expects the caller to pass the complete size of the record and returns it fully.
CORE-8725
svn path=/trunk/; revision=65166
Add a simple application (à la NTFSInfo & nfi) to dump the data about a NTFS volume using the FSCTL control codes.
This is for both demonstrating how it works and to assess our driver behavior against Windows behavior of these FSCTL.
CORE-8725
svn path=/trunk/; revision=65165
- rewrite the GEO_FRIENDLYNAME case of GetGeoInfoW which was completely wrong, intl.cpl now shows a valid list of locations
- note: The geo location resource strings have an arbitrary offset so that they don't collide with the nls resources (which shouldn't be there). Once these are moved out of kernel32 into locale.nls this offset alongside with the hack already present in GetLocalisedText can be removed.
svn path=/trunk/; revision=65157
- pszEnvironment is a multi-sz string so make sure to double-null terminate it. Fixes a buffer overrun in winlogon; powered by DPH.
svn path=/trunk/; revision=65153
- fall back to english if a localized resource is not found in GetLocalisedText
- this prevents debug spam and loss of functionality (see CORE-8720) for languages that don't have kernel32 translated
svn path=/trunk/; revision=65152
* Move TRASH_CanTrashFile and TRASH_TrashFile declarations to recyclebin.h.
* Do not include xdg.h in shlfileop.cpp anymore.
* Remove the now unneeded xdg.h.
svn path=/trunk/; revision=65146
- Move FSCTL_GET_NTFS_VOLUME_DATA implementation in its own function GetNfsVolumeData(), so remove it from NtfsUserFsRequest()
- Add support for FSCTL_GET_NTFS_FILE_RECORD in NtfsUserFsRequest()
- Implement GetNtfsFileRecord() which returns a file record. Not fully sure about its implementation, this is to be checked.
This does not really improve the nfi situation in any kind yet...
CORE-8725
svn path=/trunk/; revision=65145
Testman confirms that IopParseDevice() hack is no longer required.
So, this is the end of this 8y old hack.
Now ReactOS is able to properly use its storage stack (who said, amazing?)
Bisous IopParseDevice() hack :-)
svn path=/trunk/; revision=65144
Revert r65120...
Which was reverting r65119...
Which was reverting r65117...
Which was reverting r65110...
Which was reverting r65107...
Which was reverting r65104!
To make it short? Fuck you IopParseDevice() hack!
And this time, I hope this is the right one!
I couldn't reproduce any of the previous issues in my WC.
svn path=/trunk/; revision=65142
Some fixes and missing implementations in RawFS:
- Remove some magic values for VCB state
- Add a local VPB in our VCB. This will allow diverting the IO VPB with ours in case of dismount (and thus, will fix dismount)
- Implement support for said VPB in both RawCheckForDismount() and RawInitializeVcb()
- Now, dismount is handled in RawCheckForDismount() directly, and not in functions calling it
- Handle streams in RawClose
- Handle VCB being dismounted in RawCreate()
- Handle dismount also in RawCleanup() (not only in RawClose())
This fixes volume locking and dismounting in RawFS. And thus, allows FS swap in case of raw volume formating in first stage :-).
More to follow!
svn path=/trunk/; revision=65141
- Fix renaming files in case the target file already exists.
- Plug a few reference count leaks on FCBs.
CORE-8660 #comment fixed in r65140 #resolve
CORE-8633 #comment fixed in r65140 #resolve
CORE-4758 #comment fixed in r65140 #resolve
svn path=/trunk/; revision=65140
On the road for getting rid of IopParseDevice() hack....
- Add support for FSCTL_LOCK_VOLUME, FSCTL_UNLOCK_VOLUME, FSCTL_DISMOUNT_VOLUME in VfatFileSystemControl()
- Implement VfatLockOrUnlockVolume(). This brings some kind of volume locking to FastFAT FSD. It's likely not fully accurate and welcome to race conditions. But, its main purpose is to be used in usetup. So, this is fine for now.
- Stubplement VfatDismountVolume(). This will be implemented later on, but should be relatively straightforward as it requires volume to be locked (so, we're the last ones on it :-)).
svn path=/trunk/; revision=65129
