Minifilter drivers call FltCreateFileEx to create a new file or open an existing file.
NTSTATUS FLTAPI FltCreateFileEx( PFLT_FILTER Filter, PFLT_INSTANCE Instance, PHANDLE FileHandle, PFILE_OBJECT *FileObject, ACCESS_MASK DesiredAccess, POBJECT_ATTRIBUTES ObjectAttributes, PIO_STATUS_BLOCK IoStatusBlock, PLARGE_INTEGER AllocationSize, ULONG FileAttributes, ULONG ShareAccess, ULONG CreateDisposition, ULONG CreateOptions, PVOID EaBuffer, ULONG EaLength, ULONG Flags );
An opaque filter pointer for the caller.
An opaque instance pointer for the minifilter driver instance that the create request is to be sent to. The instance must be attached to the volume where the file or directory resides. This parameter is optional and can be NULL. If this parameter is NULL, the request is sent to the device object at the top of the file system driver stack for the volume. If it is non-NULL, the request is sent only to minifilter driver instances that are attached below the specified instance.
A pointer to a caller-allocated variable that receives the file handle if the call to FltCreateFileEx is successful.
A pointer to a caller-allocated variable that receives the file object pointer if the call to FltCreateFileEx is successful. This parameter is optional and can be NULL.
A bitmask of flags specifying the type of access that the caller requires to the file or directory. The set of system-defined DesiredAccess flags determines the following specific access rights for file objects.
Do not specify FILE_READ_DATA, FILE_WRITE_DATA, FILE_EXECUTE, or FILE_APPEND_DATA when creating or opening a directory.
|DELETE||The file can be deleted.|
|FILE_READ_DATA||Data can be read from the file.|
|FILE_READ_ATTRIBUTES||FileAttributes flags, described later, can be read.|
|FILE_READ_EA||Extended attributes associated with the file can be read.|
|READ_CONTROL||The access control list (ACL) and ownership information associated with the file can be read.|
|FILE_WRITE_DATA||Data can be written to the file.|
|FILE_WRITE_ATTRIBUTES||FileAttributes flags can be written.|
|FILE_WRITE_EA||Extended attributes associated with the file can be written.|
|FILE_APPEND_DATA||Data can be appended to the file.|
|WRITE_DAC||The discretionary access control list (DACL) associated with the file can be written.|
|WRITE_OWNER||Ownership information associated with the file can be written.|
|SYNCHRONIZE||The caller can synchronize the completion of an I/O operation by waiting for the returned FileHandle to be set to the Signaled state. This flag must be set if the CreateOptions FILE_SYNCHRONOUS_IO_ALERT or FILE_SYNCHRONOUS_IO_NONALERT flag is set.|
|FILE_EXECUTE||Use system paging I/O to read data from the file into system memory.|
Alternatively, for any file object that does not represent a directory, you can specify one or more of the following generic ACCESS_MASK flags. (The STANDARD_RIGHTS_XXX flags are predefined system values that are used to enforce security on system objects.) You can also combine these generic flags with additional flags from the preceding table.
|DesiredAccess to File Values||Maps to DesiredAccess Flags|
|GENERIC_READ||STANDARD_RIGHTS_READ, FILE_READ_DATA, FILE_READ_ATTRIBUTES, FILE_READ_EA, and SYNCHRONIZE.|
|GENERIC_WRITE||STANDARD_RIGHTS_WRITE, FILE_WRITE_DATA, FILE_WRITE_ATTRIBUTES, FILE_WRITE_EA, FILE_APPEND_DATA, and SYNCHRONIZE.|
|GENERIC_EXECUTE||STANDARD_RIGHTS_EXECUTE, SYNCHRONIZE, FILE_READ_ATTRIBUTES, and FILE_EXECUTE.|
For directories, you can specify one or more of the following ACCESS_MASK flags, which you can also combine with any compatible flags that were described earlier.
|DesiredAccess to Directory Values||Meaning|
|FILE_LIST_DIRECTORY||Files in the directory can be listed.|
|FILE_TRAVERSE||The directory can be traversed: that is, it can be part of the pathname of a file.|
A pointer to a structure already initialized with InitializeObjectAttributes. If the caller is running in the system process context, this parameter can be NULL. Otherwise, the caller must set the OBJ_KERNEL_HANDLE attribute in the call to InitializeObjectAttributes. Members of this structure for a file object include the following.
|ULONGLength||Specifies the number of bytes of ObjectAttributes data supplied. This value must be at least sizeof(OBJECT_ATTRIBUTES).|
|PUNICODE_STRING ObjectName||Pointer to a UNICODE_STRING structure containing the name of the file to be created or opened. This name must be a fully qualified file specification or the name of a device object unless it is the name of a file relative to the directory specified by RootDirectory. For example, \Device\Floppy1\myfile.dat or \??\B:\myfile.dat could be the fully qualified file specification, if the floppy driver and overlying file system are already loaded. (Note: \?? replaces \DosDevices as the name of the Win32 object namespace. \DosDevices still works, but \?? is translated faster by the object manager.)|
|HANDLERootDirectory||Optionally specifies a handle to a directory obtained by a preceding call to FltCreateFileEx. If this value is NULL, the ObjectName member must be a fully qualified file specification that includes the full path to the target file. If this value is non-NULL, the ObjectName member specifies a file name relative to this directory.|
|PSECURITY_DESCRIPTOR SecurityDescriptor||Optionally specifies a security descriptor (SECURITY_DESCRIPTOR) to be applied to a file. ACLs specified by such a security descriptor are only applied to the file when it is created. If the value is NULL when a file is created, the ACL placed on the file is file-system-dependent; most file systems propagate some part of such an ACL from the parent directory file combined with the caller's default ACL.|
|ULONGAttributes||Is a set of flags that controls the file object attributes. If the caller is running in the system process context, this parameter can be zero. Otherwise, the caller must set the OBJ_KERNEL_HANDLE flag. The caller can also optionally set the OBJ_CASE_INSENSITIVE flag, which indicates that name-lookup code should ignore the case of ObjectName rather than performing an exact-match search.|
A pointer to an IO_STATUS_BLOCK structure that receives the final completion status and information about the requested operation. On return from FltCreateFileEx, the Information member contains one of the following values.
Optionally specifies the initial allocation size, in bytes, for the file stream. A nonzero value has no effect unless the file is being created, overwritten, or superseded.
Specifies one or more of the following FILE_ATTRIBUTE_XXX flags, which represent the file attributes to set if you are creating, superseding, or overwriting a file. Normally, you specify FILE_ATTRIBUTE_NORMAL, which sets the default attributes.
|FILE_ATTRIBUTE_NORMAL||A file with standard attributes should be created.|
|FILE_ATTRIBUTE_READONLY||A read-only file should be created.|
|FILE_ATTRIBUTE_HIDDEN||A hidden file should be created.|
|FILE_ATTRIBUTE_SYSTEM||A system file should be created.|
|FILE_ATTRIBUTE_ARCHIVE||An archive file should be created. This attribute is used to mark files for backup or removal.|
|FILE_ATTRIBUTE_TEMPORARY||A temporary file should be created.|
Specifies the type of share access to the file that the caller requires, as zero or one, or a combination of the following flags. If the IO_IGNORE_SHARE_ACCESS_CHECK flag is specified in the Flags parameter, the I/O manager ignores this parameter. However, the file system might still perform access checks. Thus, it is important to specify the sharing mode you would like for this parameter, even when using the IO_IGNORE_SHARE_ACCESS_CHECK flag. For the greatest chance of avoiding sharing violation errors, specify all of the following share access flags.
|FILE_SHARE_READ||The file can be opened for read access by other threads' calls to FltCreateFileEx.|
|FILE_SHARE_WRITE||The file can be opened for write access by other threads' calls to FltCreateFileEx.|
|FILE_SHARE_DELETE||The file can be opened for delete access by other threads' calls to FltCreateFileEx.|
Specifies a value that determines the action to be taken, depending on whether the file already exists. The value can be any of those described following.
|FILE_SUPERSEDE||If the file already exists, replace it with the given file. If it does not, create the given file.|
|FILE_CREATE||If the file already exists, fail the request and do not create or open the given file. If it does not, create the given file.|
|FILE_OPEN||If the file already exists, open it instead of creating a new file. If it does not, fail the request and do not create a new file.|
|FILE_OPEN_IF||If the file already exists, open it. If it does not, create the given file.|
|FILE_OVERWRITE||If the file already exists, open it and overwrite it. If it does not, fail the request.|
|FILE_OVERWRITE_IF||If the file already exists, open it and overwrite it. If it does not, create the given file.|
Specifies the options to be applied when creating or opening the file, as a compatible combination of the following flags.
|FILE_DIRECTORY_FILE (0x00000001)||The file being created or opened is a directory file. With this flag, the CreateDisposition parameter must be set to one of FILE_CREATE, FILE_OPEN, or FILE_OPEN_IF. With this flag, other compatible CreateOptions flags include only the following: FILE_SYNCHRONOUS_IO_ALERT, FILE_SYNCHRONOUS_IO_NONALERT, FILE_WRITE_THROUGH, FILE_OPEN_FOR_BACKUP_INTENT, and FILE_OPEN_BY_FILE_ID.|
|FILE_WRITE_THROUGH (0x00000002)||System services, file systems, and drivers that write data to the file must actually transfer the data into the file before any requested write operation is considered complete.|
|FILE_SEQUENTIAL_ONLY (0x00000004)||All accesses to the file will be sequential.|
|FILE_NO_INTERMEDIATE_BUFFERING (0x00000008)||The file cannot be cached or buffered in a driver's internal buffers. This flag is incompatible with the *DesiredAccess *FILE_APPEND_DATA flag.|
|FILE_SYNCHRONOUS_IO_ALERT (0x00000010)||All operations on the file are performed synchronously. Any wait on behalf of the caller is subject to premature termination from alerts. This flag also causes the I/O system to maintain the file position context. If this flag is set, the DesiredAccess SYNCHRONIZE flag also must be set so that the I/O Manager uses the file object as a synchronization object.|
|FILE_SYNCHRONOUS_IO_NONALERT (0x00000020)||All operations on the file are performed synchronously. Waits in the system to synchronize I/O queuing and completion are not subject to alerts. This flag also causes the I/O system to maintain the file position context. If this flag is set, the DesiredAccess SYNCHRONIZE flag also must be set so that the I/O Manager uses the file object as a synchronization object.|
|FILE_NON_DIRECTORY_FILE (0x00000040)||The file being opened must not be a directory file or this call fails. The file object being opened can represent a data file; a logical, virtual, or physical device; or a volume.|
|FILE_CREATE_TREE_CONNECTION (0x00000080)||Create a tree connection for this file in order to open it over the network.|
|FILE_COMPLETE_IF_OPLOCKED (0x00000100)||Complete this operation immediately with an alternate success code if the target file is oplocked, rather than blocking the caller's thread. If the file is oplocked, another caller already has access to the file over the network.|
|FILE_NO_EA_KNOWLEDGE (0x00000200)||If the extended attributes on an existing file being opened indicate that the caller must understand extended attributes to properly interpret the file, fail this request because the caller does not understand how to deal with extended attributes.|
|FILE_OPEN_REMOTE_INSTANCE (0x00000400)||Reserved for system use; do not use.|
|FILE_RANDOM_ACCESS (0x00000800)||Accesses to the file can be random, so no sequential read-ahead operations should be performed on the file by file systems or the operating system.|
|FILE_DELETE_ON_CLOSE (0x00001000)||Delete the file when the last handle to it is passed to FltClose.|
|FILE_OPEN_BY_FILE_ID (0x00002000)||The file is being opened by ID. The file name contains the name of a device and a 64-bit ID to be used to open the file.|
|FILE_OPEN_FOR_BACKUP_INTENT (0x000004000)||The file is being opened for backup intent; therefore, the system should check for certain access rights and grant the caller the appropriate accesses to the file before checking the input DesiredAccess against the file's security descriptor.|
|FILE_NO_COMPRESSION (0x00008000)||Suppress inheritance of FILE_ATTRIBUTE_COMPRESSED from the parent directory. This allows creation of a non-compressed file in a directory that is marked compressed.|
|FILE_OPEN_REQUIRING_OPLOCK (0x00010000)||The file is being opened and an opportunistic lock (oplock) on the file is being requested as a single atomic operation. The file system checks for oplocks before it performs the create operation, and the create operation will fail with a return code of STATUS_CANNOT_BREAK_OPLOCK if the create operation would break an existing oplock. This flag is available in Windows 7, Windows Server 2008 R2 and later Windows operating systems.|
|FILE_DISALLOW_EXCLUSIVE (0x00020000)||When opening an existing file, if FILE_SHARE_READ is not specified and file system access checks would not grant the caller write access to the file, fail this open with STATUS_ACCESS_DENIED. This was default behavior prior to Windows 7.|
|FILE_SESSION_AWARE (0x00040000)||The file or device is being opened with session awareness. If this flag is not specified, then per-session devices (such as a device using RemoteFX USB Redirection) cannot be opened by processes running in session 0. This flag has no effect for callers not in session 0. This flag is supported only on server editions of Windows. This flag is not supported before Windows Server 2012.|
|FILE_RESERVE_OPFILTER (0x00100000)||This flag allows an application to request a filter opportunistic lock (oplock) to prevent other applications from getting share violations. If there are already open handles, the create request will fail with STATUS_OPLOCK_NOT_GRANTED. For more information, see the following Remarks section.|
|FILE_OPEN_REPARSE_POINT (0x00200000)||Open a file with a reparse point and bypass normal reparse point processing for the file. For more information, see the following Remarks section.|
|FILE_OPEN_NO_RECALL (0x00400000)||Instructs any filters that perform offline storage or virtualization to not recall the contents of the file as a result of this open.|
|FILE_OPEN_FOR_FREE_SPACE_QUERY (0x00800000)||This flag instructs the file system to capture the user associated with the calling thread. Any subsequent calls to FltQueryVolumeInformation or ZwQueryVolumeInformationFile using the returned handle will assume the captured user, rather than the calling user at the time, for purposes of computing the free space available to the caller. This applies to the following FsInformationClass values: FileFsSizeInformation, FileFsFullSizeInformation, and FileFsFullSizeInformationEx.|
A pointer to a caller-supplied FILE_FULL_EA_INFORMATION-structured buffer containing extended attribute (EA) information to be applied to the file.
Length, in bytes, of EaBuffer.
Specifies options to be used during the creation of the create request. The following table lists the available options.
|IO_FORCE_ACCESS_CHECK||Indicates that the I/O manager must check the create request against the file's security descriptor.|
|IO_IGNORE_SHARE_ACCESS_CHECK||Indicates that the I/O manager should not perform share-access checks on the file object after it is created. However, the file system might still perform these checks.|
|IO_STOP_ON_SYMLINK||The I/O manager or the file system will return STATUS_STOPPED_ON_SYMLINK if a symbolic link is encountered while opening or creating the file.|
FltCreateFileEx returns STATUS_SUCCESS or an appropriate NTSTATUS value such as one of the following:
||The specified Filter or Instance is being torn down. This status code can be received if the open crosses a volume mount point and the Instance parameter is non-NULL. This is an error code.|
FltCreateFileEx encountered a pool allocation failure. This is an error code.
||The file or directory name contains a mount point that resolves to a volume other than the one that the specified minifilter driver instance is attached to. This is an error code.|
||The ObjectAttributes parameter did not contain a RootDirectory member, but the ObjectName member in the OBJECT_ATTRIBUTES structure was an empty string or did not contain an OBJECT_NAME_PATH_SEPARATOR character. This error code indicates incorrect syntax for the object path.|
FltCreateFileEx is identical to FltCreateFile except for the addition of an optional FileObject output parameter.
FltCreateFileEx sends the create request only to the instances attached below the specified minifilter driver instance and to the file system. The specified instance and the instances attached above it do not receive the create request. If no instance is specified, the request goes to the top of the stack and is received by all instances and the file system.
There are two alternate ways to specify the name of the file to be created or opened with FltCreateFileEx:
- As a fully qualified pathname, supplied in the ObjectName member of the input ObjectAttributes
- As a pathname relative to the directory file represented by the handle in the RootDirectory member of the input ObjectAttributes
Driver routines that do not run in the system process context must set the OBJ_KERNEL_HANDLE attribute for the ObjectAttributes parameter of FltCreateFileEx. Setting this attribute restricts the use of the handle that is returned by FltCreateFileEx to processes running in kernel mode. Otherwise, the handle can be accessed by the process in whose context the driver is running.
Certain DesiredAccess flags and combinations of flags have the following effects:
- For a caller to synchronize an I/O completion by waiting for the returned FileHandle to be set to the Signaled state, the SYNCHRONIZE flag must be set.
- If only the FILE_APPEND_DATA and SYNCHRONIZE flags are set, the caller can write only to the end of the file, and any offset information about writes to the file is ignored. However, the file is automatically extended as necessary for this type of write operation.
- Setting the FILE_WRITE_DATA flag for a file also allows writes beyond the end of the file to occur. The file is automatically extended for this type of write, as well.
- If only the FILE_EXECUTE and SYNCHRONIZE flags are set, the caller cannot use the returned FileHandle to directly read or write any data to or from the file. That is, all operations on the file must use system paging I/O to read or write file data.
For a shared file to be successfully opened, the requested DesiredAccess to the file must be compatible with both the DesiredAccess and ShareAccess specifications of all preceding opens that have not yet been released with FltClose. That is, the DesiredAccess specified to FltCreateFileEx for a given file must not conflict with the accesses that other openers of the file have disallowed.
The CreateDisposition values FILE_OVERWRITE_IF and FILE_SUPERSEDE are similar. If FltCreateFileEx is called with an existing file and either of these CreateDisposition values, the file is replaced.
Overwriting a file is semantically equivalent to a supersede operation, except for the following:
- The caller must have write access to the file, rather than delete access. This implies that, if the file has already been opened by another thread, it opened the file with the FILE_SHARE_WRITE flag set in the input ShareAccess.
- The specified file attributes are logically ORed with those already on the file. This implies that if the file has already been opened by another thread, a subsequent caller of FltCreateFileEx cannot disable existing FileAttributes flags but can enable additional flags for the same file. Note that this style of overwriting files is consistent with MS-DOS, Windows 3.1, and OS/2.
The CreateOptions FILE_NO_INTERMEDIATE_BUFFERING flag prevents the file system from performing any intermediate buffering on behalf of the caller. Specifying this value places certain restrictions on the caller's parameters to other Flt..File routines or Zw..File routines, including the following:
- Any byte Offset value passed to FltReadFile, ZwReadFile, FltWriteFile, or ZwWriteFile must be a multiple of the sector size.
- The Length passed to FltReadFile, ZwReadFile, FltWriteFile, or ZwWriteFile must be a multiple of the sector size. Note that specifying a read operation to a buffer whose length is exactly the sector size might result in fewer significant bytes being transferred to that buffer if the end of the file was reached during the transfer.
- Buffers must be aligned in accordance with the alignment requirement of the underlying storage device. This information can be obtained by calling FltCreateFileEx to get a handle for the file object that represents the physical device and then calling ZwQueryInformationFile with that handle, specifying FileAlignmentInformation as the FileInformationClass. For more information about the system FILE_XXX_ALIGNMENT values, which are defined in ntifs.h, see DEVICE_OBJECT and Initializing a Device Object.
- Calls to FltSetInformationFile or ZwSetInformationFile with the FileInformationClass parameter set to FilePositionInformation must specify an offset that is a multiple of the sector size.
If the CreateOptions FILE_OPEN_REPARSE_POINT flag is not specified and FltCreateFileEx attempts to open a file with a reparse point, normal reparse point processing occurs for the file. If, on the other hand, the FILE_OPEN_REPARSE_POINT flag is specified, normal reparse processing does not occur and FltCreateFileEx attempts to directly open the reparse point file. In either case, if the open operation was successful, FltCreateFileEx returns STATUS_SUCCESS; otherwise, the routine returns an NTSTATUS error code. FltCreateFileEx never returns STATUS_REPARSE.
The CreateOptions FILE_OPEN_REQUIRING_OPLOCK flag eliminates the time between when you open the file and request an oplock that could potentially enable a third party to open the file and get a sharing violation. An application can use the FILE_OPEN_REQUIRING_OPLOCK flag on FltCreateFileEx and then request any oplock. This ensures that an oplock owner will be notified of any later open request that causes a sharing violation.
In Windows 7, if other handles exist on the file when an application uses the FILE_OPEN_REQUIRING_OPLOCK flag, the create operation will fail with STATUS_OPLOCK_NOT_GRANTED. This restriction no longer exists starting with Windows 8.
If this create operation would break an oplock that already exists on the file, then setting the FILE_OPEN_REQUIRING_OPLOCK flag will cause the create operation to fail with STATUS_CANNOT_BREAK_OPLOCK. The existing oplock will not be broken by this create operation.
An application that uses this flag must request an oplock after this call succeeds, or all later attempts to open the file will be blocked without the benefit of typical oplock processing. Similarly, if this call succeeds but the later oplock request fails, an application that uses this flag must close its handle after it detects that the oplock request has failed.
- Issue a create request with CreateOptions of FILE_RESERVE_OPFILTER, DesiredAccess of exactly FILE_READ_ATTRIBUTES, and ShareAccess of exactly FILE_SHARE_READ | FILE_SHARE_WRITE | FILE_SHARE_DELETE.
- If there are already open handles, the create request fails with STATUS_OPLOCK_NOT_GRANTED, and the next requested oplock also fails.
- If you open with more access or less sharing will also cause a failure of STATUS_OPLOCK_NOT_GRANTED.
- If the create request succeeds, request an oplock.
- Open another handle to the file to do I/O.
NTFS is the only Microsoft file system that implements FILE_RESERVE_OPFILTER.
You must not use FltCreateFileEx to open a handle with direct access to the storage device for the volume or you will leak system resources. If you want to open a handle with direct access to a storage device, call the IoCreateFileEx, IoCreateFileSpecifyDeviceObjectHint, or ZwCreateFile function instead.
|Minimum supported client||Available in Microsoft Windows 2000 Update Rollup 1 for SP4, Windows XP SP3, Windows Server 2003 SP1, and later versions of the Windows operating system.|
|Header||fltkernel.h (include FltKernel.h)|