struct NFS41_V_NET_ROOT_EXTENSION now stores only the FILE_FS_ATTRIBUTE_INFORMATION (without the extra buffer space for a name). on QueryVolumeInfo() for FileFsAttributeInformation on the root directory, the FILE_FS_ATTRIBUTE_INFORMATION is copied into the output buffer, and the name is added there. QueryVolumeInfo() only makes upcalls when FileFsAttributeInformation queries are not for the root directory
new function is_root_directory() uses the logic from Set/GetReparsePoint() to determine whether it's operating on the root directory
moved logic from volume.c:handle_volume_attributes() to superblock.c:nfs41_superblock_fs_attributes(). the mount downcall copies the FILE_FS_ATTRIBUTE_INFORMATION buffer down to the driver. the driver reads this buffer directly into VNetRootContext->FsAttrs
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
if non-cygwin EAs are given to Create, SetEaInfo, or QueryEaInfo and the server does not support named attributes, fail with STATUS_EAS_NOT_SUPPORTED. also adds access checks for FILE_READ_EA and FILE_WRITE_EA
Create will accept all 3 cygwin EAs: NfsV3Attributes, NfsActOnLink, and NfsSymlinkTargetName. it now handles NfsActOnLink by adding FILE_OPEN_REPARSE_POINT to the create options
SetEaInfo will only accept NfsV3Attributes for setting the file mode. the other two will fail with STATUS_INVALID_PARAMETER. removed handling of NfsActOnLink in setattr.c
QueryEaInfo will accept all 3 cygwin EAs
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
adds a new function nfs41_superblock_getattr(), which adds an OPENATTR call with createdir=0 after the GETATTR to check for named attribute support. OPENATTR will either return ERR_NOTSUPP if not supported, or OK/ERR_NOENT depending on whether named attributes are present on the given file. we can't query for the 'named_attr' attribute, because it only tells us whether the given file contains named attributes
if named attributes are supported on a superblock, it will return FILE_SUPPORTS_EXTENDED_ATTRIBUTES for associated volume queries
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
zero-initialize the paths for named attrs, and initialize the file.name instead of using a temporary dst_name
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
reverting patch 4d18cf9ce7.
it was wrong to append the timestamp to create a silly renamed name
because it was not a reproducable name.
instead, take an fh and md5(fh) then append as before.
assign a superblock to named attribute files on nfs41_open(), to prevent crashing in nfs41_write()
also avoid updating the attribute cache with named attribute files on close and write
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
on creation of a new superblock, construct a bitmap for the default attribute mask to be used for GETATTR and READDIR requests on that filesystem. mask out any unsupported attributes, and store the bitmap in the field nfs41_superblock.default_getattr
replaced function init_getattr_request() with nfs41_superblock_getattr_mask(), which returns a copy of superblock->default_getattr
removed the locking in nfs41_superblock_supported_attrs() and nfs41_superblock_supported_attrs_exclcreat(), as the supported_attrs and suppattr_exclcreat fields are read-only after the superblock is first initialized. also factored out their common code into a bitmap_intersect() function in util.h
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
the upcall to set an EA was crashing in nfs41_open() on a null superblock, because nfs41_rpc_openattr() was returning a filehandle without a superblock. copy the parent file's superblock to the returned filehandle
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
allow callers of nfs41_open() and nfs41_create() to pass arbitrary attributes fo
r use with open_args.openhow.how.createattrs and create_args.createattrs
instead of appending a fh value when creating a silly rename name,
we'll append a timestamp. using fh value was problematic because
it was creating a filename longer than 64char long which is
currently out max_filename_size.
removing notepad opens file as a directory ifdef
for volume query notepad supplies insufficient buffer space but
doesn't like buffer_too_small error code. it seems to be fine
with partial buffer filled and buffer_overflow error.
windows can send a create irp with dispostion representing creation
of the file and specify DesiredAccess of 0.
we were treating no data access as lookup but in this case, we still
need to create a file. so send an open with a read access and
request no delegation to be returned.
in comparing server identities, in previous commit we removed
comparison of the returned clientids.
however, running against the emc server, we ran into issues of data
servers retuning the same server major, minor, and scope identities
but different clientids. we then decided that it's the same data
server.
adding -o timeout=value (in sec) mount option to override a default
timeout value for the upcalls for non-io upcalls. Default timeout is
at least the lease_time seconds of an established mount or 20secs.
read/write upcall also take into account number of bytes for the request
and 100MB/s network speed and 100MB/s disk speed
when we are doing non-pnfs and writing to a netapp filer, even though we
write UNSTABLE it returns us FILE_SYNC4. since we were doing unstable
writes we were not sending getattrs, and since the data servers we returning
stable commits we didn't send commit. in doing so, we never update our
attribute cache and had a wrong file size.
client was previously only sending RECLAIM_COMPLETE during grace period. the spec mandates that RECLAIM_COMPLETE be sent before any non-reclaim locking operations, regardless of grace period. when recovery code detects a NFS4ERR_NO_GRACE error, send RECLAIM_COMPLETE immediately before attempting any out-of-grace recovery
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
CB_GETATTR specifies the delegation by filehandle, but deleg_fh_cmp() was comparing superblock and fileid. renamed deleg_fh_cmp() to deleg_file_cmp(), and created new deleg_fh_cmp() to compare actual filehandles
call to nfs41_attr_cache_lookup() was not setting flags in info.attrmask, so the change and size attributes were not being encoded in the CB_GETATTR response
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
fixes an issue introduced by commit 1f7e560a9a, "pnfs: fix for short write on ds error"
when a write is satisfied completely by pnfs, the upcall was returning out_len=0. the only time i noticed this was against python server, where cp in cygwin would succeed but print 'No space left on device'
Signed-off-by: Casey Bodley <cbodley@citi.umich.edu>
