On 3/6/25 3:52 AM, Jeff Layton wrote:
On Wed, 2025-03-05 at 12:59 -0800, Dai Ngo wrote:
On 3/5/25 12:47 PM, Dai Ngo wrote:
On 3/5/25 8:08 AM, Jeff Layton wrote:
On Wed, 2025-03-05 at 09:46 -0500, Chuck Lever wrote:
On 3/5/25 9:34 AM, Jeff Layton wrote:
On Tue, 2025-03-04 at 12:38 -0800, Dai Ngo wrote:
Allow READ using write delegation stateid granted on OPENs with
OPEN4_SHARE_ACCESS_WRITE only, to accommodate clients whose WRITE
implementation may unavoidably do (e.g., due to buffer cache
constraints).
For write delegation granted for OPEN with OPEN4_SHARE_ACCESS_WRITE
a new nfsd_file and a struct file are allocated to use for reads.
The nfsd_file is freed when the file is closed by release_all_access.
Signed-off-by: Dai Ngo <dai.ngo@xxxxxxxxxx>
---
fs/nfsd/nfs4state.c | 44
++++++++++++++++++++++++++++++++++++--------
1 file changed, 36 insertions(+), 8 deletions(-)
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index b533225e57cf..35018af4e7fb 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -6126,6 +6126,34 @@ nfs4_delegation_stat(struct nfs4_delegation
*dp, struct svc_fh *currentfh,
return rc == 0;
}
+/*
+ * Add NFS4_SHARE_ACCESS_READ to the write delegation granted on
OPEN
+ * with NFS4_SHARE_ACCESS_WRITE by allocating separate nfsd_file and
+ * struct file to be used for read with delegation stateid.
+ *
+ */
+static bool
+nfsd4_add_rdaccess_to_wrdeleg(struct svc_rqst *rqstp, struct
nfsd4_open *open,
+ struct svc_fh *fh, struct nfs4_ol_stateid *stp)
+{
+ struct nfs4_file *fp;
+ struct nfsd_file *nf = NULL;
+
+ if ((open->op_share_access & NFS4_SHARE_ACCESS_BOTH) ==
+ NFS4_SHARE_ACCESS_WRITE) {
+ if (nfsd_file_acquire_opened(rqstp, fh, NFSD_MAY_READ,
NULL, &nf))
+ return (false);
+ fp = stp->st_stid.sc_file;
+ spin_lock(&fp->fi_lock);
+ __nfs4_file_get_access(fp, NFS4_SHARE_ACCESS_READ);
+ set_access(NFS4_SHARE_ACCESS_READ, stp);
The only other (minor) issue is that this might be problematic vs.
DENY_READ modes:
Suppose someone opens the file SHARE_ACCESS_WRITE and gets back a r/w
delegation. Then someone else tries to open the file
SHARE_ACCESS_READ|SHARE_DENY_READ. That should succeed, AFAICT, but I
think with this patch that would fail because we check the deny mode
before doing the open (and revoking the delegation).
It'd be good to test and see if that's the case.
Yes, you're correct. The 2nd OPEN fails due to the read access set
to the file in nfsd4_add_rdaccess_to_wrdeleg().
I think the deny mode is used only by SMB and not Linux client, not
sure though. What should we do about this, any thought?
Deny modes are a Windows/DOS thing, but they are part of the NFSv4 spec
too. Linux doesn't have a userland interface that allows you to set
them, and they aren't plumbed through the VFS layer, so you can still
do an open locally on the box, even if a deny mode is set. I _think_
BSD might also have support at the VFS layer for share/deny locking but
I don't know for sure.
Without this patch, nfsd does not hand out the write delegation and don't
set the read access so the 2nd OPEN would work. But is that the correct
behavior because the open stateid of the 1st OPEN is allowed to do read?
That's a good question.
The main reason the server might allow reads on an O_WRONLY open is
because the client may need to do a RMW cycle if it's doing page-
aligned buffered I/Os. The client really shouldn't allow userland to do
an O_WRONLY open and start issuing read() calls on it, however. So,
from that standpoint I think the original behavior of knfsd does
conform to the spec.
To fix this the right way, we probably need to make the implicit
O_WRONLY -> O_RDRW upgrade for a delegation take some sort of "shadow"
reference. IOW, we need to be able to use the O_RDONLY file internally
and put its reference when the file is closed, but we don't want to
count that reference toward share/deny mode enforcement.
The fix to support deny mode turns out to be simple. When we upgrade
the write delegation from WRONLY to RDWR, don't set the read access in
st_access_bmap, don't need it. The nfsd always allow file opened with
write to do read anyway, this is done in access_permit_read().
-Dai
+ fp = stp->st_stid.sc_file;
+ fp->fi_fds[O_RDONLY] = nf;
+ spin_unlock(&fp->fi_lock);
+ }
+ return (true);
no need for parenthesis here ^^^
Fixed.
+}
+
/*
* The Linux NFS server does not offer write delegations to NFSv4.0
* clients in order to avoid conflicts between write delegations
and
@@ -6151,8 +6179,9 @@ nfs4_delegation_stat(struct nfs4_delegation
*dp, struct svc_fh *currentfh,
* open or lock state.
*/
static void
-nfs4_open_delegation(struct nfsd4_open *open, struct
nfs4_ol_stateid *stp,
- struct svc_fh *currentfh)
+nfs4_open_delegation(struct svc_rqst *rqstp, struct nfsd4_open
*open,
+ struct nfs4_ol_stateid *stp, struct svc_fh *currentfh,
+ struct svc_fh *fh)
{
bool deleg_ts = open->op_deleg_want &
OPEN4_SHARE_ACCESS_WANT_DELEG_TIMESTAMPS;
struct nfs4_openowner *oo = openowner(stp->st_stateowner);
@@ -6197,7 +6226,8 @@ nfs4_open_delegation(struct nfsd4_open
*open, struct nfs4_ol_stateid *stp,
memcpy(&open->op_delegate_stateid, &dp->dl_stid.sc_stateid,
sizeof(dp->dl_stid.sc_stateid));
if (open->op_share_access & NFS4_SHARE_ACCESS_WRITE) {
- if (!nfs4_delegation_stat(dp, currentfh, &stat)) {
+ if ((!nfsd4_add_rdaccess_to_wrdeleg(rqstp, open, fh,
stp)) ||
extra set of parens above too ^^^
Fixed.
+ !nfs4_delegation_stat(dp, currentfh, &stat)) {
nfs4_put_stid(&dp->dl_stid);
destroy_delegation(dp);
goto out_no_deleg;
@@ -6353,7 +6383,8 @@ nfsd4_process_open2(struct svc_rqst *rqstp,
struct svc_fh *current_fh, struct nf
* Attempt to hand out a delegation. No error return, because
the
* OPEN succeeds even if we fail.
*/
- nfs4_open_delegation(open, stp, &resp->cstate.current_fh);
+ nfs4_open_delegation(rqstp, open, stp,
+ &resp->cstate.current_fh, current_fh);
/*
* If there is an existing open stateid, it must be updated and
@@ -7098,10 +7129,6 @@ nfs4_find_file(struct nfs4_stid *s, int flags)
switch (s->sc_type) {
case SC_TYPE_DELEG:
- spin_lock(&s->sc_file->fi_lock);
- ret = nfsd_file_get(s->sc_file->fi_deleg_file);
- spin_unlock(&s->sc_file->fi_lock);
- break;
case SC_TYPE_OPEN:
case SC_TYPE_LOCK:
if (flags & RD_STATE)
@@ -7277,6 +7304,7 @@ nfs4_preprocess_stateid_op(struct svc_rqst
*rqstp,
status = find_cpntf_state(nn, stateid, &s);
if (status)
return status;
+
status = nfsd4_stid_check_stateid_generation(stateid, s,
nfsd4_has_session(cstate));
if (status)
Patch itself looks good though, so with the nits fixed up, you can
add:
Reviewed-by: Jeff Layton <jlayton@xxxxxxxxxx>
Dai, I can fix the parentheses in my tree, no need for a v5.
Thanks Chuck, I will fold these patches into one to avoid potential
bisect issue before sending out v5.
-Dai
