[PATCH 1/2] nfsd4: common stateid-printing code

[PATCH 1/2] nfsd4: common stateid-printing code

[J. Bruce Fields]

From: "J. Bruce Fields" <bfields@redhat.com>

There's a problem with how I'm formatting stateids.  Before I fix it,
I'd like to move the stateid formatting into a common helper.

Signed-off-by: J. Bruce Fields <bfields@redhat.com>
---
 fs/nfsd/nfs4state.c | 21 +++++++++++++++++----
 1 file changed, 17 insertions(+), 4 deletions(-)

diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index e32ecedece0f..7537f2f5156e 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -2420,6 +2420,11 @@ static void nfs4_show_owner(struct seq_file *s, struct nfs4_stateowner *oo)
 	seq_quote_mem(s, oo->so_owner.data, oo->so_owner.len);
 }
 
+static void nfs4_show_stateid(struct seq_file *s, stateid_t *stid)
+{
+	seq_printf(s, "0x%16phN", stid);
+}
+
 static int nfs4_show_open(struct seq_file *s, struct nfs4_stid *st)
 {
 	struct nfs4_ol_stateid *ols;
@@ -2435,7 +2440,9 @@ static int nfs4_show_open(struct seq_file *s, struct nfs4_stid *st)
 	nf = st->sc_file;
 	file = find_any_file(nf);
 
-	seq_printf(s, "- 0x%16phN: { type: open, ", &st->sc_stateid);
+	seq_printf(s, "- ");
+	nfs4_show_stateid(s, &st->sc_stateid);
+	seq_printf(s, ": { type: open, ");
 
 	access = bmap_to_share_mode(ols->st_access_bmap);
 	deny   = bmap_to_share_mode(ols->st_deny_bmap);
@@ -2468,7 +2475,9 @@ static int nfs4_show_lock(struct seq_file *s, struct nfs4_stid *st)
 	nf = st->sc_file;
 	file = find_any_file(nf);
 
-	seq_printf(s, "- 0x%16phN: { type: lock, ", &st->sc_stateid);
+	seq_printf(s, "- ");
+	nfs4_show_stateid(s, &st->sc_stateid);
+	seq_printf(s, ": { type: lock, ");
 
 	/*
 	 * Note: a lock stateid isn't really the same thing as a lock,
@@ -2497,7 +2506,9 @@ static int nfs4_show_deleg(struct seq_file *s, struct nfs4_stid *st)
 	nf = st->sc_file;
 	file = nf->fi_deleg_file;
 
-	seq_printf(s, "- 0x%16phN: { type: deleg, ", &st->sc_stateid);
+	seq_printf(s, "- ");
+	nfs4_show_stateid(s, &st->sc_stateid);
+	seq_printf(s, ": { type: deleg, ");
 
 	/* Kinda dead code as long as we only support read delegs: */
 	seq_printf(s, "access: %s, ",
@@ -2519,7 +2530,9 @@ static int nfs4_show_layout(struct seq_file *s, struct nfs4_stid *st)
 	ls = container_of(st, struct nfs4_layout_stateid, ls_stid);
 	file = ls->ls_file;
 
-	seq_printf(s, "- 0x%16phN: { type: layout, ", &st->sc_stateid);
+	seq_printf(s, "- ");
+	nfs4_show_stateid(s, &st->sc_stateid);
+	seq_printf(s, ": { type: layout, ");
 
 	/* XXX: What else would be useful? */
 
-- 
2.25.3

[PATCH 2/2] nfsd4: stid display should preserve on-the-wire byte order

[J. Bruce Fields]

From: "J. Bruce Fields" <bfields@redhat.com>

When we decode the stateid we byte-swap si_generation.

But for simplicity's sake and ease of comparison with network traces,
it's better to display the whole thing in network order.

Reported-by: Scott Mayhew <smayhew@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
---
 fs/nfsd/nfs4state.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 7537f2f5156e..a6e0a7f77eb0 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -2422,7 +2422,8 @@ static void nfs4_show_owner(struct seq_file *s, struct nfs4_stateowner *oo)
 
 static void nfs4_show_stateid(struct seq_file *s, stateid_t *stid)
 {
-	seq_printf(s, "0x%16phN", stid);
+	seq_printf(s, "0x%.8x", stid->si_generation);
+	seq_printf(s, "%12phN", &stid->si_opaque);
 }
 
 static int nfs4_show_open(struct seq_file *s, struct nfs4_stid *st)
-- 
2.25.3

Re: [PATCH 2/2] nfsd4: stid display should preserve on-the-wire byte order

[J. Bruce Fields]

By the way, one other thing I noticed, in the "states" file there's a
field like:

	superblock: "fd:10:114"

That's major:minor:inode number, which is the same thing /proc/locks
uses to identify files.  "superblock" makes sense for the "major:minor"
part, but the inode number is for the one file, not the superblock.  So
that probably should have been

	superblock: "fd:10:114", inode: 114

or something.  Oh well.  It's been that way in a few kernel versions now
so I guess it's not worth breaking backwards compatibility.

--b.