Received-SPF: pass (google.com: domain of linux-nfs-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Subject: [PATCH] NFSD: Fix trace_nfsd_fh_verify_err() crasher
From:   Chuck Lever <chuck.lever@oracle.com>
To:     linux-nfs@vger.kernel.org
Date:   Sat, 12 Nov 2022 15:06:07 -0500
Message-ID: <166828356768.42842.2412588018662481301.stgit@klimt.1015granger.net>
User-Agent: StGit/1.5.dev3+g9561319
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
Precedence: bulk

Now that the nfsd_fh_verify_err() tracepoint is always called on
error, it needs to handle cases where the filehandle is not yet
fully formed.

Fixes: 93c128e709ae ("nfsd: ensure we always call fh_verify_error tracepoint")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
---
 fs/nfsd/trace.h |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/fs/nfsd/trace.h b/fs/nfsd/trace.h
index ef01ecd3eec6..331a33a8f1f8 100644
--- a/fs/nfsd/trace.h
+++ b/fs/nfsd/trace.h
@@ -254,7 +254,10 @@ TRACE_EVENT_CONDITION(nfsd_fh_verify_err,
 				  rqstp->rq_xprt->xpt_remotelen);
 		__entry->xid = be32_to_cpu(rqstp->rq_xid);
 		__entry->fh_hash = knfsd_fh_hash(&fhp->fh_handle);
-		__entry->inode = d_inode(fhp->fh_dentry);
+		if (fhp->fh_dentry)
+			__entry->inode = d_inode(fhp->fh_dentry);
+		else
+			__entry->inode = NULL;
 		__entry->type = type;
 		__entry->access = access;
 		__entry->error = be32_to_cpu(error);