From: Chuck Lever <chuck.lever@oracle.com>
Date: Fri, 1 Apr 2022 17:08:21 -0400
Subject: [PATCH] SUNRPC: Fix NFSD's request deferral on RDMA transports
Git-commit: 773f91b2cf3f52df0d7508fdbf60f37567cdaee4
Patch-mainline: v5.18
References: git-fixes

Trond Myklebust reports an NFSD crash in svc_rdma_sendto(). Further
investigation shows that the crash occurred while NFSD was handling
a deferred request.

This patch addresses two inter-related issues that prevent request
deferral from working correctly for RPC/RDMA requests:

1. Prevent the crash by ensuring that the original
   svc_rqst::rq_xprt_ctxt value is available when the request is
   revisited. Otherwise svc_rdma_sendto() does not have a Receive
   context available with which to construct its reply.

2. Possibly since before commit 71641d99ce03 ("svcrdma: Properly
   compute .len and .buflen for received RPC Calls"),
   svc_rdma_recvfrom() did not include the transport header in the
   returned xdr_buf. There should have been no need for svc_defer()
   and friends to save and restore that header, as of that commit.
   This issue is addressed in a backport-friendly way by simply
   having svc_rdma_recvfrom() set rq_xprt_hlen to zero
   unconditionally, just as svc_tcp_recvfrom() does. This enables
   svc_deferred_recv() to correctly reconstruct an RPC message
   received via RPC/RDMA.

Reported-by: Trond Myklebust <trondmy@hammerspace.com>
Link: https://lore.kernel.org/linux-nfs/82662b7190f26fb304eb0ab1bb04279072439d4e.camel@hammerspace.com/
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Cc: <stable@vger.kernel.org>
Acked-by: NeilBrown <neilb@suse.com>

---
 include/linux/sunrpc/svc.h              |    1 +
 net/sunrpc/svc_xprt.c                   |    3 +++
 net/sunrpc/xprtrdma/svc_rdma_recvfrom.c |    2 +-
 3 files changed, 5 insertions(+), 1 deletion(-)

--- a/include/linux/sunrpc/svc.h
+++ b/include/linux/sunrpc/svc.h
@@ -382,6 +382,7 @@ struct svc_deferred_req {
 	size_t			addrlen;
 	struct sockaddr_storage	daddr;	/* where reply must come from */
 	size_t			daddrlen;
+	void			*xprt_ctxt;
 	struct cache_deferred_req handle;
 	size_t			xprt_hlen;
 	int			argslen;
--- a/net/sunrpc/svc_xprt.c
+++ b/net/sunrpc/svc_xprt.c
@@ -1210,6 +1210,8 @@ static struct cache_deferred_req *svc_de
 		dr->daddr = rqstp->rq_daddr;
 		dr->argslen = rqstp->rq_arg.len >> 2;
 		dr->xprt_hlen = rqstp->rq_xprt_hlen;
+		dr->xprt_ctxt = rqstp->rq_xprt_ctxt;
+		rqstp->rq_xprt_ctxt = NULL;
 
 		/* back up head to the start of the buffer and copy */
 		skip = rqstp->rq_arg.len - rqstp->rq_arg.head[0].iov_len;
@@ -1248,6 +1250,7 @@ static noinline int svc_deferred_recv(st
 	rqstp->rq_xprt_hlen   = dr->xprt_hlen;
 	rqstp->rq_daddr       = dr->daddr;
 	rqstp->rq_respages    = rqstp->rq_pages;
+	rqstp->rq_xprt_ctxt   = dr->xprt_ctxt;
 	svc_xprt_received(rqstp->rq_xprt);
 	return (dr->argslen<<2) - dr->xprt_hlen;
 }
--- a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
@@ -826,7 +826,7 @@ int svc_rdma_recvfrom(struct svc_rqst *r
 		goto out_err;
 	if (ret == 0)
 		goto out_drop;
-	rqstp->rq_xprt_hlen = ret;
+	rqstp->rq_xprt_hlen = 0;
 
 	if (svc_rdma_is_reverse_direction_reply(xprt, ctxt))
 		goto out_backchannel;