From ec482cc1c1180c65da582e17f4e79d204bb338e0 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:23:44 -0500 Subject: [PATCH 01/90] xprtrdma: Fix sparse warnings linux/net/sunrpc/xprtrdma/rpc_rdma.c:375:63: warning: incorrect type in argument 5 (different base types) linux/net/sunrpc/xprtrdma/rpc_rdma.c:375:63: expected unsigned int [usertype] xid linux/net/sunrpc/xprtrdma/rpc_rdma.c:375:63: got restricted __be32 [usertype] rq_xid linux/net/sunrpc/xprtrdma/rpc_rdma.c:432:62: warning: incorrect type in argument 5 (different base types) linux/net/sunrpc/xprtrdma/rpc_rdma.c:432:62: expected unsigned int [usertype] xid linux/net/sunrpc/xprtrdma/rpc_rdma.c:432:62: got restricted __be32 [usertype] rq_xid linux/net/sunrpc/xprtrdma/rpc_rdma.c:489:62: warning: incorrect type in argument 5 (different base types) linux/net/sunrpc/xprtrdma/rpc_rdma.c:489:62: expected unsigned int [usertype] xid linux/net/sunrpc/xprtrdma/rpc_rdma.c:489:62: got restricted __be32 [usertype] rq_xid Fixes: 0a93fbcb16e6 ("xprtrdma: Plant XID in on-the-wire RDMA ... ") Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- net/sunrpc/xprtrdma/frwr_ops.c | 4 ++-- net/sunrpc/xprtrdma/xprt_rdma.h | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/net/sunrpc/xprtrdma/frwr_ops.c b/net/sunrpc/xprtrdma/frwr_ops.c index 6a561056b538..52cb6c1b0c2b 100644 --- a/net/sunrpc/xprtrdma/frwr_ops.c +++ b/net/sunrpc/xprtrdma/frwr_ops.c @@ -391,7 +391,7 @@ frwr_wc_localinv_wake(struct ib_cq *cq, struct ib_wc *wc) */ struct rpcrdma_mr_seg *frwr_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, - int nsegs, bool writing, u32 xid, + int nsegs, bool writing, __be32 xid, struct rpcrdma_mr **out) { struct rpcrdma_ia *ia = &r_xprt->rx_ia; @@ -446,7 +446,7 @@ struct rpcrdma_mr_seg *frwr_map(struct rpcrdma_xprt *r_xprt, goto out_mapmr_err; ibmr->iova &= 0x00000000ffffffff; - ibmr->iova |= ((u64)cpu_to_be32(xid)) << 32; + ibmr->iova |= ((u64)be32_to_cpu(xid)) << 32; key = (u8)(ibmr->rkey & 0x000000FF); ib_update_fast_reg_key(ibmr, ++key); diff --git a/net/sunrpc/xprtrdma/xprt_rdma.h b/net/sunrpc/xprtrdma/xprt_rdma.h index 5a18472f2c9c..33db208fa9a8 100644 --- a/net/sunrpc/xprtrdma/xprt_rdma.h +++ b/net/sunrpc/xprtrdma/xprt_rdma.h @@ -577,7 +577,7 @@ void frwr_release_mr(struct rpcrdma_mr *mr); size_t frwr_maxpages(struct rpcrdma_xprt *r_xprt); struct rpcrdma_mr_seg *frwr_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, - int nsegs, bool writing, u32 xid, + int nsegs, bool writing, __be32 xid, struct rpcrdma_mr **mr); int frwr_send(struct rpcrdma_ia *ia, struct rpcrdma_req *req); void frwr_reminv(struct rpcrdma_rep *rep, struct list_head *mrs); From d4550bbee66f4ba5a5e9bbe8055006332ebfc58b Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:23:49 -0500 Subject: [PATCH 02/90] xprtrdma: Check inline size before providing a Write chunk In very rare cases, an NFS READ operation might predict that the non-payload part of the RPC Call is large. For instance, an NFSv4 COMPOUND with a large GETATTR result, in combination with a large Kerberos credential, could push the non-payload part to be several kilobytes. If the non-payload part is larger than the connection's inline threshold, the client is required to provision a Reply chunk. The current Linux client does not check for this case. There are two obvious ways to handle it: a. Provision a Write chunk for the payload and a Reply chunk for the non-payload part b. Provision a Reply chunk for the whole RPC Reply Some testing at a recent NFS bake-a-thon showed that servers can mostly handle a. but there are some corner cases that do not work yet. b. already works (it has to, to handle krb5i/p), but could be somewhat less efficient. However, I expect this scenario to be very rare -- no-one has reported a problem yet. So I'm going to implement b. Sometime later I will provide some patches to help make b. a little more efficient by more carefully choosing the Reply chunk's segment sizes to ensure the payload is optimally aligned. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- net/sunrpc/xprtrdma/rpc_rdma.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/net/sunrpc/xprtrdma/rpc_rdma.c b/net/sunrpc/xprtrdma/rpc_rdma.c index d18614e02b4e..7774aee7c013 100644 --- a/net/sunrpc/xprtrdma/rpc_rdma.c +++ b/net/sunrpc/xprtrdma/rpc_rdma.c @@ -164,6 +164,21 @@ static bool rpcrdma_results_inline(struct rpcrdma_xprt *r_xprt, return rqst->rq_rcv_buf.buflen <= ia->ri_max_inline_read; } +/* The client is required to provide a Reply chunk if the maximum + * size of the non-payload part of the RPC Reply is larger than + * the inline threshold. + */ +static bool +rpcrdma_nonpayload_inline(const struct rpcrdma_xprt *r_xprt, + const struct rpc_rqst *rqst) +{ + const struct xdr_buf *buf = &rqst->rq_rcv_buf; + const struct rpcrdma_ia *ia = &r_xprt->rx_ia; + + return buf->head[0].iov_len + buf->tail[0].iov_len < + ia->ri_max_inline_read; +} + /* Split @vec on page boundaries into SGEs. FMR registers pages, not * a byte range. Other modes coalesce these SGEs into a single MR * when they can. @@ -762,7 +777,8 @@ rpcrdma_marshal_req(struct rpcrdma_xprt *r_xprt, struct rpc_rqst *rqst) */ if (rpcrdma_results_inline(r_xprt, rqst)) wtype = rpcrdma_noch; - else if (ddp_allowed && rqst->rq_rcv_buf.flags & XDRBUF_READ) + else if ((ddp_allowed && rqst->rq_rcv_buf.flags & XDRBUF_READ) && + rpcrdma_nonpayload_inline(r_xprt, rqst)) wtype = rpcrdma_writech; else wtype = rpcrdma_replych; From e340c2d6ef2a8cdcc11672f8cc839c30ad360e52 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:23:54 -0500 Subject: [PATCH 03/90] xprtrdma: Reduce the doorbell rate (Receive) Post RECV WRs in batches to reduce the hardware doorbell rate per transport. This helps the RPC-over-RDMA client scale better in number of transports. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- net/sunrpc/xprtrdma/verbs.c | 2 ++ net/sunrpc/xprtrdma/xprt_rdma.h | 10 ++++++++++ 2 files changed, 12 insertions(+) diff --git a/net/sunrpc/xprtrdma/verbs.c b/net/sunrpc/xprtrdma/verbs.c index 4994e75945b8..b4e997d53ec7 100644 --- a/net/sunrpc/xprtrdma/verbs.c +++ b/net/sunrpc/xprtrdma/verbs.c @@ -1480,6 +1480,8 @@ rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, bool temp) if (ep->rep_receive_count > needed) goto out; needed -= ep->rep_receive_count; + if (!temp) + needed += RPCRDMA_MAX_RECV_BATCH; count = 0; wr = NULL; diff --git a/net/sunrpc/xprtrdma/xprt_rdma.h b/net/sunrpc/xprtrdma/xprt_rdma.h index 33db208fa9a8..10f6593e1a6a 100644 --- a/net/sunrpc/xprtrdma/xprt_rdma.h +++ b/net/sunrpc/xprtrdma/xprt_rdma.h @@ -205,6 +205,16 @@ struct rpcrdma_rep { struct ib_recv_wr rr_recv_wr; }; +/* To reduce the rate at which a transport invokes ib_post_recv + * (and thus the hardware doorbell rate), xprtrdma posts Receive + * WRs in batches. + * + * Setting this to zero disables Receive post batching. + */ +enum { + RPCRDMA_MAX_RECV_BATCH = 7, +}; + /* struct rpcrdma_sendctx - DMA mapped SGEs to unmap after Send completes */ struct rpcrdma_req; From 6f701383368d54e8d43fad2a0bac30f877899568 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:00 -0500 Subject: [PATCH 04/90] SUNRPC: Display symbolic flag names in RPC trace events Human-readable flags make it easier to observe RPC scheduling decisions and other operational details. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/trace/events/sunrpc.h | 56 +++++++++++++++++++++++++++++++---- 1 file changed, 50 insertions(+), 6 deletions(-) diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 0d5d0d91f861..f88b0f52aa7e 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -77,6 +77,50 @@ TRACE_EVENT(rpc_request, ) ); +TRACE_DEFINE_ENUM(RPC_TASK_ASYNC); +TRACE_DEFINE_ENUM(RPC_TASK_SWAPPER); +TRACE_DEFINE_ENUM(RPC_CALL_MAJORSEEN); +TRACE_DEFINE_ENUM(RPC_TASK_ROOTCREDS); +TRACE_DEFINE_ENUM(RPC_TASK_DYNAMIC); +TRACE_DEFINE_ENUM(RPC_TASK_KILLED); +TRACE_DEFINE_ENUM(RPC_TASK_SOFT); +TRACE_DEFINE_ENUM(RPC_TASK_SOFTCONN); +TRACE_DEFINE_ENUM(RPC_TASK_SENT); +TRACE_DEFINE_ENUM(RPC_TASK_TIMEOUT); +TRACE_DEFINE_ENUM(RPC_TASK_NOCONNECT); +TRACE_DEFINE_ENUM(RPC_TASK_NO_RETRANS_TIMEOUT); + +#define rpc_show_task_flags(flags) \ + __print_flags(flags, "|", \ + { RPC_TASK_ASYNC, "ASYNC" }, \ + { RPC_TASK_SWAPPER, "SWAPPER" }, \ + { RPC_CALL_MAJORSEEN, "MAJORSEEN" }, \ + { RPC_TASK_ROOTCREDS, "ROOTCREDS" }, \ + { RPC_TASK_DYNAMIC, "DYNAMIC" }, \ + { RPC_TASK_KILLED, "KILLED" }, \ + { RPC_TASK_SOFT, "SOFT" }, \ + { RPC_TASK_SOFTCONN, "SOFTCONN" }, \ + { RPC_TASK_SENT, "SENT" }, \ + { RPC_TASK_TIMEOUT, "TIMEOUT" }, \ + { RPC_TASK_NOCONNECT, "NOCONNECT" }, \ + { RPC_TASK_NO_RETRANS_TIMEOUT, "NORTO" }) + +TRACE_DEFINE_ENUM(RPC_TASK_RUNNING); +TRACE_DEFINE_ENUM(RPC_TASK_QUEUED); +TRACE_DEFINE_ENUM(RPC_TASK_ACTIVE); +TRACE_DEFINE_ENUM(RPC_TASK_NEED_XMIT); +TRACE_DEFINE_ENUM(RPC_TASK_NEED_RECV); +TRACE_DEFINE_ENUM(RPC_TASK_MSG_PIN_WAIT); + +#define rpc_show_runstate(flags) \ + __print_flags(flags, "|", \ + { (1UL << RPC_TASK_RUNNING), "RUNNING" }, \ + { (1UL << RPC_TASK_QUEUED), "QUEUED" }, \ + { (1UL << RPC_TASK_ACTIVE), "ACTIVE" }, \ + { (1UL << RPC_TASK_NEED_XMIT), "NEED_XMIT" }, \ + { (1UL << RPC_TASK_NEED_RECV), "NEED_RECV" }, \ + { (1UL << RPC_TASK_MSG_PIN_WAIT), "MSG_PIN_WAIT" }) + DECLARE_EVENT_CLASS(rpc_task_running, TP_PROTO(const struct rpc_task *task, const void *action), @@ -102,10 +146,10 @@ DECLARE_EVENT_CLASS(rpc_task_running, __entry->flags = task->tk_flags; ), - TP_printk("task:%u@%d flags=%4.4x state=%4.4lx status=%d action=%pf", + TP_printk("task:%u@%d flags=%s runstate=%s status=%d action=%pf", __entry->task_id, __entry->client_id, - __entry->flags, - __entry->runstate, + rpc_show_task_flags(__entry->flags), + rpc_show_runstate(__entry->runstate), __entry->status, __entry->action ) @@ -149,10 +193,10 @@ DECLARE_EVENT_CLASS(rpc_task_queued, __assign_str(q_name, rpc_qname(q)); ), - TP_printk("task:%u@%d flags=%4.4x state=%4.4lx status=%d timeout=%lu queue=%s", + TP_printk("task:%u@%d flags=%s runstate=%s status=%d timeout=%lu queue=%s", __entry->task_id, __entry->client_id, - __entry->flags, - __entry->runstate, + rpc_show_task_flags(__entry->flags), + rpc_show_runstate(__entry->runstate), __entry->status, __entry->timeout, __get_str(q_name) From 0ccc61b1c76e5163c6fea6cf83bd18e7ea244c5b Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:05 -0500 Subject: [PATCH 05/90] SUNRPC: Add xdr_stream::rqst field Having access to the controlling rpc_rqst means a trace point in the XDR code can report: - the XID - the task ID and client ID - the p_name of RPC being processed Subsequent patches will introduce such trace points. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- fs/nfs/callback_xdr.c | 5 +++-- fs/nfs/flexfilelayout/flexfilelayout.c | 2 +- include/linux/sunrpc/xdr.h | 8 ++++++-- net/sunrpc/auth.c | 4 ++-- net/sunrpc/auth_gss/auth_gss.c | 4 ++-- net/sunrpc/xdr.c | 12 +++++++++--- net/sunrpc/xprtrdma/backchannel.c | 2 +- net/sunrpc/xprtrdma/rpc_rdma.c | 4 ++-- 8 files changed, 26 insertions(+), 15 deletions(-) diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index a87a56273407..bc7c1766a2be 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -943,10 +943,11 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp) }; unsigned int nops = 0; - xdr_init_decode(&xdr_in, &rqstp->rq_arg, rqstp->rq_arg.head[0].iov_base); + xdr_init_decode(&xdr_in, &rqstp->rq_arg, + rqstp->rq_arg.head[0].iov_base, NULL); p = (__be32*)((char *)rqstp->rq_res.head[0].iov_base + rqstp->rq_res.head[0].iov_len); - xdr_init_encode(&xdr_out, &rqstp->rq_res, p); + xdr_init_encode(&xdr_out, &rqstp->rq_res, p, NULL); status = decode_compound_hdr_arg(&xdr_in, &hdr_arg); if (status == htonl(NFS4ERR_RESOURCE)) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 63abe705f4ca..32701b6a9566 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -2036,7 +2036,7 @@ ff_layout_encode_layoutreturn(struct xdr_stream *xdr, dprintk("%s: Begin\n", __func__); - xdr_init_encode(&tmp_xdr, &tmp_buf, NULL); + xdr_init_encode(&tmp_xdr, &tmp_buf, NULL, NULL); ff_layout_encode_ioerr(&tmp_xdr, args, ff_args); ff_layout_encode_iostats_array(&tmp_xdr, args, ff_args); diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index 2ec128060239..787939d13643 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -217,6 +217,8 @@ struct xdr_stream { struct kvec scratch; /* Scratch buffer */ struct page **page_ptr; /* pointer to the current page */ unsigned int nwords; /* Remaining decode buffer length */ + + struct rpc_rqst *rqst; /* For debugging */ }; /* @@ -227,7 +229,8 @@ typedef void (*kxdreproc_t)(struct rpc_rqst *rqstp, struct xdr_stream *xdr, typedef int (*kxdrdproc_t)(struct rpc_rqst *rqstp, struct xdr_stream *xdr, void *obj); -extern void xdr_init_encode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p); +extern void xdr_init_encode(struct xdr_stream *xdr, struct xdr_buf *buf, + __be32 *p, struct rpc_rqst *rqst); extern __be32 *xdr_reserve_space(struct xdr_stream *xdr, size_t nbytes); extern void xdr_commit_encode(struct xdr_stream *xdr); extern void xdr_truncate_encode(struct xdr_stream *xdr, size_t len); @@ -235,7 +238,8 @@ extern int xdr_restrict_buflen(struct xdr_stream *xdr, int newbuflen); extern void xdr_write_pages(struct xdr_stream *xdr, struct page **pages, unsigned int base, unsigned int len); extern unsigned int xdr_stream_pos(const struct xdr_stream *xdr); -extern void xdr_init_decode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p); +extern void xdr_init_decode(struct xdr_stream *xdr, struct xdr_buf *buf, + __be32 *p, struct rpc_rqst *rqst); extern void xdr_init_decode_pages(struct xdr_stream *xdr, struct xdr_buf *buf, struct page **pages, unsigned int len); extern void xdr_set_scratch_buffer(struct xdr_stream *xdr, void *buf, size_t buflen); diff --git a/net/sunrpc/auth.c b/net/sunrpc/auth.c index f3023bbc0b7f..8dfab6119e6a 100644 --- a/net/sunrpc/auth.c +++ b/net/sunrpc/auth.c @@ -798,7 +798,7 @@ static void rpcauth_wrap_req_encode(kxdreproc_t encode, struct rpc_rqst *rqstp, { struct xdr_stream xdr; - xdr_init_encode(&xdr, &rqstp->rq_snd_buf, data); + xdr_init_encode(&xdr, &rqstp->rq_snd_buf, data, rqstp); encode(rqstp, &xdr, obj); } @@ -823,7 +823,7 @@ rpcauth_unwrap_req_decode(kxdrdproc_t decode, struct rpc_rqst *rqstp, { struct xdr_stream xdr; - xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, data); + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, data, rqstp); return decode(rqstp, &xdr, obj); } diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 1531b0219344..a42672e81792 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -1722,7 +1722,7 @@ static void gss_wrap_req_encode(kxdreproc_t encode, struct rpc_rqst *rqstp, { struct xdr_stream xdr; - xdr_init_encode(&xdr, &rqstp->rq_snd_buf, p); + xdr_init_encode(&xdr, &rqstp->rq_snd_buf, p, rqstp); encode(rqstp, &xdr, obj); } @@ -1998,7 +1998,7 @@ gss_unwrap_req_decode(kxdrdproc_t decode, struct rpc_rqst *rqstp, { struct xdr_stream xdr; - xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p, rqstp); return decode(rqstp, &xdr, obj); } diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index f302c6eb8779..345f08b634ee 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -483,6 +483,7 @@ EXPORT_SYMBOL_GPL(xdr_stream_pos); * @xdr: pointer to xdr_stream struct * @buf: pointer to XDR buffer in which to encode data * @p: current pointer inside XDR buffer + * @rqst: pointer to controlling rpc_rqst, for debugging * * Note: at the moment the RPC client only passes the length of our * scratch buffer in the xdr_buf's header kvec. Previously this @@ -491,7 +492,8 @@ EXPORT_SYMBOL_GPL(xdr_stream_pos); * of the buffer length, and takes care of adjusting the kvec * length for us. */ -void xdr_init_encode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p) +void xdr_init_encode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p, + struct rpc_rqst *rqst) { struct kvec *iov = buf->head; int scratch_len = buf->buflen - buf->page_len - buf->tail[0].iov_len; @@ -513,6 +515,7 @@ void xdr_init_encode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p) buf->len += len; iov->iov_len += len; } + xdr->rqst = rqst; } EXPORT_SYMBOL_GPL(xdr_init_encode); @@ -819,8 +822,10 @@ static bool xdr_set_next_buffer(struct xdr_stream *xdr) * @xdr: pointer to xdr_stream struct * @buf: pointer to XDR buffer from which to decode data * @p: current pointer inside XDR buffer + * @rqst: pointer to controlling rpc_rqst, for debugging */ -void xdr_init_decode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p) +void xdr_init_decode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p, + struct rpc_rqst *rqst) { xdr->buf = buf; xdr->scratch.iov_base = NULL; @@ -836,6 +841,7 @@ void xdr_init_decode(struct xdr_stream *xdr, struct xdr_buf *buf, __be32 *p) xdr->nwords -= p - xdr->p; xdr->p = p; } + xdr->rqst = rqst; } EXPORT_SYMBOL_GPL(xdr_init_decode); @@ -854,7 +860,7 @@ void xdr_init_decode_pages(struct xdr_stream *xdr, struct xdr_buf *buf, buf->page_len = len; buf->buflen = len; buf->len = len; - xdr_init_decode(xdr, buf, NULL); + xdr_init_decode(xdr, buf, NULL, NULL); } EXPORT_SYMBOL_GPL(xdr_init_decode_pages); diff --git a/net/sunrpc/xprtrdma/backchannel.c b/net/sunrpc/xprtrdma/backchannel.c index 0de9b3e63770..98c1e43eb7b1 100644 --- a/net/sunrpc/xprtrdma/backchannel.c +++ b/net/sunrpc/xprtrdma/backchannel.c @@ -123,7 +123,7 @@ static int rpcrdma_bc_marshal_reply(struct rpc_rqst *rqst) rpcrdma_set_xdrlen(&req->rl_hdrbuf, 0); xdr_init_encode(&req->rl_stream, &req->rl_hdrbuf, - req->rl_rdmabuf->rg_base); + req->rl_rdmabuf->rg_base, rqst); p = xdr_reserve_space(&req->rl_stream, 28); if (unlikely(!p)) diff --git a/net/sunrpc/xprtrdma/rpc_rdma.c b/net/sunrpc/xprtrdma/rpc_rdma.c index 7774aee7c013..6c1fb270f127 100644 --- a/net/sunrpc/xprtrdma/rpc_rdma.c +++ b/net/sunrpc/xprtrdma/rpc_rdma.c @@ -748,7 +748,7 @@ rpcrdma_marshal_req(struct rpcrdma_xprt *r_xprt, struct rpc_rqst *rqst) rpcrdma_set_xdrlen(&req->rl_hdrbuf, 0); xdr_init_encode(xdr, &req->rl_hdrbuf, - req->rl_rdmabuf->rg_base); + req->rl_rdmabuf->rg_base, rqst); /* Fixed header fields */ ret = -EMSGSIZE; @@ -1329,7 +1329,7 @@ void rpcrdma_reply_handler(struct rpcrdma_rep *rep) /* Fixed transport header fields */ xdr_init_decode(&rep->rr_stream, &rep->rr_hdrbuf, - rep->rr_hdrbuf.head[0].iov_base); + rep->rr_hdrbuf.head[0].iov_base, NULL); p = xdr_inline_decode(&rep->rr_stream, 4 * sizeof(*p)); if (unlikely(!p)) goto out_shortreply; From 5582863f450ca44c472dbf4812a62f4e0e68b0c8 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:10 -0500 Subject: [PATCH 06/90] SUNRPC: Add XDR overflow trace event This can help field troubleshooting without needing the overhead of a full network capture (ie, tcpdump). Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/trace/events/sunrpc.h | 67 +++++++++++++++++++++++++++++++++++ net/sunrpc/xdr.c | 24 +++++++++---- 2 files changed, 84 insertions(+), 7 deletions(-) diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index f88b0f52aa7e..fbc41b8142d3 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -254,6 +254,73 @@ TRACE_EVENT(rpc_stats_latency, __entry->backlog, __entry->rtt, __entry->execute) ); +TRACE_EVENT(rpc_xdr_overflow, + TP_PROTO( + const struct xdr_stream *xdr, + size_t requested + ), + + TP_ARGS(xdr, requested), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + __field(int, version) + __field(size_t, requested) + __field(const void *, end) + __field(const void *, p) + __field(const void *, head_base) + __field(size_t, head_len) + __field(const void *, tail_base) + __field(size_t, tail_len) + __field(unsigned int, page_len) + __field(unsigned int, len) + __string(progname, + xdr->rqst->rq_task->tk_client->cl_program->name) + __string(procedure, + xdr->rqst->rq_task->tk_msg.rpc_proc->p_name) + ), + + TP_fast_assign( + if (xdr->rqst) { + const struct rpc_task *task = xdr->rqst->rq_task; + + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; + __assign_str(progname, + task->tk_client->cl_program->name) + __entry->version = task->tk_client->cl_vers; + __assign_str(procedure, task->tk_msg.rpc_proc->p_name) + } else { + __entry->task_id = 0; + __entry->client_id = 0; + __assign_str(progname, "unknown") + __entry->version = 0; + __assign_str(procedure, "unknown") + } + __entry->requested = requested; + __entry->end = xdr->end; + __entry->p = xdr->p; + __entry->head_base = xdr->buf->head[0].iov_base, + __entry->head_len = xdr->buf->head[0].iov_len, + __entry->page_len = xdr->buf->page_len, + __entry->tail_base = xdr->buf->tail[0].iov_base, + __entry->tail_len = xdr->buf->tail[0].iov_len, + __entry->len = xdr->buf->len; + ), + + TP_printk( + "task:%u@%u %sv%d %s requested=%zu p=%p end=%p xdr=[%p,%zu]/%u/[%p,%zu]/%u\n", + __entry->task_id, __entry->client_id, + __get_str(progname), __entry->version, __get_str(procedure), + __entry->requested, __entry->p, __entry->end, + __entry->head_base, __entry->head_len, + __entry->page_len, + __entry->tail_base, __entry->tail_len, + __entry->len + ) +); + /* * First define the enums in the below macros to be exported to userspace * via TRACE_DEFINE_ENUM(). diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 345f08b634ee..6d0b615a02ae 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -16,6 +16,7 @@ #include #include #include +#include /* * XDR functions for basic NFS types @@ -554,9 +555,9 @@ static __be32 *xdr_get_next_encode_buffer(struct xdr_stream *xdr, int frag1bytes, frag2bytes; if (nbytes > PAGE_SIZE) - return NULL; /* Bigger buffers require special handling */ + goto out_overflow; /* Bigger buffers require special handling */ if (xdr->buf->len + nbytes > xdr->buf->buflen) - return NULL; /* Sorry, we're totally out of space */ + goto out_overflow; /* Sorry, we're totally out of space */ frag1bytes = (xdr->end - xdr->p) << 2; frag2bytes = nbytes - frag1bytes; if (xdr->iov) @@ -585,6 +586,9 @@ static __be32 *xdr_get_next_encode_buffer(struct xdr_stream *xdr, xdr->buf->page_len += frag2bytes; xdr->buf->len += nbytes; return p; +out_overflow: + trace_rpc_xdr_overflow(xdr, nbytes); + return NULL; } /** @@ -902,20 +906,23 @@ static __be32 *xdr_copy_to_scratch(struct xdr_stream *xdr, size_t nbytes) size_t cplen = (char *)xdr->end - (char *)xdr->p; if (nbytes > xdr->scratch.iov_len) - return NULL; + goto out_overflow; p = __xdr_inline_decode(xdr, cplen); if (p == NULL) return NULL; memcpy(cpdest, p, cplen); + if (!xdr_set_next_buffer(xdr)) + goto out_overflow; cpdest += cplen; nbytes -= cplen; - if (!xdr_set_next_buffer(xdr)) - return NULL; p = __xdr_inline_decode(xdr, nbytes); if (p == NULL) return NULL; memcpy(cpdest, p, nbytes); return xdr->scratch.iov_base; +out_overflow: + trace_rpc_xdr_overflow(xdr, nbytes); + return NULL; } /** @@ -932,14 +939,17 @@ __be32 * xdr_inline_decode(struct xdr_stream *xdr, size_t nbytes) { __be32 *p; - if (nbytes == 0) + if (unlikely(nbytes == 0)) return xdr->p; if (xdr->p == xdr->end && !xdr_set_next_buffer(xdr)) - return NULL; + goto out_overflow; p = __xdr_inline_decode(xdr, nbytes); if (p != NULL) return p; return xdr_copy_to_scratch(xdr, nbytes); +out_overflow: + trace_rpc_xdr_overflow(xdr, nbytes); + return NULL; } EXPORT_SYMBOL_GPL(xdr_inline_decode); From 7be9cea3600ba24f45733f823cb4ecb5a40f5db7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:16 -0500 Subject: [PATCH 07/90] SUNRPC: Add trace event that reports reply page vector alignment We don't want READ payloads that are partially in the head iovec and in the page buffer because this requires pull-up, which can be expensive. The NFS/RPC client tries hard to predict the size of the head iovec so that the incoming READ data payload lands only in the page vector, but it doesn't always get it right. To help diagnose such problems, add a trace point in the logic that decodes READ-like operations that reports whether pull-up is being done. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/trace/events/sunrpc.h | 59 +++++++++++++++++++++++++++++++++++ net/sunrpc/xdr.c | 33 ++++++++++++++++---- 2 files changed, 86 insertions(+), 6 deletions(-) diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index fbc41b8142d3..627650800676 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -321,6 +321,65 @@ TRACE_EVENT(rpc_xdr_overflow, ) ); +TRACE_EVENT(rpc_xdr_alignment, + TP_PROTO( + const struct xdr_stream *xdr, + size_t offset, + unsigned int copied + ), + + TP_ARGS(xdr, offset, copied), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + __field(int, version) + __field(size_t, offset) + __field(unsigned int, copied) + __field(const void *, head_base) + __field(size_t, head_len) + __field(const void *, tail_base) + __field(size_t, tail_len) + __field(unsigned int, page_len) + __field(unsigned int, len) + __string(progname, + xdr->rqst->rq_task->tk_client->cl_program->name) + __string(procedure, + xdr->rqst->rq_task->tk_msg.rpc_proc->p_name) + ), + + TP_fast_assign( + const struct rpc_task *task = xdr->rqst->rq_task; + + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; + __assign_str(progname, + task->tk_client->cl_program->name) + __entry->version = task->tk_client->cl_vers; + __assign_str(procedure, task->tk_msg.rpc_proc->p_name) + + __entry->offset = offset; + __entry->copied = copied; + __entry->head_base = xdr->buf->head[0].iov_base, + __entry->head_len = xdr->buf->head[0].iov_len, + __entry->page_len = xdr->buf->page_len, + __entry->tail_base = xdr->buf->tail[0].iov_base, + __entry->tail_len = xdr->buf->tail[0].iov_len, + __entry->len = xdr->buf->len; + ), + + TP_printk( + "task:%u@%u %sv%d %s offset=%zu copied=%u xdr=[%p,%zu]/%u/[%p,%zu]/%u\n", + __entry->task_id, __entry->client_id, + __get_str(progname), __entry->version, __get_str(procedure), + __entry->offset, __entry->copied, + __entry->head_base, __entry->head_len, + __entry->page_len, + __entry->tail_base, __entry->tail_len, + __entry->len + ) +); + /* * First define the enums in the below macros to be exported to userspace * via TRACE_DEFINE_ENUM(). diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 6d0b615a02ae..5f0aa53fa4ae 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -347,13 +347,15 @@ EXPORT_SYMBOL_GPL(_copy_from_pages); * 'len' bytes. The extra data is not lost, but is instead * moved into the inlined pages and/or the tail. */ -static void +static unsigned int xdr_shrink_bufhead(struct xdr_buf *buf, size_t len) { struct kvec *head, *tail; size_t copy, offs; unsigned int pglen = buf->page_len; + unsigned int result; + result = 0; tail = buf->tail; head = buf->head; @@ -367,6 +369,7 @@ xdr_shrink_bufhead(struct xdr_buf *buf, size_t len) copy = tail->iov_len - len; memmove((char *)tail->iov_base + len, tail->iov_base, copy); + result += copy; } /* Copy from the inlined pages into the tail */ copy = len; @@ -377,11 +380,13 @@ xdr_shrink_bufhead(struct xdr_buf *buf, size_t len) copy = 0; else if (copy > tail->iov_len - offs) copy = tail->iov_len - offs; - if (copy != 0) + if (copy != 0) { _copy_from_pages((char *)tail->iov_base + offs, buf->pages, buf->page_base + pglen + offs - len, copy); + result += copy; + } /* Do we also need to copy data from the head into the tail ? */ if (len > pglen) { offs = copy = len - pglen; @@ -391,6 +396,7 @@ xdr_shrink_bufhead(struct xdr_buf *buf, size_t len) (char *)head->iov_base + head->iov_len - offs, copy); + result += copy; } } /* Now handle pages */ @@ -406,12 +412,15 @@ xdr_shrink_bufhead(struct xdr_buf *buf, size_t len) _copy_to_pages(buf->pages, buf->page_base, (char *)head->iov_base + head->iov_len - len, copy); + result += copy; } head->iov_len -= len; buf->buflen -= len; /* Have we truncated the message? */ if (buf->len > buf->buflen) buf->len = buf->buflen; + + return result; } /** @@ -423,14 +432,16 @@ xdr_shrink_bufhead(struct xdr_buf *buf, size_t len) * 'len' bytes. The extra data is not lost, but is instead * moved into the tail. */ -static void +static unsigned int xdr_shrink_pagelen(struct xdr_buf *buf, size_t len) { struct kvec *tail; size_t copy; unsigned int pglen = buf->page_len; unsigned int tailbuf_len; + unsigned int result; + result = 0; tail = buf->tail; BUG_ON (len > pglen); @@ -448,18 +459,22 @@ xdr_shrink_pagelen(struct xdr_buf *buf, size_t len) if (tail->iov_len > len) { char *p = (char *)tail->iov_base + len; memmove(p, tail->iov_base, tail->iov_len - len); + result += tail->iov_len - len; } else copy = tail->iov_len; /* Copy from the inlined pages into the tail */ _copy_from_pages((char *)tail->iov_base, buf->pages, buf->page_base + pglen - len, copy); + result += copy; } buf->page_len -= len; buf->buflen -= len; /* Have we truncated the message? */ if (buf->len > buf->buflen) buf->len = buf->buflen; + + return result; } void @@ -959,13 +974,17 @@ static unsigned int xdr_align_pages(struct xdr_stream *xdr, unsigned int len) struct kvec *iov; unsigned int nwords = XDR_QUADLEN(len); unsigned int cur = xdr_stream_pos(xdr); + unsigned int copied, offset; if (xdr->nwords == 0) return 0; + /* Realign pages to current pointer position */ - iov = buf->head; + iov = buf->head; if (iov->iov_len > cur) { - xdr_shrink_bufhead(buf, iov->iov_len - cur); + offset = iov->iov_len - cur; + copied = xdr_shrink_bufhead(buf, offset); + trace_rpc_xdr_alignment(xdr, offset, copied); xdr->nwords = XDR_QUADLEN(buf->len - cur); } @@ -977,7 +996,9 @@ static unsigned int xdr_align_pages(struct xdr_stream *xdr, unsigned int len) len = buf->page_len; else if (nwords < xdr->nwords) { /* Truncate page data and move it into the tail */ - xdr_shrink_pagelen(buf, buf->page_len - len); + offset = buf->page_len - len; + copied = xdr_shrink_pagelen(buf, offset); + trace_rpc_xdr_alignment(xdr, offset, copied); xdr->nwords = XDR_QUADLEN(buf->len - cur); } return len; From eb72f484a5eb94c53a241e6a7811270fb25200ad Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:21 -0500 Subject: [PATCH 08/90] NFS: Remove print_overflow_msg() This issue is now captured by a trace point in the RPC client. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- fs/lockd/clnt4xdr.c | 14 -- fs/lockd/clntxdr.c | 14 -- fs/nfs/callback_xdr.c | 59 +++--- fs/nfs/nfs2xdr.c | 84 +++----- fs/nfs/nfs3xdr.c | 163 +++++---------- fs/nfs/nfs42xdr.c | 21 +- fs/nfs/nfs4xdr.c | 451 +++++++++++------------------------------ fs/nfsd/nfs4callback.c | 13 -- 8 files changed, 219 insertions(+), 600 deletions(-) diff --git a/fs/lockd/clnt4xdr.c b/fs/lockd/clnt4xdr.c index 214a2fa1f1e3..7df6324ccb8a 100644 --- a/fs/lockd/clnt4xdr.c +++ b/fs/lockd/clnt4xdr.c @@ -74,17 +74,6 @@ static void nlm4_compute_offsets(const struct nlm_lock *lock, *l_len = loff_t_to_s64(fl->fl_end - fl->fl_start + 1); } -/* - * Handle decode buffer overflows out-of-line. - */ -static void print_overflow_msg(const char *func, const struct xdr_stream *xdr) -{ - dprintk("lockd: %s prematurely hit the end of our receive buffer. " - "Remaining buffer length is %tu words.\n", - func, xdr->end - xdr->p); -} - - /* * Encode/decode NLMv4 basic data types * @@ -176,7 +165,6 @@ out_size: dprintk("NFS: returned cookie was too long: %u\n", length); return -EIO; out_overflow: - print_overflow_msg(__func__, xdr); return -EIO; } @@ -236,7 +224,6 @@ out_bad_xdr: __func__, be32_to_cpup(p)); return -EIO; out_overflow: - print_overflow_msg(__func__, xdr); return -EIO; } @@ -309,7 +296,6 @@ static int decode_nlm4_holder(struct xdr_stream *xdr, struct nlm_res *result) out: return error; out_overflow: - print_overflow_msg(__func__, xdr); return -EIO; } diff --git a/fs/lockd/clntxdr.c b/fs/lockd/clntxdr.c index 747b9c8c940a..4df62f635529 100644 --- a/fs/lockd/clntxdr.c +++ b/fs/lockd/clntxdr.c @@ -70,17 +70,6 @@ static void nlm_compute_offsets(const struct nlm_lock *lock, *l_len = loff_t_to_s32(fl->fl_end - fl->fl_start + 1); } -/* - * Handle decode buffer overflows out-of-line. - */ -static void print_overflow_msg(const char *func, const struct xdr_stream *xdr) -{ - dprintk("lockd: %s prematurely hit the end of our receive buffer. " - "Remaining buffer length is %tu words.\n", - func, xdr->end - xdr->p); -} - - /* * Encode/decode NLMv3 basic data types * @@ -173,7 +162,6 @@ out_size: dprintk("NFS: returned cookie was too long: %u\n", length); return -EIO; out_overflow: - print_overflow_msg(__func__, xdr); return -EIO; } @@ -231,7 +219,6 @@ out_enum: __func__, be32_to_cpup(p)); return -EIO; out_overflow: - print_overflow_msg(__func__, xdr); return -EIO; } @@ -303,7 +290,6 @@ static int decode_nlm_holder(struct xdr_stream *xdr, struct nlm_res *result) out: return error; out_overflow: - print_overflow_msg(__func__, xdr); return -EIO; } diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index bc7c1766a2be..06233bfa6d73 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -72,16 +72,6 @@ static int nfs4_encode_void(struct svc_rqst *rqstp, __be32 *p) return xdr_ressize_check(rqstp, p); } -static __be32 *read_buf(struct xdr_stream *xdr, size_t nbytes) -{ - __be32 *p; - - p = xdr_inline_decode(xdr, nbytes); - if (unlikely(p == NULL)) - printk(KERN_WARNING "NFS: NFSv4 callback reply buffer overflowed!\n"); - return p; -} - static __be32 decode_string(struct xdr_stream *xdr, unsigned int *len, const char **str, size_t maxlen) { @@ -98,13 +88,13 @@ static __be32 decode_fh(struct xdr_stream *xdr, struct nfs_fh *fh) { __be32 *p; - p = read_buf(xdr, 4); + p = xdr_inline_decode(xdr, 4); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); fh->size = ntohl(*p); if (fh->size > NFS4_FHSIZE) return htonl(NFS4ERR_BADHANDLE); - p = read_buf(xdr, fh->size); + p = xdr_inline_decode(xdr, fh->size); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); memcpy(&fh->data[0], p, fh->size); @@ -117,11 +107,11 @@ static __be32 decode_bitmap(struct xdr_stream *xdr, uint32_t *bitmap) __be32 *p; unsigned int attrlen; - p = read_buf(xdr, 4); + p = xdr_inline_decode(xdr, 4); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); attrlen = ntohl(*p); - p = read_buf(xdr, attrlen << 2); + p = xdr_inline_decode(xdr, attrlen << 2); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); if (likely(attrlen > 0)) @@ -135,7 +125,7 @@ static __be32 decode_stateid(struct xdr_stream *xdr, nfs4_stateid *stateid) { __be32 *p; - p = read_buf(xdr, NFS4_STATEID_SIZE); + p = xdr_inline_decode(xdr, NFS4_STATEID_SIZE); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); memcpy(stateid->data, p, NFS4_STATEID_SIZE); @@ -156,7 +146,7 @@ static __be32 decode_compound_hdr_arg(struct xdr_stream *xdr, struct cb_compound status = decode_string(xdr, &hdr->taglen, &hdr->tag, CB_OP_TAGLEN_MAXSZ); if (unlikely(status != 0)) return status; - p = read_buf(xdr, 12); + p = xdr_inline_decode(xdr, 12); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); hdr->minorversion = ntohl(*p++); @@ -176,7 +166,7 @@ static __be32 decode_compound_hdr_arg(struct xdr_stream *xdr, struct cb_compound static __be32 decode_op_hdr(struct xdr_stream *xdr, unsigned int *op) { __be32 *p; - p = read_buf(xdr, 4); + p = xdr_inline_decode(xdr, 4); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE_HDR); *op = ntohl(*p); @@ -205,7 +195,7 @@ static __be32 decode_recall_args(struct svc_rqst *rqstp, status = decode_delegation_stateid(xdr, &args->stateid); if (unlikely(status != 0)) return status; - p = read_buf(xdr, 4); + p = xdr_inline_decode(xdr, 4); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); args->truncate = ntohl(*p); @@ -227,7 +217,7 @@ static __be32 decode_layoutrecall_args(struct svc_rqst *rqstp, __be32 status = 0; uint32_t iomode; - p = read_buf(xdr, 4 * sizeof(uint32_t)); + p = xdr_inline_decode(xdr, 4 * sizeof(uint32_t)); if (unlikely(p == NULL)) return htonl(NFS4ERR_BADXDR); @@ -245,14 +235,14 @@ static __be32 decode_layoutrecall_args(struct svc_rqst *rqstp, if (unlikely(status != 0)) return status; - p = read_buf(xdr, 2 * sizeof(uint64_t)); + p = xdr_inline_decode(xdr, 2 * sizeof(uint64_t)); if (unlikely(p == NULL)) return htonl(NFS4ERR_BADXDR); p = xdr_decode_hyper(p, &args->cbl_range.offset); p = xdr_decode_hyper(p, &args->cbl_range.length); return decode_layout_stateid(xdr, &args->cbl_stateid); } else if (args->cbl_recall_type == RETURN_FSID) { - p = read_buf(xdr, 2 * sizeof(uint64_t)); + p = xdr_inline_decode(xdr, 2 * sizeof(uint64_t)); if (unlikely(p == NULL)) return htonl(NFS4ERR_BADXDR); p = xdr_decode_hyper(p, &args->cbl_fsid.major); @@ -275,7 +265,7 @@ __be32 decode_devicenotify_args(struct svc_rqst *rqstp, args->ndevs = 0; /* Num of device notifications */ - p = read_buf(xdr, sizeof(uint32_t)); + p = xdr_inline_decode(xdr, sizeof(uint32_t)); if (unlikely(p == NULL)) { status = htonl(NFS4ERR_BADXDR); goto out; @@ -298,7 +288,8 @@ __be32 decode_devicenotify_args(struct svc_rqst *rqstp, for (i = 0; i < n; i++) { struct cb_devicenotifyitem *dev = &args->devs[i]; - p = read_buf(xdr, (4 * sizeof(uint32_t)) + NFS4_DEVICEID4_SIZE); + p = xdr_inline_decode(xdr, (4 * sizeof(uint32_t)) + + NFS4_DEVICEID4_SIZE); if (unlikely(p == NULL)) { status = htonl(NFS4ERR_BADXDR); goto err; @@ -329,7 +320,7 @@ __be32 decode_devicenotify_args(struct svc_rqst *rqstp, p += XDR_QUADLEN(NFS4_DEVICEID4_SIZE); if (dev->cbd_layout_type == NOTIFY_DEVICEID4_CHANGE) { - p = read_buf(xdr, sizeof(uint32_t)); + p = xdr_inline_decode(xdr, sizeof(uint32_t)); if (unlikely(p == NULL)) { status = htonl(NFS4ERR_BADXDR); goto err; @@ -359,7 +350,7 @@ static __be32 decode_sessionid(struct xdr_stream *xdr, { __be32 *p; - p = read_buf(xdr, NFS4_MAX_SESSIONID_LEN); + p = xdr_inline_decode(xdr, NFS4_MAX_SESSIONID_LEN); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); @@ -379,13 +370,13 @@ static __be32 decode_rc_list(struct xdr_stream *xdr, goto out; status = htonl(NFS4ERR_RESOURCE); - p = read_buf(xdr, sizeof(uint32_t)); + p = xdr_inline_decode(xdr, sizeof(uint32_t)); if (unlikely(p == NULL)) goto out; rc_list->rcl_nrefcalls = ntohl(*p++); if (rc_list->rcl_nrefcalls) { - p = read_buf(xdr, + p = xdr_inline_decode(xdr, rc_list->rcl_nrefcalls * 2 * sizeof(uint32_t)); if (unlikely(p == NULL)) goto out; @@ -418,7 +409,7 @@ static __be32 decode_cb_sequence_args(struct svc_rqst *rqstp, if (status) return status; - p = read_buf(xdr, 5 * sizeof(uint32_t)); + p = xdr_inline_decode(xdr, 5 * sizeof(uint32_t)); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); @@ -461,7 +452,7 @@ static __be32 decode_recallany_args(struct svc_rqst *rqstp, uint32_t bitmap[2]; __be32 *p, status; - p = read_buf(xdr, 4); + p = xdr_inline_decode(xdr, 4); if (unlikely(p == NULL)) return htonl(NFS4ERR_BADXDR); args->craa_objs_to_keep = ntohl(*p++); @@ -480,7 +471,7 @@ static __be32 decode_recallslot_args(struct svc_rqst *rqstp, struct cb_recallslotargs *args = argp; __be32 *p; - p = read_buf(xdr, 4); + p = xdr_inline_decode(xdr, 4); if (unlikely(p == NULL)) return htonl(NFS4ERR_BADXDR); args->crsa_target_highest_slotid = ntohl(*p++); @@ -492,14 +483,14 @@ static __be32 decode_lockowner(struct xdr_stream *xdr, struct cb_notify_lock_arg __be32 *p; unsigned int len; - p = read_buf(xdr, 12); + p = xdr_inline_decode(xdr, 12); if (unlikely(p == NULL)) return htonl(NFS4ERR_BADXDR); p = xdr_decode_hyper(p, &args->cbnl_owner.clientid); len = be32_to_cpu(*p); - p = read_buf(xdr, len); + p = xdr_inline_decode(xdr, len); if (unlikely(p == NULL)) return htonl(NFS4ERR_BADXDR); @@ -537,7 +528,7 @@ static __be32 decode_write_response(struct xdr_stream *xdr, __be32 *p; /* skip the always zero field */ - p = read_buf(xdr, 4); + p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) goto out; p++; @@ -577,7 +568,7 @@ static __be32 decode_offload_args(struct svc_rqst *rqstp, return status; /* decode status */ - p = read_buf(xdr, 4); + p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) goto out; args->error = ntohl(*p++); diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index 350675e3ed47..76614311f957 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -79,17 +79,6 @@ static void prepare_reply_buffer(struct rpc_rqst *req, struct page **pages, xdr_inline_pages(&req->rq_rcv_buf, replen << 2, pages, base, len); } -/* - * Handle decode buffer overflows out-of-line. - */ -static void print_overflow_msg(const char *func, const struct xdr_stream *xdr) -{ - dprintk("NFS: %s prematurely hit the end of our receive buffer. " - "Remaining buffer length is %tu words.\n", - func, xdr->end - xdr->p); -} - - /* * Encode/decode NFSv2 basic data types * @@ -110,8 +99,8 @@ static int decode_nfsdata(struct xdr_stream *xdr, struct nfs_pgio_res *result) __be32 *p; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; count = be32_to_cpup(p); recvd = xdr_read_pages(xdr, count); if (unlikely(count > recvd)) @@ -125,9 +114,6 @@ out_cheating: "count %u > recvd %u\n", count, recvd); count = recvd; goto out; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -157,13 +143,10 @@ static int decode_stat(struct xdr_stream *xdr, enum nfs_stat *status) __be32 *p; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; *status = be32_to_cpup(p); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -205,14 +188,11 @@ static int decode_fhandle(struct xdr_stream *xdr, struct nfs_fh *fh) __be32 *p; p = xdr_inline_decode(xdr, NFS2_FHSIZE); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; fh->size = NFS2_FHSIZE; memcpy(fh->data, p, NFS2_FHSIZE); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -282,8 +262,8 @@ static int decode_fattr(struct xdr_stream *xdr, struct nfs_fattr *fattr) __be32 *p; p = xdr_inline_decode(xdr, NFS_fattr_sz << 2); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; fattr->valid |= NFS_ATTR_FATTR_V2; @@ -325,9 +305,6 @@ out_uid: out_gid: dprintk("NFS: returned invalid gid\n"); return -EINVAL; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -416,23 +393,20 @@ static int decode_filename_inline(struct xdr_stream *xdr, u32 count; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; count = be32_to_cpup(p); if (count > NFS3_MAXNAMLEN) goto out_nametoolong; p = xdr_inline_decode(xdr, count); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; *name = (const char *)p; *length = count; return 0; out_nametoolong: dprintk("NFS: returned filename too long: %u\n", count); return -ENAMETOOLONG; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -455,8 +429,8 @@ static int decode_path(struct xdr_stream *xdr) __be32 *p; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; length = be32_to_cpup(p); if (unlikely(length >= xdr->buf->page_len || length > NFS_MAXPATHLEN)) goto out_size; @@ -472,9 +446,6 @@ out_cheating: dprintk("NFS: server cheating in pathname result: " "length %u > received %u\n", length, recvd); return -EIO; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -951,12 +922,12 @@ int nfs2_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, int error; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EAGAIN; if (*p++ == xdr_zero) { p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EAGAIN; if (*p++ == xdr_zero) return -EAGAIN; entry->eof = 1; @@ -964,8 +935,8 @@ int nfs2_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, } p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EAGAIN; entry->ino = be32_to_cpup(p); error = decode_filename_inline(xdr, &entry->name, &entry->len); @@ -978,17 +949,13 @@ int nfs2_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, */ entry->prev_cookie = entry->cookie; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EAGAIN; entry->cookie = be32_to_cpup(p); entry->d_type = DT_UNKNOWN; return 0; - -out_overflow: - print_overflow_msg(__func__, xdr); - return -EAGAIN; } /* @@ -1052,17 +1019,14 @@ static int decode_info(struct xdr_stream *xdr, struct nfs2_fsstat *result) __be32 *p; p = xdr_inline_decode(xdr, NFS_info_sz << 2); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; result->tsize = be32_to_cpup(p++); result->bsize = be32_to_cpup(p++); result->blocks = be32_to_cpup(p++); result->bfree = be32_to_cpup(p++); result->bavail = be32_to_cpup(p); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int nfs2_xdr_dec_statfsres(struct rpc_rqst *req, struct xdr_stream *xdr, diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index 78df4eb60f85..e5619803a4f6 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -118,17 +118,6 @@ static void prepare_reply_buffer(struct rpc_rqst *req, struct page **pages, xdr_inline_pages(&req->rq_rcv_buf, replen << 2, pages, base, len); } -/* - * Handle decode buffer overflows out-of-line. - */ -static void print_overflow_msg(const char *func, const struct xdr_stream *xdr) -{ - dprintk("NFS: %s prematurely hit the end of our receive buffer. " - "Remaining buffer length is %tu words.\n", - func, xdr->end - xdr->p); -} - - /* * Encode/decode NFSv3 basic data types * @@ -151,13 +140,10 @@ static int decode_uint32(struct xdr_stream *xdr, u32 *value) __be32 *p; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; *value = be32_to_cpup(p); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_uint64(struct xdr_stream *xdr, u64 *value) @@ -165,13 +151,10 @@ static int decode_uint64(struct xdr_stream *xdr, u64 *value) __be32 *p; p = xdr_inline_decode(xdr, 8); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; xdr_decode_hyper(p, value); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -211,14 +194,14 @@ static int decode_inline_filename3(struct xdr_stream *xdr, u32 count; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; count = be32_to_cpup(p); if (count > NFS3_MAXNAMLEN) goto out_nametoolong; p = xdr_inline_decode(xdr, count); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; *name = (const char *)p; *length = count; return 0; @@ -226,9 +209,6 @@ static int decode_inline_filename3(struct xdr_stream *xdr, out_nametoolong: dprintk("NFS: returned filename too long: %u\n", count); return -ENAMETOOLONG; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -249,8 +229,8 @@ static int decode_nfspath3(struct xdr_stream *xdr) __be32 *p; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; count = be32_to_cpup(p); if (unlikely(count >= xdr->buf->page_len || count > NFS3_MAXPATHLEN)) goto out_nametoolong; @@ -267,9 +247,6 @@ out_cheating: dprintk("NFS: server cheating in pathname result: " "count %u > recvd %u\n", count, recvd); return -EIO; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -303,13 +280,10 @@ static int decode_cookieverf3(struct xdr_stream *xdr, __be32 *verifier) __be32 *p; p = xdr_inline_decode(xdr, NFS3_COOKIEVERFSIZE); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; memcpy(verifier, p, NFS3_COOKIEVERFSIZE); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -330,13 +304,10 @@ static int decode_writeverf3(struct xdr_stream *xdr, struct nfs_write_verifier * __be32 *p; p = xdr_inline_decode(xdr, NFS3_WRITEVERFSIZE); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; memcpy(verifier->data, p, NFS3_WRITEVERFSIZE); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -364,13 +335,10 @@ static int decode_nfsstat3(struct xdr_stream *xdr, enum nfs_stat *status) __be32 *p; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; *status = be32_to_cpup(p); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -453,23 +421,20 @@ static int decode_nfs_fh3(struct xdr_stream *xdr, struct nfs_fh *fh) __be32 *p; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; length = be32_to_cpup(p++); if (unlikely(length > NFS3_FHSIZE)) goto out_toobig; p = xdr_inline_decode(xdr, length); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; fh->size = length; memcpy(fh->data, p, length); return 0; out_toobig: dprintk("NFS: file handle size (%u) too big\n", length); return -E2BIG; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static void zero_nfs_fh3(struct nfs_fh *fh) @@ -655,8 +620,8 @@ static int decode_fattr3(struct xdr_stream *xdr, struct nfs_fattr *fattr) __be32 *p; p = xdr_inline_decode(xdr, NFS3_fattr_sz << 2); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; p = xdr_decode_ftype3(p, &fmode); @@ -690,9 +655,6 @@ out_uid: out_gid: dprintk("NFS: returned invalid gid\n"); return -EINVAL; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -710,14 +672,11 @@ static int decode_post_op_attr(struct xdr_stream *xdr, struct nfs_fattr *fattr) __be32 *p; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; if (*p != xdr_zero) return decode_fattr3(xdr, fattr); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -733,8 +692,8 @@ static int decode_wcc_attr(struct xdr_stream *xdr, struct nfs_fattr *fattr) __be32 *p; p = xdr_inline_decode(xdr, NFS3_wcc_attr_sz << 2); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; fattr->valid |= NFS_ATTR_FATTR_PRESIZE | NFS_ATTR_FATTR_PRECHANGE @@ -747,9 +706,6 @@ static int decode_wcc_attr(struct xdr_stream *xdr, struct nfs_fattr *fattr) fattr->pre_change_attr = nfs_timespec_to_change_attr(&fattr->pre_ctime); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -773,14 +729,11 @@ static int decode_pre_op_attr(struct xdr_stream *xdr, struct nfs_fattr *fattr) __be32 *p; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; if (*p != xdr_zero) return decode_wcc_attr(xdr, fattr); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_wcc_data(struct xdr_stream *xdr, struct nfs_fattr *fattr) @@ -808,15 +761,12 @@ out: static int decode_post_op_fh3(struct xdr_stream *xdr, struct nfs_fh *fh) { __be32 *p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; if (*p != xdr_zero) return decode_nfs_fh3(xdr, fh); zero_nfs_fh3(fh); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -1643,8 +1593,8 @@ static int decode_read3resok(struct xdr_stream *xdr, __be32 *p; p = xdr_inline_decode(xdr, 4 + 4 + 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; count = be32_to_cpup(p++); eof = be32_to_cpup(p++); ocount = be32_to_cpup(p++); @@ -1667,9 +1617,6 @@ out_cheating: count = recvd; eof = 0; goto out; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int nfs3_xdr_dec_read3res(struct rpc_rqst *req, struct xdr_stream *xdr, @@ -1731,22 +1678,18 @@ static int decode_write3resok(struct xdr_stream *xdr, __be32 *p; p = xdr_inline_decode(xdr, 4 + 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; result->count = be32_to_cpup(p++); result->verf->committed = be32_to_cpup(p++); if (unlikely(result->verf->committed > NFS_FILE_SYNC)) goto out_badvalue; if (decode_writeverf3(xdr, &result->verf->verifier)) - goto out_eio; + return -EIO; return result->count; out_badvalue: dprintk("NFS: bad stable_how value: %u\n", result->verf->committed); return -EIO; -out_overflow: - print_overflow_msg(__func__, xdr); -out_eio: - return -EIO; } static int nfs3_xdr_dec_write3res(struct rpc_rqst *req, struct xdr_stream *xdr, @@ -2010,12 +1953,12 @@ int nfs3_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, u64 new_cookie; p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EAGAIN; if (*p == xdr_zero) { p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EAGAIN; if (*p == xdr_zero) return -EAGAIN; entry->eof = 1; @@ -2051,8 +1994,8 @@ int nfs3_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, /* In fact, a post_op_fh3: */ p = xdr_inline_decode(xdr, 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EAGAIN; if (*p != xdr_zero) { error = decode_nfs_fh3(xdr, entry->fh); if (unlikely(error)) { @@ -2069,9 +2012,6 @@ int nfs3_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EAGAIN; out_truncated: dprintk("NFS: directory entry contains invalid file handle\n"); *entry = old; @@ -2183,8 +2123,8 @@ static int decode_fsstat3resok(struct xdr_stream *xdr, __be32 *p; p = xdr_inline_decode(xdr, 8 * 6 + 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; p = xdr_decode_size3(p, &result->tbytes); p = xdr_decode_size3(p, &result->fbytes); p = xdr_decode_size3(p, &result->abytes); @@ -2193,9 +2133,6 @@ static int decode_fsstat3resok(struct xdr_stream *xdr, xdr_decode_size3(p, &result->afiles); /* ignore invarsec */ return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int nfs3_xdr_dec_fsstat3res(struct rpc_rqst *req, @@ -2255,8 +2192,8 @@ static int decode_fsinfo3resok(struct xdr_stream *xdr, __be32 *p; p = xdr_inline_decode(xdr, 4 * 7 + 8 + 8 + 4); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; result->rtmax = be32_to_cpup(p++); result->rtpref = be32_to_cpup(p++); result->rtmult = be32_to_cpup(p++); @@ -2270,9 +2207,6 @@ static int decode_fsinfo3resok(struct xdr_stream *xdr, /* ignore properties */ result->lease_time = 0; return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int nfs3_xdr_dec_fsinfo3res(struct rpc_rqst *req, @@ -2328,15 +2262,12 @@ static int decode_pathconf3resok(struct xdr_stream *xdr, __be32 *p; p = xdr_inline_decode(xdr, 4 * 6); - if (unlikely(p == NULL)) - goto out_overflow; + if (unlikely(!p)) + return -EIO; result->max_link = be32_to_cpup(p++); result->max_namelen = be32_to_cpup(p); /* ignore remaining fields */ return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int nfs3_xdr_dec_pathconf3res(struct rpc_rqst *req, diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c index 69f72ed2bf87..22b3425df08f 100644 --- a/fs/nfs/nfs42xdr.c +++ b/fs/nfs/nfs42xdr.c @@ -394,7 +394,7 @@ static int decode_write_response(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; count = be32_to_cpup(p); if (count > 1) return -EREMOTEIO; @@ -402,18 +402,14 @@ static int decode_write_response(struct xdr_stream *xdr, status = decode_opaque_fixed(xdr, &res->stateid, NFS4_STATEID_SIZE); if (unlikely(status)) - goto out_overflow; + return -EIO; } p = xdr_inline_decode(xdr, 8 + 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; p = xdr_decode_hyper(p, &res->count); res->verifier.committed = be32_to_cpup(p); return decode_verifier(xdr, &res->verifier.verifier); - -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_copy_requirements(struct xdr_stream *xdr, @@ -422,14 +418,11 @@ static int decode_copy_requirements(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 4 + 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->consecutive = be32_to_cpup(p++); res->synchronous = be32_to_cpup(p++); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_copy(struct xdr_stream *xdr, struct nfs42_copy_res *res) @@ -474,15 +467,11 @@ static int decode_seek(struct xdr_stream *xdr, struct nfs42_seek_res *res) p = xdr_inline_decode(xdr, 4 + 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->sr_eof = be32_to_cpup(p++); p = xdr_decode_hyper(p, &res->sr_offset); return 0; - -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_layoutstats(struct xdr_stream *xdr) diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 2fc8f6fa25e4..24e6a45a5726 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -3144,22 +3144,12 @@ static void nfs4_xdr_enc_free_stateid(struct rpc_rqst *req, } #endif /* CONFIG_NFS_V4_1 */ -static void print_overflow_msg(const char *func, const struct xdr_stream *xdr) -{ - dprintk("nfs: %s: prematurely hit end of receive buffer. " - "Remaining buffer length is %tu words.\n", - func, xdr->end - xdr->p); -} - static int decode_opaque_inline(struct xdr_stream *xdr, unsigned int *len, char **string) { ssize_t ret = xdr_stream_decode_opaque_inline(xdr, (void **)string, NFS4_OPAQUE_LIMIT); - if (unlikely(ret < 0)) { - if (ret == -EBADMSG) - print_overflow_msg(__func__, xdr); + if (unlikely(ret < 0)) return -EIO; - } *len = ret; return 0; } @@ -3170,22 +3160,19 @@ static int decode_compound_hdr(struct xdr_stream *xdr, struct compound_hdr *hdr) p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; hdr->status = be32_to_cpup(p++); hdr->taglen = be32_to_cpup(p); p = xdr_inline_decode(xdr, hdr->taglen + 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; hdr->tag = (char *)p; p += XDR_QUADLEN(hdr->taglen); hdr->nops = be32_to_cpup(p); if (unlikely(hdr->nops < 1)) return nfs4_stat_to_errno(hdr->status); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static bool __decode_op_hdr(struct xdr_stream *xdr, enum nfs_opnum4 expected, @@ -3214,7 +3201,6 @@ out_bad_operation: *nfs_retval = -EREMOTEIO; return false; out_overflow: - print_overflow_msg(__func__, xdr); *nfs_retval = -EIO; return false; } @@ -3235,10 +3221,9 @@ static int decode_ace(struct xdr_stream *xdr, void *ace) char *str; p = xdr_inline_decode(xdr, 12); - if (likely(p)) - return decode_opaque_inline(xdr, &strlen, &str); - print_overflow_msg(__func__, xdr); - return -EIO; + if (unlikely(!p)) + return -EIO; + return decode_opaque_inline(xdr, &strlen, &str); } static ssize_t @@ -3249,10 +3234,9 @@ decode_bitmap4(struct xdr_stream *xdr, uint32_t *bitmap, size_t sz) ret = xdr_stream_decode_uint32_array(xdr, bitmap, sz); if (likely(ret >= 0)) return ret; - if (ret == -EMSGSIZE) - return sz; - print_overflow_msg(__func__, xdr); - return -EIO; + if (ret != -EMSGSIZE) + return -EIO; + return sz; } static int decode_attr_bitmap(struct xdr_stream *xdr, uint32_t *bitmap) @@ -3268,13 +3252,10 @@ static int decode_attr_length(struct xdr_stream *xdr, uint32_t *attrlen, unsigne p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; *attrlen = be32_to_cpup(p); *savep = xdr_stream_pos(xdr); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_supported(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t *bitmask) @@ -3303,7 +3284,7 @@ static int decode_attr_type(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t * if (likely(bitmap[0] & FATTR4_WORD0_TYPE)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; *type = be32_to_cpup(p); if (*type < NF4REG || *type > NF4NAMEDATTR) { dprintk("%s: bad type %d\n", __func__, *type); @@ -3314,9 +3295,6 @@ static int decode_attr_type(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t * } dprintk("%s: type=0%o\n", __func__, nfs_type2fmt[*type]); return ret; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_fh_expire_type(struct xdr_stream *xdr, @@ -3330,15 +3308,12 @@ static int decode_attr_fh_expire_type(struct xdr_stream *xdr, if (likely(bitmap[0] & FATTR4_WORD0_FH_EXPIRE_TYPE)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; *type = be32_to_cpup(p); bitmap[0] &= ~FATTR4_WORD0_FH_EXPIRE_TYPE; } dprintk("%s: expire type=0x%x\n", __func__, *type); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_change(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *change) @@ -3352,7 +3327,7 @@ static int decode_attr_change(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t if (likely(bitmap[0] & FATTR4_WORD0_CHANGE)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, change); bitmap[0] &= ~FATTR4_WORD0_CHANGE; ret = NFS_ATTR_FATTR_CHANGE; @@ -3360,9 +3335,6 @@ static int decode_attr_change(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t dprintk("%s: change attribute=%Lu\n", __func__, (unsigned long long)*change); return ret; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_size(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *size) @@ -3376,16 +3348,13 @@ static int decode_attr_size(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t * if (likely(bitmap[0] & FATTR4_WORD0_SIZE)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, size); bitmap[0] &= ~FATTR4_WORD0_SIZE; ret = NFS_ATTR_FATTR_SIZE; } dprintk("%s: file size=%Lu\n", __func__, (unsigned long long)*size); return ret; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_link_support(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t *res) @@ -3398,15 +3367,12 @@ static int decode_attr_link_support(struct xdr_stream *xdr, uint32_t *bitmap, ui if (likely(bitmap[0] & FATTR4_WORD0_LINK_SUPPORT)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; *res = be32_to_cpup(p); bitmap[0] &= ~FATTR4_WORD0_LINK_SUPPORT; } dprintk("%s: link support=%s\n", __func__, *res == 0 ? "false" : "true"); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_symlink_support(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t *res) @@ -3419,15 +3385,12 @@ static int decode_attr_symlink_support(struct xdr_stream *xdr, uint32_t *bitmap, if (likely(bitmap[0] & FATTR4_WORD0_SYMLINK_SUPPORT)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; *res = be32_to_cpup(p); bitmap[0] &= ~FATTR4_WORD0_SYMLINK_SUPPORT; } dprintk("%s: symlink support=%s\n", __func__, *res == 0 ? "false" : "true"); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_fsid(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs_fsid *fsid) @@ -3442,7 +3405,7 @@ static int decode_attr_fsid(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs if (likely(bitmap[0] & FATTR4_WORD0_FSID)) { p = xdr_inline_decode(xdr, 16); if (unlikely(!p)) - goto out_overflow; + return -EIO; p = xdr_decode_hyper(p, &fsid->major); xdr_decode_hyper(p, &fsid->minor); bitmap[0] &= ~FATTR4_WORD0_FSID; @@ -3452,9 +3415,6 @@ static int decode_attr_fsid(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs (unsigned long long)fsid->major, (unsigned long long)fsid->minor); return ret; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_lease_time(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t *res) @@ -3467,15 +3427,12 @@ static int decode_attr_lease_time(struct xdr_stream *xdr, uint32_t *bitmap, uint if (likely(bitmap[0] & FATTR4_WORD0_LEASE_TIME)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; *res = be32_to_cpup(p); bitmap[0] &= ~FATTR4_WORD0_LEASE_TIME; } dprintk("%s: file size=%u\n", __func__, (unsigned int)*res); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_error(struct xdr_stream *xdr, uint32_t *bitmap, int32_t *res) @@ -3487,14 +3444,11 @@ static int decode_attr_error(struct xdr_stream *xdr, uint32_t *bitmap, int32_t * if (likely(bitmap[0] & FATTR4_WORD0_RDATTR_ERROR)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; bitmap[0] &= ~FATTR4_WORD0_RDATTR_ERROR; *res = -be32_to_cpup(p); } return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_exclcreat_supported(struct xdr_stream *xdr, @@ -3526,13 +3480,13 @@ static int decode_attr_filehandle(struct xdr_stream *xdr, uint32_t *bitmap, stru if (likely(bitmap[0] & FATTR4_WORD0_FILEHANDLE)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; len = be32_to_cpup(p); if (len > NFS4_FHSIZE) return -EIO; p = xdr_inline_decode(xdr, len); if (unlikely(!p)) - goto out_overflow; + return -EIO; if (fh != NULL) { memcpy(fh->data, p, len); fh->size = len; @@ -3540,9 +3494,6 @@ static int decode_attr_filehandle(struct xdr_stream *xdr, uint32_t *bitmap, stru bitmap[0] &= ~FATTR4_WORD0_FILEHANDLE; } return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_aclsupport(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t *res) @@ -3555,15 +3506,12 @@ static int decode_attr_aclsupport(struct xdr_stream *xdr, uint32_t *bitmap, uint if (likely(bitmap[0] & FATTR4_WORD0_ACLSUPPORT)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; *res = be32_to_cpup(p); bitmap[0] &= ~FATTR4_WORD0_ACLSUPPORT; } dprintk("%s: ACLs supported=%u\n", __func__, (unsigned int)*res); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_fileid(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *fileid) @@ -3577,16 +3525,13 @@ static int decode_attr_fileid(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t if (likely(bitmap[0] & FATTR4_WORD0_FILEID)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, fileid); bitmap[0] &= ~FATTR4_WORD0_FILEID; ret = NFS_ATTR_FATTR_FILEID; } dprintk("%s: fileid=%Lu\n", __func__, (unsigned long long)*fileid); return ret; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_mounted_on_fileid(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *fileid) @@ -3600,16 +3545,13 @@ static int decode_attr_mounted_on_fileid(struct xdr_stream *xdr, uint32_t *bitma if (likely(bitmap[1] & FATTR4_WORD1_MOUNTED_ON_FILEID)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, fileid); bitmap[1] &= ~FATTR4_WORD1_MOUNTED_ON_FILEID; ret = NFS_ATTR_FATTR_MOUNTED_ON_FILEID; } dprintk("%s: fileid=%Lu\n", __func__, (unsigned long long)*fileid); return ret; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_files_avail(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *res) @@ -3623,15 +3565,12 @@ static int decode_attr_files_avail(struct xdr_stream *xdr, uint32_t *bitmap, uin if (likely(bitmap[0] & FATTR4_WORD0_FILES_AVAIL)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, res); bitmap[0] &= ~FATTR4_WORD0_FILES_AVAIL; } dprintk("%s: files avail=%Lu\n", __func__, (unsigned long long)*res); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_files_free(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *res) @@ -3645,15 +3584,12 @@ static int decode_attr_files_free(struct xdr_stream *xdr, uint32_t *bitmap, uint if (likely(bitmap[0] & FATTR4_WORD0_FILES_FREE)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, res); bitmap[0] &= ~FATTR4_WORD0_FILES_FREE; } dprintk("%s: files free=%Lu\n", __func__, (unsigned long long)*res); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_files_total(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *res) @@ -3667,15 +3603,12 @@ static int decode_attr_files_total(struct xdr_stream *xdr, uint32_t *bitmap, uin if (likely(bitmap[0] & FATTR4_WORD0_FILES_TOTAL)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, res); bitmap[0] &= ~FATTR4_WORD0_FILES_TOTAL; } dprintk("%s: files total=%Lu\n", __func__, (unsigned long long)*res); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_pathname(struct xdr_stream *xdr, struct nfs4_pathname *path) @@ -3686,7 +3619,7 @@ static int decode_pathname(struct xdr_stream *xdr, struct nfs4_pathname *path) p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; n = be32_to_cpup(p); if (n == 0) goto root_path; @@ -3718,9 +3651,6 @@ out_eio: dprintk(" status %d", status); status = -EIO; goto out; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_fs_locations(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs4_fs_locations *res) @@ -3745,7 +3675,7 @@ static int decode_attr_fs_locations(struct xdr_stream *xdr, uint32_t *bitmap, st goto out; p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + goto out_eio; n = be32_to_cpup(p); if (n <= 0) goto out_eio; @@ -3758,7 +3688,7 @@ static int decode_attr_fs_locations(struct xdr_stream *xdr, uint32_t *bitmap, st loc = &res->locations[res->nlocations]; p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + goto out_eio; m = be32_to_cpup(p); dprintk("%s: servers:\n", __func__); @@ -3796,8 +3726,6 @@ static int decode_attr_fs_locations(struct xdr_stream *xdr, uint32_t *bitmap, st out: dprintk("%s: fs_locations done, error = %d\n", __func__, status); return status; -out_overflow: - print_overflow_msg(__func__, xdr); out_eio: status = -EIO; goto out; @@ -3814,15 +3742,12 @@ static int decode_attr_maxfilesize(struct xdr_stream *xdr, uint32_t *bitmap, uin if (likely(bitmap[0] & FATTR4_WORD0_MAXFILESIZE)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, res); bitmap[0] &= ~FATTR4_WORD0_MAXFILESIZE; } dprintk("%s: maxfilesize=%Lu\n", __func__, (unsigned long long)*res); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_maxlink(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t *maxlink) @@ -3836,15 +3761,12 @@ static int decode_attr_maxlink(struct xdr_stream *xdr, uint32_t *bitmap, uint32_ if (likely(bitmap[0] & FATTR4_WORD0_MAXLINK)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; *maxlink = be32_to_cpup(p); bitmap[0] &= ~FATTR4_WORD0_MAXLINK; } dprintk("%s: maxlink=%u\n", __func__, *maxlink); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_maxname(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t *maxname) @@ -3858,15 +3780,12 @@ static int decode_attr_maxname(struct xdr_stream *xdr, uint32_t *bitmap, uint32_ if (likely(bitmap[0] & FATTR4_WORD0_MAXNAME)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; *maxname = be32_to_cpup(p); bitmap[0] &= ~FATTR4_WORD0_MAXNAME; } dprintk("%s: maxname=%u\n", __func__, *maxname); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_maxread(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t *res) @@ -3881,7 +3800,7 @@ static int decode_attr_maxread(struct xdr_stream *xdr, uint32_t *bitmap, uint32_ uint64_t maxread; p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, &maxread); if (maxread > 0x7FFFFFFF) maxread = 0x7FFFFFFF; @@ -3890,9 +3809,6 @@ static int decode_attr_maxread(struct xdr_stream *xdr, uint32_t *bitmap, uint32_ } dprintk("%s: maxread=%lu\n", __func__, (unsigned long)*res); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_maxwrite(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t *res) @@ -3907,7 +3823,7 @@ static int decode_attr_maxwrite(struct xdr_stream *xdr, uint32_t *bitmap, uint32 uint64_t maxwrite; p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, &maxwrite); if (maxwrite > 0x7FFFFFFF) maxwrite = 0x7FFFFFFF; @@ -3916,9 +3832,6 @@ static int decode_attr_maxwrite(struct xdr_stream *xdr, uint32_t *bitmap, uint32 } dprintk("%s: maxwrite=%lu\n", __func__, (unsigned long)*res); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_mode(struct xdr_stream *xdr, uint32_t *bitmap, umode_t *mode) @@ -3933,7 +3846,7 @@ static int decode_attr_mode(struct xdr_stream *xdr, uint32_t *bitmap, umode_t *m if (likely(bitmap[1] & FATTR4_WORD1_MODE)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; tmp = be32_to_cpup(p); *mode = tmp & ~S_IFMT; bitmap[1] &= ~FATTR4_WORD1_MODE; @@ -3941,9 +3854,6 @@ static int decode_attr_mode(struct xdr_stream *xdr, uint32_t *bitmap, umode_t *m } dprintk("%s: file mode=0%o\n", __func__, (unsigned int)*mode); return ret; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_nlink(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t *nlink) @@ -3957,16 +3867,13 @@ static int decode_attr_nlink(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t if (likely(bitmap[1] & FATTR4_WORD1_NUMLINKS)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; *nlink = be32_to_cpup(p); bitmap[1] &= ~FATTR4_WORD1_NUMLINKS; ret = NFS_ATTR_FATTR_NLINK; } dprintk("%s: nlink=%u\n", __func__, (unsigned int)*nlink); return ret; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static ssize_t decode_nfs4_string(struct xdr_stream *xdr, @@ -4011,10 +3918,9 @@ static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, return NFS_ATTR_FATTR_OWNER; } out: - if (len != -EBADMSG) - return 0; - print_overflow_msg(__func__, xdr); - return -EIO; + if (len == -EBADMSG) + return -EIO; + return 0; } static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, @@ -4046,10 +3952,9 @@ static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, return NFS_ATTR_FATTR_GROUP; } out: - if (len != -EBADMSG) - return 0; - print_overflow_msg(__func__, xdr); - return -EIO; + if (len == -EBADMSG) + return -EIO; + return 0; } static int decode_attr_rdev(struct xdr_stream *xdr, uint32_t *bitmap, dev_t *rdev) @@ -4066,7 +3971,7 @@ static int decode_attr_rdev(struct xdr_stream *xdr, uint32_t *bitmap, dev_t *rde p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; major = be32_to_cpup(p++); minor = be32_to_cpup(p); tmp = MKDEV(major, minor); @@ -4077,9 +3982,6 @@ static int decode_attr_rdev(struct xdr_stream *xdr, uint32_t *bitmap, dev_t *rde } dprintk("%s: rdev=(0x%x:0x%x)\n", __func__, major, minor); return ret; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_space_avail(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *res) @@ -4093,15 +3995,12 @@ static int decode_attr_space_avail(struct xdr_stream *xdr, uint32_t *bitmap, uin if (likely(bitmap[1] & FATTR4_WORD1_SPACE_AVAIL)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, res); bitmap[1] &= ~FATTR4_WORD1_SPACE_AVAIL; } dprintk("%s: space avail=%Lu\n", __func__, (unsigned long long)*res); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_space_free(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *res) @@ -4115,15 +4014,12 @@ static int decode_attr_space_free(struct xdr_stream *xdr, uint32_t *bitmap, uint if (likely(bitmap[1] & FATTR4_WORD1_SPACE_FREE)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, res); bitmap[1] &= ~FATTR4_WORD1_SPACE_FREE; } dprintk("%s: space free=%Lu\n", __func__, (unsigned long long)*res); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_space_total(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *res) @@ -4137,15 +4033,12 @@ static int decode_attr_space_total(struct xdr_stream *xdr, uint32_t *bitmap, uin if (likely(bitmap[1] & FATTR4_WORD1_SPACE_TOTAL)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, res); bitmap[1] &= ~FATTR4_WORD1_SPACE_TOTAL; } dprintk("%s: space total=%Lu\n", __func__, (unsigned long long)*res); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_space_used(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *used) @@ -4159,7 +4052,7 @@ static int decode_attr_space_used(struct xdr_stream *xdr, uint32_t *bitmap, uint if (likely(bitmap[1] & FATTR4_WORD1_SPACE_USED)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, used); bitmap[1] &= ~FATTR4_WORD1_SPACE_USED; ret = NFS_ATTR_FATTR_SPACE_USED; @@ -4167,9 +4060,6 @@ static int decode_attr_space_used(struct xdr_stream *xdr, uint32_t *bitmap, uint dprintk("%s: space used=%Lu\n", __func__, (unsigned long long)*used); return ret; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static __be32 * @@ -4189,12 +4079,9 @@ static int decode_attr_time(struct xdr_stream *xdr, struct timespec *time) p = xdr_inline_decode(xdr, nfstime4_maxsz << 2); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_nfstime4(p, time); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_time_access(struct xdr_stream *xdr, uint32_t *bitmap, struct timespec *time) @@ -4265,19 +4152,19 @@ static int decode_attr_security_label(struct xdr_stream *xdr, uint32_t *bitmap, if (likely(bitmap[2] & FATTR4_WORD2_SECURITY_LABEL)) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; lfs = be32_to_cpup(p++); p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; pi = be32_to_cpup(p++); p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; len = be32_to_cpup(p++); p = xdr_inline_decode(xdr, len); if (unlikely(!p)) - goto out_overflow; + return -EIO; if (len < NFS4_MAXLABELLEN) { if (label) { memcpy(label->label, p, len); @@ -4295,10 +4182,6 @@ static int decode_attr_security_label(struct xdr_stream *xdr, uint32_t *bitmap, dprintk("%s: label=%s, len=%d, PI=%d, LFS=%d\n", __func__, (char *)label->label, label->len, label->pi, label->lfs); return status; - -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_attr_time_modify(struct xdr_stream *xdr, uint32_t *bitmap, struct timespec *time) @@ -4342,14 +4225,11 @@ static int decode_change_info(struct xdr_stream *xdr, struct nfs4_change_info *c p = xdr_inline_decode(xdr, 20); if (unlikely(!p)) - goto out_overflow; + return -EIO; cinfo->atomic = be32_to_cpup(p++); p = xdr_decode_hyper(p, &cinfo->before); xdr_decode_hyper(p, &cinfo->after); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_access(struct xdr_stream *xdr, u32 *supported, u32 *access) @@ -4363,24 +4243,19 @@ static int decode_access(struct xdr_stream *xdr, u32 *supported, u32 *access) return status; p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; supp = be32_to_cpup(p++); acc = be32_to_cpup(p); *supported = supp; *access = acc; return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_opaque_fixed(struct xdr_stream *xdr, void *buf, size_t len) { ssize_t ret = xdr_stream_decode_opaque_fixed(xdr, buf, len); - if (unlikely(ret < 0)) { - print_overflow_msg(__func__, xdr); + if (unlikely(ret < 0)) return -EIO; - } return 0; } @@ -4460,13 +4335,11 @@ static int decode_create(struct xdr_stream *xdr, struct nfs4_change_info *cinfo) return status; p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; bmlen = be32_to_cpup(p); p = xdr_inline_decode(xdr, bmlen << 2); if (likely(p)) return 0; -out_overflow: - print_overflow_msg(__func__, xdr); return -EIO; } @@ -4574,13 +4447,10 @@ static int decode_threshold_hint(struct xdr_stream *xdr, if (likely(bitmap[0] & hint_bit)) { p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, res); } return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_first_threshold_item4(struct xdr_stream *xdr, @@ -4593,10 +4463,8 @@ static int decode_first_threshold_item4(struct xdr_stream *xdr, /* layout type */ p = xdr_inline_decode(xdr, 4); - if (unlikely(!p)) { - print_overflow_msg(__func__, xdr); + if (unlikely(!p)) return -EIO; - } res->l_type = be32_to_cpup(p); /* thi_hintset bitmap */ @@ -4654,7 +4522,7 @@ static int decode_attr_mdsthreshold(struct xdr_stream *xdr, return -EREMOTEIO; p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; num = be32_to_cpup(p); if (num == 0) return 0; @@ -4667,9 +4535,6 @@ static int decode_attr_mdsthreshold(struct xdr_stream *xdr, bitmap[2] &= ~FATTR4_WORD2_MDSTHRESHOLD; } return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_getfattr_attrs(struct xdr_stream *xdr, uint32_t *bitmap, @@ -4857,7 +4722,7 @@ static int decode_pnfs_layout_types(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; fsinfo->nlayouttypes = be32_to_cpup(p); /* pNFS is not supported by the underlying file system */ @@ -4867,7 +4732,7 @@ static int decode_pnfs_layout_types(struct xdr_stream *xdr, /* Decode and set first layout type, move xdr->p past unused types */ p = xdr_inline_decode(xdr, fsinfo->nlayouttypes * 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; /* If we get too many, then just cap it at the max */ if (fsinfo->nlayouttypes > NFS_MAX_LAYOUT_TYPES) { @@ -4879,9 +4744,6 @@ static int decode_pnfs_layout_types(struct xdr_stream *xdr, for(i = 0; i < fsinfo->nlayouttypes; ++i) fsinfo->layouttype[i] = be32_to_cpup(p++); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } /* @@ -4915,10 +4777,8 @@ static int decode_attr_layout_blksize(struct xdr_stream *xdr, uint32_t *bitmap, *res = 0; if (bitmap[2] & FATTR4_WORD2_LAYOUT_BLKSIZE) { p = xdr_inline_decode(xdr, 4); - if (unlikely(!p)) { - print_overflow_msg(__func__, xdr); + if (unlikely(!p)) return -EIO; - } *res = be32_to_cpup(p); bitmap[2] &= ~FATTR4_WORD2_LAYOUT_BLKSIZE; } @@ -4937,10 +4797,8 @@ static int decode_attr_clone_blksize(struct xdr_stream *xdr, uint32_t *bitmap, *res = 0; if (bitmap[2] & FATTR4_WORD2_CLONE_BLKSIZE) { p = xdr_inline_decode(xdr, 4); - if (unlikely(!p)) { - print_overflow_msg(__func__, xdr); + if (unlikely(!p)) return -EIO; - } *res = be32_to_cpup(p); bitmap[2] &= ~FATTR4_WORD2_CLONE_BLKSIZE; } @@ -5016,19 +4874,16 @@ static int decode_getfh(struct xdr_stream *xdr, struct nfs_fh *fh) p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; len = be32_to_cpup(p); if (len > NFS4_FHSIZE) return -EIO; fh->size = len; p = xdr_inline_decode(xdr, len); if (unlikely(!p)) - goto out_overflow; + return -EIO; memcpy(fh->data, p, len); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_link(struct xdr_stream *xdr, struct nfs4_change_info *cinfo) @@ -5052,7 +4907,7 @@ static int decode_lock_denied (struct xdr_stream *xdr, struct file_lock *fl) p = xdr_inline_decode(xdr, 32); /* read 32 bytes */ if (unlikely(!p)) - goto out_overflow; + return -EIO; p = xdr_decode_hyper(p, &offset); /* read 2 8-byte long words */ p = xdr_decode_hyper(p, &length); type = be32_to_cpup(p++); /* 4 byte read */ @@ -5069,11 +4924,9 @@ static int decode_lock_denied (struct xdr_stream *xdr, struct file_lock *fl) p = xdr_decode_hyper(p, &clientid); /* read 8 bytes */ namelen = be32_to_cpup(p); /* read 4 bytes */ /* have read all 32 bytes now */ p = xdr_inline_decode(xdr, namelen); /* variable size field */ - if (likely(p)) - return -NFS4ERR_DENIED; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; + if (likely(!p)) + return -EIO; + return -NFS4ERR_DENIED; } static int decode_lock(struct xdr_stream *xdr, struct nfs_lock_res *res) @@ -5142,7 +4995,7 @@ static int decode_space_limit(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 12); if (unlikely(!p)) - goto out_overflow; + return -EIO; limit_type = be32_to_cpup(p++); switch (limit_type) { case NFS4_LIMIT_SIZE: @@ -5156,9 +5009,6 @@ static int decode_space_limit(struct xdr_stream *xdr, maxsize >>= PAGE_SHIFT; *pagemod_limit = min_t(u64, maxsize, ULONG_MAX); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_rw_delegation(struct xdr_stream *xdr, @@ -5173,7 +5023,7 @@ static int decode_rw_delegation(struct xdr_stream *xdr, return status; p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->do_recall = be32_to_cpup(p); switch (delegation_type) { @@ -5186,9 +5036,6 @@ static int decode_rw_delegation(struct xdr_stream *xdr, return -EIO; } return decode_ace(xdr, NULL); -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_no_delegation(struct xdr_stream *xdr, struct nfs_openres *res) @@ -5198,7 +5045,7 @@ static int decode_no_delegation(struct xdr_stream *xdr, struct nfs_openres *res) p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; why_no_delegation = be32_to_cpup(p); switch (why_no_delegation) { case WND4_CONTENTION: @@ -5207,9 +5054,6 @@ static int decode_no_delegation(struct xdr_stream *xdr, struct nfs_openres *res) /* Ignore for now */ } return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res) @@ -5219,7 +5063,7 @@ static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res) p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; delegation_type = be32_to_cpup(p); res->delegation_type = 0; switch (delegation_type) { @@ -5232,9 +5076,6 @@ static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res) return decode_no_delegation(xdr, res); } return -EIO; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res) @@ -5256,7 +5097,7 @@ static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res) p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->rflags = be32_to_cpup(p++); bmlen = be32_to_cpup(p); if (bmlen > 10) @@ -5264,7 +5105,7 @@ static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res) p = xdr_inline_decode(xdr, bmlen << 2); if (unlikely(!p)) - goto out_overflow; + return -EIO; savewords = min_t(uint32_t, bmlen, NFS4_BITMAP_SIZE); for (i = 0; i < savewords; ++i) res->attrset[i] = be32_to_cpup(p++); @@ -5275,9 +5116,6 @@ static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res) xdr_error: dprintk("%s: Bitmap too large! Length = %u\n", __func__, bmlen); return -EIO; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_open_confirm(struct xdr_stream *xdr, struct nfs_open_confirmres *res) @@ -5326,7 +5164,7 @@ static int decode_read(struct xdr_stream *xdr, struct rpc_rqst *req, return status; p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; eof = be32_to_cpup(p++); count = be32_to_cpup(p); recvd = xdr_read_pages(xdr, count); @@ -5339,9 +5177,6 @@ static int decode_read(struct xdr_stream *xdr, struct rpc_rqst *req, res->eof = eof; res->count = count; return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_readdir(struct xdr_stream *xdr, struct rpc_rqst *req, struct nfs4_readdir_res *readdir) @@ -5374,7 +5209,7 @@ static int decode_readlink(struct xdr_stream *xdr, struct rpc_rqst *req) /* Convert length of symlink */ p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; len = be32_to_cpup(p); if (len >= rcvbuf->page_len || len <= 0) { dprintk("nfs: server returned giant symlink!\n"); @@ -5395,9 +5230,6 @@ static int decode_readlink(struct xdr_stream *xdr, struct rpc_rqst *req) */ xdr_terminate_string(rcvbuf, len); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_remove(struct xdr_stream *xdr, struct nfs4_change_info *cinfo) @@ -5500,7 +5332,6 @@ static int decode_setattr(struct xdr_stream *xdr) return status; if (decode_bitmap4(xdr, NULL, 0) >= 0) return 0; - print_overflow_msg(__func__, xdr); return -EIO; } @@ -5512,7 +5343,7 @@ static int decode_setclientid(struct xdr_stream *xdr, struct nfs4_setclientid_re p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; opnum = be32_to_cpup(p++); if (opnum != OP_SETCLIENTID) { dprintk("nfs: decode_setclientid: Server returned operation" @@ -5523,7 +5354,7 @@ static int decode_setclientid(struct xdr_stream *xdr, struct nfs4_setclientid_re if (nfserr == NFS_OK) { p = xdr_inline_decode(xdr, 8 + NFS4_VERIFIER_SIZE); if (unlikely(!p)) - goto out_overflow; + return -EIO; p = xdr_decode_hyper(p, &res->clientid); memcpy(res->confirm.data, p, NFS4_VERIFIER_SIZE); } else if (nfserr == NFSERR_CLID_INUSE) { @@ -5532,28 +5363,25 @@ static int decode_setclientid(struct xdr_stream *xdr, struct nfs4_setclientid_re /* skip netid string */ p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; len = be32_to_cpup(p); p = xdr_inline_decode(xdr, len); if (unlikely(!p)) - goto out_overflow; + return -EIO; /* skip uaddr string */ p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; len = be32_to_cpup(p); p = xdr_inline_decode(xdr, len); if (unlikely(!p)) - goto out_overflow; + return -EIO; return -NFSERR_CLID_INUSE; } else return nfs4_stat_to_errno(nfserr); return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_setclientid_confirm(struct xdr_stream *xdr) @@ -5572,13 +5400,10 @@ static int decode_write(struct xdr_stream *xdr, struct nfs_pgio_res *res) p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->count = be32_to_cpup(p++); res->verf->committed = be32_to_cpup(p++); return decode_write_verifier(xdr, &res->verf->verifier); -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_delegreturn(struct xdr_stream *xdr) @@ -5594,30 +5419,24 @@ static int decode_secinfo_gss(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; oid_len = be32_to_cpup(p); if (oid_len > GSS_OID_MAX_LEN) - goto out_err; + return -EINVAL; p = xdr_inline_decode(xdr, oid_len); if (unlikely(!p)) - goto out_overflow; + return -EIO; memcpy(flavor->flavor_info.oid.data, p, oid_len); flavor->flavor_info.oid.len = oid_len; p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; flavor->flavor_info.qop = be32_to_cpup(p++); flavor->flavor_info.service = be32_to_cpup(p); return 0; - -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; -out_err: - return -EINVAL; } static int decode_secinfo_common(struct xdr_stream *xdr, struct nfs4_secinfo_res *res) @@ -5629,7 +5448,7 @@ static int decode_secinfo_common(struct xdr_stream *xdr, struct nfs4_secinfo_res p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->flavors->num_flavors = 0; num_flavors = be32_to_cpup(p); @@ -5641,7 +5460,7 @@ static int decode_secinfo_common(struct xdr_stream *xdr, struct nfs4_secinfo_res p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; sec_flavor->flavor = be32_to_cpup(p); if (sec_flavor->flavor == RPC_AUTH_GSS) { @@ -5655,9 +5474,6 @@ static int decode_secinfo_common(struct xdr_stream *xdr, struct nfs4_secinfo_res status = 0; out: return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_secinfo(struct xdr_stream *xdr, struct nfs4_secinfo_res *res) @@ -5711,11 +5527,11 @@ static int decode_exchange_id(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; xdr_decode_hyper(p, &res->clientid); p = xdr_inline_decode(xdr, 12); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->seqid = be32_to_cpup(p++); res->flags = be32_to_cpup(p++); @@ -5739,7 +5555,7 @@ static int decode_exchange_id(struct xdr_stream *xdr, /* server_owner4.so_minor_id */ p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; p = xdr_decode_hyper(p, &res->server_owner->minor_id); /* server_owner4.so_major_id */ @@ -5759,7 +5575,7 @@ static int decode_exchange_id(struct xdr_stream *xdr, /* Implementation Id */ p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; impl_id_count = be32_to_cpup(p++); if (impl_id_count) { @@ -5778,16 +5594,13 @@ static int decode_exchange_id(struct xdr_stream *xdr, /* nii_date */ p = xdr_inline_decode(xdr, 12); if (unlikely(!p)) - goto out_overflow; + return -EIO; p = xdr_decode_hyper(p, &res->impl_id->date.seconds); res->impl_id->date.nseconds = be32_to_cpup(p); /* if there's more than one entry, ignore the rest */ } return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_chan_attrs(struct xdr_stream *xdr, @@ -5798,7 +5611,7 @@ static int decode_chan_attrs(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 28); if (unlikely(!p)) - goto out_overflow; + return -EIO; val = be32_to_cpup(p++); /* headerpadsz */ if (val) return -EINVAL; /* no support for header padding yet */ @@ -5816,12 +5629,9 @@ static int decode_chan_attrs(struct xdr_stream *xdr, if (nr_attrs == 1) { p = xdr_inline_decode(xdr, 4); /* skip rdma_attrs */ if (unlikely(!p)) - goto out_overflow; + return -EIO; } return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_sessionid(struct xdr_stream *xdr, struct nfs4_sessionid *sid) @@ -5844,7 +5654,7 @@ static int decode_bind_conn_to_session(struct xdr_stream *xdr, /* dir flags, rdma mode bool */ p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->dir = be32_to_cpup(p++); if (res->dir == 0 || res->dir > NFS4_CDFS4_BOTH) @@ -5855,9 +5665,6 @@ static int decode_bind_conn_to_session(struct xdr_stream *xdr, res->use_conn_in_rdma_mode = true; return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_create_session(struct xdr_stream *xdr, @@ -5875,7 +5682,7 @@ static int decode_create_session(struct xdr_stream *xdr, /* seqid, flags */ p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->seqid = be32_to_cpup(p++); res->flags = be32_to_cpup(p); @@ -5884,9 +5691,6 @@ static int decode_create_session(struct xdr_stream *xdr, if (!status) status = decode_chan_attrs(xdr, &res->bc_attrs); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_destroy_session(struct xdr_stream *xdr, void *dummy) @@ -5967,7 +5771,6 @@ out_err: res->sr_status = status; return status; out_overflow: - print_overflow_msg(__func__, xdr); status = -EIO; goto out_err; #else /* CONFIG_NFS_V4_1 */ @@ -5995,7 +5798,7 @@ static int decode_getdeviceinfo(struct xdr_stream *xdr, if (status == -ETOOSMALL) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; pdev->mincount = be32_to_cpup(p); dprintk("%s: Min count too small. mincnt = %u\n", __func__, pdev->mincount); @@ -6005,7 +5808,7 @@ static int decode_getdeviceinfo(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; type = be32_to_cpup(p++); if (type != pdev->layout_type) { dprintk("%s: layout mismatch req: %u pdev: %u\n", @@ -6019,19 +5822,19 @@ static int decode_getdeviceinfo(struct xdr_stream *xdr, */ pdev->mincount = be32_to_cpup(p); if (xdr_read_pages(xdr, pdev->mincount) != pdev->mincount) - goto out_overflow; + return -EIO; /* Parse notification bitmap, verifying that it is zero. */ p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; len = be32_to_cpup(p); if (len) { uint32_t i; p = xdr_inline_decode(xdr, 4 * len); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->notification = be32_to_cpup(p++); for (i = 1; i < len; i++) { @@ -6043,9 +5846,6 @@ static int decode_getdeviceinfo(struct xdr_stream *xdr, } } return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_layoutget(struct xdr_stream *xdr, struct rpc_rqst *req, @@ -6115,7 +5915,6 @@ out: res->status = status; return status; out_overflow: - print_overflow_msg(__func__, xdr); status = -EIO; goto out; } @@ -6131,16 +5930,13 @@ static int decode_layoutreturn(struct xdr_stream *xdr, return status; p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->lrs_present = be32_to_cpup(p); if (res->lrs_present) status = decode_layout_stateid(xdr, &res->stateid); else nfs4_stateid_copy(&res->stateid, &invalid_stateid); return status; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_layoutcommit(struct xdr_stream *xdr, @@ -6158,19 +5954,16 @@ static int decode_layoutcommit(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; sizechanged = be32_to_cpup(p); if (sizechanged) { /* throw away new size */ p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) - goto out_overflow; + return -EIO; } return 0; -out_overflow: - print_overflow_msg(__func__, xdr); - return -EIO; } static int decode_test_stateid(struct xdr_stream *xdr, @@ -6186,21 +5979,17 @@ static int decode_test_stateid(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; num_res = be32_to_cpup(p++); if (num_res != 1) - goto out; + return -EIO; p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EIO; res->status = be32_to_cpup(p++); return status; -out_overflow: - print_overflow_msg(__func__, xdr); -out: - return -EIO; } static int decode_free_stateid(struct xdr_stream *xdr, @@ -7570,11 +7359,11 @@ int nfs4_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, uint64_t new_cookie; __be32 *p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EAGAIN; if (*p == xdr_zero) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) - goto out_overflow; + return -EAGAIN; if (*p == xdr_zero) return -EAGAIN; entry->eof = 1; @@ -7583,13 +7372,13 @@ int nfs4_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, p = xdr_inline_decode(xdr, 12); if (unlikely(!p)) - goto out_overflow; + return -EAGAIN; p = xdr_decode_hyper(p, &new_cookie); entry->len = be32_to_cpup(p); p = xdr_inline_decode(xdr, entry->len); if (unlikely(!p)) - goto out_overflow; + return -EAGAIN; entry->name = (const char *) p; /* @@ -7601,14 +7390,14 @@ int nfs4_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, entry->fattr->valid = 0; if (decode_attr_bitmap(xdr, bitmap) < 0) - goto out_overflow; + return -EAGAIN; if (decode_attr_length(xdr, &len, &savep) < 0) - goto out_overflow; + return -EAGAIN; if (decode_getfattr_attrs(xdr, bitmap, entry->fattr, entry->fh, NULL, entry->label, entry->server) < 0) - goto out_overflow; + return -EAGAIN; if (entry->fattr->valid & NFS_ATTR_FATTR_MOUNTED_ON_FILEID) entry->ino = entry->fattr->mounted_on_fileid; else if (entry->fattr->valid & NFS_ATTR_FATTR_FILEID) @@ -7622,10 +7411,6 @@ int nfs4_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, entry->cookie = new_cookie; return 0; - -out_overflow: - print_overflow_msg(__func__, xdr); - return -EAGAIN; } /* diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index c74e4538d0eb..a9d24d5a967c 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -60,16 +60,6 @@ struct nfs4_cb_compound_hdr { int status; }; -/* - * Handle decode buffer overflows out-of-line. - */ -static void print_overflow_msg(const char *func, const struct xdr_stream *xdr) -{ - dprintk("NFS: %s prematurely hit the end of our receive buffer. " - "Remaining buffer length is %tu words.\n", - func, xdr->end - xdr->p); -} - static __be32 *xdr_encode_empty_array(__be32 *p) { *p++ = xdr_zero; @@ -240,7 +230,6 @@ static int decode_cb_op_status(struct xdr_stream *xdr, *status = nfs_cb_stat_to_errno(be32_to_cpup(p)); return 0; out_overflow: - print_overflow_msg(__func__, xdr); return -EIO; out_unexpected: dprintk("NFSD: Callback server returned operation %d but " @@ -309,7 +298,6 @@ static int decode_cb_compound4res(struct xdr_stream *xdr, hdr->nops = be32_to_cpup(p); return 0; out_overflow: - print_overflow_msg(__func__, xdr); return -EIO; } @@ -437,7 +425,6 @@ out: cb->cb_seq_status = status; return status; out_overflow: - print_overflow_msg(__func__, xdr); status = -EIO; goto out; } From f23f658404229278af49284871518174ebbe3580 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:26 -0500 Subject: [PATCH 09/90] NFS: Add trace events to report non-zero NFS status codes These can help field troubleshooting without needing the overhead of a full network capture (ie, tcpdump). Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- fs/nfs/nfs2xdr.c | 7 ++++ fs/nfs/nfs3xdr.c | 7 ++++ fs/nfs/nfs4trace.h | 25 ++++++++++++++ fs/nfs/nfs4xdr.c | 12 ++++--- fs/nfs/nfstrace.c | 1 + fs/nfs/nfstrace.h | 85 ++++++++++++++++++++++++++++++++++++++++++++++ 6 files changed, 133 insertions(+), 4 deletions(-) diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index 76614311f957..bac3a4e2cb5d 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -22,6 +22,7 @@ #include #include #include +#include "nfstrace.h" #include "internal.h" #define NFSDBG_FACILITY NFSDBG_XDR @@ -145,7 +146,13 @@ static int decode_stat(struct xdr_stream *xdr, enum nfs_stat *status) p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) return -EIO; + if (unlikely(*p != cpu_to_be32(NFS_OK))) + goto out_status; + *status = 0; + return 0; +out_status: *status = be32_to_cpup(p); + trace_nfs_xdr_status((int)*status); return 0; } diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index e5619803a4f6..4aa3ffe1800e 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -21,6 +21,7 @@ #include #include #include +#include "nfstrace.h" #include "internal.h" #define NFSDBG_FACILITY NFSDBG_XDR @@ -337,7 +338,13 @@ static int decode_nfsstat3(struct xdr_stream *xdr, enum nfs_stat *status) p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) return -EIO; + if (unlikely(*p != cpu_to_be32(NFS3_OK))) + goto out_status; + *status = 0; + return 0; +out_status: *status = be32_to_cpup(p); + trace_nfs_xdr_status((int)*status); return 0; } diff --git a/fs/nfs/nfs4trace.h b/fs/nfs/nfs4trace.h index b4557cf685fb..cd1a5c08da9a 100644 --- a/fs/nfs/nfs4trace.h +++ b/fs/nfs/nfs4trace.h @@ -524,6 +524,31 @@ TRACE_EVENT(nfs4_setup_sequence, ) ); +TRACE_EVENT(nfs4_xdr_status, + TP_PROTO( + u32 op, + int error + ), + + TP_ARGS(op, error), + + TP_STRUCT__entry( + __field(u32, op) + __field(int, error) + ), + + TP_fast_assign( + __entry->op = op; + __entry->error = -error; + ), + + TP_printk( + "operation %d: nfs status %d (%s)", + __entry->op, + __entry->error, show_nfsv4_errors(__entry->error) + ) +); + DECLARE_EVENT_CLASS(nfs4_open_event, TP_PROTO( const struct nfs_open_context *ctx, diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 24e6a45a5726..38a4cbc18657 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -54,6 +54,7 @@ #include #include "nfs4_fs.h" +#include "nfs4trace.h" #include "internal.h" #include "nfs4idmap.h" #include "nfs4session.h" @@ -3188,11 +3189,14 @@ static bool __decode_op_hdr(struct xdr_stream *xdr, enum nfs_opnum4 expected, opnum = be32_to_cpup(p++); if (unlikely(opnum != expected)) goto out_bad_operation; + if (unlikely(*p != cpu_to_be32(NFS_OK))) + goto out_status; + *nfs_retval = 0; + return true; +out_status: nfserr = be32_to_cpup(p); - if (nfserr == NFS_OK) - *nfs_retval = 0; - else - *nfs_retval = nfs4_stat_to_errno(nfserr); + trace_nfs4_xdr_status(opnum, nfserr); + *nfs_retval = nfs4_stat_to_errno(nfserr); return true; out_bad_operation: dprintk("nfs: Server returned operation" diff --git a/fs/nfs/nfstrace.c b/fs/nfs/nfstrace.c index b60d5fbd7727..a90b363500c2 100644 --- a/fs/nfs/nfstrace.c +++ b/fs/nfs/nfstrace.c @@ -11,3 +11,4 @@ EXPORT_TRACEPOINT_SYMBOL_GPL(nfs_fsync_enter); EXPORT_TRACEPOINT_SYMBOL_GPL(nfs_fsync_exit); +EXPORT_TRACEPOINT_SYMBOL_GPL(nfs_xdr_status); diff --git a/fs/nfs/nfstrace.h b/fs/nfs/nfstrace.h index bd60f8d1e181..a0d6910aa03a 100644 --- a/fs/nfs/nfstrace.h +++ b/fs/nfs/nfstrace.h @@ -969,6 +969,91 @@ TRACE_EVENT(nfs_commit_done, ) ); +TRACE_DEFINE_ENUM(NFS_OK); +TRACE_DEFINE_ENUM(NFSERR_PERM); +TRACE_DEFINE_ENUM(NFSERR_NOENT); +TRACE_DEFINE_ENUM(NFSERR_IO); +TRACE_DEFINE_ENUM(NFSERR_NXIO); +TRACE_DEFINE_ENUM(NFSERR_ACCES); +TRACE_DEFINE_ENUM(NFSERR_EXIST); +TRACE_DEFINE_ENUM(NFSERR_XDEV); +TRACE_DEFINE_ENUM(NFSERR_NODEV); +TRACE_DEFINE_ENUM(NFSERR_NOTDIR); +TRACE_DEFINE_ENUM(NFSERR_ISDIR); +TRACE_DEFINE_ENUM(NFSERR_INVAL); +TRACE_DEFINE_ENUM(NFSERR_FBIG); +TRACE_DEFINE_ENUM(NFSERR_NOSPC); +TRACE_DEFINE_ENUM(NFSERR_ROFS); +TRACE_DEFINE_ENUM(NFSERR_MLINK); +TRACE_DEFINE_ENUM(NFSERR_NAMETOOLONG); +TRACE_DEFINE_ENUM(NFSERR_NOTEMPTY); +TRACE_DEFINE_ENUM(NFSERR_DQUOT); +TRACE_DEFINE_ENUM(NFSERR_STALE); +TRACE_DEFINE_ENUM(NFSERR_REMOTE); +TRACE_DEFINE_ENUM(NFSERR_WFLUSH); +TRACE_DEFINE_ENUM(NFSERR_BADHANDLE); +TRACE_DEFINE_ENUM(NFSERR_NOT_SYNC); +TRACE_DEFINE_ENUM(NFSERR_BAD_COOKIE); +TRACE_DEFINE_ENUM(NFSERR_NOTSUPP); +TRACE_DEFINE_ENUM(NFSERR_TOOSMALL); +TRACE_DEFINE_ENUM(NFSERR_SERVERFAULT); +TRACE_DEFINE_ENUM(NFSERR_BADTYPE); +TRACE_DEFINE_ENUM(NFSERR_JUKEBOX); + +#define nfs_show_status(x) \ + __print_symbolic(x, \ + { NFS_OK, "OK" }, \ + { NFSERR_PERM, "PERM" }, \ + { NFSERR_NOENT, "NOENT" }, \ + { NFSERR_IO, "IO" }, \ + { NFSERR_NXIO, "NXIO" }, \ + { NFSERR_ACCES, "ACCES" }, \ + { NFSERR_EXIST, "EXIST" }, \ + { NFSERR_XDEV, "XDEV" }, \ + { NFSERR_NODEV, "NODEV" }, \ + { NFSERR_NOTDIR, "NOTDIR" }, \ + { NFSERR_ISDIR, "ISDIR" }, \ + { NFSERR_INVAL, "INVAL" }, \ + { NFSERR_FBIG, "FBIG" }, \ + { NFSERR_NOSPC, "NOSPC" }, \ + { NFSERR_ROFS, "ROFS" }, \ + { NFSERR_MLINK, "MLINK" }, \ + { NFSERR_NAMETOOLONG, "NAMETOOLONG" }, \ + { NFSERR_NOTEMPTY, "NOTEMPTY" }, \ + { NFSERR_DQUOT, "DQUOT" }, \ + { NFSERR_STALE, "STALE" }, \ + { NFSERR_REMOTE, "REMOTE" }, \ + { NFSERR_WFLUSH, "WFLUSH" }, \ + { NFSERR_BADHANDLE, "BADHANDLE" }, \ + { NFSERR_NOT_SYNC, "NOTSYNC" }, \ + { NFSERR_BAD_COOKIE, "BADCOOKIE" }, \ + { NFSERR_NOTSUPP, "NOTSUPP" }, \ + { NFSERR_TOOSMALL, "TOOSMALL" }, \ + { NFSERR_SERVERFAULT, "REMOTEIO" }, \ + { NFSERR_BADTYPE, "BADTYPE" }, \ + { NFSERR_JUKEBOX, "JUKEBOX" }) + +TRACE_EVENT(nfs_xdr_status, + TP_PROTO( + int error + ), + + TP_ARGS(error), + + TP_STRUCT__entry( + __field(int, error) + ), + + TP_fast_assign( + __entry->error = error; + ), + + TP_printk( + "error=%d (%s)", + __entry->error, nfs_show_status(__entry->error) + ) +); + #endif /* _TRACE_NFS_H */ #undef TRACE_INCLUDE_PATH From 80125d4ae70aff049b170dd21ea5d007694d629a Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:32 -0500 Subject: [PATCH 10/90] SUNRPC: Remove some dprintk() call sites from auth functions Clean up: Reduce dprintk noise by removing dprintk() call sites from hot path that do not report exceptions. These are usually replaceable with function graph tracing. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- net/sunrpc/auth.c | 29 ----------------------------- net/sunrpc/auth_unix.c | 9 +-------- 2 files changed, 1 insertion(+), 37 deletions(-) diff --git a/net/sunrpc/auth.c b/net/sunrpc/auth.c index 8dfab6119e6a..275e84e817b7 100644 --- a/net/sunrpc/auth.c +++ b/net/sunrpc/auth.c @@ -17,10 +17,6 @@ #include #include -#if IS_ENABLED(CONFIG_SUNRPC_DEBUG) -# define RPCDBG_FACILITY RPCDBG_AUTH -#endif - #define RPC_CREDCACHE_DEFAULT_HASHBITS (4) struct rpc_cred_cache { struct hlist_head *hashtable; @@ -267,8 +263,6 @@ rpcauth_list_flavors(rpc_authflavor_t *array, int size) } } rcu_read_unlock(); - - dprintk("RPC: %s returns %d\n", __func__, result); return result; } EXPORT_SYMBOL_GPL(rpcauth_list_flavors); @@ -636,9 +630,6 @@ rpcauth_lookupcred(struct rpc_auth *auth, int flags) struct rpc_cred *ret; const struct cred *cred = current_cred(); - dprintk("RPC: looking up %s cred\n", - auth->au_ops->au_name); - memset(&acred, 0, sizeof(acred)); acred.cred = cred; ret = auth->au_ops->lookup_cred(auth, &acred, flags); @@ -670,8 +661,6 @@ rpcauth_bind_root_cred(struct rpc_task *task, int lookupflags) }; struct rpc_cred *ret; - dprintk("RPC: %5u looking up %s cred\n", - task->tk_pid, task->tk_client->cl_auth->au_ops->au_name); ret = auth->au_ops->lookup_cred(auth, &acred, lookupflags); put_cred(acred.cred); return ret; @@ -688,8 +677,6 @@ rpcauth_bind_machine_cred(struct rpc_task *task, int lookupflags) if (!acred.principal) return NULL; - dprintk("RPC: %5u looking up %s machine cred\n", - task->tk_pid, task->tk_client->cl_auth->au_ops->au_name); return auth->au_ops->lookup_cred(auth, &acred, lookupflags); } @@ -698,8 +685,6 @@ rpcauth_bind_new_cred(struct rpc_task *task, int lookupflags) { struct rpc_auth *auth = task->tk_client->cl_auth; - dprintk("RPC: %5u looking up %s cred\n", - task->tk_pid, auth->au_ops->au_name); return rpcauth_lookupcred(auth, lookupflags); } @@ -776,9 +761,6 @@ rpcauth_marshcred(struct rpc_task *task, __be32 *p) { struct rpc_cred *cred = task->tk_rqstp->rq_cred; - dprintk("RPC: %5u marshaling %s cred %p\n", - task->tk_pid, cred->cr_auth->au_ops->au_name, cred); - return cred->cr_ops->crmarshal(task, p); } @@ -787,9 +769,6 @@ rpcauth_checkverf(struct rpc_task *task, __be32 *p) { struct rpc_cred *cred = task->tk_rqstp->rq_cred; - dprintk("RPC: %5u validating %s cred %p\n", - task->tk_pid, cred->cr_auth->au_ops->au_name, cred); - return cred->cr_ops->crvalidate(task, p); } @@ -808,8 +787,6 @@ rpcauth_wrap_req(struct rpc_task *task, kxdreproc_t encode, void *rqstp, { struct rpc_cred *cred = task->tk_rqstp->rq_cred; - dprintk("RPC: %5u using %s cred %p to wrap rpc data\n", - task->tk_pid, cred->cr_ops->cr_name, cred); if (cred->cr_ops->crwrap_req) return cred->cr_ops->crwrap_req(task, encode, rqstp, data, obj); /* By default, we encode the arguments normally. */ @@ -833,8 +810,6 @@ rpcauth_unwrap_resp(struct rpc_task *task, kxdrdproc_t decode, void *rqstp, { struct rpc_cred *cred = task->tk_rqstp->rq_cred; - dprintk("RPC: %5u using %s cred %p to unwrap rpc data\n", - task->tk_pid, cred->cr_ops->cr_name, cred); if (cred->cr_ops->crunwrap_resp) return cred->cr_ops->crunwrap_resp(task, decode, rqstp, data, obj); @@ -865,8 +840,6 @@ rpcauth_refreshcred(struct rpc_task *task) goto out; cred = task->tk_rqstp->rq_cred; } - dprintk("RPC: %5u refreshing %s cred %p\n", - task->tk_pid, cred->cr_auth->au_ops->au_name, cred); err = cred->cr_ops->crrefresh(task); out: @@ -880,8 +853,6 @@ rpcauth_invalcred(struct rpc_task *task) { struct rpc_cred *cred = task->tk_rqstp->rq_cred; - dprintk("RPC: %5u invalidating %s cred %p\n", - task->tk_pid, cred->cr_auth->au_ops->au_name, cred); if (cred) clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); } diff --git a/net/sunrpc/auth_unix.c b/net/sunrpc/auth_unix.c index 387f6b3ffbea..fc8a59134640 100644 --- a/net/sunrpc/auth_unix.c +++ b/net/sunrpc/auth_unix.c @@ -28,8 +28,6 @@ static mempool_t *unix_pool; static struct rpc_auth * unx_create(const struct rpc_auth_create_args *args, struct rpc_clnt *clnt) { - dprintk("RPC: creating UNIX authenticator for client %p\n", - clnt); refcount_inc(&unix_auth.au_count); return &unix_auth; } @@ -37,7 +35,6 @@ unx_create(const struct rpc_auth_create_args *args, struct rpc_clnt *clnt) static void unx_destroy(struct rpc_auth *auth) { - dprintk("RPC: destroying UNIX authenticator %p\n", auth); } /* @@ -48,10 +45,6 @@ unx_lookup_cred(struct rpc_auth *auth, struct auth_cred *acred, int flags) { struct rpc_cred *ret = mempool_alloc(unix_pool, GFP_NOFS); - dprintk("RPC: allocating UNIX cred for uid %d gid %d\n", - from_kuid(&init_user_ns, acred->cred->fsuid), - from_kgid(&init_user_ns, acred->cred->fsgid)); - rpcauth_init_cred(ret, acred, auth, &unix_credops); ret->cr_flags = 1UL << RPCAUTH_CRED_UPTODATE; return ret; @@ -61,7 +54,7 @@ static void unx_free_cred_callback(struct rcu_head *head) { struct rpc_cred *rpc_cred = container_of(head, struct rpc_cred, cr_rcu); - dprintk("RPC: unx_free_cred %p\n", rpc_cred); + put_cred(rpc_cred->cr_cred); mempool_free(rpc_cred, unix_pool); } From 067fb11b12af1448f7bbcacca41e470cb775e9fa Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:37 -0500 Subject: [PATCH 11/90] SUNRPC: Remove rpc_xprt::tsh_size tsh_size was added to accommodate transports that send a pre-amble before each RPC message. However, this assumes the pre-amble is fixed in size, which isn't true for some transports. That makes tsh_size not very generic. Also I'd like to make the estimation of RPC send and receive buffer sizes more precise. tsh_size doesn't currently appear to be accounted for at all by call_allocate. Therefore let's just remove the tsh_size concept, and make the only transports that have a non-zero tsh_size employ a direct approach. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xprt.h | 7 -- net/sunrpc/auth_gss/auth_gss.c | 3 +- net/sunrpc/clnt.c | 1 - net/sunrpc/svc.c | 19 +---- net/sunrpc/xprtrdma/svc_rdma_backchannel.c | 1 - net/sunrpc/xprtrdma/transport.c | 1 - net/sunrpc/xprtsock.c | 91 ++++++++++++++-------- 7 files changed, 65 insertions(+), 58 deletions(-) diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index ad7e910b119d..3a391544299e 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -196,8 +196,6 @@ struct rpc_xprt { size_t max_payload; /* largest RPC payload size, in bytes */ - unsigned int tsh_size; /* size of transport specific - header */ struct rpc_wait_queue binding; /* requests waiting on rpcbind */ struct rpc_wait_queue sending; /* requests waiting to send */ @@ -362,11 +360,6 @@ struct rpc_xprt * xprt_alloc(struct net *net, size_t size, unsigned int max_req); void xprt_free(struct rpc_xprt *); -static inline __be32 *xprt_skip_transport_header(struct rpc_xprt *xprt, __be32 *p) -{ - return p + xprt->tsh_size; -} - static inline int xprt_enable_swap(struct rpc_xprt *xprt) { diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index a42672e81792..4b52e2b11c58 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -1563,8 +1563,7 @@ gss_marshal(struct rpc_task *task, __be32 *p) /* We compute the checksum for the verifier over the xdr-encoded bytes * starting with the xid and ending at the end of the credential: */ - iov.iov_base = xprt_skip_transport_header(req->rq_xprt, - req->rq_snd_buf.head[0].iov_base); + iov.iov_base = req->rq_snd_buf.head[0].iov_base; iov.iov_len = (u8 *)p - (u8 *)iov.iov_base; xdr_buf_from_iov(&iov, &verf_buf); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index d7ec6132c046..c4203f6138ef 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -2331,7 +2331,6 @@ rpc_encode_header(struct rpc_task *task) /* FIXME: check buffer size? */ - p = xprt_skip_transport_header(req->rq_xprt, p); *p++ = req->rq_xid; /* XID */ *p++ = htonl(RPC_CALL); /* CALL */ *p++ = htonl(RPC_VERSION); /* RPC version */ diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index e87ddb9f7feb..dbd19697ee38 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -1144,17 +1144,6 @@ void svc_printk(struct svc_rqst *rqstp, const char *fmt, ...) static __printf(2,3) void svc_printk(struct svc_rqst *rqstp, const char *fmt, ...) {} #endif -/* - * Setup response header for TCP, it has a 4B record length field. - */ -static void svc_tcp_prep_reply_hdr(struct svc_rqst *rqstp) -{ - struct kvec *resv = &rqstp->rq_res.head[0]; - - /* tcp needs a space for the record length... */ - svc_putnl(resv, 0); -} - /* * Common routine for processing the RPC request. */ @@ -1182,10 +1171,6 @@ svc_process_common(struct svc_rqst *rqstp, struct kvec *argv, struct kvec *resv) set_bit(RQ_USEDEFERRAL, &rqstp->rq_flags); clear_bit(RQ_DROPME, &rqstp->rq_flags); - /* Setup reply header */ - if (rqstp->rq_prot == IPPROTO_TCP) - svc_tcp_prep_reply_hdr(rqstp); - svc_putu32(resv, rqstp->rq_xid); vers = svc_getnl(argv); @@ -1443,6 +1428,10 @@ svc_process(struct svc_rqst *rqstp) goto out_drop; } + /* Reserve space for the record marker */ + if (rqstp->rq_prot == IPPROTO_TCP) + svc_putnl(resv, 0); + /* Returns 1 for send, 0 for drop */ if (likely(svc_process_common(rqstp, argv, resv))) return svc_send(rqstp); diff --git a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c index b908f2ca08fd..907464c2a9f0 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c +++ b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c @@ -304,7 +304,6 @@ xprt_setup_rdma_bc(struct xprt_create *args) xprt->idle_timeout = RPCRDMA_IDLE_DISC_TO; xprt->prot = XPRT_TRANSPORT_BC_RDMA; - xprt->tsh_size = 0; xprt->ops = &xprt_rdma_bc_procs; memcpy(&xprt->addr, args->dstaddr, args->addrlen); diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c index fbc171ebfe91..e7274dc10120 100644 --- a/net/sunrpc/xprtrdma/transport.c +++ b/net/sunrpc/xprtrdma/transport.c @@ -332,7 +332,6 @@ xprt_setup_rdma(struct xprt_create *args) xprt->idle_timeout = RPCRDMA_IDLE_DISC_TO; xprt->resvport = 0; /* privileged port not needed */ - xprt->tsh_size = 0; /* RPC-RDMA handles framing */ xprt->ops = &xprt_rdma_procs; /* diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 7754aa3e434f..ae09d850cd11 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -696,6 +696,40 @@ xs_stream_reset_connect(struct sock_xprt *transport) #define XS_SENDMSG_FLAGS (MSG_DONTWAIT | MSG_NOSIGNAL) +/* Common case: + * - stream transport + * - sending from byte 0 of the message + * - the message is wholly contained in @xdr's head iovec + */ +static int xs_send_rm_and_kvec(struct socket *sock, struct xdr_buf *xdr, + unsigned int remainder) +{ + struct msghdr msg = { + .msg_flags = XS_SENDMSG_FLAGS | (remainder ? MSG_MORE : 0) + }; + rpc_fraghdr marker = cpu_to_be32(RPC_LAST_STREAM_FRAGMENT | + (u32)xdr->len); + struct kvec iov[2] = { + { + .iov_base = &marker, + .iov_len = sizeof(marker) + }, + { + .iov_base = xdr->head[0].iov_base, + .iov_len = xdr->head[0].iov_len + }, + }; + int ret; + + ret = kernel_sendmsg(sock, &msg, iov, 2, + iov[0].iov_len + iov[1].iov_len); + if (ret < 0) + return ret; + if (ret < iov[0].iov_len) + return -EPIPE; + return ret - iov[0].iov_len; +} + static int xs_send_kvec(struct socket *sock, struct sockaddr *addr, int addrlen, struct kvec *vec, unsigned int base, int more) { struct msghdr msg = { @@ -779,7 +813,11 @@ static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, if (base < xdr->head[0].iov_len || addr != NULL) { unsigned int len = xdr->head[0].iov_len - base; remainder -= len; - err = xs_send_kvec(sock, addr, addrlen, &xdr->head[0], base, remainder != 0); + if (!base && !addr) + err = xs_send_rm_and_kvec(sock, xdr, remainder); + else + err = xs_send_kvec(sock, addr, addrlen, &xdr->head[0], + base, remainder != 0); if (remainder == 0 || err != len) goto out; *sent_p += err; @@ -869,16 +907,6 @@ xs_send_request_was_aborted(struct sock_xprt *transport, struct rpc_rqst *req) return transport->xmit.offset != 0 && req->rq_bytes_sent == 0; } -/* - * Construct a stream transport record marker in @buf. - */ -static inline void xs_encode_stream_record_marker(struct xdr_buf *buf) -{ - u32 reclen = buf->len - sizeof(rpc_fraghdr); - rpc_fraghdr *base = buf->head[0].iov_base; - *base = cpu_to_be32(RPC_LAST_STREAM_FRAGMENT | reclen); -} - /** * xs_local_send_request - write an RPC request to an AF_LOCAL socket * @req: pointer to RPC request @@ -905,8 +933,6 @@ static int xs_local_send_request(struct rpc_rqst *req) return -ENOTCONN; } - xs_encode_stream_record_marker(&req->rq_snd_buf); - xs_pktdump("packet data:", req->rq_svec->iov_base, req->rq_svec->iov_len); @@ -1057,8 +1083,6 @@ static int xs_tcp_send_request(struct rpc_rqst *req) return -ENOTCONN; } - xs_encode_stream_record_marker(&req->rq_snd_buf); - xs_pktdump("packet data:", req->rq_svec->iov_base, req->rq_svec->iov_len); @@ -2534,26 +2558,35 @@ static int bc_sendto(struct rpc_rqst *req) { int len; struct xdr_buf *xbufp = &req->rq_snd_buf; - struct rpc_xprt *xprt = req->rq_xprt; struct sock_xprt *transport = - container_of(xprt, struct sock_xprt, xprt); - struct socket *sock = transport->sock; + container_of(req->rq_xprt, struct sock_xprt, xprt); unsigned long headoff; unsigned long tailoff; + struct page *tailpage; + struct msghdr msg = { + .msg_flags = MSG_MORE + }; + rpc_fraghdr marker = cpu_to_be32(RPC_LAST_STREAM_FRAGMENT | + (u32)xbufp->len); + struct kvec iov = { + .iov_base = &marker, + .iov_len = sizeof(marker), + }; - xs_encode_stream_record_marker(xbufp); + len = kernel_sendmsg(transport->sock, &msg, &iov, 1, iov.iov_len); + if (len != iov.iov_len) + return -EAGAIN; + tailpage = NULL; + if (xbufp->tail[0].iov_len) + tailpage = virt_to_page(xbufp->tail[0].iov_base); tailoff = (unsigned long)xbufp->tail[0].iov_base & ~PAGE_MASK; headoff = (unsigned long)xbufp->head[0].iov_base & ~PAGE_MASK; - len = svc_send_common(sock, xbufp, + len = svc_send_common(transport->sock, xbufp, virt_to_page(xbufp->head[0].iov_base), headoff, - xbufp->tail[0].iov_base, tailoff); - - if (len != xbufp->len) { - printk(KERN_NOTICE "Error sending entire callback!\n"); - len = -EAGAIN; - } - + tailpage, tailoff); + if (len != xbufp->len) + return -EAGAIN; return len; } @@ -2793,7 +2826,6 @@ static struct rpc_xprt *xs_setup_local(struct xprt_create *args) transport = container_of(xprt, struct sock_xprt, xprt); xprt->prot = 0; - xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32); xprt->max_payload = RPC_MAX_FRAGMENT_SIZE; xprt->bind_timeout = XS_BIND_TO; @@ -2862,7 +2894,6 @@ static struct rpc_xprt *xs_setup_udp(struct xprt_create *args) transport = container_of(xprt, struct sock_xprt, xprt); xprt->prot = IPPROTO_UDP; - xprt->tsh_size = 0; /* XXX: header size can vary due to auth type, IPv6, etc. */ xprt->max_payload = (1U << 16) - (MAX_HEADER << 3); @@ -2942,7 +2973,6 @@ static struct rpc_xprt *xs_setup_tcp(struct xprt_create *args) transport = container_of(xprt, struct sock_xprt, xprt); xprt->prot = IPPROTO_TCP; - xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32); xprt->max_payload = RPC_MAX_FRAGMENT_SIZE; xprt->bind_timeout = XS_BIND_TO; @@ -3015,7 +3045,6 @@ static struct rpc_xprt *xs_setup_bc_tcp(struct xprt_create *args) transport = container_of(xprt, struct sock_xprt, xprt); xprt->prot = IPPROTO_TCP; - xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32); xprt->max_payload = RPC_MAX_FRAGMENT_SIZE; xprt->timeout = &xs_tcp_default_timeout; From fe9a270519c72bccb3af524db7ea6c7b67700d50 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:43 -0500 Subject: [PATCH 12/90] SUNRPC: Add build option to disable support for insecure enctypes Enable distributions to enforce the rejection of ancient and insecure Kerberos enctypes in the kernel's RPCSEC_GSS implementation. These are the single-DES encryption types that were deprecated in 2012 by RFC 6649. Enctypes that were deprecated more recently (by RFC 8429) remain fully supported for now because they are still likely to be widely used. Signed-off-by: Chuck Lever Acked-by: Simo Sorce Signed-off-by: Anna Schumaker --- include/linux/sunrpc/gss_krb5_enctypes.h | 42 +++++++++++++++++++++++- net/sunrpc/Kconfig | 16 +++++++++ net/sunrpc/auth_gss/gss_krb5_mech.c | 2 ++ 3 files changed, 59 insertions(+), 1 deletion(-) diff --git a/include/linux/sunrpc/gss_krb5_enctypes.h b/include/linux/sunrpc/gss_krb5_enctypes.h index ec6234eee89c..981c89cef19d 100644 --- a/include/linux/sunrpc/gss_krb5_enctypes.h +++ b/include/linux/sunrpc/gss_krb5_enctypes.h @@ -1,4 +1,44 @@ +/* SPDX-License-Identifier: GPL-2.0 */ /* - * Dumb way to share this static piece of information with nfsd + * Define the string that exports the set of kernel-supported + * Kerberos enctypes. This list is sent via upcall to gssd, and + * is also exposed via the nfsd /proc API. The consumers generally + * treat this as an ordered list, where the first item in the list + * is the most preferred. + */ + +#ifndef _LINUX_SUNRPC_GSS_KRB5_ENCTYPES_H +#define _LINUX_SUNRPC_GSS_KRB5_ENCTYPES_H + +#ifdef CONFIG_SUNRPC_DISABLE_INSECURE_ENCTYPES + +/* + * NB: This list includes encryption types that were deprecated + * by RFC 8429 (DES3_CBC_SHA1 and ARCFOUR_HMAC). + * + * ENCTYPE_AES256_CTS_HMAC_SHA1_96 + * ENCTYPE_AES128_CTS_HMAC_SHA1_96 + * ENCTYPE_DES3_CBC_SHA1 + * ENCTYPE_ARCFOUR_HMAC + */ +#define KRB5_SUPPORTED_ENCTYPES "18,17,16,23" + +#else /* CONFIG_SUNRPC_DISABLE_INSECURE_ENCTYPES */ + +/* + * NB: This list includes encryption types that were deprecated + * by RFC 8429 and RFC 6649. + * + * ENCTYPE_AES256_CTS_HMAC_SHA1_96 + * ENCTYPE_AES128_CTS_HMAC_SHA1_96 + * ENCTYPE_DES3_CBC_SHA1 + * ENCTYPE_ARCFOUR_HMAC + * ENCTYPE_DES_CBC_MD5 + * ENCTYPE_DES_CBC_CRC + * ENCTYPE_DES_CBC_MD4 */ #define KRB5_SUPPORTED_ENCTYPES "18,17,16,23,3,1,2" + +#endif /* CONFIG_SUNRPC_DISABLE_INSECURE_ENCTYPES */ + +#endif /* _LINUX_SUNRPC_GSS_KRB5_ENCTYPES_H */ diff --git a/net/sunrpc/Kconfig b/net/sunrpc/Kconfig index ac09ca803296..83f5617bae07 100644 --- a/net/sunrpc/Kconfig +++ b/net/sunrpc/Kconfig @@ -34,6 +34,22 @@ config RPCSEC_GSS_KRB5 If unsure, say Y. +config CONFIG_SUNRPC_DISABLE_INSECURE_ENCTYPES + bool "Secure RPC: Disable insecure Kerberos encryption types" + depends on RPCSEC_GSS_KRB5 + default n + help + Choose Y here to disable the use of deprecated encryption types + with the Kerberos version 5 GSS-API mechanism (RFC 1964). The + deprecated encryption types include DES-CBC-MD5, DES-CBC-CRC, + and DES-CBC-MD4. These types were deprecated by RFC 6649 because + they were found to be insecure. + + N is the default because many sites have deployed KDCs and + keytabs that contain only these deprecated encryption types. + Choosing Y prevents the use of known-insecure encryption types + but might result in compatibility problems. + config SUNRPC_DEBUG bool "RPC: Enable dprintk debugging" depends on SUNRPC && SYSCTL diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index eab71fc7af3e..be31a58d54e0 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -53,6 +53,7 @@ static struct gss_api_mech gss_kerberos_mech; /* forward declaration */ static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { +#ifndef CONFIG_SUNRPC_DISABLE_INSECURE_ENCTYPES /* * DES (All DES enctypes are mapped to the same gss functionality) */ @@ -74,6 +75,7 @@ static const struct gss_krb5_enctype supported_gss_krb5_enctypes[] = { .cksumlength = 8, .keyed_cksum = 0, }, +#endif /* CONFIG_SUNRPC_DISABLE_INSECURE_ENCTYPES */ /* * RC4-HMAC */ From e8680a24a269bd6dcb533f4e4a5faba9ae58925c Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:48 -0500 Subject: [PATCH 13/90] SUNRPC: Use struct xdr_stream when constructing RPC Call header Modernize and harden the code path that constructs each RPC Call message. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/auth.h | 15 ++- include/linux/sunrpc/xdr.h | 6 + include/trace/events/sunrpc.h | 29 +++++ net/sunrpc/auth.c | 62 +++++++---- net/sunrpc/auth_gss/auth_gss.c | 193 ++++++++++++++++----------------- net/sunrpc/auth_null.c | 21 ++-- net/sunrpc/auth_unix.c | 57 ++++++---- net/sunrpc/clnt.c | 62 ++++++----- 8 files changed, 265 insertions(+), 180 deletions(-) diff --git a/include/linux/sunrpc/auth.h b/include/linux/sunrpc/auth.h index eed3cb16ccf1..96e237f8e60b 100644 --- a/include/linux/sunrpc/auth.h +++ b/include/linux/sunrpc/auth.h @@ -131,11 +131,12 @@ struct rpc_credops { void (*crdestroy)(struct rpc_cred *); int (*crmatch)(struct auth_cred *, struct rpc_cred *, int); - __be32 * (*crmarshal)(struct rpc_task *, __be32 *); + int (*crmarshal)(struct rpc_task *task, + struct xdr_stream *xdr); int (*crrefresh)(struct rpc_task *); __be32 * (*crvalidate)(struct rpc_task *, __be32 *); - int (*crwrap_req)(struct rpc_task *, kxdreproc_t, - void *, __be32 *, void *); + int (*crwrap_req)(struct rpc_task *task, + struct xdr_stream *xdr); int (*crunwrap_resp)(struct rpc_task *, kxdrdproc_t, void *, __be32 *, void *); int (*crkey_timeout)(struct rpc_cred *); @@ -165,9 +166,13 @@ struct rpc_cred * rpcauth_lookup_credcache(struct rpc_auth *, struct auth_cred * void rpcauth_init_cred(struct rpc_cred *, const struct auth_cred *, struct rpc_auth *, const struct rpc_credops *); struct rpc_cred * rpcauth_lookupcred(struct rpc_auth *, int); void put_rpccred(struct rpc_cred *); -__be32 * rpcauth_marshcred(struct rpc_task *, __be32 *); +int rpcauth_marshcred(struct rpc_task *task, + struct xdr_stream *xdr); __be32 * rpcauth_checkverf(struct rpc_task *, __be32 *); -int rpcauth_wrap_req(struct rpc_task *task, kxdreproc_t encode, void *rqstp, __be32 *data, void *obj); +int rpcauth_wrap_req_encode(struct rpc_task *task, + struct xdr_stream *xdr); +int rpcauth_wrap_req(struct rpc_task *task, + struct xdr_stream *xdr); int rpcauth_unwrap_resp(struct rpc_task *task, kxdrdproc_t decode, void *rqstp, __be32 *data, void *obj); bool rpcauth_xmit_need_reencode(struct rpc_task *task); int rpcauth_refreshcred(struct rpc_task *); diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index 787939d13643..6df9ac1ca471 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -87,6 +87,12 @@ xdr_buf_init(struct xdr_buf *buf, void *start, size_t len) #define xdr_one cpu_to_be32(1) #define xdr_two cpu_to_be32(2) +#define rpc_auth_null cpu_to_be32(RPC_AUTH_NULL) +#define rpc_auth_unix cpu_to_be32(RPC_AUTH_UNIX) +#define rpc_auth_gss cpu_to_be32(RPC_AUTH_GSS) + +#define rpc_call cpu_to_be32(RPC_CALL) + #define rpc_success cpu_to_be32(RPC_SUCCESS) #define rpc_prog_unavail cpu_to_be32(RPC_PROG_UNAVAIL) #define rpc_prog_mismatch cpu_to_be32(RPC_PROG_MISMATCH) diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 627650800676..2b3f9d139e75 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -213,6 +213,35 @@ DECLARE_EVENT_CLASS(rpc_task_queued, DEFINE_RPC_QUEUED_EVENT(sleep); DEFINE_RPC_QUEUED_EVENT(wakeup); +DECLARE_EVENT_CLASS(rpc_failure, + + TP_PROTO(const struct rpc_task *task), + + TP_ARGS(task), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + ), + + TP_fast_assign( + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; + ), + + TP_printk("task:%u@%u", + __entry->task_id, __entry->client_id) +); + +#define DEFINE_RPC_FAILURE(name) \ + DEFINE_EVENT(rpc_failure, rpc_bad_##name, \ + TP_PROTO( \ + const struct rpc_task *task \ + ), \ + TP_ARGS(task)) + +DEFINE_RPC_FAILURE(callhdr); + TRACE_EVENT(rpc_stats_latency, TP_PROTO( diff --git a/net/sunrpc/auth.c b/net/sunrpc/auth.c index 275e84e817b7..add2135d9b01 100644 --- a/net/sunrpc/auth.c +++ b/net/sunrpc/auth.c @@ -756,12 +756,21 @@ destroy: } EXPORT_SYMBOL_GPL(put_rpccred); -__be32 * -rpcauth_marshcred(struct rpc_task *task, __be32 *p) +/** + * rpcauth_marshcred - Append RPC credential to end of @xdr + * @task: controlling RPC task + * @xdr: xdr_stream containing initial portion of RPC Call header + * + * On success, an appropriate verifier is added to @xdr, @xdr is + * updated to point past the verifier, and zero is returned. + * Otherwise, @xdr is in an undefined state and a negative errno + * is returned. + */ +int rpcauth_marshcred(struct rpc_task *task, struct xdr_stream *xdr) { - struct rpc_cred *cred = task->tk_rqstp->rq_cred; + const struct rpc_credops *ops = task->tk_rqstp->rq_cred->cr_ops; - return cred->cr_ops->crmarshal(task, p); + return ops->crmarshal(task, xdr); } __be32 * @@ -772,27 +781,38 @@ rpcauth_checkverf(struct rpc_task *task, __be32 *p) return cred->cr_ops->crvalidate(task, p); } -static void rpcauth_wrap_req_encode(kxdreproc_t encode, struct rpc_rqst *rqstp, - __be32 *data, void *obj) +/** + * rpcauth_wrap_req_encode - XDR encode the RPC procedure + * @task: controlling RPC task + * @xdr: stream where on-the-wire bytes are to be marshalled + * + * On success, @xdr contains the encoded and wrapped message. + * Otherwise, @xdr is in an undefined state. + */ +int rpcauth_wrap_req_encode(struct rpc_task *task, struct xdr_stream *xdr) { - struct xdr_stream xdr; + kxdreproc_t encode = task->tk_msg.rpc_proc->p_encode; - xdr_init_encode(&xdr, &rqstp->rq_snd_buf, data, rqstp); - encode(rqstp, &xdr, obj); -} - -int -rpcauth_wrap_req(struct rpc_task *task, kxdreproc_t encode, void *rqstp, - __be32 *data, void *obj) -{ - struct rpc_cred *cred = task->tk_rqstp->rq_cred; - - if (cred->cr_ops->crwrap_req) - return cred->cr_ops->crwrap_req(task, encode, rqstp, data, obj); - /* By default, we encode the arguments normally. */ - rpcauth_wrap_req_encode(encode, rqstp, data, obj); + encode(task->tk_rqstp, xdr, task->tk_msg.rpc_argp); return 0; } +EXPORT_SYMBOL_GPL(rpcauth_wrap_req_encode); + +/** + * rpcauth_wrap_req - XDR encode and wrap the RPC procedure + * @task: controlling RPC task + * @xdr: stream where on-the-wire bytes are to be marshalled + * + * On success, @xdr contains the encoded and wrapped message, + * and zero is returned. Otherwise, @xdr is in an undefined + * state and a negative errno is returned. + */ +int rpcauth_wrap_req(struct rpc_task *task, struct xdr_stream *xdr) +{ + const struct rpc_credops *ops = task->tk_rqstp->rq_cred->cr_ops; + + return ops->crwrap_req(task, xdr); +} static int rpcauth_unwrap_req_decode(kxdrdproc_t decode, struct rpc_rqst *rqstp, diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 4b52e2b11c58..b333b1bdad45 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -1526,18 +1526,20 @@ out: } /* -* Marshal credentials. -* Maybe we should keep a cached credential for performance reasons. -*/ -static __be32 * -gss_marshal(struct rpc_task *task, __be32 *p) + * Marshal credentials. + * + * The expensive part is computing the verifier. We can't cache a + * pre-computed version of the verifier because the seqno, which + * is different every time, is included in the MIC. + */ +static int gss_marshal(struct rpc_task *task, struct xdr_stream *xdr) { struct rpc_rqst *req = task->tk_rqstp; struct rpc_cred *cred = req->rq_cred; struct gss_cred *gss_cred = container_of(cred, struct gss_cred, gc_base); struct gss_cl_ctx *ctx = gss_cred_get_ctx(cred); - __be32 *cred_len; + __be32 *p, *cred_len; u32 maj_stat = 0; struct xdr_netobj mic; struct kvec iov; @@ -1545,7 +1547,13 @@ gss_marshal(struct rpc_task *task, __be32 *p) dprintk("RPC: %5u %s\n", task->tk_pid, __func__); - *p++ = htonl(RPC_AUTH_GSS); + /* Credential */ + + p = xdr_reserve_space(xdr, 7 * sizeof(*p) + + ctx->gc_wire_ctx.len); + if (!p) + goto out_put_ctx; + *p++ = rpc_auth_gss; cred_len = p++; spin_lock(&ctx->gc_seq_lock); @@ -1554,12 +1562,14 @@ gss_marshal(struct rpc_task *task, __be32 *p) if (req->rq_seqno == MAXSEQ) goto out_expired; - *p++ = htonl((u32) RPC_GSS_VERSION); - *p++ = htonl((u32) ctx->gc_proc); - *p++ = htonl((u32) req->rq_seqno); - *p++ = htonl((u32) gss_cred->gc_service); + *p++ = cpu_to_be32(RPC_GSS_VERSION); + *p++ = cpu_to_be32(ctx->gc_proc); + *p++ = cpu_to_be32(req->rq_seqno); + *p++ = cpu_to_be32(gss_cred->gc_service); p = xdr_encode_netobj(p, &ctx->gc_wire_ctx); - *cred_len = htonl((p - (cred_len + 1)) << 2); + *cred_len = cpu_to_be32((p - (cred_len + 1)) << 2); + + /* Verifier */ /* We compute the checksum for the verifier over the xdr-encoded bytes * starting with the xid and ending at the end of the credential: */ @@ -1567,27 +1577,27 @@ gss_marshal(struct rpc_task *task, __be32 *p) iov.iov_len = (u8 *)p - (u8 *)iov.iov_base; xdr_buf_from_iov(&iov, &verf_buf); - /* set verifier flavor*/ - *p++ = htonl(RPC_AUTH_GSS); - + p = xdr_reserve_space(xdr, sizeof(*p)); + if (!p) + goto out_put_ctx; + *p++ = rpc_auth_gss; mic.data = (u8 *)(p + 1); maj_stat = gss_get_mic(ctx->gc_gss_ctx, &verf_buf, &mic); - if (maj_stat == GSS_S_CONTEXT_EXPIRED) { + if (maj_stat == GSS_S_CONTEXT_EXPIRED) goto out_expired; - } else if (maj_stat != 0) { - pr_warn("gss_marshal: gss_get_mic FAILED (%d)\n", maj_stat); - task->tk_status = -EIO; + else if (maj_stat != 0) + goto out_put_ctx; + if (xdr_stream_encode_opaque_inline(xdr, (void **)&p, mic.len) < 0) goto out_put_ctx; - } - p = xdr_encode_opaque(p, NULL, mic.len); gss_put_ctx(ctx); - return p; + return 0; out_expired: + gss_put_ctx(ctx); clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); - task->tk_status = -EKEYEXPIRED; + return -EKEYEXPIRED; out_put_ctx: gss_put_ctx(ctx); - return NULL; + return -EMSGSIZE; } static int gss_renew_cred(struct rpc_task *task) @@ -1716,61 +1726,45 @@ out_bad: return ret; } -static void gss_wrap_req_encode(kxdreproc_t encode, struct rpc_rqst *rqstp, - __be32 *p, void *obj) +static int gss_wrap_req_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, + struct rpc_task *task, struct xdr_stream *xdr) { - struct xdr_stream xdr; - - xdr_init_encode(&xdr, &rqstp->rq_snd_buf, p, rqstp); - encode(rqstp, &xdr, obj); -} - -static inline int -gss_wrap_req_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, - kxdreproc_t encode, struct rpc_rqst *rqstp, - __be32 *p, void *obj) -{ - struct xdr_buf *snd_buf = &rqstp->rq_snd_buf; - struct xdr_buf integ_buf; - __be32 *integ_len = NULL; + struct rpc_rqst *rqstp = task->tk_rqstp; + struct xdr_buf integ_buf, *snd_buf = &rqstp->rq_snd_buf; struct xdr_netobj mic; - u32 offset; - __be32 *q; - struct kvec *iov; - u32 maj_stat = 0; - int status = -EIO; + __be32 *p, *integ_len; + u32 offset, maj_stat; + p = xdr_reserve_space(xdr, 2 * sizeof(*p)); + if (!p) + goto wrap_failed; integ_len = p++; + *p = cpu_to_be32(rqstp->rq_seqno); + + if (rpcauth_wrap_req_encode(task, xdr)) + goto wrap_failed; + offset = (u8 *)p - (u8 *)snd_buf->head[0].iov_base; - *p++ = htonl(rqstp->rq_seqno); - - gss_wrap_req_encode(encode, rqstp, p, obj); - if (xdr_buf_subsegment(snd_buf, &integ_buf, offset, snd_buf->len - offset)) - return status; - *integ_len = htonl(integ_buf.len); + goto wrap_failed; + *integ_len = cpu_to_be32(integ_buf.len); - /* guess whether we're in the head or the tail: */ - if (snd_buf->page_len || snd_buf->tail[0].iov_len) - iov = snd_buf->tail; - else - iov = snd_buf->head; - p = iov->iov_base + iov->iov_len; + p = xdr_reserve_space(xdr, 0); + if (!p) + goto wrap_failed; mic.data = (u8 *)(p + 1); - maj_stat = gss_get_mic(ctx->gc_gss_ctx, &integ_buf, &mic); - status = -EIO; /* XXX? */ if (maj_stat == GSS_S_CONTEXT_EXPIRED) clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); else if (maj_stat) - return status; - q = xdr_encode_opaque(p, NULL, mic.len); - - offset = (u8 *)q - (u8 *)p; - iov->iov_len += offset; - snd_buf->len += offset; + goto wrap_failed; + /* Check that the trailing MIC fit in the buffer, after the fact */ + if (xdr_stream_encode_opaque_inline(xdr, (void **)&p, mic.len) < 0) + goto wrap_failed; return 0; +wrap_failed: + return -EMSGSIZE; } static void @@ -1821,61 +1815,63 @@ out: return -EAGAIN; } -static inline int -gss_wrap_req_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, - kxdreproc_t encode, struct rpc_rqst *rqstp, - __be32 *p, void *obj) +static int gss_wrap_req_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, + struct rpc_task *task, struct xdr_stream *xdr) { + struct rpc_rqst *rqstp = task->tk_rqstp; struct xdr_buf *snd_buf = &rqstp->rq_snd_buf; - u32 offset; - u32 maj_stat; + u32 pad, offset, maj_stat; int status; - __be32 *opaque_len; + __be32 *p, *opaque_len; struct page **inpages; int first; - int pad; struct kvec *iov; - char *tmp; + status = -EIO; + p = xdr_reserve_space(xdr, 2 * sizeof(*p)); + if (!p) + goto wrap_failed; opaque_len = p++; - offset = (u8 *)p - (u8 *)snd_buf->head[0].iov_base; - *p++ = htonl(rqstp->rq_seqno); + *p = cpu_to_be32(rqstp->rq_seqno); - gss_wrap_req_encode(encode, rqstp, p, obj); + if (rpcauth_wrap_req_encode(task, xdr)) + goto wrap_failed; status = alloc_enc_pages(rqstp); - if (status) - return status; + if (unlikely(status)) + goto wrap_failed; first = snd_buf->page_base >> PAGE_SHIFT; inpages = snd_buf->pages + first; snd_buf->pages = rqstp->rq_enc_pages; snd_buf->page_base -= first << PAGE_SHIFT; /* - * Give the tail its own page, in case we need extra space in the - * head when wrapping: + * Move the tail into its own page, in case gss_wrap needs + * more space in the head when wrapping. * - * call_allocate() allocates twice the slack space required - * by the authentication flavor to rq_callsize. - * For GSS, slack is GSS_CRED_SLACK. + * Still... Why can't gss_wrap just slide the tail down? */ if (snd_buf->page_len || snd_buf->tail[0].iov_len) { + char *tmp; + tmp = page_address(rqstp->rq_enc_pages[rqstp->rq_enc_pages_num - 1]); memcpy(tmp, snd_buf->tail[0].iov_base, snd_buf->tail[0].iov_len); snd_buf->tail[0].iov_base = tmp; } + status = -EIO; + offset = (u8 *)p - (u8 *)snd_buf->head[0].iov_base; maj_stat = gss_wrap(ctx->gc_gss_ctx, offset, snd_buf, inpages); /* slack space should prevent this ever happening: */ - BUG_ON(snd_buf->len > snd_buf->buflen); - status = -EIO; + if (unlikely(snd_buf->len > snd_buf->buflen)) + goto wrap_failed; /* We're assuming that when GSS_S_CONTEXT_EXPIRED, the encryption was * done anyway, so it's safe to put the request on the wire: */ if (maj_stat == GSS_S_CONTEXT_EXPIRED) clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); else if (maj_stat) - return status; + goto wrap_failed; - *opaque_len = htonl(snd_buf->len - offset); - /* guess whether we're in the head or the tail: */ + *opaque_len = cpu_to_be32(snd_buf->len - offset); + /* guess whether the pad goes into the head or the tail: */ if (snd_buf->page_len || snd_buf->tail[0].iov_len) iov = snd_buf->tail; else @@ -1887,37 +1883,36 @@ gss_wrap_req_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, snd_buf->len += pad; return 0; +wrap_failed: + return status; } -static int -gss_wrap_req(struct rpc_task *task, - kxdreproc_t encode, void *rqstp, __be32 *p, void *obj) +static int gss_wrap_req(struct rpc_task *task, struct xdr_stream *xdr) { struct rpc_cred *cred = task->tk_rqstp->rq_cred; struct gss_cred *gss_cred = container_of(cred, struct gss_cred, gc_base); struct gss_cl_ctx *ctx = gss_cred_get_ctx(cred); - int status = -EIO; + int status; dprintk("RPC: %5u %s\n", task->tk_pid, __func__); + status = -EIO; if (ctx->gc_proc != RPC_GSS_PROC_DATA) { /* The spec seems a little ambiguous here, but I think that not * wrapping context destruction requests makes the most sense. */ - gss_wrap_req_encode(encode, rqstp, p, obj); - status = 0; + status = rpcauth_wrap_req_encode(task, xdr); goto out; } switch (gss_cred->gc_service) { case RPC_GSS_SVC_NONE: - gss_wrap_req_encode(encode, rqstp, p, obj); - status = 0; + status = rpcauth_wrap_req_encode(task, xdr); break; case RPC_GSS_SVC_INTEGRITY: - status = gss_wrap_req_integ(cred, ctx, encode, rqstp, p, obj); + status = gss_wrap_req_integ(cred, ctx, task, xdr); break; case RPC_GSS_SVC_PRIVACY: - status = gss_wrap_req_priv(cred, ctx, encode, rqstp, p, obj); + status = gss_wrap_req_priv(cred, ctx, task, xdr); break; } out: diff --git a/net/sunrpc/auth_null.c b/net/sunrpc/auth_null.c index d0ceac57c06e..797f8472c21b 100644 --- a/net/sunrpc/auth_null.c +++ b/net/sunrpc/auth_null.c @@ -59,15 +59,21 @@ nul_match(struct auth_cred *acred, struct rpc_cred *cred, int taskflags) /* * Marshal credential. */ -static __be32 * -nul_marshal(struct rpc_task *task, __be32 *p) +static int +nul_marshal(struct rpc_task *task, struct xdr_stream *xdr) { - *p++ = htonl(RPC_AUTH_NULL); - *p++ = 0; - *p++ = htonl(RPC_AUTH_NULL); - *p++ = 0; + __be32 *p; - return p; + p = xdr_reserve_space(xdr, 4 * sizeof(*p)); + if (!p) + return -EMSGSIZE; + /* Credential */ + *p++ = rpc_auth_null; + *p++ = xdr_zero; + /* Verifier */ + *p++ = rpc_auth_null; + *p = xdr_zero; + return 0; } /* @@ -125,6 +131,7 @@ const struct rpc_credops null_credops = { .crdestroy = nul_destroy_cred, .crmatch = nul_match, .crmarshal = nul_marshal, + .crwrap_req = rpcauth_wrap_req_encode, .crrefresh = nul_refresh, .crvalidate = nul_validate, }; diff --git a/net/sunrpc/auth_unix.c b/net/sunrpc/auth_unix.c index fc8a59134640..1d5b7ed9c6f7 100644 --- a/net/sunrpc/auth_unix.c +++ b/net/sunrpc/auth_unix.c @@ -99,37 +99,55 @@ unx_match(struct auth_cred *acred, struct rpc_cred *cred, int flags) * Marshal credentials. * Maybe we should keep a cached credential for performance reasons. */ -static __be32 * -unx_marshal(struct rpc_task *task, __be32 *p) +static int +unx_marshal(struct rpc_task *task, struct xdr_stream *xdr) { struct rpc_clnt *clnt = task->tk_client; struct rpc_cred *cred = task->tk_rqstp->rq_cred; - __be32 *base, *hold; + __be32 *p, *cred_len, *gidarr_len; int i; struct group_info *gi = cred->cr_cred->group_info; - *p++ = htonl(RPC_AUTH_UNIX); - base = p++; - *p++ = htonl(jiffies/HZ); + /* Credential */ - /* - * Copy the UTS nodename captured when the client was created. - */ - p = xdr_encode_array(p, clnt->cl_nodename, clnt->cl_nodelen); + p = xdr_reserve_space(xdr, 3 * sizeof(*p)); + if (!p) + goto marshal_failed; + *p++ = rpc_auth_unix; + cred_len = p++; + *p++ = xdr_zero; /* stamp */ + if (xdr_stream_encode_opaque(xdr, clnt->cl_nodename, + clnt->cl_nodelen) < 0) + goto marshal_failed; + p = xdr_reserve_space(xdr, 3 * sizeof(*p)); + if (!p) + goto marshal_failed; + *p++ = cpu_to_be32(from_kuid(&init_user_ns, cred->cr_cred->fsuid)); + *p++ = cpu_to_be32(from_kgid(&init_user_ns, cred->cr_cred->fsgid)); - *p++ = htonl((u32) from_kuid(&init_user_ns, cred->cr_cred->fsuid)); - *p++ = htonl((u32) from_kgid(&init_user_ns, cred->cr_cred->fsgid)); - hold = p++; + gidarr_len = p++; if (gi) for (i = 0; i < UNX_NGROUPS && i < gi->ngroups; i++) - *p++ = htonl((u32) from_kgid(&init_user_ns, gi->gid[i])); - *hold = htonl(p - hold - 1); /* gid array length */ - *base = htonl((p - base - 1) << 2); /* cred length */ + *p++ = cpu_to_be32(from_kgid(&init_user_ns, + gi->gid[i])); + *gidarr_len = cpu_to_be32(p - gidarr_len - 1); + *cred_len = cpu_to_be32((p - cred_len - 1) << 2); + p = xdr_reserve_space(xdr, (p - gidarr_len - 1) << 2); + if (!p) + goto marshal_failed; - *p++ = htonl(RPC_AUTH_NULL); - *p++ = htonl(0); + /* Verifier */ - return p; + p = xdr_reserve_space(xdr, 2 * sizeof(*p)); + if (!p) + goto marshal_failed; + *p++ = rpc_auth_null; + *p = xdr_zero; + + return 0; + +marshal_failed: + return -EMSGSIZE; } /* @@ -202,6 +220,7 @@ const struct rpc_credops unix_credops = { .crdestroy = unx_destroy_cred, .crmatch = unx_match, .crmarshal = unx_marshal, + .crwrap_req = rpcauth_wrap_req_encode, .crrefresh = unx_refresh, .crvalidate = unx_validate, }; diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index c4203f6138ef..d6750b7f169a 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -77,7 +77,8 @@ static void call_timeout(struct rpc_task *task); static void call_connect(struct rpc_task *task); static void call_connect_status(struct rpc_task *task); -static __be32 *rpc_encode_header(struct rpc_task *task); +static int rpc_encode_header(struct rpc_task *task, + struct xdr_stream *xdr); static __be32 *rpc_verify_header(struct rpc_task *task); static int rpc_ping(struct rpc_clnt *clnt); @@ -1728,10 +1729,7 @@ static void rpc_xdr_encode(struct rpc_task *task) { struct rpc_rqst *req = task->tk_rqstp; - kxdreproc_t encode; - __be32 *p; - - dprint_status(task); + struct xdr_stream xdr; xdr_buf_init(&req->rq_snd_buf, req->rq_buffer, @@ -1740,18 +1738,13 @@ rpc_xdr_encode(struct rpc_task *task) req->rq_rbuffer, req->rq_rcvsize); - p = rpc_encode_header(task); - if (p == NULL) + req->rq_snd_buf.head[0].iov_len = 0; + xdr_init_encode(&xdr, &req->rq_snd_buf, + req->rq_snd_buf.head[0].iov_base, req); + if (rpc_encode_header(task, &xdr)) return; - encode = task->tk_msg.rpc_proc->p_encode; - if (encode == NULL) - return; - - task->tk_status = rpcauth_wrap_req(task, encode, req, p, - task->tk_msg.rpc_argp); - if (task->tk_status == 0) - xprt_request_prepare(req); + task->tk_status = rpcauth_wrap_req(task, &xdr); } /* @@ -1762,6 +1755,7 @@ call_encode(struct rpc_task *task) { if (!rpc_task_need_encode(task)) goto out; + dprint_status(task); /* Encode here so that rpcsec_gss can use correct sequence number. */ rpc_xdr_encode(task); /* Did the encode result in an error condition? */ @@ -1779,6 +1773,8 @@ call_encode(struct rpc_task *task) rpc_exit(task, task->tk_status); } return; + } else { + xprt_request_prepare(task->tk_rqstp); } /* Add task to reply queue before transmission to avoid races */ @@ -2322,25 +2318,33 @@ out_retry: } } -static __be32 * -rpc_encode_header(struct rpc_task *task) +static int +rpc_encode_header(struct rpc_task *task, struct xdr_stream *xdr) { struct rpc_clnt *clnt = task->tk_client; struct rpc_rqst *req = task->tk_rqstp; - __be32 *p = req->rq_svec[0].iov_base; + __be32 *p; + int error; - /* FIXME: check buffer size? */ + error = -EMSGSIZE; + p = xdr_reserve_space(xdr, RPC_CALLHDRSIZE << 2); + if (!p) + goto out_fail; + *p++ = req->rq_xid; + *p++ = rpc_call; + *p++ = cpu_to_be32(RPC_VERSION); + *p++ = cpu_to_be32(clnt->cl_prog); + *p++ = cpu_to_be32(clnt->cl_vers); + *p = cpu_to_be32(task->tk_msg.rpc_proc->p_proc); - *p++ = req->rq_xid; /* XID */ - *p++ = htonl(RPC_CALL); /* CALL */ - *p++ = htonl(RPC_VERSION); /* RPC version */ - *p++ = htonl(clnt->cl_prog); /* program number */ - *p++ = htonl(clnt->cl_vers); /* program version */ - *p++ = htonl(task->tk_msg.rpc_proc->p_proc); /* procedure */ - p = rpcauth_marshcred(task, p); - if (p) - req->rq_slen = xdr_adjust_iovec(&req->rq_svec[0], p); - return p; + error = rpcauth_marshcred(task, xdr); + if (error < 0) + goto out_fail; + return 0; +out_fail: + trace_rpc_bad_callhdr(task); + rpc_exit(task, error); + return error; } static __be32 * From 7f5667a5f8c4ff85b14ccce9d41f9244bd30ab68 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:53 -0500 Subject: [PATCH 14/90] SUNRPC: Clean up rpc_verify_header() - Recover some instruction count because I'm about to introduce a few xdr_inline_decode call sites - Replace dprintk() call sites with trace points - Reduce the hot path so it fits in fewer cachelines I've also renamed it rpc_decode_header() to match everything else in the RPC client. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xdr.h | 7 +- include/trace/events/sunrpc.h | 52 ++++++++ net/sunrpc/clnt.c | 223 +++++++++++++++------------------- 3 files changed, 154 insertions(+), 128 deletions(-) diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index 6df9ac1ca471..c54041950cc0 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -92,6 +92,9 @@ xdr_buf_init(struct xdr_buf *buf, void *start, size_t len) #define rpc_auth_gss cpu_to_be32(RPC_AUTH_GSS) #define rpc_call cpu_to_be32(RPC_CALL) +#define rpc_reply cpu_to_be32(RPC_REPLY) + +#define rpc_msg_accepted cpu_to_be32(RPC_MSG_ACCEPTED) #define rpc_success cpu_to_be32(RPC_SUCCESS) #define rpc_prog_unavail cpu_to_be32(RPC_PROG_UNAVAIL) @@ -101,6 +104,9 @@ xdr_buf_init(struct xdr_buf *buf, void *start, size_t len) #define rpc_system_err cpu_to_be32(RPC_SYSTEM_ERR) #define rpc_drop_reply cpu_to_be32(RPC_DROP_REPLY) +#define rpc_mismatch cpu_to_be32(RPC_MISMATCH) +#define rpc_auth_error cpu_to_be32(RPC_AUTH_ERROR) + #define rpc_auth_ok cpu_to_be32(RPC_AUTH_OK) #define rpc_autherr_badcred cpu_to_be32(RPC_AUTH_BADCRED) #define rpc_autherr_rejectedcred cpu_to_be32(RPC_AUTH_REJECTEDCRED) @@ -109,7 +115,6 @@ xdr_buf_init(struct xdr_buf *buf, void *start, size_t len) #define rpc_autherr_tooweak cpu_to_be32(RPC_AUTH_TOOWEAK) #define rpcsec_gsserr_credproblem cpu_to_be32(RPCSEC_GSS_CREDPROBLEM) #define rpcsec_gsserr_ctxproblem cpu_to_be32(RPCSEC_GSS_CTXPROBLEM) -#define rpc_autherr_oldseqnum cpu_to_be32(101) /* * Miscellaneous XDR helper functions diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 2b3f9d139e75..0654e9c50371 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -241,6 +241,58 @@ DECLARE_EVENT_CLASS(rpc_failure, TP_ARGS(task)) DEFINE_RPC_FAILURE(callhdr); +DEFINE_RPC_FAILURE(verifier); + +DECLARE_EVENT_CLASS(rpc_reply_event, + + TP_PROTO( + const struct rpc_task *task + ), + + TP_ARGS(task), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + __field(u32, xid) + __string(progname, task->tk_client->cl_program->name) + __field(u32, version) + __string(procname, rpc_proc_name(task)) + __string(servername, task->tk_xprt->servername) + ), + + TP_fast_assign( + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; + __entry->xid = be32_to_cpu(task->tk_rqstp->rq_xid); + __assign_str(progname, task->tk_client->cl_program->name) + __entry->version = task->tk_client->cl_vers; + __assign_str(procname, rpc_proc_name(task)) + __assign_str(servername, task->tk_xprt->servername) + ), + + TP_printk("task:%u@%d server=%s xid=0x%08x %sv%d %s", + __entry->task_id, __entry->client_id, __get_str(servername), + __entry->xid, __get_str(progname), __entry->version, + __get_str(procname)) +) + +#define DEFINE_RPC_REPLY_EVENT(name) \ + DEFINE_EVENT(rpc_reply_event, rpc__##name, \ + TP_PROTO( \ + const struct rpc_task *task \ + ), \ + TP_ARGS(task)) + +DEFINE_RPC_REPLY_EVENT(prog_unavail); +DEFINE_RPC_REPLY_EVENT(prog_mismatch); +DEFINE_RPC_REPLY_EVENT(proc_unavail); +DEFINE_RPC_REPLY_EVENT(garbage_args); +DEFINE_RPC_REPLY_EVENT(unparsable); +DEFINE_RPC_REPLY_EVENT(mismatch); +DEFINE_RPC_REPLY_EVENT(stale_creds); +DEFINE_RPC_REPLY_EVENT(bad_creds); +DEFINE_RPC_REPLY_EVENT(auth_tooweak); TRACE_EVENT(rpc_stats_latency, diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index d6750b7f169a..e9735089bd66 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -79,7 +79,7 @@ static void call_connect_status(struct rpc_task *task); static int rpc_encode_header(struct rpc_task *task, struct xdr_stream *xdr); -static __be32 *rpc_verify_header(struct rpc_task *task); +static __be32 *rpc_decode_header(struct rpc_task *task); static int rpc_ping(struct rpc_clnt *clnt); static void rpc_register_client(struct rpc_clnt *clnt) @@ -2292,7 +2292,7 @@ call_decode(struct rpc_task *task) goto out_retry; } - p = rpc_verify_header(task); + p = rpc_decode_header(task); if (IS_ERR(p)) { if (p == ERR_PTR(-EAGAIN)) goto out_retry; @@ -2308,7 +2308,7 @@ call_decode(struct rpc_task *task) return; out_retry: task->tk_status = 0; - /* Note: rpc_verify_header() may have freed the RPC slot */ + /* Note: rpc_decode_header() may have freed the RPC slot */ if (task->tk_rqstp == req) { xdr_free_bvec(&req->rq_rcv_buf); req->rq_reply_bytes_recvd = req->rq_rcv_buf.len = 0; @@ -2347,164 +2347,133 @@ out_fail: return error; } -static __be32 * -rpc_verify_header(struct rpc_task *task) +static noinline __be32 * +rpc_decode_header(struct rpc_task *task) { struct rpc_clnt *clnt = task->tk_client; struct kvec *iov = &task->tk_rqstp->rq_rcv_buf.head[0]; int len = task->tk_rqstp->rq_rcv_buf.len >> 2; __be32 *p = iov->iov_base; - u32 n; int error = -EACCES; - if ((task->tk_rqstp->rq_rcv_buf.len & 3) != 0) { - /* RFC-1014 says that the representation of XDR data must be a - * multiple of four bytes - * - if it isn't pointer subtraction in the NFS client may give - * undefined results - */ - dprintk("RPC: %5u %s: XDR representation not a multiple of" - " 4 bytes: 0x%x\n", task->tk_pid, __func__, - task->tk_rqstp->rq_rcv_buf.len); - error = -EIO; - goto out_err; - } + /* RFC-1014 says that the representation of XDR data must be a + * multiple of four bytes + * - if it isn't pointer subtraction in the NFS client may give + * undefined results + */ + if (task->tk_rqstp->rq_rcv_buf.len & 3) + goto out_badlen; if ((len -= 3) < 0) - goto out_overflow; + goto out_unparsable; - p += 1; /* skip XID */ - if ((n = ntohl(*p++)) != RPC_REPLY) { - dprintk("RPC: %5u %s: not an RPC reply: %x\n", - task->tk_pid, __func__, n); - error = -EIO; - goto out_garbage; - } + p++; /* skip XID */ + if (*p++ != rpc_reply) + goto out_unparsable; + if (*p++ != rpc_msg_accepted) + goto out_msg_denied; - if ((n = ntohl(*p++)) != RPC_MSG_ACCEPTED) { - if (--len < 0) - goto out_overflow; - switch ((n = ntohl(*p++))) { - case RPC_AUTH_ERROR: - break; - case RPC_MISMATCH: - dprintk("RPC: %5u %s: RPC call version mismatch!\n", - task->tk_pid, __func__); - error = -EPROTONOSUPPORT; - goto out_err; - default: - dprintk("RPC: %5u %s: RPC call rejected, " - "unknown error: %x\n", - task->tk_pid, __func__, n); - error = -EIO; - goto out_err; - } - if (--len < 0) - goto out_overflow; - switch ((n = ntohl(*p++))) { - case RPC_AUTH_REJECTEDCRED: - case RPC_AUTH_REJECTEDVERF: - case RPCSEC_GSS_CREDPROBLEM: - case RPCSEC_GSS_CTXPROBLEM: - if (!task->tk_cred_retry) - break; - task->tk_cred_retry--; - dprintk("RPC: %5u %s: retry stale creds\n", - task->tk_pid, __func__); - rpcauth_invalcred(task); - /* Ensure we obtain a new XID! */ - xprt_release(task); - task->tk_action = call_reserve; - goto out_retry; - case RPC_AUTH_BADCRED: - case RPC_AUTH_BADVERF: - /* possibly garbled cred/verf? */ - if (!task->tk_garb_retry) - break; - task->tk_garb_retry--; - dprintk("RPC: %5u %s: retry garbled creds\n", - task->tk_pid, __func__); - task->tk_action = call_encode; - goto out_retry; - case RPC_AUTH_TOOWEAK: - printk(KERN_NOTICE "RPC: server %s requires stronger " - "authentication.\n", - task->tk_xprt->servername); - break; - default: - dprintk("RPC: %5u %s: unknown auth error: %x\n", - task->tk_pid, __func__, n); - error = -EIO; - } - dprintk("RPC: %5u %s: call rejected %d\n", - task->tk_pid, __func__, n); - goto out_err; - } p = rpcauth_checkverf(task, p); - if (IS_ERR(p)) { - error = PTR_ERR(p); - dprintk("RPC: %5u %s: auth check failed with %d\n", - task->tk_pid, __func__, error); - goto out_garbage; /* bad verifier, retry */ - } + if (IS_ERR(p)) + goto out_verifier; + len = p - (__be32 *)iov->iov_base - 1; if (len < 0) - goto out_overflow; - switch ((n = ntohl(*p++))) { - case RPC_SUCCESS: + goto out_unparsable; + switch (*p++) { + case rpc_success: return p; - case RPC_PROG_UNAVAIL: - dprintk("RPC: %5u %s: program %u is unsupported " - "by server %s\n", task->tk_pid, __func__, - (unsigned int)clnt->cl_prog, - task->tk_xprt->servername); + case rpc_prog_unavail: + trace_rpc__prog_unavail(task); error = -EPFNOSUPPORT; goto out_err; - case RPC_PROG_MISMATCH: - dprintk("RPC: %5u %s: program %u, version %u unsupported " - "by server %s\n", task->tk_pid, __func__, - (unsigned int)clnt->cl_prog, - (unsigned int)clnt->cl_vers, - task->tk_xprt->servername); + case rpc_prog_mismatch: + trace_rpc__prog_mismatch(task); error = -EPROTONOSUPPORT; goto out_err; - case RPC_PROC_UNAVAIL: - dprintk("RPC: %5u %s: proc %s unsupported by program %u, " - "version %u on server %s\n", - task->tk_pid, __func__, - rpc_proc_name(task), - clnt->cl_prog, clnt->cl_vers, - task->tk_xprt->servername); + case rpc_proc_unavail: + trace_rpc__proc_unavail(task); error = -EOPNOTSUPP; goto out_err; - case RPC_GARBAGE_ARGS: - dprintk("RPC: %5u %s: server saw garbage\n", - task->tk_pid, __func__); - break; /* retry */ + case rpc_garbage_args: + trace_rpc__garbage_args(task); + break; default: - dprintk("RPC: %5u %s: server accept status: %x\n", - task->tk_pid, __func__, n); - /* Also retry */ + trace_rpc__unparsable(task); } out_garbage: clnt->cl_stats->rpcgarbage++; if (task->tk_garb_retry) { task->tk_garb_retry--; - dprintk("RPC: %5u %s: retrying\n", - task->tk_pid, __func__); task->tk_action = call_encode; -out_retry: return ERR_PTR(-EAGAIN); } out_err: rpc_exit(task, error); - dprintk("RPC: %5u %s: call failed with error %d\n", task->tk_pid, - __func__, error); return ERR_PTR(error); -out_overflow: - dprintk("RPC: %5u %s: server reply was truncated.\n", task->tk_pid, - __func__); + +out_badlen: + trace_rpc__unparsable(task); + error = -EIO; + goto out_err; + +out_unparsable: + trace_rpc__unparsable(task); + error = -EIO; goto out_garbage; + +out_verifier: + trace_rpc_bad_verifier(task); + error = PTR_ERR(p); + goto out_garbage; + +out_msg_denied: + switch (*p++) { + case rpc_auth_error: + break; + case rpc_mismatch: + trace_rpc__mismatch(task); + error = -EPROTONOSUPPORT; + goto out_err; + default: + trace_rpc__unparsable(task); + error = -EIO; + goto out_err; + } + + switch (*p++) { + case rpc_autherr_rejectedcred: + case rpc_autherr_rejectedverf: + case rpcsec_gsserr_credproblem: + case rpcsec_gsserr_ctxproblem: + if (!task->tk_cred_retry) + break; + task->tk_cred_retry--; + trace_rpc__stale_creds(task); + rpcauth_invalcred(task); + /* Ensure we obtain a new XID! */ + xprt_release(task); + task->tk_action = call_reserve; + return ERR_PTR(-EAGAIN); + case rpc_autherr_badcred: + case rpc_autherr_badverf: + /* possibly garbled cred/verf? */ + if (!task->tk_garb_retry) + break; + task->tk_garb_retry--; + trace_rpc__bad_creds(task); + task->tk_action = call_encode; + return ERR_PTR(-EAGAIN); + case rpc_autherr_tooweak: + trace_rpc__auth_tooweak(task); + pr_warn("RPC: server %s requires stronger authentication.\n", + task->tk_xprt->servername); + break; + default: + trace_rpc__unparsable(task); + error = -EIO; + } + goto out_err; } static void rpcproc_encode_null(struct rpc_rqst *rqstp, struct xdr_stream *xdr, From a0584ee9aed805446b044ce855e67264f0dc619e Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:24:58 -0500 Subject: [PATCH 15/90] SUNRPC: Use struct xdr_stream when decoding RPC Reply header Modernize and harden the code path that parses an RPC Reply message. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/auth.h | 15 ++- include/linux/sunrpc/xdr.h | 1 + net/sunrpc/auth.c | 63 ++++++---- net/sunrpc/auth_gss/auth_gss.c | 204 ++++++++++++++++++--------------- net/sunrpc/auth_null.c | 29 ++--- net/sunrpc/auth_unix.c | 40 ++++--- net/sunrpc/clnt.c | 88 +++++++------- 7 files changed, 241 insertions(+), 199 deletions(-) diff --git a/include/linux/sunrpc/auth.h b/include/linux/sunrpc/auth.h index 96e237f8e60b..c51e1893f77e 100644 --- a/include/linux/sunrpc/auth.h +++ b/include/linux/sunrpc/auth.h @@ -134,11 +134,12 @@ struct rpc_credops { int (*crmarshal)(struct rpc_task *task, struct xdr_stream *xdr); int (*crrefresh)(struct rpc_task *); - __be32 * (*crvalidate)(struct rpc_task *, __be32 *); + int (*crvalidate)(struct rpc_task *task, + struct xdr_stream *xdr); int (*crwrap_req)(struct rpc_task *task, struct xdr_stream *xdr); - int (*crunwrap_resp)(struct rpc_task *, kxdrdproc_t, - void *, __be32 *, void *); + int (*crunwrap_resp)(struct rpc_task *task, + struct xdr_stream *xdr); int (*crkey_timeout)(struct rpc_cred *); char * (*crstringify_acceptor)(struct rpc_cred *); bool (*crneed_reencode)(struct rpc_task *); @@ -168,12 +169,16 @@ struct rpc_cred * rpcauth_lookupcred(struct rpc_auth *, int); void put_rpccred(struct rpc_cred *); int rpcauth_marshcred(struct rpc_task *task, struct xdr_stream *xdr); -__be32 * rpcauth_checkverf(struct rpc_task *, __be32 *); +int rpcauth_checkverf(struct rpc_task *task, + struct xdr_stream *xdr); int rpcauth_wrap_req_encode(struct rpc_task *task, struct xdr_stream *xdr); int rpcauth_wrap_req(struct rpc_task *task, struct xdr_stream *xdr); -int rpcauth_unwrap_resp(struct rpc_task *task, kxdrdproc_t decode, void *rqstp, __be32 *data, void *obj); +int rpcauth_unwrap_resp_decode(struct rpc_task *task, + struct xdr_stream *xdr); +int rpcauth_unwrap_resp(struct rpc_task *task, + struct xdr_stream *xdr); bool rpcauth_xmit_need_reencode(struct rpc_task *task); int rpcauth_refreshcred(struct rpc_task *); void rpcauth_invalcred(struct rpc_task *); diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index c54041950cc0..65af6a204b75 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -89,6 +89,7 @@ xdr_buf_init(struct xdr_buf *buf, void *start, size_t len) #define rpc_auth_null cpu_to_be32(RPC_AUTH_NULL) #define rpc_auth_unix cpu_to_be32(RPC_AUTH_UNIX) +#define rpc_auth_short cpu_to_be32(RPC_AUTH_SHORT) #define rpc_auth_gss cpu_to_be32(RPC_AUTH_GSS) #define rpc_call cpu_to_be32(RPC_CALL) diff --git a/net/sunrpc/auth.c b/net/sunrpc/auth.c index add2135d9b01..e7861026b9e5 100644 --- a/net/sunrpc/auth.c +++ b/net/sunrpc/auth.c @@ -17,6 +17,8 @@ #include #include +#include + #define RPC_CREDCACHE_DEFAULT_HASHBITS (4) struct rpc_cred_cache { struct hlist_head *hashtable; @@ -773,14 +775,6 @@ int rpcauth_marshcred(struct rpc_task *task, struct xdr_stream *xdr) return ops->crmarshal(task, xdr); } -__be32 * -rpcauth_checkverf(struct rpc_task *task, __be32 *p) -{ - struct rpc_cred *cred = task->tk_rqstp->rq_cred; - - return cred->cr_ops->crvalidate(task, p); -} - /** * rpcauth_wrap_req_encode - XDR encode the RPC procedure * @task: controlling RPC task @@ -814,27 +808,52 @@ int rpcauth_wrap_req(struct rpc_task *task, struct xdr_stream *xdr) return ops->crwrap_req(task, xdr); } -static int -rpcauth_unwrap_req_decode(kxdrdproc_t decode, struct rpc_rqst *rqstp, - __be32 *data, void *obj) +/** + * rpcauth_checkverf - Validate verifier in RPC Reply header + * @task: controlling RPC task + * @xdr: xdr_stream containing RPC Reply header + * + * On success, @xdr is updated to point past the verifier and + * zero is returned. Otherwise, @xdr is in an undefined state + * and a negative errno is returned. + */ +int +rpcauth_checkverf(struct rpc_task *task, struct xdr_stream *xdr) { - struct xdr_stream xdr; + const struct rpc_credops *ops = task->tk_rqstp->rq_cred->cr_ops; - xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, data, rqstp); - return decode(rqstp, &xdr, obj); + return ops->crvalidate(task, xdr); } +/** + * rpcauth_unwrap_resp_decode - Invoke XDR decode function + * @task: controlling RPC task + * @xdr: stream where the Reply message resides + * + * Returns zero on success; otherwise a negative errno is returned. + */ int -rpcauth_unwrap_resp(struct rpc_task *task, kxdrdproc_t decode, void *rqstp, - __be32 *data, void *obj) +rpcauth_unwrap_resp_decode(struct rpc_task *task, struct xdr_stream *xdr) { - struct rpc_cred *cred = task->tk_rqstp->rq_cred; + kxdrdproc_t decode = task->tk_msg.rpc_proc->p_decode; - if (cred->cr_ops->crunwrap_resp) - return cred->cr_ops->crunwrap_resp(task, decode, rqstp, - data, obj); - /* By default, we decode the arguments normally. */ - return rpcauth_unwrap_req_decode(decode, rqstp, data, obj); + return decode(task->tk_rqstp, xdr, task->tk_msg.rpc_resp); +} +EXPORT_SYMBOL_GPL(rpcauth_unwrap_resp_decode); + +/** + * rpcauth_unwrap_resp - Invoke unwrap and decode function for the cred + * @task: controlling RPC task + * @xdr: stream where the Reply message resides + * + * Returns zero on success; otherwise a negative errno is returned. + */ +int +rpcauth_unwrap_resp(struct rpc_task *task, struct xdr_stream *xdr) +{ + const struct rpc_credops *ops = task->tk_rqstp->rq_cred->cr_ops; + + return ops->crunwrap_resp(task, xdr); } bool diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index b333b1bdad45..206788e8b787 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -1671,59 +1671,62 @@ gss_refresh_null(struct rpc_task *task) return 0; } -static __be32 * -gss_validate(struct rpc_task *task, __be32 *p) +static int +gss_validate(struct rpc_task *task, struct xdr_stream *xdr) { struct rpc_cred *cred = task->tk_rqstp->rq_cred; struct gss_cl_ctx *ctx = gss_cred_get_ctx(cred); - __be32 *seq = NULL; + __be32 *p, *seq = NULL; struct kvec iov; struct xdr_buf verf_buf; struct xdr_netobj mic; - u32 flav,len; - u32 maj_stat; - __be32 *ret = ERR_PTR(-EIO); + u32 len, maj_stat; + int status; - dprintk("RPC: %5u %s\n", task->tk_pid, __func__); + p = xdr_inline_decode(xdr, 2 * sizeof(*p)); + if (!p) + goto validate_failed; + if (*p++ != rpc_auth_gss) + goto validate_failed; + len = be32_to_cpup(p); + if (len > RPC_MAX_AUTH_SIZE) + goto validate_failed; + p = xdr_inline_decode(xdr, len); + if (!p) + goto validate_failed; - flav = ntohl(*p++); - if ((len = ntohl(*p++)) > RPC_MAX_AUTH_SIZE) - goto out_bad; - if (flav != RPC_AUTH_GSS) - goto out_bad; seq = kmalloc(4, GFP_NOFS); if (!seq) - goto out_bad; - *seq = htonl(task->tk_rqstp->rq_seqno); + goto validate_failed; + *seq = cpu_to_be32(task->tk_rqstp->rq_seqno); iov.iov_base = seq; iov.iov_len = 4; xdr_buf_from_iov(&iov, &verf_buf); mic.data = (u8 *)p; mic.len = len; - - ret = ERR_PTR(-EACCES); maj_stat = gss_verify_mic(ctx->gc_gss_ctx, &verf_buf, &mic); if (maj_stat == GSS_S_CONTEXT_EXPIRED) clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); - if (maj_stat) { - dprintk("RPC: %5u %s: gss_verify_mic returned error 0x%08x\n", - task->tk_pid, __func__, maj_stat); - goto out_bad; - } + if (maj_stat) + goto bad_mic; + /* We leave it to unwrap to calculate au_rslack. For now we just * calculate the length of the verifier: */ cred->cr_auth->au_verfsize = XDR_QUADLEN(len) + 2; + status = 0; +out: gss_put_ctx(ctx); - dprintk("RPC: %5u %s: gss_verify_mic succeeded.\n", - task->tk_pid, __func__); kfree(seq); - return p + XDR_QUADLEN(len); -out_bad: - gss_put_ctx(ctx); - dprintk("RPC: %5u %s failed ret %ld.\n", task->tk_pid, __func__, - PTR_ERR(ret)); - kfree(seq); - return ret; + return status; + +validate_failed: + status = -EIO; + goto out; +bad_mic: + dprintk("RPC: %5u %s: gss_verify_mic returned error 0x%08x\n", + task->tk_pid, __func__, maj_stat); + status = -EACCES; + goto out; } static int gss_wrap_req_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, @@ -1921,79 +1924,98 @@ out: return status; } -static inline int -gss_unwrap_resp_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, - struct rpc_rqst *rqstp, __be32 **p) +static int +gss_unwrap_resp_auth(struct rpc_cred *cred) { - struct xdr_buf *rcv_buf = &rqstp->rq_rcv_buf; - struct xdr_buf integ_buf; - struct xdr_netobj mic; - u32 data_offset, mic_offset; - u32 integ_len; - u32 maj_stat; - int status = -EIO; + cred->cr_auth->au_rslack = cred->cr_auth->au_verfsize; + return 0; +} - integ_len = ntohl(*(*p)++); +static int +gss_unwrap_resp_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, + struct rpc_rqst *rqstp, struct xdr_stream *xdr) +{ + struct xdr_buf integ_buf, *rcv_buf = &rqstp->rq_rcv_buf; + u32 data_offset, mic_offset, integ_len, maj_stat; + struct xdr_netobj mic; + __be32 *p; + + p = xdr_inline_decode(xdr, 2 * sizeof(*p)); + if (unlikely(!p)) + goto unwrap_failed; + integ_len = be32_to_cpup(p++); if (integ_len & 3) - return status; - data_offset = (u8 *)(*p) - (u8 *)rcv_buf->head[0].iov_base; + goto unwrap_failed; + data_offset = (u8 *)(p) - (u8 *)rcv_buf->head[0].iov_base; mic_offset = integ_len + data_offset; if (mic_offset > rcv_buf->len) - return status; - if (ntohl(*(*p)++) != rqstp->rq_seqno) - return status; - - if (xdr_buf_subsegment(rcv_buf, &integ_buf, data_offset, - mic_offset - data_offset)) - return status; + goto unwrap_failed; + if (be32_to_cpup(p) != rqstp->rq_seqno) + goto unwrap_failed; + if (xdr_buf_subsegment(rcv_buf, &integ_buf, data_offset, integ_len)) + goto unwrap_failed; if (xdr_buf_read_netobj(rcv_buf, &mic, mic_offset)) - return status; - + goto unwrap_failed; maj_stat = gss_verify_mic(ctx->gc_gss_ctx, &integ_buf, &mic); if (maj_stat == GSS_S_CONTEXT_EXPIRED) clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); if (maj_stat != GSS_S_COMPLETE) - return status; + goto bad_mic; + + cred->cr_auth->au_rslack = cred->cr_auth->au_verfsize + 2 + + 1 + XDR_QUADLEN(mic.len); return 0; +unwrap_failed: + return -EIO; +bad_mic: + dprintk("RPC: %s: gss_verify_mic returned error 0x%08x\n", + __func__, maj_stat); + return -EIO; } -static inline int +static int gss_unwrap_resp_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, - struct rpc_rqst *rqstp, __be32 **p) + struct rpc_rqst *rqstp, struct xdr_stream *xdr) { - struct xdr_buf *rcv_buf = &rqstp->rq_rcv_buf; - u32 offset; - u32 opaque_len; - u32 maj_stat; - int status = -EIO; + struct xdr_buf *rcv_buf = &rqstp->rq_rcv_buf; + struct kvec *head = rqstp->rq_rcv_buf.head; + unsigned int savedlen = rcv_buf->len; + u32 offset, opaque_len, maj_stat; + __be32 *p; - opaque_len = ntohl(*(*p)++); - offset = (u8 *)(*p) - (u8 *)rcv_buf->head[0].iov_base; + p = xdr_inline_decode(xdr, 2 * sizeof(*p)); + if (unlikely(!p)) + goto unwrap_failed; + opaque_len = be32_to_cpup(p++); + offset = (u8 *)(p) - (u8 *)head->iov_base; if (offset + opaque_len > rcv_buf->len) - return status; - /* remove padding: */ + goto unwrap_failed; rcv_buf->len = offset + opaque_len; maj_stat = gss_unwrap(ctx->gc_gss_ctx, offset, rcv_buf); if (maj_stat == GSS_S_CONTEXT_EXPIRED) clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); if (maj_stat != GSS_S_COMPLETE) - return status; - if (ntohl(*(*p)++) != rqstp->rq_seqno) - return status; + goto bad_unwrap; + /* gss_unwrap decrypted the sequence number */ + if (be32_to_cpup(p++) != rqstp->rq_seqno) + goto unwrap_failed; + /* gss_unwrap redacts the opaque blob from the head iovec. + * rcv_buf has changed, thus the stream needs to be reset. + */ + xdr_init_decode(xdr, rcv_buf, p, rqstp); + + cred->cr_auth->au_rslack = cred->cr_auth->au_verfsize + 2 + + XDR_QUADLEN(savedlen - rcv_buf->len); return 0; -} - -static int -gss_unwrap_req_decode(kxdrdproc_t decode, struct rpc_rqst *rqstp, - __be32 *p, void *obj) -{ - struct xdr_stream xdr; - - xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p, rqstp); - return decode(rqstp, &xdr, obj); +unwrap_failed: + return -EIO; +bad_unwrap: + dprintk("RPC: %s: gss_unwrap returned error 0x%08x\n", + __func__, maj_stat); + return -EIO; } static bool @@ -2037,39 +2059,33 @@ out: } static int -gss_unwrap_resp(struct rpc_task *task, - kxdrdproc_t decode, void *rqstp, __be32 *p, void *obj) +gss_unwrap_resp(struct rpc_task *task, struct xdr_stream *xdr) { - struct rpc_cred *cred = task->tk_rqstp->rq_cred; + struct rpc_rqst *rqstp = task->tk_rqstp; + struct rpc_cred *cred = rqstp->rq_cred; struct gss_cred *gss_cred = container_of(cred, struct gss_cred, gc_base); struct gss_cl_ctx *ctx = gss_cred_get_ctx(cred); - __be32 *savedp = p; - struct kvec *head = ((struct rpc_rqst *)rqstp)->rq_rcv_buf.head; - int savedlen = head->iov_len; - int status = -EIO; + int status = -EIO; if (ctx->gc_proc != RPC_GSS_PROC_DATA) goto out_decode; switch (gss_cred->gc_service) { case RPC_GSS_SVC_NONE: + status = gss_unwrap_resp_auth(cred); break; case RPC_GSS_SVC_INTEGRITY: - status = gss_unwrap_resp_integ(cred, ctx, rqstp, &p); - if (status) - goto out; + status = gss_unwrap_resp_integ(cred, ctx, rqstp, xdr); break; case RPC_GSS_SVC_PRIVACY: - status = gss_unwrap_resp_priv(cred, ctx, rqstp, &p); - if (status) - goto out; + status = gss_unwrap_resp_priv(cred, ctx, rqstp, xdr); break; } - /* take into account extra slack for integrity and privacy cases: */ - cred->cr_auth->au_rslack = cred->cr_auth->au_verfsize + (p - savedp) - + (savedlen - head->iov_len); + if (status) + goto out; + out_decode: - status = gss_unwrap_req_decode(decode, rqstp, p, obj); + status = rpcauth_unwrap_resp_decode(task, xdr); out: gss_put_ctx(ctx); dprintk("RPC: %5u %s returning %d\n", diff --git a/net/sunrpc/auth_null.c b/net/sunrpc/auth_null.c index 797f8472c21b..bf96975ffc4b 100644 --- a/net/sunrpc/auth_null.c +++ b/net/sunrpc/auth_null.c @@ -86,25 +86,19 @@ nul_refresh(struct rpc_task *task) return 0; } -static __be32 * -nul_validate(struct rpc_task *task, __be32 *p) +static int +nul_validate(struct rpc_task *task, struct xdr_stream *xdr) { - rpc_authflavor_t flavor; - u32 size; + __be32 *p; - flavor = ntohl(*p++); - if (flavor != RPC_AUTH_NULL) { - printk("RPC: bad verf flavor: %u\n", flavor); - return ERR_PTR(-EIO); - } - - size = ntohl(*p++); - if (size != 0) { - printk("RPC: bad verf size: %u\n", size); - return ERR_PTR(-EIO); - } - - return p; + p = xdr_inline_decode(xdr, 2 * sizeof(*p)); + if (!p) + return -EIO; + if (*p++ != rpc_auth_null) + return -EIO; + if (*p != xdr_zero) + return -EIO; + return 0; } const struct rpc_authops authnull_ops = { @@ -134,6 +128,7 @@ const struct rpc_credops null_credops = { .crwrap_req = rpcauth_wrap_req_encode, .crrefresh = nul_refresh, .crvalidate = nul_validate, + .crunwrap_resp = rpcauth_unwrap_resp_decode, }; static diff --git a/net/sunrpc/auth_unix.c b/net/sunrpc/auth_unix.c index 1d5b7ed9c6f7..5ea84a96f96e 100644 --- a/net/sunrpc/auth_unix.c +++ b/net/sunrpc/auth_unix.c @@ -160,29 +160,32 @@ unx_refresh(struct rpc_task *task) return 0; } -static __be32 * -unx_validate(struct rpc_task *task, __be32 *p) +static int +unx_validate(struct rpc_task *task, struct xdr_stream *xdr) { - rpc_authflavor_t flavor; - u32 size; + __be32 *p; + u32 size; - flavor = ntohl(*p++); - if (flavor != RPC_AUTH_NULL && - flavor != RPC_AUTH_UNIX && - flavor != RPC_AUTH_SHORT) { - printk("RPC: bad verf flavor: %u\n", flavor); - return ERR_PTR(-EIO); + p = xdr_inline_decode(xdr, 2 * sizeof(*p)); + if (!p) + return -EIO; + switch (*p++) { + case rpc_auth_null: + case rpc_auth_unix: + case rpc_auth_short: + break; + default: + return -EIO; } + size = be32_to_cpup(p); + if (size > RPC_MAX_AUTH_SIZE) + return -EIO; + p = xdr_inline_decode(xdr, size); + if (!p) + return -EIO; - size = ntohl(*p++); - if (size > RPC_MAX_AUTH_SIZE) { - printk("RPC: giant verf size: %u\n", size); - return ERR_PTR(-EIO); - } task->tk_rqstp->rq_cred->cr_auth->au_rslack = (size >> 2) + 2; - p += (size >> 2); - - return p; + return 0; } int __init rpc_init_authunix(void) @@ -223,4 +226,5 @@ const struct rpc_credops unix_credops = { .crwrap_req = rpcauth_wrap_req_encode, .crrefresh = unx_refresh, .crvalidate = unx_validate, + .crunwrap_resp = rpcauth_unwrap_resp_decode, }; diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index e9735089bd66..803e93105af1 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -79,7 +79,8 @@ static void call_connect_status(struct rpc_task *task); static int rpc_encode_header(struct rpc_task *task, struct xdr_stream *xdr); -static __be32 *rpc_decode_header(struct rpc_task *task); +static int rpc_decode_header(struct rpc_task *task, + struct xdr_stream *xdr); static int rpc_ping(struct rpc_clnt *clnt); static void rpc_register_client(struct rpc_clnt *clnt) @@ -2251,12 +2252,11 @@ call_decode(struct rpc_task *task) { struct rpc_clnt *clnt = task->tk_client; struct rpc_rqst *req = task->tk_rqstp; - kxdrdproc_t decode = task->tk_msg.rpc_proc->p_decode; - __be32 *p; + struct xdr_stream xdr; dprint_status(task); - if (!decode) { + if (!task->tk_msg.rpc_proc->p_decode) { task->tk_action = rpc_exit_task; return; } @@ -2292,29 +2292,27 @@ call_decode(struct rpc_task *task) goto out_retry; } - p = rpc_decode_header(task); - if (IS_ERR(p)) { - if (p == ERR_PTR(-EAGAIN)) - goto out_retry; + xdr_init_decode(&xdr, &req->rq_rcv_buf, + req->rq_rcv_buf.head[0].iov_base, req); + switch (rpc_decode_header(task, &xdr)) { + case 0: + task->tk_action = rpc_exit_task; + task->tk_status = rpcauth_unwrap_resp(task, &xdr); + dprintk("RPC: %5u %s result %d\n", + task->tk_pid, __func__, task->tk_status); return; - } - task->tk_action = rpc_exit_task; - - task->tk_status = rpcauth_unwrap_resp(task, decode, req, p, - task->tk_msg.rpc_resp); - - dprintk("RPC: %5u call_decode result %d\n", task->tk_pid, - task->tk_status); - return; + case -EAGAIN: out_retry: - task->tk_status = 0; - /* Note: rpc_decode_header() may have freed the RPC slot */ - if (task->tk_rqstp == req) { - xdr_free_bvec(&req->rq_rcv_buf); - req->rq_reply_bytes_recvd = req->rq_rcv_buf.len = 0; - if (task->tk_client->cl_discrtry) - xprt_conditional_disconnect(req->rq_xprt, - req->rq_connect_cookie); + task->tk_status = 0; + /* Note: rpc_decode_header() may have freed the RPC slot */ + if (task->tk_rqstp == req) { + xdr_free_bvec(&req->rq_rcv_buf); + req->rq_reply_bytes_recvd = 0; + req->rq_rcv_buf.len = 0; + if (task->tk_client->cl_discrtry) + xprt_conditional_disconnect(req->rq_xprt, + req->rq_connect_cookie); + } } } @@ -2347,14 +2345,12 @@ out_fail: return error; } -static noinline __be32 * -rpc_decode_header(struct rpc_task *task) +static noinline int +rpc_decode_header(struct rpc_task *task, struct xdr_stream *xdr) { struct rpc_clnt *clnt = task->tk_client; - struct kvec *iov = &task->tk_rqstp->rq_rcv_buf.head[0]; - int len = task->tk_rqstp->rq_rcv_buf.len >> 2; - __be32 *p = iov->iov_base; int error = -EACCES; + __be32 *p; /* RFC-1014 says that the representation of XDR data must be a * multiple of four bytes @@ -2363,25 +2359,26 @@ rpc_decode_header(struct rpc_task *task) */ if (task->tk_rqstp->rq_rcv_buf.len & 3) goto out_badlen; - if ((len -= 3) < 0) - goto out_unparsable; + p = xdr_inline_decode(xdr, 3 * sizeof(*p)); + if (!p) + goto out_unparsable; p++; /* skip XID */ if (*p++ != rpc_reply) goto out_unparsable; if (*p++ != rpc_msg_accepted) goto out_msg_denied; - p = rpcauth_checkverf(task, p); - if (IS_ERR(p)) + error = rpcauth_checkverf(task, xdr); + if (error) goto out_verifier; - len = p - (__be32 *)iov->iov_base - 1; - if (len < 0) + p = xdr_inline_decode(xdr, sizeof(*p)); + if (!p) goto out_unparsable; - switch (*p++) { + switch (*p) { case rpc_success: - return p; + return 0; case rpc_prog_unavail: trace_rpc__prog_unavail(task); error = -EPFNOSUPPORT; @@ -2406,11 +2403,11 @@ out_garbage: if (task->tk_garb_retry) { task->tk_garb_retry--; task->tk_action = call_encode; - return ERR_PTR(-EAGAIN); + return -EAGAIN; } out_err: rpc_exit(task, error); - return ERR_PTR(error); + return error; out_badlen: trace_rpc__unparsable(task); @@ -2424,10 +2421,12 @@ out_unparsable: out_verifier: trace_rpc_bad_verifier(task); - error = PTR_ERR(p); goto out_garbage; out_msg_denied: + p = xdr_inline_decode(xdr, sizeof(*p)); + if (!p) + goto out_unparsable; switch (*p++) { case rpc_auth_error: break; @@ -2441,6 +2440,9 @@ out_msg_denied: goto out_err; } + p = xdr_inline_decode(xdr, sizeof(*p)); + if (!p) + goto out_unparsable; switch (*p++) { case rpc_autherr_rejectedcred: case rpc_autherr_rejectedverf: @@ -2454,7 +2456,7 @@ out_msg_denied: /* Ensure we obtain a new XID! */ xprt_release(task); task->tk_action = call_reserve; - return ERR_PTR(-EAGAIN); + return -EAGAIN; case rpc_autherr_badcred: case rpc_autherr_badverf: /* possibly garbled cred/verf? */ @@ -2463,7 +2465,7 @@ out_msg_denied: task->tk_garb_retry--; trace_rpc__bad_creds(task); task->tk_action = call_encode; - return ERR_PTR(-EAGAIN); + return -EAGAIN; case rpc_autherr_tooweak: trace_rpc__auth_tooweak(task); pr_warn("RPC: server %s requires stronger authentication.\n", From 0c77668ddb4e7bdfbca462c6185d154d0b8889ae Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:25:04 -0500 Subject: [PATCH 16/90] SUNRPC: Introduce trace points in rpc_auth_gss.ko Add infrastructure for trace points in the RPC_AUTH_GSS kernel module, and add a few sample trace points. These report exceptional or unexpected events, and observe the assignment of GSS sequence numbers. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/trace/events/rpcgss.h | 361 +++++++++++++++++++++++++++++++++ include/trace/events/rpcrdma.h | 12 +- include/trace/events/sunrpc.h | 61 +++++- net/sunrpc/auth_gss/Makefile | 2 +- net/sunrpc/auth_gss/auth_gss.c | 165 ++++++++------- net/sunrpc/auth_gss/trace.c | 11 + net/sunrpc/xprt.c | 10 +- 7 files changed, 530 insertions(+), 92 deletions(-) create mode 100644 include/trace/events/rpcgss.h create mode 100644 net/sunrpc/auth_gss/trace.c diff --git a/include/trace/events/rpcgss.h b/include/trace/events/rpcgss.h new file mode 100644 index 000000000000..d1f7fe1b6fe4 --- /dev/null +++ b/include/trace/events/rpcgss.h @@ -0,0 +1,361 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Copyright (c) 2018 Oracle. All rights reserved. + * + * Trace point definitions for the "rpcgss" subsystem. + */ + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM rpcgss + +#if !defined(_TRACE_RPCRDMA_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_RPCGSS_H + +#include + +/** + ** GSS-API related trace events + **/ + +TRACE_DEFINE_ENUM(GSS_S_BAD_MECH); +TRACE_DEFINE_ENUM(GSS_S_BAD_NAME); +TRACE_DEFINE_ENUM(GSS_S_BAD_NAMETYPE); +TRACE_DEFINE_ENUM(GSS_S_BAD_BINDINGS); +TRACE_DEFINE_ENUM(GSS_S_BAD_STATUS); +TRACE_DEFINE_ENUM(GSS_S_BAD_SIG); +TRACE_DEFINE_ENUM(GSS_S_NO_CRED); +TRACE_DEFINE_ENUM(GSS_S_NO_CONTEXT); +TRACE_DEFINE_ENUM(GSS_S_DEFECTIVE_TOKEN); +TRACE_DEFINE_ENUM(GSS_S_DEFECTIVE_CREDENTIAL); +TRACE_DEFINE_ENUM(GSS_S_CREDENTIALS_EXPIRED); +TRACE_DEFINE_ENUM(GSS_S_CONTEXT_EXPIRED); +TRACE_DEFINE_ENUM(GSS_S_FAILURE); +TRACE_DEFINE_ENUM(GSS_S_BAD_QOP); +TRACE_DEFINE_ENUM(GSS_S_UNAUTHORIZED); +TRACE_DEFINE_ENUM(GSS_S_UNAVAILABLE); +TRACE_DEFINE_ENUM(GSS_S_DUPLICATE_ELEMENT); +TRACE_DEFINE_ENUM(GSS_S_NAME_NOT_MN); +TRACE_DEFINE_ENUM(GSS_S_CONTINUE_NEEDED); +TRACE_DEFINE_ENUM(GSS_S_DUPLICATE_TOKEN); +TRACE_DEFINE_ENUM(GSS_S_OLD_TOKEN); +TRACE_DEFINE_ENUM(GSS_S_UNSEQ_TOKEN); +TRACE_DEFINE_ENUM(GSS_S_GAP_TOKEN); + +#define show_gss_status(x) \ + __print_flags(x, "|", \ + { GSS_S_BAD_MECH, "GSS_S_BAD_MECH" }, \ + { GSS_S_BAD_NAME, "GSS_S_BAD_NAME" }, \ + { GSS_S_BAD_NAMETYPE, "GSS_S_BAD_NAMETYPE" }, \ + { GSS_S_BAD_BINDINGS, "GSS_S_BAD_BINDINGS" }, \ + { GSS_S_BAD_STATUS, "GSS_S_BAD_STATUS" }, \ + { GSS_S_BAD_SIG, "GSS_S_BAD_SIG" }, \ + { GSS_S_NO_CRED, "GSS_S_NO_CRED" }, \ + { GSS_S_NO_CONTEXT, "GSS_S_NO_CONTEXT" }, \ + { GSS_S_DEFECTIVE_TOKEN, "GSS_S_DEFECTIVE_TOKEN" }, \ + { GSS_S_DEFECTIVE_CREDENTIAL, "GSS_S_DEFECTIVE_CREDENTIAL" }, \ + { GSS_S_CREDENTIALS_EXPIRED, "GSS_S_CREDENTIALS_EXPIRED" }, \ + { GSS_S_CONTEXT_EXPIRED, "GSS_S_CONTEXT_EXPIRED" }, \ + { GSS_S_FAILURE, "GSS_S_FAILURE" }, \ + { GSS_S_BAD_QOP, "GSS_S_BAD_QOP" }, \ + { GSS_S_UNAUTHORIZED, "GSS_S_UNAUTHORIZED" }, \ + { GSS_S_UNAVAILABLE, "GSS_S_UNAVAILABLE" }, \ + { GSS_S_DUPLICATE_ELEMENT, "GSS_S_DUPLICATE_ELEMENT" }, \ + { GSS_S_NAME_NOT_MN, "GSS_S_NAME_NOT_MN" }, \ + { GSS_S_CONTINUE_NEEDED, "GSS_S_CONTINUE_NEEDED" }, \ + { GSS_S_DUPLICATE_TOKEN, "GSS_S_DUPLICATE_TOKEN" }, \ + { GSS_S_OLD_TOKEN, "GSS_S_OLD_TOKEN" }, \ + { GSS_S_UNSEQ_TOKEN, "GSS_S_UNSEQ_TOKEN" }, \ + { GSS_S_GAP_TOKEN, "GSS_S_GAP_TOKEN" }) + + +DECLARE_EVENT_CLASS(rpcgss_gssapi_event, + TP_PROTO( + const struct rpc_task *task, + u32 maj_stat + ), + + TP_ARGS(task, maj_stat), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + __field(u32, maj_stat) + + ), + + TP_fast_assign( + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; + __entry->maj_stat = maj_stat; + ), + + TP_printk("task:%u@%u maj_stat=%s", + __entry->task_id, __entry->client_id, + __entry->maj_stat == 0 ? + "GSS_S_COMPLETE" : show_gss_status(__entry->maj_stat)) +); + +#define DEFINE_GSSAPI_EVENT(name) \ + DEFINE_EVENT(rpcgss_gssapi_event, rpcgss_##name, \ + TP_PROTO( \ + const struct rpc_task *task, \ + u32 maj_stat \ + ), \ + TP_ARGS(task, maj_stat)) + +TRACE_EVENT(rpcgss_import_ctx, + TP_PROTO( + int status + ), + + TP_ARGS(status), + + TP_STRUCT__entry( + __field(int, status) + ), + + TP_fast_assign( + __entry->status = status; + ), + + TP_printk("status=%d", __entry->status) +); + +DEFINE_GSSAPI_EVENT(get_mic); +DEFINE_GSSAPI_EVENT(verify_mic); +DEFINE_GSSAPI_EVENT(wrap); +DEFINE_GSSAPI_EVENT(unwrap); + + +/** + ** GSS auth unwrap failures + **/ + +TRACE_EVENT(rpcgss_unwrap_failed, + TP_PROTO( + const struct rpc_task *task + ), + + TP_ARGS(task), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + ), + + TP_fast_assign( + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; + ), + + TP_printk("task:%u@%u", __entry->task_id, __entry->client_id) +); + +TRACE_EVENT(rpcgss_bad_seqno, + TP_PROTO( + const struct rpc_task *task, + u32 expected, + u32 received + ), + + TP_ARGS(task, expected, received), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + __field(u32, expected) + __field(u32, received) + ), + + TP_fast_assign( + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; + __entry->expected = expected; + __entry->received = received; + ), + + TP_printk("task:%u@%u expected seqno %u, received seqno %u", + __entry->task_id, __entry->client_id, + __entry->expected, __entry->received) +); + +TRACE_EVENT(rpcgss_seqno, + TP_PROTO( + const struct rpc_task *task + ), + + TP_ARGS(task), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + __field(u32, xid) + __field(u32, seqno) + ), + + TP_fast_assign( + const struct rpc_rqst *rqst = task->tk_rqstp; + + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; + __entry->xid = be32_to_cpu(rqst->rq_xid); + __entry->seqno = rqst->rq_seqno; + ), + + TP_printk("task:%u@%u xid=0x%08x seqno=%u", + __entry->task_id, __entry->client_id, + __entry->xid, __entry->seqno) +); + +TRACE_EVENT(rpcgss_need_reencode, + TP_PROTO( + const struct rpc_task *task, + u32 seq_xmit, + bool ret + ), + + TP_ARGS(task, seq_xmit, ret), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + __field(u32, xid) + __field(u32, seq_xmit) + __field(u32, seqno) + __field(bool, ret) + ), + + TP_fast_assign( + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; + __entry->xid = be32_to_cpu(task->tk_rqstp->rq_xid); + __entry->seq_xmit = seq_xmit; + __entry->seqno = task->tk_rqstp->rq_seqno; + __entry->ret = ret; + ), + + TP_printk("task:%u@%u xid=0x%08x rq_seqno=%u seq_xmit=%u reencode %sneeded", + __entry->task_id, __entry->client_id, + __entry->xid, __entry->seqno, __entry->seq_xmit, + __entry->ret ? "" : "un") +); + +/** + ** gssd upcall related trace events + **/ + +TRACE_EVENT(rpcgss_upcall_msg, + TP_PROTO( + const char *buf + ), + + TP_ARGS(buf), + + TP_STRUCT__entry( + __string(msg, buf) + ), + + TP_fast_assign( + __assign_str(msg, buf) + ), + + TP_printk("msg='%s'", __get_str(msg)) +); + +TRACE_EVENT(rpcgss_upcall_result, + TP_PROTO( + u32 uid, + int result + ), + + TP_ARGS(uid, result), + + TP_STRUCT__entry( + __field(u32, uid) + __field(int, result) + + ), + + TP_fast_assign( + __entry->uid = uid; + __entry->result = result; + ), + + TP_printk("for uid %u, result=%d", __entry->uid, __entry->result) +); + +TRACE_EVENT(rpcgss_context, + TP_PROTO( + unsigned long expiry, + unsigned long now, + unsigned int timeout, + unsigned int len, + const u8 *data + ), + + TP_ARGS(expiry, now, timeout, len, data), + + TP_STRUCT__entry( + __field(unsigned long, expiry) + __field(unsigned long, now) + __field(unsigned int, timeout) + __field(int, len) + __string(acceptor, data) + ), + + TP_fast_assign( + __entry->expiry = expiry; + __entry->now = now; + __entry->timeout = timeout; + __entry->len = len; + strncpy(__get_str(acceptor), data, len); + ), + + TP_printk("gc_expiry=%lu now=%lu timeout=%u acceptor=%.*s", + __entry->expiry, __entry->now, __entry->timeout, + __entry->len, __get_str(acceptor)) +); + + +/** + ** Miscellaneous events + */ + +TRACE_DEFINE_ENUM(RPC_AUTH_GSS_KRB5); +TRACE_DEFINE_ENUM(RPC_AUTH_GSS_KRB5I); +TRACE_DEFINE_ENUM(RPC_AUTH_GSS_KRB5P); + +#define show_pseudoflavor(x) \ + __print_symbolic(x, \ + { RPC_AUTH_GSS_KRB5, "RPC_AUTH_GSS_KRB5" }, \ + { RPC_AUTH_GSS_KRB5I, "RPC_AUTH_GSS_KRB5I" }, \ + { RPC_AUTH_GSS_KRB5P, "RPC_AUTH_GSS_KRB5P" }) + + +TRACE_EVENT(rpcgss_createauth, + TP_PROTO( + unsigned int flavor, + int error + ), + + TP_ARGS(flavor, error), + + TP_STRUCT__entry( + __field(unsigned int, flavor) + __field(int, error) + + ), + + TP_fast_assign( + __entry->flavor = flavor; + __entry->error = error; + ), + + TP_printk("flavor=%s error=%d", + show_pseudoflavor(__entry->flavor), __entry->error) +); + + +#endif /* _TRACE_RPCGSS_H */ + +#include diff --git a/include/trace/events/rpcrdma.h b/include/trace/events/rpcrdma.h index 399b1aedc927..962975b4313f 100644 --- a/include/trace/events/rpcrdma.h +++ b/include/trace/events/rpcrdma.h @@ -521,12 +521,18 @@ TRACE_EVENT(xprtrdma_post_send, TP_STRUCT__entry( __field(const void *, req) + __field(unsigned int, task_id) + __field(unsigned int, client_id) __field(int, num_sge) __field(int, signaled) __field(int, status) ), TP_fast_assign( + const struct rpc_rqst *rqst = &req->rl_slot; + + __entry->task_id = rqst->rq_task->tk_pid; + __entry->client_id = rqst->rq_task->tk_client->cl_clid; __entry->req = req; __entry->num_sge = req->rl_sendctx->sc_wr.num_sge; __entry->signaled = req->rl_sendctx->sc_wr.send_flags & @@ -534,9 +540,11 @@ TRACE_EVENT(xprtrdma_post_send, __entry->status = status; ), - TP_printk("req=%p, %d SGEs%s, status=%d", + TP_printk("task:%u@%u req=%p (%d SGE%s) %sstatus=%d", + __entry->task_id, __entry->client_id, __entry->req, __entry->num_sge, - (__entry->signaled ? ", signaled" : ""), + (__entry->num_sge == 1 ? "" : "s"), + (__entry->signaled ? "signaled " : ""), __entry->status ) ); diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 0654e9c50371..e58dda8e038c 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -655,9 +655,68 @@ DECLARE_EVENT_CLASS(rpc_xprt_event, DEFINE_RPC_XPRT_EVENT(timer); DEFINE_RPC_XPRT_EVENT(lookup_rqst); -DEFINE_RPC_XPRT_EVENT(transmit); DEFINE_RPC_XPRT_EVENT(complete_rqst); +TRACE_EVENT(xprt_transmit, + TP_PROTO( + const struct rpc_rqst *rqst, + int status + ), + + TP_ARGS(rqst, status), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + __field(u32, xid) + __field(u32, seqno) + __field(int, status) + ), + + TP_fast_assign( + __entry->task_id = rqst->rq_task->tk_pid; + __entry->client_id = rqst->rq_task->tk_client->cl_clid; + __entry->xid = be32_to_cpu(rqst->rq_xid); + __entry->seqno = rqst->rq_seqno; + __entry->status = status; + ), + + TP_printk( + "task:%u@%u xid=0x%08x seqno=%u status=%d", + __entry->task_id, __entry->client_id, __entry->xid, + __entry->seqno, __entry->status) +); + +TRACE_EVENT(xprt_enq_xmit, + TP_PROTO( + const struct rpc_task *task, + int stage + ), + + TP_ARGS(task, stage), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + __field(u32, xid) + __field(u32, seqno) + __field(int, stage) + ), + + TP_fast_assign( + __entry->task_id = task->tk_pid; + __entry->client_id = task->tk_client->cl_clid; + __entry->xid = be32_to_cpu(task->tk_rqstp->rq_xid); + __entry->seqno = task->tk_rqstp->rq_seqno; + __entry->stage = stage; + ), + + TP_printk( + "task:%u@%u xid=0x%08x seqno=%u stage=%d", + __entry->task_id, __entry->client_id, __entry->xid, + __entry->seqno, __entry->stage) +); + TRACE_EVENT(xprt_ping, TP_PROTO(const struct rpc_xprt *xprt, int status), diff --git a/net/sunrpc/auth_gss/Makefile b/net/sunrpc/auth_gss/Makefile index c374268b008f..4a29f4c5dac4 100644 --- a/net/sunrpc/auth_gss/Makefile +++ b/net/sunrpc/auth_gss/Makefile @@ -7,7 +7,7 @@ obj-$(CONFIG_SUNRPC_GSS) += auth_rpcgss.o auth_rpcgss-y := auth_gss.o gss_generic_token.o \ gss_mech_switch.o svcauth_gss.o \ - gss_rpc_upcall.o gss_rpc_xdr.o + gss_rpc_upcall.o gss_rpc_xdr.o trace.o obj-$(CONFIG_RPCSEC_GSS_KRB5) += rpcsec_gss_krb5.o diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 206788e8b787..3d1fbd6d3370 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -55,6 +55,8 @@ #include "../netns.h" +#include + static const struct rpc_authops authgss_ops; static const struct rpc_credops gss_credops; @@ -260,6 +262,7 @@ gss_fill_context(const void *p, const void *end, struct gss_cl_ctx *ctx, struct } ret = gss_import_sec_context(p, seclen, gm, &ctx->gc_gss_ctx, NULL, GFP_NOFS); if (ret < 0) { + trace_rpcgss_import_ctx(ret); p = ERR_PTR(ret); goto err; } @@ -275,12 +278,9 @@ gss_fill_context(const void *p, const void *end, struct gss_cl_ctx *ctx, struct if (IS_ERR(p)) goto err; done: - dprintk("RPC: %s Success. gc_expiry %lu now %lu timeout %u acceptor %.*s\n", - __func__, ctx->gc_expiry, now, timeout, ctx->gc_acceptor.len, - ctx->gc_acceptor.data); - return p; + trace_rpcgss_context(ctx->gc_expiry, now, timeout, + ctx->gc_acceptor.len, ctx->gc_acceptor.data); err: - dprintk("RPC: %s returns error %ld\n", __func__, -PTR_ERR(p)); return p; } @@ -354,10 +354,8 @@ __gss_find_upcall(struct rpc_pipe *pipe, kuid_t uid, const struct gss_auth *auth if (auth && pos->auth->service != auth->service) continue; refcount_inc(&pos->count); - dprintk("RPC: %s found msg %p\n", __func__, pos); return pos; } - dprintk("RPC: %s found nothing\n", __func__); return NULL; } @@ -456,7 +454,7 @@ static int gss_encode_v1_msg(struct gss_upcall_msg *gss_msg, size_t buflen = sizeof(gss_msg->databuf); int len; - len = scnprintf(p, buflen, "mech=%s uid=%d ", mech->gm_name, + len = scnprintf(p, buflen, "mech=%s uid=%d", mech->gm_name, from_kuid(&init_user_ns, gss_msg->uid)); buflen -= len; p += len; @@ -467,7 +465,7 @@ static int gss_encode_v1_msg(struct gss_upcall_msg *gss_msg, * identity that we are authenticating to. */ if (target_name) { - len = scnprintf(p, buflen, "target=%s ", target_name); + len = scnprintf(p, buflen, " target=%s", target_name); buflen -= len; p += len; gss_msg->msg.len += len; @@ -487,11 +485,11 @@ static int gss_encode_v1_msg(struct gss_upcall_msg *gss_msg, char *c = strchr(service_name, '@'); if (!c) - len = scnprintf(p, buflen, "service=%s ", + len = scnprintf(p, buflen, " service=%s", service_name); else len = scnprintf(p, buflen, - "service=%.*s srchost=%s ", + " service=%.*s srchost=%s", (int)(c - service_name), service_name, c + 1); buflen -= len; @@ -500,17 +498,17 @@ static int gss_encode_v1_msg(struct gss_upcall_msg *gss_msg, } if (mech->gm_upcall_enctypes) { - len = scnprintf(p, buflen, "enctypes=%s ", + len = scnprintf(p, buflen, " enctypes=%s", mech->gm_upcall_enctypes); buflen -= len; p += len; gss_msg->msg.len += len; } + trace_rpcgss_upcall_msg(gss_msg->databuf); len = scnprintf(p, buflen, "\n"); if (len == 0) goto out_overflow; gss_msg->msg.len += len; - gss_msg->msg.data = gss_msg->databuf; return 0; out_overflow: @@ -603,8 +601,6 @@ gss_refresh_upcall(struct rpc_task *task) struct rpc_pipe *pipe; int err = 0; - dprintk("RPC: %5u %s for uid %u\n", - task->tk_pid, __func__, from_kuid(&init_user_ns, cred->cr_cred->fsuid)); gss_msg = gss_setup_upcall(gss_auth, cred); if (PTR_ERR(gss_msg) == -EAGAIN) { /* XXX: warning on the first, under the assumption we @@ -612,7 +608,8 @@ gss_refresh_upcall(struct rpc_task *task) warn_gssd(); task->tk_timeout = 15*HZ; rpc_sleep_on(&pipe_version_rpc_waitqueue, task, NULL); - return -EAGAIN; + err = -EAGAIN; + goto out; } if (IS_ERR(gss_msg)) { err = PTR_ERR(gss_msg); @@ -635,9 +632,8 @@ gss_refresh_upcall(struct rpc_task *task) spin_unlock(&pipe->lock); gss_release_msg(gss_msg); out: - dprintk("RPC: %5u %s for uid %u result %d\n", - task->tk_pid, __func__, - from_kuid(&init_user_ns, cred->cr_cred->fsuid), err); + trace_rpcgss_upcall_result(from_kuid(&init_user_ns, + cred->cr_cred->fsuid), err); return err; } @@ -652,14 +648,13 @@ gss_create_upcall(struct gss_auth *gss_auth, struct gss_cred *gss_cred) DEFINE_WAIT(wait); int err; - dprintk("RPC: %s for uid %u\n", - __func__, from_kuid(&init_user_ns, cred->cr_cred->fsuid)); retry: err = 0; /* if gssd is down, just skip upcalling altogether */ if (!gssd_running(net)) { warn_gssd(); - return -EACCES; + err = -EACCES; + goto out; } gss_msg = gss_setup_upcall(gss_auth, cred); if (PTR_ERR(gss_msg) == -EAGAIN) { @@ -700,8 +695,8 @@ out_intr: finish_wait(&gss_msg->waitqueue, &wait); gss_release_msg(gss_msg); out: - dprintk("RPC: %s for uid %u result %d\n", - __func__, from_kuid(&init_user_ns, cred->cr_cred->fsuid), err); + trace_rpcgss_upcall_result(from_kuid(&init_user_ns, + cred->cr_cred->fsuid), err); return err; } @@ -794,7 +789,6 @@ err_put_ctx: err: kfree(buf); out: - dprintk("RPC: %s returning %zd\n", __func__, err); return err; } @@ -863,8 +857,6 @@ gss_pipe_destroy_msg(struct rpc_pipe_msg *msg) struct gss_upcall_msg *gss_msg = container_of(msg, struct gss_upcall_msg, msg); if (msg->errno < 0) { - dprintk("RPC: %s releasing msg %p\n", - __func__, gss_msg); refcount_inc(&gss_msg->count); gss_unhash_msg(gss_msg); if (msg->errno == -ETIMEDOUT) @@ -1024,8 +1016,6 @@ gss_create_new(const struct rpc_auth_create_args *args, struct rpc_clnt *clnt) struct rpc_auth * auth; int err = -ENOMEM; /* XXX? */ - dprintk("RPC: creating GSS authenticator for client %p\n", clnt); - if (!try_module_get(THIS_MODULE)) return ERR_PTR(err); if (!(gss_auth = kmalloc(sizeof(*gss_auth), GFP_KERNEL))) @@ -1041,10 +1031,8 @@ gss_create_new(const struct rpc_auth_create_args *args, struct rpc_clnt *clnt) gss_auth->net = get_net(rpc_net_ns(clnt)); err = -EINVAL; gss_auth->mech = gss_mech_get_by_pseudoflavor(flavor); - if (!gss_auth->mech) { - dprintk("RPC: Pseudoflavor %d not found!\n", flavor); + if (!gss_auth->mech) goto err_put_net; - } gss_auth->service = gss_pseudoflavor_to_service(gss_auth->mech, flavor); if (gss_auth->service == 0) goto err_put_mech; @@ -1099,6 +1087,7 @@ err_free: kfree(gss_auth); out_dec: module_put(THIS_MODULE); + trace_rpcgss_createauth(flavor, err); return ERR_PTR(err); } @@ -1135,9 +1124,6 @@ gss_destroy(struct rpc_auth *auth) struct gss_auth *gss_auth = container_of(auth, struct gss_auth, rpc_auth); - dprintk("RPC: destroying GSS authenticator %p flavor %d\n", - auth, auth->au_flavor); - if (hash_hashed(&gss_auth->hash)) { spin_lock(&gss_auth_hash_lock); hash_del(&gss_auth->hash); @@ -1300,8 +1286,6 @@ gss_send_destroy_context(struct rpc_cred *cred) static void gss_do_free_ctx(struct gss_cl_ctx *ctx) { - dprintk("RPC: %s\n", __func__); - gss_delete_sec_context(&ctx->gc_gss_ctx); kfree(ctx->gc_wire_ctx.data); kfree(ctx->gc_acceptor.data); @@ -1324,7 +1308,6 @@ gss_free_ctx(struct gss_cl_ctx *ctx) static void gss_free_cred(struct gss_cred *gss_cred) { - dprintk("RPC: %s cred=%p\n", __func__, gss_cred); kfree(gss_cred); } @@ -1381,10 +1364,6 @@ gss_create_cred(struct rpc_auth *auth, struct auth_cred *acred, int flags, gfp_t struct gss_cred *cred = NULL; int err = -ENOMEM; - dprintk("RPC: %s for uid %d, flavor %d\n", - __func__, from_kuid(&init_user_ns, acred->cred->fsuid), - auth->au_flavor); - if (!(cred = kzalloc(sizeof(*cred), gfp))) goto out_err; @@ -1400,7 +1379,6 @@ gss_create_cred(struct rpc_auth *auth, struct auth_cred *acred, int flags, gfp_t return &cred->gc_base; out_err: - dprintk("RPC: %s failed with error %d\n", __func__, err); return ERR_PTR(err); } @@ -1544,15 +1522,14 @@ static int gss_marshal(struct rpc_task *task, struct xdr_stream *xdr) struct xdr_netobj mic; struct kvec iov; struct xdr_buf verf_buf; - - dprintk("RPC: %5u %s\n", task->tk_pid, __func__); + int status; /* Credential */ p = xdr_reserve_space(xdr, 7 * sizeof(*p) + ctx->gc_wire_ctx.len); if (!p) - goto out_put_ctx; + goto marshal_failed; *p++ = rpc_auth_gss; cred_len = p++; @@ -1560,7 +1537,8 @@ static int gss_marshal(struct rpc_task *task, struct xdr_stream *xdr) req->rq_seqno = (ctx->gc_seq < MAXSEQ) ? ctx->gc_seq++ : MAXSEQ; spin_unlock(&ctx->gc_seq_lock); if (req->rq_seqno == MAXSEQ) - goto out_expired; + goto expired; + trace_rpcgss_seqno(task); *p++ = cpu_to_be32(RPC_GSS_VERSION); *p++ = cpu_to_be32(ctx->gc_proc); @@ -1579,25 +1557,31 @@ static int gss_marshal(struct rpc_task *task, struct xdr_stream *xdr) p = xdr_reserve_space(xdr, sizeof(*p)); if (!p) - goto out_put_ctx; + goto marshal_failed; *p++ = rpc_auth_gss; mic.data = (u8 *)(p + 1); maj_stat = gss_get_mic(ctx->gc_gss_ctx, &verf_buf, &mic); if (maj_stat == GSS_S_CONTEXT_EXPIRED) - goto out_expired; + goto expired; else if (maj_stat != 0) - goto out_put_ctx; + goto bad_mic; if (xdr_stream_encode_opaque_inline(xdr, (void **)&p, mic.len) < 0) - goto out_put_ctx; - gss_put_ctx(ctx); - return 0; -out_expired: + goto marshal_failed; + status = 0; +out: gss_put_ctx(ctx); + return status; +expired: clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); - return -EKEYEXPIRED; -out_put_ctx: - gss_put_ctx(ctx); - return -EMSGSIZE; + status = -EKEYEXPIRED; + goto out; +marshal_failed: + status = -EMSGSIZE; + goto out; +bad_mic: + trace_rpcgss_get_mic(task, maj_stat); + status = -EIO; + goto out; } static int gss_renew_cred(struct rpc_task *task) @@ -1723,8 +1707,7 @@ validate_failed: status = -EIO; goto out; bad_mic: - dprintk("RPC: %5u %s: gss_verify_mic returned error 0x%08x\n", - task->tk_pid, __func__, maj_stat); + trace_rpcgss_verify_mic(task, maj_stat); status = -EACCES; goto out; } @@ -1761,13 +1744,16 @@ static int gss_wrap_req_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, if (maj_stat == GSS_S_CONTEXT_EXPIRED) clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); else if (maj_stat) - goto wrap_failed; + goto bad_mic; /* Check that the trailing MIC fit in the buffer, after the fact */ if (xdr_stream_encode_opaque_inline(xdr, (void **)&p, mic.len) < 0) goto wrap_failed; return 0; wrap_failed: return -EMSGSIZE; +bad_mic: + trace_rpcgss_get_mic(task, maj_stat); + return -EIO; } static void @@ -1860,7 +1846,6 @@ static int gss_wrap_req_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, memcpy(tmp, snd_buf->tail[0].iov_base, snd_buf->tail[0].iov_len); snd_buf->tail[0].iov_base = tmp; } - status = -EIO; offset = (u8 *)p - (u8 *)snd_buf->head[0].iov_base; maj_stat = gss_wrap(ctx->gc_gss_ctx, offset, snd_buf, inpages); /* slack space should prevent this ever happening: */ @@ -1871,7 +1856,7 @@ static int gss_wrap_req_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, if (maj_stat == GSS_S_CONTEXT_EXPIRED) clear_bit(RPCAUTH_CRED_UPTODATE, &cred->cr_flags); else if (maj_stat) - goto wrap_failed; + goto bad_wrap; *opaque_len = cpu_to_be32(snd_buf->len - offset); /* guess whether the pad goes into the head or the tail: */ @@ -1888,6 +1873,9 @@ static int gss_wrap_req_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, return 0; wrap_failed: return status; +bad_wrap: + trace_rpcgss_wrap(task, maj_stat); + return -EIO; } static int gss_wrap_req(struct rpc_task *task, struct xdr_stream *xdr) @@ -1898,7 +1886,6 @@ static int gss_wrap_req(struct rpc_task *task, struct xdr_stream *xdr) struct gss_cl_ctx *ctx = gss_cred_get_ctx(cred); int status; - dprintk("RPC: %5u %s\n", task->tk_pid, __func__); status = -EIO; if (ctx->gc_proc != RPC_GSS_PROC_DATA) { /* The spec seems a little ambiguous here, but I think that not @@ -1917,10 +1904,11 @@ static int gss_wrap_req(struct rpc_task *task, struct xdr_stream *xdr) case RPC_GSS_SVC_PRIVACY: status = gss_wrap_req_priv(cred, ctx, task, xdr); break; + default: + status = -EIO; } out: gss_put_ctx(ctx); - dprintk("RPC: %5u %s returning %d\n", task->tk_pid, __func__, status); return status; } @@ -1932,8 +1920,9 @@ gss_unwrap_resp_auth(struct rpc_cred *cred) } static int -gss_unwrap_resp_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, - struct rpc_rqst *rqstp, struct xdr_stream *xdr) +gss_unwrap_resp_integ(struct rpc_task *task, struct rpc_cred *cred, + struct gss_cl_ctx *ctx, struct rpc_rqst *rqstp, + struct xdr_stream *xdr) { struct xdr_buf integ_buf, *rcv_buf = &rqstp->rq_rcv_buf; u32 data_offset, mic_offset, integ_len, maj_stat; @@ -1951,7 +1940,7 @@ gss_unwrap_resp_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, if (mic_offset > rcv_buf->len) goto unwrap_failed; if (be32_to_cpup(p) != rqstp->rq_seqno) - goto unwrap_failed; + goto bad_seqno; if (xdr_buf_subsegment(rcv_buf, &integ_buf, data_offset, integ_len)) goto unwrap_failed; @@ -1967,16 +1956,20 @@ gss_unwrap_resp_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, 1 + XDR_QUADLEN(mic.len); return 0; unwrap_failed: + trace_rpcgss_unwrap_failed(task); + return -EIO; +bad_seqno: + trace_rpcgss_bad_seqno(task, rqstp->rq_seqno, be32_to_cpup(p)); return -EIO; bad_mic: - dprintk("RPC: %s: gss_verify_mic returned error 0x%08x\n", - __func__, maj_stat); + trace_rpcgss_verify_mic(task, maj_stat); return -EIO; } static int -gss_unwrap_resp_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, - struct rpc_rqst *rqstp, struct xdr_stream *xdr) +gss_unwrap_resp_priv(struct rpc_task *task, struct rpc_cred *cred, + struct gss_cl_ctx *ctx, struct rpc_rqst *rqstp, + struct xdr_stream *xdr) { struct xdr_buf *rcv_buf = &rqstp->rq_rcv_buf; struct kvec *head = rqstp->rq_rcv_buf.head; @@ -2000,7 +1993,7 @@ gss_unwrap_resp_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, goto bad_unwrap; /* gss_unwrap decrypted the sequence number */ if (be32_to_cpup(p++) != rqstp->rq_seqno) - goto unwrap_failed; + goto bad_seqno; /* gss_unwrap redacts the opaque blob from the head iovec. * rcv_buf has changed, thus the stream needs to be reset. @@ -2011,10 +2004,13 @@ gss_unwrap_resp_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, XDR_QUADLEN(savedlen - rcv_buf->len); return 0; unwrap_failed: + trace_rpcgss_unwrap_failed(task); + return -EIO; +bad_seqno: + trace_rpcgss_bad_seqno(task, rqstp->rq_seqno, be32_to_cpup(--p)); return -EIO; bad_unwrap: - dprintk("RPC: %s: gss_unwrap returned error 0x%08x\n", - __func__, maj_stat); + trace_rpcgss_unwrap(task, maj_stat); return -EIO; } @@ -2030,14 +2026,14 @@ gss_xmit_need_reencode(struct rpc_task *task) struct rpc_rqst *req = task->tk_rqstp; struct rpc_cred *cred = req->rq_cred; struct gss_cl_ctx *ctx = gss_cred_get_ctx(cred); - u32 win, seq_xmit; + u32 win, seq_xmit = 0; bool ret = true; if (!ctx) - return true; + goto out; if (gss_seq_is_newer(req->rq_seqno, READ_ONCE(ctx->gc_seq))) - goto out; + goto out_ctx; seq_xmit = READ_ONCE(ctx->gc_seq_xmit); while (gss_seq_is_newer(req->rq_seqno, seq_xmit)) { @@ -2046,15 +2042,18 @@ gss_xmit_need_reencode(struct rpc_task *task) seq_xmit = cmpxchg(&ctx->gc_seq_xmit, tmp, req->rq_seqno); if (seq_xmit == tmp) { ret = false; - goto out; + goto out_ctx; } } win = ctx->gc_win; if (win > 0) ret = !gss_seq_is_newer(req->rq_seqno, seq_xmit - win); -out: + +out_ctx: gss_put_ctx(ctx); +out: + trace_rpcgss_need_reencode(task, seq_xmit, ret); return ret; } @@ -2075,10 +2074,10 @@ gss_unwrap_resp(struct rpc_task *task, struct xdr_stream *xdr) status = gss_unwrap_resp_auth(cred); break; case RPC_GSS_SVC_INTEGRITY: - status = gss_unwrap_resp_integ(cred, ctx, rqstp, xdr); + status = gss_unwrap_resp_integ(task, cred, ctx, rqstp, xdr); break; case RPC_GSS_SVC_PRIVACY: - status = gss_unwrap_resp_priv(cred, ctx, rqstp, xdr); + status = gss_unwrap_resp_priv(task, cred, ctx, rqstp, xdr); break; } if (status) @@ -2088,8 +2087,6 @@ out_decode: status = rpcauth_unwrap_resp_decode(task, xdr); out: gss_put_ctx(ctx); - dprintk("RPC: %5u %s returning %d\n", - task->tk_pid, __func__, status); return status; } diff --git a/net/sunrpc/auth_gss/trace.c b/net/sunrpc/auth_gss/trace.c new file mode 100644 index 000000000000..5576f1e66de9 --- /dev/null +++ b/net/sunrpc/auth_gss/trace.c @@ -0,0 +1,11 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (c) 2018, 2019 Oracle. All rights reserved. + */ + +#include +#include +#include + +#define CREATE_TRACE_POINTS +#include diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index f1ec2110efeb..bc7489f1fe55 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -1165,6 +1165,7 @@ xprt_request_enqueue_transmit(struct rpc_task *task) /* Note: req is added _before_ pos */ list_add_tail(&req->rq_xmit, &pos->rq_xmit); INIT_LIST_HEAD(&req->rq_xmit2); + trace_xprt_enq_xmit(task, 1); goto out; } } else if (RPC_IS_SWAPPER(task)) { @@ -1176,6 +1177,7 @@ xprt_request_enqueue_transmit(struct rpc_task *task) /* Note: req is added _before_ pos */ list_add_tail(&req->rq_xmit, &pos->rq_xmit); INIT_LIST_HEAD(&req->rq_xmit2); + trace_xprt_enq_xmit(task, 2); goto out; } } else if (!req->rq_seqno) { @@ -1184,11 +1186,13 @@ xprt_request_enqueue_transmit(struct rpc_task *task) continue; list_add_tail(&req->rq_xmit2, &pos->rq_xmit2); INIT_LIST_HEAD(&req->rq_xmit); + trace_xprt_enq_xmit(task, 3); goto out; } } list_add_tail(&req->rq_xmit, &xprt->xmit_queue); INIT_LIST_HEAD(&req->rq_xmit2); + trace_xprt_enq_xmit(task, 4); out: set_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate); spin_unlock(&xprt->queue_lock); @@ -1313,8 +1317,6 @@ xprt_request_transmit(struct rpc_rqst *req, struct rpc_task *snd_task) int is_retrans = RPC_WAS_SENT(task); int status; - dprintk("RPC: %5u xprt_transmit(%u)\n", task->tk_pid, req->rq_slen); - if (!req->rq_bytes_sent) { if (xprt_request_data_received(task)) { status = 0; @@ -1336,9 +1338,9 @@ xprt_request_transmit(struct rpc_rqst *req, struct rpc_task *snd_task) connect_cookie = xprt->connect_cookie; status = xprt->ops->send_request(req); - trace_xprt_transmit(xprt, req->rq_xid, status); if (status != 0) { req->rq_ntrans--; + trace_xprt_transmit(req, status); return status; } @@ -1347,7 +1349,6 @@ xprt_request_transmit(struct rpc_rqst *req, struct rpc_task *snd_task) xprt_inject_disconnect(xprt); - dprintk("RPC: %5u xmit complete\n", task->tk_pid); task->tk_flags |= RPC_TASK_SENT; spin_lock_bh(&xprt->transport_lock); @@ -1360,6 +1361,7 @@ xprt_request_transmit(struct rpc_rqst *req, struct rpc_task *snd_task) req->rq_connect_cookie = connect_cookie; out_dequeue: + trace_xprt_transmit(req, status); xprt_request_dequeue_transmit(task); rpc_wake_up_queued_task_set_status(&xprt->sending, task, status); return status; From 241b1f419f0ea9966d574d7cc67377c74982a125 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:25:09 -0500 Subject: [PATCH 17/90] SUNRPC: Remove xdr_buf_trim() The key action of xdr_buf_trim() is that it shortens buf->len, the length of the xdr_buf's content. The other actions -- shortening the head, pages, and tail components -- are actually not necessary. In particular, changing the size of those components can corrupt the RPC message contained in the buffer. This is an accident waiting to happen rather than a current bug, as far as we know. Signed-off-by: Chuck Lever Acked-by: Bruce Fields Signed-off-by: Anna Schumaker --- include/linux/sunrpc/xdr.h | 1 - net/sunrpc/auth_gss/gss_krb5_wrap.c | 8 +++--- net/sunrpc/auth_gss/svcauth_gss.c | 2 +- net/sunrpc/xdr.c | 41 ----------------------------- 4 files changed, 6 insertions(+), 46 deletions(-) diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index 65af6a204b75..9ee3970ba59c 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -179,7 +179,6 @@ xdr_adjust_iovec(struct kvec *iov, __be32 *p) extern void xdr_shift_buf(struct xdr_buf *, size_t); extern void xdr_buf_from_iov(struct kvec *, struct xdr_buf *); extern int xdr_buf_subsegment(struct xdr_buf *, struct xdr_buf *, unsigned int, unsigned int); -extern void xdr_buf_trim(struct xdr_buf *, unsigned int); extern int xdr_buf_read_netobj(struct xdr_buf *, struct xdr_netobj *, unsigned int); extern int read_bytes_from_xdr_buf(struct xdr_buf *, unsigned int, void *, unsigned int); extern int write_bytes_to_xdr_buf(struct xdr_buf *, unsigned int, void *, unsigned int); diff --git a/net/sunrpc/auth_gss/gss_krb5_wrap.c b/net/sunrpc/auth_gss/gss_krb5_wrap.c index 5cdde6cb703a..14a0aff0cd84 100644 --- a/net/sunrpc/auth_gss/gss_krb5_wrap.c +++ b/net/sunrpc/auth_gss/gss_krb5_wrap.c @@ -570,14 +570,16 @@ gss_unwrap_kerberos_v2(struct krb5_ctx *kctx, int offset, struct xdr_buf *buf) */ movelen = min_t(unsigned int, buf->head[0].iov_len, buf->len); movelen -= offset + GSS_KRB5_TOK_HDR_LEN + headskip; - BUG_ON(offset + GSS_KRB5_TOK_HDR_LEN + headskip + movelen > - buf->head[0].iov_len); + if (offset + GSS_KRB5_TOK_HDR_LEN + headskip + movelen > + buf->head[0].iov_len) + return GSS_S_FAILURE; memmove(ptr, ptr + GSS_KRB5_TOK_HDR_LEN + headskip, movelen); buf->head[0].iov_len -= GSS_KRB5_TOK_HDR_LEN + headskip; buf->len -= GSS_KRB5_TOK_HDR_LEN + headskip; /* Trim off the trailing "extra count" and checksum blob */ - xdr_buf_trim(buf, ec + GSS_KRB5_TOK_HDR_LEN + tailskip); + buf->len -= ec + GSS_KRB5_TOK_HDR_LEN + tailskip; + return GSS_S_COMPLETE; } diff --git a/net/sunrpc/auth_gss/svcauth_gss.c b/net/sunrpc/auth_gss/svcauth_gss.c index 152790ed309c..f1aabab4a4c2 100644 --- a/net/sunrpc/auth_gss/svcauth_gss.c +++ b/net/sunrpc/auth_gss/svcauth_gss.c @@ -896,7 +896,7 @@ unwrap_integ_data(struct svc_rqst *rqstp, struct xdr_buf *buf, u32 seq, struct g if (svc_getnl(&buf->head[0]) != seq) goto out; /* trim off the mic and padding at the end before returning */ - xdr_buf_trim(buf, round_up_to_quad(mic.len) + 4); + buf->len -= 4 + round_up_to_quad(mic.len); stat = 0; out: kfree(mic.data); diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 5f0aa53fa4ae..4bce61978062 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -1139,47 +1139,6 @@ xdr_buf_subsegment(struct xdr_buf *buf, struct xdr_buf *subbuf, } EXPORT_SYMBOL_GPL(xdr_buf_subsegment); -/** - * xdr_buf_trim - lop at most "len" bytes off the end of "buf" - * @buf: buf to be trimmed - * @len: number of bytes to reduce "buf" by - * - * Trim an xdr_buf by the given number of bytes by fixing up the lengths. Note - * that it's possible that we'll trim less than that amount if the xdr_buf is - * too small, or if (for instance) it's all in the head and the parser has - * already read too far into it. - */ -void xdr_buf_trim(struct xdr_buf *buf, unsigned int len) -{ - size_t cur; - unsigned int trim = len; - - if (buf->tail[0].iov_len) { - cur = min_t(size_t, buf->tail[0].iov_len, trim); - buf->tail[0].iov_len -= cur; - trim -= cur; - if (!trim) - goto fix_len; - } - - if (buf->page_len) { - cur = min_t(unsigned int, buf->page_len, trim); - buf->page_len -= cur; - trim -= cur; - if (!trim) - goto fix_len; - } - - if (buf->head[0].iov_len) { - cur = min_t(size_t, buf->head[0].iov_len, trim); - buf->head[0].iov_len -= cur; - trim -= cur; - } -fix_len: - buf->len -= (len - trim); -} -EXPORT_SYMBOL_GPL(xdr_buf_trim); - static void __read_bytes_from_xdr_buf(struct xdr_buf *subbuf, void *obj, unsigned int len) { unsigned int this_len; From 2573a46499956c2ff311e1a65052364243161e78 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:25:15 -0500 Subject: [PATCH 18/90] SUNRPC: Add SPDX IDs to some net/sunrpc/auth_gss/ files Files under net/sunrpc/auth_gss/ do not yet have SPDX ID tags. This directory is somewhat complicated because most of these files have license boilerplate that is not strictly GPL 2.0. In this patch I add ID tags where there is an obvious match. The less recognizable licenses are still under research. For reference, SPDX IDs added in this patch correspond to the following license text: GPL-2.0 https://spdx.org/licenses/GPL-2.0.html GPL-2.0+ https://spdx.org/licenses/GPL-2.0+.html BSD-3-Clause https://spdx.org/licenses/BSD-3-Clause.html Cc: Simo Sorce Cc: Kate Stewart Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- net/sunrpc/auth_gss/auth_gss.c | 27 +-------------------------- net/sunrpc/auth_gss/gss_krb5_mech.c | 27 +-------------------------- net/sunrpc/auth_gss/gss_mech_switch.c | 27 +-------------------------- net/sunrpc/auth_gss/gss_rpc_upcall.c | 15 +-------------- net/sunrpc/auth_gss/gss_rpc_upcall.h | 16 ++-------------- net/sunrpc/auth_gss/gss_rpc_xdr.c | 15 +-------------- net/sunrpc/auth_gss/gss_rpc_xdr.h | 17 +---------------- net/sunrpc/auth_gss/svcauth_gss.c | 1 + 8 files changed, 9 insertions(+), 136 deletions(-) diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 3d1fbd6d3370..fda454c9b594 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -1,3 +1,4 @@ +// SPDX-License-Identifier: BSD-3-Clause /* * linux/net/sunrpc/auth_gss/auth_gss.c * @@ -8,34 +9,8 @@ * * Dug Song * Andy Adamson - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * 1. Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * 2. Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in the - * documentation and/or other materials provided with the distribution. - * 3. Neither the name of the University nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED - * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF - * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE - * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE - * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR - * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF - * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR - * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF - * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING - * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS - * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ - #include #include #include diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index be31a58d54e0..56cc85c5bc06 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -1,3 +1,4 @@ +// SPDX-License-Identifier: BSD-3-Clause /* * linux/net/sunrpc/gss_krb5_mech.c * @@ -6,32 +7,6 @@ * * Andy Adamson * J. Bruce Fields - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * 1. Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * 2. Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in the - * documentation and/or other materials provided with the distribution. - * 3. Neither the name of the University nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED - * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF - * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE - * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE - * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR - * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF - * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR - * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF - * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING - * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS - * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * */ #include diff --git a/net/sunrpc/auth_gss/gss_mech_switch.c b/net/sunrpc/auth_gss/gss_mech_switch.c index 379318dff534..82060099a429 100644 --- a/net/sunrpc/auth_gss/gss_mech_switch.c +++ b/net/sunrpc/auth_gss/gss_mech_switch.c @@ -1,3 +1,4 @@ +// SPDX-License-Identifier: BSD-3-Clause /* * linux/net/sunrpc/gss_mech_switch.c * @@ -5,32 +6,6 @@ * All rights reserved. * * J. Bruce Fields - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * 1. Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * 2. Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in the - * documentation and/or other materials provided with the distribution. - * 3. Neither the name of the University nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED - * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF - * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE - * DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE - * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR - * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF - * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR - * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF - * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING - * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS - * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * */ #include diff --git a/net/sunrpc/auth_gss/gss_rpc_upcall.c b/net/sunrpc/auth_gss/gss_rpc_upcall.c index 73dcda060335..0349f455a862 100644 --- a/net/sunrpc/auth_gss/gss_rpc_upcall.c +++ b/net/sunrpc/auth_gss/gss_rpc_upcall.c @@ -1,21 +1,8 @@ +// SPDX-License-Identifier: GPL-2.0+ /* * linux/net/sunrpc/gss_rpc_upcall.c * * Copyright (C) 2012 Simo Sorce - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ #include diff --git a/net/sunrpc/auth_gss/gss_rpc_upcall.h b/net/sunrpc/auth_gss/gss_rpc_upcall.h index 1e542aded90a..31e96344167e 100644 --- a/net/sunrpc/auth_gss/gss_rpc_upcall.h +++ b/net/sunrpc/auth_gss/gss_rpc_upcall.h @@ -1,21 +1,8 @@ +/* SPDX-License-Identifier: GPL-2.0+ */ /* * linux/net/sunrpc/gss_rpc_upcall.h * * Copyright (C) 2012 Simo Sorce - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ #ifndef _GSS_RPC_UPCALL_H @@ -45,4 +32,5 @@ void gssp_free_upcall_data(struct gssp_upcall_data *data); void init_gssp_clnt(struct sunrpc_net *); int set_gssp_clnt(struct net *); void clear_gssp_clnt(struct sunrpc_net *); + #endif /* _GSS_RPC_UPCALL_H */ diff --git a/net/sunrpc/auth_gss/gss_rpc_xdr.c b/net/sunrpc/auth_gss/gss_rpc_xdr.c index 006062ad5f58..2ff7b7083eba 100644 --- a/net/sunrpc/auth_gss/gss_rpc_xdr.c +++ b/net/sunrpc/auth_gss/gss_rpc_xdr.c @@ -1,21 +1,8 @@ +// SPDX-License-Identifier: GPL-2.0+ /* * GSS Proxy upcall module * * Copyright (C) 2012 Simo Sorce - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ #include diff --git a/net/sunrpc/auth_gss/gss_rpc_xdr.h b/net/sunrpc/auth_gss/gss_rpc_xdr.h index 146c31032917..3f17411b7e65 100644 --- a/net/sunrpc/auth_gss/gss_rpc_xdr.h +++ b/net/sunrpc/auth_gss/gss_rpc_xdr.h @@ -1,21 +1,8 @@ +/* SPDX-License-Identifier: GPL-2.0+ */ /* * GSS Proxy upcall module * * Copyright (C) 2012 Simo Sorce - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ #ifndef _LINUX_GSS_RPC_XDR_H @@ -262,6 +249,4 @@ int gssx_dec_accept_sec_context(struct rpc_rqst *rqstp, #define GSSX_ARG_wrap_size_limit_sz 0 #define GSSX_RES_wrap_size_limit_sz 0 - - #endif /* _LINUX_GSS_RPC_XDR_H */ diff --git a/net/sunrpc/auth_gss/svcauth_gss.c b/net/sunrpc/auth_gss/svcauth_gss.c index f1aabab4a4c2..0c5d7896d6dd 100644 --- a/net/sunrpc/auth_gss/svcauth_gss.c +++ b/net/sunrpc/auth_gss/svcauth_gss.c @@ -1,3 +1,4 @@ +// SPDX-License-Identifier: GPL-2.0 /* * Neil Brown * J. Bruce Fields From cf500bac8fd48b57f38ece890235923d4ed5ee91 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:25:20 -0500 Subject: [PATCH 19/90] SUNRPC: Introduce rpc_prepare_reply_pages() prepare_reply_buffer() and its NFSv4 equivalents expose the details of the RPC header and the auth slack values to upper layer consumers, creating a layering violation, and duplicating code. Remedy these issues by adding a new RPC client API that hides those details from upper layers in a common helper function. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- fs/nfs/nfs2xdr.c | 27 +++++-------------- fs/nfs/nfs3xdr.c | 29 +++++--------------- fs/nfs/nfs4xdr.c | 51 +++++++++++++++-------------------- include/linux/sunrpc/clnt.h | 3 +++ include/trace/events/sunrpc.h | 37 +++++++++++++++++++++++++ net/sunrpc/clnt.c | 19 +++++++++++++ net/sunrpc/xdr.c | 9 +++++++ 7 files changed, 102 insertions(+), 73 deletions(-) diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index bac3a4e2cb5d..1dcd0feda32d 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -65,21 +65,6 @@ static int nfs_stat_to_errno(enum nfs_stat); -/* - * While encoding arguments, set up the reply buffer in advance to - * receive reply data directly into the page cache. - */ -static void prepare_reply_buffer(struct rpc_rqst *req, struct page **pages, - unsigned int base, unsigned int len, - unsigned int bufsize) -{ - struct rpc_auth *auth = req->rq_cred->cr_auth; - unsigned int replen; - - replen = RPC_REPHDRSIZE + auth->au_rslack + bufsize; - xdr_inline_pages(&req->rq_rcv_buf, replen << 2, pages, base, len); -} - /* * Encode/decode NFSv2 basic data types * @@ -593,8 +578,8 @@ static void nfs2_xdr_enc_readlinkargs(struct rpc_rqst *req, const struct nfs_readlinkargs *args = data; encode_fhandle(xdr, args->fh); - prepare_reply_buffer(req, args->pages, args->pgbase, - args->pglen, NFS_readlinkres_sz); + rpc_prepare_reply_pages(req, args->pages, args->pgbase, + args->pglen, NFS_readlinkres_sz); } /* @@ -629,8 +614,8 @@ static void nfs2_xdr_enc_readargs(struct rpc_rqst *req, const struct nfs_pgio_args *args = data; encode_readargs(xdr, args); - prepare_reply_buffer(req, args->pages, args->pgbase, - args->count, NFS_readres_sz); + rpc_prepare_reply_pages(req, args->pages, args->pgbase, + args->count, NFS_readres_sz); req->rq_rcv_buf.flags |= XDRBUF_READ; } @@ -787,8 +772,8 @@ static void nfs2_xdr_enc_readdirargs(struct rpc_rqst *req, const struct nfs_readdirargs *args = data; encode_readdirargs(xdr, args); - prepare_reply_buffer(req, args->pages, 0, - args->count, NFS_readdirres_sz); + rpc_prepare_reply_pages(req, args->pages, 0, + args->count, NFS_readdirres_sz); } /* diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index 4aa3ffe1800e..a54dcf4bfb1d 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -104,21 +104,6 @@ static const umode_t nfs_type2fmt[] = { [NF3FIFO] = S_IFIFO, }; -/* - * While encoding arguments, set up the reply buffer in advance to - * receive reply data directly into the page cache. - */ -static void prepare_reply_buffer(struct rpc_rqst *req, struct page **pages, - unsigned int base, unsigned int len, - unsigned int bufsize) -{ - struct rpc_auth *auth = req->rq_cred->cr_auth; - unsigned int replen; - - replen = RPC_REPHDRSIZE + auth->au_rslack + bufsize; - xdr_inline_pages(&req->rq_rcv_buf, replen << 2, pages, base, len); -} - /* * Encode/decode NFSv3 basic data types * @@ -910,8 +895,8 @@ static void nfs3_xdr_enc_readlink3args(struct rpc_rqst *req, const struct nfs3_readlinkargs *args = data; encode_nfs_fh3(xdr, args->fh); - prepare_reply_buffer(req, args->pages, args->pgbase, - args->pglen, NFS3_readlinkres_sz); + rpc_prepare_reply_pages(req, args->pages, args->pgbase, + args->pglen, NFS3_readlinkres_sz); } /* @@ -943,8 +928,8 @@ static void nfs3_xdr_enc_read3args(struct rpc_rqst *req, unsigned int replen = args->replen ? args->replen : NFS3_readres_sz; encode_read3args(xdr, args); - prepare_reply_buffer(req, args->pages, args->pgbase, - args->count, replen); + rpc_prepare_reply_pages(req, args->pages, args->pgbase, + args->count, replen); req->rq_rcv_buf.flags |= XDRBUF_READ; } @@ -1236,7 +1221,7 @@ static void nfs3_xdr_enc_readdir3args(struct rpc_rqst *req, const struct nfs3_readdirargs *args = data; encode_readdir3args(xdr, args); - prepare_reply_buffer(req, args->pages, 0, + rpc_prepare_reply_pages(req, args->pages, 0, args->count, NFS3_readdirres_sz); } @@ -1278,7 +1263,7 @@ static void nfs3_xdr_enc_readdirplus3args(struct rpc_rqst *req, const struct nfs3_readdirargs *args = data; encode_readdirplus3args(xdr, args); - prepare_reply_buffer(req, args->pages, 0, + rpc_prepare_reply_pages(req, args->pages, 0, args->count, NFS3_readdirres_sz); } @@ -1323,7 +1308,7 @@ static void nfs3_xdr_enc_getacl3args(struct rpc_rqst *req, encode_nfs_fh3(xdr, args->fh); encode_uint32(xdr, args->mask); if (args->mask & (NFS_ACL | NFS_DFACL)) { - prepare_reply_buffer(req, args->pages, 0, + rpc_prepare_reply_pages(req, args->pages, 0, NFSACL_MAXPAGES << PAGE_SHIFT, ACL3_getaclres_sz); req->rq_rcv_buf.flags |= XDRBUF_SPARSE_PAGES; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 38a4cbc18657..d0fa18df32ea 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -1016,12 +1016,11 @@ static void encode_compound_hdr(struct xdr_stream *xdr, struct compound_hdr *hdr) { __be32 *p; - struct rpc_auth *auth = req->rq_cred->cr_auth; /* initialize running count of expected bytes in reply. * NOTE: the replied tag SHOULD be the same is the one sent, * but this is not required as a MUST for the server to do so. */ - hdr->replen = RPC_REPHDRSIZE + auth->au_rslack + 3 + hdr->taglen; + hdr->replen = 3 + hdr->taglen; WARN_ON_ONCE(hdr->taglen > NFS4_MAXTAGLEN); encode_string(xdr, hdr->taglen, hdr->tag); @@ -2341,9 +2340,9 @@ static void nfs4_xdr_enc_open(struct rpc_rqst *req, struct xdr_stream *xdr, encode_getfattr_open(xdr, args->bitmask, args->open_bitmap, &hdr); if (args->lg_args) { encode_layoutget(xdr, args->lg_args, &hdr); - xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2, - args->lg_args->layout.pages, - 0, args->lg_args->layout.pglen); + rpc_prepare_reply_pages(req, args->lg_args->layout.pages, 0, + args->lg_args->layout.pglen, + hdr.replen); } encode_nops(&hdr); } @@ -2387,9 +2386,9 @@ static void nfs4_xdr_enc_open_noattr(struct rpc_rqst *req, encode_getfattr_open(xdr, args->bitmask, args->open_bitmap, &hdr); if (args->lg_args) { encode_layoutget(xdr, args->lg_args, &hdr); - xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2, - args->lg_args->layout.pages, - 0, args->lg_args->layout.pglen); + rpc_prepare_reply_pages(req, args->lg_args->layout.pages, 0, + args->lg_args->layout.pglen, + hdr.replen); } encode_nops(&hdr); } @@ -2499,8 +2498,8 @@ static void nfs4_xdr_enc_readlink(struct rpc_rqst *req, struct xdr_stream *xdr, encode_putfh(xdr, args->fh, &hdr); encode_readlink(xdr, args, req, &hdr); - xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2, args->pages, - args->pgbase, args->pglen); + rpc_prepare_reply_pages(req, args->pages, args->pgbase, + args->pglen, hdr.replen); encode_nops(&hdr); } @@ -2520,11 +2519,8 @@ static void nfs4_xdr_enc_readdir(struct rpc_rqst *req, struct xdr_stream *xdr, encode_putfh(xdr, args->fh, &hdr); encode_readdir(xdr, args, req, &hdr); - xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2, args->pages, - args->pgbase, args->count); - dprintk("%s: inlined page args = (%u, %p, %u, %u)\n", - __func__, hdr.replen << 2, args->pages, - args->pgbase, args->count); + rpc_prepare_reply_pages(req, args->pages, args->pgbase, + args->count, hdr.replen); encode_nops(&hdr); } @@ -2544,8 +2540,8 @@ static void nfs4_xdr_enc_read(struct rpc_rqst *req, struct xdr_stream *xdr, encode_putfh(xdr, args->fh, &hdr); encode_read(xdr, args, &hdr); - xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2, - args->pages, args->pgbase, args->count); + rpc_prepare_reply_pages(req, args->pages, args->pgbase, + args->count, hdr.replen); req->rq_rcv_buf.flags |= XDRBUF_READ; encode_nops(&hdr); } @@ -2591,9 +2587,8 @@ static void nfs4_xdr_enc_getacl(struct rpc_rqst *req, struct xdr_stream *xdr, encode_getattr(xdr, nfs4_acl_bitmap, NULL, ARRAY_SIZE(nfs4_acl_bitmap), &hdr); - xdr_inline_pages(&req->rq_rcv_buf, replen << 2, - args->acl_pages, 0, args->acl_len); - + rpc_prepare_reply_pages(req, args->acl_pages, 0, + args->acl_len, replen); encode_nops(&hdr); } @@ -2814,9 +2809,8 @@ static void nfs4_xdr_enc_fs_locations(struct rpc_rqst *req, encode_fs_locations(xdr, args->bitmask, &hdr); } - /* Set up reply kvec to capture returned fs_locations array. */ - xdr_inline_pages(&req->rq_rcv_buf, replen << 2, - (struct page **)&args->page, 0, PAGE_SIZE); + rpc_prepare_reply_pages(req, (struct page **)&args->page, 0, + PAGE_SIZE, replen); encode_nops(&hdr); } @@ -3018,10 +3012,8 @@ static void nfs4_xdr_enc_getdeviceinfo(struct rpc_rqst *req, /* set up reply kvec. Subtract notification bitmap max size (2) * so that notification bitmap is put in xdr_buf tail */ - xdr_inline_pages(&req->rq_rcv_buf, (hdr.replen - 2) << 2, - args->pdev->pages, args->pdev->pgbase, - args->pdev->pglen); - + rpc_prepare_reply_pages(req, args->pdev->pages, args->pdev->pgbase, + args->pdev->pglen, hdr.replen - 2); encode_nops(&hdr); } @@ -3042,9 +3034,8 @@ static void nfs4_xdr_enc_layoutget(struct rpc_rqst *req, encode_putfh(xdr, NFS_FH(args->inode), &hdr); encode_layoutget(xdr, args, &hdr); - xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2, - args->layout.pages, 0, args->layout.pglen); - + rpc_prepare_reply_pages(req, args->layout.pages, 0, + args->layout.pglen, hdr.replen); encode_nops(&hdr); } diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 1c441714d569..98bc9883b230 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -169,6 +169,9 @@ int rpcb_v4_register(struct net *net, const u32 program, const char *netid); void rpcb_getport_async(struct rpc_task *); +void rpc_prepare_reply_pages(struct rpc_rqst *req, struct page **pages, + unsigned int base, unsigned int len, + unsigned int hdrsize); void rpc_call_start(struct rpc_task *); int rpc_call_async(struct rpc_clnt *clnt, const struct rpc_message *msg, int flags, diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index e58dda8e038c..8451f30c6a0f 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -461,6 +461,43 @@ TRACE_EVENT(rpc_xdr_alignment, ) ); +TRACE_EVENT(rpc_reply_pages, + TP_PROTO( + const struct rpc_rqst *req + ), + + TP_ARGS(req), + + TP_STRUCT__entry( + __field(unsigned int, task_id) + __field(unsigned int, client_id) + __field(const void *, head_base) + __field(size_t, head_len) + __field(const void *, tail_base) + __field(size_t, tail_len) + __field(unsigned int, page_len) + ), + + TP_fast_assign( + __entry->task_id = req->rq_task->tk_pid; + __entry->client_id = req->rq_task->tk_client->cl_clid; + + __entry->head_base = req->rq_rcv_buf.head[0].iov_base; + __entry->head_len = req->rq_rcv_buf.head[0].iov_len; + __entry->page_len = req->rq_rcv_buf.page_len; + __entry->tail_base = req->rq_rcv_buf.tail[0].iov_base; + __entry->tail_len = req->rq_rcv_buf.tail[0].iov_len; + ), + + TP_printk( + "task:%u@%u xdr=[%p,%zu]/%u/[%p,%zu]\n", + __entry->task_id, __entry->client_id, + __entry->head_base, __entry->head_len, + __entry->page_len, + __entry->tail_base, __entry->tail_len + ) +); + /* * First define the enums in the below macros to be exported to userspace * via TRACE_DEFINE_ENUM(). diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 803e93105af1..f780605fffe0 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1164,6 +1164,25 @@ struct rpc_task *rpc_run_bc_task(struct rpc_rqst *req) } #endif /* CONFIG_SUNRPC_BACKCHANNEL */ +/** + * rpc_prepare_reply_pages - Prepare to receive a reply data payload into pages + * @req: RPC request to prepare + * @pages: vector of struct page pointers + * @base: offset in first page where receive should start, in bytes + * @len: expected size of the upper layer data payload, in bytes + * @hdrsize: expected size of upper layer reply header, in XDR words + * + */ +void rpc_prepare_reply_pages(struct rpc_rqst *req, struct page **pages, + unsigned int base, unsigned int len, + unsigned int hdrsize) +{ + hdrsize += RPC_REPHDRSIZE + req->rq_cred->cr_auth->au_rslack; + xdr_inline_pages(&req->rq_rcv_buf, hdrsize << 2, pages, base, len); + trace_rpc_reply_pages(req); +} +EXPORT_SYMBOL_GPL(rpc_prepare_reply_pages); + void rpc_call_start(struct rpc_task *task) { diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 4bce61978062..7cca51560442 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -163,6 +163,15 @@ xdr_free_bvec(struct xdr_buf *buf) buf->bvec = NULL; } +/** + * xdr_inline_pages - Prepare receive buffer for a large reply + * @xdr: xdr_buf into which reply will be placed + * @offset: expected offset where data payload will start, in bytes + * @pages: vector of struct page pointers + * @base: offset in first page where receive should start, in bytes + * @len: expected size of the upper layer data payload, in bytes + * + */ void xdr_inline_pages(struct xdr_buf *xdr, unsigned int offset, struct page **pages, unsigned int base, unsigned int len) From 02ef04e432babf8fc703104212314e54112ecd2d Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:25:25 -0500 Subject: [PATCH 20/90] NFS: Account for XDR pad of buf->pages Certain NFS results (eg. READLINK) might expect a data payload that is not an exact multiple of 4 bytes. In this case, XDR encoding is required to pad that payload so its length on the wire is a multiple of 4 bytes. The constants that define the maximum size of each NFS result do not appear to account for this extra word. In each case where the data payload is to be received into pages: - 1 word is added to the size of the receive buffer allocated by call_allocate - rpc_inline_rcv_pages subtracts 1 word from @hdrsize so that the extra buffer space falls into the rcv_buf's tail iovec - If buf->pagelen is word-aligned, an XDR pad is not needed and is thus removed from the tail Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- fs/nfs/nfs2xdr.c | 6 +++--- fs/nfs/nfs3xdr.c | 10 +++++----- fs/nfs/nfs4xdr.c | 15 ++++++++------- net/sunrpc/clnt.c | 6 +++++- net/sunrpc/xdr.c | 2 ++ 5 files changed, 23 insertions(+), 16 deletions(-) diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index 1dcd0feda32d..a7ed29de0a40 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -56,11 +56,11 @@ #define NFS_attrstat_sz (1+NFS_fattr_sz) #define NFS_diropres_sz (1+NFS_fhandle_sz+NFS_fattr_sz) -#define NFS_readlinkres_sz (2) -#define NFS_readres_sz (1+NFS_fattr_sz+1) +#define NFS_readlinkres_sz (2+1) +#define NFS_readres_sz (1+NFS_fattr_sz+1+1) #define NFS_writeres_sz (NFS_attrstat_sz) #define NFS_stat_sz (1) -#define NFS_readdirres_sz (1) +#define NFS_readdirres_sz (1+1) #define NFS_statfsres_sz (1+NFS_info_sz) static int nfs_stat_to_errno(enum nfs_stat); diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index a54dcf4bfb1d..110358f4986d 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -69,13 +69,13 @@ #define NFS3_removeres_sz (NFS3_setattrres_sz) #define NFS3_lookupres_sz (1+NFS3_fh_sz+(2 * NFS3_post_op_attr_sz)) #define NFS3_accessres_sz (1+NFS3_post_op_attr_sz+1) -#define NFS3_readlinkres_sz (1+NFS3_post_op_attr_sz+1) -#define NFS3_readres_sz (1+NFS3_post_op_attr_sz+3) +#define NFS3_readlinkres_sz (1+NFS3_post_op_attr_sz+1+1) +#define NFS3_readres_sz (1+NFS3_post_op_attr_sz+3+1) #define NFS3_writeres_sz (1+NFS3_wcc_data_sz+4) #define NFS3_createres_sz (1+NFS3_fh_sz+NFS3_post_op_attr_sz+NFS3_wcc_data_sz) #define NFS3_renameres_sz (1+(2 * NFS3_wcc_data_sz)) #define NFS3_linkres_sz (1+NFS3_post_op_attr_sz+NFS3_wcc_data_sz) -#define NFS3_readdirres_sz (1+NFS3_post_op_attr_sz+2) +#define NFS3_readdirres_sz (1+NFS3_post_op_attr_sz+2+1) #define NFS3_fsstatres_sz (1+NFS3_post_op_attr_sz+13) #define NFS3_fsinfores_sz (1+NFS3_post_op_attr_sz+12) #define NFS3_pathconfres_sz (1+NFS3_post_op_attr_sz+6) @@ -85,7 +85,7 @@ #define ACL3_setaclargs_sz (NFS3_fh_sz+1+ \ XDR_QUADLEN(NFS_ACL_INLINE_BUFSIZE)) #define ACL3_getaclres_sz (1+NFS3_post_op_attr_sz+1+ \ - XDR_QUADLEN(NFS_ACL_INLINE_BUFSIZE)) + XDR_QUADLEN(NFS_ACL_INLINE_BUFSIZE)+1) #define ACL3_setaclres_sz (1+NFS3_post_op_attr_sz) static int nfs3_stat_to_errno(enum nfs_stat); @@ -1629,7 +1629,7 @@ static int nfs3_xdr_dec_read3res(struct rpc_rqst *req, struct xdr_stream *xdr, result->op_status = status; if (status != NFS3_OK) goto out_status; - result->replen = 3 + ((xdr_stream_pos(xdr) - pos) >> 2); + result->replen = 4 + ((xdr_stream_pos(xdr) - pos) >> 2); error = decode_read3resok(xdr, result); out: return error; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index d0fa18df32ea..6d9d5e2f6308 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -215,14 +215,14 @@ static int decode_layoutget(struct xdr_stream *xdr, struct rpc_rqst *req, nfs4_fattr_bitmap_maxsz) #define encode_read_maxsz (op_encode_hdr_maxsz + \ encode_stateid_maxsz + 3) -#define decode_read_maxsz (op_decode_hdr_maxsz + 2) +#define decode_read_maxsz (op_decode_hdr_maxsz + 2 + 1) #define encode_readdir_maxsz (op_encode_hdr_maxsz + \ 2 + encode_verifier_maxsz + 5 + \ nfs4_label_maxsz) #define decode_readdir_maxsz (op_decode_hdr_maxsz + \ - decode_verifier_maxsz) + decode_verifier_maxsz + 1) #define encode_readlink_maxsz (op_encode_hdr_maxsz) -#define decode_readlink_maxsz (op_decode_hdr_maxsz + 1) +#define decode_readlink_maxsz (op_decode_hdr_maxsz + 1 + 1) #define encode_write_maxsz (op_encode_hdr_maxsz + \ encode_stateid_maxsz + 4) #define decode_write_maxsz (op_decode_hdr_maxsz + \ @@ -284,14 +284,14 @@ static int decode_layoutget(struct xdr_stream *xdr, struct rpc_rqst *req, #define decode_delegreturn_maxsz (op_decode_hdr_maxsz) #define encode_getacl_maxsz (encode_getattr_maxsz) #define decode_getacl_maxsz (op_decode_hdr_maxsz + \ - nfs4_fattr_bitmap_maxsz + 1) + nfs4_fattr_bitmap_maxsz + 1 + 1) #define encode_setacl_maxsz (op_encode_hdr_maxsz + \ encode_stateid_maxsz + 3) #define decode_setacl_maxsz (decode_setattr_maxsz) #define encode_fs_locations_maxsz \ (encode_getattr_maxsz) #define decode_fs_locations_maxsz \ - (0) + (1) #define encode_secinfo_maxsz (op_encode_hdr_maxsz + nfs4_name_maxsz) #define decode_secinfo_maxsz (op_decode_hdr_maxsz + 1 + ((NFS_MAX_SECFLAVORS * (16 + GSS_OID_MAX_LEN)) / 4)) @@ -392,12 +392,13 @@ static int decode_layoutget(struct xdr_stream *xdr, struct rpc_rqst *req, 1 /* opaque devaddr4 length */ + \ /* devaddr4 payload is read into page */ \ 1 /* notification bitmap length */ + \ - 1 /* notification bitmap, word 0 */) + 1 /* notification bitmap, word 0 */ + \ + 1 /* possible XDR padding */) #define encode_layoutget_maxsz (op_encode_hdr_maxsz + 10 + \ encode_stateid_maxsz) #define decode_layoutget_maxsz (op_decode_hdr_maxsz + 8 + \ decode_stateid_maxsz + \ - XDR_QUADLEN(PNFS_LAYOUT_MAXSIZE)) + XDR_QUADLEN(PNFS_LAYOUT_MAXSIZE) + 1) #define encode_layoutcommit_maxsz (op_encode_hdr_maxsz + \ 2 /* offset */ + \ 2 /* length */ + \ diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index f780605fffe0..4ea38b029e2f 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1177,7 +1177,11 @@ void rpc_prepare_reply_pages(struct rpc_rqst *req, struct page **pages, unsigned int base, unsigned int len, unsigned int hdrsize) { - hdrsize += RPC_REPHDRSIZE + req->rq_cred->cr_auth->au_rslack; + /* Subtract one to force an extra word of buffer space for the + * payload's XDR pad to fall into the rcv_buf's tail iovec. + */ + hdrsize += RPC_REPHDRSIZE + req->rq_cred->cr_auth->au_rslack - 1; + xdr_inline_pages(&req->rq_rcv_buf, hdrsize << 2, pages, base, len); trace_rpc_reply_pages(req); } diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 7cca51560442..aa8177ddcbda 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -189,6 +189,8 @@ xdr_inline_pages(struct xdr_buf *xdr, unsigned int offset, tail->iov_base = buf + offset; tail->iov_len = buflen - offset; + if ((xdr->page_len & 3) == 0) + tail->iov_len -= sizeof(__be32); xdr->buflen += len; } From a00275baa68e1ee226cc659f54dc3a571f3ad600 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:25:31 -0500 Subject: [PATCH 21/90] SUNRPC: Make AUTH_SYS and AUTH_NULL set au_verfsize au_verfsize will be needed for a non-flavor-specific computation in a subsequent patch. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/auth.h | 3 +-- net/sunrpc/auth_gss/auth_gss.c | 1 + net/sunrpc/auth_null.c | 1 + net/sunrpc/auth_unix.c | 5 ++++- 4 files changed, 7 insertions(+), 3 deletions(-) diff --git a/include/linux/sunrpc/auth.h b/include/linux/sunrpc/auth.h index c51e1893f77e..359dfdd04e77 100644 --- a/include/linux/sunrpc/auth.h +++ b/include/linux/sunrpc/auth.h @@ -77,8 +77,7 @@ struct rpc_auth { /* guess at number of u32's auth adds before * reply data; normally the verifier size: */ unsigned int au_rslack; - /* for gss, used to calculate au_rslack: */ - unsigned int au_verfsize; + unsigned int au_verfsize; /* size of reply verifier */ unsigned int au_flags; /* various flags */ const struct rpc_authops *au_ops; /* operations */ diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index fda454c9b594..731e7a482e18 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -1016,6 +1016,7 @@ gss_create_new(const struct rpc_auth_create_args *args, struct rpc_clnt *clnt) auth = &gss_auth->rpc_auth; auth->au_cslack = GSS_CRED_SLACK >> 2; auth->au_rslack = GSS_VERF_SLACK >> 2; + auth->au_verfsize = GSS_VERF_SLACK >> 2; auth->au_flags = 0; auth->au_ops = &authgss_ops; auth->au_flavor = flavor; diff --git a/net/sunrpc/auth_null.c b/net/sunrpc/auth_null.c index bf96975ffc4b..9ae08248a9e1 100644 --- a/net/sunrpc/auth_null.c +++ b/net/sunrpc/auth_null.c @@ -114,6 +114,7 @@ static struct rpc_auth null_auth = { .au_cslack = NUL_CALLSLACK, .au_rslack = NUL_REPLYSLACK, + .au_verfsize = NUL_REPLYSLACK, .au_ops = &authnull_ops, .au_flavor = RPC_AUTH_NULL, .au_count = REFCOUNT_INIT(1), diff --git a/net/sunrpc/auth_unix.c b/net/sunrpc/auth_unix.c index 5ea84a96f96e..a93c56442487 100644 --- a/net/sunrpc/auth_unix.c +++ b/net/sunrpc/auth_unix.c @@ -163,6 +163,7 @@ unx_refresh(struct rpc_task *task) static int unx_validate(struct rpc_task *task, struct xdr_stream *xdr) { + struct rpc_auth *auth = task->tk_rqstp->rq_cred->cr_auth; __be32 *p; u32 size; @@ -184,7 +185,8 @@ unx_validate(struct rpc_task *task, struct xdr_stream *xdr) if (!p) return -EIO; - task->tk_rqstp->rq_cred->cr_auth->au_rslack = (size >> 2) + 2; + auth->au_verfsize = XDR_QUADLEN(size) + 2; + auth->au_rslack = XDR_QUADLEN(size) + 2; return 0; } @@ -212,6 +214,7 @@ static struct rpc_auth unix_auth = { .au_cslack = UNX_CALLSLACK, .au_rslack = NUL_REPLYSLACK, + .au_verfsize = NUL_REPLYSLACK, .au_ops = &authunix_ops, .au_flavor = RPC_AUTH_UNIX, .au_count = REFCOUNT_INIT(1), From 35e77d21baa04b554bf3dc9a08dfa7e569286e51 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:25:36 -0500 Subject: [PATCH 22/90] SUNRPC: Add rpc_auth::au_ralign field Currently rpc_inline_rcv_pages() uses au_rslack to estimate the size of the upper layer reply header. This is fine for auth flavors where au_verfsize == au_rslack. However, some auth flavors have more going on. krb5i for example has two more words after the verifier, and another blob following the RPC message. The calculation involving au_rslack pushes the upper layer reply header too far into the rcv_buf. au_rslack is still valuable: it's the amount of buffer space needed for the reply, and is used when allocating the reply buffer. We'll keep that. But, add a new field that can be used to properly estimate the location of the upper layer header in each RPC reply, based on the auth flavor in use. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/linux/sunrpc/auth.h | 9 ++++----- net/sunrpc/auth_gss/auth_gss.c | 18 +++++++++++++----- net/sunrpc/auth_null.c | 1 + net/sunrpc/auth_unix.c | 1 + net/sunrpc/clnt.c | 2 +- 5 files changed, 20 insertions(+), 11 deletions(-) diff --git a/include/linux/sunrpc/auth.h b/include/linux/sunrpc/auth.h index 359dfdd04e77..5f9076fdb090 100644 --- a/include/linux/sunrpc/auth.h +++ b/include/linux/sunrpc/auth.h @@ -74,13 +74,12 @@ struct rpc_cred_cache; struct rpc_authops; struct rpc_auth { unsigned int au_cslack; /* call cred size estimate */ - /* guess at number of u32's auth adds before - * reply data; normally the verifier size: */ - unsigned int au_rslack; + unsigned int au_rslack; /* reply cred size estimate */ unsigned int au_verfsize; /* size of reply verifier */ + unsigned int au_ralign; /* words before UL header */ - unsigned int au_flags; /* various flags */ - const struct rpc_authops *au_ops; /* operations */ + unsigned int au_flags; + const struct rpc_authops *au_ops; rpc_authflavor_t au_flavor; /* pseudoflavor (note may * differ from the flavor in * au_ops->au_flavor in gss diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 731e7a482e18..c67e2ad151ae 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -1017,6 +1017,7 @@ gss_create_new(const struct rpc_auth_create_args *args, struct rpc_clnt *clnt) auth->au_cslack = GSS_CRED_SLACK >> 2; auth->au_rslack = GSS_VERF_SLACK >> 2; auth->au_verfsize = GSS_VERF_SLACK >> 2; + auth->au_ralign = GSS_VERF_SLACK >> 2; auth->au_flags = 0; auth->au_ops = &authgss_ops; auth->au_flavor = flavor; @@ -1891,7 +1892,10 @@ out: static int gss_unwrap_resp_auth(struct rpc_cred *cred) { - cred->cr_auth->au_rslack = cred->cr_auth->au_verfsize; + struct rpc_auth *auth = cred->cr_auth; + + auth->au_rslack = auth->au_verfsize; + auth->au_ralign = auth->au_verfsize; return 0; } @@ -1902,6 +1906,7 @@ gss_unwrap_resp_integ(struct rpc_task *task, struct rpc_cred *cred, { struct xdr_buf integ_buf, *rcv_buf = &rqstp->rq_rcv_buf; u32 data_offset, mic_offset, integ_len, maj_stat; + struct rpc_auth *auth = cred->cr_auth; struct xdr_netobj mic; __be32 *p; @@ -1928,8 +1933,8 @@ gss_unwrap_resp_integ(struct rpc_task *task, struct rpc_cred *cred, if (maj_stat != GSS_S_COMPLETE) goto bad_mic; - cred->cr_auth->au_rslack = cred->cr_auth->au_verfsize + 2 + - 1 + XDR_QUADLEN(mic.len); + auth->au_rslack = auth->au_verfsize + 2 + 1 + XDR_QUADLEN(mic.len); + auth->au_ralign = auth->au_verfsize + 2; return 0; unwrap_failed: trace_rpcgss_unwrap_failed(task); @@ -1949,6 +1954,7 @@ gss_unwrap_resp_priv(struct rpc_task *task, struct rpc_cred *cred, { struct xdr_buf *rcv_buf = &rqstp->rq_rcv_buf; struct kvec *head = rqstp->rq_rcv_buf.head; + struct rpc_auth *auth = cred->cr_auth; unsigned int savedlen = rcv_buf->len; u32 offset, opaque_len, maj_stat; __be32 *p; @@ -1976,8 +1982,10 @@ gss_unwrap_resp_priv(struct rpc_task *task, struct rpc_cred *cred, */ xdr_init_decode(xdr, rcv_buf, p, rqstp); - cred->cr_auth->au_rslack = cred->cr_auth->au_verfsize + 2 + - XDR_QUADLEN(savedlen - rcv_buf->len); + auth->au_rslack = auth->au_verfsize + 2 + + XDR_QUADLEN(savedlen - rcv_buf->len); + auth->au_ralign = auth->au_verfsize + 2 + + XDR_QUADLEN(savedlen - rcv_buf->len); return 0; unwrap_failed: trace_rpcgss_unwrap_failed(task); diff --git a/net/sunrpc/auth_null.c b/net/sunrpc/auth_null.c index 9ae08248a9e1..41a633a4049e 100644 --- a/net/sunrpc/auth_null.c +++ b/net/sunrpc/auth_null.c @@ -115,6 +115,7 @@ struct rpc_auth null_auth = { .au_cslack = NUL_CALLSLACK, .au_rslack = NUL_REPLYSLACK, .au_verfsize = NUL_REPLYSLACK, + .au_ralign = NUL_REPLYSLACK, .au_ops = &authnull_ops, .au_flavor = RPC_AUTH_NULL, .au_count = REFCOUNT_INIT(1), diff --git a/net/sunrpc/auth_unix.c b/net/sunrpc/auth_unix.c index a93c56442487..c048eb6deaaf 100644 --- a/net/sunrpc/auth_unix.c +++ b/net/sunrpc/auth_unix.c @@ -187,6 +187,7 @@ unx_validate(struct rpc_task *task, struct xdr_stream *xdr) auth->au_verfsize = XDR_QUADLEN(size) + 2; auth->au_rslack = XDR_QUADLEN(size) + 2; + auth->au_ralign = XDR_QUADLEN(size) + 2; return 0; } diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 4ea38b029e2f..99bfeb17367c 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1180,7 +1180,7 @@ void rpc_prepare_reply_pages(struct rpc_rqst *req, struct page **pages, /* Subtract one to force an extra word of buffer space for the * payload's XDR pad to fall into the rcv_buf's tail iovec. */ - hdrsize += RPC_REPHDRSIZE + req->rq_cred->cr_auth->au_rslack - 1; + hdrsize += RPC_REPHDRSIZE + req->rq_cred->cr_auth->au_ralign - 1; xdr_inline_pages(&req->rq_rcv_buf, hdrsize << 2, pages, base, len); trace_rpc_reply_pages(req); From 2c94b8eca1a26cd46010d6e73a23da5f2e93a19d Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 11 Feb 2019 11:25:41 -0500 Subject: [PATCH 23/90] SUNRPC: Use au_rslack when computing reply buffer size au_rslack is significantly smaller than (au_cslack << 2). Using that value results in smaller receive buffers. In some cases this eliminates an extra segment in Reply chunks (RPC/RDMA). Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- net/sunrpc/clnt.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 99bfeb17367c..241e8423fd0c 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1690,7 +1690,7 @@ call_refreshresult(struct rpc_task *task) static void call_allocate(struct rpc_task *task) { - unsigned int slack = task->tk_rqstp->rq_cred->cr_auth->au_cslack; + const struct rpc_auth *auth = task->tk_rqstp->rq_cred->cr_auth; struct rpc_rqst *req = task->tk_rqstp; struct rpc_xprt *xprt = req->rq_xprt; const struct rpc_procinfo *proc = task->tk_msg.rpc_proc; @@ -1715,9 +1715,10 @@ call_allocate(struct rpc_task *task) * and reply headers, and convert both values * to byte sizes. */ - req->rq_callsize = RPC_CALLHDRSIZE + (slack << 1) + proc->p_arglen; + req->rq_callsize = RPC_CALLHDRSIZE + (auth->au_cslack << 1) + + proc->p_arglen; req->rq_callsize <<= 2; - req->rq_rcvsize = RPC_REPHDRSIZE + slack + proc->p_replen; + req->rq_rcvsize = RPC_REPHDRSIZE + auth->au_rslack + proc->p_replen; req->rq_rcvsize <<= 2; status = xprt->ops->buf_alloc(task); From f57dcf4c72113c745d83f1c65f7291299f65c14f Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 13 Feb 2019 09:21:38 -0500 Subject: [PATCH 24/90] NFS: Fix I/O request leakages When we fail to add the request to the I/O queue, we currently leave it to the caller to free the failed request. However since some of the requests that fail are actually created by nfs_pageio_add_request() itself, and are not passed back the caller, this leads to a leakage issue, which can again cause page locks to leak. This commit addresses the leakage by freeing the created requests on error, using desc->pg_completion_ops->error_cleanup() Signed-off-by: Trond Myklebust Fixes: a7d42ddb30997 ("nfs: add mirroring support to pgio layer") Cc: stable@vger.kernel.org # v4.0: c18b96a1b862: nfs: clean up rest of reqs Cc: stable@vger.kernel.org # v4.0: d600ad1f2bdb: NFS41: pop some layoutget Cc: stable@vger.kernel.org # v4.0+ --- fs/nfs/pagelist.c | 26 +++++++++++++++++++++----- 1 file changed, 21 insertions(+), 5 deletions(-) diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index e54d899c1848..301880a3ad8e 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -988,6 +988,17 @@ static void nfs_pageio_doio(struct nfs_pageio_descriptor *desc) } } +static void +nfs_pageio_cleanup_request(struct nfs_pageio_descriptor *desc, + struct nfs_page *req) +{ + LIST_HEAD(head); + + nfs_list_remove_request(req); + nfs_list_add_request(req, &head); + desc->pg_completion_ops->error_cleanup(&head); +} + /** * nfs_pageio_add_request - Attempt to coalesce a request into a page list. * @desc: destination io descriptor @@ -1025,10 +1036,8 @@ static int __nfs_pageio_add_request(struct nfs_pageio_descriptor *desc, nfs_page_group_unlock(req); desc->pg_moreio = 1; nfs_pageio_doio(desc); - if (desc->pg_error < 0) - return 0; - if (mirror->pg_recoalesce) - return 0; + if (desc->pg_error < 0 || mirror->pg_recoalesce) + goto out_cleanup_subreq; /* retry add_request for this subreq */ nfs_page_group_lock(req); continue; @@ -1061,6 +1070,10 @@ err_ptr: desc->pg_error = PTR_ERR(subreq); nfs_page_group_unlock(req); return 0; +out_cleanup_subreq: + if (req != subreq) + nfs_pageio_cleanup_request(desc, subreq); + return 0; } static int nfs_do_recoalesce(struct nfs_pageio_descriptor *desc) @@ -1168,11 +1181,14 @@ int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc, if (nfs_pgio_has_mirroring(desc)) desc->pg_mirror_idx = midx; if (!nfs_pageio_add_request_mirror(desc, dupreq)) - goto out_failed; + goto out_cleanup_subreq; } return 1; +out_cleanup_subreq: + if (req != dupreq) + nfs_pageio_cleanup_request(desc, dupreq); out_failed: nfs_pageio_error_cleanup(desc); return 0; From 4d91969ed4dbcefd0e78f77494f0cb8fada9048a Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 15 Feb 2019 14:59:52 -0500 Subject: [PATCH 25/90] NFS: Fix an I/O request leakage in nfs_do_recoalesce Whether we need to exit early, or just reprocess the list, we must not lost track of the request which failed to get recoalesced. Fixes: 03d5eb65b538 ("NFS: Fix a memory leak in nfs_do_recoalesce") Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org # v4.0+ --- fs/nfs/pagelist.c | 1 - 1 file changed, 1 deletion(-) diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index 301880a3ad8e..03bde9a41451 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -1092,7 +1092,6 @@ static int nfs_do_recoalesce(struct nfs_pageio_descriptor *desc) struct nfs_page *req; req = list_first_entry(&head, struct nfs_page, wb_list); - nfs_list_remove_request(req); if (__nfs_pageio_add_request(desc, req)) continue; if (desc->pg_error < 0) { From 8127d82705998568b52ac724e28e00941538083d Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 15 Feb 2019 16:08:25 -0500 Subject: [PATCH 26/90] NFS: Don't recoalesce on error in nfs_pageio_complete_mirror() If the I/O completion failed with a fatal error, then we should just exit nfs_pageio_complete_mirror() rather than try to recoalesce. Fixes: a7d42ddb3099 ("nfs: add mirroring support to pgio layer") Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org # v4.0+ --- fs/nfs/pagelist.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index 03bde9a41451..a8951f1f7b4e 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -1209,7 +1209,7 @@ static void nfs_pageio_complete_mirror(struct nfs_pageio_descriptor *desc, desc->pg_mirror_idx = mirror_idx; for (;;) { nfs_pageio_doio(desc); - if (!mirror->pg_recoalesce) + if (desc->pg_error < 0 || !mirror->pg_recoalesce) break; if (!nfs_do_recoalesce(desc)) break; From 078b5fd92c4913dd367361db6c28568386077c89 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 18 Feb 2019 11:35:54 -0500 Subject: [PATCH 27/90] NFS: Clean up list moves of struct nfs_page In several places we're just moving the struct nfs_page from one list to another by first removing from the existing list, then adding to the new one. Signed-off-by: Trond Myklebust --- fs/nfs/direct.c | 3 +-- fs/nfs/pagelist.c | 12 ++++-------- include/linux/nfs_page.h | 10 ++++++++++ 3 files changed, 15 insertions(+), 10 deletions(-) diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 33824a0a57bf..1377ee20ecf9 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -664,8 +664,7 @@ static void nfs_direct_write_reschedule(struct nfs_direct_req *dreq) list_for_each_entry_safe(req, tmp, &reqs, wb_list) { if (!nfs_pageio_add_request(&desc, req)) { - nfs_list_remove_request(req); - nfs_list_add_request(req, &failed); + nfs_list_move_request(req, &failed); spin_lock(&cinfo.inode->i_lock); dreq->flags = 0; if (desc.pg_error < 0) diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index a8951f1f7b4e..9cbfdb979992 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -768,8 +768,7 @@ int nfs_generic_pgio(struct nfs_pageio_descriptor *desc, pageused = 0; while (!list_empty(head)) { req = nfs_list_entry(head->next); - nfs_list_remove_request(req); - nfs_list_add_request(req, &hdr->pages); + nfs_list_move_request(req, &hdr->pages); if (!last_page || last_page != req->wb_page) { pageused++; @@ -961,8 +960,7 @@ static int nfs_pageio_do_add_request(struct nfs_pageio_descriptor *desc, } if (!nfs_can_coalesce_requests(prev, req, desc)) return 0; - nfs_list_remove_request(req); - nfs_list_add_request(req, &mirror->pg_list); + nfs_list_move_request(req, &mirror->pg_list); mirror->pg_count += req->wb_bytes; return 1; } @@ -994,8 +992,7 @@ nfs_pageio_cleanup_request(struct nfs_pageio_descriptor *desc, { LIST_HEAD(head); - nfs_list_remove_request(req); - nfs_list_add_request(req, &head); + nfs_list_move_request(req, &head); desc->pg_completion_ops->error_cleanup(&head); } @@ -1237,9 +1234,8 @@ int nfs_pageio_resend(struct nfs_pageio_descriptor *desc, while (!list_empty(&hdr->pages)) { struct nfs_page *req = nfs_list_entry(hdr->pages.next); - nfs_list_remove_request(req); if (!nfs_pageio_add_request(desc, req)) - nfs_list_add_request(req, &failed); + nfs_list_move_request(req, &failed); } nfs_pageio_complete(desc); if (!list_empty(&failed)) { diff --git a/include/linux/nfs_page.h b/include/linux/nfs_page.h index e27572d30d97..ad69430fd0eb 100644 --- a/include/linux/nfs_page.h +++ b/include/linux/nfs_page.h @@ -164,6 +164,16 @@ nfs_list_add_request(struct nfs_page *req, struct list_head *head) list_add_tail(&req->wb_list, head); } +/** + * nfs_list_move_request - Move a request to a new list + * @req: request + * @head: head of list into which to insert the request. + */ +static inline void +nfs_list_move_request(struct nfs_page *req, struct list_head *head) +{ + list_move_tail(&req->wb_list, head); +} /** * nfs_list_remove_request - Remove a request from its wb_list From df3accb849607a86278a37c35e6b313635ccc48b Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 13 Feb 2019 10:39:39 -0500 Subject: [PATCH 28/90] NFS: Pass error information to the pgio error cleanup routine Allow the caller to pass error information when cleaning up a failed I/O request so that we can conditionally take action to cancel the request altogether if the error turned out to be fatal. Signed-off-by: Trond Myklebust --- fs/nfs/direct.c | 4 ++-- fs/nfs/pagelist.c | 5 +++-- fs/nfs/read.c | 2 +- fs/nfs/write.c | 11 +++++++++-- include/linux/nfs_xdr.h | 2 +- 5 files changed, 16 insertions(+), 8 deletions(-) diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 1377ee20ecf9..0fd811ac08b5 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -428,7 +428,7 @@ out_put: hdr->release(hdr); } -static void nfs_read_sync_pgio_error(struct list_head *head) +static void nfs_read_sync_pgio_error(struct list_head *head, int error) { struct nfs_page *req; @@ -820,7 +820,7 @@ out_put: hdr->release(hdr); } -static void nfs_write_sync_pgio_error(struct list_head *head) +static void nfs_write_sync_pgio_error(struct list_head *head, int error) { struct nfs_page *req; diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index 9cbfdb979992..695afb7de3a7 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -993,7 +993,7 @@ nfs_pageio_cleanup_request(struct nfs_pageio_descriptor *desc, LIST_HEAD(head); nfs_list_move_request(req, &head); - desc->pg_completion_ops->error_cleanup(&head); + desc->pg_completion_ops->error_cleanup(&head, desc->pg_error); } /** @@ -1129,7 +1129,8 @@ static void nfs_pageio_error_cleanup(struct nfs_pageio_descriptor *desc) for (midx = 0; midx < desc->pg_mirror_count; midx++) { mirror = &desc->pg_mirrors[midx]; - desc->pg_completion_ops->error_cleanup(&mirror->pg_list); + desc->pg_completion_ops->error_cleanup(&mirror->pg_list, + desc->pg_error); } } diff --git a/fs/nfs/read.c b/fs/nfs/read.c index f9f19784db82..1d95a60b2586 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -205,7 +205,7 @@ static void nfs_initiate_read(struct nfs_pgio_header *hdr, } static void -nfs_async_read_error(struct list_head *head) +nfs_async_read_error(struct list_head *head, int error) { struct nfs_page *req; diff --git a/fs/nfs/write.c b/fs/nfs/write.c index d09c9f878141..11df9f03245f 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -1412,20 +1412,27 @@ static void nfs_redirty_request(struct nfs_page *req) nfs_release_request(req); } -static void nfs_async_write_error(struct list_head *head) +static void nfs_async_write_error(struct list_head *head, int error) { struct nfs_page *req; while (!list_empty(head)) { req = nfs_list_entry(head->next); nfs_list_remove_request(req); + if (nfs_error_is_fatal(error)) { + nfs_context_set_write_error(req->wb_context, error); + if (nfs_error_is_fatal_on_server(error)) { + nfs_write_error_remove_page(req); + continue; + } + } nfs_redirty_request(req); } } static void nfs_async_write_reschedule_io(struct nfs_pgio_header *hdr) { - nfs_async_write_error(&hdr->pages); + nfs_async_write_error(&hdr->pages, 0); filemap_fdatawrite_range(hdr->inode->i_mapping, hdr->args.offset, hdr->args.offset + hdr->args.count - 1); } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 441a93ebcac0..b4bd2bf5f585 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -1549,7 +1549,7 @@ struct nfs_commit_data { }; struct nfs_pgio_completion_ops { - void (*error_cleanup)(struct list_head *head); + void (*error_cleanup)(struct list_head *head, int); void (*init_hdr)(struct nfs_pgio_header *hdr); void (*completion)(struct nfs_pgio_header *hdr); void (*reschedule_io)(struct nfs_pgio_header *hdr); From 875bc3fbf2724134234ddb3069c8e9862b0b19b3 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 18 Feb 2019 13:06:54 -0500 Subject: [PATCH 29/90] NFS: Ensure NFS writeback allocations don't recurse back into NFS. All the allocations that we can hit in the NFS layer and sunrpc layers themselves are already marked as GFP_NOFS, but we need to ensure that any calls to generic kernel functionality do the right thing as well. Signed-off-by: Trond Myklebust --- fs/nfs/write.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 11df9f03245f..d1bc0384ac95 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -26,6 +26,7 @@ #include #include +#include #include "delegation.h" #include "internal.h" @@ -712,11 +713,13 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc) { struct inode *inode = mapping->host; struct nfs_pageio_descriptor pgio; - struct nfs_io_completion *ioc = nfs_io_completion_alloc(GFP_NOFS); + struct nfs_io_completion *ioc; + unsigned int pflags = memalloc_nofs_save(); int err; nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGES); + ioc = nfs_io_completion_alloc(GFP_NOFS); if (ioc) nfs_io_completion_init(ioc, nfs_io_completion_commit, inode); @@ -727,6 +730,8 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc) nfs_pageio_complete(&pgio); nfs_io_completion_put(ioc); + memalloc_nofs_restore(pflags); + if (err < 0) goto out_err; err = pgio.pg_error; From 7dc58ca5d8cd3d81ff329500a502cbe64208d4a1 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 22 Jan 2019 07:39:09 -0500 Subject: [PATCH 30/90] NFS: EINTR is also a fatal error. Signed-off-by: Trond Myklebust --- fs/nfs/internal.h | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index b1e577302518..1cb9670bb8b5 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -755,6 +755,7 @@ static inline bool nfs_error_is_fatal(int err) { switch (err) { case -ERESTARTSYS: + case -EINTR: case -EACCES: case -EDQUOT: case -EFBIG: From 2dc23afffbcac7c8935422beca1ba8636ff310ad Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 13 Feb 2019 08:29:27 -0500 Subject: [PATCH 31/90] NFS: ENOMEM should also be a fatal error. Signed-off-by: Trond Myklebust --- fs/nfs/internal.h | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 1cb9670bb8b5..b26622d9686f 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -764,6 +764,7 @@ static inline bool nfs_error_is_fatal(int err) case -EROFS: case -ESTALE: case -E2BIG: + case -ENOMEM: return true; default: return false; From 302fad7bd5c36de0eb4e94db5c15e12c68a9c842 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 18 Feb 2019 13:32:38 -0500 Subject: [PATCH 32/90] NFS: Fix up documentation warnings Fix up some compiler warnings about function parameters, etc not being correctly described or formatted. Signed-off-by: Trond Myklebust --- fs/nfs/delegation.c | 2 +- fs/nfs/dir.c | 2 +- fs/nfs/file.c | 4 +-- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 12 ++++++--- fs/nfs/inode.c | 33 ++++++++++++----------- fs/nfs/io.c | 12 ++++----- fs/nfs/namespace.c | 8 +++--- fs/nfs/nfs4client.c | 32 ++++++++++++---------- fs/nfs/nfs4namespace.c | 5 ++-- fs/nfs/nfs4proc.c | 8 +++--- fs/nfs/nfs4session.c | 2 +- fs/nfs/nfs4state.c | 1 + fs/nfs/pagelist.c | 7 +++-- fs/nfs/unlink.c | 8 +++--- fs/nfs/write.c | 1 - 15 files changed, 75 insertions(+), 62 deletions(-) diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index 885363ca8569..70e5931f3c60 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -681,7 +681,7 @@ void nfs_expire_all_delegations(struct nfs_client *clp) /** * nfs_super_return_all_delegations - return delegations for one superblock - * @sb: sb to process + * @server: pointer to nfs_server to process * */ void nfs_server_return_all_delegations(struct nfs_server *server) diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 6bf4471850c8..0fb37fac88c9 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -945,7 +945,7 @@ static int nfs_fsync_dir(struct file *filp, loff_t start, loff_t end, /** * nfs_force_lookup_revalidate - Mark the directory as having changed - * @dir - pointer to directory inode + * @dir: pointer to directory inode * * This forces the revalidation code in nfs_lookup_revalidate() to do a * full lookup on all child dentries of 'dir' whenever a change occurs diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 29553fdba8af..e4a33e66f417 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -89,8 +89,8 @@ EXPORT_SYMBOL_GPL(nfs_file_release); /** * nfs_revalidate_size - Revalidate the file size - * @inode - pointer to inode struct - * @file - pointer to struct file + * @inode: pointer to inode struct + * @filp: pointer to struct file * * Revalidates the file length. This is basically a wrapper around * nfs_revalidate_inode() that takes into account the fact that we may diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index 11766a74216d..ca7a6203b3cb 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -483,9 +483,15 @@ ff_layout_get_ds_cred(struct pnfs_layout_segment *lseg, u32 ds_idx, } /** -* Find or create a DS rpc client with th MDS server rpc client auth flavor -* in the nfs_client cl_ds_clients list. -*/ + * nfs4_ff_find_or_create_ds_client - Find or create a DS rpc client + * @lseg: pointer to layout segment + * @ds_idx: mirror index + * @ds_clp: nfs_client for the DS + * @inode: pointer to inode + * + * Find or create a DS rpc client with th MDS server rpc client auth flavor + * in the nfs_client cl_ds_clients list. + */ struct rpc_clnt * nfs4_ff_find_or_create_ds_client(struct pnfs_layout_segment *lseg, u32 ds_idx, struct nfs_client *ds_clp, struct inode *inode) diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 094775ea0781..414a90d48493 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -143,6 +143,7 @@ EXPORT_SYMBOL_GPL(nfs_sync_inode); /** * nfs_sync_mapping - helper to flush all mmapped dirty data to disk + * @mapping: pointer to struct address_space */ int nfs_sync_mapping(struct address_space *mapping) { @@ -1184,8 +1185,8 @@ int nfs_attribute_cache_expired(struct inode *inode) /** * nfs_revalidate_inode - Revalidate the inode attributes - * @server - pointer to nfs_server struct - * @inode - pointer to inode struct + * @server: pointer to nfs_server struct + * @inode: pointer to inode struct * * Updates inode attribute information by retrieving the data from the server. */ @@ -1255,8 +1256,8 @@ out: /** * nfs_revalidate_mapping - Revalidate the pagecache - * @inode - pointer to host inode - * @mapping - pointer to mapping + * @inode: pointer to host inode + * @mapping: pointer to mapping */ int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping) @@ -1371,8 +1372,8 @@ static void nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr *fattr) /** * nfs_check_inode_attributes - verify consistency of the inode attribute cache - * @inode - pointer to inode - * @fattr - updated attributes + * @inode: pointer to inode + * @fattr: updated attributes * * Verifies the attribute cache. If we have just changed the attributes, * so that fattr carries weak cache consistency data, then it may @@ -1572,8 +1573,8 @@ EXPORT_SYMBOL_GPL(_nfs_display_fhandle); /** * nfs_inode_attrs_need_update - check if the inode attributes need updating - * @inode - pointer to inode - * @fattr - attributes + * @inode: pointer to inode + * @fattr: attributes * * Attempt to divine whether or not an RPC call reply carrying stale * attributes got scheduled after another call carrying updated ones. @@ -1614,8 +1615,8 @@ static int nfs_refresh_inode_locked(struct inode *inode, struct nfs_fattr *fattr /** * nfs_refresh_inode - try to update the inode attribute cache - * @inode - pointer to inode - * @fattr - updated attributes + * @inode: pointer to inode + * @fattr: updated attributes * * Check that an RPC call that returned attributes has not overlapped with * other recent updates of the inode metadata, then decide whether it is @@ -1649,8 +1650,8 @@ static int nfs_post_op_update_inode_locked(struct inode *inode, /** * nfs_post_op_update_inode - try to update the inode attribute cache - * @inode - pointer to inode - * @fattr - updated attributes + * @inode: pointer to inode + * @fattr: updated attributes * * After an operation that has changed the inode metadata, mark the * attribute cache as being invalid, then try to update it. @@ -1679,8 +1680,8 @@ EXPORT_SYMBOL_GPL(nfs_post_op_update_inode); /** * nfs_post_op_update_inode_force_wcc_locked - update the inode attribute cache - * @inode - pointer to inode - * @fattr - updated attributes + * @inode: pointer to inode + * @fattr: updated attributes * * After an operation that has changed the inode metadata, mark the * attribute cache as being invalid, then try to update it. Fake up @@ -1731,8 +1732,8 @@ out_noforce: /** * nfs_post_op_update_inode_force_wcc - try to update the inode attribute cache - * @inode - pointer to inode - * @fattr - updated attributes + * @inode: pointer to inode + * @fattr: updated attributes * * After an operation that has changed the inode metadata, mark the * attribute cache as being invalid, then try to update it. Fake up diff --git a/fs/nfs/io.c b/fs/nfs/io.c index 9034b4926909..5088fda9b453 100644 --- a/fs/nfs/io.c +++ b/fs/nfs/io.c @@ -25,7 +25,7 @@ static void nfs_block_o_direct(struct nfs_inode *nfsi, struct inode *inode) /** * nfs_start_io_read - declare the file is being used for buffered reads - * @inode - file inode + * @inode: file inode * * Declare that a buffered read operation is about to start, and ensure * that we block all direct I/O. @@ -56,7 +56,7 @@ nfs_start_io_read(struct inode *inode) /** * nfs_end_io_read - declare that the buffered read operation is done - * @inode - file inode + * @inode: file inode * * Declare that a buffered read operation is done, and release the shared * lock on inode->i_rwsem. @@ -69,7 +69,7 @@ nfs_end_io_read(struct inode *inode) /** * nfs_start_io_write - declare the file is being used for buffered writes - * @inode - file inode + * @inode: file inode * * Declare that a buffered read operation is about to start, and ensure * that we block all direct I/O. @@ -83,7 +83,7 @@ nfs_start_io_write(struct inode *inode) /** * nfs_end_io_write - declare that the buffered write operation is done - * @inode - file inode + * @inode: file inode * * Declare that a buffered write operation is done, and release the * lock on inode->i_rwsem. @@ -105,7 +105,7 @@ static void nfs_block_buffered(struct nfs_inode *nfsi, struct inode *inode) /** * nfs_end_io_direct - declare the file is being used for direct i/o - * @inode - file inode + * @inode: file inode * * Declare that a direct I/O operation is about to start, and ensure * that we block all buffered I/O. @@ -136,7 +136,7 @@ nfs_start_io_direct(struct inode *inode) /** * nfs_end_io_direct - declare that the direct i/o operation is done - * @inode - file inode + * @inode: file inode * * Declare that a direct I/O operation is done, and release the shared * lock on inode->i_rwsem. diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c index e5686be67be8..15f099a24c29 100644 --- a/fs/nfs/namespace.c +++ b/fs/nfs/namespace.c @@ -221,10 +221,10 @@ static struct vfsmount *nfs_do_clone_mount(struct nfs_server *server, /** * nfs_do_submount - set up mountpoint when crossing a filesystem boundary - * @dentry - parent directory - * @fh - filehandle for new root dentry - * @fattr - attributes for new root inode - * @authflavor - security flavor to use when performing the mount + * @dentry: parent directory + * @fh: filehandle for new root dentry + * @fattr: attributes for new root inode + * @authflavor: security flavor to use when performing the mount * */ struct vfsmount *nfs_do_submount(struct dentry *dentry, struct nfs_fh *fh, diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index 2548405da1f7..170115fc43c5 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -42,7 +42,7 @@ static int nfs_get_cb_ident_idr(struct nfs_client *clp, int minorversion) } #ifdef CONFIG_NFS_V4_1 -/** +/* * Per auth flavor data server rpc clients */ struct nfs4_ds_server { @@ -51,7 +51,9 @@ struct nfs4_ds_server { }; /** - * Common lookup case for DS I/O + * nfs4_find_ds_client - Common lookup case for DS I/O + * @ds_clp: pointer to the DS's nfs_client + * @flavor: rpc auth flavour to match */ static struct nfs4_ds_server * nfs4_find_ds_client(struct nfs_client *ds_clp, rpc_authflavor_t flavor) @@ -118,9 +120,13 @@ nfs4_free_ds_server(struct nfs4_ds_server *dss) } /** -* Find or create a DS rpc client with th MDS server rpc client auth flavor -* in the nfs_client cl_ds_clients list. -*/ + * nfs4_find_or_create_ds_client - Find or create a DS rpc client + * @ds_clp: pointer to the DS's nfs_client + * @inode: pointer to the inode + * + * Find or create a DS rpc client with th MDS server rpc client auth flavor + * in the nfs_client cl_ds_clients list. + */ struct rpc_clnt * nfs4_find_or_create_ds_client(struct nfs_client *ds_clp, struct inode *inode) { @@ -284,7 +290,7 @@ static int nfs4_init_callback(struct nfs_client *clp) /** * nfs40_init_client - nfs_client initialization tasks for NFSv4.0 - * @clp - nfs_client to initialize + * @clp: nfs_client to initialize * * Returns zero on success, or a negative errno if some error occurred. */ @@ -312,7 +318,7 @@ int nfs40_init_client(struct nfs_client *clp) /** * nfs41_init_client - nfs_client initialization tasks for NFSv4.1+ - * @clp - nfs_client to initialize + * @clp: nfs_client to initialize * * Returns zero on success, or a negative errno if some error occurred. */ @@ -360,9 +366,7 @@ static int nfs4_init_client_minor_version(struct nfs_client *clp) * nfs4_init_client - Initialise an NFS4 client record * * @clp: nfs_client to initialise - * @timeparms: timeout parameters for underlying RPC transport - * @ip_addr: callback IP address in presentation format - * @authflavor: authentication flavor for underlying RPC transport + * @cl_init: pointer to nfs_client_initdata * * Returns pointer to an NFS client, or an ERR_PTR value. */ @@ -649,13 +653,13 @@ nfs4_check_server_scope(struct nfs41_server_scope *s1, /** * nfs4_detect_session_trunking - Checks for session trunking. - * - * Called after a successful EXCHANGE_ID on a multi-addr connection. - * Upon success, add the transport. - * * @clp: original mount nfs_client * @res: result structure from an exchange_id using the original mount * nfs_client with a new multi_addr transport + * @xprt: pointer to the transport to add. + * + * Called after a successful EXCHANGE_ID on a multi-addr connection. + * Upon success, add the transport. * * Returns zero on success, otherwise -EINVAL * diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c index 24f06dcc2b08..2e460c33ae48 100644 --- a/fs/nfs/nfs4namespace.c +++ b/fs/nfs/nfs4namespace.c @@ -137,6 +137,7 @@ static size_t nfs_parse_server_name(char *string, size_t len, /** * nfs_find_best_sec - Find a security mechanism supported locally + * @clnt: pointer to rpc_clnt * @server: NFS server struct * @flavors: List of security tuples returned by SECINFO procedure * @@ -288,8 +289,8 @@ static struct vfsmount *try_location(struct nfs_clone_mount *mountdata, /** * nfs_follow_referral - set up mountpoint when hitting a referral on moved error - * @dentry - parent directory - * @locations - array of NFSv4 server location information + * @dentry: parent directory + * @locations: array of NFSv4 server location information * */ static struct vfsmount *nfs_follow_referral(struct dentry *dentry, diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 557a5d636183..e8e12faa8604 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -5963,7 +5963,7 @@ out: /** * nfs4_proc_setclientid_confirm - Confirm client ID * @clp: state data structure - * @res: result of a previous SETCLIENTID + * @arg: result of a previous SETCLIENTID * @cred: credential to use for this call * * Returns zero, a negative errno, or a negative NFS4ERR status code. @@ -7527,7 +7527,7 @@ int nfs4_proc_fsid_present(struct inode *inode, const struct cred *cred) return status; } -/** +/* * If 'use_integrity' is true and the state managment nfs_client * cl_rpcclient is using krb5i/p, use the integrity protected cl_rpcclient * and the machine credential as per RFC3530bis and RFC5661 Security @@ -9219,7 +9219,7 @@ nfs4_proc_layoutcommit(struct nfs4_layoutcommit_data *data, bool sync) return status; } -/** +/* * Use the state managment nfs_client cl_rpcclient, which uses krb5i (if * possible) as per RFC3530bis and RFC5661 Security Considerations sections */ @@ -9484,7 +9484,7 @@ static const struct rpc_call_ops nfs41_free_stateid_ops = { * @server: server / transport on which to perform the operation * @stateid: state ID to release * @cred: credential - * @is_recovery: set to true if this call needs to be privileged + * @privileged: set to true if this call needs to be privileged * * Note: this function is always asynchronous. */ diff --git a/fs/nfs/nfs4session.c b/fs/nfs/nfs4session.c index a5489d70a724..25c5255a395c 100644 --- a/fs/nfs/nfs4session.c +++ b/fs/nfs/nfs4session.c @@ -55,7 +55,7 @@ static void nfs4_shrink_slot_table(struct nfs4_slot_table *tbl, u32 newsize) /** * nfs4_slot_tbl_drain_complete - wake waiters when drain is complete - * @tbl - controlling slot table + * @tbl: controlling slot table * */ void nfs4_slot_tbl_drain_complete(struct nfs4_slot_table *tbl) diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 02488b50534a..3de36479ed7a 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -563,6 +563,7 @@ static void nfs4_gc_state_owners(struct nfs_server *server) * nfs4_get_state_owner - Look up a state owner given a credential * @server: nfs_server to search * @cred: RPC credential to match + * @gfp_flags: allocation mode * * Returns a pointer to an instantiated nfs4_state_owner struct, or NULL. */ diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index 695afb7de3a7..e9f39fa5964b 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -350,7 +350,7 @@ nfs_create_request(struct nfs_open_context *ctx, struct page *page, /** * nfs_unlock_request - Unlock request and wake up sleepers. - * @req: + * @req: pointer to request */ void nfs_unlock_request(struct nfs_page *req) { @@ -368,7 +368,7 @@ void nfs_unlock_request(struct nfs_page *req) /** * nfs_unlock_and_release_request - Unlock request and release the nfs_page - * @req: + * @req: pointer to request */ void nfs_unlock_and_release_request(struct nfs_page *req) { @@ -531,7 +531,6 @@ EXPORT_SYMBOL_GPL(nfs_pgio_header_free); * nfs_pgio_rpcsetup - Set up arguments for a pageio call * @hdr: The pageio hdr * @count: Number of bytes to read - * @offset: Initial offset * @how: How to commit data (writes only) * @cinfo: Commit information for the call (writes only) */ @@ -634,7 +633,6 @@ EXPORT_SYMBOL_GPL(nfs_initiate_pgio); /** * nfs_pgio_error - Clean up from a pageio error - * @desc: IO descriptor * @hdr: pageio header */ static void nfs_pgio_error(struct nfs_pgio_header *hdr) @@ -892,6 +890,7 @@ static bool nfs_match_lock_context(const struct nfs_lock_context *l1, * nfs_can_coalesce_requests - test two requests for compatibility * @prev: pointer to nfs_page * @req: pointer to nfs_page + * @pgio: pointer to nfs_pagio_descriptor * * The nfs_page structures 'prev' and 'req' are compared to ensure that the * page data area they describe is contiguous, and that their RPC diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index 79b97b3c4427..52d533967485 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -39,6 +39,7 @@ nfs_free_unlinkdata(struct nfs_unlinkdata *data) /** * nfs_async_unlink_done - Sillydelete post-processing * @task: rpc_task of the sillydelete + * @calldata: pointer to nfs_unlinkdata * * Do the directory attribute update. */ @@ -54,7 +55,7 @@ static void nfs_async_unlink_done(struct rpc_task *task, void *calldata) /** * nfs_async_unlink_release - Release the sillydelete data. - * @task: rpc_task of the sillydelete + * @calldata: struct nfs_unlinkdata to release * * We need to call nfs_put_unlinkdata as a 'tk_release' task since the * rpc_task would be freed too. @@ -159,8 +160,8 @@ static int nfs_call_unlink(struct dentry *dentry, struct inode *inode, struct nf /** * nfs_async_unlink - asynchronous unlinking of a file - * @dir: parent directory of dentry - * @dentry: dentry to unlink + * @dentry: parent directory of dentry + * @name: name of dentry to unlink */ static int nfs_async_unlink(struct dentry *dentry, const struct qstr *name) @@ -324,6 +325,7 @@ static const struct rpc_call_ops nfs_rename_ops = { * @new_dir: target directory for the rename * @old_dentry: original dentry to be renamed * @new_dentry: dentry to which the old_dentry should be renamed + * @complete: Function to run on successful completion * * It's expected that valid references to the dentries and inodes are held */ diff --git a/fs/nfs/write.c b/fs/nfs/write.c index d1bc0384ac95..f3ebabaa291d 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -870,7 +870,6 @@ EXPORT_SYMBOL_GPL(nfs_request_add_commit_list_locked); /** * nfs_request_add_commit_list - add request to a commit list * @req: pointer to a struct nfs_page - * @dst: commit list head * @cinfo: holds list lock and accounting info * * This sets the PG_CLEAN bit, updates the cinfo count of From e9acf2105fdba95953e786612243a4a3a1d0b166 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 22 Jan 2019 14:01:16 -0500 Subject: [PATCH 33/90] NFS: Fix sparse annotations for nfs_set_open_stateid_locked() Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index e8e12faa8604..1dbdc85e6885 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1555,6 +1555,10 @@ static void nfs_clear_open_stateid(struct nfs4_state *state, static void nfs_set_open_stateid_locked(struct nfs4_state *state, const nfs4_stateid *stateid, nfs4_stateid *freeme) + __must_hold(&state->owner->so_lock) + __must_hold(&state->seqlock) + __must_hold(RCU) + { DEFINE_WAIT(wait); int status = 0; From a1231fda7e944adf37d8368b2e182041a39ea1ca Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 18 Feb 2019 10:02:29 -0500 Subject: [PATCH 34/90] SUNRPC: Set memalloc_nofs_save() on all rpciod/xprtiod jobs Set memalloc_nofs_save() on all the rpciod/xprtiod jobs so that we ensure memory allocations for asynchronous rpc calls don't ever end up recursing back to the NFS layer for memory reclaim. Signed-off-by: Trond Myklebust --- net/sunrpc/sched.c | 7 +++++++ net/sunrpc/xprt.c | 3 +++ net/sunrpc/xprtsock.c | 10 +++++++--- 3 files changed, 17 insertions(+), 3 deletions(-) diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c index adc3c40cc733..2168d4d9c09f 100644 --- a/net/sunrpc/sched.c +++ b/net/sunrpc/sched.c @@ -19,6 +19,7 @@ #include #include #include +#include #include @@ -902,7 +903,10 @@ void rpc_execute(struct rpc_task *task) static void rpc_async_schedule(struct work_struct *work) { + unsigned int pflags = memalloc_nofs_save(); + __rpc_execute(container_of(work, struct rpc_task, u.tk_work)); + memalloc_nofs_restore(pflags); } /** @@ -1067,7 +1071,10 @@ static void rpc_free_task(struct rpc_task *task) static void rpc_async_release(struct work_struct *work) { + unsigned int pflags = memalloc_nofs_save(); + rpc_free_task(container_of(work, struct rpc_task, u.tk_work)); + memalloc_nofs_restore(pflags); } static void rpc_release_resources_task(struct rpc_task *task) diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index f1ec2110efeb..b95f4452dbc6 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -49,6 +49,7 @@ #include #include #include +#include #include @@ -643,11 +644,13 @@ static void xprt_autoclose(struct work_struct *work) { struct rpc_xprt *xprt = container_of(work, struct rpc_xprt, task_cleanup); + unsigned int pflags = memalloc_nofs_save(); clear_bit(XPRT_CLOSE_WAIT, &xprt->state); xprt->ops->close(xprt); xprt_release_write(xprt, NULL); wake_up_bit(&xprt->state, XPRT_LOCKED); + memalloc_nofs_restore(pflags); } /** diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 7754aa3e434f..f5d7dcd9e8d9 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -50,6 +50,7 @@ #include #include #include +#include #include @@ -680,7 +681,10 @@ static void xs_stream_data_receive_workfn(struct work_struct *work) { struct sock_xprt *transport = container_of(work, struct sock_xprt, recv_worker); + unsigned int pflags = memalloc_nofs_save(); + xs_stream_data_receive(transport); + memalloc_nofs_restore(pflags); } static void @@ -1378,7 +1382,10 @@ static void xs_udp_data_receive_workfn(struct work_struct *work) { struct sock_xprt *transport = container_of(work, struct sock_xprt, recv_worker); + unsigned int pflags = memalloc_nofs_save(); + xs_udp_data_receive(transport); + memalloc_nofs_restore(pflags); } /** @@ -1869,7 +1876,6 @@ static int xs_local_finish_connecting(struct rpc_xprt *xprt, sk->sk_write_space = xs_udp_write_space; sock_set_flag(sk, SOCK_FASYNC); sk->sk_error_report = xs_error_report; - sk->sk_allocation = GFP_NOIO; xprt_clear_connected(xprt); @@ -2057,7 +2063,6 @@ static void xs_udp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock) sk->sk_data_ready = xs_data_ready; sk->sk_write_space = xs_udp_write_space; sock_set_flag(sk, SOCK_FASYNC); - sk->sk_allocation = GFP_NOIO; xprt_set_connected(xprt); @@ -2220,7 +2225,6 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock) sk->sk_write_space = xs_tcp_write_space; sock_set_flag(sk, SOCK_FASYNC); sk->sk_error_report = xs_error_report; - sk->sk_allocation = GFP_NOIO; /* socket options */ sock_reset_flag(sk, SOCK_LINGER); From 0ffe86f48026b7f34db22d1004bc9992f0db8b33 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 30 Jan 2019 14:51:26 -0500 Subject: [PATCH 35/90] SUNRPC: Use poll() to fix up the socket requeue races Because we clear XPRT_SOCK_DATA_READY before reading, we can end up with a situation where new data arrives, causing xs_data_ready() to queue up a second receive worker job for the same socket, which then immediately gets stuck waiting on the transport receive mutex. The fix is to only clear XPRT_SOCK_DATA_READY once we're done reading, and then to use poll() to check if we might need to queue up a new job in order to deal with any new data. Signed-off-by: Trond Myklebust --- net/sunrpc/xprtsock.c | 26 ++++++++++++++++++++++++-- 1 file changed, 24 insertions(+), 2 deletions(-) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index f5d7dcd9e8d9..da45bb1e931e 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -656,13 +656,34 @@ out_err: return ret != 0 ? ret : -ESHUTDOWN; } +static __poll_t xs_poll_socket(struct sock_xprt *transport) +{ + return transport->sock->ops->poll(NULL, transport->sock, NULL); +} + +static bool xs_poll_socket_readable(struct sock_xprt *transport) +{ + __poll_t events = xs_poll_socket(transport); + + return (events & (EPOLLIN | EPOLLRDNORM)) && !(events & EPOLLRDHUP); +} + +static void xs_poll_check_readable(struct sock_xprt *transport) +{ + + clear_bit(XPRT_SOCK_DATA_READY, &transport->sock_state); + if (!xs_poll_socket_readable(transport)) + return; + if (!test_and_set_bit(XPRT_SOCK_DATA_READY, &transport->sock_state)) + queue_work(xprtiod_workqueue, &transport->recv_worker); +} + static void xs_stream_data_receive(struct sock_xprt *transport) { size_t read = 0; ssize_t ret = 0; mutex_lock(&transport->recv_mutex); - clear_bit(XPRT_SOCK_DATA_READY, &transport->sock_state); if (transport->sock == NULL) goto out; for (;;) { @@ -672,6 +693,7 @@ static void xs_stream_data_receive(struct sock_xprt *transport) read += ret; cond_resched(); } + xs_poll_check_readable(transport); out: mutex_unlock(&transport->recv_mutex); trace_xs_stream_read_data(&transport->xprt, ret, read); @@ -1362,7 +1384,6 @@ static void xs_udp_data_receive(struct sock_xprt *transport) int err; mutex_lock(&transport->recv_mutex); - clear_bit(XPRT_SOCK_DATA_READY, &transport->sock_state); sk = transport->inet; if (sk == NULL) goto out; @@ -1374,6 +1395,7 @@ static void xs_udp_data_receive(struct sock_xprt *transport) consume_skb(skb); cond_resched(); } + xs_poll_check_readable(transport); out: mutex_unlock(&transport->recv_mutex); } From b9779a54bb224cd2a993c9bf8a1910a7494c062e Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 2 Jan 2019 15:54:42 -0500 Subject: [PATCH 36/90] SUNRPC: Ensure rq_bytes_sent is reset before request transmission When we resend a request, ensure that the 'rq_bytes_sent' is reset to zero. Signed-off-by: Trond Myklebust --- net/sunrpc/backchannel_rqst.c | 1 - net/sunrpc/xprt.c | 2 -- net/sunrpc/xprtrdma/backchannel.c | 1 - net/sunrpc/xprtrdma/transport.c | 1 - net/sunrpc/xprtsock.c | 2 -- 5 files changed, 7 deletions(-) diff --git a/net/sunrpc/backchannel_rqst.c b/net/sunrpc/backchannel_rqst.c index ec451b8114b0..b9313c15ee3a 100644 --- a/net/sunrpc/backchannel_rqst.c +++ b/net/sunrpc/backchannel_rqst.c @@ -252,7 +252,6 @@ static struct rpc_rqst *xprt_alloc_bc_request(struct rpc_xprt *xprt, __be32 xid) req = list_first_entry(&xprt->bc_pa_list, struct rpc_rqst, rq_bc_pa_list); req->rq_reply_bytes_recvd = 0; - req->rq_bytes_sent = 0; memcpy(&req->rq_private_buf, &req->rq_rcv_buf, sizeof(req->rq_private_buf)); req->rq_xid = xid; diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index b95f4452dbc6..1587728f40d1 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -1602,7 +1602,6 @@ xprt_request_init(struct rpc_task *task) req->rq_buffer = NULL; req->rq_xid = xprt_alloc_xid(xprt); xprt_init_connect_cookie(req, xprt); - req->rq_bytes_sent = 0; req->rq_snd_buf.len = 0; req->rq_snd_buf.buflen = 0; req->rq_rcv_buf.len = 0; @@ -1752,7 +1751,6 @@ xprt_init_bc_request(struct rpc_rqst *req, struct rpc_task *task) */ xbufp->len = xbufp->head[0].iov_len + xbufp->page_len + xbufp->tail[0].iov_len; - req->rq_bytes_sent = 0; } #endif diff --git a/net/sunrpc/xprtrdma/backchannel.c b/net/sunrpc/xprtrdma/backchannel.c index 0de9b3e63770..749452724e6e 100644 --- a/net/sunrpc/xprtrdma/backchannel.c +++ b/net/sunrpc/xprtrdma/backchannel.c @@ -267,7 +267,6 @@ void rpcrdma_bc_receive_call(struct rpcrdma_xprt *r_xprt, /* Prepare rqst */ rqst->rq_reply_bytes_recvd = 0; - rqst->rq_bytes_sent = 0; rqst->rq_xid = *p; rqst->rq_private_buf.len = size; diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c index fbc171ebfe91..3c490172d383 100644 --- a/net/sunrpc/xprtrdma/transport.c +++ b/net/sunrpc/xprtrdma/transport.c @@ -738,7 +738,6 @@ xprt_rdma_send_request(struct rpc_rqst *rqst) goto drop_connection; rqst->rq_xmit_bytes_sent += rqst->rq_snd_buf.len; - rqst->rq_bytes_sent = 0; /* An RPC with no reply will throw off credit accounting, * so drop the connection to reset the credit grant. diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index da45bb1e931e..5d22c175c7dd 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -951,7 +951,6 @@ static int xs_local_send_request(struct rpc_rqst *req) req->rq_bytes_sent = transport->xmit.offset; if (likely(req->rq_bytes_sent >= req->rq_slen)) { req->rq_xmit_bytes_sent += transport->xmit.offset; - req->rq_bytes_sent = 0; transport->xmit.offset = 0; return 0; } @@ -1117,7 +1116,6 @@ static int xs_tcp_send_request(struct rpc_rqst *req) req->rq_bytes_sent = transport->xmit.offset; if (likely(req->rq_bytes_sent >= req->rq_slen)) { req->rq_xmit_bytes_sent += transport->xmit.offset; - req->rq_bytes_sent = 0; transport->xmit.offset = 0; return 0; } From 45bb8d802742842fa974b0d7d474d115df1d07db Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Sun, 23 Dec 2018 09:57:10 +0100 Subject: [PATCH 37/90] NFS: drop useless LIST_HEAD Drop LIST_HEAD where the variable it declares has never been used. The semantic patch that fixes this problem is as follows: (http://coccinelle.lip6.fr/) // @@ identifier x; @@ - LIST_HEAD(x); ... when != x // Fixes: 0e20162ed1e9 ("NFSv4.1 Use MDS auth flavor for data server connection") Signed-off-by: Julia Lawall Signed-off-by: Trond Myklebust --- fs/nfs/nfs4client.c | 1 - 1 file changed, 1 deletion(-) diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index 170115fc43c5..1339ede979af 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -151,7 +151,6 @@ static void nfs4_shutdown_ds_clients(struct nfs_client *clp) { struct nfs4_ds_server *dss; - LIST_HEAD(shutdown_list); while (!list_empty(&clp->cl_ds_clients)) { dss = list_entry(clp->cl_ds_clients.next, From e3735c89985415f4fb509c67963e3d05969fbdb1 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 7 Jan 2019 17:53:52 +1100 Subject: [PATCH 38/90] SUNRPC: remove pointless test in unx_match() As reported by Dan Carpenter, this test for acred->cred being set is inconsistent with the dereference of the pointer a few lines earlier. An 'auth_cred' *always* has ->cred set - every place that creates one initializes this field, often as the first thing done. So remove this test. Reported-by: Dan Carpenter Signed-off-by: NeilBrown Signed-off-by: Trond Myklebust --- net/sunrpc/auth_unix.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sunrpc/auth_unix.c b/net/sunrpc/auth_unix.c index 387f6b3ffbea..770e338a9b11 100644 --- a/net/sunrpc/auth_unix.c +++ b/net/sunrpc/auth_unix.c @@ -87,7 +87,7 @@ unx_match(struct auth_cred *acred, struct rpc_cred *cred, int flags) if (!uid_eq(cred->cr_cred->fsuid, acred->cred->fsuid) || !gid_eq(cred->cr_cred->fsgid, acred->cred->fsgid)) return 0; - if (acred->cred && acred->cred->group_info != NULL) + if (acred->cred->group_info != NULL) groups = acred->cred->group_info->ngroups; if (groups > UNX_NGROUPS) groups = UNX_NGROUPS; From 40cc394be1aa18848b8757e03bd8ed23281f572e Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 30 Jan 2019 07:58:38 -0600 Subject: [PATCH 39/90] fs/nfs: Fix nfs_parse_devname to not modify it's argument In the rare and unsupported case of a hostname list nfs_parse_devname will modify dev_name. There is no need to modify dev_name as the all that is being computed is the length of the hostname, so the computed length can just be shorted. Fixes: dc04589827f7 ("NFS: Use common device name parsing logic for NFSv4 and NFSv2/v3") Signed-off-by: "Eric W. Biederman" Signed-off-by: Trond Myklebust --- fs/nfs/super.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 0570391eaa16..23790c7b2289 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -1919,7 +1919,7 @@ static int nfs_parse_devname(const char *dev_name, /* kill possible hostname list: not supported */ comma = strchr(dev_name, ','); if (comma != NULL && comma < end) - *comma = 0; + len = comma - dev_name; } if (len > maxnamlen) From be4c2d4723a4a637f0d1b4f7c66447141a4b3564 Mon Sep 17 00:00:00 2001 From: luanshi Date: Tue, 29 Jan 2019 15:34:17 +0800 Subject: [PATCH 40/90] NFS: readdirplus optimization by cache mechanism When listing very large directories via NFS, clients may take a long time to complete. There are about three factors involved: First of all, ls and practically every other method of listing a directory including python os.listdir and find rely on libc readdir(). However readdir() only reads 32K of directory entries at a time, which means that if you have a lot of files in the same directory, it is going to take an insanely long time to read all the directory entries. Secondly, libc readdir() reads 32K of directory entries at a time, in kernel space 32K buffer split into 8 pages. One NFS readdirplus rpc will be called for one page, which introduces many readdirplus rpc calls. Lastly, one NFS readdirplus rpc asks for 32K data (filled by nfs_dentry) to fill one page (filled by dentry), we found that nearly one third of data was wasted. To solve above problems, pagecache mechanism was introduced. One NFS readdirplus rpc will ask for a large data (more than 32k), the data can fill more than one page, the cached pages can be used for next readdir call. This can reduce many readdirplus rpc calls and improve readdirplus performance. TESTING: When listing very large directories(include 300 thousand files) via NFS time ls -l /nfs_mount | wc -l without the patch: 300001 real 1m53.524s user 0m2.314s sys 0m2.599s with the patch: 300001 real 0m23.487s user 0m2.305s sys 0m2.558s Improved performance: 79.6% readdirplus rpc calls decrease: 85% Signed-off-by: Liguang Zhang Signed-off-by: Trond Myklebust --- fs/nfs/dir.c | 90 +++++++++++++++++++++++++++++++++++++++++++---- fs/nfs/internal.h | 3 +- 2 files changed, 86 insertions(+), 7 deletions(-) diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 0fb37fac88c9..6b3c547154b3 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -139,12 +139,19 @@ struct nfs_cache_array { struct nfs_cache_array_entry array[0]; }; +struct readdirvec { + unsigned long nr; + unsigned long index; + struct page *pages[NFS_MAX_READDIR_RAPAGES]; +}; + typedef int (*decode_dirent_t)(struct xdr_stream *, struct nfs_entry *, bool); typedef struct { struct file *file; struct page *page; struct dir_context *ctx; unsigned long page_index; + struct readdirvec pvec; u64 *dir_cookie; u64 last_cookie; loff_t current_index; @@ -524,6 +531,10 @@ int nfs_readdir_page_filler(nfs_readdir_descriptor_t *desc, struct nfs_entry *en struct nfs_cache_array *array; unsigned int count = 0; int status; + int max_rapages = NFS_MAX_READDIR_RAPAGES; + + desc->pvec.index = desc->page_index; + desc->pvec.nr = 0; scratch = alloc_page(GFP_KERNEL); if (scratch == NULL) @@ -548,20 +559,40 @@ int nfs_readdir_page_filler(nfs_readdir_descriptor_t *desc, struct nfs_entry *en if (desc->plus) nfs_prime_dcache(file_dentry(desc->file), entry); - status = nfs_readdir_add_to_array(entry, page); + status = nfs_readdir_add_to_array(entry, desc->pvec.pages[desc->pvec.nr]); + if (status == -ENOSPC) { + desc->pvec.nr++; + if (desc->pvec.nr == max_rapages) + break; + status = nfs_readdir_add_to_array(entry, desc->pvec.pages[desc->pvec.nr]); + } if (status != 0) break; } while (!entry->eof); + /* + * page and desc->pvec.pages[0] are valid, don't need to check + * whether or not to be NULL. + */ + copy_highpage(page, desc->pvec.pages[0]); + out_nopages: if (count == 0 || (status == -EBADCOOKIE && entry->eof != 0)) { - array = kmap(page); + array = kmap_atomic(desc->pvec.pages[desc->pvec.nr]); array->eof_index = array->size; status = 0; - kunmap(page); + kunmap_atomic(array); } put_page(scratch); + + /* + * desc->pvec.nr > 0 means at least one page was completely filled, + * we should return -ENOSPC. Otherwise function + * nfs_readdir_xdr_to_array will enter infinite loop. + */ + if (desc->pvec.nr > 0) + return -ENOSPC; return status; } @@ -595,6 +626,24 @@ out_freepages: return -ENOMEM; } +/* + * nfs_readdir_rapages_init initialize rapages by nfs_cache_array structure. + */ +static +void nfs_readdir_rapages_init(nfs_readdir_descriptor_t *desc) +{ + struct nfs_cache_array *array; + int max_rapages = NFS_MAX_READDIR_RAPAGES; + int index; + + for (index = 0; index < max_rapages; index++) { + array = kmap_atomic(desc->pvec.pages[index]); + memset(array, 0, sizeof(struct nfs_cache_array)); + array->eof_index = -1; + kunmap_atomic(array); + } +} + static int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page, struct inode *inode) { @@ -605,6 +654,12 @@ int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page, int status = -ENOMEM; unsigned int array_size = ARRAY_SIZE(pages); + /* + * This means we hit readdir rdpages miss, the preallocated rdpages + * are useless, the preallocate rdpages should be reinitialized. + */ + nfs_readdir_rapages_init(desc); + entry.prev_cookie = 0; entry.cookie = desc->last_cookie; entry.eof = 0; @@ -664,9 +719,24 @@ int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page* page) struct inode *inode = file_inode(desc->file); int ret; - ret = nfs_readdir_xdr_to_array(desc, page, inode); - if (ret < 0) - goto error; + /* + * If desc->page_index in range desc->pvec.index and + * desc->pvec.index + desc->pvec.nr, we get readdir cache hit. + */ + if (desc->page_index >= desc->pvec.index && + desc->page_index < (desc->pvec.index + desc->pvec.nr)) { + /* + * page and desc->pvec.pages[x] are valid, don't need to check + * whether or not to be NULL. + */ + copy_highpage(page, desc->pvec.pages[desc->page_index - desc->pvec.index]); + ret = 0; + } else { + ret = nfs_readdir_xdr_to_array(desc, page, inode); + if (ret < 0) + goto error; + } + SetPageUptodate(page); if (invalidate_inode_pages2_range(inode->i_mapping, page->index + 1, -1) < 0) { @@ -831,6 +901,7 @@ static int nfs_readdir(struct file *file, struct dir_context *ctx) *desc = &my_desc; struct nfs_open_dir_context *dir_ctx = file->private_data; int res = 0; + int max_rapages = NFS_MAX_READDIR_RAPAGES; dfprintk(FILE, "NFS: readdir(%pD2) starting at cookie %llu\n", file, (long long)ctx->pos); @@ -850,6 +921,12 @@ static int nfs_readdir(struct file *file, struct dir_context *ctx) desc->decode = NFS_PROTO(inode)->decode_dirent; desc->plus = nfs_use_readdirplus(inode, ctx); + res = nfs_readdir_alloc_pages(desc->pvec.pages, max_rapages); + if (res < 0) + return -ENOMEM; + + nfs_readdir_rapages_init(desc); + if (ctx->pos == 0 || nfs_attribute_cache_expired(inode)) res = nfs_revalidate_mapping(inode, file->f_mapping); if (res < 0) @@ -885,6 +962,7 @@ static int nfs_readdir(struct file *file, struct dir_context *ctx) break; } while (!desc->eof); out: + nfs_readdir_free_pages(desc->pvec.pages, max_rapages); if (res > 0) res = 0; dfprintk(FILE, "NFS: readdir(%pD2) returns %d\n", file, res); diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index b26622d9686f..c7cf23ae6597 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -69,7 +69,8 @@ struct nfs_clone_mount { * Maximum number of pages that readdir can use for creating * a vmapped array of pages. */ -#define NFS_MAX_READDIR_PAGES 8 +#define NFS_MAX_READDIR_PAGES 64 +#define NFS_MAX_READDIR_RAPAGES 8 struct nfs_client_initdata { unsigned long init_flags; From 42f72cf368c502c435af4e206e26d651cfb7d9ad Mon Sep 17 00:00:00 2001 From: zhangliguang Date: Tue, 12 Feb 2019 09:38:33 +0800 Subject: [PATCH 41/90] NFS: Remove redundant semicolon This removes redundant semicolon for ending code. Fixes: c7944ebb9ce9 ("NFSv4: Fix lookup revalidate of regular files") Signed-off-by: Liguang Zhang Signed-off-by: Trond Myklebust --- fs/nfs/dir.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 6b3c547154b3..a112b67a3f8f 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1727,7 +1727,7 @@ nfs4_do_lookup_revalidate(struct inode *dir, struct dentry *dentry, reval_dentry: if (flags & LOOKUP_RCU) return -ECHILD; - return nfs_lookup_revalidate_dentry(dir, dentry, inode);; + return nfs_lookup_revalidate_dentry(dir, dentry, inode); full_reval: return nfs_do_lookup_revalidate(dir, dentry, flags); From bf211ca1a84c570607137f4a556870ac034e318c Mon Sep 17 00:00:00 2001 From: zhangliguang Date: Sat, 16 Feb 2019 08:38:40 +0800 Subject: [PATCH 42/90] NFS: Fix typo in comments of nfs_readdir_alloc_pages() This fixes the typo in comments of nfs_readdir_alloc_pages(). Because nfs_readdir_large_page and nfs_readdir_free_pagearray had been renamed. Signed-off-by: Liguang Zhang Signed-off-by: Trond Myklebust --- fs/nfs/dir.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index a112b67a3f8f..a71d0b42d160 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -605,8 +605,8 @@ void nfs_readdir_free_pages(struct page **pages, unsigned int npages) } /* - * nfs_readdir_large_page will allocate pages that must be freed with a call - * to nfs_readdir_free_pagearray + * nfs_readdir_alloc_pages() will allocate pages that must be freed with a call + * to nfs_readdir_free_pages() */ static int nfs_readdir_alloc_pages(struct page **pages, unsigned int npages) From 97ae91bbf3a70fc8cee3c9030564cfc892cc8cee Mon Sep 17 00:00:00 2001 From: Kazuo Ito Date: Thu, 14 Feb 2019 18:36:58 +0900 Subject: [PATCH 43/90] pNFS: Fix potential corruption of page being written nfs_want_read_modify_write() didn't check for !PagePrivate when pNFS block or SCSI layout was in use, therefore we could lose data forever if the page being written was filled by a read before completion. Signed-off-by: Kazuo Ito Signed-off-by: Trond Myklebust --- fs/nfs/file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nfs/file.c b/fs/nfs/file.c index e4a33e66f417..7086d5677ed3 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -299,7 +299,7 @@ static int nfs_want_read_modify_write(struct file *file, struct page *page, unsigned int end = offset + len; if (pnfs_ld_read_whole_page(file->f_mapping->host)) { - if (!PageUptodate(page)) + if (!PageUptodate(page) && !PagePrivate(page)) return 1; return 0; } From 2cde04e90d5be46b4b6655b965b496e6b6f18e49 Mon Sep 17 00:00:00 2001 From: Kazuo Ito Date: Thu, 14 Feb 2019 18:39:03 +0900 Subject: [PATCH 44/90] pNFS: Avoid read/modify/write when it is not necessary As the block and SCSI layouts can only read/write fixed-length blocks, we must perform read-modify-write when data to be written is not aligned to a block boundary or smaller than the block size. (612aa983a0410 pnfs: add flag to force read-modify-write in ->write_begin) The current code tries to see if we have to do read-modify-write on block-oriented pNFS layouts by just checking !PageUptodate(page), but the same condition also applies for overwriting of any uncached potions of existing files, making such operations excessively slow even it is block-aligned. The change does not affect the optimization for modify-write-read cases (38c73044f5f4d NFS: read-modify-write page updating), because partial update of !PageUptodate() pages can only happen in layouts that can do arbitrary length read/write and never in block-based ones. Testing results: We ran fio on one of the pNFS clients running 4.20 kernel (vanilla and patched) in this configuration to read/write/overwrite files on the storage array, exported as pnfs share by the server. pNFS clients ---1G Ethernet--- pNFS server (HP DL360 G8) (HP DL360 G8) | | | | +------8G Fiber Channel--------+ | Storage Array (HP P6350) Throughput of overwrite (both buffered and O_SYNC) is noticeably improved. Ops. |block size| Throughput | | (KiB) | (MiB/s) | | | 4.20 | patched| ---------+----------+----------------+ buffered | 4| 21.3 | 232 | overwrite| 32| 22.2 | 256 | | 512| 22.4 | 260 | ---------+----------+----------------+ O_SYNC | 4| 3.84| 4.77| overwrite| 32| 12.2 | 32.0 | | 512| 18.5 | 152 | ---------+----------+----------------+ Read and write (buffered and O_SYNC) by the same client remain unchanged by the patch either negatively or positively, as they should do. Ops. |block size| Throughput | | (KiB) | (MiB/s) | | | 4.20 | patched| ---------+----------+----------------+ read | 4| 548 | 550 | | 32| 547 | 551 | | 512| 548 | 551 | ---------+----------+----------------+ buffered | 4| 237 | 244 | write | 32| 261 | 268 | | 512| 265 | 272 | ---------+----------+----------------+ O_SYNC | 4| 0.46| 0.46| write | 32| 3.60| 3.57| | 512| 105 | 106 | ---------+----------+----------------+ Signed-off-by: Kazuo Ito Tested-by: Hiroyuki Watanabe Signed-off-by: Trond Myklebust --- fs/nfs/file.c | 40 ++++++++++++++++++++++++++-------------- 1 file changed, 26 insertions(+), 14 deletions(-) diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 7086d5677ed3..4899b85f9b3c 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -276,6 +276,12 @@ EXPORT_SYMBOL_GPL(nfs_file_fsync); * then a modify/write/read cycle when writing to a page in the * page cache. * + * Some pNFS layout drivers can only read/write at a certain block + * granularity like all block devices and therefore we must perform + * read/modify/write whenever a page hasn't read yet and the data + * to be written there is not aligned to a block boundary and/or + * smaller than the block size. + * * The modify/write/read cycle may occur if a page is read before * being completely filled by the writer. In this situation, the * page must be completely written to stable storage on the server @@ -291,26 +297,32 @@ EXPORT_SYMBOL_GPL(nfs_file_fsync); * and that the new data won't completely replace the old data in * that range of the file. */ -static int nfs_want_read_modify_write(struct file *file, struct page *page, - loff_t pos, unsigned len) +static bool nfs_full_page_write(struct page *page, loff_t pos, unsigned int len) { unsigned int pglen = nfs_page_length(page); unsigned int offset = pos & (PAGE_SIZE - 1); unsigned int end = offset + len; - if (pnfs_ld_read_whole_page(file->f_mapping->host)) { - if (!PageUptodate(page) && !PagePrivate(page)) - return 1; - return 0; - } + return !pglen || (end >= pglen && !offset); +} - if ((file->f_mode & FMODE_READ) && /* open for read? */ - !PageUptodate(page) && /* Uptodate? */ - !PagePrivate(page) && /* i/o request already? */ - pglen && /* valid bytes of file? */ - (end < pglen || offset)) /* replace all valid bytes? */ - return 1; - return 0; +static bool nfs_want_read_modify_write(struct file *file, struct page *page, + loff_t pos, unsigned int len) +{ + /* + * Up-to-date pages, those with ongoing or full-page write + * don't need read/modify/write + */ + if (PageUptodate(page) || PagePrivate(page) || + nfs_full_page_write(page, pos, len)) + return false; + + if (pnfs_ld_read_whole_page(file->f_mapping->host)) + return true; + /* Open for reading too? */ + if (file->f_mode & FMODE_READ) + return true; + return false; } /* From ded52fbe7020a5696b0b0a0fdbf234e37bf16f94 Mon Sep 17 00:00:00 2001 From: ZhangXiaoxu Date: Mon, 18 Feb 2019 22:56:43 +0800 Subject: [PATCH 45/90] nfs: fix xfstest generic/099 failed on nfsv3 After setxattr, the nfsv3 cached the acl which set by user. But at the backend, the shared file system (eg. ext4) will check the acl, if it can merged with mode, it won't add acl to the file. So, the nfsv3 cached acl is redundant. Don't 'set_cached_acl' when setxattr. Signed-off-by: ZhangXiaoxu Signed-off-by: Trond Myklebust --- fs/nfs/nfs3acl.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c index 9fce18548f7e..c5c3fc6e6c60 100644 --- a/fs/nfs/nfs3acl.c +++ b/fs/nfs/nfs3acl.c @@ -222,8 +222,6 @@ static int __nfs3_proc_setacls(struct inode *inode, struct posix_acl *acl, switch (status) { case 0: status = nfs_refresh_inode(inode, fattr); - set_cached_acl(inode, ACL_TYPE_ACCESS, acl); - set_cached_acl(inode, ACL_TYPE_DEFAULT, dfacl); break; case -EPFNOSUPPORT: case -EPROTONOSUPPORT: From ae0535515161d3a131f2acabb7b883e3f3d45235 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 20 Feb 2019 14:56:20 -0500 Subject: [PATCH 46/90] SUNRPC: Don't reset the stream record info when the receive worker is running To ensure that the receive worker has exclusive access to the stream record info, we must not reset the contents other than when holding the transport->recv_mutex. Signed-off-by: Trond Myklebust --- net/sunrpc/xprtsock.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 5d22c175c7dd..ec0bb153c140 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -716,6 +716,11 @@ xs_stream_reset_connect(struct sock_xprt *transport) transport->recv.len = 0; transport->recv.copied = 0; transport->xmit.offset = 0; +} + +static void +xs_stream_start_connect(struct sock_xprt *transport) +{ transport->xprt.stat.connect_count++; transport->xprt.stat.connect_start = jiffies; } @@ -1255,6 +1260,8 @@ static void xs_reset_transport(struct sock_xprt *transport) xprt_clear_connected(xprt); write_unlock_bh(&sk->sk_callback_lock); xs_sock_reset_connection_flags(xprt); + /* Reset stream record info */ + xs_stream_reset_connect(transport); mutex_unlock(&transport->recv_mutex); trace_rpc_socket_close(xprt, sock); @@ -1906,7 +1913,7 @@ static int xs_local_finish_connecting(struct rpc_xprt *xprt, write_unlock_bh(&sk->sk_callback_lock); } - xs_stream_reset_connect(transport); + xs_stream_start_connect(transport); return kernel_connect(sock, xs_addr(xprt), xprt->addrlen, 0); } @@ -2264,8 +2271,7 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock) xs_set_memalloc(xprt); - /* Reset TCP record info */ - xs_stream_reset_connect(transport); + xs_stream_start_connect(transport); /* Tell the socket layer to start connecting... */ set_bit(XPRT_SOCK_CONNECTING, &transport->sock_state); From e92053a52e68b841d1314ca04da7cbeb326ee1a0 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 15 Feb 2019 18:14:27 -0500 Subject: [PATCH 47/90] SUNRPC: Handle zero length fragments correctly A zero length fragment is really a bug, but let's ensure we don't go nuts when one turns up. Signed-off-by: Trond Myklebust --- net/sunrpc/xprtsock.c | 43 +++++++++++++++++++++++++++++-------------- 1 file changed, 29 insertions(+), 14 deletions(-) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index ec0bb153c140..492cec3f1451 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -405,8 +405,8 @@ xs_read_xdr_buf(struct socket *sock, struct msghdr *msg, int flags, size_t want, seek_init = seek, offset = 0; ssize_t ret; - if (seek < buf->head[0].iov_len) { - want = min_t(size_t, count, buf->head[0].iov_len); + want = min_t(size_t, count, buf->head[0].iov_len); + if (seek < want) { ret = xs_read_kvec(sock, msg, flags, &buf->head[0], want, seek); if (ret <= 0) goto sock_err; @@ -417,8 +417,8 @@ xs_read_xdr_buf(struct socket *sock, struct msghdr *msg, int flags, goto out; seek = 0; } else { - seek -= buf->head[0].iov_len; - offset += buf->head[0].iov_len; + seek -= want; + offset += want; } want = xs_alloc_sparse_pages(buf, @@ -443,8 +443,8 @@ xs_read_xdr_buf(struct socket *sock, struct msghdr *msg, int flags, offset += want; } - if (seek < buf->tail[0].iov_len) { - want = min_t(size_t, count - offset, buf->tail[0].iov_len); + want = min_t(size_t, count - offset, buf->tail[0].iov_len); + if (seek < want) { ret = xs_read_kvec(sock, msg, flags, &buf->tail[0], want, seek); if (ret <= 0) goto sock_err; @@ -454,7 +454,7 @@ xs_read_xdr_buf(struct socket *sock, struct msghdr *msg, int flags, if (ret != want) goto out; } else - offset += buf->tail[0].iov_len; + offset = seek_init; ret = -EMSGSIZE; out: *read = offset - seek_init; @@ -482,6 +482,14 @@ xs_read_stream_request_done(struct sock_xprt *transport) return transport->recv.fraghdr & cpu_to_be32(RPC_LAST_STREAM_FRAGMENT); } +static void +xs_read_stream_check_eor(struct sock_xprt *transport, + struct msghdr *msg) +{ + if (xs_read_stream_request_done(transport)) + msg->msg_flags |= MSG_EOR; +} + static ssize_t xs_read_stream_request(struct sock_xprt *transport, struct msghdr *msg, int flags, struct rpc_rqst *req) @@ -493,17 +501,24 @@ xs_read_stream_request(struct sock_xprt *transport, struct msghdr *msg, xs_read_header(transport, buf); want = transport->recv.len - transport->recv.offset; - ret = xs_read_xdr_buf(transport->sock, msg, flags, buf, - transport->recv.copied + want, transport->recv.copied, - &read); - transport->recv.offset += read; - transport->recv.copied += read; + if (want != 0) { + ret = xs_read_xdr_buf(transport->sock, msg, flags, buf, + transport->recv.copied + want, + transport->recv.copied, + &read); + transport->recv.offset += read; + transport->recv.copied += read; + } else + read = 0; + if (transport->recv.offset == transport->recv.len) { - if (xs_read_stream_request_done(transport)) - msg->msg_flags |= MSG_EOR; + xs_read_stream_check_eor(transport, msg); return read; } + if (want == 0) + return 0; + switch (ret) { default: break; From 727fcc64a0f8d346bdd60a7c92e9e720228ce037 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 15 Feb 2019 16:53:04 -0500 Subject: [PATCH 48/90] SUNRPC: Don't suppress socket errors when a message read completes If the message read completes, but the socket returned an error condition, we should ensure to propagate that error. Signed-off-by: Trond Myklebust --- net/sunrpc/xprtsock.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 492cec3f1451..e51716e88899 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -508,14 +508,11 @@ xs_read_stream_request(struct sock_xprt *transport, struct msghdr *msg, &read); transport->recv.offset += read; transport->recv.copied += read; - } else - read = 0; - - if (transport->recv.offset == transport->recv.len) { - xs_read_stream_check_eor(transport, msg); - return read; } + if (transport->recv.offset == transport->recv.len) + xs_read_stream_check_eor(transport, msg); + if (want == 0) return 0; From 5f52a9d429b8ba8f9d669730adf16bc534eb74ea Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 16 Feb 2019 09:31:47 -0500 Subject: [PATCH 49/90] SUNRPC: Initiate a connection close on an ESHUTDOWN error in stream receive If the client stream receive code receives an ESHUTDOWN error either because the server closed the connection, or because it sent a callback which cannot be processed, then we should shut down the connection. Signed-off-by: Trond Myklebust --- net/sunrpc/xprtsock.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index e51716e88899..a3aadc04808f 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -705,7 +705,10 @@ static void xs_stream_data_receive(struct sock_xprt *transport) read += ret; cond_resched(); } - xs_poll_check_readable(transport); + if (ret == -ESHUTDOWN) + kernel_sock_shutdown(transport->sock, SHUT_RDWR); + else + xs_poll_check_readable(transport); out: mutex_unlock(&transport->recv_mutex); trace_xs_stream_read_data(&transport->xprt, ret, read); From e791f8e9380d945ee9ed75aba2d65a9369f01820 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 19 Feb 2019 12:39:40 -0500 Subject: [PATCH 50/90] SUNRPC: Convert xs_send_kvec() to use iov_iter_kvec() Prepare to the socket transmission code to use iov_iter. Signed-off-by: Trond Myklebust --- net/sunrpc/xprtsock.c | 38 +++++++++++++++++++++++--------------- 1 file changed, 23 insertions(+), 15 deletions(-) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index a3aadc04808f..e1546dd6aacc 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -742,21 +742,21 @@ xs_stream_start_connect(struct sock_xprt *transport) #define XS_SENDMSG_FLAGS (MSG_DONTWAIT | MSG_NOSIGNAL) -static int xs_send_kvec(struct socket *sock, struct sockaddr *addr, int addrlen, struct kvec *vec, unsigned int base, int more) +static int xs_sendmsg(struct socket *sock, struct msghdr *msg, size_t seek) { - struct msghdr msg = { - .msg_name = addr, - .msg_namelen = addrlen, - .msg_flags = XS_SENDMSG_FLAGS | (more ? MSG_MORE : 0), - }; - struct kvec iov = { - .iov_base = vec->iov_base + base, - .iov_len = vec->iov_len - base, - }; + if (seek) + iov_iter_advance(&msg->msg_iter, seek); + return sock_sendmsg(sock, msg); +} - if (iov.iov_len != 0) - return kernel_sendmsg(sock, &msg, &iov, 1, iov.iov_len); - return kernel_sendmsg(sock, &msg, NULL, 0, 0); +static int xs_send_kvec(struct socket *sock, struct msghdr *msg, struct kvec *vec, size_t seek) +{ + if (!vec) { + iov_iter_kvec(&msg->msg_iter, WRITE, NULL, 0, 0); + return sock_sendmsg(sock, msg); + } + iov_iter_kvec(&msg->msg_iter, WRITE, vec, 1, vec->iov_len); + return xs_sendmsg(sock, msg, seek); } static int xs_send_pagedata(struct socket *sock, struct xdr_buf *xdr, unsigned int base, int more, bool zerocopy, int *sent_p) @@ -810,6 +810,11 @@ static int xs_send_pagedata(struct socket *sock, struct xdr_buf *xdr, unsigned i */ static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, struct xdr_buf *xdr, unsigned int base, bool zerocopy, int *sent_p) { + struct msghdr msg = { + .msg_name = addr, + .msg_namelen = addrlen, + .msg_flags = XS_SENDMSG_FLAGS | MSG_MORE, + }; unsigned int remainder = xdr->len - base; int err = 0; int sent = 0; @@ -825,7 +830,9 @@ static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, if (base < xdr->head[0].iov_len || addr != NULL) { unsigned int len = xdr->head[0].iov_len - base; remainder -= len; - err = xs_send_kvec(sock, addr, addrlen, &xdr->head[0], base, remainder != 0); + if (remainder == 0) + msg.msg_flags &= ~MSG_MORE; + err = xs_send_kvec(sock, &msg, &xdr->head[0], base); if (remainder == 0 || err != len) goto out; *sent_p += err; @@ -846,7 +853,8 @@ static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, if (base >= xdr->tail[0].iov_len) return 0; - err = xs_send_kvec(sock, NULL, 0, &xdr->tail[0], base, 0); + msg.msg_flags &= ~MSG_MORE; + err = xs_send_kvec(sock, &msg, &xdr->tail[0], base); out: if (err > 0) { *sent_p += err; From 0472e476604998c127f3c80d291113e77c5676ac Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 19 Feb 2019 13:00:13 -0500 Subject: [PATCH 51/90] SUNRPC: Convert socket page send code to use iov_iter() Simplify the page send code using iov_iter and bvecs. Signed-off-by: Trond Myklebust --- net/sunrpc/xprt.c | 1 + net/sunrpc/xprtsock.c | 49 ++++++++++++------------------------------- 2 files changed, 14 insertions(+), 36 deletions(-) diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 1587728f40d1..2af6be9d6574 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -1723,6 +1723,7 @@ void xprt_release(struct rpc_task *task) xprt->ops->buf_free(task); xprt_inject_disconnect(xprt); xdr_free_bvec(&req->rq_rcv_buf); + xdr_free_bvec(&req->rq_snd_buf); if (req->rq_cred != NULL) put_rpccred(req->rq_cred); task->tk_rqstp = NULL; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index e1546dd6aacc..35d1e81b6e5e 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -759,42 +759,18 @@ static int xs_send_kvec(struct socket *sock, struct msghdr *msg, struct kvec *ve return xs_sendmsg(sock, msg, seek); } -static int xs_send_pagedata(struct socket *sock, struct xdr_buf *xdr, unsigned int base, int more, bool zerocopy, int *sent_p) +static int xs_send_pagedata(struct socket *sock, struct msghdr *msg, struct xdr_buf *xdr, size_t base) { - ssize_t (*do_sendpage)(struct socket *sock, struct page *page, - int offset, size_t size, int flags); - struct page **ppage; - unsigned int remainder; int err; - remainder = xdr->page_len - base; - base += xdr->page_base; - ppage = xdr->pages + (base >> PAGE_SHIFT); - base &= ~PAGE_MASK; - do_sendpage = sock->ops->sendpage; - if (!zerocopy) - do_sendpage = sock_no_sendpage; - for(;;) { - unsigned int len = min_t(unsigned int, PAGE_SIZE - base, remainder); - int flags = XS_SENDMSG_FLAGS; + err = xdr_alloc_bvec(xdr, GFP_KERNEL); + if (err < 0) + return err; - remainder -= len; - if (more) - flags |= MSG_MORE; - if (remainder != 0) - flags |= MSG_SENDPAGE_NOTLAST | MSG_MORE; - err = do_sendpage(sock, *ppage, base, len, flags); - if (remainder == 0 || err != len) - break; - *sent_p += err; - ppage++; - base = 0; - } - if (err > 0) { - *sent_p += err; - err = 0; - } - return err; + iov_iter_bvec(&msg->msg_iter, WRITE, xdr->bvec, + xdr_buf_pagecount(xdr), + xdr->page_len + xdr->page_base); + return xs_sendmsg(sock, msg, base + xdr->page_base); } /** @@ -817,7 +793,6 @@ static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, }; unsigned int remainder = xdr->len - base; int err = 0; - int sent = 0; if (unlikely(!sock)) return -ENOTSOCK; @@ -843,10 +818,12 @@ static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, if (base < xdr->page_len) { unsigned int len = xdr->page_len - base; remainder -= len; - err = xs_send_pagedata(sock, xdr, base, remainder != 0, zerocopy, &sent); - *sent_p += sent; - if (remainder == 0 || sent != len) + if (remainder == 0) + msg.msg_flags &= ~MSG_MORE; + err = xs_send_pagedata(sock, &msg, xdr, base); + if (remainder == 0 || err != len) goto out; + *sent_p += err; base = 0; } else base -= xdr->page_len; From c87dc4c73bb04cb3f86c7d60f9f576eb08514ad8 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 19 Feb 2019 13:13:40 -0500 Subject: [PATCH 52/90] SUNRPC: Further cleanups of xs_sendpages() Now that we send the pages using a struct msghdr, instead of using sendpage(), we no longer need to 'prime the socket' with an address for unconnected UDP messages. Signed-off-by: Trond Myklebust --- net/sunrpc/xprtsock.c | 11 +---------- 1 file changed, 1 insertion(+), 10 deletions(-) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 35d1e81b6e5e..9c6eb76b3f46 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -751,10 +751,6 @@ static int xs_sendmsg(struct socket *sock, struct msghdr *msg, size_t seek) static int xs_send_kvec(struct socket *sock, struct msghdr *msg, struct kvec *vec, size_t seek) { - if (!vec) { - iov_iter_kvec(&msg->msg_iter, WRITE, NULL, 0, 0); - return sock_sendmsg(sock, msg); - } iov_iter_kvec(&msg->msg_iter, WRITE, vec, 1, vec->iov_len); return xs_sendmsg(sock, msg, seek); } @@ -797,12 +793,7 @@ static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, if (unlikely(!sock)) return -ENOTSOCK; - if (base != 0) { - addr = NULL; - addrlen = 0; - } - - if (base < xdr->head[0].iov_len || addr != NULL) { + if (base < xdr->head[0].iov_len) { unsigned int len = xdr->head[0].iov_len - base; remainder -= len; if (remainder == 0) From 6f903b111e33c4f0c349402efbd5319f52362f8e Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 19 Feb 2019 13:21:38 -0500 Subject: [PATCH 53/90] SUNRPC: Remove the redundant 'zerocopy' argument to xs_sendpages() Signed-off-by: Trond Myklebust --- net/sunrpc/xprtsock.c | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 9c6eb76b3f46..8fa74c8a4fd8 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -776,11 +776,10 @@ static int xs_send_pagedata(struct socket *sock, struct msghdr *msg, struct xdr_ * @addrlen: UDP only -- length of destination address * @xdr: buffer containing this request * @base: starting position in the buffer - * @zerocopy: true if it is safe to use sendpage() * @sent_p: return the total number of bytes successfully queued for sending * */ -static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, struct xdr_buf *xdr, unsigned int base, bool zerocopy, int *sent_p) +static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, struct xdr_buf *xdr, unsigned int base, int *sent_p) { struct msghdr msg = { .msg_name = addr, @@ -935,7 +934,7 @@ static int xs_local_send_request(struct rpc_rqst *req) req->rq_xtime = ktime_get(); status = xs_sendpages(transport->sock, NULL, 0, xdr, transport->xmit.offset, - true, &sent); + &sent); dprintk("RPC: %s(%u) = %d\n", __func__, xdr->len - transport->xmit.offset, status); @@ -1002,7 +1001,7 @@ static int xs_udp_send_request(struct rpc_rqst *req) req->rq_xtime = ktime_get(); status = xs_sendpages(transport->sock, xs_addr(xprt), xprt->addrlen, - xdr, 0, true, &sent); + xdr, 0, &sent); dprintk("RPC: xs_udp_send_request(%u) = %d\n", xdr->len, status); @@ -1066,7 +1065,6 @@ static int xs_tcp_send_request(struct rpc_rqst *req) struct rpc_xprt *xprt = req->rq_xprt; struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt); struct xdr_buf *xdr = &req->rq_snd_buf; - bool zerocopy = true; bool vm_wait = false; int status; int sent; @@ -1083,12 +1081,6 @@ static int xs_tcp_send_request(struct rpc_rqst *req) xs_pktdump("packet data:", req->rq_svec->iov_base, req->rq_svec->iov_len); - /* Don't use zero copy if this is a resend. If the RPC call - * completes while the socket holds a reference to the pages, - * then we may end up resending corrupted data. - */ - if (req->rq_task->tk_flags & RPC_TASK_SENT) - zerocopy = false; if (test_bit(XPRT_SOCK_UPD_TIMEOUT, &transport->sock_state)) xs_tcp_set_socket_timeouts(xprt, transport->sock); @@ -1101,7 +1093,7 @@ static int xs_tcp_send_request(struct rpc_rqst *req) sent = 0; status = xs_sendpages(transport->sock, NULL, 0, xdr, transport->xmit.offset, - zerocopy, &sent); + &sent); dprintk("RPC: xs_tcp_send_request(%u) = %d\n", xdr->len - transport->xmit.offset, status); From 3453d5708b33efe76f40eca1c0ed60923094b971 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 20 Jun 2018 17:53:34 -0400 Subject: [PATCH 54/90] NFSv4.1: Avoid false retries when RPC calls are interrupted A 'false retry' in NFSv4.1 occurs when the client attempts to transmit a new RPC call using a slot+sequence number combination that references an already cached one. Currently, the Linux NFS client will do this if a user process interrupts an RPC call that is in progress. The problem with doing so is that we defeat the main mechanism used by the server to differentiate between a new call and a replayed one. Even if the server is able to perfectly cache the arguments of the old call, it cannot know if the client intended to replay or send a new call. The obvious fix is to bump the sequence number pre-emptively if an RPC call is interrupted, but in order to deal with the corner cases where the interrupted call is not actually received and processed by the server, we need to interpret the error NFS4ERR_SEQ_MISORDERED as a sign that we need to either wait or locate a correct sequence number that lies between the value we sent, and the last value that was acked by a SEQUENCE call on that slot. Signed-off-by: Trond Myklebust Tested-by: Jason Tibbitts --- fs/nfs/nfs4proc.c | 105 ++++++++++++++++++++----------------------- fs/nfs/nfs4session.c | 5 ++- fs/nfs/nfs4session.h | 5 ++- 3 files changed, 55 insertions(+), 60 deletions(-) diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 1dbdc85e6885..77c6e2d3f3fc 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -730,13 +730,25 @@ static void nfs41_sequence_free_slot(struct nfs4_sequence_res *res) res->sr_slot = NULL; } +static void nfs4_slot_sequence_record_sent(struct nfs4_slot *slot, + u32 seqnr) +{ + if ((s32)(seqnr - slot->seq_nr_highest_sent) > 0) + slot->seq_nr_highest_sent = seqnr; +} +static void nfs4_slot_sequence_acked(struct nfs4_slot *slot, + u32 seqnr) +{ + slot->seq_nr_highest_sent = seqnr; + slot->seq_nr_last_acked = seqnr; +} + static int nfs41_sequence_process(struct rpc_task *task, struct nfs4_sequence_res *res) { struct nfs4_session *session; struct nfs4_slot *slot = res->sr_slot; struct nfs_client *clp; - bool interrupted = false; int ret = 1; if (slot == NULL) @@ -747,16 +759,12 @@ static int nfs41_sequence_process(struct rpc_task *task, session = slot->table->session; - if (slot->interrupted) { - if (res->sr_status != -NFS4ERR_DELAY) - slot->interrupted = 0; - interrupted = true; - } - trace_nfs4_sequence_done(session, res); /* Check the SEQUENCE operation status */ switch (res->sr_status) { case 0: + /* Mark this sequence number as having been acked */ + nfs4_slot_sequence_acked(slot, slot->seq_nr); /* Update the slot's sequence and clientid lease timer */ slot->seq_done = 1; clp = session->clp; @@ -771,9 +779,9 @@ static int nfs41_sequence_process(struct rpc_task *task, * sr_status remains 1 if an RPC level error occurred. * The server may or may not have processed the sequence * operation.. - * Mark the slot as having hosted an interrupted RPC call. */ - slot->interrupted = 1; + nfs4_slot_sequence_record_sent(slot, slot->seq_nr); + slot->seq_done = 1; goto out; case -NFS4ERR_DELAY: /* The server detected a resend of the RPC call and @@ -784,6 +792,7 @@ static int nfs41_sequence_process(struct rpc_task *task, __func__, slot->slot_nr, slot->seq_nr); + nfs4_slot_sequence_acked(slot, slot->seq_nr); goto out_retry; case -NFS4ERR_RETRY_UNCACHED_REP: case -NFS4ERR_SEQ_FALSE_RETRY: @@ -791,6 +800,7 @@ static int nfs41_sequence_process(struct rpc_task *task, * The server thinks we tried to replay a request. * Retry the call after bumping the sequence ID. */ + nfs4_slot_sequence_acked(slot, slot->seq_nr); goto retry_new_seq; case -NFS4ERR_BADSLOT: /* @@ -801,21 +811,28 @@ static int nfs41_sequence_process(struct rpc_task *task, goto session_recover; goto retry_nowait; case -NFS4ERR_SEQ_MISORDERED: + nfs4_slot_sequence_record_sent(slot, slot->seq_nr); /* - * Was the last operation on this sequence interrupted? - * If so, retry after bumping the sequence number. + * Were one or more calls using this slot interrupted? + * If the server never received the request, then our + * transmitted slot sequence number may be too high. */ - if (interrupted) - goto retry_new_seq; - /* - * Could this slot have been previously retired? - * If so, then the server may be expecting seq_nr = 1! - */ - if (slot->seq_nr != 1) { - slot->seq_nr = 1; + if ((s32)(slot->seq_nr - slot->seq_nr_last_acked) > 1) { + slot->seq_nr--; goto retry_nowait; } - goto session_recover; + /* + * RFC5661: + * A retry might be sent while the original request is + * still in progress on the replier. The replier SHOULD + * deal with the issue by returning NFS4ERR_DELAY as the + * reply to SEQUENCE or CB_SEQUENCE operation, but + * implementations MAY return NFS4ERR_SEQ_MISORDERED. + * + * Restart the search after a delay. + */ + slot->seq_nr = slot->seq_nr_highest_sent; + goto out_retry; default: /* Just update the slot sequence no. */ slot->seq_done = 1; @@ -906,17 +923,6 @@ static const struct rpc_call_ops nfs41_call_sync_ops = { .rpc_call_done = nfs41_call_sync_done, }; -static void -nfs4_sequence_process_interrupted(struct nfs_client *client, - struct nfs4_slot *slot, const struct cred *cred) -{ - struct rpc_task *task; - - task = _nfs41_proc_sequence(client, cred, slot, true); - if (!IS_ERR(task)) - rpc_put_task_async(task); -} - #else /* !CONFIG_NFS_V4_1 */ static int nfs4_sequence_process(struct rpc_task *task, struct nfs4_sequence_res *res) @@ -937,14 +943,6 @@ int nfs4_sequence_done(struct rpc_task *task, } EXPORT_SYMBOL_GPL(nfs4_sequence_done); -static void -nfs4_sequence_process_interrupted(struct nfs_client *client, - struct nfs4_slot *slot, const struct cred *cred) -{ - WARN_ON_ONCE(1); - slot->interrupted = 0; -} - #endif /* !CONFIG_NFS_V4_1 */ static @@ -982,26 +980,19 @@ int nfs4_setup_sequence(struct nfs_client *client, task->tk_timeout = 0; } - for (;;) { - spin_lock(&tbl->slot_tbl_lock); - /* The state manager will wait until the slot table is empty */ - if (nfs4_slot_tbl_draining(tbl) && !args->sa_privileged) - goto out_sleep; + spin_lock(&tbl->slot_tbl_lock); + /* The state manager will wait until the slot table is empty */ + if (nfs4_slot_tbl_draining(tbl) && !args->sa_privileged) + goto out_sleep; - slot = nfs4_alloc_slot(tbl); - if (IS_ERR(slot)) { - /* Try again in 1/4 second */ - if (slot == ERR_PTR(-ENOMEM)) - task->tk_timeout = HZ >> 2; - goto out_sleep; - } - spin_unlock(&tbl->slot_tbl_lock); - - if (likely(!slot->interrupted)) - break; - nfs4_sequence_process_interrupted(client, - slot, task->tk_msg.rpc_cred); + slot = nfs4_alloc_slot(tbl); + if (IS_ERR(slot)) { + /* Try again in 1/4 second */ + if (slot == ERR_PTR(-ENOMEM)) + task->tk_timeout = HZ >> 2; + goto out_sleep; } + spin_unlock(&tbl->slot_tbl_lock); nfs4_sequence_attach_slot(args, res, slot); diff --git a/fs/nfs/nfs4session.c b/fs/nfs/nfs4session.c index 25c5255a395c..bcb532def9e2 100644 --- a/fs/nfs/nfs4session.c +++ b/fs/nfs/nfs4session.c @@ -110,6 +110,8 @@ static struct nfs4_slot *nfs4_new_slot(struct nfs4_slot_table *tbl, slot->table = tbl; slot->slot_nr = slotid; slot->seq_nr = seq_init; + slot->seq_nr_highest_sent = seq_init; + slot->seq_nr_last_acked = seq_init - 1; } return slot; } @@ -276,7 +278,8 @@ static void nfs4_reset_slot_table(struct nfs4_slot_table *tbl, p = &tbl->slots; while (*p) { (*p)->seq_nr = ivalue; - (*p)->interrupted = 0; + (*p)->seq_nr_highest_sent = ivalue; + (*p)->seq_nr_last_acked = ivalue - 1; p = &(*p)->next; } tbl->highest_used_slotid = NFS4_NO_SLOT; diff --git a/fs/nfs/nfs4session.h b/fs/nfs/nfs4session.h index 3c550f297561..230509b77121 100644 --- a/fs/nfs/nfs4session.h +++ b/fs/nfs/nfs4session.h @@ -23,8 +23,9 @@ struct nfs4_slot { unsigned long generation; u32 slot_nr; u32 seq_nr; - unsigned int interrupted : 1, - privileged : 1, + u32 seq_nr_last_acked; + u32 seq_nr_highest_sent; + unsigned int privileged : 1, seq_done : 1; }; From 6f9449be53f3ce383caed797708b332ede8d952c Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 21 Feb 2019 14:51:25 -0500 Subject: [PATCH 55/90] NFS: Fix a soft lockup in the delegation recovery code Fix a soft lockup when NFS client delegation recovery is attempted but the inode is in the process of being freed. When the igrab(inode) call fails, and we have to restart the recovery process, we need to ensure that we won't attempt to recover the same delegation again. Fixes: 45870d6909d5a ("NFSv4.1: Test delegation stateids when server...") Signed-off-by: Trond Myklebust --- fs/nfs/delegation.c | 20 ++++++++++++-------- fs/nfs/delegation.h | 1 + 2 files changed, 13 insertions(+), 8 deletions(-) diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index 70e5931f3c60..2f6b447cdd82 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -229,6 +229,8 @@ static struct inode *nfs_delegation_grab_inode(struct nfs_delegation *delegation spin_lock(&delegation->lock); if (delegation->inode != NULL) inode = igrab(delegation->inode); + if (!inode) + set_bit(NFS_DELEGATION_INODE_FREEING, &delegation->flags); spin_unlock(&delegation->lock); return inode; } @@ -944,10 +946,11 @@ restart: list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) { list_for_each_entry_rcu(delegation, &server->delegations, super_list) { - if (test_bit(NFS_DELEGATION_RETURNING, - &delegation->flags)) - continue; - if (test_bit(NFS_DELEGATION_NEED_RECLAIM, + if (test_bit(NFS_DELEGATION_INODE_FREEING, + &delegation->flags) || + test_bit(NFS_DELEGATION_RETURNING, + &delegation->flags) || + test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags) == 0) continue; if (!nfs_sb_active(server->super)) @@ -1053,10 +1056,11 @@ restart: list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) { list_for_each_entry_rcu(delegation, &server->delegations, super_list) { - if (test_bit(NFS_DELEGATION_RETURNING, - &delegation->flags)) - continue; - if (test_bit(NFS_DELEGATION_TEST_EXPIRED, + if (test_bit(NFS_DELEGATION_INODE_FREEING, + &delegation->flags) || + test_bit(NFS_DELEGATION_RETURNING, + &delegation->flags) || + test_bit(NFS_DELEGATION_TEST_EXPIRED, &delegation->flags) == 0) continue; if (!nfs_sb_active(server->super)) diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h index dcbf3394ba0e..35b4b02c1ae0 100644 --- a/fs/nfs/delegation.h +++ b/fs/nfs/delegation.h @@ -34,6 +34,7 @@ enum { NFS_DELEGATION_RETURNING, NFS_DELEGATION_REVOKED, NFS_DELEGATION_TEST_EXPIRED, + NFS_DELEGATION_INODE_FREEING, }; int nfs_inode_set_delegation(struct inode *inode, const struct cred *cred, From 5085607d209102b37b169bc94d0aa39566a9842a Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 22 Feb 2019 14:20:27 -0500 Subject: [PATCH 56/90] NFS/pnfs: Bulk destroy of layouts needs to be safe w.r.t. umount If a bulk layout recall or a metadata server reboot coincides with a umount, then holding a reference to an inode is unsafe unless we also hold a reference to the super block. Fixes: fd9a8d7160937 ("NFSv4.1: Fix bulk recall and destroy of layouts") Signed-off-by: Trond Myklebust --- fs/nfs/pnfs.c | 33 +++++++++++++++++++++++---------- fs/nfs/pnfs.h | 1 + 2 files changed, 24 insertions(+), 10 deletions(-) diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index 53726da5c010..8247bd1634cb 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -758,22 +758,35 @@ static int pnfs_layout_bulk_destroy_byserver_locked(struct nfs_client *clp, struct nfs_server *server, struct list_head *layout_list) + __must_hold(&clp->cl_lock) + __must_hold(RCU) { struct pnfs_layout_hdr *lo, *next; struct inode *inode; list_for_each_entry_safe(lo, next, &server->layouts, plh_layouts) { - if (test_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags)) + if (test_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags) || + test_bit(NFS_LAYOUT_INODE_FREEING, &lo->plh_flags) || + !list_empty(&lo->plh_bulk_destroy)) continue; + /* If the sb is being destroyed, just bail */ + if (!nfs_sb_active(server->super)) + break; inode = igrab(lo->plh_inode); - if (inode == NULL) - continue; - list_del_init(&lo->plh_layouts); - if (pnfs_layout_add_bulk_destroy_list(inode, layout_list)) - continue; - rcu_read_unlock(); - spin_unlock(&clp->cl_lock); - iput(inode); + if (inode != NULL) { + list_del_init(&lo->plh_layouts); + if (pnfs_layout_add_bulk_destroy_list(inode, + layout_list)) + continue; + rcu_read_unlock(); + spin_unlock(&clp->cl_lock); + iput(inode); + } else { + rcu_read_unlock(); + spin_unlock(&clp->cl_lock); + set_bit(NFS_LAYOUT_INODE_FREEING, &lo->plh_flags); + } + nfs_sb_deactive(server->super); spin_lock(&clp->cl_lock); rcu_read_lock(); return -EAGAIN; @@ -811,7 +824,7 @@ pnfs_layout_free_bulk_destroy_list(struct list_head *layout_list, /* Free all lsegs that are attached to commit buckets */ nfs_commit_inode(inode, 0); pnfs_put_layout_hdr(lo); - iput(inode); + nfs_iput_and_deactive(inode); } return ret; } diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h index 5e80a07b7bea..56659ccce1d8 100644 --- a/fs/nfs/pnfs.h +++ b/fs/nfs/pnfs.h @@ -104,6 +104,7 @@ enum { NFS_LAYOUT_RETURN_REQUESTED, /* Return this layout ASAP */ NFS_LAYOUT_INVALID_STID, /* layout stateid id is invalid */ NFS_LAYOUT_FIRST_LAYOUTGET, /* Serialize first layoutget */ + NFS_LAYOUT_INODE_FREEING, /* The inode is being freed */ }; enum layoutdriver_policy_flags { From a73881c96d73ee72b7dbbd38a6eeef66182a8ef7 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 26 Feb 2019 06:33:02 -0500 Subject: [PATCH 57/90] SUNRPC: Fix an Oops in udp_poll() udp_poll() checks the struct file for the O_NONBLOCK flag, so we must not call it with a NULL file pointer. Fixes: 0ffe86f48026 ("SUNRPC: Use poll() to fix up the socket requeue races") Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprtsock.h | 1 + net/sunrpc/xprtsock.c | 21 +++++++++++++++++++-- 2 files changed, 20 insertions(+), 2 deletions(-) diff --git a/include/linux/sunrpc/xprtsock.h b/include/linux/sunrpc/xprtsock.h index 458bfe0137f5..b81d0b3e0799 100644 --- a/include/linux/sunrpc/xprtsock.h +++ b/include/linux/sunrpc/xprtsock.h @@ -26,6 +26,7 @@ struct sock_xprt { */ struct socket * sock; struct sock * inet; + struct file * file; /* * State of TCP reply receive diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 53de72d2dded..e829036ed81f 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -670,7 +670,8 @@ out_err: static __poll_t xs_poll_socket(struct sock_xprt *transport) { - return transport->sock->ops->poll(NULL, transport->sock, NULL); + return transport->sock->ops->poll(transport->file, transport->sock, + NULL); } static bool xs_poll_socket_readable(struct sock_xprt *transport) @@ -1253,6 +1254,7 @@ static void xs_reset_transport(struct sock_xprt *transport) struct socket *sock = transport->sock; struct sock *sk = transport->inet; struct rpc_xprt *xprt = &transport->xprt; + struct file *filp = transport->file; if (sk == NULL) return; @@ -1266,6 +1268,7 @@ static void xs_reset_transport(struct sock_xprt *transport) write_lock_bh(&sk->sk_callback_lock); transport->inet = NULL; transport->sock = NULL; + transport->file = NULL; sk->sk_user_data = NULL; @@ -1278,7 +1281,7 @@ static void xs_reset_transport(struct sock_xprt *transport) mutex_unlock(&transport->recv_mutex); trace_rpc_socket_close(xprt, sock); - sock_release(sock); + fput(filp); xprt_disconnect_done(xprt); } @@ -1873,6 +1876,7 @@ static struct socket *xs_create_sock(struct rpc_xprt *xprt, struct sock_xprt *transport, int family, int type, int protocol, bool reuseport) { + struct file *filp; struct socket *sock; int err; @@ -1893,6 +1897,11 @@ static struct socket *xs_create_sock(struct rpc_xprt *xprt, goto out; } + filp = sock_alloc_file(sock, O_NONBLOCK, NULL); + if (IS_ERR(filp)) + return ERR_CAST(filp); + transport->file = filp; + return sock; out: return ERR_PTR(err); @@ -1938,6 +1947,7 @@ static int xs_local_finish_connecting(struct rpc_xprt *xprt, static int xs_local_setup_socket(struct sock_xprt *transport) { struct rpc_xprt *xprt = &transport->xprt; + struct file *filp; struct socket *sock; int status = -EIO; @@ -1950,6 +1960,13 @@ static int xs_local_setup_socket(struct sock_xprt *transport) } xs_reclassify_socket(AF_LOCAL, sock); + filp = sock_alloc_file(sock, O_NONBLOCK, NULL); + if (IS_ERR(filp)) { + status = PTR_ERR(filp); + goto out; + } + transport->file = filp; + dprintk("RPC: worker connecting xprt %p via AF_LOCAL to %s\n", xprt, xprt->address_strings[RPC_DISPLAY_ADDR]); From c1dffe0bf7f9c3d57d9f237a7cb2a81e62babd2b Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 1 Mar 2019 12:13:34 -0500 Subject: [PATCH 58/90] NFSv4.1: Reinitialise sequence results before retransmitting a request If we have to retransmit a request, we should ensure that we reinitialise the sequence results structure, since in the event of a signal we need to treat the request as if it had not been sent. Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org --- fs/nfs/nfs4proc.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 77c6e2d3f3fc..5f8517d1a0a5 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -945,6 +945,13 @@ EXPORT_SYMBOL_GPL(nfs4_sequence_done); #endif /* !CONFIG_NFS_V4_1 */ +static void nfs41_sequence_res_init(struct nfs4_sequence_res *res) +{ + res->sr_timestamp = jiffies; + res->sr_status_flags = 0; + res->sr_status = 1; +} + static void nfs4_sequence_attach_slot(struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, @@ -956,10 +963,6 @@ void nfs4_sequence_attach_slot(struct nfs4_sequence_args *args, args->sa_slot = slot; res->sr_slot = slot; - res->sr_timestamp = jiffies; - res->sr_status_flags = 0; - res->sr_status = 1; - } int nfs4_setup_sequence(struct nfs_client *client, @@ -998,6 +1001,7 @@ int nfs4_setup_sequence(struct nfs_client *client, trace_nfs4_setup_sequence(session, args); out_start: + nfs41_sequence_res_init(res); rpc_call_start(task); return 0; From c71c46f0157348a6c1c66f7d382c770ffc212724 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 1 Mar 2019 11:40:05 -0500 Subject: [PATCH 59/90] NFSv4.1: Don't process the sequence op more than once. Ensure that if we call nfs41_sequence_process() a second time for the same rpc_task, then we only process the results once. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 5f8517d1a0a5..5b980246b035 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -754,7 +754,7 @@ static int nfs41_sequence_process(struct rpc_task *task, if (slot == NULL) goto out_noaction; /* don't increment the sequence number if the task wasn't sent */ - if (!RPC_WAS_SENT(task)) + if (!RPC_WAS_SENT(task) || slot->seq_done) goto out; session = slot->table->session; From 1a3466aed3a17eed41cd9411f89eb637f58349b0 Mon Sep 17 00:00:00 2001 From: Anna Schumaker Date: Fri, 1 Mar 2019 16:09:56 -0500 Subject: [PATCH 60/90] NFS: Add missing encode / decode sequence_maxsz to v4.2 operations These really should have been there from the beginning, but we never noticed because there was enough slack in the RPC request for the extra bytes. Chuck's recent patch to use au_cslack and au_rslack to compute buffer size shrunk the buffer enough that this was now a problem for SEEK operations on my test client. Fixes: f4ac1674f5da4 ("nfs: Add ALLOCATE support") Fixes: 2e72448b07dc3 ("NFS: Add COPY nfs operation") Fixes: cb95deea0b4aa ("NFS OFFLOAD_CANCEL xdr") Fixes: 624bd5b7b683c ("nfs: Add DEALLOCATE support") Fixes: 1c6dcbe5ceff8 ("NFS: Implement SEEK") Signed-off-by: Anna Schumaker Signed-off-by: Trond Myklebust --- fs/nfs/nfs42xdr.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c index 22b3425df08f..7d596e8a0941 100644 --- a/fs/nfs/nfs42xdr.c +++ b/fs/nfs/nfs42xdr.c @@ -59,43 +59,53 @@ #define decode_clone_maxsz (op_decode_hdr_maxsz) #define NFS4_enc_allocate_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_allocate_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_allocate_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_allocate_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_copy_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_savefh_maxsz + \ encode_putfh_maxsz + \ encode_copy_maxsz + \ encode_commit_maxsz) #define NFS4_dec_copy_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_savefh_maxsz + \ decode_putfh_maxsz + \ decode_copy_maxsz + \ decode_commit_maxsz) #define NFS4_enc_offload_cancel_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_offload_cancel_maxsz) #define NFS4_dec_offload_cancel_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_offload_cancel_maxsz) #define NFS4_enc_deallocate_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_deallocate_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_deallocate_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_deallocate_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_seek_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_seek_maxsz) #define NFS4_dec_seek_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_seek_maxsz) #define NFS4_enc_layoutstats_sz (compound_encode_hdr_maxsz + \ From 108bb4afd351d65826648a47f11fa3104e250d9b Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 26 Feb 2019 11:19:46 -0500 Subject: [PATCH 61/90] NFSv4/flexfiles: Fix invalid deref in FF_LAYOUT_DEVID_NODE() If the attempt to instantiate the mirror's layout DS pointer failed, then that pointer may hold a value of type ERR_PTR(), so we need to check that before we dereference it. Fixes: 65990d1afbd2d ("pNFS/flexfiles: Fix a deadlock on LAYOUTGET") Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.h | 32 +++++++++++++++----------- 1 file changed, 19 insertions(+), 13 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h index c2626bad466b..8a2d5d630af9 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.h +++ b/fs/nfs/flexfilelayout/flexfilelayout.h @@ -132,16 +132,6 @@ FF_LAYOUT_LSEG(struct pnfs_layout_segment *lseg) generic_hdr); } -static inline struct nfs4_deviceid_node * -FF_LAYOUT_DEVID_NODE(struct pnfs_layout_segment *lseg, u32 idx) -{ - if (idx >= FF_LAYOUT_LSEG(lseg)->mirror_array_cnt || - FF_LAYOUT_LSEG(lseg)->mirror_array[idx] == NULL || - FF_LAYOUT_LSEG(lseg)->mirror_array[idx]->mirror_ds == NULL) - return NULL; - return &FF_LAYOUT_LSEG(lseg)->mirror_array[idx]->mirror_ds->id_node; -} - static inline struct nfs4_ff_layout_ds * FF_LAYOUT_MIRROR_DS(struct nfs4_deviceid_node *node) { @@ -151,9 +141,25 @@ FF_LAYOUT_MIRROR_DS(struct nfs4_deviceid_node *node) static inline struct nfs4_ff_layout_mirror * FF_LAYOUT_COMP(struct pnfs_layout_segment *lseg, u32 idx) { - if (idx >= FF_LAYOUT_LSEG(lseg)->mirror_array_cnt) - return NULL; - return FF_LAYOUT_LSEG(lseg)->mirror_array[idx]; + struct nfs4_ff_layout_segment *fls = FF_LAYOUT_LSEG(lseg); + + if (idx < fls->mirror_array_cnt) + return fls->mirror_array[idx]; + return NULL; +} + +static inline struct nfs4_deviceid_node * +FF_LAYOUT_DEVID_NODE(struct pnfs_layout_segment *lseg, u32 idx) +{ + struct nfs4_ff_layout_mirror *mirror = FF_LAYOUT_COMP(lseg, idx); + + if (mirror != NULL) { + struct nfs4_ff_layout_ds *mirror_ds = mirror->mirror_ds; + + if (!IS_ERR_OR_NULL(mirror_ds)) + return &mirror_ds->id_node; + } + return NULL; } static inline u32 From 762bb7e97351b37c6267509d5806b4c151250ab1 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 26 Feb 2019 11:56:12 -0500 Subject: [PATCH 62/90] NFS/flexfiles: Fix up sparse RCU annotations Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 32701b6a9566..244a03c22b31 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -410,7 +410,7 @@ ff_layout_alloc_lseg(struct pnfs_layout_hdr *lh, for (i = 0; i < fls->mirror_array_cnt; i++) { struct nfs4_ff_layout_mirror *mirror; struct cred *kcred; - const struct cred *cred; + const struct cred __rcu *cred; kuid_t uid; kgid_t gid; u32 ds_count, fh_count, id; @@ -501,7 +501,7 @@ ff_layout_alloc_lseg(struct pnfs_layout_hdr *lh, goto out_err_free; kcred->fsuid = uid; kcred->fsgid = gid; - cred = kcred; + cred = RCU_INITIALIZER(kcred); if (lgr->range.iomode == IOMODE_READ) rcu_assign_pointer(fls->mirror_array[i]->ro_cred, cred); From 39a5201a2b450c2bf0447fb65305654c65f7f954 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 26 Feb 2019 14:27:58 -0500 Subject: [PATCH 63/90] NFSv4/pnfs: Fix barriers in nfs4_mark_deviceid_unavailable() Fix the memory barriers in nfs4_mark_deviceid_unavailable() and nfs4_test_deviceid_unavailable(). Signed-off-by: Trond Myklebust --- fs/nfs/pnfs_dev.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/nfs/pnfs_dev.c b/fs/nfs/pnfs_dev.c index 7fb59487ee90..ec6b607ff1d7 100644 --- a/fs/nfs/pnfs_dev.c +++ b/fs/nfs/pnfs_dev.c @@ -287,7 +287,9 @@ void nfs4_mark_deviceid_unavailable(struct nfs4_deviceid_node *node) { node->timestamp_unavailable = jiffies; + smp_mb__before_atomic(); set_bit(NFS_DEVICEID_UNAVAILABLE, &node->flags); + smp_mb__after_atomic(); } EXPORT_SYMBOL_GPL(nfs4_mark_deviceid_unavailable); @@ -302,6 +304,7 @@ nfs4_test_deviceid_unavailable(struct nfs4_deviceid_node *node) if (time_in_range(node->timestamp_unavailable, start, end)) return true; clear_bit(NFS_DEVICEID_UNAVAILABLE, &node->flags); + smp_mb__after_atomic(); } return false; } From a79f194aa4879e9baad118c3f8bb2ca24dbef765 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 27 Feb 2019 15:37:36 -0500 Subject: [PATCH 64/90] NFSv4/flexfiles: Abort I/O early if the layout segment was invalidated If a layout segment gets invalidated while a pNFS I/O operation is queued for transmission, then we ideally want to abort immediately. This is particularly the case when there is a large number of I/O related RPCs queued in the RPC layer, and the layout segment gets invalidated due to an ENOSPC error, or an EACCES (because the client was fenced). We may end up forced to spam the MDS with a lot of otherwise unnecessary LAYOUTERRORs after that I/O fails. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 17 +++++++++++++++++ include/linux/sunrpc/sched.h | 1 + net/sunrpc/xprt.c | 7 +++++++ 3 files changed, 25 insertions(+) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 244a03c22b31..a8e9bdd978e7 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -1071,6 +1071,8 @@ static int ff_layout_async_handle_error_v4(struct rpc_task *task, break; case -NFS4ERR_RETRY_UNCACHED_REP: break; + case -EAGAIN: + return -NFS4ERR_RESET_TO_PNFS; /* Invalidate Layout errors */ case -NFS4ERR_PNFS_NO_LAYOUT: case -ESTALE: /* mapped NFS4ERR_STALE */ @@ -1131,6 +1133,7 @@ static int ff_layout_async_handle_error_v3(struct rpc_task *task, case -EBADHANDLE: case -ELOOP: case -ENOSPC: + case -EAGAIN: break; case -EJUKEBOX: nfs_inc_stats(lseg->pls_layout->plh_inode, NFSIOS_DELAY); @@ -1369,6 +1372,16 @@ static void ff_layout_read_prepare_v4(struct rpc_task *task, void *data) ff_layout_read_prepare_common(task, hdr); } +static void +ff_layout_io_prepare_transmit(struct rpc_task *task, + void *data) +{ + struct nfs_pgio_header *hdr = data; + + if (!pnfs_is_valid_lseg(hdr->lseg)) + rpc_exit(task, -EAGAIN); +} + static void ff_layout_read_call_done(struct rpc_task *task, void *data) { struct nfs_pgio_header *hdr = data; @@ -1657,6 +1670,7 @@ static void ff_layout_commit_release(void *data) static const struct rpc_call_ops ff_layout_read_call_ops_v3 = { .rpc_call_prepare = ff_layout_read_prepare_v3, + .rpc_call_prepare_transmit = ff_layout_io_prepare_transmit, .rpc_call_done = ff_layout_read_call_done, .rpc_count_stats = ff_layout_read_count_stats, .rpc_release = ff_layout_read_release, @@ -1664,6 +1678,7 @@ static const struct rpc_call_ops ff_layout_read_call_ops_v3 = { static const struct rpc_call_ops ff_layout_read_call_ops_v4 = { .rpc_call_prepare = ff_layout_read_prepare_v4, + .rpc_call_prepare_transmit = ff_layout_io_prepare_transmit, .rpc_call_done = ff_layout_read_call_done, .rpc_count_stats = ff_layout_read_count_stats, .rpc_release = ff_layout_read_release, @@ -1671,6 +1686,7 @@ static const struct rpc_call_ops ff_layout_read_call_ops_v4 = { static const struct rpc_call_ops ff_layout_write_call_ops_v3 = { .rpc_call_prepare = ff_layout_write_prepare_v3, + .rpc_call_prepare_transmit = ff_layout_io_prepare_transmit, .rpc_call_done = ff_layout_write_call_done, .rpc_count_stats = ff_layout_write_count_stats, .rpc_release = ff_layout_write_release, @@ -1678,6 +1694,7 @@ static const struct rpc_call_ops ff_layout_write_call_ops_v3 = { static const struct rpc_call_ops ff_layout_write_call_ops_v4 = { .rpc_call_prepare = ff_layout_write_prepare_v4, + .rpc_call_prepare_transmit = ff_layout_io_prepare_transmit, .rpc_call_done = ff_layout_write_call_done, .rpc_count_stats = ff_layout_write_count_stats, .rpc_release = ff_layout_write_release, diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h index 219aa3910a0c..52d41d0c1ae1 100644 --- a/include/linux/sunrpc/sched.h +++ b/include/linux/sunrpc/sched.h @@ -97,6 +97,7 @@ typedef void (*rpc_action)(struct rpc_task *); struct rpc_call_ops { void (*rpc_call_prepare)(struct rpc_task *, void *); + void (*rpc_call_prepare_transmit)(struct rpc_task *, void *); void (*rpc_call_done)(struct rpc_task *, void *); void (*rpc_count_stats)(struct rpc_task *, void *); void (*rpc_release)(void *); diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 1cf4e379be7b..e096c5a725df 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -1330,6 +1330,13 @@ xprt_request_transmit(struct rpc_rqst *req, struct rpc_task *snd_task) status = -EBADMSG; goto out_dequeue; } + if (task->tk_ops->rpc_call_prepare_transmit) { + task->tk_ops->rpc_call_prepare_transmit(task, + task->tk_calldata); + status = task->tk_status; + if (status < 0) + goto out_dequeue; + } } /* From 3eb86093ea400c58f444eac0debcf6c50d617418 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 8 Feb 2019 10:31:05 -0500 Subject: [PATCH 65/90] NFSv4.2: Add client support for the generic 'layouterror' RPC call Signed-off-by: Trond Myklebust --- fs/nfs/nfs42.h | 3 + fs/nfs/nfs42proc.c | 164 ++++++++++++++++++++++++++++++++++++++ fs/nfs/nfs42xdr.c | 99 +++++++++++++++++++++++ fs/nfs/nfs4proc.c | 3 +- fs/nfs/nfs4xdr.c | 1 + include/linux/nfs4.h | 1 + include/linux/nfs_fs_sb.h | 1 + include/linux/nfs_xdr.h | 35 ++++++++ 8 files changed, 306 insertions(+), 1 deletion(-) diff --git a/fs/nfs/nfs42.h b/fs/nfs/nfs42.h index 19ec38f85ce0..901cca7542f9 100644 --- a/fs/nfs/nfs42.h +++ b/fs/nfs/nfs42.h @@ -20,5 +20,8 @@ loff_t nfs42_proc_llseek(struct file *, loff_t, int); int nfs42_proc_layoutstats_generic(struct nfs_server *, struct nfs42_layoutstat_data *); int nfs42_proc_clone(struct file *, struct file *, loff_t, loff_t, loff_t); +int nfs42_proc_layouterror(struct pnfs_layout_segment *lseg, + const struct nfs42_layout_error *errors, + size_t n); #endif /* __LINUX_FS_NFS_NFS4_2_H */ diff --git a/fs/nfs/nfs42proc.c b/fs/nfs/nfs42proc.c index fed06fd9998d..ff6f85fb676b 100644 --- a/fs/nfs/nfs42proc.c +++ b/fs/nfs/nfs42proc.c @@ -672,6 +672,170 @@ int nfs42_proc_layoutstats_generic(struct nfs_server *server, return 0; } +static struct nfs42_layouterror_data * +nfs42_alloc_layouterror_data(struct pnfs_layout_segment *lseg, gfp_t gfp_flags) +{ + struct nfs42_layouterror_data *data; + struct inode *inode = lseg->pls_layout->plh_inode; + + data = kzalloc(sizeof(*data), gfp_flags); + if (data) { + data->args.inode = data->inode = nfs_igrab_and_active(inode); + if (data->inode) { + data->lseg = pnfs_get_lseg(lseg); + if (data->lseg) + return data; + nfs_iput_and_deactive(data->inode); + } + kfree(data); + } + return NULL; +} + +static void +nfs42_free_layouterror_data(struct nfs42_layouterror_data *data) +{ + pnfs_put_lseg(data->lseg); + nfs_iput_and_deactive(data->inode); + kfree(data); +} + +static void +nfs42_layouterror_prepare(struct rpc_task *task, void *calldata) +{ + struct nfs42_layouterror_data *data = calldata; + struct inode *inode = data->inode; + struct nfs_server *server = NFS_SERVER(inode); + struct pnfs_layout_hdr *lo = data->lseg->pls_layout; + unsigned i; + + spin_lock(&inode->i_lock); + if (!pnfs_layout_is_valid(lo)) { + spin_unlock(&inode->i_lock); + rpc_exit(task, 0); + return; + } + for (i = 0; i < data->args.num_errors; i++) + nfs4_stateid_copy(&data->args.errors[i].stateid, + &lo->plh_stateid); + spin_unlock(&inode->i_lock); + nfs4_setup_sequence(server->nfs_client, &data->args.seq_args, + &data->res.seq_res, task); +} + +static void +nfs42_layouterror_done(struct rpc_task *task, void *calldata) +{ + struct nfs42_layouterror_data *data = calldata; + struct inode *inode = data->inode; + struct pnfs_layout_hdr *lo = data->lseg->pls_layout; + + if (!nfs4_sequence_done(task, &data->res.seq_res)) + return; + + switch (task->tk_status) { + case 0: + break; + case -NFS4ERR_BADHANDLE: + case -ESTALE: + pnfs_destroy_layout(NFS_I(inode)); + break; + case -NFS4ERR_EXPIRED: + case -NFS4ERR_ADMIN_REVOKED: + case -NFS4ERR_DELEG_REVOKED: + case -NFS4ERR_STALE_STATEID: + case -NFS4ERR_BAD_STATEID: + spin_lock(&inode->i_lock); + if (pnfs_layout_is_valid(lo) && + nfs4_stateid_match(&data->args.errors[0].stateid, + &lo->plh_stateid)) { + LIST_HEAD(head); + + /* + * Mark the bad layout state as invalid, then retry + * with the current stateid. + */ + pnfs_mark_layout_stateid_invalid(lo, &head); + spin_unlock(&inode->i_lock); + pnfs_free_lseg_list(&head); + nfs_commit_inode(inode, 0); + } else + spin_unlock(&inode->i_lock); + break; + case -NFS4ERR_OLD_STATEID: + spin_lock(&inode->i_lock); + if (pnfs_layout_is_valid(lo) && + nfs4_stateid_match_other(&data->args.errors[0].stateid, + &lo->plh_stateid)) { + /* Do we need to delay before resending? */ + if (!nfs4_stateid_is_newer(&lo->plh_stateid, + &data->args.errors[0].stateid)) + rpc_delay(task, HZ); + rpc_restart_call_prepare(task); + } + spin_unlock(&inode->i_lock); + break; + case -ENOTSUPP: + case -EOPNOTSUPP: + NFS_SERVER(inode)->caps &= ~NFS_CAP_LAYOUTERROR; + } +} + +static void +nfs42_layouterror_release(void *calldata) +{ + struct nfs42_layouterror_data *data = calldata; + + nfs42_free_layouterror_data(data); +} + +static const struct rpc_call_ops nfs42_layouterror_ops = { + .rpc_call_prepare = nfs42_layouterror_prepare, + .rpc_call_done = nfs42_layouterror_done, + .rpc_release = nfs42_layouterror_release, +}; + +int nfs42_proc_layouterror(struct pnfs_layout_segment *lseg, + const struct nfs42_layout_error *errors, size_t n) +{ + struct inode *inode = lseg->pls_layout->plh_inode; + struct nfs42_layouterror_data *data; + struct rpc_task *task; + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LAYOUTERROR], + }; + struct rpc_task_setup task_setup = { + .rpc_message = &msg, + .callback_ops = &nfs42_layouterror_ops, + .flags = RPC_TASK_ASYNC, + }; + unsigned int i; + + if (!nfs_server_capable(inode, NFS_CAP_LAYOUTERROR)) + return -EOPNOTSUPP; + if (n > NFS42_LAYOUTERROR_MAX) + return -EINVAL; + data = nfs42_alloc_layouterror_data(lseg, GFP_NOFS); + if (!data) + return -ENOMEM; + for (i = 0; i < n; i++) { + data->args.errors[i] = errors[i]; + data->args.num_errors++; + data->res.num_errors++; + } + msg.rpc_argp = &data->args; + msg.rpc_resp = &data->res; + task_setup.callback_data = data; + task_setup.rpc_client = NFS_SERVER(inode)->client; + nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 0, 0); + task = rpc_run_task(&task_setup); + if (IS_ERR(task)) + return PTR_ERR(task); + rpc_put_task(task); + return 0; +} +EXPORT_SYMBOL_GPL(nfs42_proc_layouterror); + static int _nfs42_proc_clone(struct rpc_message *msg, struct file *src_f, struct file *dst_f, struct nfs_lock_context *src_lock, struct nfs_lock_context *dst_lock, loff_t src_offset, diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c index 7d596e8a0941..aed865a84629 100644 --- a/fs/nfs/nfs42xdr.c +++ b/fs/nfs/nfs42xdr.c @@ -51,6 +51,15 @@ 1 /* opaque devaddr4 length */ + \ XDR_QUADLEN(PNFS_LAYOUTSTATS_MAXSIZE)) #define decode_layoutstats_maxsz (op_decode_hdr_maxsz) +#define encode_device_error_maxsz (XDR_QUADLEN(NFS4_DEVICEID4_SIZE) + \ + 1 /* status */ + 1 /* opnum */) +#define encode_layouterror_maxsz (op_decode_hdr_maxsz + \ + 2 /* offset */ + \ + 2 /* length */ + \ + encode_stateid_maxsz + \ + 1 /* Array size */ + \ + encode_device_error_maxsz) +#define decode_layouterror_maxsz (op_decode_hdr_maxsz) #define encode_clone_maxsz (encode_stateid_maxsz + \ encode_stateid_maxsz + \ 2 /* src offset */ + \ @@ -116,6 +125,16 @@ decode_sequence_maxsz + \ decode_putfh_maxsz + \ PNFS_LAYOUTSTATS_MAXDEV * decode_layoutstats_maxsz) +#define NFS4_enc_layouterror_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ + encode_putfh_maxsz + \ + NFS42_LAYOUTERROR_MAX * \ + encode_layouterror_maxsz) +#define NFS4_dec_layouterror_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ + decode_putfh_maxsz + \ + NFS42_LAYOUTERROR_MAX * \ + decode_layouterror_maxsz) #define NFS4_enc_clone_sz (compound_encode_hdr_maxsz + \ encode_sequence_maxsz + \ encode_putfh_maxsz + \ @@ -233,6 +252,34 @@ static void encode_clone(struct xdr_stream *xdr, xdr_encode_hyper(p, args->count); } +static void encode_device_error(struct xdr_stream *xdr, + const struct nfs42_device_error *error) +{ + __be32 *p; + + p = reserve_space(xdr, NFS4_DEVICEID4_SIZE + 2*4); + p = xdr_encode_opaque_fixed(p, error->dev_id.data, + NFS4_DEVICEID4_SIZE); + *p++ = cpu_to_be32(error->status); + *p = cpu_to_be32(error->opnum); +} + +static void encode_layouterror(struct xdr_stream *xdr, + const struct nfs42_layout_error *args, + struct compound_hdr *hdr) +{ + __be32 *p; + + encode_op_hdr(xdr, OP_LAYOUTERROR, decode_layouterror_maxsz, hdr); + p = reserve_space(xdr, 8 + 8); + p = xdr_encode_hyper(p, args->offset); + p = xdr_encode_hyper(p, args->length); + encode_nfs4_stateid(xdr, &args->stateid); + p = reserve_space(xdr, 4); + *p = cpu_to_be32(1); + encode_device_error(xdr, &args->errors[0]); +} + /* * Encode ALLOCATE request */ @@ -391,6 +438,27 @@ static void nfs4_xdr_enc_clone(struct rpc_rqst *req, encode_nops(&hdr); } +/* + * Encode LAYOUTERROR request + */ +static void nfs4_xdr_enc_layouterror(struct rpc_rqst *req, + struct xdr_stream *xdr, + const void *data) +{ + const struct nfs42_layouterror_args *args = data; + struct compound_hdr hdr = { + .minorversion = nfs4_xdr_minorversion(&args->seq_args), + }; + int i; + + encode_compound_hdr(xdr, req, &hdr); + encode_sequence(xdr, &args->seq_args, &hdr); + encode_putfh(xdr, NFS_FH(args->inode), &hdr); + for (i = 0; i < args->num_errors; i++) + encode_layouterror(xdr, &args->errors[i], &hdr); + encode_nops(&hdr); +} + static int decode_allocate(struct xdr_stream *xdr, struct nfs42_falloc_res *res) { return decode_op_hdr(xdr, OP_ALLOCATE); @@ -494,6 +562,11 @@ static int decode_clone(struct xdr_stream *xdr) return decode_op_hdr(xdr, OP_CLONE); } +static int decode_layouterror(struct xdr_stream *xdr) +{ + return decode_op_hdr(xdr, OP_LAYOUTERROR); +} + /* * Decode ALLOCATE request */ @@ -703,4 +776,30 @@ out: return status; } +/* + * Decode LAYOUTERROR request + */ +static int nfs4_xdr_dec_layouterror(struct rpc_rqst *rqstp, + struct xdr_stream *xdr, + void *data) +{ + struct nfs42_layouterror_res *res = data; + struct compound_hdr hdr; + int status, i; + + status = decode_compound_hdr(xdr, &hdr); + if (status) + goto out; + status = decode_sequence(xdr, &res->seq_res, rqstp); + if (status) + goto out; + status = decode_putfh(xdr); + + for (i = 0; i < res->num_errors && status == 0; i++) + status = decode_layouterror(xdr); +out: + res->rpc_status = status; + return status; +} + #endif /* __LINUX_FS_NFS_NFS4_2XDR_H */ diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 5b980246b035..73889ea7d196 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -9690,7 +9690,8 @@ static const struct nfs4_minor_version_ops nfs_v4_2_minor_ops = { | NFS_CAP_DEALLOCATE | NFS_CAP_SEEK | NFS_CAP_LAYOUTSTATS - | NFS_CAP_CLONE, + | NFS_CAP_CLONE + | NFS_CAP_LAYOUTERROR, .init_client = nfs41_init_client, .shutdown_client = nfs41_shutdown_client, .match_stateid = nfs41_match_stateid, diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 6d9d5e2f6308..cfcabc33e24d 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -7572,6 +7572,7 @@ const struct rpc_procinfo nfs4_procedures[] = { PROC42(COPY, enc_copy, dec_copy), PROC42(OFFLOAD_CANCEL, enc_offload_cancel, dec_offload_cancel), PROC(LOOKUPP, enc_lookupp, dec_lookupp), + PROC42(LAYOUTERROR, enc_layouterror, dec_layouterror), }; static unsigned int nfs_version4_counts[ARRAY_SIZE(nfs4_procedures)]; diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h index 1b06f0b28453..22494d170619 100644 --- a/include/linux/nfs4.h +++ b/include/linux/nfs4.h @@ -538,6 +538,7 @@ enum { NFSPROC4_CLNT_OFFLOAD_CANCEL, NFSPROC4_CLNT_LOOKUPP, + NFSPROC4_CLNT_LAYOUTERROR, }; /* nfs41 types */ diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index 6aa8cc83c3b6..c827d31298cc 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -261,5 +261,6 @@ struct nfs_server { #define NFS_CAP_CLONE (1U << 23) #define NFS_CAP_COPY (1U << 24) #define NFS_CAP_OFFLOAD_CANCEL (1U << 25) +#define NFS_CAP_LAYOUTERROR (1U << 26) #endif diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index b4bd2bf5f585..9b8324ec08f3 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -383,6 +383,41 @@ struct nfs42_layoutstat_data { struct nfs42_layoutstat_res res; }; +struct nfs42_device_error { + struct nfs4_deviceid dev_id; + int status; + enum nfs_opnum4 opnum; +}; + +struct nfs42_layout_error { + __u64 offset; + __u64 length; + nfs4_stateid stateid; + struct nfs42_device_error errors[1]; +}; + +#define NFS42_LAYOUTERROR_MAX 5 + +struct nfs42_layouterror_args { + struct nfs4_sequence_args seq_args; + struct inode *inode; + unsigned int num_errors; + struct nfs42_layout_error errors[NFS42_LAYOUTERROR_MAX]; +}; + +struct nfs42_layouterror_res { + struct nfs4_sequence_res seq_res; + unsigned int num_errors; + int rpc_status; +}; + +struct nfs42_layouterror_data { + struct nfs42_layouterror_args args; + struct nfs42_layouterror_res res; + struct inode *inode; + struct pnfs_layout_segment *lseg; +}; + struct nfs42_clone_args { struct nfs4_sequence_args seq_args; struct nfs_fh *src_fh; From f0922a6c0cdb928d21c2599f9671206237e7e505 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 10 Feb 2019 22:38:43 -0500 Subject: [PATCH 66/90] NFS/flexfiles: Send LAYOUTERROR when failing over mirrored reads When a read to the preferred mirror returns an error, the flexfiles driver records the error in the inode list and currently marks the layout for return before failing over the attempted read to the next mirror. What we actually want to do is fire off a LAYOUTERROR to notify the MDS that there is an issue with the preferred mirror, then we fail over. Only once we've failed to read from all mirrors should we return the layout. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 60 +++++++++++++++++++++-- fs/nfs/flexfilelayout/flexfilelayout.h | 1 + fs/nfs/flexfilelayout/flexfilelayoutdev.c | 2 +- 3 files changed, 57 insertions(+), 6 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index a8e9bdd978e7..eb86da30d4b9 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -1252,7 +1252,7 @@ static int ff_layout_read_done_cb(struct rpc_task *task, if (ff_layout_choose_best_ds_for_read(hdr->lseg, hdr->pgio_mirror_idx + 1, &hdr->pgio_mirror_idx)) - goto out_eagain; + goto out_layouterror; set_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags); return task->tk_status; case -NFS4ERR_RESET_TO_MDS: @@ -1263,6 +1263,8 @@ static int ff_layout_read_done_cb(struct rpc_task *task, } return 0; +out_layouterror: + ff_layout_send_layouterror(hdr->lseg); out_eagain: rpc_restart_call_prepare(task); return -EAGAIN; @@ -1412,9 +1414,10 @@ static void ff_layout_read_release(void *data) struct nfs_pgio_header *hdr = data; ff_layout_read_record_layoutstats_done(&hdr->task, hdr); - if (test_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags)) + if (test_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags)) { + ff_layout_send_layouterror(hdr->lseg); pnfs_read_resend_pnfs(hdr); - else if (test_bit(NFS_IOHDR_RESEND_MDS, &hdr->flags)) + } else if (test_bit(NFS_IOHDR_RESEND_MDS, &hdr->flags)) ff_layout_reset_read(hdr); pnfs_generic_rw_release(data); } @@ -1586,9 +1589,10 @@ static void ff_layout_write_release(void *data) struct nfs_pgio_header *hdr = data; ff_layout_write_record_layoutstats_done(&hdr->task, hdr); - if (test_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags)) + if (test_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags)) { + ff_layout_send_layouterror(hdr->lseg); ff_layout_reset_write(hdr, true); - else if (test_bit(NFS_IOHDR_RESEND_MDS, &hdr->flags)) + } else if (test_bit(NFS_IOHDR_RESEND_MDS, &hdr->flags)) ff_layout_reset_write(hdr, false); pnfs_generic_rw_release(data); } @@ -2119,6 +2123,52 @@ out_nomem: return -ENOMEM; } +#ifdef CONFIG_NFS_V4_2 +void +ff_layout_send_layouterror(struct pnfs_layout_segment *lseg) +{ + struct pnfs_layout_hdr *lo = lseg->pls_layout; + struct nfs42_layout_error *errors; + LIST_HEAD(head); + + if (!nfs_server_capable(lo->plh_inode, NFS_CAP_LAYOUTERROR)) + return; + ff_layout_fetch_ds_ioerr(lo, &lseg->pls_range, &head, -1); + if (list_empty(&head)) + return; + + errors = kmalloc_array(NFS42_LAYOUTERROR_MAX, + sizeof(*errors), GFP_NOFS); + if (errors != NULL) { + const struct nfs4_ff_layout_ds_err *pos; + size_t n = 0; + + list_for_each_entry(pos, &head, list) { + errors[n].offset = pos->offset; + errors[n].length = pos->length; + nfs4_stateid_copy(&errors[n].stateid, &pos->stateid); + errors[n].errors[0].dev_id = pos->deviceid; + errors[n].errors[0].status = pos->status; + errors[n].errors[0].opnum = pos->opnum; + n++; + if (!list_is_last(&pos->list, &head) && + n < NFS42_LAYOUTERROR_MAX) + continue; + if (nfs42_proc_layouterror(lseg, errors, n) < 0) + break; + n = 0; + } + kfree(errors); + } + ff_layout_free_ds_ioerr(&head); +} +#else +void +ff_layout_send_layouterror(struct pnfs_layout_segment *lseg) +{ +} +#endif + static int ff_layout_ntop4(const struct sockaddr *sap, char *buf, const size_t buflen) { diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h index 8a2d5d630af9..31a62820a5c6 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.h +++ b/fs/nfs/flexfilelayout/flexfilelayout.h @@ -213,6 +213,7 @@ int ff_layout_track_ds_error(struct nfs4_flexfile_layout *flo, struct nfs4_ff_layout_mirror *mirror, u64 offset, u64 length, int status, enum nfs_opnum4 opnum, gfp_t gfp_flags); +void ff_layout_send_layouterror(struct pnfs_layout_segment *lseg); int ff_layout_encode_ds_ioerr(struct xdr_stream *xdr, const struct list_head *head); void ff_layout_free_ds_ioerr(struct list_head *head); unsigned int ff_layout_fetch_ds_ioerr(struct pnfs_layout_hdr *lo, diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index ca7a6203b3cb..c174f23afc6d 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -326,7 +326,6 @@ int ff_layout_track_ds_error(struct nfs4_flexfile_layout *flo, spin_lock(&flo->generic_hdr.plh_inode->i_lock); ff_layout_add_ds_error_locked(flo, dserr); spin_unlock(&flo->generic_hdr.plh_inode->i_lock); - return 0; } @@ -458,6 +457,7 @@ out_fail: mirror, lseg->pls_range.offset, lseg->pls_range.length, NFS4ERR_NXIO, OP_ILLEGAL, GFP_NOIO); + ff_layout_send_layouterror(lseg); if (fail_return || !ff_layout_has_available_ds(lseg)) pnfs_error_mark_layout_for_return(ino, lseg); ds = NULL; From 18c0778a650b21f8729576b7bc382447d2027d4d Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 13 Feb 2019 07:55:31 -0500 Subject: [PATCH 67/90] NFSv4: Handle early exit in layoutget by returning an error If the LAYOUTGET rpc call exits early without an error, convert it to EAGAIN. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 73889ea7d196..4dbb0ee23432 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -8936,10 +8936,12 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, long *timeout) if (status != 0) goto out; - /* if layoutp->len is 0, nfs4_layoutget_prepare called rpc_exit */ - if (task->tk_status < 0 || lgp->res.layoutp->len == 0) { + if (task->tk_status < 0) { status = nfs4_layoutget_handle_exception(task, lgp, &exception); *timeout = exception.timeout; + } else if (lgp->res.layoutp->len == 0) { + status = -EAGAIN; + *timeout = nfs4_update_delay(&exception.timeout); } else lseg = pnfs_layout_process(lgp); out: From 2444ff277a686d743b5479dd52d8357c21d5ce13 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 14 Feb 2019 13:45:45 -0500 Subject: [PATCH 68/90] NFS/flexfiles: refactor calls to fs4_ff_layout_prepare_ds() While we may want to skip attempting to connect to a downed mirror when we're deciding which mirror to select for a read, we do not want to do so once we've committed to attempting the I/O in ff_layout_read/write_pagelist(), or ff_layout_initiate_commit() Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 26 +++++++++++++++++------ fs/nfs/flexfilelayout/flexfilelayout.h | 3 ++- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 21 ++++++------------ 3 files changed, 28 insertions(+), 22 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index eb86da30d4b9..0134bcac46a7 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -794,6 +794,7 @@ ff_layout_choose_best_ds_for_read(struct pnfs_layout_segment *lseg, int *best_idx) { struct nfs4_ff_layout_segment *fls = FF_LAYOUT_LSEG(lseg); + struct nfs4_ff_layout_mirror *mirror; struct nfs4_pnfs_ds *ds; bool fail_return = false; int idx; @@ -802,7 +803,12 @@ ff_layout_choose_best_ds_for_read(struct pnfs_layout_segment *lseg, for (idx = start_idx; idx < fls->mirror_array_cnt; idx++) { if (idx+1 == fls->mirror_array_cnt) fail_return = true; - ds = nfs4_ff_layout_prepare_ds(lseg, idx, fail_return); + + mirror = FF_LAYOUT_COMP(lseg, idx); + if (ff_layout_test_devid_unavailable(&mirror->mirror_ds->id_node)) + continue; + + ds = nfs4_ff_layout_prepare_ds(lseg, mirror, fail_return); if (ds) { *best_idx = idx; return ds; @@ -925,7 +931,10 @@ retry: goto out_mds; for (i = 0; i < pgio->pg_mirror_count; i++) { - ds = nfs4_ff_layout_prepare_ds(pgio->pg_lseg, i, true); + mirror = FF_LAYOUT_COMP(pgio->pg_lseg, i); + if (ff_layout_test_devid_unavailable(&mirror->mirror_ds->id_node)) + continue; + ds = nfs4_ff_layout_prepare_ds(pgio->pg_lseg, mirror, true); if (!ds) { if (!ff_layout_no_fallback_to_mds(pgio->pg_lseg)) goto out_mds; @@ -936,7 +945,6 @@ retry: goto retry; } pgm = &pgio->pg_mirrors[i]; - mirror = FF_LAYOUT_COMP(pgio->pg_lseg, i); pgm->pg_bsize = mirror->mirror_ds->ds_versions[0].wsize; } @@ -1724,6 +1732,7 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr) struct pnfs_layout_segment *lseg = hdr->lseg; struct nfs4_pnfs_ds *ds; struct rpc_clnt *ds_clnt; + struct nfs4_ff_layout_mirror *mirror; const struct cred *ds_cred; loff_t offset = hdr->args.offset; u32 idx = hdr->pgio_mirror_idx; @@ -1734,7 +1743,8 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr) __func__, hdr->inode->i_ino, hdr->args.pgbase, (size_t)hdr->args.count, offset); - ds = nfs4_ff_layout_prepare_ds(lseg, idx, false); + mirror = FF_LAYOUT_COMP(lseg, idx); + ds = nfs4_ff_layout_prepare_ds(lseg, mirror, false); if (!ds) goto out_failed; @@ -1791,13 +1801,15 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync) struct pnfs_layout_segment *lseg = hdr->lseg; struct nfs4_pnfs_ds *ds; struct rpc_clnt *ds_clnt; + struct nfs4_ff_layout_mirror *mirror; const struct cred *ds_cred; loff_t offset = hdr->args.offset; int vers; struct nfs_fh *fh; int idx = hdr->pgio_mirror_idx; - ds = nfs4_ff_layout_prepare_ds(lseg, idx, true); + mirror = FF_LAYOUT_COMP(lseg, idx); + ds = nfs4_ff_layout_prepare_ds(lseg, mirror, true); if (!ds) goto out_failed; @@ -1870,6 +1882,7 @@ static int ff_layout_initiate_commit(struct nfs_commit_data *data, int how) struct pnfs_layout_segment *lseg = data->lseg; struct nfs4_pnfs_ds *ds; struct rpc_clnt *ds_clnt; + struct nfs4_ff_layout_mirror *mirror; const struct cred *ds_cred; u32 idx; int vers, ret; @@ -1880,7 +1893,8 @@ static int ff_layout_initiate_commit(struct nfs_commit_data *data, int how) goto out_err; idx = calc_ds_index_from_commit(lseg, data->ds_commit_index); - ds = nfs4_ff_layout_prepare_ds(lseg, idx, true); + mirror = FF_LAYOUT_COMP(lseg, idx); + ds = nfs4_ff_layout_prepare_ds(lseg, mirror, true); if (!ds) goto out_err; diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h index 31a62820a5c6..93508112486a 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.h +++ b/fs/nfs/flexfilelayout/flexfilelayout.h @@ -228,7 +228,8 @@ nfs4_ff_layout_select_ds_stateid(struct pnfs_layout_segment *lseg, nfs4_stateid *stateid); struct nfs4_pnfs_ds * -nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx, +nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, + struct nfs4_ff_layout_mirror *mirror, bool fail_return); struct rpc_clnt * diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index c174f23afc6d..f9a13e1fb1a8 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -391,7 +391,7 @@ out: /** * nfs4_ff_layout_prepare_ds - prepare a DS connection for an RPC call * @lseg: the layout segment we're operating on - * @ds_idx: index of the DS to use + * @mirror: layout mirror describing the DS to use * @fail_return: return layout on connect failure? * * Try to prepare a DS connection to accept an RPC call. This involves @@ -406,26 +406,18 @@ out: * Returns a pointer to a connected DS object on success or NULL on failure. */ struct nfs4_pnfs_ds * -nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx, +nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, + struct nfs4_ff_layout_mirror *mirror, bool fail_return) { - struct nfs4_ff_layout_mirror *mirror = FF_LAYOUT_COMP(lseg, ds_idx); struct nfs4_pnfs_ds *ds = NULL; - struct nfs4_deviceid_node *devid; struct inode *ino = lseg->pls_layout->plh_inode; struct nfs_server *s = NFS_SERVER(ino); unsigned int max_payload; int status; - if (!ff_layout_mirror_valid(lseg, mirror, true)) { - pr_err_ratelimited("NFS: %s: No data server for offset index %d\n", - __func__, ds_idx); + if (!ff_layout_mirror_valid(lseg, mirror, true)) goto out; - } - - devid = &mirror->mirror_ds->id_node; - if (ff_layout_test_devid_unavailable(devid)) - goto out_fail; ds = mirror->mirror_ds->ds; /* matching smp_wmb() in _nfs4_pnfs_v3/4_ds_connect */ @@ -436,8 +428,8 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx, /* FIXME: For now we assume the server sent only one version of NFS * to use for the DS. */ - status = nfs4_pnfs_ds_connect(s, ds, devid, dataserver_timeo, - dataserver_retrans, + status = nfs4_pnfs_ds_connect(s, ds, &mirror->mirror_ds->id_node, + dataserver_timeo, dataserver_retrans, mirror->mirror_ds->ds_versions[0].version, mirror->mirror_ds->ds_versions[0].minor_version); @@ -452,7 +444,6 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx, mirror->mirror_ds->ds_versions[0].wsize = max_payload; goto out; } -out_fail: ff_layout_track_ds_error(FF_LAYOUT_FROM_HDR(lseg->pls_layout), mirror, lseg->pls_range.offset, lseg->pls_range.length, NFS4ERR_NXIO, From 0a156dd58274b0b847118966365203955288f4c6 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 27 Feb 2019 23:46:04 -0500 Subject: [PATCH 69/90] NFS/flexfiles: Avoid unnecessary layout invalidations In ff_layout_mirror_valid() we may not want to invalidate the layout segment despite the call to GETDEVICEINFO failing. The reason is that a read may still be able to make progress on another mirror. So instead we let the caller (in this case nfs4_ff_layout_prepare_ds()) decide whether or not it needs to invalidate. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index f9a13e1fb1a8..92efb5692eb3 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -196,7 +196,7 @@ static bool ff_layout_mirror_valid(struct pnfs_layout_segment *lseg, struct nfs4_ff_layout_mirror *mirror, bool create) { - if (mirror == NULL || IS_ERR(mirror->mirror_ds)) + if (mirror == NULL) goto outerr; if (mirror->mirror_ds == NULL) { if (create) { @@ -229,7 +229,6 @@ static bool ff_layout_mirror_valid(struct pnfs_layout_segment *lseg, } return true; outerr: - pnfs_error_mark_layout_for_return(lseg->pls_layout->plh_inode, lseg); return false; } @@ -417,7 +416,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, int status; if (!ff_layout_mirror_valid(lseg, mirror, true)) - goto out; + goto noconnect; ds = mirror->mirror_ds->ds; /* matching smp_wmb() in _nfs4_pnfs_v3/4_ds_connect */ @@ -444,6 +443,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, mirror->mirror_ds->ds_versions[0].wsize = max_payload; goto out; } +noconnect: ff_layout_track_ds_error(FF_LAYOUT_FROM_HDR(lseg->pls_layout), mirror, lseg->pls_range.offset, lseg->pls_range.length, NFS4ERR_NXIO, From d082d4b5a0ddf56d9a63eae66119f2ef53783d5f Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 26 Feb 2019 12:57:06 -0500 Subject: [PATCH 70/90] NFS/flexfiles: Remove bogus checks for invalid deviceids We already check the deviceids before we start the RPC call. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 20 -------------------- 1 file changed, 20 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 0134bcac46a7..747ab2fafb69 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -932,8 +932,6 @@ retry: for (i = 0; i < pgio->pg_mirror_count; i++) { mirror = FF_LAYOUT_COMP(pgio->pg_lseg, i); - if (ff_layout_test_devid_unavailable(&mirror->mirror_ds->id_node)) - continue; ds = nfs4_ff_layout_prepare_ds(pgio->pg_lseg, mirror, true); if (!ds) { if (!ff_layout_no_fallback_to_mds(pgio->pg_lseg)) @@ -1306,15 +1304,6 @@ ff_layout_set_layoutcommit(struct inode *inode, (unsigned long long) NFS_I(inode)->layout->plh_lwb); } -static bool -ff_layout_device_unavailable(struct pnfs_layout_segment *lseg, int idx) -{ - /* No mirroring for now */ - struct nfs4_deviceid_node *node = FF_LAYOUT_DEVID_NODE(lseg, idx); - - return ff_layout_test_devid_unavailable(node); -} - static void ff_layout_read_record_layoutstats_start(struct rpc_task *task, struct nfs_pgio_header *hdr) { @@ -1345,10 +1334,6 @@ static int ff_layout_read_prepare_common(struct rpc_task *task, rpc_exit(task, -EIO); return -EIO; } - if (ff_layout_device_unavailable(hdr->lseg, hdr->pgio_mirror_idx)) { - rpc_exit(task, -EHOSTDOWN); - return -EAGAIN; - } ff_layout_read_record_layoutstats_start(task, hdr); return 0; @@ -1537,11 +1522,6 @@ static int ff_layout_write_prepare_common(struct rpc_task *task, return -EIO; } - if (ff_layout_device_unavailable(hdr->lseg, hdr->pgio_mirror_idx)) { - rpc_exit(task, -EHOSTDOWN); - return -EAGAIN; - } - ff_layout_write_record_layoutstats_start(task, hdr); return 0; } From 17aaec816751d77e443dee66d11d9e6ac00ac95c Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 26 Feb 2019 13:52:50 -0500 Subject: [PATCH 71/90] NFS/flexfiles: Don't invalidate DS deviceids for being unresponsive If the DS is unresponsive, we want to just mark it as such, while reporting the errors. If the server later returns the same deviceid in a new layout, then we don't want to have to look it up again. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.h | 18 ------------------ fs/nfs/flexfilelayout/flexfilelayoutdev.c | 6 +++--- 2 files changed, 3 insertions(+), 21 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h index 93508112486a..2e3c606625ef 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.h +++ b/fs/nfs/flexfilelayout/flexfilelayout.h @@ -180,24 +180,6 @@ ff_layout_no_read_on_rw(struct pnfs_layout_segment *lseg) return FF_LAYOUT_LSEG(lseg)->flags & FF_FLAGS_NO_READ_IO; } -static inline bool -ff_layout_test_devid_unavailable(struct nfs4_deviceid_node *node) -{ - /* - * Flexfiles should never mark a DS unavailable, but if it does - * print a (ratelimited) warning as this can affect performance. - */ - if (nfs4_test_deviceid_unavailable(node)) { - u32 *p = (u32 *)node->deviceid.data; - - pr_warn_ratelimited("NFS: flexfiles layout referencing an " - "unavailable device [%x%x%x%x]\n", - p[0], p[1], p[2], p[3]); - return true; - } - return false; -} - static inline int nfs4_ff_layout_ds_version(struct pnfs_layout_segment *lseg, u32 ds_idx) { diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index 92efb5692eb3..7fd92226e4b9 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -186,7 +186,7 @@ out_err: static void ff_layout_mark_devid_invalid(struct pnfs_layout_segment *lseg, struct nfs4_deviceid_node *devid) { - nfs4_delete_deviceid(devid->ld, devid->nfs_client, &devid->deviceid); + nfs4_mark_deviceid_unavailable(devid); if (!ff_layout_has_available_ds(lseg)) pnfs_error_mark_layout_for_return(lseg->pls_layout->plh_inode, lseg); @@ -605,7 +605,7 @@ static bool ff_read_layout_has_available_ds(struct pnfs_layout_segment *lseg) if (IS_ERR(mirror->mirror_ds)) continue; devid = &mirror->mirror_ds->id_node; - if (!ff_layout_test_devid_unavailable(devid)) + if (!nfs4_test_deviceid_unavailable(devid)) return true; } } @@ -626,7 +626,7 @@ static bool ff_rw_layout_has_available_ds(struct pnfs_layout_segment *lseg) if (!mirror->mirror_ds) continue; devid = &mirror->mirror_ds->id_node; - if (ff_layout_test_devid_unavailable(devid)) + if (nfs4_test_deviceid_unavailable(devid)) return false; } From 76c6690522bb3e335ce1e201360df8776cab4d2c Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 14 Feb 2019 17:32:40 -0500 Subject: [PATCH 72/90] NFS/flexfiles: Speed up read failover when DSes are down If we notice that a DS may be down, we should attempt to read from the other mirrors first before we go back to retry the dead DS. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 74 +++++++++++++++++++++----- fs/nfs/pnfs.h | 1 + fs/nfs/pnfs_dev.c | 10 ++++ 3 files changed, 73 insertions(+), 12 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 747ab2fafb69..44ccfce3784e 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -788,10 +788,28 @@ ff_layout_alloc_commit_info(struct pnfs_layout_segment *lseg, } } +static void +ff_layout_mark_ds_unreachable(struct pnfs_layout_segment *lseg, int idx) +{ + struct nfs4_deviceid_node *devid = FF_LAYOUT_DEVID_NODE(lseg, idx); + + if (devid) + nfs4_mark_deviceid_unavailable(devid); +} + +static void +ff_layout_mark_ds_reachable(struct pnfs_layout_segment *lseg, int idx) +{ + struct nfs4_deviceid_node *devid = FF_LAYOUT_DEVID_NODE(lseg, idx); + + if (devid) + nfs4_mark_deviceid_available(devid); +} + static struct nfs4_pnfs_ds * -ff_layout_choose_best_ds_for_read(struct pnfs_layout_segment *lseg, - int start_idx, - int *best_idx) +ff_layout_choose_ds_for_read(struct pnfs_layout_segment *lseg, + int start_idx, int *best_idx, + bool check_device) { struct nfs4_ff_layout_segment *fls = FF_LAYOUT_LSEG(lseg); struct nfs4_ff_layout_mirror *mirror; @@ -799,25 +817,53 @@ ff_layout_choose_best_ds_for_read(struct pnfs_layout_segment *lseg, bool fail_return = false; int idx; - /* mirrors are sorted by efficiency */ + /* mirrors are initially sorted by efficiency */ for (idx = start_idx; idx < fls->mirror_array_cnt; idx++) { if (idx+1 == fls->mirror_array_cnt) - fail_return = true; + fail_return = !check_device; mirror = FF_LAYOUT_COMP(lseg, idx); - if (ff_layout_test_devid_unavailable(&mirror->mirror_ds->id_node)) + ds = nfs4_ff_layout_prepare_ds(lseg, mirror, fail_return); + if (!ds) continue; - ds = nfs4_ff_layout_prepare_ds(lseg, mirror, fail_return); - if (ds) { - *best_idx = idx; - return ds; - } + if (check_device && + nfs4_test_deviceid_unavailable(&mirror->mirror_ds->id_node)) + continue; + + *best_idx = idx; + return ds; } return NULL; } +static struct nfs4_pnfs_ds * +ff_layout_choose_any_ds_for_read(struct pnfs_layout_segment *lseg, + int start_idx, int *best_idx) +{ + return ff_layout_choose_ds_for_read(lseg, start_idx, best_idx, false); +} + +static struct nfs4_pnfs_ds * +ff_layout_choose_valid_ds_for_read(struct pnfs_layout_segment *lseg, + int start_idx, int *best_idx) +{ + return ff_layout_choose_ds_for_read(lseg, start_idx, best_idx, true); +} + +static struct nfs4_pnfs_ds * +ff_layout_choose_best_ds_for_read(struct pnfs_layout_segment *lseg, + int start_idx, int *best_idx) +{ + struct nfs4_pnfs_ds *ds; + + ds = ff_layout_choose_valid_ds_for_read(lseg, start_idx, best_idx); + if (ds) + return ds; + return ff_layout_choose_any_ds_for_read(lseg, start_idx, best_idx); +} + static void ff_layout_pg_get_read(struct nfs_pageio_descriptor *pgio, struct nfs_page *req, @@ -1167,8 +1213,10 @@ static int ff_layout_async_handle_error(struct rpc_task *task, { int vers = clp->cl_nfs_mod->rpc_vers->number; - if (task->tk_status >= 0) + if (task->tk_status >= 0) { + ff_layout_mark_ds_reachable(lseg, idx); return 0; + } /* Handle the case of an invalid layout segment */ if (!pnfs_is_valid_lseg(lseg)) @@ -1231,6 +1279,8 @@ static void ff_layout_io_track_ds_error(struct pnfs_layout_segment *lseg, err = ff_layout_track_ds_error(FF_LAYOUT_FROM_HDR(lseg->pls_layout), mirror, offset, length, status, opnum, GFP_NOIO); + if (status == NFS4ERR_NXIO) + ff_layout_mark_ds_unreachable(lseg, idx); pnfs_error_mark_layout_for_return(lseg->pls_layout->plh_inode, lseg); dprintk("%s: err %d op %d status %u\n", __func__, err, opnum, status); } diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h index 56659ccce1d8..c0420b979d88 100644 --- a/fs/nfs/pnfs.h +++ b/fs/nfs/pnfs.h @@ -350,6 +350,7 @@ void nfs4_delete_deviceid(const struct pnfs_layoutdriver_type *, const struct nf void nfs4_init_deviceid_node(struct nfs4_deviceid_node *, struct nfs_server *, const struct nfs4_deviceid *); bool nfs4_put_deviceid_node(struct nfs4_deviceid_node *); +void nfs4_mark_deviceid_available(struct nfs4_deviceid_node *node); void nfs4_mark_deviceid_unavailable(struct nfs4_deviceid_node *node); bool nfs4_test_deviceid_unavailable(struct nfs4_deviceid_node *node); void nfs4_deviceid_purge_client(const struct nfs_client *); diff --git a/fs/nfs/pnfs_dev.c b/fs/nfs/pnfs_dev.c index ec6b607ff1d7..537b80d693f1 100644 --- a/fs/nfs/pnfs_dev.c +++ b/fs/nfs/pnfs_dev.c @@ -283,6 +283,16 @@ nfs4_put_deviceid_node(struct nfs4_deviceid_node *d) } EXPORT_SYMBOL_GPL(nfs4_put_deviceid_node); +void +nfs4_mark_deviceid_available(struct nfs4_deviceid_node *node) +{ + if (test_bit(NFS_DEVICEID_UNAVAILABLE, &node->flags)) { + clear_bit(NFS_DEVICEID_UNAVAILABLE, &node->flags); + smp_mb__after_atomic(); + } +} +EXPORT_SYMBOL_GPL(nfs4_mark_deviceid_available); + void nfs4_mark_deviceid_unavailable(struct nfs4_deviceid_node *node) { From 749da527b3e7bff92cc76b907bc222ef2df6fcdc Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 28 Feb 2019 10:34:13 -0500 Subject: [PATCH 73/90] NFS/flexfiles: Simplify nfs4_ff_layout_select_ds_fh() Pass in a pointer to the mirror rather than having to retrieve it from the array and then verify the resulting pointer. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 4 ++-- fs/nfs/flexfilelayout/flexfilelayout.h | 2 +- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 15 ++------------- 3 files changed, 5 insertions(+), 16 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 44ccfce3784e..f57cb38bd286 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -1795,7 +1795,7 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr) hdr->pgio_done_cb = ff_layout_read_done_cb; refcount_inc(&ds->ds_clp->cl_count); hdr->ds_clp = ds->ds_clp; - fh = nfs4_ff_layout_select_ds_fh(lseg, idx); + fh = nfs4_ff_layout_select_ds_fh(mirror); if (fh) hdr->args.fh = fh; @@ -1863,7 +1863,7 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync) refcount_inc(&ds->ds_clp->cl_count); hdr->ds_clp = ds->ds_clp; hdr->ds_commit_idx = idx; - fh = nfs4_ff_layout_select_ds_fh(lseg, idx); + fh = nfs4_ff_layout_select_ds_fh(mirror); if (fh) hdr->args.fh = fh; diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h index 2e3c606625ef..935cba942834 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.h +++ b/fs/nfs/flexfilelayout/flexfilelayout.h @@ -203,7 +203,7 @@ unsigned int ff_layout_fetch_ds_ioerr(struct pnfs_layout_hdr *lo, struct list_head *head, unsigned int maxnum); struct nfs_fh * -nfs4_ff_layout_select_ds_fh(struct pnfs_layout_segment *lseg, u32 mirror_idx); +nfs4_ff_layout_select_ds_fh(struct nfs4_ff_layout_mirror *mirror); int nfs4_ff_layout_select_ds_stateid(struct pnfs_layout_segment *lseg, u32 mirror_idx, diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index 7fd92226e4b9..3a2419313fe9 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -351,21 +351,10 @@ ff_layout_get_mirror_cred(struct nfs4_ff_layout_mirror *mirror, u32 iomode) } struct nfs_fh * -nfs4_ff_layout_select_ds_fh(struct pnfs_layout_segment *lseg, u32 mirror_idx) +nfs4_ff_layout_select_ds_fh(struct nfs4_ff_layout_mirror *mirror) { - struct nfs4_ff_layout_mirror *mirror = FF_LAYOUT_COMP(lseg, mirror_idx); - struct nfs_fh *fh = NULL; - - if (!ff_layout_mirror_valid(lseg, mirror, false)) { - pr_err_ratelimited("NFS: %s: No data server for mirror offset index %d\n", - __func__, mirror_idx); - goto out; - } - /* FIXME: For now assume there is only 1 version available for the DS */ - fh = &mirror->fh_versions[0]; -out: - return fh; + return &mirror->fh_versions[0]; } int From 561d6f8aaffcc9d346e452361a6d80cb8461799d Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 28 Feb 2019 10:38:41 -0500 Subject: [PATCH 74/90] NFS/flexfiles: Simplify nfs4_ff_find_or_create_ds_client() Pass in a pointer to the mirror rather than forcing another array access. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 6 +++--- fs/nfs/flexfilelayout/flexfilelayout.h | 3 +-- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 7 ++----- 3 files changed, 6 insertions(+), 10 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index f57cb38bd286..88b9190b761b 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -1778,7 +1778,7 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr) if (!ds) goto out_failed; - ds_clnt = nfs4_ff_find_or_create_ds_client(lseg, idx, ds->ds_clp, + ds_clnt = nfs4_ff_find_or_create_ds_client(mirror, ds->ds_clp, hdr->inode); if (IS_ERR(ds_clnt)) goto out_failed; @@ -1843,7 +1843,7 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync) if (!ds) goto out_failed; - ds_clnt = nfs4_ff_find_or_create_ds_client(lseg, idx, ds->ds_clp, + ds_clnt = nfs4_ff_find_or_create_ds_client(mirror, ds->ds_clp, hdr->inode); if (IS_ERR(ds_clnt)) goto out_failed; @@ -1928,7 +1928,7 @@ static int ff_layout_initiate_commit(struct nfs_commit_data *data, int how) if (!ds) goto out_err; - ds_clnt = nfs4_ff_find_or_create_ds_client(lseg, idx, ds->ds_clp, + ds_clnt = nfs4_ff_find_or_create_ds_client(mirror, ds->ds_clp, data->inode); if (IS_ERR(ds_clnt)) goto out_err; diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h index 935cba942834..2bda4baa655a 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.h +++ b/fs/nfs/flexfilelayout/flexfilelayout.h @@ -215,8 +215,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, bool fail_return); struct rpc_clnt * -nfs4_ff_find_or_create_ds_client(struct pnfs_layout_segment *lseg, - u32 ds_idx, +nfs4_ff_find_or_create_ds_client(struct nfs4_ff_layout_mirror *mirror, struct nfs_client *ds_clp, struct inode *inode); const struct cred *ff_layout_get_ds_cred(struct pnfs_layout_segment *lseg, diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index 3a2419313fe9..1fe66745f452 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -464,8 +464,7 @@ ff_layout_get_ds_cred(struct pnfs_layout_segment *lseg, u32 ds_idx, /** * nfs4_ff_find_or_create_ds_client - Find or create a DS rpc client - * @lseg: pointer to layout segment - * @ds_idx: mirror index + * @mirror: pointer to the mirror * @ds_clp: nfs_client for the DS * @inode: pointer to inode * @@ -473,11 +472,9 @@ ff_layout_get_ds_cred(struct pnfs_layout_segment *lseg, u32 ds_idx, * in the nfs_client cl_ds_clients list. */ struct rpc_clnt * -nfs4_ff_find_or_create_ds_client(struct pnfs_layout_segment *lseg, u32 ds_idx, +nfs4_ff_find_or_create_ds_client(struct nfs4_ff_layout_mirror *mirror, struct nfs_client *ds_clp, struct inode *inode) { - struct nfs4_ff_layout_mirror *mirror = FF_LAYOUT_COMP(lseg, ds_idx); - switch (mirror->mirror_ds->ds_versions[0].version) { case 3: /* For NFSv3 DS, flavor is set when creating DS connections */ From 312cd4cb126a7936ad25306d14209fa6fd00f30a Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 28 Feb 2019 10:41:57 -0500 Subject: [PATCH 75/90] NFS/flexfiles: Simplify ff_layout_get_ds_cred() Pass in a pointer to the mirror rather than forcing another array access. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 6 +++--- fs/nfs/flexfilelayout/flexfilelayout.h | 5 +++-- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 6 +++--- 3 files changed, 9 insertions(+), 8 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 88b9190b761b..24a7d74743ff 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -1783,7 +1783,7 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr) if (IS_ERR(ds_clnt)) goto out_failed; - ds_cred = ff_layout_get_ds_cred(lseg, idx, hdr->cred); + ds_cred = ff_layout_get_ds_cred(mirror, &lseg->pls_range, hdr->cred); if (!ds_cred) goto out_failed; @@ -1848,7 +1848,7 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync) if (IS_ERR(ds_clnt)) goto out_failed; - ds_cred = ff_layout_get_ds_cred(lseg, idx, hdr->cred); + ds_cred = ff_layout_get_ds_cred(mirror, &lseg->pls_range, hdr->cred); if (!ds_cred) goto out_failed; @@ -1933,7 +1933,7 @@ static int ff_layout_initiate_commit(struct nfs_commit_data *data, int how) if (IS_ERR(ds_clnt)) goto out_err; - ds_cred = ff_layout_get_ds_cred(lseg, idx, data->cred); + ds_cred = ff_layout_get_ds_cred(mirror, &lseg->pls_range, data->cred); if (!ds_cred) goto out_err; diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h index 2bda4baa655a..e0b2c3deae5a 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.h +++ b/fs/nfs/flexfilelayout/flexfilelayout.h @@ -218,8 +218,9 @@ struct rpc_clnt * nfs4_ff_find_or_create_ds_client(struct nfs4_ff_layout_mirror *mirror, struct nfs_client *ds_clp, struct inode *inode); -const struct cred *ff_layout_get_ds_cred(struct pnfs_layout_segment *lseg, - u32 ds_idx, const struct cred *mdscred); +const struct cred *ff_layout_get_ds_cred(struct nfs4_ff_layout_mirror *mirror, + const struct pnfs_layout_range *range, + const struct cred *mdscred); bool ff_layout_avoid_mds_available_ds(struct pnfs_layout_segment *lseg); bool ff_layout_avoid_read_on_rw(struct pnfs_layout_segment *lseg); diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index 1fe66745f452..1673935074cc 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -446,14 +446,14 @@ out: } const struct cred * -ff_layout_get_ds_cred(struct pnfs_layout_segment *lseg, u32 ds_idx, +ff_layout_get_ds_cred(struct nfs4_ff_layout_mirror *mirror, + const struct pnfs_layout_range *range, const struct cred *mdscred) { - struct nfs4_ff_layout_mirror *mirror = FF_LAYOUT_COMP(lseg, ds_idx); const struct cred *cred; if (mirror && !mirror->mirror_ds->ds_versions[0].tightly_coupled) { - cred = ff_layout_get_mirror_cred(mirror, lseg->pls_range.iomode); + cred = ff_layout_get_mirror_cred(mirror, range->iomode); if (!cred) cred = get_cred(mdscred); } else { From 626d48b12cde3ae9740e1da16fc9c1d6d7bbed14 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 28 Feb 2019 10:44:15 -0500 Subject: [PATCH 76/90] NFS/flexfile: Simplify nfs4_ff_layout_ds_version() Pass in a pointer to the mirror rather than forcing another array access. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 6 +++--- fs/nfs/flexfilelayout/flexfilelayout.h | 4 ++-- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 24a7d74743ff..650df8d109a1 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -1787,7 +1787,7 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr) if (!ds_cred) goto out_failed; - vers = nfs4_ff_layout_ds_version(lseg, idx); + vers = nfs4_ff_layout_ds_version(mirror); dprintk("%s USE DS: %s cl_count %d vers %d\n", __func__, ds->ds_remotestr, refcount_read(&ds->ds_clp->cl_count), vers); @@ -1852,7 +1852,7 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync) if (!ds_cred) goto out_failed; - vers = nfs4_ff_layout_ds_version(lseg, idx); + vers = nfs4_ff_layout_ds_version(mirror); dprintk("%s ino %lu sync %d req %zu@%llu DS: %s cl_count %d vers %d\n", __func__, hdr->inode->i_ino, sync, (size_t) hdr->args.count, @@ -1937,7 +1937,7 @@ static int ff_layout_initiate_commit(struct nfs_commit_data *data, int how) if (!ds_cred) goto out_err; - vers = nfs4_ff_layout_ds_version(lseg, idx); + vers = nfs4_ff_layout_ds_version(mirror); dprintk("%s ino %lu, how %d cl_count %d vers %d\n", __func__, data->inode->i_ino, how, refcount_read(&ds->ds_clp->cl_count), diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h index e0b2c3deae5a..99cfe2e8f32a 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.h +++ b/fs/nfs/flexfilelayout/flexfilelayout.h @@ -181,9 +181,9 @@ ff_layout_no_read_on_rw(struct pnfs_layout_segment *lseg) } static inline int -nfs4_ff_layout_ds_version(struct pnfs_layout_segment *lseg, u32 ds_idx) +nfs4_ff_layout_ds_version(const struct nfs4_ff_layout_mirror *mirror) { - return FF_LAYOUT_COMP(lseg, ds_idx)->mirror_ds->ds_versions[0].version; + return mirror->mirror_ds->ds_versions[0].version; } struct nfs4_ff_layout_ds * From 4cbc8a571c24133a8b645c62188205908ef2ea83 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 28 Feb 2019 10:49:11 -0500 Subject: [PATCH 77/90] NFS/flexfile: Simplify nfs4_ff_layout_select_ds_stateid() Pass in a pointer to the mirror rather than forcing another array access. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayout.c | 8 ++------ fs/nfs/flexfilelayout/flexfilelayout.h | 7 +++---- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 21 +++++---------------- 3 files changed, 10 insertions(+), 26 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 650df8d109a1..f9264e1922a2 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -1799,9 +1799,7 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr) if (fh) hdr->args.fh = fh; - if (vers == 4 && - !nfs4_ff_layout_select_ds_stateid(lseg, idx, &hdr->args.stateid)) - goto out_failed; + nfs4_ff_layout_select_ds_stateid(mirror, &hdr->args.stateid); /* * Note that if we ever decide to split across DSes, @@ -1867,9 +1865,7 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync) if (fh) hdr->args.fh = fh; - if (vers == 4 && - !nfs4_ff_layout_select_ds_stateid(lseg, idx, &hdr->args.stateid)) - goto out_failed; + nfs4_ff_layout_select_ds_stateid(mirror, &hdr->args.stateid); /* * Note that if we ever decide to split across DSes, diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h index 99cfe2e8f32a..2f369966abf7 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.h +++ b/fs/nfs/flexfilelayout/flexfilelayout.h @@ -204,10 +204,9 @@ unsigned int ff_layout_fetch_ds_ioerr(struct pnfs_layout_hdr *lo, unsigned int maxnum); struct nfs_fh * nfs4_ff_layout_select_ds_fh(struct nfs4_ff_layout_mirror *mirror); -int -nfs4_ff_layout_select_ds_stateid(struct pnfs_layout_segment *lseg, - u32 mirror_idx, - nfs4_stateid *stateid); +void +nfs4_ff_layout_select_ds_stateid(const struct nfs4_ff_layout_mirror *mirror, + nfs4_stateid *stateid); struct nfs4_pnfs_ds * nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index 1673935074cc..d8ef4042f25a 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -357,23 +357,12 @@ nfs4_ff_layout_select_ds_fh(struct nfs4_ff_layout_mirror *mirror) return &mirror->fh_versions[0]; } -int -nfs4_ff_layout_select_ds_stateid(struct pnfs_layout_segment *lseg, - u32 mirror_idx, - nfs4_stateid *stateid) +void +nfs4_ff_layout_select_ds_stateid(const struct nfs4_ff_layout_mirror *mirror, + nfs4_stateid *stateid) { - struct nfs4_ff_layout_mirror *mirror = FF_LAYOUT_COMP(lseg, mirror_idx); - - if (!ff_layout_mirror_valid(lseg, mirror, false)) { - pr_err_ratelimited("NFS: %s: No data server for mirror offset index %d\n", - __func__, mirror_idx); - goto out; - } - - nfs4_stateid_copy(stateid, &mirror->stateid); - return 1; -out: - return 0; + if (nfs4_ff_layout_ds_version(mirror) == 4) + nfs4_stateid_copy(stateid, &mirror->stateid); } /** From 29a23909e41f6bfc892b362a75d7ed5fb74af1a2 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 28 Feb 2019 11:35:33 -0500 Subject: [PATCH 78/90] NFS/flexfiles: Remove dead code in ff_layout_mirror_valid() nfs4_ff_alloc_deviceid_node() guarantees that if mirror->mirror_ds is a valid pointer, then so is mirror->mirror_ds->ds. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 15 --------------- 1 file changed, 15 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index d8ef4042f25a..69fec9fe7f6c 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -183,15 +183,6 @@ out_err: return NULL; } -static void ff_layout_mark_devid_invalid(struct pnfs_layout_segment *lseg, - struct nfs4_deviceid_node *devid) -{ - nfs4_mark_deviceid_unavailable(devid); - if (!ff_layout_has_available_ds(lseg)) - pnfs_error_mark_layout_for_return(lseg->pls_layout->plh_inode, - lseg); -} - static bool ff_layout_mirror_valid(struct pnfs_layout_segment *lseg, struct nfs4_ff_layout_mirror *mirror, bool create) @@ -221,12 +212,6 @@ static bool ff_layout_mirror_valid(struct pnfs_layout_segment *lseg, if (IS_ERR(mirror->mirror_ds)) goto outerr; - if (mirror->mirror_ds->ds == NULL) { - struct nfs4_deviceid_node *devid; - devid = &mirror->mirror_ds->id_node; - ff_layout_mark_devid_invalid(lseg, devid); - return false; - } return true; outerr: return false; From cefa587a40bb5333901486632d4062f40a146585 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 28 Feb 2019 11:01:05 -0500 Subject: [PATCH 79/90] NFS/flexfiles: Clean up mirror DS initialisation Get rid of the redundant parameter and rename the function ff_layout_mirror_valid() to ff_layout_init_mirror_ds() for clarity. Signed-off-by: Trond Myklebust --- fs/nfs/flexfilelayout/flexfilelayoutdev.c | 66 +++++++++++------------ 1 file changed, 31 insertions(+), 35 deletions(-) diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index 69fec9fe7f6c..a809989807d6 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -183,40 +183,6 @@ out_err: return NULL; } -static bool ff_layout_mirror_valid(struct pnfs_layout_segment *lseg, - struct nfs4_ff_layout_mirror *mirror, - bool create) -{ - if (mirror == NULL) - goto outerr; - if (mirror->mirror_ds == NULL) { - if (create) { - struct nfs4_deviceid_node *node; - struct pnfs_layout_hdr *lh = lseg->pls_layout; - struct nfs4_ff_layout_ds *mirror_ds = ERR_PTR(-ENODEV); - - node = nfs4_find_get_deviceid(NFS_SERVER(lh->plh_inode), - &mirror->devid, lh->plh_lc_cred, - GFP_KERNEL); - if (node) - mirror_ds = FF_LAYOUT_MIRROR_DS(node); - - /* check for race with another call to this function */ - if (cmpxchg(&mirror->mirror_ds, NULL, mirror_ds) && - mirror_ds != ERR_PTR(-ENODEV)) - nfs4_put_deviceid_node(node); - } else - goto outerr; - } - - if (IS_ERR(mirror->mirror_ds)) - goto outerr; - - return true; -outerr: - return false; -} - static void extend_ds_error(struct nfs4_ff_layout_ds_err *err, u64 offset, u64 length) { @@ -350,6 +316,36 @@ nfs4_ff_layout_select_ds_stateid(const struct nfs4_ff_layout_mirror *mirror, nfs4_stateid_copy(stateid, &mirror->stateid); } +static bool +ff_layout_init_mirror_ds(struct pnfs_layout_hdr *lo, + struct nfs4_ff_layout_mirror *mirror) +{ + if (mirror == NULL) + goto outerr; + if (mirror->mirror_ds == NULL) { + struct nfs4_deviceid_node *node; + struct nfs4_ff_layout_ds *mirror_ds = ERR_PTR(-ENODEV); + + node = nfs4_find_get_deviceid(NFS_SERVER(lo->plh_inode), + &mirror->devid, lo->plh_lc_cred, + GFP_KERNEL); + if (node) + mirror_ds = FF_LAYOUT_MIRROR_DS(node); + + /* check for race with another call to this function */ + if (cmpxchg(&mirror->mirror_ds, NULL, mirror_ds) && + mirror_ds != ERR_PTR(-ENODEV)) + nfs4_put_deviceid_node(node); + } + + if (IS_ERR(mirror->mirror_ds)) + goto outerr; + + return true; +outerr: + return false; +} + /** * nfs4_ff_layout_prepare_ds - prepare a DS connection for an RPC call * @lseg: the layout segment we're operating on @@ -378,7 +374,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, unsigned int max_payload; int status; - if (!ff_layout_mirror_valid(lseg, mirror, true)) + if (!ff_layout_init_mirror_ds(lseg->pls_layout, mirror)) goto noconnect; ds = mirror->mirror_ds->ds; From 12a3ad6184f86ba48f2269198c1a4520085f3002 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 2 Mar 2019 10:14:02 -0500 Subject: [PATCH 80/90] SUNRPC: Convert remaining GFP_NOIO, and GFP_NOWAIT sites in sunrpc Convert the remaining gfp_flags arguments in sunrpc to standard reclaiming allocations, now that we set memalloc_nofs_save() as appropriate. Signed-off-by: Trond Myklebust --- net/sunrpc/auth_gss/auth_gss.c | 2 +- net/sunrpc/sched.c | 7 ++----- net/sunrpc/xprtsock.c | 4 ++-- 3 files changed, 5 insertions(+), 8 deletions(-) diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index c67e2ad151ae..3fd56c0c90ae 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -1208,7 +1208,7 @@ gss_dup_cred(struct gss_auth *gss_auth, struct gss_cred *gss_cred) struct gss_cred *new; /* Make a copy of the cred so that we can reference count it */ - new = kzalloc(sizeof(*gss_cred), GFP_NOIO); + new = kzalloc(sizeof(*gss_cred), GFP_NOFS); if (new) { struct auth_cred acred = { .cred = gss_cred->gc_base.cr_cred, diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c index 2168d4d9c09f..f21557213a43 100644 --- a/net/sunrpc/sched.c +++ b/net/sunrpc/sched.c @@ -925,16 +925,13 @@ static void rpc_async_schedule(struct work_struct *work) * Most requests are 'small' (under 2KiB) and can be serviced from a * mempool, ensuring that NFS reads and writes can always proceed, * and that there is good locality of reference for these buffers. - * - * In order to avoid memory starvation triggering more writebacks of - * NFS requests, we avoid using GFP_KERNEL. */ int rpc_malloc(struct rpc_task *task) { struct rpc_rqst *rqst = task->tk_rqstp; size_t size = rqst->rq_callsize + rqst->rq_rcvsize; struct rpc_buffer *buf; - gfp_t gfp = GFP_NOIO | __GFP_NOWARN; + gfp_t gfp = GFP_NOFS; if (RPC_IS_SWAPPER(task)) gfp = __GFP_MEMALLOC | GFP_NOWAIT | __GFP_NOWARN; @@ -1015,7 +1012,7 @@ static void rpc_init_task(struct rpc_task *task, const struct rpc_task_setup *ta static struct rpc_task * rpc_alloc_task(void) { - return (struct rpc_task *)mempool_alloc(rpc_task_mempool, GFP_NOIO); + return (struct rpc_task *)mempool_alloc(rpc_task_mempool, GFP_NOFS); } /* diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index e829036ed81f..42f45d33dc56 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -423,7 +423,7 @@ xs_read_xdr_buf(struct socket *sock, struct msghdr *msg, int flags, want = xs_alloc_sparse_pages(buf, min_t(size_t, count - offset, buf->page_len), - GFP_NOWAIT); + GFP_KERNEL); if (seek < want) { ret = xs_read_bvec(sock, msg, flags, buf->bvec, xdr_buf_pagecount(buf), @@ -909,7 +909,7 @@ static int xs_nospace(struct rpc_rqst *req) static void xs_stream_prepare_request(struct rpc_rqst *req) { - req->rq_task->tk_status = xdr_alloc_bvec(&req->rq_rcv_buf, GFP_NOIO); + req->rq_task->tk_status = xdr_alloc_bvec(&req->rq_rcv_buf, GFP_KERNEL); } /* From 067c46967160b5c7052492734d695c501ef376fe Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 2 Mar 2019 10:22:56 -0500 Subject: [PATCH 81/90] NFSv4.1: Bump the default callback session slot count to 16 Users can still control this value explicitly using the max_session_cb_slots module parameter, but let's bump the default up to 16 for now. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4session.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/nfs/nfs4session.h b/fs/nfs/nfs4session.h index 230509b77121..b996ee23f1ba 100644 --- a/fs/nfs/nfs4session.h +++ b/fs/nfs/nfs4session.h @@ -10,7 +10,7 @@ /* maximum number of slots to use */ #define NFS4_DEF_SLOT_TABLE_SIZE (64U) -#define NFS4_DEF_CB_SLOT_TABLE_SIZE (1U) +#define NFS4_DEF_CB_SLOT_TABLE_SIZE (16U) #define NFS4_MAX_SLOT_TABLE (1024U) #define NFS4_NO_SLOT ((u32)-1) From 0d1bf3407c4ae887a464d136aaa5e9ef609834f0 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 2 Mar 2019 13:12:22 -0500 Subject: [PATCH 82/90] SUNRPC: Allow dynamic allocation of back channel slots Now that the reads happen in a process context rather than a softirq, it is safe to allocate back channel slots using a reclaiming allocation. Signed-off-by: Trond Myklebust --- net/sunrpc/backchannel_rqst.c | 39 +++++++++++++++++++++-------------- 1 file changed, 24 insertions(+), 15 deletions(-) diff --git a/net/sunrpc/backchannel_rqst.c b/net/sunrpc/backchannel_rqst.c index b9313c15ee3a..c47d82622fd1 100644 --- a/net/sunrpc/backchannel_rqst.c +++ b/net/sunrpc/backchannel_rqst.c @@ -235,7 +235,8 @@ out: list_empty(&xprt->bc_pa_list) ? "true" : "false"); } -static struct rpc_rqst *xprt_alloc_bc_request(struct rpc_xprt *xprt, __be32 xid) +static struct rpc_rqst *xprt_get_bc_request(struct rpc_xprt *xprt, __be32 xid, + struct rpc_rqst *new) { struct rpc_rqst *req = NULL; @@ -243,10 +244,9 @@ static struct rpc_rqst *xprt_alloc_bc_request(struct rpc_xprt *xprt, __be32 xid) if (atomic_read(&xprt->bc_free_slots) <= 0) goto not_found; if (list_empty(&xprt->bc_pa_list)) { - req = xprt_alloc_bc_req(xprt, GFP_ATOMIC); - if (!req) + if (!new) goto not_found; - list_add_tail(&req->rq_bc_pa_list, &xprt->bc_pa_list); + list_add_tail(&new->rq_bc_pa_list, &xprt->bc_pa_list); xprt->bc_alloc_count++; } req = list_first_entry(&xprt->bc_pa_list, struct rpc_rqst, @@ -256,8 +256,8 @@ static struct rpc_rqst *xprt_alloc_bc_request(struct rpc_xprt *xprt, __be32 xid) sizeof(req->rq_private_buf)); req->rq_xid = xid; req->rq_connect_cookie = xprt->connect_cookie; -not_found: dprintk("RPC: backchannel req=%p\n", req); +not_found: return req; } @@ -320,18 +320,27 @@ void xprt_free_bc_rqst(struct rpc_rqst *req) */ struct rpc_rqst *xprt_lookup_bc_request(struct rpc_xprt *xprt, __be32 xid) { - struct rpc_rqst *req; + struct rpc_rqst *req, *new = NULL; - spin_lock(&xprt->bc_pa_lock); - list_for_each_entry(req, &xprt->bc_pa_list, rq_bc_pa_list) { - if (req->rq_connect_cookie != xprt->connect_cookie) - continue; - if (req->rq_xid == xid) - goto found; - } - req = xprt_alloc_bc_request(xprt, xid); + do { + spin_lock(&xprt->bc_pa_lock); + list_for_each_entry(req, &xprt->bc_pa_list, rq_bc_pa_list) { + if (req->rq_connect_cookie != xprt->connect_cookie) + continue; + if (req->rq_xid == xid) + goto found; + } + req = xprt_get_bc_request(xprt, xid, new); found: - spin_unlock(&xprt->bc_pa_lock); + spin_unlock(&xprt->bc_pa_lock); + if (new) { + if (req != new) + xprt_free_bc_rqst(new); + break; + } else if (req) + break; + new = xprt_alloc_bc_req(xprt, GFP_KERNEL); + } while (new); return req; } From ed7dc973bd91da234d93aff6d033a5206a6c9885 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 4 Mar 2019 14:19:31 -0500 Subject: [PATCH 83/90] SUNRPC: Prevent thundering herd when the socket is not connected If the socket is not connected, then we want to initiate a reconnect rather that trying to transmit requests. If there is a large number of requests queued and waiting for the lock in call_transmit(), then it can take a while for one of the to loop back and retake the lock in call_connect. Fixes: 89f90fe1ad8b ("SUNRPC: Allow calls to xprt_transmit() to drain...") Signed-off-by: Trond Myklebust --- net/sunrpc/clnt.c | 21 +++++++++++++++++---- 1 file changed, 17 insertions(+), 4 deletions(-) diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 241e8423fd0c..7ab4da342ab5 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1807,7 +1807,12 @@ call_encode(struct rpc_task *task) xprt_request_enqueue_receive(task); xprt_request_enqueue_transmit(task); out: - task->tk_action = call_bind; + task->tk_action = call_transmit; + /* Check that the connection is OK */ + if (!xprt_bound(task->tk_xprt)) + task->tk_action = call_bind; + else if (!xprt_connected(task->tk_xprt)) + task->tk_action = call_connect; } /* @@ -1999,13 +2004,19 @@ call_transmit(struct rpc_task *task) { dprint_status(task); - task->tk_status = 0; + task->tk_action = call_transmit_status; if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) { if (!xprt_prepare_transmit(task)) return; - xprt_transmit(task); + task->tk_status = 0; + if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) { + if (!xprt_connected(task->tk_xprt)) { + task->tk_status = -ENOTCONN; + return; + } + xprt_transmit(task); + } } - task->tk_action = call_transmit_status; xprt_end_transmit(task); } @@ -2067,6 +2078,8 @@ call_transmit_status(struct rpc_task *task) case -EADDRINUSE: case -ENOTCONN: case -EPIPE: + task->tk_action = call_bind; + task->tk_status = 0; break; } } From 477687e1116ad16180caf8633dd830b296a5ce73 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 5 Mar 2019 07:30:48 -0500 Subject: [PATCH 84/90] SUNRPC: Fix up RPC back channel transmission Now that transmissions happen through a queue, we require the RPC tasks to handle error conditions that may have been set while they were sleeping. The back channel does not currently do this, but assumes that any error condition happens during its own call to xprt_transmit(). The solution is to ensure that the back channel splits out the error handling just like the forward channel does. Fixes: 89f90fe1ad8b ("SUNRPC: Allow calls to xprt_transmit() to drain...") Signed-off-by: Trond Myklebust --- net/sunrpc/clnt.c | 63 +++++++++++++++++++++++++---------------------- 1 file changed, 34 insertions(+), 29 deletions(-) diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 7ab4da342ab5..b9558e10c6c1 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -66,9 +66,6 @@ static void call_decode(struct rpc_task *task); static void call_bind(struct rpc_task *task); static void call_bind_status(struct rpc_task *task); static void call_transmit(struct rpc_task *task); -#if defined(CONFIG_SUNRPC_BACKCHANNEL) -static void call_bc_transmit(struct rpc_task *task); -#endif /* CONFIG_SUNRPC_BACKCHANNEL */ static void call_status(struct rpc_task *task); static void call_transmit_status(struct rpc_task *task); static void call_refresh(struct rpc_task *task); @@ -1133,6 +1130,8 @@ rpc_call_async(struct rpc_clnt *clnt, const struct rpc_message *msg, int flags, EXPORT_SYMBOL_GPL(rpc_call_async); #if defined(CONFIG_SUNRPC_BACKCHANNEL) +static void call_bc_encode(struct rpc_task *task); + /** * rpc_run_bc_task - Allocate a new RPC task for backchannel use, then run * rpc_execute against it @@ -1154,7 +1153,7 @@ struct rpc_task *rpc_run_bc_task(struct rpc_rqst *req) task = rpc_new_task(&task_setup_data); xprt_init_bc_request(req, task); - task->tk_action = call_bc_transmit; + task->tk_action = call_bc_encode; atomic_inc(&task->tk_count); WARN_ON_ONCE(atomic_read(&task->tk_count) != 2); rpc_execute(task); @@ -2085,33 +2084,40 @@ call_transmit_status(struct rpc_task *task) } #if defined(CONFIG_SUNRPC_BACKCHANNEL) +static void call_bc_transmit(struct rpc_task *task); +static void call_bc_transmit_status(struct rpc_task *task); + +static void +call_bc_encode(struct rpc_task *task) +{ + xprt_request_enqueue_transmit(task); + task->tk_action = call_bc_transmit; +} + /* * 5b. Send the backchannel RPC reply. On error, drop the reply. In * addition, disconnect on connectivity errors. */ static void call_bc_transmit(struct rpc_task *task) +{ + task->tk_action = call_bc_transmit_status; + if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) { + if (!xprt_prepare_transmit(task)) + return; + task->tk_status = 0; + xprt_transmit(task); + } + xprt_end_transmit(task); +} + +static void +call_bc_transmit_status(struct rpc_task *task) { struct rpc_rqst *req = task->tk_rqstp; - if (rpc_task_need_encode(task)) - xprt_request_enqueue_transmit(task); - if (!test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) - goto out_wakeup; - - if (!xprt_prepare_transmit(task)) - goto out_retry; - - if (task->tk_status < 0) { - printk(KERN_NOTICE "RPC: Could not send backchannel reply " - "error: %d\n", task->tk_status); - goto out_done; - } - - xprt_transmit(task); - - xprt_end_transmit(task); dprint_status(task); + switch (task->tk_status) { case 0: /* Success */ @@ -2125,8 +2131,14 @@ call_bc_transmit(struct rpc_task *task) case -ENOTCONN: case -EPIPE: break; + case -ENOBUFS: + rpc_delay(task, HZ>>2); + /* fall through */ + case -EBADSLT: case -EAGAIN: - goto out_retry; + task->tk_status = 0; + task->tk_action = call_bc_transmit; + return; case -ETIMEDOUT: /* * Problem reaching the server. Disconnect and let the @@ -2145,18 +2157,11 @@ call_bc_transmit(struct rpc_task *task) * We were unable to reply and will have to drop the * request. The server should reconnect and retransmit. */ - WARN_ON_ONCE(task->tk_status == -EAGAIN); printk(KERN_NOTICE "RPC: Could not send backchannel reply " "error: %d\n", task->tk_status); break; } -out_wakeup: - rpc_wake_up_queued_task(&req->rq_xprt->pending, task); -out_done: task->tk_action = rpc_exit_task; - return; -out_retry: - task->tk_status = 0; } #endif /* CONFIG_SUNRPC_BACKCHANNEL */ From 7b3fef8e4157ed424bcde039a60a730aa0dfb0eb Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 7 Mar 2019 14:10:32 -0500 Subject: [PATCH 85/90] SUNRPC: Respect RPC call timeouts when retrying transmission Fix a regression where soft and softconn requests are not timing out as expected. Fixes: 89f90fe1ad8b ("SUNRPC: Allow calls to xprt_transmit() to drain...") Signed-off-by: Trond Myklebust --- net/sunrpc/clnt.c | 42 ++++++++++++++++++++++++------------------ 1 file changed, 24 insertions(+), 18 deletions(-) diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index b9558e10c6c1..311029b7c33a 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -79,6 +79,7 @@ static int rpc_encode_header(struct rpc_task *task, static int rpc_decode_header(struct rpc_task *task, struct xdr_stream *xdr); static int rpc_ping(struct rpc_clnt *clnt); +static void rpc_check_timeout(struct rpc_task *task); static void rpc_register_client(struct rpc_clnt *clnt) { @@ -1962,8 +1963,7 @@ call_connect_status(struct rpc_task *task) break; if (clnt->cl_autobind) { rpc_force_rebind(clnt); - task->tk_action = call_bind; - return; + goto out_retry; } /* fall through */ case -ECONNRESET: @@ -1983,16 +1983,19 @@ call_connect_status(struct rpc_task *task) /* fall through */ case -ENOTCONN: case -EAGAIN: - /* Check for timeouts before looping back to call_bind */ case -ETIMEDOUT: - task->tk_action = call_timeout; - return; + goto out_retry; case 0: clnt->cl_stats->netreconn++; task->tk_action = call_transmit; return; } rpc_exit(task, status); + return; +out_retry: + /* Check for timeouts before looping back to call_bind */ + task->tk_action = call_bind; + rpc_check_timeout(task); } /* @@ -2069,7 +2072,7 @@ call_transmit_status(struct rpc_task *task) trace_xprt_ping(task->tk_xprt, task->tk_status); rpc_exit(task, task->tk_status); - break; + return; } /* fall through */ case -ECONNRESET: @@ -2081,6 +2084,7 @@ call_transmit_status(struct rpc_task *task) task->tk_status = 0; break; } + rpc_check_timeout(task); } #if defined(CONFIG_SUNRPC_BACKCHANNEL) @@ -2217,7 +2221,7 @@ call_status(struct rpc_task *task) case -EPIPE: case -ENOTCONN: case -EAGAIN: - task->tk_action = call_encode; + task->tk_action = call_timeout; break; case -EIO: /* shutdown or soft timeout */ @@ -2231,20 +2235,13 @@ call_status(struct rpc_task *task) } } -/* - * 6a. Handle RPC timeout - * We do not release the request slot, so we keep using the - * same XID for all retransmits. - */ static void -call_timeout(struct rpc_task *task) +rpc_check_timeout(struct rpc_task *task) { struct rpc_clnt *clnt = task->tk_client; - if (xprt_adjust_timeout(task->tk_rqstp) == 0) { - dprintk("RPC: %5u call_timeout (minor)\n", task->tk_pid); - goto retry; - } + if (xprt_adjust_timeout(task->tk_rqstp) == 0) + return; dprintk("RPC: %5u call_timeout (major)\n", task->tk_pid); task->tk_timeouts++; @@ -2280,10 +2277,19 @@ call_timeout(struct rpc_task *task) * event? RFC2203 requires the server to drop all such requests. */ rpcauth_invalcred(task); +} -retry: +/* + * 6a. Handle RPC timeout + * We do not release the request slot, so we keep using the + * same XID for all retransmits. + */ +static void +call_timeout(struct rpc_task *task) +{ task->tk_action = call_encode; task->tk_status = 0; + rpc_check_timeout(task); } /* From cea57789e4081870ac3498fbefabbbd0d0fd8434 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 9 Mar 2019 16:06:47 -0500 Subject: [PATCH 86/90] SUNRPC: Clean up Replace remaining callers of call_timeout() with rpc_check_timeout(). Signed-off-by: Trond Myklebust --- net/sunrpc/clnt.c | 47 ++++++++++++++--------------------------------- 1 file changed, 14 insertions(+), 33 deletions(-) diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 311029b7c33a..a0e7322fc171 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -70,7 +70,6 @@ static void call_status(struct rpc_task *task); static void call_transmit_status(struct rpc_task *task); static void call_refresh(struct rpc_task *task); static void call_refreshresult(struct rpc_task *task); -static void call_timeout(struct rpc_task *task); static void call_connect(struct rpc_task *task); static void call_connect_status(struct rpc_task *task); @@ -1908,7 +1907,8 @@ call_bind_status(struct rpc_task *task) retry_timeout: task->tk_status = 0; - task->tk_action = call_timeout; + task->tk_action = call_encode; + rpc_check_timeout(task); } /* @@ -2197,10 +2197,8 @@ call_status(struct rpc_task *task) case -EHOSTUNREACH: case -ENETUNREACH: case -EPERM: - if (RPC_IS_SOFTCONN(task)) { - rpc_exit(task, status); - break; - } + if (RPC_IS_SOFTCONN(task)) + goto out_exit; /* * Delay any retries for 3 seconds, then handle as if it * were a timeout. @@ -2208,7 +2206,6 @@ call_status(struct rpc_task *task) rpc_delay(task, 3*HZ); /* fall through */ case -ETIMEDOUT: - task->tk_action = call_timeout; break; case -ECONNREFUSED: case -ECONNRESET: @@ -2221,18 +2218,21 @@ call_status(struct rpc_task *task) case -EPIPE: case -ENOTCONN: case -EAGAIN: - task->tk_action = call_timeout; break; case -EIO: /* shutdown or soft timeout */ - rpc_exit(task, status); - break; + goto out_exit; default: if (clnt->cl_chatty) printk("%s: RPC call returned error %d\n", clnt->cl_program->name, -status); - rpc_exit(task, status); + goto out_exit; } + task->tk_action = call_encode; + rpc_check_timeout(task); + return; +out_exit: + rpc_exit(task, status); } static void @@ -2279,19 +2279,6 @@ rpc_check_timeout(struct rpc_task *task) rpcauth_invalcred(task); } -/* - * 6a. Handle RPC timeout - * We do not release the request slot, so we keep using the - * same XID for all retransmits. - */ -static void -call_timeout(struct rpc_task *task) -{ - task->tk_action = call_encode; - task->tk_status = 0; - rpc_check_timeout(task); -} - /* * 7. Decode the RPC reply */ @@ -2329,16 +2316,8 @@ call_decode(struct rpc_task *task) WARN_ON(memcmp(&req->rq_rcv_buf, &req->rq_private_buf, sizeof(req->rq_rcv_buf)) != 0); - if (req->rq_rcv_buf.len < 12) { - if (!RPC_IS_SOFT(task)) { - task->tk_action = call_encode; - goto out_retry; - } - dprintk("RPC: %s: too small RPC reply size (%d bytes)\n", - clnt->cl_program->name, task->tk_status); - task->tk_action = call_timeout; + if (req->rq_rcv_buf.len < 12) goto out_retry; - } xdr_init_decode(&xdr, &req->rq_rcv_buf, req->rq_rcv_buf.head[0].iov_base, req); @@ -2361,6 +2340,8 @@ out_retry: xprt_conditional_disconnect(req->rq_xprt, req->rq_connect_cookie); } + task->tk_action = call_encode; + rpc_check_timeout(task); } } From 6b5f590016870a989a94d61c984e3bfc7a37bfe7 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 9 Mar 2019 16:20:11 -0500 Subject: [PATCH 87/90] SUNRPC: Remove redundant calls to RPC_IS_QUEUED() The RPC task wakeup calls all check for RPC_IS_QUEUED() before taking any locks. In addition, rpc_exit() already calls rpc_wake_up_queued_task(). Signed-off-by: Trond Myklebust --- net/sunrpc/clnt.c | 3 --- net/sunrpc/sched.c | 3 +-- 2 files changed, 1 insertion(+), 5 deletions(-) diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index a0e7322fc171..9cf63e6339f4 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -833,9 +833,6 @@ void rpc_killall_tasks(struct rpc_clnt *clnt) if (!(rovr->tk_flags & RPC_TASK_KILLED)) { rovr->tk_flags |= RPC_TASK_KILLED; rpc_exit(rovr, -EIO); - if (RPC_IS_QUEUED(rovr)) - rpc_wake_up_queued_task(rovr->tk_waitqueue, - rovr); } } spin_unlock(&clnt->cl_lock); diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c index f21557213a43..28956c70100a 100644 --- a/net/sunrpc/sched.c +++ b/net/sunrpc/sched.c @@ -785,8 +785,7 @@ void rpc_exit(struct rpc_task *task, int status) { task->tk_status = status; task->tk_action = rpc_exit_task; - if (RPC_IS_QUEUED(task)) - rpc_wake_up_queued_task(task->tk_waitqueue, task); + rpc_wake_up_queued_task(task->tk_waitqueue, task); } EXPORT_SYMBOL_GPL(rpc_exit); From 03e51d32da995030a16697038232171807eeb0f2 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 10 Mar 2019 11:17:13 -0400 Subject: [PATCH 88/90] SUNRPC: Check whether the task was transmitted before rebind/reconnect Before initiating transport actions that require putting the task to sleep, such as rebinding or reconnecting, we should check whether or not the task was already transmitted. Signed-off-by: Trond Myklebust --- net/sunrpc/clnt.c | 44 ++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 40 insertions(+), 4 deletions(-) diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 9cf63e6339f4..67c955d8b21b 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1811,6 +1811,24 @@ out: task->tk_action = call_connect; } +/* + * Helpers to check if the task was already transmitted, and + * to take action when that is the case. + */ +static bool +rpc_task_transmitted(struct rpc_task *task) +{ + return !test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate); +} + +static void +rpc_task_handle_transmitted(struct rpc_task *task) +{ + xprt_end_transmit(task); + task->tk_action = call_transmit_status; + call_transmit_status(task); +} + /* * 4. Get the server port number if not yet set */ @@ -1819,6 +1837,11 @@ call_bind(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_rqstp->rq_xprt; + if (rpc_task_transmitted(task)) { + rpc_task_handle_transmitted(task); + return; + } + dprint_status(task); task->tk_action = call_connect; @@ -1837,6 +1860,11 @@ call_bind_status(struct rpc_task *task) { int status = -EIO; + if (rpc_task_transmitted(task)) { + rpc_task_handle_transmitted(task); + return; + } + if (task->tk_status >= 0) { dprint_status(task); task->tk_status = 0; @@ -1916,6 +1944,11 @@ call_connect(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_rqstp->rq_xprt; + if (rpc_task_transmitted(task)) { + rpc_task_handle_transmitted(task); + return; + } + dprintk("RPC: %5u call_connect xprt %p %s connected\n", task->tk_pid, xprt, (xprt_connected(xprt) ? "is" : "is not")); @@ -1942,10 +1975,8 @@ call_connect_status(struct rpc_task *task) struct rpc_clnt *clnt = task->tk_client; int status = task->tk_status; - /* Check if the task was already transmitted */ - if (!test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) { - xprt_end_transmit(task); - task->tk_action = call_transmit_status; + if (rpc_task_transmitted(task)) { + rpc_task_handle_transmitted(task); return; } @@ -2001,6 +2032,11 @@ out_retry: static void call_transmit(struct rpc_task *task) { + if (rpc_task_transmitted(task)) { + rpc_task_handle_transmitted(task); + return; + } + dprint_status(task); task->tk_action = call_transmit_status; From 009a82f6437490c262584d65a14094a818bcb747 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sat, 9 Mar 2019 12:07:17 -0500 Subject: [PATCH 89/90] SUNRPC: Micro-optimise when the task is known not to be sleeping In cases where we know the task is not sleeping, try to optimise away the indirect call to task->tk_action() by replacing it with a direct call. Only change tail calls, to allow gcc to perform tail call elimination. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/sched.h | 8 +++ net/sunrpc/clnt.c | 99 +++++++++++++++++++++++------------- 2 files changed, 73 insertions(+), 34 deletions(-) diff --git a/include/linux/sunrpc/sched.h b/include/linux/sunrpc/sched.h index 52d41d0c1ae1..ec861cd0cfe8 100644 --- a/include/linux/sunrpc/sched.h +++ b/include/linux/sunrpc/sched.h @@ -304,4 +304,12 @@ rpc_clnt_swap_deactivate(struct rpc_clnt *clnt) } #endif /* CONFIG_SUNRPC_SWAP */ +static inline bool +rpc_task_need_resched(const struct rpc_task *task) +{ + if (RPC_IS_QUEUED(task) || task->tk_callback) + return true; + return false; +} + #endif /* _LINUX_SUNRPC_SCHED_H_ */ diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 67c955d8b21b..498dd6ad5bc5 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1540,6 +1540,7 @@ call_start(struct rpc_task *task) clnt->cl_stats->rpccnt++; task->tk_action = call_reserve; rpc_task_set_transport(task, clnt); + call_reserve(task); } /* @@ -1553,6 +1554,9 @@ call_reserve(struct rpc_task *task) task->tk_status = 0; task->tk_action = call_reserveresult; xprt_reserve(task); + if (rpc_task_need_resched(task)) + return; + call_reserveresult(task); } static void call_retry_reserve(struct rpc_task *task); @@ -1575,6 +1579,7 @@ call_reserveresult(struct rpc_task *task) if (status >= 0) { if (task->tk_rqstp) { task->tk_action = call_refresh; + call_refresh(task); return; } @@ -1600,6 +1605,7 @@ call_reserveresult(struct rpc_task *task) /* fall through */ case -EAGAIN: /* woken up; retry */ task->tk_action = call_retry_reserve; + call_retry_reserve(task); return; case -EIO: /* probably a shutdown */ break; @@ -1622,6 +1628,9 @@ call_retry_reserve(struct rpc_task *task) task->tk_status = 0; task->tk_action = call_reserveresult; xprt_retry_reserve(task); + if (rpc_task_need_resched(task)) + return; + call_reserveresult(task); } /* @@ -1636,6 +1645,9 @@ call_refresh(struct rpc_task *task) task->tk_status = 0; task->tk_client->cl_stats->rpcauthrefresh++; rpcauth_refreshcred(task); + if (rpc_task_need_resched(task)) + return; + call_refreshresult(task); } /* @@ -1654,6 +1666,7 @@ call_refreshresult(struct rpc_task *task) case 0: if (rpcauth_uptodatecred(task)) { task->tk_action = call_allocate; + call_allocate(task); return; } /* Use rate-limiting and a max number of retries if refresh @@ -1672,6 +1685,7 @@ call_refreshresult(struct rpc_task *task) task->tk_cred_retry--; dprintk("RPC: %5u %s: retry refresh creds\n", task->tk_pid, __func__); + call_refresh(task); return; } dprintk("RPC: %5u %s: refresh creds failed with error %d\n", @@ -1697,8 +1711,10 @@ call_allocate(struct rpc_task *task) task->tk_status = 0; task->tk_action = call_encode; - if (req->rq_buffer) + if (req->rq_buffer) { + call_encode(task); return; + } if (proc->p_proc != 0) { BUG_ON(proc->p_arglen == 0); @@ -1719,8 +1735,12 @@ call_allocate(struct rpc_task *task) status = xprt->ops->buf_alloc(task); xprt_inject_disconnect(xprt); - if (status == 0) + if (status == 0) { + if (rpc_task_need_resched(task)) + return; + call_encode(task); return; + } if (status != -ENOMEM) { rpc_exit(task, status); return; @@ -1803,12 +1823,8 @@ call_encode(struct rpc_task *task) xprt_request_enqueue_receive(task); xprt_request_enqueue_transmit(task); out: - task->tk_action = call_transmit; - /* Check that the connection is OK */ - if (!xprt_bound(task->tk_xprt)) - task->tk_action = call_bind; - else if (!xprt_connected(task->tk_xprt)) - task->tk_action = call_connect; + task->tk_action = call_bind; + call_bind(task); } /* @@ -1842,14 +1858,17 @@ call_bind(struct rpc_task *task) return; } + if (xprt_bound(xprt)) { + task->tk_action = call_connect; + call_connect(task); + return; + } + dprint_status(task); - task->tk_action = call_connect; - if (!xprt_bound(xprt)) { - task->tk_action = call_bind_status; - task->tk_timeout = xprt->bind_timeout; - xprt->ops->rpcbind(task); - } + task->tk_action = call_bind_status; + task->tk_timeout = xprt->bind_timeout; + xprt->ops->rpcbind(task); } /* @@ -1869,6 +1888,7 @@ call_bind_status(struct rpc_task *task) dprint_status(task); task->tk_status = 0; task->tk_action = call_connect; + call_connect(task); return; } @@ -1949,21 +1969,24 @@ call_connect(struct rpc_task *task) return; } + if (xprt_connected(xprt)) { + task->tk_action = call_transmit; + call_transmit(task); + return; + } + dprintk("RPC: %5u call_connect xprt %p %s connected\n", task->tk_pid, xprt, (xprt_connected(xprt) ? "is" : "is not")); - task->tk_action = call_transmit; - if (!xprt_connected(xprt)) { - task->tk_action = call_connect_status; - if (task->tk_status < 0) - return; - if (task->tk_flags & RPC_TASK_NOCONNECT) { - rpc_exit(task, -ENOTCONN); - return; - } - xprt_connect(task); + task->tk_action = call_connect_status; + if (task->tk_status < 0) + return; + if (task->tk_flags & RPC_TASK_NOCONNECT) { + rpc_exit(task, -ENOTCONN); + return; } + xprt_connect(task); } /* @@ -2016,6 +2039,7 @@ call_connect_status(struct rpc_task *task) case 0: clnt->cl_stats->netreconn++; task->tk_action = call_transmit; + call_transmit(task); return; } rpc_exit(task, status); @@ -2040,19 +2064,20 @@ call_transmit(struct rpc_task *task) dprint_status(task); task->tk_action = call_transmit_status; + if (!xprt_prepare_transmit(task)) + return; + task->tk_status = 0; if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) { - if (!xprt_prepare_transmit(task)) + if (!xprt_connected(task->tk_xprt)) { + task->tk_status = -ENOTCONN; return; - task->tk_status = 0; - if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) { - if (!xprt_connected(task->tk_xprt)) { - task->tk_status = -ENOTCONN; - return; - } - xprt_transmit(task); } + xprt_transmit(task); } xprt_end_transmit(task); + if (rpc_task_need_resched(task)) + return; + call_transmit_status(task); } /* @@ -2067,8 +2092,12 @@ call_transmit_status(struct rpc_task *task) * Common case: success. Force the compiler to put this * test first. */ - if (task->tk_status == 0) { - xprt_request_wait_receive(task); + if (rpc_task_transmitted(task)) { + if (task->tk_status == 0) + xprt_request_wait_receive(task); + if (rpc_task_need_resched(task)) + return; + call_status(task); return; } @@ -2129,6 +2158,7 @@ call_bc_encode(struct rpc_task *task) { xprt_request_enqueue_transmit(task); task->tk_action = call_bc_transmit; + call_bc_transmit(task); } /* @@ -2219,6 +2249,7 @@ call_status(struct rpc_task *task) status = task->tk_status; if (status >= 0) { task->tk_action = call_decode; + call_decode(task); return; } From 4d6c671ace569d4b0d3f8d92ab3aef18a5d166bc Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Sun, 10 Mar 2019 12:22:39 -0400 Subject: [PATCH 90/90] SUNRPC: Take the transport send lock before binding+connecting Before trying to bind a port, ensure we grab the send lock to ensure that we don't change the port while another task is busy transmitting requests. The connect code already takes the send lock in xprt_connect(), but it is harmless to take it before that. Signed-off-by: Trond Myklebust --- net/sunrpc/clnt.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 498dd6ad5bc5..4216fe33204a 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1867,6 +1867,9 @@ call_bind(struct rpc_task *task) dprint_status(task); task->tk_action = call_bind_status; + if (!xprt_prepare_transmit(task)) + return; + task->tk_timeout = xprt->bind_timeout; xprt->ops->rpcbind(task); } @@ -1911,6 +1914,8 @@ call_bind_status(struct rpc_task *task) task->tk_rebind_retry--; rpc_delay(task, 3*HZ); goto retry_timeout; + case -EAGAIN: + goto retry_timeout; case -ETIMEDOUT: dprintk("RPC: %5u rpcbind request timed out\n", task->tk_pid); @@ -1952,7 +1957,7 @@ call_bind_status(struct rpc_task *task) retry_timeout: task->tk_status = 0; - task->tk_action = call_encode; + task->tk_action = call_bind; rpc_check_timeout(task); } @@ -1986,6 +1991,8 @@ call_connect(struct rpc_task *task) rpc_exit(task, -ENOTCONN); return; } + if (!xprt_prepare_transmit(task)) + return; xprt_connect(task); }