Merge tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-nextgrafted

Pull networking updates from Jakub Kicinski:
 "Core:

   - Add dedicated kmem_cache for typical/small skb->head, avoid having
     to access struct page at kfree time, and improve memory use.

   - Introduce sysctl to set default RPS configuration for new netdevs.

   - Define Netlink protocol specification format which can be used to
     describe messages used by each family and auto-generate parsers.
     Add tools for generating kernel data structures and uAPI headers.

   - Expose all net/core sysctls inside netns.

   - Remove 4s sleep in netpoll if carrier is instantly detected on
     boot.

   - Add configurable limit of MDB entries per port, and port-vlan.

   - Continue populating drop reasons throughout the stack.

   - Retire a handful of legacy Qdiscs and classifiers.

  Protocols:

   - Support IPv4 big TCP (TSO frames larger than 64kB).

   - Add IP_LOCAL_PORT_RANGE socket option, to control local port range
     on socket by socket basis.

   - Track and report in procfs number of MPTCP sockets used.

   - Support mixing IPv4 and IPv6 flows in the in-kernel MPTCP path
     manager.

   - IPv6: don't check net.ipv6.route.max_size and rely on garbage
     collection to free memory (similarly to IPv4).

   - Support Penultimate Segment Pop (PSP) flavor in SRv6 (RFC8986).

   - ICMP: add per-rate limit counters.

   - Add support for user scanning requests in ieee802154.

   - Remove static WEP support.

   - Support minimal Wi-Fi 7 Extremely High Throughput (EHT) rate
     reporting.

   - WiFi 7 EHT channel puncturing support (client & AP).

  BPF:

   - Add a rbtree data structure following the "next-gen data structure"
     precedent set by recently added linked list, that is, by using
     kfunc + kptr instead of adding a new BPF map type.

   - Expose XDP hints via kfuncs with initial support for RX hash and
     timestamp metadata.

   - Add BPF_F_NO_TUNNEL_KEY extension to bpf_skb_set_tunnel_key to
     better support decap on GRE tunnel devices not operating in collect
     metadata.

   - Improve x86 JIT's codegen for PROBE_MEM runtime error checks.

   - Remove the need for trace_printk_lock for bpf_trace_printk and
     bpf_trace_vprintk helpers.

   - Extend libbpf's bpf_tracing.h support for tracing arguments of
     kprobes/uprobes and syscall as a special case.

   - Significantly reduce the search time for module symbols by
     livepatch and BPF.

   - Enable cpumasks to be used as kptrs, which is useful for tracing
     programs tracking which tasks end up running on which CPUs in
     different time intervals.

   - Add support for BPF trampoline on s390x and riscv64.

   - Add capability to export the XDP features supported by the NIC.

   - Add __bpf_kfunc tag for marking kernel functions as kfuncs.

   - Add cgroup.memory=nobpf kernel parameter option to disable BPF
     memory accounting for container environments.

  Netfilter:

   - Remove the CLUSTERIP target. It has been marked as obsolete for
     years, and we still have WARN splats wrt races of the out-of-band
     /proc interface installed by this target.

   - Add 'destroy' commands to nf_tables. They are identical to the
     existing 'delete' commands, but do not return an error if the
     referenced object (set, chain, rule...) did not exist.

  Driver API:

   - Improve cpumask_local_spread() locality to help NICs set the right
     IRQ affinity on AMD platforms.

   - Separate C22 and C45 MDIO bus transactions more clearly.

   - Introduce new DCB table to control DSCP rewrite on egress.

   - Support configuration of Physical Layer Collision Avoidance (PLCA)
     Reconciliation Sublayer (RS) (802.3cg-2019). Modern version of
     shared medium Ethernet.

   - Support for MAC Merge layer (IEEE 802.3-2018 clause 99). Allowing
     preemption of low priority frames by high priority frames.

   - Add support for controlling MACSec offload using netlink SET.

   - Rework devlink instance refcounts to allow registration and
     de-registration under the instance lock. Split the code into
     multiple files, drop some of the unnecessarily granular locks and
     factor out common parts of netlink operation handling.

   - Add TX frame aggregation parameters (for USB drivers).

   - Add a new attr TCA_EXT_WARN_MSG to report TC (offload) warning
     messages with notifications for debug.

   - Allow offloading of UDP NEW connections via act_ct.

   - Add support for per action HW stats in TC.

   - Support hardware miss to TC action (continue processing in SW from
     a specific point in the action chain).

   - Warn if old Wireless Extension user space interface is used with
     modern cfg80211/mac80211 drivers. Do not support Wireless
     Extensions for Wi-Fi 7 devices at all. Everyone should switch to
     using nl80211 interface instead.

   - Improve the CAN bit timing configuration. Use extack to return
     error messages directly to user space, update the SJW handling,
     including the definition of a new default value that will benefit
     CAN-FD controllers, by increasing their oscillator tolerance.

  New hardware / drivers:

   - Ethernet:
      - nVidia BlueField-3 support (control traffic driver)
      - Ethernet support for imx93 SoCs
      - Motorcomm yt8531 gigabit Ethernet PHY
      - onsemi NCN26000 10BASE-T1S PHY (with support for PLCA)
      - Microchip LAN8841 PHY (incl. cable diagnostics and PTP)
      - Amlogic gxl MDIO mux

   - WiFi:
      - RealTek RTL8188EU (rtl8xxxu)
      - Qualcomm Wi-Fi 7 devices (ath12k)

   - CAN:
      - Renesas R-Car V4H

  Drivers:

   - Bluetooth:
      - Set Per Platform Antenna Gain (PPAG) for Intel controllers.

   - Ethernet NICs:
      - Intel (1G, igc):
         - support TSN / Qbv / packet scheduling features of i226 model
      - Intel (100G, ice):
         - use GNSS subsystem instead of TTY
         - multi-buffer XDP support
         - extend support for GPIO pins to E823 devices
      - nVidia/Mellanox:
         - update the shared buffer configuration on PFC commands
         - implement PTP adjphase function for HW offset control
         - TC support for Geneve and GRE with VF tunnel offload
         - more efficient crypto key management method
         - multi-port eswitch support
      - Netronome/Corigine:
         - add DCB IEEE support
         - support IPsec offloading for NFP3800
      - Freescale/NXP (enetc):
         - support XDP_REDIRECT for XDP non-linear buffers
         - improve reconfig, avoid link flap and waiting for idle
         - support MAC Merge layer
      - Other NICs:
         - sfc/ef100: add basic devlink support for ef100
         - ionic: rx_push mode operation (writing descriptors via MMIO)
         - bnxt: use the auxiliary bus abstraction for RDMA
         - r8169: disable ASPM and reset bus in case of tx timeout
         - cpsw: support QSGMII mode for J721e CPSW9G
         - cpts: support pulse-per-second output
         - ngbe: add an mdio bus driver
         - usbnet: optimize usbnet_bh() by avoiding unnecessary queuing
         - r8152: handle devices with FW with NCM support
         - amd-xgbe: support 10Mbps, 2.5GbE speeds and rx-adaptation
         - virtio-net: support multi buffer XDP
         - virtio/vsock: replace virtio_vsock_pkt with sk_buff
         - tsnep: XDP support

   - Ethernet high-speed switches:
      - nVidia/Mellanox (mlxsw):
         - add support for latency TLV (in FW control messages)
      - Microchip (sparx5):
         - separate explicit and implicit traffic forwarding rules, make
           the implicit rules always active
         - add support for egress DSCP rewrite
         - IS0 VCAP support (Ingress Classification)
         - IS2 VCAP filters (protos, L3 addrs, L4 ports, flags, ToS
           etc.)
         - ES2 VCAP support (Egress Access Control)
         - support for Per-Stream Filtering and Policing (802.1Q,
           8.6.5.1)

   - Ethernet embedded switches:
      - Marvell (mv88e6xxx):
         - add MAB (port auth) offload support
         - enable PTP receive for mv88e6390
      - NXP (ocelot):
         - support MAC Merge layer
         - support for the the vsc7512 internal copper phys
      - Microchip:
         - lan9303: convert to PHYLINK
         - lan966x: support TC flower filter statistics
         - lan937x: PTP support for KSZ9563/KSZ8563 and LAN937x
         - lan937x: support Credit Based Shaper configuration
         - ksz9477: support Energy Efficient Ethernet
      - other:
         - qca8k: convert to regmap read/write API, use bulk operations
         - rswitch: Improve TX timestamp accuracy

   - Intel WiFi (iwlwifi):
      - EHT (Wi-Fi 7) rate reporting
      - STEP equalizer support: transfer some STEP (connection to radio
        on platforms with integrated wifi) related parameters from the
        BIOS to the firmware.

   - Qualcomm 802.11ax WiFi (ath11k):
      - IPQ5018 support
      - Fine Timing Measurement (FTM) responder role support
      - channel 177 support

   - MediaTek WiFi (mt76):
      - per-PHY LED support
      - mt7996: EHT (Wi-Fi 7) support
      - Wireless Ethernet Dispatch (WED) reset support
      - switch to using page pool allocator

   - RealTek WiFi (rtw89):
      - support new version of Bluetooth co-existance

   - Mobile:
      - rmnet: support TX aggregation"

* tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1872 commits)
  page_pool: add a comment explaining the fragment counter usage
  net: ethtool: fix __ethtool_dev_mm_supported() implementation
  ethtool: pse-pd: Fix double word in comments
  xsk: add linux/vmalloc.h to xsk.c
  sefltests: netdevsim: wait for devlink instance after netns removal
  selftest: fib_tests: Always cleanup before exit
  net/mlx5e: Align IPsec ASO result memory to be as required by hardware
  net/mlx5e: TC, Set CT miss to the specific ct action instance
  net/mlx5e: Rename CHAIN_TO_REG to MAPPED_OBJ_TO_REG
  net/mlx5: Refactor tc miss handling to a single function
  net/mlx5: Kconfig: Make tc offload depend on tc skb extension
  net/sched: flower: Support hardware miss to tc action
  net/sched: flower: Move filter handle initialization earlier
  net/sched: cls_api: Support hardware miss to tc action
  net/sched: Rename user cookie and act cookie
  sfc: fix builds without CONFIG_RTC_LIB
  sfc: clean up some inconsistent indentings
  net/mlx4_en: Introduce flexible array to silence overflow warning
  net: lan966x: Fix possible deadlock inside PTP
  net/ulp: Remove redundant ->clone() test in inet_clone_ulp().
  ...

1 files changed, 1165 insertions, 0 deletions

diff --git a/net/sunrpc/xprtrdma/svc_rdma_rw.c b/net/sunrpc/xprtrdma/svc_rdma_rw.c
new file mode 100644
index 000000000..11cf7c646
--- /dev/null
+++ b/net/sunrpc/xprtrdma/svc_rdma_rw.c
@@ -0,0 +1,1165 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2016-2018 Oracle.  All rights reserved.
+ *
+ * Use the core R/W API to move RPC-over-RDMA Read and Write chunks.
+ */
+
+#include <rdma/rw.h>
+
+#include <linux/sunrpc/xdr.h>
+#include <linux/sunrpc/rpc_rdma.h>
+#include <linux/sunrpc/svc_rdma.h>
+
+#include "xprt_rdma.h"
+#include <trace/events/rpcrdma.h>
+
+static void svc_rdma_write_done(struct ib_cq *cq, struct ib_wc *wc);
+static void svc_rdma_wc_read_done(struct ib_cq *cq, struct ib_wc *wc);
+
+/* Each R/W context contains state for one chain of RDMA Read or
+ * Write Work Requests.
+ *
+ * Each WR chain handles a single contiguous server-side buffer,
+ * because scatterlist entries after the first have to start on
+ * page alignment. xdr_buf iovecs cannot guarantee alignment.
+ *
+ * Each WR chain handles only one R_key. Each RPC-over-RDMA segment
+ * from a client may contain a unique R_key, so each WR chain moves
+ * up to one segment at a time.
+ *
+ * The scatterlist makes this data structure over 4KB in size. To
+ * make it less likely to fail, and to handle the allocation for
+ * smaller I/O requests without disabling bottom-halves, these
+ * contexts are created on demand, but cached and reused until the
+ * controlling svcxprt_rdma is destroyed.
+ */
+struct svc_rdma_rw_ctxt {
+	struct llist_node	rw_node;
+	struct list_head	rw_list;
+	struct rdma_rw_ctx	rw_ctx;
+	unsigned int		rw_nents;
+	struct sg_table		rw_sg_table;
+	struct scatterlist	rw_first_sgl[];
+};
+
+static inline struct svc_rdma_rw_ctxt *
+svc_rdma_next_ctxt(struct list_head *list)
+{
+	return list_first_entry_or_null(list, struct svc_rdma_rw_ctxt,
+					rw_list);
+}
+
+static struct svc_rdma_rw_ctxt *
+svc_rdma_get_rw_ctxt(struct svcxprt_rdma *rdma, unsigned int sges)
+{
+	struct svc_rdma_rw_ctxt *ctxt;
+	struct llist_node *node;
+
+	spin_lock(&rdma->sc_rw_ctxt_lock);
+	node = llist_del_first(&rdma->sc_rw_ctxts);
+	spin_unlock(&rdma->sc_rw_ctxt_lock);
+	if (node) {
+		ctxt = llist_entry(node, struct svc_rdma_rw_ctxt, rw_node);
+	} else {
+		ctxt = kmalloc(struct_size(ctxt, rw_first_sgl, SG_CHUNK_SIZE),
+			       GFP_KERNEL);
+		if (!ctxt)
+			goto out_noctx;
+
+		INIT_LIST_HEAD(&ctxt->rw_list);
+	}
+
+	ctxt->rw_sg_table.sgl = ctxt->rw_first_sgl;
+	if (sg_alloc_table_chained(&ctxt->rw_sg_table, sges,
+				   ctxt->rw_sg_table.sgl,
+				   SG_CHUNK_SIZE))
+		goto out_free;
+	return ctxt;
+
+out_free:
+	kfree(ctxt);
+out_noctx:
+	trace_svcrdma_no_rwctx_err(rdma, sges);
+	return NULL;
+}
+
+static void __svc_rdma_put_rw_ctxt(struct svcxprt_rdma *rdma,
+				   struct svc_rdma_rw_ctxt *ctxt,
+				   struct llist_head *list)
+{
+	sg_free_table_chained(&ctxt->rw_sg_table, SG_CHUNK_SIZE);
+	llist_add(&ctxt->rw_node, list);
+}
+
+static void svc_rdma_put_rw_ctxt(struct svcxprt_rdma *rdma,
+				 struct svc_rdma_rw_ctxt *ctxt)
+{
+	__svc_rdma_put_rw_ctxt(rdma, ctxt, &rdma->sc_rw_ctxts);
+}
+
+/**
+ * svc_rdma_destroy_rw_ctxts - Free accumulated R/W contexts
+ * @rdma: transport about to be destroyed
+ *
+ */
+void svc_rdma_destroy_rw_ctxts(struct svcxprt_rdma *rdma)
+{
+	struct svc_rdma_rw_ctxt *ctxt;
+	struct llist_node *node;
+
+	while ((node = llist_del_first(&rdma->sc_rw_ctxts)) != NULL) {
+		ctxt = llist_entry(node, struct svc_rdma_rw_ctxt, rw_node);
+		kfree(ctxt);
+	}
+}
+
+/**
+ * svc_rdma_rw_ctx_init - Prepare a R/W context for I/O
+ * @rdma: controlling transport instance
+ * @ctxt: R/W context to prepare
+ * @offset: RDMA offset
+ * @handle: RDMA tag/handle
+ * @direction: I/O direction
+ *
+ * Returns on success, the number of WQEs that will be needed
+ * on the workqueue, or a negative errno.
+ */
+static int svc_rdma_rw_ctx_init(struct svcxprt_rdma *rdma,
+				struct svc_rdma_rw_ctxt *ctxt,
+				u64 offset, u32 handle,
+				enum dma_data_direction direction)
+{
+	int ret;
+
+	ret = rdma_rw_ctx_init(&ctxt->rw_ctx, rdma->sc_qp, rdma->sc_port_num,
+			       ctxt->rw_sg_table.sgl, ctxt->rw_nents,
+			       0, offset, handle, direction);
+	if (unlikely(ret < 0)) {
+		svc_rdma_put_rw_ctxt(rdma, ctxt);
+		trace_svcrdma_dma_map_rw_err(rdma, ctxt->rw_nents, ret);
+	}
+	return ret;
+}
+
+/* A chunk context tracks all I/O for moving one Read or Write
+ * chunk. This is a set of rdma_rw's that handle data movement
+ * for all segments of one chunk.
+ *
+ * These are small, acquired with a single allocator call, and
+ * no more than one is needed per chunk. They are allocated on
+ * demand, and not cached.
+ */
+struct svc_rdma_chunk_ctxt {
+	struct rpc_rdma_cid	cc_cid;
+	struct ib_cqe		cc_cqe;
+	struct svcxprt_rdma	*cc_rdma;
+	struct list_head	cc_rwctxts;
+	ktime_t			cc_posttime;
+	int			cc_sqecount;
+	enum ib_wc_status	cc_status;
+	struct completion	cc_done;
+};
+
+static void svc_rdma_cc_cid_init(struct svcxprt_rdma *rdma,
+				 struct rpc_rdma_cid *cid)
+{
+	cid->ci_queue_id = rdma->sc_sq_cq->res.id;
+	cid->ci_completion_id = atomic_inc_return(&rdma->sc_completion_ids);
+}
+
+static void svc_rdma_cc_init(struct svcxprt_rdma *rdma,
+			     struct svc_rdma_chunk_ctxt *cc)
+{
+	svc_rdma_cc_cid_init(rdma, &cc->cc_cid);
+	cc->cc_rdma = rdma;
+
+	INIT_LIST_HEAD(&cc->cc_rwctxts);
+	cc->cc_sqecount = 0;
+}
+
+/*
+ * The consumed rw_ctx's are cleaned and placed on a local llist so
+ * that only one atomic llist operation is needed to put them all
+ * back on the free list.
+ */
+static void svc_rdma_cc_release(struct svc_rdma_chunk_ctxt *cc,
+				enum dma_data_direction dir)
+{
+	struct svcxprt_rdma *rdma = cc->cc_rdma;
+	struct llist_node *first, *last;
+	struct svc_rdma_rw_ctxt *ctxt;
+	LLIST_HEAD(free);
+
+	first = last = NULL;
+	while ((ctxt = svc_rdma_next_ctxt(&cc->cc_rwctxts)) != NULL) {
+		list_del(&ctxt->rw_list);
+
+		rdma_rw_ctx_destroy(&ctxt->rw_ctx, rdma->sc_qp,
+				    rdma->sc_port_num, ctxt->rw_sg_table.sgl,
+				    ctxt->rw_nents, dir);
+		__svc_rdma_put_rw_ctxt(rdma, ctxt, &free);
+
+		ctxt->rw_node.next = first;
+		first = &ctxt->rw_node;
+		if (!last)
+			last = first;
+	}
+	if (first)
+		llist_add_batch(first, last, &rdma->sc_rw_ctxts);
+}
+
+/* State for sending a Write or Reply chunk.
+ *  - Tracks progress of writing one chunk over all its segments
+ *  - Stores arguments for the SGL constructor functions
+ */
+struct svc_rdma_write_info {
+	const struct svc_rdma_chunk	*wi_chunk;
+
+	/* write state of this chunk */
+	unsigned int		wi_seg_off;
+	unsigned int		wi_seg_no;
+
+	/* SGL constructor arguments */
+	const struct xdr_buf	*wi_xdr;
+	unsigned char		*wi_base;
+	unsigned int		wi_next_off;
+
+	struct svc_rdma_chunk_ctxt	wi_cc;
+};
+
+static struct svc_rdma_write_info *
+svc_rdma_write_info_alloc(struct svcxprt_rdma *rdma,
+			  const struct svc_rdma_chunk *chunk)
+{
+	struct svc_rdma_write_info *info;
+
+	info = kmalloc(sizeof(*info), GFP_KERNEL);
+	if (!info)
+		return info;
+
+	info->wi_chunk = chunk;
+	info->wi_seg_off = 0;
+	info->wi_seg_no = 0;
+	svc_rdma_cc_init(rdma, &info->wi_cc);
+	info->wi_cc.cc_cqe.done = svc_rdma_write_done;
+	return info;
+}
+
+static void svc_rdma_write_info_free(struct svc_rdma_write_info *info)
+{
+	svc_rdma_cc_release(&info->wi_cc, DMA_TO_DEVICE);
+	kfree(info);
+}
+
+/**
+ * svc_rdma_write_done - Write chunk completion
+ * @cq: controlling Completion Queue
+ * @wc: Work Completion
+ *
+ * Pages under I/O are freed by a subsequent Send completion.
+ */
+static void svc_rdma_write_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+	struct ib_cqe *cqe = wc->wr_cqe;
+	struct svc_rdma_chunk_ctxt *cc =
+			container_of(cqe, struct svc_rdma_chunk_ctxt, cc_cqe);
+	struct svcxprt_rdma *rdma = cc->cc_rdma;
+	struct svc_rdma_write_info *info =
+			container_of(cc, struct svc_rdma_write_info, wi_cc);
+
+	switch (wc->status) {
+	case IB_WC_SUCCESS:
+		trace_svcrdma_wc_write(wc, &cc->cc_cid);
+		break;
+	case IB_WC_WR_FLUSH_ERR:
+		trace_svcrdma_wc_write_flush(wc, &cc->cc_cid);
+		break;
+	default:
+		trace_svcrdma_wc_write_err(wc, &cc->cc_cid);
+	}
+
+	svc_rdma_wake_send_waiters(rdma, cc->cc_sqecount);
+
+	if (unlikely(wc->status != IB_WC_SUCCESS))
+		svc_xprt_deferred_close(&rdma->sc_xprt);
+
+	svc_rdma_write_info_free(info);
+}
+
+/* State for pulling a Read chunk.
+ */
+struct svc_rdma_read_info {
+	struct svc_rqst			*ri_rqst;
+	struct svc_rdma_recv_ctxt	*ri_readctxt;
+	unsigned int			ri_pageno;
+	unsigned int			ri_pageoff;
+	unsigned int			ri_totalbytes;
+
+	struct svc_rdma_chunk_ctxt	ri_cc;
+};
+
+static struct svc_rdma_read_info *
+svc_rdma_read_info_alloc(struct svcxprt_rdma *rdma)
+{
+	struct svc_rdma_read_info *info;
+
+	info = kmalloc(sizeof(*info), GFP_KERNEL);
+	if (!info)
+		return info;
+
+	svc_rdma_cc_init(rdma, &info->ri_cc);
+	info->ri_cc.cc_cqe.done = svc_rdma_wc_read_done;
+	return info;
+}
+
+static void svc_rdma_read_info_free(struct svc_rdma_read_info *info)
+{
+	svc_rdma_cc_release(&info->ri_cc, DMA_FROM_DEVICE);
+	kfree(info);
+}
+
+/**
+ * svc_rdma_wc_read_done - Handle completion of an RDMA Read ctx
+ * @cq: controlling Completion Queue
+ * @wc: Work Completion
+ *
+ */
+static void svc_rdma_wc_read_done(struct ib_cq *cq, struct ib_wc *wc)
+{
+	struct ib_cqe *cqe = wc->wr_cqe;
+	struct svc_rdma_chunk_ctxt *cc =
+			container_of(cqe, struct svc_rdma_chunk_ctxt, cc_cqe);
+	struct svc_rdma_read_info *info;
+
+	switch (wc->status) {
+	case IB_WC_SUCCESS:
+		info = container_of(cc, struct svc_rdma_read_info, ri_cc);
+		trace_svcrdma_wc_read(wc, &cc->cc_cid, info->ri_totalbytes,
+				      cc->cc_posttime);
+		break;
+	case IB_WC_WR_FLUSH_ERR:
+		trace_svcrdma_wc_read_flush(wc, &cc->cc_cid);
+		break;
+	default:
+		trace_svcrdma_wc_read_err(wc, &cc->cc_cid);
+	}
+
+	svc_rdma_wake_send_waiters(cc->cc_rdma, cc->cc_sqecount);
+	cc->cc_status = wc->status;
+	complete(&cc->cc_done);
+	return;
+}
+
+/* This function sleeps when the transport's Send Queue is congested.
+ *
+ * Assumptions:
+ * - If ib_post_send() succeeds, only one completion is expected,
+ *   even if one or more WRs are flushed. This is true when posting
+ *   an rdma_rw_ctx or when posting a single signaled WR.
+ */
+static int svc_rdma_post_chunk_ctxt(struct svc_rdma_chunk_ctxt *cc)
+{
+	struct svcxprt_rdma *rdma = cc->cc_rdma;
+	struct ib_send_wr *first_wr;
+	const struct ib_send_wr *bad_wr;
+	struct list_head *tmp;
+	struct ib_cqe *cqe;
+	int ret;
+
+	if (cc->cc_sqecount > rdma->sc_sq_depth)
+		return -EINVAL;
+
+	first_wr = NULL;
+	cqe = &cc->cc_cqe;
+	list_for_each(tmp, &cc->cc_rwctxts) {
+		struct svc_rdma_rw_ctxt *ctxt;
+
+		ctxt = list_entry(tmp, struct svc_rdma_rw_ctxt, rw_list);
+		first_wr = rdma_rw_ctx_wrs(&ctxt->rw_ctx, rdma->sc_qp,
+					   rdma->sc_port_num, cqe, first_wr);
+		cqe = NULL;
+	}
+
+	do {
+		if (atomic_sub_return(cc->cc_sqecount,
+				      &rdma->sc_sq_avail) > 0) {
+			cc->cc_posttime = ktime_get();
+			ret = ib_post_send(rdma->sc_qp, first_wr, &bad_wr);
+			if (ret)
+				break;
+			return 0;
+		}
+
+		percpu_counter_inc(&svcrdma_stat_sq_starve);
+		trace_svcrdma_sq_full(rdma);
+		atomic_add(cc->cc_sqecount, &rdma->sc_sq_avail);
+		wait_event(rdma->sc_send_wait,
+			   atomic_read(&rdma->sc_sq_avail) > cc->cc_sqecount);
+		trace_svcrdma_sq_retry(rdma);
+	} while (1);
+
+	trace_svcrdma_sq_post_err(rdma, ret);
+	svc_xprt_deferred_close(&rdma->sc_xprt);
+
+	/* If even one was posted, there will be a completion. */
+	if (bad_wr != first_wr)
+		return 0;
+
+	atomic_add(cc->cc_sqecount, &rdma->sc_sq_avail);
+	wake_up(&rdma->sc_send_wait);
+	return -ENOTCONN;
+}
+
+/* Build and DMA-map an SGL that covers one kvec in an xdr_buf
+ */
+static void svc_rdma_vec_to_sg(struct svc_rdma_write_info *info,
+			       unsigned int len,
+			       struct svc_rdma_rw_ctxt *ctxt)
+{
+	struct scatterlist *sg = ctxt->rw_sg_table.sgl;
+
+	sg_set_buf(&sg[0], info->wi_base, len);
+	info->wi_base += len;
+
+	ctxt->rw_nents = 1;
+}
+
+/* Build and DMA-map an SGL that covers part of an xdr_buf's pagelist.
+ */
+static void svc_rdma_pagelist_to_sg(struct svc_rdma_write_info *info,
+				    unsigned int remaining,
+				    struct svc_rdma_rw_ctxt *ctxt)
+{
+	unsigned int sge_no, sge_bytes, page_off, page_no;
+	const struct xdr_buf *xdr = info->wi_xdr;
+	struct scatterlist *sg;
+	struct page **page;
+
+	page_off = info->wi_next_off + xdr->page_base;
+	page_no = page_off >> PAGE_SHIFT;
+	page_off = offset_in_page(page_off);
+	page = xdr->pages + page_no;
+	info->wi_next_off += remaining;
+	sg = ctxt->rw_sg_table.sgl;
+	sge_no = 0;
+	do {
+		sge_bytes = min_t(unsigned int, remaining,
+				  PAGE_SIZE - page_off);
+		sg_set_page(sg, *page, sge_bytes, page_off);
+
+		remaining -= sge_bytes;
+		sg = sg_next(sg);
+		page_off = 0;
+		sge_no++;
+		page++;
+	} while (remaining);
+
+	ctxt->rw_nents = sge_no;
+}
+
+/* Construct RDMA Write WRs to send a portion of an xdr_buf containing
+ * an RPC Reply.
+ */
+static int
+svc_rdma_build_writes(struct svc_rdma_write_info *info,
+		      void (*constructor)(struct svc_rdma_write_info *info,
+					  unsigned int len,
+					  struct svc_rdma_rw_ctxt *ctxt),
+		      unsigned int remaining)
+{
+	struct svc_rdma_chunk_ctxt *cc = &info->wi_cc;
+	struct svcxprt_rdma *rdma = cc->cc_rdma;
+	const struct svc_rdma_segment *seg;
+	struct svc_rdma_rw_ctxt *ctxt;
+	int ret;
+
+	do {
+		unsigned int write_len;
+		u64 offset;
+
+		if (info->wi_seg_no >= info->wi_chunk->ch_segcount)
+			goto out_overflow;
+
+		seg = &info->wi_chunk->ch_segments[info->wi_seg_no];
+		write_len = min(remaining, seg->rs_length - info->wi_seg_off);
+		if (!write_len)
+			goto out_overflow;
+		ctxt = svc_rdma_get_rw_ctxt(rdma,
+					    (write_len >> PAGE_SHIFT) + 2);
+		if (!ctxt)
+			return -ENOMEM;
+
+		constructor(info, write_len, ctxt);
+		offset = seg->rs_offset + info->wi_seg_off;
+		ret = svc_rdma_rw_ctx_init(rdma, ctxt, offset, seg->rs_handle,
+					   DMA_TO_DEVICE);
+		if (ret < 0)
+			return -EIO;
+		percpu_counter_inc(&svcrdma_stat_write);
+
+		list_add(&ctxt->rw_list, &cc->cc_rwctxts);
+		cc->cc_sqecount += ret;
+		if (write_len == seg->rs_length - info->wi_seg_off) {
+			info->wi_seg_no++;
+			info->wi_seg_off = 0;
+		} else {
+			info->wi_seg_off += write_len;
+		}
+		remaining -= write_len;
+	} while (remaining);
+
+	return 0;
+
+out_overflow:
+	trace_svcrdma_small_wrch_err(rdma, remaining, info->wi_seg_no,
+				     info->wi_chunk->ch_segcount);
+	return -E2BIG;
+}
+
+/**
+ * svc_rdma_iov_write - Construct RDMA Writes from an iov
+ * @info: pointer to write arguments
+ * @iov: kvec to write
+ *
+ * Returns:
+ *   On success, returns zero
+ *   %-E2BIG if the client-provided Write chunk is too small
+ *   %-ENOMEM if a resource has been exhausted
+ *   %-EIO if an rdma-rw error occurred
+ */
+static int svc_rdma_iov_write(struct svc_rdma_write_info *info,
+			      const struct kvec *iov)
+{
+	info->wi_base = iov->iov_base;
+	return svc_rdma_build_writes(info, svc_rdma_vec_to_sg,
+				     iov->iov_len);
+}
+
+/**
+ * svc_rdma_pages_write - Construct RDMA Writes from pages
+ * @info: pointer to write arguments
+ * @xdr: xdr_buf with pages to write
+ * @offset: offset into the content of @xdr
+ * @length: number of bytes to write
+ *
+ * Returns:
+ *   On success, returns zero
+ *   %-E2BIG if the client-provided Write chunk is too small
+ *   %-ENOMEM if a resource has been exhausted
+ *   %-EIO if an rdma-rw error occurred
+ */
+static int svc_rdma_pages_write(struct svc_rdma_write_info *info,
+				const struct xdr_buf *xdr,
+				unsigned int offset,
+				unsigned long length)
+{
+	info->wi_xdr = xdr;
+	info->wi_next_off = offset - xdr->head[0].iov_len;
+	return svc_rdma_build_writes(info, svc_rdma_pagelist_to_sg,
+				     length);
+}
+
+/**
+ * svc_rdma_xb_write - Construct RDMA Writes to write an xdr_buf
+ * @xdr: xdr_buf to write
+ * @data: pointer to write arguments
+ *
+ * Returns:
+ *   On success, returns zero
+ *   %-E2BIG if the client-provided Write chunk is too small
+ *   %-ENOMEM if a resource has been exhausted
+ *   %-EIO if an rdma-rw error occurred
+ */
+static int svc_rdma_xb_write(const struct xdr_buf *xdr, void *data)
+{
+	struct svc_rdma_write_info *info = data;
+	int ret;
+
+	if (xdr->head[0].iov_len) {
+		ret = svc_rdma_iov_write(info, &xdr->head[0]);
+		if (ret < 0)
+			return ret;
+	}
+
+	if (xdr->page_len) {
+		ret = svc_rdma_pages_write(info, xdr, xdr->head[0].iov_len,
+					   xdr->page_len);
+		if (ret < 0)
+			return ret;
+	}
+
+	if (xdr->tail[0].iov_len) {
+		ret = svc_rdma_iov_write(info, &xdr->tail[0]);
+		if (ret < 0)
+			return ret;
+	}
+
+	return xdr->len;
+}
+
+/**
+ * svc_rdma_send_write_chunk - Write all segments in a Write chunk
+ * @rdma: controlling RDMA transport
+ * @chunk: Write chunk provided by the client
+ * @xdr: xdr_buf containing the data payload
+ *
+ * Returns a non-negative number of bytes the chunk consumed, or
+ *	%-E2BIG if the payload was larger than the Write chunk,
+ *	%-EINVAL if client provided too many segments,
+ *	%-ENOMEM if rdma_rw context pool was exhausted,
+ *	%-ENOTCONN if posting failed (connection is lost),
+ *	%-EIO if rdma_rw initialization failed (DMA mapping, etc).
+ */
+int svc_rdma_send_write_chunk(struct svcxprt_rdma *rdma,
+			      const struct svc_rdma_chunk *chunk,
+			      const struct xdr_buf *xdr)
+{
+	struct svc_rdma_write_info *info;
+	struct svc_rdma_chunk_ctxt *cc;
+	int ret;
+
+	info = svc_rdma_write_info_alloc(rdma, chunk);
+	if (!info)
+		return -ENOMEM;
+	cc = &info->wi_cc;
+
+	ret = svc_rdma_xb_write(xdr, info);
+	if (ret != xdr->len)
+		goto out_err;
+
+	trace_svcrdma_post_write_chunk(&cc->cc_cid, cc->cc_sqecount);
+	ret = svc_rdma_post_chunk_ctxt(cc);
+	if (ret < 0)
+		goto out_err;
+	return xdr->len;
+
+out_err:
+	svc_rdma_write_info_free(info);
+	return ret;
+}
+
+/**
+ * svc_rdma_send_reply_chunk - Write all segments in the Reply chunk
+ * @rdma: controlling RDMA transport
+ * @rctxt: Write and Reply chunks from client
+ * @xdr: xdr_buf containing an RPC Reply
+ *
+ * Returns a non-negative number of bytes the chunk consumed, or
+ *	%-E2BIG if the payload was larger than the Reply chunk,
+ *	%-EINVAL if client provided too many segments,
+ *	%-ENOMEM if rdma_rw context pool was exhausted,
+ *	%-ENOTCONN if posting failed (connection is lost),
+ *	%-EIO if rdma_rw initialization failed (DMA mapping, etc).
+ */
+int svc_rdma_send_reply_chunk(struct svcxprt_rdma *rdma,
+			      const struct svc_rdma_recv_ctxt *rctxt,
+			      const struct xdr_buf *xdr)
+{
+	struct svc_rdma_write_info *info;
+	struct svc_rdma_chunk_ctxt *cc;
+	struct svc_rdma_chunk *chunk;
+	int ret;
+
+	if (pcl_is_empty(&rctxt->rc_reply_pcl))
+		return 0;
+
+	chunk = pcl_first_chunk(&rctxt->rc_reply_pcl);
+	info = svc_rdma_write_info_alloc(rdma, chunk);
+	if (!info)
+		return -ENOMEM;
+	cc = &info->wi_cc;
+
+	ret = pcl_process_nonpayloads(&rctxt->rc_write_pcl, xdr,
+				      svc_rdma_xb_write, info);
+	if (ret < 0)
+		goto out_err;
+
+	trace_svcrdma_post_reply_chunk(&cc->cc_cid, cc->cc_sqecount);
+	ret = svc_rdma_post_chunk_ctxt(cc);
+	if (ret < 0)
+		goto out_err;
+
+	return xdr->len;
+
+out_err:
+	svc_rdma_write_info_free(info);
+	return ret;
+}
+
+/**
+ * svc_rdma_build_read_segment - Build RDMA Read WQEs to pull one RDMA segment
+ * @info: context for ongoing I/O
+ * @segment: co-ordinates of remote memory to be read
+ *
+ * Returns:
+ *   %0: the Read WR chain was constructed successfully
+ *   %-EINVAL: there were not enough rq_pages to finish
+ *   %-ENOMEM: allocating a local resources failed
+ *   %-EIO: a DMA mapping error occurred
+ */
+static int svc_rdma_build_read_segment(struct svc_rdma_read_info *info,
+				       const struct svc_rdma_segment *segment)
+{
+	struct svc_rdma_recv_ctxt *head = info->ri_readctxt;
+	struct svc_rdma_chunk_ctxt *cc = &info->ri_cc;
+	struct svc_rqst *rqstp = info->ri_rqst;
+	unsigned int sge_no, seg_len, len;
+	struct svc_rdma_rw_ctxt *ctxt;
+	struct scatterlist *sg;
+	int ret;
+
+	len = segment->rs_length;
+	sge_no = PAGE_ALIGN(info->ri_pageoff + len) >> PAGE_SHIFT;
+	ctxt = svc_rdma_get_rw_ctxt(cc->cc_rdma, sge_no);
+	if (!ctxt)
+		return -ENOMEM;
+	ctxt->rw_nents = sge_no;
+
+	sg = ctxt->rw_sg_table.sgl;
+	for (sge_no = 0; sge_no < ctxt->rw_nents; sge_no++) {
+		seg_len = min_t(unsigned int, len,
+				PAGE_SIZE - info->ri_pageoff);
+
+		if (!info->ri_pageoff)
+			head->rc_page_count++;
+
+		sg_set_page(sg, rqstp->rq_pages[info->ri_pageno],
+			    seg_len, info->ri_pageoff);
+		sg = sg_next(sg);
+
+		info->ri_pageoff += seg_len;
+		if (info->ri_pageoff == PAGE_SIZE) {
+			info->ri_pageno++;
+			info->ri_pageoff = 0;
+		}
+		len -= seg_len;
+
+		/* Safety check */
+		if (len &&
+		    &rqstp->rq_pages[info->ri_pageno + 1] > rqstp->rq_page_end)
+			goto out_overrun;
+	}
+
+	ret = svc_rdma_rw_ctx_init(cc->cc_rdma, ctxt, segment->rs_offset,
+				   segment->rs_handle, DMA_FROM_DEVICE);
+	if (ret < 0)
+		return -EIO;
+	percpu_counter_inc(&svcrdma_stat_read);
+
+	list_add(&ctxt->rw_list, &cc->cc_rwctxts);
+	cc->cc_sqecount += ret;
+	return 0;
+
+out_overrun:
+	trace_svcrdma_page_overrun_err(cc->cc_rdma, rqstp, info->ri_pageno);
+	return -EINVAL;
+}
+
+/**
+ * svc_rdma_build_read_chunk - Build RDMA Read WQEs to pull one RDMA chunk
+ * @info: context for ongoing I/O
+ * @chunk: Read chunk to pull
+ *
+ * Return values:
+ *   %0: the Read WR chain was constructed successfully
+ *   %-EINVAL: there were not enough resources to finish
+ *   %-ENOMEM: allocating a local resources failed
+ *   %-EIO: a DMA mapping error occurred
+ */
+static int svc_rdma_build_read_chunk(struct svc_rdma_read_info *info,
+				     const struct svc_rdma_chunk *chunk)
+{
+	const struct svc_rdma_segment *segment;
+	int ret;
+
+	ret = -EINVAL;
+	pcl_for_each_segment(segment, chunk) {
+		ret = svc_rdma_build_read_segment(info, segment);
+		if (ret < 0)
+			break;
+		info->ri_totalbytes += segment->rs_length;
+	}
+	return ret;
+}
+
+/**
+ * svc_rdma_copy_inline_range - Copy part of the inline content into pages
+ * @info: context for RDMA Reads
+ * @offset: offset into the Receive buffer of region to copy
+ * @remaining: length of region to copy
+ *
+ * Take a page at a time from rqstp->rq_pages and copy the inline
+ * content from the Receive buffer into that page. Update
+ * info->ri_pageno and info->ri_pageoff so that the next RDMA Read
+ * result will land contiguously with the copied content.
+ *
+ * Return values:
+ *   %0: Inline content was successfully copied
+ *   %-EINVAL: offset or length was incorrect
+ */
+static int svc_rdma_copy_inline_range(struct svc_rdma_read_info *info,
+				      unsigned int offset,
+				      unsigned int remaining)
+{
+	struct svc_rdma_recv_ctxt *head = info->ri_readctxt;
+	unsigned char *dst, *src = head->rc_recv_buf;
+	struct svc_rqst *rqstp = info->ri_rqst;
+	unsigned int page_no, numpages;
+
+	numpages = PAGE_ALIGN(info->ri_pageoff + remaining) >> PAGE_SHIFT;
+	for (page_no = 0; page_no < numpages; page_no++) {
+		unsigned int page_len;
+
+		page_len = min_t(unsigned int, remaining,
+				 PAGE_SIZE - info->ri_pageoff);
+
+		if (!info->ri_pageoff)
+			head->rc_page_count++;
+
+		dst = page_address(rqstp->rq_pages[info->ri_pageno]);
+		memcpy(dst + info->ri_pageno, src + offset, page_len);
+
+		info->ri_totalbytes += page_len;
+		info->ri_pageoff += page_len;
+		if (info->ri_pageoff == PAGE_SIZE) {
+			info->ri_pageno++;
+			info->ri_pageoff = 0;
+		}
+		remaining -= page_len;
+		offset += page_len;
+	}
+
+	return -EINVAL;
+}
+
+/**
+ * svc_rdma_read_multiple_chunks - Construct RDMA Reads to pull data item Read chunks
+ * @info: context for RDMA Reads
+ *
+ * The chunk data lands in rqstp->rq_arg as a series of contiguous pages,
+ * like an incoming TCP call.
+ *
+ * Return values:
+ *   %0: RDMA Read WQEs were successfully built
+ *   %-EINVAL: client provided too many chunks or segments,
+ *   %-ENOMEM: rdma_rw context pool was exhausted,
+ *   %-ENOTCONN: posting failed (connection is lost),
+ *   %-EIO: rdma_rw initialization failed (DMA mapping, etc).
+ */
+static noinline int svc_rdma_read_multiple_chunks(struct svc_rdma_read_info *info)
+{
+	struct svc_rdma_recv_ctxt *head = info->ri_readctxt;
+	const struct svc_rdma_pcl *pcl = &head->rc_read_pcl;
+	struct xdr_buf *buf = &info->ri_rqst->rq_arg;
+	struct svc_rdma_chunk *chunk, *next;
+	unsigned int start, length;
+	int ret;
+
+	start = 0;
+	chunk = pcl_first_chunk(pcl);
+	length = chunk->ch_position;
+	ret = svc_rdma_copy_inline_range(info, start, length);
+	if (ret < 0)
+		return ret;
+
+	pcl_for_each_chunk(chunk, pcl) {
+		ret = svc_rdma_build_read_chunk(info, chunk);
+		if (ret < 0)
+			return ret;
+
+		next = pcl_next_chunk(pcl, chunk);
+		if (!next)
+			break;
+
+		start += length;
+		length = next->ch_position - info->ri_totalbytes;
+		ret = svc_rdma_copy_inline_range(info, start, length);
+		if (ret < 0)
+			return ret;
+	}
+
+	start += length;
+	length = head->rc_byte_len - start;
+	ret = svc_rdma_copy_inline_range(info, start, length);
+	if (ret < 0)
+		return ret;
+
+	buf->len += info->ri_totalbytes;
+	buf->buflen += info->ri_totalbytes;
+
+	buf->head[0].iov_base = page_address(info->ri_rqst->rq_pages[0]);
+	buf->head[0].iov_len = min_t(size_t, PAGE_SIZE, info->ri_totalbytes);
+	buf->pages = &info->ri_rqst->rq_pages[1];
+	buf->page_len = info->ri_totalbytes - buf->head[0].iov_len;
+	return 0;
+}
+
+/**
+ * svc_rdma_read_data_item - Construct RDMA Reads to pull data item Read chunks
+ * @info: context for RDMA Reads
+ *
+ * The chunk data lands in the page list of rqstp->rq_arg.pages.
+ *
+ * Currently NFSD does not look at the rqstp->rq_arg.tail[0] kvec.
+ * Therefore, XDR round-up of the Read chunk and trailing
+ * inline content must both be added at the end of the pagelist.
+ *
+ * Return values:
+ *   %0: RDMA Read WQEs were successfully built
+ *   %-EINVAL: client provided too many chunks or segments,
+ *   %-ENOMEM: rdma_rw context pool was exhausted,
+ *   %-ENOTCONN: posting failed (connection is lost),
+ *   %-EIO: rdma_rw initialization failed (DMA mapping, etc).
+ */
+static int svc_rdma_read_data_item(struct svc_rdma_read_info *info)
+{
+	struct svc_rdma_recv_ctxt *head = info->ri_readctxt;
+	struct xdr_buf *buf = &info->ri_rqst->rq_arg;
+	struct svc_rdma_chunk *chunk;
+	unsigned int length;
+	int ret;
+
+	chunk = pcl_first_chunk(&head->rc_read_pcl);
+	ret = svc_rdma_build_read_chunk(info, chunk);
+	if (ret < 0)
+		goto out;
+
+	/* Split the Receive buffer between the head and tail
+	 * buffers at Read chunk's position. XDR roundup of the
+	 * chunk is not included in either the pagelist or in
+	 * the tail.
+	 */
+	buf->tail[0].iov_base = buf->head[0].iov_base + chunk->ch_position;
+	buf->tail[0].iov_len = buf->head[0].iov_len - chunk->ch_position;
+	buf->head[0].iov_len = chunk->ch_position;
+
+	/* Read chunk may need XDR roundup (see RFC 8166, s. 3.4.5.2).
+	 *
+	 * If the client already rounded up the chunk length, the
+	 * length does not change. Otherwise, the length of the page
+	 * list is increased to include XDR round-up.
+	 *
+	 * Currently these chunks always start at page offset 0,
+	 * thus the rounded-up length never crosses a page boundary.
+	 */
+	buf->pages = &info->ri_rqst->rq_pages[0];
+	length = xdr_align_size(chunk->ch_length);
+	buf->page_len = length;
+	buf->len += length;
+	buf->buflen += length;
+
+out:
+	return ret;
+}
+
+/**
+ * svc_rdma_read_chunk_range - Build RDMA Read WQEs for portion of a chunk
+ * @info: context for RDMA Reads
+ * @chunk: parsed Call chunk to pull
+ * @offset: offset of region to pull
+ * @length: length of region to pull
+ *
+ * Return values:
+ *   %0: RDMA Read WQEs were successfully built
+ *   %-EINVAL: there were not enough resources to finish
+ *   %-ENOMEM: rdma_rw context pool was exhausted,
+ *   %-ENOTCONN: posting failed (connection is lost),
+ *   %-EIO: rdma_rw initialization failed (DMA mapping, etc).
+ */
+static int svc_rdma_read_chunk_range(struct svc_rdma_read_info *info,
+				     const struct svc_rdma_chunk *chunk,
+				     unsigned int offset, unsigned int length)
+{
+	const struct svc_rdma_segment *segment;
+	int ret;
+
+	ret = -EINVAL;
+	pcl_for_each_segment(segment, chunk) {
+		struct svc_rdma_segment dummy;
+
+		if (offset > segment->rs_length) {
+			offset -= segment->rs_length;
+			continue;
+		}
+
+		dummy.rs_handle = segment->rs_handle;
+		dummy.rs_length = min_t(u32, length, segment->rs_length) - offset;
+		dummy.rs_offset = segment->rs_offset + offset;
+
+		ret = svc_rdma_build_read_segment(info, &dummy);
+		if (ret < 0)
+			break;
+
+		info->ri_totalbytes += dummy.rs_length;
+		length -= dummy.rs_length;
+		offset = 0;
+	}
+	return ret;
+}
+
+/**
+ * svc_rdma_read_call_chunk - Build RDMA Read WQEs to pull a Long Message
+ * @info: context for RDMA Reads
+ *
+ * Return values:
+ *   %0: RDMA Read WQEs were successfully built
+ *   %-EINVAL: there were not enough resources to finish
+ *   %-ENOMEM: rdma_rw context pool was exhausted,
+ *   %-ENOTCONN: posting failed (connection is lost),
+ *   %-EIO: rdma_rw initialization failed (DMA mapping, etc).
+ */
+static int svc_rdma_read_call_chunk(struct svc_rdma_read_info *info)
+{
+	struct svc_rdma_recv_ctxt *head = info->ri_readctxt;
+	const struct svc_rdma_chunk *call_chunk =
+			pcl_first_chunk(&head->rc_call_pcl);
+	const struct svc_rdma_pcl *pcl = &head->rc_read_pcl;
+	struct svc_rdma_chunk *chunk, *next;
+	unsigned int start, length;
+	int ret;
+
+	if (pcl_is_empty(pcl))
+		return svc_rdma_build_read_chunk(info, call_chunk);
+
+	start = 0;
+	chunk = pcl_first_chunk(pcl);
+	length = chunk->ch_position;
+	ret = svc_rdma_read_chunk_range(info, call_chunk, start, length);
+	if (ret < 0)
+		return ret;
+
+	pcl_for_each_chunk(chunk, pcl) {
+		ret = svc_rdma_build_read_chunk(info, chunk);
+		if (ret < 0)
+			return ret;
+
+		next = pcl_next_chunk(pcl, chunk);
+		if (!next)
+			break;
+
+		start += length;
+		length = next->ch_position - info->ri_totalbytes;
+		ret = svc_rdma_read_chunk_range(info, call_chunk,
+						start, length);
+		if (ret < 0)
+			return ret;
+	}
+
+	start += length;
+	length = call_chunk->ch_length - start;
+	return svc_rdma_read_chunk_range(info, call_chunk, start, length);
+}
+
+/**
+ * svc_rdma_read_special - Build RDMA Read WQEs to pull a Long Message
+ * @info: context for RDMA Reads
+ *
+ * The start of the data lands in the first page just after the
+ * Transport header, and the rest lands in rqstp->rq_arg.pages.
+ *
+ * Assumptions:
+ *	- A PZRC is never sent in an RDMA_MSG message, though it's
+ *	  allowed by spec.
+ *
+ * Return values:
+ *   %0: RDMA Read WQEs were successfully built
+ *   %-EINVAL: client provided too many chunks or segments,
+ *   %-ENOMEM: rdma_rw context pool was exhausted,
+ *   %-ENOTCONN: posting failed (connection is lost),
+ *   %-EIO: rdma_rw initialization failed (DMA mapping, etc).
+ */
+static noinline int svc_rdma_read_special(struct svc_rdma_read_info *info)
+{
+	struct xdr_buf *buf = &info->ri_rqst->rq_arg;
+	int ret;
+
+	ret = svc_rdma_read_call_chunk(info);
+	if (ret < 0)
+		goto out;
+
+	buf->len += info->ri_totalbytes;
+	buf->buflen += info->ri_totalbytes;
+
+	buf->head[0].iov_base = page_address(info->ri_rqst->rq_pages[0]);
+	buf->head[0].iov_len = min_t(size_t, PAGE_SIZE, info->ri_totalbytes);
+	buf->pages = &info->ri_rqst->rq_pages[1];
+	buf->page_len = info->ri_totalbytes - buf->head[0].iov_len;
+
+out:
+	return ret;
+}
+
+/**
+ * svc_rdma_process_read_list - Pull list of Read chunks from the client
+ * @rdma: controlling RDMA transport
+ * @rqstp: set of pages to use as Read sink buffers
+ * @head: pages under I/O collect here
+ *
+ * The RPC/RDMA protocol assumes that the upper layer's XDR decoders
+ * pull each Read chunk as they decode an incoming RPC message.
+ *
+ * On Linux, however, the server needs to have a fully-constructed RPC
+ * message in rqstp->rq_arg when there is a positive return code from
+ * ->xpo_recvfrom. So the Read list is safety-checked immediately when
+ * it is received, then here the whole Read list is pulled all at once.
+ * The ingress RPC message is fully reconstructed once all associated
+ * RDMA Reads have completed.
+ *
+ * Return values:
+ *   %1: all needed RDMA Reads were posted successfully,
+ *   %-EINVAL: client provided too many chunks or segments,
+ *   %-ENOMEM: rdma_rw context pool was exhausted,
+ *   %-ENOTCONN: posting failed (connection is lost),
+ *   %-EIO: rdma_rw initialization failed (DMA mapping, etc).
+ */
+int svc_rdma_process_read_list(struct svcxprt_rdma *rdma,
+			       struct svc_rqst *rqstp,
+			       struct svc_rdma_recv_ctxt *head)
+{
+	struct svc_rdma_read_info *info;
+	struct svc_rdma_chunk_ctxt *cc;
+	int ret;
+
+	info = svc_rdma_read_info_alloc(rdma);
+	if (!info)
+		return -ENOMEM;
+	cc = &info->ri_cc;
+	info->ri_rqst = rqstp;
+	info->ri_readctxt = head;
+	info->ri_pageno = 0;
+	info->ri_pageoff = 0;
+	info->ri_totalbytes = 0;
+
+	if (pcl_is_empty(&head->rc_call_pcl)) {
+		if (head->rc_read_pcl.cl_count == 1)
+			ret = svc_rdma_read_data_item(info);
+		else
+			ret = svc_rdma_read_multiple_chunks(info);
+	} else
+		ret = svc_rdma_read_special(info);
+	if (ret < 0)
+		goto out_err;
+
+	trace_svcrdma_post_read_chunk(&cc->cc_cid, cc->cc_sqecount);
+	init_completion(&cc->cc_done);
+	ret = svc_rdma_post_chunk_ctxt(cc);
+	if (ret < 0)
+		goto out_err;
+
+	ret = 1;
+	wait_for_completion(&cc->cc_done);
+	if (cc->cc_status != IB_WC_SUCCESS)
+		ret = -EIO;
+
+	/* rq_respages starts after the last arg page */
+	rqstp->rq_respages = &rqstp->rq_pages[head->rc_page_count];
+	rqstp->rq_next_page = rqstp->rq_respages + 1;
+
+	/* Ensure svc_rdma_recv_ctxt_put() does not try to release pages */
+	head->rc_page_count = 0;
+
+out_err:
+	svc_rdma_read_info_free(info);
+	return ret;
+}