RE: [PATCH v2 08/13] SoftiWarp Queue Pair methods

["Steve Wise" <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>]

> Signed-off-by: Bernard Metzler <bmt-OA+xvbQnYDHMbYB6QlFGEg@public.gmane.org>
> ---
>  drivers/infiniband/sw/siw/siw_qp.c | 1173
> ++++++++++++++++++++++++++++++++++++
>  1 file changed, 1173 insertions(+)
>  create mode 100644 drivers/infiniband/sw/siw/siw_qp.c
> 
> diff --git a/drivers/infiniband/sw/siw/siw_qp.c
b/drivers/infiniband/sw/siw/siw_qp.c
> new file mode 100644
> index 000000000000..dc33d8fd93f8
> --- /dev/null
> +++ b/drivers/infiniband/sw/siw/siw_qp.c
> @@ -0,0 +1,1173 @@
> +/*
> + * Software iWARP device driver for Linux
> + *
> + * Authors: Bernard Metzler <bmt-OA+xvbQnYDHMbYB6QlFGEg@public.gmane.org>
> + *          Fredy Neeser <nfd-OA+xvbQnYDHMbYB6QlFGEg@public.gmane.org>
> + *
> + * Copyright (c) 2008-2017, IBM Corporation
> + *
> + * This software is available to you under a choice of one of two
> + * licenses. You may choose to be licensed under the terms of the GNU
> + * General Public License (GPL) Version 2, available from the file
> + * COPYING in the main directory of this source tree, or the
> + * BSD license below:
> + *
> + *   Redistribution and use in source and binary forms, with or
> + *   without modification, are permitted provided that the following
> + *   conditions are met:
> + *
> + *   - Redistributions of source code must retain the above copyright notice,
> + *     this list of conditions and the following disclaimer.
> + *
> + *   - 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.
> + *
> + *   - Neither the name of IBM nor the names of its contributors may be
> + *     used to endorse or promote products derived from this software without
> + *     specific prior written permission.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
> + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
> + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
> + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
> HOLDERS
> + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
> + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
> + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
> + * SOFTWARE.
> + */
> +
> +#include <linux/errno.h>
> +#include <linux/types.h>
> +#include <linux/net.h>
> +#include <linux/file.h>
> +#include <linux/scatterlist.h>
> +#include <linux/highmem.h>
> +#include <linux/vmalloc.h>
> +#include <asm/barrier.h>
> +#include <net/sock.h>
> +#include <net/tcp_states.h>
> +#include <net/tcp.h>
> +
> +#include <rdma/iw_cm.h>
> +#include <rdma/ib_verbs.h>
> +#include <rdma/ib_smi.h>
> +#include <rdma/ib_user_verbs.h>
> +
> +#include "siw.h"
> +#include "siw_obj.h"
> +#include "siw_cm.h"
> +
> +
> +#if DPRINT_MASK > 0
> +static char siw_qp_state_to_string[SIW_QP_STATE_COUNT][sizeof "TERMINATE"]
> = {
> +	[SIW_QP_STATE_IDLE]		= "IDLE",
> +	[SIW_QP_STATE_RTR]		= "RTR",
> +	[SIW_QP_STATE_RTS]		= "RTS",
> +	[SIW_QP_STATE_CLOSING]		= "CLOSING",
> +	[SIW_QP_STATE_TERMINATE]	= "TERMINATE",
> +	[SIW_QP_STATE_ERROR]		= "ERROR"
> +};
> +#endif
> +
> +extern struct crypto_shash *siw_crypto_shash;
> +
> +/*
> + * iWARP (RDMAP, DDP and MPA) parameters as well as Softiwarp settings on a
> + * per-RDMAP message basis. Please keep order of initializer. All MPA len
> + * is initialized to minimum packet size.
> + */
> +struct iwarp_msg_info iwarp_pktinfo[RDMAP_TERMINATE + 1] = { {
> +	/* RDMAP_RDMA_WRITE */
> +	.hdr_len = sizeof(struct iwarp_rdma_write),
> +	.ctrl.mpa_len = htons(sizeof(struct iwarp_rdma_write) - 2),
> +	.ctrl.ddp_rdmap_ctrl = DDP_FLAG_TAGGED | DDP_FLAG_LAST
> +		| cpu_to_be16(DDP_VERSION << 8)
> +		| cpu_to_be16(RDMAP_VERSION << 6)
> +		| cpu_to_be16(RDMAP_RDMA_WRITE),
> +	.proc_data = siw_proc_write
> +},
> +{	/* RDMAP_RDMA_READ_REQ */
> +	.hdr_len = sizeof(struct iwarp_rdma_rreq),
> +	.ctrl.mpa_len = htons(sizeof(struct iwarp_rdma_rreq) - 2),
> +	.ctrl.ddp_rdmap_ctrl = DDP_FLAG_LAST
> +		| cpu_to_be16(DDP_VERSION << 8)
> +		| cpu_to_be16(RDMAP_VERSION << 6)
> +		| cpu_to_be16(RDMAP_RDMA_READ_REQ),
> +	.proc_data = siw_proc_rreq
> +},
> +{	/* RDMAP_RDMA_READ_RESP */
> +	.hdr_len = sizeof(struct iwarp_rdma_rresp),
> +	.ctrl.mpa_len = htons(sizeof(struct iwarp_rdma_rresp) - 2),
> +	.ctrl.ddp_rdmap_ctrl = DDP_FLAG_TAGGED | DDP_FLAG_LAST
> +		| cpu_to_be16(DDP_VERSION << 8)
> +		| cpu_to_be16(RDMAP_VERSION << 6)
> +		| cpu_to_be16(RDMAP_RDMA_READ_RESP),
> +	.proc_data = siw_proc_rresp
> +},
> +{	/* RDMAP_SEND */
> +	.hdr_len = sizeof(struct iwarp_send),
> +	.ctrl.mpa_len = htons(sizeof(struct iwarp_send) - 2),
> +	.ctrl.ddp_rdmap_ctrl = DDP_FLAG_LAST
> +		| cpu_to_be16(DDP_VERSION << 8)
> +		| cpu_to_be16(RDMAP_VERSION << 6)
> +		| cpu_to_be16(RDMAP_SEND),
> +	.proc_data = siw_proc_send
> +},
> +{	/* RDMAP_SEND_INVAL */
> +	.hdr_len = sizeof(struct iwarp_send_inv),
> +	.ctrl.mpa_len = htons(sizeof(struct iwarp_send_inv) - 2),
> +	.ctrl.ddp_rdmap_ctrl = DDP_FLAG_LAST
> +		| cpu_to_be16(DDP_VERSION << 8)
> +		| cpu_to_be16(RDMAP_VERSION << 6)
> +		| cpu_to_be16(RDMAP_SEND_INVAL),
> +	.proc_data = siw_proc_send
> +},
> +{	/* RDMAP_SEND_SE */
> +	.hdr_len = sizeof(struct iwarp_send),
> +	.ctrl.mpa_len = htons(sizeof(struct iwarp_send) - 2),
> +	.ctrl.ddp_rdmap_ctrl = DDP_FLAG_LAST
> +		| cpu_to_be16(DDP_VERSION << 8)
> +		| cpu_to_be16(RDMAP_VERSION << 6)
> +		| cpu_to_be16(RDMAP_SEND_SE),
> +	.proc_data = siw_proc_send
> +},
> +{	/* RDMAP_SEND_SE_INVAL */
> +	.hdr_len = sizeof(struct iwarp_send_inv),
> +	.ctrl.mpa_len = htons(sizeof(struct iwarp_send_inv) - 2),
> +	.ctrl.ddp_rdmap_ctrl = DDP_FLAG_LAST
> +		| cpu_to_be16(DDP_VERSION << 8)
> +		| cpu_to_be16(RDMAP_VERSION << 6)
> +		| cpu_to_be16(RDMAP_SEND_SE_INVAL),
> +	.proc_data = siw_proc_send
> +},
> +{	/* RDMAP_TERMINATE */
> +	.hdr_len = sizeof(struct iwarp_terminate),
> +	.ctrl.mpa_len = htons(sizeof(struct iwarp_terminate) - 2),
> +	.ctrl.ddp_rdmap_ctrl = DDP_FLAG_LAST
> +		| cpu_to_be16(DDP_VERSION << 8)
> +		| cpu_to_be16(RDMAP_VERSION << 6)
> +		| cpu_to_be16(RDMAP_TERMINATE),
> +	.proc_data = siw_proc_terminate
> +} };
> +
> +void siw_qp_llp_data_ready(struct sock *sk)
> +{
> +	struct siw_qp		*qp;
> +
> +	read_lock(&sk->sk_callback_lock);
> +
> +	if (unlikely(!sk->sk_user_data || !sk_to_qp(sk))) {
> +		dprint(DBG_ON, " No QP: %p\n", sk->sk_user_data);
> +		goto done;
> +	}
> +	qp = sk_to_qp(sk);
> +
> +	if (likely(!qp->rx_ctx.rx_suspend &&
> +		   down_read_trylock(&qp->state_lock))) {
> +		read_descriptor_t rd_desc = {.arg.data = qp, .count = 1};
> +
> +		dprint(DBG_SK|DBG_RX, "(QP%d): state (before read_sock)=%d\n",
> +			QP_ID(qp), qp->attrs.state);
> +
> +		if (likely(qp->attrs.state == SIW_QP_STATE_RTS))
> +			/*
> +			 * Implements data receive operation during
> +			 * socket callback. TCP gracefully catches
> +			 * the case where there is nothing to receive
> +			 * (not calling siw_tcp_rx_data() then).
> +			 */
> +			tcp_read_sock(sk, &rd_desc, siw_tcp_rx_data);
> +
> +		dprint(DBG_SK|DBG_RX, "(QP%d): state (after read_sock)=%d\n",
> +			QP_ID(qp), qp->attrs.state);
> +
> +		up_read(&qp->state_lock);
> +	} else {
> +		dprint(DBG_SK|DBG_RX, "(QP%d): Unable to RX: rx_suspend:
> %d\n",
> +			QP_ID(qp), qp->rx_ctx.rx_suspend);
> +	}
> +done:
> +	read_unlock(&sk->sk_callback_lock);
> +}
> +
> +
> +void siw_qp_llp_close(struct siw_qp *qp)
> +{
> +	dprint(DBG_CM, "(QP%d): Enter: SIW QP state = %s, cep=0x%p\n",
> +		QP_ID(qp), siw_qp_state_to_string[qp->attrs.state],
> +		qp->cep);
> +
> +	down_write(&qp->state_lock);
> +
> +	dprint(DBG_CM, "(QP%d): state locked\n", QP_ID(qp));
> +
> +	qp->rx_ctx.rx_suspend = 1;
> +	qp->tx_ctx.tx_suspend = 1;
> +	qp->attrs.llp_stream_handle = NULL;
> +
> +	switch (qp->attrs.state) {
> +
> +	case SIW_QP_STATE_RTS:
> +	case SIW_QP_STATE_RTR:
> +	case SIW_QP_STATE_IDLE:
> +	case SIW_QP_STATE_TERMINATE:
> +
> +		qp->attrs.state = SIW_QP_STATE_ERROR;
> +
> +		break;
> +	/*
> +	 * SIW_QP_STATE_CLOSING:
> +	 *
> +	 * This is a forced close. shall the QP be moved to
> +	 * ERROR or IDLE ?
> +	 */
> +	case SIW_QP_STATE_CLOSING:
> +		if (tx_wqe(qp)->wr_status == SIW_WR_IDLE)
> +			qp->attrs.state = SIW_QP_STATE_ERROR;
> +		else
> +			qp->attrs.state = SIW_QP_STATE_IDLE;
> +
> +		break;
> +
> +	default:
> +		dprint(DBG_CM, " No state transition needed: %d\n",
> +			qp->attrs.state);
> +		break;
> +	}
> +	siw_sq_flush(qp);
> +	siw_rq_flush(qp);
> +
> +	/*
> +	 * dereference closing CEP
> +	 */
> +	if (qp->cep) {
> +		siw_cep_put(qp->cep);
> +		qp->cep = NULL;
> +	}
> +
> +	up_write(&qp->state_lock);
> +	dprint(DBG_CM, "(QP%d): Exit: SIW QP state = %s, cep=0x%p\n",
> +		QP_ID(qp), siw_qp_state_to_string[qp->attrs.state],
> +		qp->cep);
> +}
> +
> +
> +/*
> + * socket callback routine informing about newly available send space.
> + * Function schedules SQ work for processing SQ items.
> + */
> +void siw_qp_llp_write_space(struct sock *sk)
> +{
> +	struct siw_cep	*cep = sk_to_cep(sk);
> +
> +	cep->sk_write_space(sk);
> +
> +	if (!test_bit(SOCK_NOSPACE, &sk->sk_socket->flags))
> +		siw_sq_start(cep->qp);
> +}
> +
> +static int siw_qp_readq_init(struct siw_qp *qp, int irq_size, int orq_size)
> +{
> +	dprint(DBG_CM|DBG_WR, "(QP%d): %d %d\n", QP_ID(qp), irq_size,
> orq_size);
> +
> +	if (!irq_size)
> +		irq_size = 1;
> +	if (!orq_size)
> +		orq_size = 1;
> +
> +	qp->attrs.irq_size = irq_size;
> +	qp->attrs.orq_size = orq_size;
> +
> +	qp->irq = vmalloc(irq_size * sizeof(struct siw_sqe));
> +	if (!qp->irq) {
> +		dprint(DBG_ON, "(QP%d): Failed\n", QP_ID(qp));
> +		qp->attrs.irq_size = 0;
> +		return -ENOMEM;
> +	}
> +	qp->orq = vmalloc(orq_size * sizeof(struct siw_sqe));
> +	if (!qp->orq) {
> +		dprint(DBG_ON, "(QP%d): Failed\n", QP_ID(qp));
> +		qp->attrs.orq_size = 0;
> +		qp->attrs.irq_size = 0;
> +		vfree(qp->irq);
> +		return -ENOMEM;
> +	}
> +	memset(qp->irq, 0, irq_size * sizeof(struct siw_sqe));
> +	memset(qp->orq, 0, orq_size * sizeof(struct siw_sqe));
> +
> +	return 0;
> +}
> +
> +
> +static int siw_qp_enable_crc(struct siw_qp *qp)
> +{
> +	struct siw_iwarp_rx *c_rx = &qp->rx_ctx;
> +	struct siw_iwarp_tx *c_tx = &qp->tx_ctx;
> +	int rv = 0;
> +
> +	if (siw_crypto_shash == NULL) {
> +		rv = -ENOENT;
> +		goto error;
> +	}
> +	c_tx->mpa_crc_hd = kzalloc(sizeof(struct shash_desc) +
> +				   crypto_shash_descsize(siw_crypto_shash),
> +				   GFP_KERNEL);
> +	c_rx->mpa_crc_hd = kzalloc(sizeof(struct shash_desc) +
> +				   crypto_shash_descsize(siw_crypto_shash),
> +				   GFP_KERNEL);
> +	if (!c_tx->mpa_crc_hd || !c_rx->mpa_crc_hd) {
> +		rv = -ENOMEM;
> +		goto error;
> +	}
> +	c_tx->mpa_crc_hd->tfm = siw_crypto_shash;
> +	c_rx->mpa_crc_hd->tfm = siw_crypto_shash;
> +
> +	return 0;
> +error:
> +	dprint(DBG_ON, "(QP%d): Failed loading crc32c: error=%d.",
> +			QP_ID(qp), rv);
> +
> +	kfree(c_tx->mpa_crc_hd);
> +	kfree(c_rx->mpa_crc_hd);
> +
> +	c_tx->mpa_crc_hd = c_rx->mpa_crc_hd = NULL;
> +
> +	return rv;
> +}
> +
> +/*
> + * Send a non signalled READ or WRITE to peer side as negotiated
> + * with MPAv2 P2P setup protocol. The work request is only created
> + * as a current active WR and does not consume Send Queue space.
> + *
> + * Caller must hold QP state lock.
> + */
> +int siw_qp_mpa_rts(struct siw_qp *qp, enum mpa_v2_ctrl ctrl)
> +{
> +	struct siw_wqe	*wqe = tx_wqe(qp);
> +	unsigned long flags;
> +	int rv = 0;
> +
> +	spin_lock_irqsave(&qp->sq_lock, flags);
> +
> +	if (unlikely(wqe->wr_status != SIW_WR_IDLE)) {
> +		spin_unlock_irqrestore(&qp->sq_lock, flags);
> +		return -EIO;
> +	}
> +	memset(wqe->mem, 0, sizeof(*wqe->mem) * SIW_MAX_SGE);
> +
> +	wqe->wr_status = SIW_WR_QUEUED;
> +	wqe->sqe.flags = 0;
> +	wqe->sqe.num_sge = 1;
> +	wqe->sqe.sge[0].length = 0;
> +	wqe->sqe.sge[0].laddr = 0;
> +	wqe->sqe.sge[0].lkey = 0;
> +	/*
> +	 * While it must not be checked for inbound zero length
> +	 * READ/WRITE, some HW may treat STag 0 special.
> +	 */
> +	wqe->sqe.rkey = 1;
> +	wqe->sqe.raddr = 0;
> +	wqe->processed = 0;
> +
> +	if (ctrl & MPA_V2_RDMA_WRITE_RTR)
> +		wqe->sqe.opcode = SIW_OP_WRITE;
> +	else if (ctrl & MPA_V2_RDMA_READ_RTR) {
> +		struct siw_sqe	*rreq;
> +
> +		wqe->sqe.opcode = SIW_OP_READ;
> +
> +		spin_lock(&qp->orq_lock);
> +
> +		rreq = orq_get_free(qp);
> +		if (rreq) {
> +			siw_read_to_orq(rreq, &wqe->sqe);
> +			qp->orq_put++;
> +		} else
> +			rv = -EIO;
> +
> +		spin_unlock(&qp->orq_lock);
> +	} else
> +		rv = -EINVAL;
> +
> +	if (rv)
> +		wqe->wr_status = SIW_WR_IDLE;
> +
> +	spin_unlock_irqrestore(&qp->sq_lock, flags);
> +
> +	if (!rv)
> +		siw_sq_start(qp);
> +
> +	return rv;
> +}
> +
> +/*
> + * handle all attrs other than state
> + */
> +static void siw_qp_modify_nonstate(struct siw_qp *qp,
> +				  struct siw_qp_attrs *attrs,
> +				  enum siw_qp_attr_mask mask)
> +{
> +	if (mask & SIW_QP_ATTR_ACCESS_FLAGS) {
> +		if (attrs->flags & SIW_RDMA_BIND_ENABLED)
> +			qp->attrs.flags |= SIW_RDMA_BIND_ENABLED;
> +		else
> +			qp->attrs.flags &= ~SIW_RDMA_BIND_ENABLED;
> +
> +		if (attrs->flags & SIW_RDMA_WRITE_ENABLED)
> +			qp->attrs.flags |= SIW_RDMA_WRITE_ENABLED;
> +		else
> +			qp->attrs.flags &= ~SIW_RDMA_WRITE_ENABLED;
> +
> +		if (attrs->flags & SIW_RDMA_READ_ENABLED)
> +			qp->attrs.flags |= SIW_RDMA_READ_ENABLED;
> +		else
> +			qp->attrs.flags &= ~SIW_RDMA_READ_ENABLED;
> +	}
> +}
> +
> +/*
> + * caller holds qp->state_lock
> + */
> +int siw_qp_modify(struct siw_qp *qp, struct siw_qp_attrs *attrs,
> +		  enum siw_qp_attr_mask mask)
> +{
> +	int	drop_conn = 0, rv = 0;
> +
> +	if (!mask)
> +		return 0;
> +
> +	dprint(DBG_CM, "(QP%d)\n", QP_ID(qp));
> +
> +	if (mask != SIW_QP_ATTR_STATE)
> +		siw_qp_modify_nonstate(qp, attrs, mask);
> +
> +	if (!(mask & SIW_QP_ATTR_STATE))
> +		return 0;
> +
> +	dprint(DBG_CM, "(QP%d): SIW QP state: %s => %s\n", QP_ID(qp),
> +		siw_qp_state_to_string[qp->attrs.state],
> +		siw_qp_state_to_string[attrs->state]);
> +
> +
> +	switch (qp->attrs.state) {
> +
> +	case SIW_QP_STATE_IDLE:
> +	case SIW_QP_STATE_RTR:
> +
> +		switch (attrs->state) {
> +
> +		case SIW_QP_STATE_RTS:
> +
> +			if (attrs->flags & SIW_MPA_CRC) {
> +				rv = siw_qp_enable_crc(qp);
> +				if (rv)
> +					break;
> +			}
> +			if (!(mask & SIW_QP_ATTR_LLP_HANDLE)) {
> +				dprint(DBG_ON, "(QP%d): socket?\n", QP_ID(qp));
> +				rv = -EINVAL;
> +				break;
> +			}
> +			if (!(mask & SIW_QP_ATTR_MPA)) {
> +				dprint(DBG_ON, "(QP%d): MPA?\n", QP_ID(qp));
> +				rv = -EINVAL;
> +				break;
> +			}
> +			dprint(DBG_CM, "(QP%d): Enter RTS\n", QP_ID(qp));
> +			dprint(DBG_CM, " peer 0x%08x, local 0x%08x\n",
> +				qp->cep->llp.raddr.sin_addr.s_addr,
> +				qp->cep->llp.laddr.sin_addr.s_addr);
> +			/*
> +			 * Initialize global iWARP TX state
> +			 */
> +			qp->tx_ctx.ddp_msn[RDMAP_UNTAGGED_QN_SEND] = 0;
> +			qp-
> >tx_ctx.ddp_msn[RDMAP_UNTAGGED_QN_RDMA_READ] = 0;
> +			qp-
> >tx_ctx.ddp_msn[RDMAP_UNTAGGED_QN_TERMINATE] = 0;
> +
> +			/*
> +			 * Initialize global iWARP RX state
> +			 */
> +			qp->rx_ctx.ddp_msn[RDMAP_UNTAGGED_QN_SEND] = 1;
> +			qp-
> >rx_ctx.ddp_msn[RDMAP_UNTAGGED_QN_RDMA_READ] = 1;
> +			qp-
> >rx_ctx.ddp_msn[RDMAP_UNTAGGED_QN_TERMINATE] = 1;
> +
> +			/*
> +			 * init IRD free queue, caller has already checked
> +			 * limits.
> +			 */
> +			rv = siw_qp_readq_init(qp, attrs->irq_size,
> +					       attrs->orq_size);
> +			if (rv)
> +				break;
> +
> +			qp->attrs.llp_stream_handle = attrs->llp_stream_handle;
> +			qp->attrs.state = SIW_QP_STATE_RTS;
> +
> +			break;
> +
> +		case SIW_QP_STATE_ERROR:
> +			siw_rq_flush(qp);

What about flushing the sq?
 
> +			qp->attrs.state = SIW_QP_STATE_ERROR;
> +			if (qp->cep) {
> +				siw_cep_put(qp->cep);
> +				qp->cep = NULL;
> +			}
> +			break;
> +
> +		case SIW_QP_STATE_RTR:
> +			/* ignore */
> +			break;
> +
> +		default:
> +			dprint(DBG_CM,
> +				" QP state transition undefined: %s => %s\n",
> +				siw_qp_state_to_string[qp->attrs.state],
> +				siw_qp_state_to_string[attrs->state]);

Shouldn't this cause an error to be returned?

> +			break;
> +		}
> +		break;
> +
> +	case SIW_QP_STATE_RTS:
> +
> +		switch (attrs->state) {
> +
> +		case SIW_QP_STATE_CLOSING:
> +			/*
> +			 * Verbs: move to IDLE if SQ and ORQ are empty.
> +			 * Move to ERROR otherwise. But first of all we must
> +			 * close the connection. So we keep CLOSING or ERROR
> +			 * as a transient state, schedule connection drop work
> +			 * and wait for the socket state change upcall to
> +			 * come back closed.
> +			 */
> +			if (tx_wqe(qp)->wr_status == SIW_WR_IDLE)
> +				qp->attrs.state = SIW_QP_STATE_CLOSING;
> +			else {
> +				qp->attrs.state = SIW_QP_STATE_ERROR;
> +				siw_sq_flush(qp);
> +			}
> +			siw_rq_flush(qp);
> +
> +			drop_conn = 1;
> +			break;
> +
> +		case SIW_QP_STATE_TERMINATE:
> +			qp->attrs.state = SIW_QP_STATE_TERMINATE;
> +			/*
> +			 * To be extended for flexible error layer,
> +			 * type and code.
> +			 */
> +			siw_send_terminate(qp, RDMAP_ERROR_LAYER_RDMA,
> +					   RDMAP_ETYPE_CATASTROPHIC,
> +					   0);
> +			drop_conn = 1;
> +
> +			break;
> +
> +		case SIW_QP_STATE_ERROR:
> +			/*
> +			 * This is an emergency close.
> +			 *
> +			 * Any in progress transmit operation will get
> +			 * cancelled.
> +			 * This will likely result in a protocol failure,
> +			 * if a TX operation is in transit. The caller
> +			 * could unconditional wait to give the current
> +			 * operation a chance to complete.
> +			 * Esp., how to handle the non-empty IRQ case?
> +			 * The peer was asking for data transfer at a valid
> +			 * point in time.
> +			 */
> +			siw_sq_flush(qp);
> +			siw_rq_flush(qp);
> +			qp->attrs.state = SIW_QP_STATE_ERROR;
> +			drop_conn = 1;
> +
> +			break;
> +
> +		default:
> +			dprint(DBG_ON,
> +				" QP state transition undefined: %s => %s\n",
> +				siw_qp_state_to_string[qp->attrs.state],
> +				siw_qp_state_to_string[attrs->state]);
> +			break;
> +		}
> +		break;
> +
> +	case SIW_QP_STATE_TERMINATE:
> +
> +		switch (attrs->state) {
> +
> +		case SIW_QP_STATE_ERROR:
> +			siw_rq_flush(qp);
> +			qp->attrs.state = SIW_QP_STATE_ERROR;
> +
> +			if (tx_wqe(qp)->wr_status != SIW_WR_IDLE)
> +				siw_sq_flush(qp);
> +
> +			break;
> +
> +		default:
> +			dprint(DBG_ON,
> +				" QP state transition undefined: %s => %s\n",
> +				siw_qp_state_to_string[qp->attrs.state],
> +				siw_qp_state_to_string[attrs->state]);
> +		}
> +		break;
> +
> +	case SIW_QP_STATE_CLOSING:
> +
> +		switch (attrs->state) {
> +
> +		case SIW_QP_STATE_IDLE:
> +			BUG_ON(tx_wqe(qp)->wr_status != SIW_WR_IDLE);
> +			qp->attrs.state = SIW_QP_STATE_IDLE;
> +
> +			break;
> +
> +		case SIW_QP_STATE_CLOSING:
> +			/*
> +			 * The LLP may already moved the QP to closing
> +			 * due to graceful peer close init
> +			 */
> +			break;
> +
> +		case SIW_QP_STATE_ERROR:
> +			/*
> +			 * QP was moved to CLOSING by LLP event
> +			 * not yet seen by user.
> +			 */
> +			qp->attrs.state = SIW_QP_STATE_ERROR;
> +
> +			if (tx_wqe(qp)->wr_status != SIW_WR_IDLE)
> +				siw_sq_flush(qp);
> +
> +			siw_rq_flush(qp);
> +
> +			break;
> +
> +		default:
> +			dprint(DBG_CM,
> +				" QP state transition undefined: %s => %s\n",
> +				siw_qp_state_to_string[qp->attrs.state],
> +				siw_qp_state_to_string[attrs->state]);
> +			return -ECONNABORTED;
> +		}
> +		break;
> +
> +	default:
> +		dprint(DBG_CM, " NOP: State: %d\n", qp->attrs.state);
> +		break;
> +	}
> +	if (drop_conn)
> +		siw_qp_cm_drop(qp, 0);
> +
> +	return rv;
> +}
> +
> +struct ib_qp *siw_get_ofaqp(struct ib_device *ofa_dev, int id)
> +{
> +	struct siw_qp *qp =  siw_qp_id2obj(siw_dev_ofa2siw(ofa_dev), id);
> +
> +	dprint(DBG_OBJ, ": dev_name: %s, OFA QPID: %d, QP: %p\n",
> +		ofa_dev->name, id, qp);
> +	if (qp) {
> +		/*
> +		 * siw_qp_id2obj() increments object reference count
> +		 */
> +		siw_qp_put(qp);
> +		dprint(DBG_OBJ, " QPID: %d\n", QP_ID(qp));
> +		return &qp->ofa_qp;
> +	}
> +	return (struct ib_qp *)NULL;
> +}
> +
> +/*
> + * siw_check_mem()
> + *
> + * Check protection domain, STAG state, access permissions and
> + * address range for memory object.
> + *
> + * @pd:		Protection Domain memory should belong to
> + * @mem:	memory to be checked
> + * @addr:	starting addr of mem
> + * @perms:	requested access permissions
> + * @len:	len of memory interval to be checked
> + *
> + */
> +int siw_check_mem(struct siw_pd *pd, struct siw_mem *mem, u64 addr,
> +		  enum siw_access_flags perms, int len)
> +{
> +	if (siw_mem2mr(mem)->pd != pd) {
> +		dprint(DBG_WR|DBG_ON, "(PD%d): PD mismatch %p : %p\n",
> +			OBJ_ID(pd),
> +			siw_mem2mr(mem)->pd, pd);
> +
> +		return -EINVAL;
> +	}
> +	if (mem->stag_valid == 0) {
> +		dprint(DBG_WR|DBG_ON, "(PD%d): STAG 0x%08x invalid\n",
> +			OBJ_ID(pd), OBJ_ID(mem));
> +		return -EPERM;
> +	}
> +	/*
> +	 * check access permissions
> +	 */
> +	if ((mem->perms & perms) < perms) {
> +		dprint(DBG_WR|DBG_ON, "(PD%d): permissions 0x%08x <
> 0x%08x\n",
> +			OBJ_ID(pd), mem->perms, perms);
> +		return -EPERM;
> +	}
> +	/*
> +	 * Check address interval: we relax check to allow memory shrinked
> +	 * from the start address _after_ placing or fetching len bytes.
> +	 * TODO: this relaxation is probably overdone
> +	 */
> +	if (addr < mem->va || addr + len > mem->va + mem->len) {
> +		dprint(DBG_WR|DBG_ON, "(PD%d): MEM interval len %d\n",
> +			OBJ_ID(pd), len);
> +		dprint(DBG_WR|DBG_ON, "[0x%016llx, 0x%016llx) out of
> bounds\n",
> +			(unsigned long long)addr,
> +			(unsigned long long)(addr + len));
> +		dprint(DBG_WR|DBG_ON, "[0x%016llx, 0x%016llx]
> LKey=0x%08x\n",
> +			(unsigned long long)mem->va,
> +			(unsigned long long)(mem->va + mem->len),
> +			OBJ_ID(mem));
> +
> +		return -EINVAL;
> +	}
> +	return 0;
> +}
> +
> +/*
> + * siw_check_sge()
> + *
> + * Check SGE for access rights in given interval
> + *
> + * @pd:		Protection Domain memory should belong to
> + * @sge:	SGE to be checked
> + * @mem:	resulting memory reference if successful
> + * @perms:	requested access permissions
> + * @off:	starting offset in SGE
> + * @len:	len of memory interval to be checked
> + *
> + * NOTE: Function references SGE's memory object (mem->obj)
> + * if not yet done. New reference is kept if check went ok and
> + * released if check failed. If mem->obj is already valid, no new
> + * lookup is being done and mem is not released it check fails.
> + */
> +int
> +siw_check_sge(struct siw_pd *pd, struct siw_sge *sge,
> +	      union siw_mem_resolved *mem, enum siw_access_flags perms,
> +	      u32 off, int len)
> +{
> +	struct siw_dev	*sdev = pd->hdr.sdev;
> +	int		new_ref = 0, rv = 0;
> +
> +	if (len + off > sge->length) {
> +		rv = -EPERM;
> +		goto fail;
> +	}
> +	if (mem->obj == NULL) {
> +		mem->obj = siw_mem_id2obj(sdev, sge->lkey >> 8);
> +		if (mem->obj == NULL) {
> +			rv = -EINVAL;
> +			goto fail;
> +		}
> +		new_ref = 1;
> +	}
> +
> +	rv = siw_check_mem(pd, mem->obj, sge->laddr + off, perms, len);
> +	if (rv)
> +		goto fail;
> +
> +	return 0;
> +
> +fail:
> +	if (new_ref) {
> +		siw_mem_put(mem->obj);
> +		mem->obj = NULL;
> +	}
> +	return rv;
> +}
> +
> +void siw_read_to_orq(struct siw_sqe *rreq, struct siw_sqe *sqe)
> +{
> +	rreq->id = sqe->id;
> +	rreq->opcode = sqe->opcode;
> +	rreq->sge[0].laddr = sqe->sge[0].laddr;
> +	rreq->sge[0].length = sqe->sge[0].length;
> +	rreq->sge[0].lkey = sqe->sge[0].lkey;
> +	rreq->sge[1].lkey = sqe->sge[1].lkey;
> +	rreq->flags = sqe->flags | SIW_WQE_VALID;
> +	rreq->num_sge = 1;
> +}
> +
> +/*
> + * Must be called with SQ locked
> + */
> +int siw_activate_tx(struct siw_qp *qp)
> +{
> +	struct siw_sqe	*sqe;
> +	struct siw_wqe	*wqe = tx_wqe(qp);
> +	int rv = 1;
> +
> +	/*
> +	 * This codes prefers pending READ Responses over SQ processing
> +	 */

Can this cause starvation of the SQ? 

> +	sqe = &qp->irq[qp->irq_get % qp->attrs.irq_size];
> +
> +	if (sqe->flags & SIW_WQE_VALID) {
> +		memset(wqe->mem, 0, sizeof(*wqe->mem) * SIW_MAX_SGE);
> +		wqe->wr_status = SIW_WR_QUEUED;
> +
> +		/* start READ RESPONSE */
> +		wqe->sqe.opcode = SIW_OP_READ_RESPONSE;
> +		wqe->sqe.flags = 0;
> +		wqe->sqe.num_sge = 1;
> +		wqe->sqe.sge[0].length = sqe->sge[0].length;
> +		wqe->sqe.sge[0].laddr = sqe->sge[0].laddr;
> +		wqe->sqe.sge[0].lkey = sqe->sge[0].lkey;
> +		wqe->sqe.rkey = sqe->rkey;
> +		wqe->sqe.raddr = sqe->raddr;
> +
> +		wqe->processed = 0;
> +		qp->irq_get++;
> +		/* mark current IRQ entry free */
> +		smp_store_mb(sqe->flags, 0);
> +
> +		goto out;
> +	}
> +
> +	sqe = sq_get_next(qp);
> +	if (sqe) {
> +		memset(wqe->mem, 0, sizeof(*wqe->mem) * SIW_MAX_SGE);
> +		wqe->wr_status = SIW_WR_QUEUED;
> +
> +		/* First copy SQE to kernel private memory */
> +		memcpy(&wqe->sqe, sqe, sizeof(*sqe));
> +
> +		if (wqe->sqe.opcode >= SIW_NUM_OPCODES) {
> +			rv = -EINVAL;
> +			goto out;
> +		}
> +
> +		if (wqe->sqe.flags & SIW_WQE_INLINE) {
> +			if (wqe->sqe.opcode != SIW_OP_SEND &&
> +			    wqe->sqe.opcode != SIW_OP_WRITE) {
> +				rv = -EINVAL;
> +				goto out;
> +			}
> +			if (wqe->sqe.sge[0].length > SIW_MAX_INLINE) {
> +				rv = -EINVAL;
> +				goto out;
> +			}
> +			wqe->sqe.sge[0].laddr = (u64)&wqe->sqe.sge[1];
> +			wqe->sqe.sge[0].lkey = 0;
> +			wqe->sqe.num_sge = 1;
> +		}
> +
> +		if (wqe->sqe.flags & SIW_WQE_READ_FENCE) {
> +			/* A READ cannot be fenced */
> +			if (unlikely(wqe->sqe.opcode == SIW_OP_READ ||
> +			    wqe->sqe.opcode == SIW_OP_READ_LOCAL_INV)) {
> +				pr_info("QP[%d]: cannot fence READ\n",
> +					QP_ID(qp));
> +				rv = -EINVAL;
> +				goto out;
> +			}
> +			spin_lock(&qp->orq_lock);
> +
> +			if (!siw_orq_empty(qp)) {
> +				qp->tx_ctx.orq_fence = 1;
> +				rv = 0;
> +			}
> +			spin_unlock(&qp->orq_lock);
> +
> +		} else if (wqe->sqe.opcode == SIW_OP_READ ||
> +			   wqe->sqe.opcode == SIW_OP_READ_LOCAL_INV) {
> +			struct siw_sqe	*rreq;
> +
> +			wqe->sqe.num_sge = 1;
> +
> +			spin_lock(&qp->orq_lock);
> +
> +			rreq = orq_get_free(qp);
> +			if (rreq) {
> +				/*
> +				 * Make an immediate copy in ORQ to be ready
> +				 * to process loopback READ reply
> +				 */
> +				siw_read_to_orq(rreq, &wqe->sqe);
> +				qp->orq_put++;
> +			} else {
> +				qp->tx_ctx.orq_fence = 1;
> +				rv = 0;
> +			}
> +			spin_unlock(&qp->orq_lock);
> +		}
> +
> +		/* Clear SQE, can be re-used by application */
> +		smp_store_mb(sqe->flags, 0);
> +		qp->sq_get++;
> +	} else
> +		rv = 0;
> +
> +out:
> +	if (unlikely(rv < 0)) {
> +		pr_warn("QP[%d]: error %d in activate_tx\n", QP_ID(qp), rv);
> +		wqe->wr_status = SIW_WR_IDLE;
> +	}
> +	return rv;
> +}
> +
> +static void siw_cq_notify(struct siw_cq *cq, u32 flags)
> +{
> +	u32 cq_notify;
> +
> +	if (unlikely(!cq->ofa_cq.comp_handler))
> +		return;
> +
> +	cq_notify = READ_ONCE(*cq->notify);
> +
> +	if ((cq_notify & SIW_NOTIFY_NEXT_COMPLETION) ||
> +	    ((cq_notify & SIW_NOTIFY_SOLICITED) &&
> +	     (flags & SIW_WQE_SOLICITED))) {
> +		/* de-arm CQ */
> +		smp_store_mb(*cq->notify, SIW_NOTIFY_NOT);
> +		(*cq->ofa_cq.comp_handler)(&cq->ofa_cq, cq->ofa_cq.cq_context);
> +	}
> +}
> +
> +int siw_sqe_complete(struct siw_qp *qp, struct siw_sqe *sqe, u32 bytes,
> +		     enum siw_wc_status status)
> +{
> +	struct siw_cq *cq = qp->scq;
> +	struct siw_cqe *cqe;
> +	u32 idx;
> +	int rv = 0;
> +
> +	if (cq) {
> +		u32 sqe_flags = sqe->flags;
> +		unsigned long flags;
> +
> +		spin_lock_irqsave(&cq->lock, flags);
> +
> +		idx = cq->cq_put % cq->num_cqe;
> +		cqe = &cq->queue[idx];
> +
> +		if (!cqe->flags) {
> +			cqe->id = sqe->id;
> +			cqe->opcode = sqe->opcode;
> +			cqe->status = status;
> +			cqe->imm_data = 0;
> +			cqe->bytes = bytes;
> +
> +			if (cq->kernel_verbs) {
> +				siw_qp_get(qp);
> +				cqe->qp = qp;
> +			} else
> +				cqe->qp_id = QP_ID(qp);
> +
> +			/* mark CQE valid for application */
> +			smp_store_mb(cqe->flags, SIW_WQE_VALID);
> +			/* recycle SQE */
> +			smp_store_mb(sqe->flags, 0);
> +
> +			cq->cq_put++;
> +			spin_unlock_irqrestore(&cq->lock, flags);
> +			siw_cq_notify(cq, sqe_flags);
> +		} else {
> +			spin_unlock_irqrestore(&cq->lock, flags);
> +			rv = -ENOMEM;
> +			siw_cq_event(cq, IB_EVENT_CQ_ERR);
> +		}
> +	} else /* recycle SQE */
> +		smp_store_mb(sqe->flags, 0);
> +
> +	return rv;
> +}
> +
> +int siw_rqe_complete(struct siw_qp *qp, struct siw_rqe *rqe, u32 bytes,
> +		     enum siw_wc_status status)
> +{
> +	struct siw_cq *cq = qp->rcq;
> +	struct siw_cqe *cqe;
> +	u32 idx;
> +	int rv = 0;
> +
> +	if (cq) {
> +		unsigned long flags;
> +
> +		spin_lock_irqsave(&cq->lock, flags);
> +
> +		idx = cq->cq_put % cq->num_cqe;
> +		cqe = &cq->queue[idx];
> +
> +		if (!cqe->flags) {
> +			cqe->id = rqe->id;
> +			cqe->opcode = SIW_OP_RECEIVE;
> +			cqe->status = status;
> +			cqe->imm_data = 0;
> +			cqe->bytes = bytes;
> +
> +			if (cq->kernel_verbs) {
> +				siw_qp_get(qp);
> +				cqe->qp = qp;
> +			} else
> +				cqe->qp_id = QP_ID(qp);
> +
> +			/* mark CQE valid for application */
> +			smp_store_mb(cqe->flags, SIW_WQE_VALID);
> +			/* recycle RQE */
> +			smp_store_mb(rqe->flags, 0);
> +
> +			cq->cq_put++;
> +			spin_unlock_irqrestore(&cq->lock, flags);
> +			siw_cq_notify(cq, SIW_WQE_SIGNALLED);
> +		} else {
> +			spin_unlock_irqrestore(&cq->lock, flags);
> +			rv = -ENOMEM;
> +			siw_cq_event(cq, IB_EVENT_CQ_ERR);
> +		}
> +	} else /* recycle RQE */
> +		smp_store_mb(rqe->flags, 0);
> +
> +	return rv;
> +}
> +
> +/*
> + * siw_sq_flush()
> + *
> + * Flush SQ and ORRQ entries to CQ.
> + *
> + * TODO: Add termination code for in-progress WQE.
> + * TODO: an in-progress WQE may have been partially
> + *       processed. It should be enforced, that transmission
> + *       of a started DDP segment must be completed if possible
> + *       by any chance.

Not sure how much the above matters.  If the SQ is getting flushed with pending
IO, it is an error condition anyway.  But I guess it's a nice-to-have.

> + *
> + * Must be called with QP state write lock held.
> + * Therefore, SQ and ORQ lock must not be taken.
> + */
> +void siw_sq_flush(struct siw_qp *qp)
> +{
> +	struct siw_sqe	*sqe;
> +	struct siw_wqe	*wqe = tx_wqe(qp);
> +	int		async_event = 0;
> +
> +	dprint(DBG_OBJ|DBG_CM|DBG_WR, "(QP%d): Enter\n", QP_ID(qp));
> +	/*
> +	 * Start with completing any work currently on the ORQ
> +	 */
> +	for (;;) {
> +		if (qp->attrs.orq_size == 0)
> +			break;
> +
> +		sqe = &qp->orq[qp->orq_get % qp->attrs.orq_size];
> +		if (!sqe->flags)
> +			break;
> +
> +		if (siw_sqe_complete(qp, sqe, 0,
> +				     SIW_WC_WR_FLUSH_ERR) != 0)
> +			break;
> +
> +		qp->orq_get++;
> +	}
> +	/*
> +	 * Flush an in-progess WQE if present
> +	 */
> +	if (wqe->wr_status != SIW_WR_IDLE) {
> +		/*
> +		 * TODO: Add iWARP Termination code
> +		 */
> +		dprint(DBG_WR,
> +			" (QP%d): Flush current SQE %p, type %d, status %d\n",
> +			QP_ID(qp), wqe, tx_type(wqe), wqe->wr_status);
> +
> +		siw_wqe_put_mem(wqe, tx_type(wqe));
> +
> +		if (tx_type(wqe) != SIW_OP_READ_RESPONSE &&
> +			((tx_type(wqe) != SIW_OP_READ &&
> +			  tx_type(wqe) != SIW_OP_READ_LOCAL_INV) ||
> +			wqe->wr_status == SIW_WR_QUEUED))
> +			/*
> +			 * An in-progress RREQUEST is already in
> +			 * the ORQ
> +			 */
> +			siw_sqe_complete(qp, &wqe->sqe, wqe->bytes,
> +					 SIW_WC_WR_FLUSH_ERR);
> +
> +		wqe->wr_status = SIW_WR_IDLE;
> +	}
> +	/*
> +	 * Flush the Send Queue
> +	 */
> +	while (qp->attrs.sq_size) {
> +		sqe = &qp->sendq[qp->sq_get % qp->attrs.sq_size];
> +		if (!sqe->flags)
> +			break;
> +
> +		async_event = 1;
> +		if (siw_sqe_complete(qp, sqe, 0, SIW_WC_WR_FLUSH_ERR) != 0)
> +			/* Shall IB_EVENT_SQ_DRAINED be supressed ? */
> +			break;
> +
> +		sqe->flags = 0;
> +		qp->sq_get++;
> +	}
> +	if (async_event)
> +		siw_qp_event(qp, IB_EVENT_SQ_DRAINED);
> +}
> +
> +/*
> + * siw_rq_flush()
> + *
> + * Flush recv queue entries to CQ.
> + *
> + * Must be called with QP state write lock held.
> + * Therefore, RQ lock must not be taken.
> + */
> +void siw_rq_flush(struct siw_qp *qp)
> +{
> +	struct siw_wqe		*wqe = rx_wqe(qp);
> +
> +	dprint(DBG_OBJ|DBG_CM|DBG_WR, "(QP%d): Enter\n", QP_ID(qp));
> +
> +	/*
> +	 * Flush an in-progess WQE if present
> +	 */
> +	if (wqe->wr_status != SIW_WR_IDLE) {
> +		dprint(DBG_WR,
> +			" (QP%d): Flush current RQE %p, type %d, status %d\n",
> +			QP_ID(qp), wqe, rx_type(wqe), wqe->wr_status);
> +		siw_wqe_put_mem(wqe, rx_type(wqe));
> +		if (rx_type(wqe) == SIW_OP_RECEIVE) {
> +			siw_rqe_complete(qp, &wqe->rqe, wqe->bytes,
> +					 SIW_WC_WR_FLUSH_ERR);
> +		} else if (rx_type(wqe) != SIW_OP_READ &&
> +			   rx_type(wqe) != SIW_OP_READ_RESPONSE &&
> +			   rx_type(wqe) != SIW_OP_WRITE) {
> +			siw_sqe_complete(qp, &wqe->sqe, 0,
> SIW_WC_WR_FLUSH_ERR);
> +		}
> +		wqe->wr_status = SIW_WR_IDLE;
> +	}
> +	/*
> +	 * Flush the Receive Queue
> +	 */
> +	while (qp->attrs.rq_size) {
> +		struct siw_rqe *rqe =
> +			&qp->recvq[qp->rq_get % qp->attrs.rq_size];
> +
> +		if (!rqe->flags)
> +			break;
> +
> +		if (siw_rqe_complete(qp, rqe, 0, SIW_WC_WR_FLUSH_ERR) != 0)
> +			break;
> +
> +		rqe->flags = 0;
> +		qp->rq_get++;
> +	}
> +}
> --
> 2.13.6
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html