root/sys/dev/cxgbe/t4_smt.c
/*-
 * SPDX-License-Identifier: BSD-2-Clause
 *
 * Copyright (c) 2018 Chelsio Communications, Inc.
 * All rights reserved.
 *
 * 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.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``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 AUTHOR 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 <sys/cdefs.h>
#include "opt_inet.h"
#include "opt_inet6.h"

#include <sys/param.h>
#include <sys/eventhandler.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/module.h>
#include <sys/bus.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/rwlock.h>
#include <sys/socket.h>
#include <sys/sbuf.h>
#include <netinet/in.h>

#include "common/common.h"
#include "common/t4_msg.h"
#include "t4_smt.h"

/*
 * Module locking notes:  There is a RW lock protecting the SMAC table as a
 * whole plus a spinlock per SMT entry.  Entry lookups and allocations happen
 * under the protection of the table lock, individual entry changes happen
 * while holding that entry's spinlock.  The table lock nests outside the
 * entry locks.  Allocations of new entries take the table lock as writers so
 * no other lookups can happen while allocating new entries.  Entry updates
 * take the table lock as readers so multiple entries can be updated in
 * parallel.  An SMT entry can be dropped by decrementing its reference count
 * and therefore can happen in parallel with entry allocation but no entry
 * can change state or increment its ref count during allocation as both of
 * these perform lookups.
 *
 * Note: We do not take references to ifnets in this module because both
 * the TOE and the sockets already hold references to the interfaces and the
 * lifetime of an SMT entry is fully contained in the lifetime of the TOE.
 */

/*
 * Allocate a free SMT entry.  Must be called with smt_data.lock held.
 */
struct smt_entry *
t4_find_or_alloc_sme(struct smt_data *s, uint8_t *smac)
{
        struct smt_entry *end, *e;
        struct smt_entry *first_free = NULL;

        rw_assert(&s->lock, RA_WLOCKED);
        for (e = &s->smtab[0], end = &s->smtab[s->smt_size]; e != end; ++e) {
                if (atomic_load_acq_int(&e->refcnt) == 0) {
                        if (!first_free)
                                first_free = e;
                } else {
                        if (e->state == SMT_STATE_SWITCHING) {
                                /*
                                 * This entry is actually in use. See if we can
                                 * re-use it?
                                 */
                                if (memcmp(e->smac, smac, ETHER_ADDR_LEN) == 0)
                                        goto found_reuse;
                        }
                }
        }
        if (first_free) {
                e = first_free;
                goto found;
        }
        return NULL;

found:
        e->state = SMT_STATE_UNUSED;
found_reuse:
        atomic_add_int(&e->refcnt, 1);
        return e;
}

/*
 * Write an SMT entry.  Must be called with the entry locked.
 */
int
t4_write_sme(struct smt_entry *e)
{
        struct smt_data *s;
        struct sge_wrq *wrq;
        struct adapter *sc;
        struct wrq_cookie cookie;
        struct cpl_smt_write_req *req;
        struct cpl_t6_smt_write_req *t6req;
        u8 row;

        mtx_assert(&e->lock, MA_OWNED);

        MPASS(e->wrq != NULL);
        wrq = e->wrq;
        sc = wrq->adapter;
        MPASS(wrq->adapter != NULL);
        s = sc->smt;


        if (chip_id(sc) <= CHELSIO_T5) {
                /* Source MAC Table (SMT) contains 256 SMAC entries
                 * organized in 128 rows of 2 entries each.
                 */
                req = start_wrq_wr(wrq, howmany(sizeof(*req), 16), &cookie);
                if (req == NULL)
                        return (ENOMEM);
                INIT_TP_WR(req, 0);
                /* Each row contains an SMAC pair.
                 * LSB selects the SMAC entry within a row
                 */
                row = (e->idx >> 1);
                if (e->idx & 1) {
                        req->pfvf1 = 0x0;
                        memcpy(req->src_mac1, e->smac, ETHER_ADDR_LEN);
                        /* fill pfvf0/src_mac0 with entry
                         * at prev index from smt-tab.
                         */
                        req->pfvf0 = 0x0;
                        memcpy(req->src_mac0, s->smtab[e->idx - 1].smac,
                                        ETHER_ADDR_LEN);
                } else {
                        req->pfvf0 = 0x0;
                        memcpy(req->src_mac0, e->smac, ETHER_ADDR_LEN);
                        /* fill pfvf1/src_mac1 with entry
                         * at next index from smt-tab
                         */
                        req->pfvf1 = 0x0;
                        memcpy(req->src_mac1, s->smtab[e->idx + 1].smac,
                                        ETHER_ADDR_LEN);
                }
        } else {
                /* Source MAC Table (SMT) contains 256 SMAC entries */
                t6req = start_wrq_wr(wrq, howmany(sizeof(*t6req), 16), &cookie);
                if (t6req == NULL)
                        return (ENOMEM);
                INIT_TP_WR(t6req, 0);
                req = (struct cpl_smt_write_req *)t6req;

                /* fill pfvf0/src_mac0 from smt-tab */
                req->pfvf0 = 0x0;
                memcpy(req->src_mac0, s->smtab[e->idx].smac, ETHER_ADDR_LEN);
                row = e->idx;
        }
        OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_SMT_WRITE_REQ, e->idx |
                                        V_TID_QID(e->iqid)));
        req->params = htonl(V_SMTW_NORPL(0) |
                        V_SMTW_IDX(row) |
                        V_SMTW_OVLAN_IDX(0));

        commit_wrq_wr(wrq, req, &cookie);

        return (0);
}

/*
 * Allocate an SMT entry for use by a switching rule.
 */
struct smt_entry *
t4_smt_alloc_switching(struct smt_data *s, uint8_t *smac)
{
        struct smt_entry *e;

        MPASS(s != NULL);
        rw_wlock(&s->lock);
        e = t4_find_or_alloc_sme(s, smac);
        rw_wunlock(&s->lock);
        return e;
}

/*
 * Sets/updates the contents of a switching SMT entry that has been allocated
 * with an earlier call to @t4_smt_alloc_switching.
 */
int
t4_smt_set_switching(struct adapter *sc, struct smt_entry *e, uint16_t pfvf,
                                                                uint8_t *smac)
{
        int rc = 0;

        if (atomic_load_acq_int(&e->refcnt) == 1) {
                /* Setup the entry for the first time */
                mtx_lock(&e->lock);
                e->wrq = &sc->sge.ctrlq[0];
                e->iqid = sc->sge.fwq.abs_id;
                e->pfvf =  pfvf;
                e->state = SMT_STATE_SWITCHING;
                memcpy(e->smac, smac, ETHER_ADDR_LEN);
                rc = t4_write_sme(e);
                mtx_unlock(&e->lock);
        }

        return (rc);
}

int
t4_init_smt(struct adapter *sc, int flags)
{
        int i, smt_size;
        struct smt_data *s;

        smt_size = SMT_SIZE;
        s = malloc(sizeof(*s) + smt_size * sizeof (struct smt_entry), M_CXGBE,
            M_ZERO | flags);
        if (!s)
                return (ENOMEM);

        s->smt_size = smt_size;
        rw_init(&s->lock, "SMT");

        for (i = 0; i < smt_size; i++) {
                struct smt_entry *e = &s->smtab[i];

                e->idx = i;
                e->state = SMT_STATE_UNUSED;
                mtx_init(&e->lock, "SMT_E", NULL, MTX_DEF);
                atomic_store_rel_int(&e->refcnt, 0);
        }

        sc->smt = s;

        return (0);
}

int
t4_free_smt(struct smt_data *s)
{
        int i;

        for (i = 0; i < s->smt_size; i++)
                mtx_destroy(&s->smtab[i].lock);
        rw_destroy(&s->lock);
        free(s, M_CXGBE);

        return (0);
}

int
do_smt_write_rpl(struct sge_iq *iq, const struct rss_header *rss,
                struct mbuf *m)
{
        struct adapter *sc = iq->adapter;
        const struct cpl_smt_write_rpl *rpl = (const void *)(rss + 1);
        unsigned int tid = GET_TID(rpl);
        unsigned int smtidx = G_TID_TID(tid);

        if (__predict_false(rpl->status != CPL_ERR_NONE)) {
                struct smt_entry *e = &sc->smt->smtab[smtidx];
                log(LOG_ERR,
                    "Unexpected SMT_WRITE_RPL (%u) for entry at hw_idx %u\n",
                    rpl->status, smtidx);
                mtx_lock(&e->lock);
                e->state = SMT_STATE_ERROR;
                mtx_unlock(&e->lock);
                return (EINVAL);
        }

        return (0);
}

static char
smt_state(const struct smt_entry *e)
{
        switch (e->state) {
        case SMT_STATE_SWITCHING: return 'X';
        case SMT_STATE_ERROR: return 'E';
        default: return 'U';
        }
}

int
sysctl_smt(SYSCTL_HANDLER_ARGS)
{
        struct adapter *sc = arg1;
        struct smt_data *smt = sc->smt;
        struct smt_entry *e;
        struct sbuf *sb;
        int rc, i, header = 0;

        if (smt == NULL)
                return (ENXIO);

        rc = sysctl_wire_old_buffer(req, 0);
        if (rc != 0)
                return (rc);

        sb = sbuf_new_for_sysctl(NULL, NULL, SMT_SIZE, req);
        if (sb == NULL)
                return (ENOMEM);

        e = &smt->smtab[0];
        for (i = 0; i < smt->smt_size; i++, e++) {
                mtx_lock(&e->lock);
                if (e->state == SMT_STATE_UNUSED)
                        goto skip;

                if (header == 0) {
                        sbuf_printf(sb, " Idx "
                            "Ethernet address  State Users");
                        header = 1;
                }
                sbuf_printf(sb, "\n%4u %02x:%02x:%02x:%02x:%02x:%02x "
                           "%c   %5u",
                           e->idx, e->smac[0], e->smac[1], e->smac[2],
                           e->smac[3], e->smac[4], e->smac[5],
                           smt_state(e), atomic_load_acq_int(&e->refcnt));
skip:
                mtx_unlock(&e->lock);
        }

        rc = sbuf_finish(sb);
        sbuf_delete(sb);

        return (rc);
}