X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=drivers%2Fnet%2Fmlx5%2Fmlx5_ethdev.c;h=d178ed6a18323cd4a7a90d2c6dbd93375fdd5964;hb=5d7b673d5fd6663b20c675dd382d9fb43b42af18;hp=6b64f4455cb0cd2870bcc2458f24c3a459013d89;hpb=859081d3fbbf81501726a5920749e47a2a65e9d3;p=dpdk.git diff --git a/drivers/net/mlx5/mlx5_ethdev.c b/drivers/net/mlx5/mlx5_ethdev.c index 6b64f4455c..d178ed6a18 100644 --- a/drivers/net/mlx5/mlx5_ethdev.c +++ b/drivers/net/mlx5/mlx5_ethdev.c @@ -1,38 +1,11 @@ -/*- - * BSD LICENSE - * - * Copyright 2015 6WIND S.A. - * Copyright 2015 Mellanox. - * - * 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 6WIND S.A. 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 BY THE COPYRIGHT HOLDERS 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 COPYRIGHT - * OWNER 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. +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright 2015 6WIND S.A. + * Copyright 2015 Mellanox Technologies, Ltd */ #include #include +#include #include #include #include @@ -43,32 +16,55 @@ #include #include #include -#include #include #include #include -#include #include +#include +#include +#include -/* DPDK headers don't like -pedantic. */ -#ifdef PEDANTIC -#pragma GCC diagnostic ignored "-Wpedantic" -#endif #include -#include +#include +#include #include #include #include -#include #include -#ifdef PEDANTIC -#pragma GCC diagnostic error "-Wpedantic" -#endif +#include +#include #include "mlx5.h" +#include "mlx5_glue.h" #include "mlx5_rxtx.h" #include "mlx5_utils.h" +/* Supported speed values found in /usr/include/linux/ethtool.h */ +#ifndef HAVE_SUPPORTED_40000baseKR4_Full +#define SUPPORTED_40000baseKR4_Full (1 << 23) +#endif +#ifndef HAVE_SUPPORTED_40000baseCR4_Full +#define SUPPORTED_40000baseCR4_Full (1 << 24) +#endif +#ifndef HAVE_SUPPORTED_40000baseSR4_Full +#define SUPPORTED_40000baseSR4_Full (1 << 25) +#endif +#ifndef HAVE_SUPPORTED_40000baseLR4_Full +#define SUPPORTED_40000baseLR4_Full (1 << 26) +#endif +#ifndef HAVE_SUPPORTED_56000baseKR4_Full +#define SUPPORTED_56000baseKR4_Full (1 << 27) +#endif +#ifndef HAVE_SUPPORTED_56000baseCR4_Full +#define SUPPORTED_56000baseCR4_Full (1 << 28) +#endif +#ifndef HAVE_SUPPORTED_56000baseSR4_Full +#define SUPPORTED_56000baseSR4_Full (1 << 29) +#endif +#ifndef HAVE_SUPPORTED_56000baseLR4_Full +#define SUPPORTED_56000baseLR4_Full (1 << 30) +#endif + /* Add defines in case the running kernel is not the same as user headers. */ #ifndef ETHTOOL_GLINKSETTINGS struct ethtool_link_settings { @@ -121,51 +117,21 @@ struct ethtool_link_settings { #endif /** - * Return private structure associated with an Ethernet device. - * - * @param dev - * Pointer to Ethernet device structure. - * - * @return - * Pointer to private structure. - */ -struct priv * -mlx5_get_priv(struct rte_eth_dev *dev) -{ - struct mlx5_secondary_data *sd; - - if (!mlx5_is_secondary()) - return dev->data->dev_private; - sd = &mlx5_secondary_data[dev->data->port_id]; - return sd->data.dev_private; -} - -/** - * Check if running as a secondary process. + * Get master interface name from private structure. * - * @return - * Nonzero if running as a secondary process. - */ -inline int -mlx5_is_secondary(void) -{ - return rte_eal_process_type() != RTE_PROC_PRIMARY; -} - -/** - * Get interface name from private structure. - * - * @param[in] priv - * Pointer to private structure. + * @param[in] dev + * Pointer to Ethernet device. * @param[out] ifname * Interface name output buffer. * * @return - * 0 on success, -1 on failure and errno is set. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ -int -priv_get_ifname(const struct priv *priv, char (*ifname)[IF_NAMESIZE]) +static int +mlx5_get_master_ifname(const struct rte_eth_dev *dev, + char (*ifname)[IF_NAMESIZE]) { + struct priv *priv = dev->data->dev_private; DIR *dir; struct dirent *dent; unsigned int dev_type = 0; @@ -173,11 +139,13 @@ priv_get_ifname(const struct priv *priv, char (*ifname)[IF_NAMESIZE]) char match[IF_NAMESIZE] = ""; { - MKSTR(path, "%s/device/net", priv->ctx->device->ibdev_path); + MKSTR(path, "%s/device/net", priv->ibdev_path); dir = opendir(path); - if (dir == NULL) - return -1; + if (dir == NULL) { + rte_errno = errno; + return -rte_errno; + } } while ((dent = readdir(dir)) != NULL) { char *name = dent->d_name; @@ -191,7 +159,7 @@ priv_get_ifname(const struct priv *priv, char (*ifname)[IF_NAMESIZE]) continue; MKSTR(path, "%s/device/net/%s/%s", - priv->ctx->device->ibdev_path, name, + priv->ibdev_path, name, (dev_type ? "dev_id" : "dev_port")); file = fopen(path, "rb"); @@ -223,360 +191,191 @@ try_dev_id: if (dev_port == dev_port_prev) goto try_dev_id; dev_port_prev = dev_port; - if (dev_port == (priv->port - 1u)) - snprintf(match, sizeof(match), "%s", name); + if (dev_port == 0) + strlcpy(match, name, sizeof(match)); } closedir(dir); - if (match[0] == '\0') - return -1; + if (match[0] == '\0') { + rte_errno = ENOENT; + return -rte_errno; + } strncpy(*ifname, match, sizeof(*ifname)); return 0; } /** - * Check if the counter is located on ib counters file. - * - * @param[in] cntr - * Counter name. + * Get interface name from private structure. * - * @return - * 1 if counter is located on ib counters file , 0 otherwise. - */ -int -priv_is_ib_cntr(const char *cntr) -{ - if (!strcmp(cntr, "out_of_buffer")) - return 1; - return 0; -} - -/** - * Read from sysfs entry. + * This is a port representor-aware version of mlx5_get_master_ifname(). * - * @param[in] priv - * Pointer to private structure. - * @param[in] entry - * Entry name relative to sysfs path. - * @param[out] buf - * Data output buffer. - * @param size - * Buffer size. + * @param[in] dev + * Pointer to Ethernet device. + * @param[out] ifname + * Interface name output buffer. * * @return - * 0 on success, -1 on failure and errno is set. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ -static int -priv_sysfs_read(const struct priv *priv, const char *entry, - char *buf, size_t size) +int +mlx5_get_ifname(const struct rte_eth_dev *dev, char (*ifname)[IF_NAMESIZE]) { - char ifname[IF_NAMESIZE]; - FILE *file; - int ret; - int err; - - if (priv_get_ifname(priv, &ifname)) - return -1; - - if (priv_is_ib_cntr(entry)) { - MKSTR(path, "%s/ports/1/hw_counters/%s", - priv->ctx->device->ibdev_path, entry); - file = fopen(path, "rb"); - } else { - MKSTR(path, "%s/device/net/%s/%s", - priv->ctx->device->ibdev_path, ifname, entry); - file = fopen(path, "rb"); + struct priv *priv = dev->data->dev_private; + unsigned int ifindex = + priv->nl_socket_rdma >= 0 ? + mlx5_nl_ifindex(priv->nl_socket_rdma, priv->ibdev_name) : 0; + + if (!ifindex) { + if (!priv->representor) + return mlx5_get_master_ifname(dev, ifname); + rte_errno = ENXIO; + return -rte_errno; } - if (file == NULL) - return -1; - ret = fread(buf, 1, size, file); - err = errno; - if (((size_t)ret < size) && (ferror(file))) - ret = -1; - else - ret = size; - fclose(file); - errno = err; - return ret; + if (if_indextoname(ifindex, &(*ifname)[0])) + return 0; + rte_errno = errno; + return -rte_errno; } /** - * Write to sysfs entry. + * Get the interface index from device name. * - * @param[in] priv - * Pointer to private structure. - * @param[in] entry - * Entry name relative to sysfs path. - * @param[in] buf - * Data buffer. - * @param size - * Buffer size. + * @param[in] dev + * Pointer to Ethernet device. * * @return - * 0 on success, -1 on failure and errno is set. + * Nonzero interface index on success, zero otherwise and rte_errno is set. */ -static int -priv_sysfs_write(const struct priv *priv, const char *entry, - char *buf, size_t size) +unsigned int +mlx5_ifindex(const struct rte_eth_dev *dev) { char ifname[IF_NAMESIZE]; - FILE *file; - int ret; - int err; - - if (priv_get_ifname(priv, &ifname)) - return -1; - - MKSTR(path, "%s/device/net/%s/%s", priv->ctx->device->ibdev_path, - ifname, entry); - - file = fopen(path, "wb"); - if (file == NULL) - return -1; - ret = fwrite(buf, 1, size, file); - err = errno; - if (((size_t)ret < size) || (ferror(file))) - ret = -1; - else - ret = size; - fclose(file); - errno = err; - return ret; -} - -/** - * Get unsigned long sysfs property. - * - * @param priv - * Pointer to private structure. - * @param[in] name - * Entry name relative to sysfs path. - * @param[out] value - * Value output buffer. - * - * @return - * 0 on success, -1 on failure and errno is set. - */ -static int -priv_get_sysfs_ulong(struct priv *priv, const char *name, unsigned long *value) -{ - int ret; - unsigned long value_ret; - char value_str[32]; - - ret = priv_sysfs_read(priv, name, value_str, (sizeof(value_str) - 1)); - if (ret == -1) { - DEBUG("cannot read %s value from sysfs: %s", - name, strerror(errno)); - return -1; - } - value_str[ret] = '\0'; - errno = 0; - value_ret = strtoul(value_str, NULL, 0); - if (errno) { - DEBUG("invalid %s value `%s': %s", name, value_str, - strerror(errno)); - return -1; - } - *value = value_ret; - return 0; -} - -/** - * Set unsigned long sysfs property. - * - * @param priv - * Pointer to private structure. - * @param[in] name - * Entry name relative to sysfs path. - * @param value - * Value to set. - * - * @return - * 0 on success, -1 on failure and errno is set. - */ -static int -priv_set_sysfs_ulong(struct priv *priv, const char *name, unsigned long value) -{ - int ret; - MKSTR(value_str, "%lu", value); + unsigned int ifindex; - ret = priv_sysfs_write(priv, name, value_str, (sizeof(value_str) - 1)); - if (ret == -1) { - DEBUG("cannot write %s `%s' (%lu) to sysfs: %s", - name, value_str, value, strerror(errno)); - return -1; - } - return 0; + if (mlx5_get_ifname(dev, &ifname)) + return 0; + ifindex = if_nametoindex(ifname); + if (!ifindex) + rte_errno = errno; + return ifindex; } /** * Perform ifreq ioctl() on associated Ethernet device. * - * @param[in] priv - * Pointer to private structure. + * @param[in] dev + * Pointer to Ethernet device. * @param req * Request number to pass to ioctl(). * @param[out] ifr * Interface request structure output buffer. * * @return - * 0 on success, -1 on failure and errno is set. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ int -priv_ifreq(const struct priv *priv, int req, struct ifreq *ifr) +mlx5_ifreq(const struct rte_eth_dev *dev, int req, struct ifreq *ifr) { int sock = socket(PF_INET, SOCK_DGRAM, IPPROTO_IP); - int ret = -1; + int ret = 0; - if (sock == -1) - return ret; - if (priv_get_ifname(priv, &ifr->ifr_name) == 0) - ret = ioctl(sock, req, ifr); + if (sock == -1) { + rte_errno = errno; + return -rte_errno; + } + ret = mlx5_get_ifname(dev, &ifr->ifr_name); + if (ret) + goto error; + ret = ioctl(sock, req, ifr); + if (ret == -1) { + rte_errno = errno; + goto error; + } close(sock); - return ret; -} - -/** - * Return the number of active VFs for the current device. - * - * @param[in] priv - * Pointer to private structure. - * @param[out] num_vfs - * Number of active VFs. - * - * @return - * 0 on success, -1 on failure and errno is set. - */ -int -priv_get_num_vfs(struct priv *priv, uint16_t *num_vfs) -{ - /* The sysfs entry name depends on the operating system. */ - const char **name = (const char *[]){ - "device/sriov_numvfs", - "device/mlx5_num_vfs", - NULL, - }; - int ret; - - do { - unsigned long ulong_num_vfs; - - ret = priv_get_sysfs_ulong(priv, *name, &ulong_num_vfs); - if (!ret) - *num_vfs = ulong_num_vfs; - } while (*(++name) && ret); - return ret; + return 0; +error: + close(sock); + return -rte_errno; } /** * Get device MTU. * - * @param priv - * Pointer to private structure. + * @param dev + * Pointer to Ethernet device. * @param[out] mtu * MTU value output buffer. * * @return - * 0 on success, -1 on failure and errno is set. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ int -priv_get_mtu(struct priv *priv, uint16_t *mtu) +mlx5_get_mtu(struct rte_eth_dev *dev, uint16_t *mtu) { - unsigned long ulong_mtu; - - if (priv_get_sysfs_ulong(priv, "mtu", &ulong_mtu) == -1) - return -1; - *mtu = ulong_mtu; - return 0; -} + struct ifreq request; + int ret = mlx5_ifreq(dev, SIOCGIFMTU, &request); -/** - * Read device counter from sysfs. - * - * @param priv - * Pointer to private structure. - * @param name - * Counter name. - * @param[out] cntr - * Counter output buffer. - * - * @return - * 0 on success, -1 on failure and errno is set. - */ -int -priv_get_cntr_sysfs(struct priv *priv, const char *name, uint64_t *cntr) -{ - unsigned long ulong_ctr; - - if (priv_get_sysfs_ulong(priv, name, &ulong_ctr) == -1) - return -1; - *cntr = ulong_ctr; + if (ret) + return ret; + *mtu = request.ifr_mtu; return 0; } /** * Set device MTU. * - * @param priv - * Pointer to private structure. + * @param dev + * Pointer to Ethernet device. * @param mtu * MTU value to set. * * @return - * 0 on success, -1 on failure and errno is set. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ static int -priv_set_mtu(struct priv *priv, uint16_t mtu) +mlx5_set_mtu(struct rte_eth_dev *dev, uint16_t mtu) { - uint16_t new_mtu; + struct ifreq request = { .ifr_mtu = mtu, }; - if (priv_set_sysfs_ulong(priv, "mtu", mtu) || - priv_get_mtu(priv, &new_mtu)) - return -1; - if (new_mtu == mtu) - return 0; - errno = EINVAL; - return -1; + return mlx5_ifreq(dev, SIOCSIFMTU, &request); } /** * Set device flags. * - * @param priv - * Pointer to private structure. + * @param dev + * Pointer to Ethernet device. * @param keep * Bitmask for flags that must remain untouched. * @param flags * Bitmask for flags to modify. * * @return - * 0 on success, -1 on failure and errno is set. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ int -priv_set_flags(struct priv *priv, unsigned int keep, unsigned int flags) +mlx5_set_flags(struct rte_eth_dev *dev, unsigned int keep, unsigned int flags) { - unsigned long tmp; + struct ifreq request; + int ret = mlx5_ifreq(dev, SIOCGIFFLAGS, &request); - if (priv_get_sysfs_ulong(priv, "flags", &tmp) == -1) - return -1; - tmp &= keep; - tmp |= (flags & (~keep)); - return priv_set_sysfs_ulong(priv, "flags", tmp); + if (ret) + return ret; + request.ifr_flags &= keep; + request.ifr_flags |= flags & ~keep; + return mlx5_ifreq(dev, SIOCSIFFLAGS, &request); } /** - * Ethernet device configuration. - * - * Prepare the driver for a given number of TX and RX queues. + * DPDK callback for Ethernet device configuration. * * @param dev * Pointer to Ethernet device structure. * * @return - * 0 on success, errno value on failure. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ -static int -dev_configure(struct rte_eth_dev *dev) +int +mlx5_dev_configure(struct rte_eth_dev *dev) { struct priv *priv = dev->data->dev_private; unsigned int rxqs_n = dev->data->nb_rx_queues; @@ -584,32 +383,61 @@ dev_configure(struct rte_eth_dev *dev) unsigned int i; unsigned int j; unsigned int reta_idx_n; + const uint8_t use_app_rss_key = + !!dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key; + int ret = 0; - priv->rss_hf = dev->data->dev_conf.rx_adv_conf.rss_conf.rss_hf; + if (use_app_rss_key && + (dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key_len != + MLX5_RSS_HASH_KEY_LEN)) { + DRV_LOG(ERR, "port %u RSS key len must be %s Bytes long", + dev->data->port_id, RTE_STR(MLX5_RSS_HASH_KEY_LEN)); + rte_errno = EINVAL; + return -rte_errno; + } + priv->rss_conf.rss_key = + rte_realloc(priv->rss_conf.rss_key, + MLX5_RSS_HASH_KEY_LEN, 0); + if (!priv->rss_conf.rss_key) { + DRV_LOG(ERR, "port %u cannot allocate RSS hash key memory (%u)", + dev->data->port_id, rxqs_n); + rte_errno = ENOMEM; + return -rte_errno; + } + memcpy(priv->rss_conf.rss_key, + use_app_rss_key ? + dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key : + rss_hash_default_key, + MLX5_RSS_HASH_KEY_LEN); + priv->rss_conf.rss_key_len = MLX5_RSS_HASH_KEY_LEN; + priv->rss_conf.rss_hf = dev->data->dev_conf.rx_adv_conf.rss_conf.rss_hf; priv->rxqs = (void *)dev->data->rx_queues; priv->txqs = (void *)dev->data->tx_queues; if (txqs_n != priv->txqs_n) { - INFO("%p: TX queues number update: %u -> %u", - (void *)dev, priv->txqs_n, txqs_n); + DRV_LOG(INFO, "port %u Tx queues number update: %u -> %u", + dev->data->port_id, priv->txqs_n, txqs_n); priv->txqs_n = txqs_n; } - if (rxqs_n > priv->ind_table_max_size) { - ERROR("cannot handle this many RX queues (%u)", rxqs_n); - return EINVAL; + if (rxqs_n > priv->config.ind_table_max_size) { + DRV_LOG(ERR, "port %u cannot handle this many Rx queues (%u)", + dev->data->port_id, rxqs_n); + rte_errno = EINVAL; + return -rte_errno; } if (rxqs_n == priv->rxqs_n) return 0; - INFO("%p: RX queues number update: %u -> %u", - (void *)dev, priv->rxqs_n, rxqs_n); + DRV_LOG(INFO, "port %u Rx queues number update: %u -> %u", + dev->data->port_id, priv->rxqs_n, rxqs_n); priv->rxqs_n = rxqs_n; /* If the requested number of RX queues is not a power of two, use the * maximum indirection table size for better balancing. * The result is always rounded to the next power of two. */ reta_idx_n = (1 << log2above((rxqs_n & (rxqs_n - 1)) ? - priv->ind_table_max_size : + priv->config.ind_table_max_size : rxqs_n)); - if (priv_rss_reta_index_resize(priv, reta_idx_n)) - return ENOMEM; + ret = mlx5_rss_reta_index_resize(dev, reta_idx_n); + if (ret) + return ret; /* When the number of RX queues is not a power of two, the remaining * table entries are padded with reused WQs and hashes are not spread * uniformly. */ @@ -622,28 +450,42 @@ dev_configure(struct rte_eth_dev *dev) } /** - * DPDK callback for Ethernet device configuration. + * Sets default tuning parameters. * * @param dev - * Pointer to Ethernet device structure. - * - * @return - * 0 on success, negative errno value on failure. + * Pointer to Ethernet device. + * @param[out] info + * Info structure output buffer. */ -int -mlx5_dev_configure(struct rte_eth_dev *dev) +static void +mlx5_set_default_params(struct rte_eth_dev *dev, struct rte_eth_dev_info *info) { struct priv *priv = dev->data->dev_private; - int ret; - if (mlx5_is_secondary()) - return -E_RTE_SECONDARY; - - priv_lock(priv); - ret = dev_configure(dev); - assert(ret >= 0); - priv_unlock(priv); - return -ret; + /* Minimum CPU utilization. */ + info->default_rxportconf.ring_size = 256; + info->default_txportconf.ring_size = 256; + info->default_rxportconf.burst_size = 64; + info->default_txportconf.burst_size = 64; + if (priv->link_speed_capa & ETH_LINK_SPEED_100G) { + info->default_rxportconf.nb_queues = 16; + info->default_txportconf.nb_queues = 16; + if (dev->data->nb_rx_queues > 2 || + dev->data->nb_tx_queues > 2) { + /* Max Throughput. */ + info->default_rxportconf.ring_size = 2048; + info->default_txportconf.ring_size = 2048; + } + } else { + info->default_rxportconf.nb_queues = 8; + info->default_txportconf.nb_queues = 8; + if (dev->data->nb_rx_queues > 2 || + dev->data->nb_tx_queues > 2) { + /* Max Throughput. */ + info->default_rxportconf.ring_size = 4096; + info->default_txportconf.ring_size = 4096; + } + } } /** @@ -657,13 +499,11 @@ mlx5_dev_configure(struct rte_eth_dev *dev) void mlx5_dev_infos_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *info) { - struct priv *priv = mlx5_get_priv(dev); + struct priv *priv = dev->data->dev_private; + struct mlx5_dev_config *config = &priv->config; unsigned int max; char ifname[IF_NAMESIZE]; - info->pci_dev = RTE_DEV_TO_PCI(dev->device); - - priv_lock(priv); /* FIXME: we should ask the device for these values. */ info->min_rx_bufsize = 32; info->max_rx_pktlen = 65536; @@ -671,57 +511,85 @@ mlx5_dev_infos_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *info) * Since we need one CQ per QP, the limit is the minimum number * between the two values. */ - max = ((priv->device_attr.max_cq > priv->device_attr.max_qp) ? - priv->device_attr.max_qp : priv->device_attr.max_cq); + max = RTE_MIN(priv->device_attr.orig_attr.max_cq, + priv->device_attr.orig_attr.max_qp); /* If max >= 65535 then max = 0, max_rx_queues is uint16_t. */ if (max >= 65535) max = 65535; info->max_rx_queues = max; info->max_tx_queues = max; - info->max_mac_addrs = RTE_DIM(priv->mac); - info->rx_offload_capa = - (priv->hw_csum ? - (DEV_RX_OFFLOAD_IPV4_CKSUM | - DEV_RX_OFFLOAD_UDP_CKSUM | - DEV_RX_OFFLOAD_TCP_CKSUM) : - 0) | - (priv->hw_vlan_strip ? DEV_RX_OFFLOAD_VLAN_STRIP : 0); - if (!priv->mps) - info->tx_offload_capa = DEV_TX_OFFLOAD_VLAN_INSERT; - if (priv->hw_csum) - info->tx_offload_capa |= - (DEV_TX_OFFLOAD_IPV4_CKSUM | - DEV_TX_OFFLOAD_UDP_CKSUM | - DEV_TX_OFFLOAD_TCP_CKSUM); - if (priv_get_ifname(priv, &ifname) == 0) + info->max_mac_addrs = MLX5_MAX_UC_MAC_ADDRESSES; + info->rx_queue_offload_capa = mlx5_get_rx_queue_offloads(dev); + info->rx_offload_capa = (mlx5_get_rx_port_offloads() | + info->rx_queue_offload_capa); + info->tx_offload_capa = mlx5_get_tx_port_offloads(dev); + if (mlx5_get_ifname(dev, &ifname) == 0) info->if_index = if_nametoindex(ifname); - /* FIXME: RETA update/query API expects the callee to know the size of - * the indirection table, for this PMD the size varies depending on - * the number of RX queues, it becomes impossible to find the correct - * size if it is not fixed. - * The API should be updated to solve this problem. */ - info->reta_size = priv->ind_table_max_size; - info->hash_key_size = ((*priv->rss_conf) ? - (*priv->rss_conf)[0]->rss_key_len : - 0); + info->reta_size = priv->reta_idx_n ? + priv->reta_idx_n : config->ind_table_max_size; + info->hash_key_size = MLX5_RSS_HASH_KEY_LEN; info->speed_capa = priv->link_speed_capa; - priv_unlock(priv); + info->flow_type_rss_offloads = ~MLX5_RSS_HF_MASK; + mlx5_set_default_params(dev, info); + info->switch_info.name = dev->data->name; + info->switch_info.domain_id = priv->domain_id; + info->switch_info.port_id = priv->representor_id; + if (priv->representor) { + unsigned int i = mlx5_dev_to_port_id(dev->device, NULL, 0); + uint16_t port_id[i]; + + i = RTE_MIN(mlx5_dev_to_port_id(dev->device, port_id, i), i); + while (i--) { + struct priv *opriv = + rte_eth_devices[port_id[i]].data->dev_private; + + if (!opriv || + opriv->representor || + opriv->domain_id != priv->domain_id) + continue; + /* + * Override switch name with that of the master + * device. + */ + info->switch_info.name = opriv->dev_data->name; + break; + } + } } +/** + * Get supported packet types. + * + * @param dev + * Pointer to Ethernet device structure. + * + * @return + * A pointer to the supported Packet types array. + */ const uint32_t * mlx5_dev_supported_ptypes_get(struct rte_eth_dev *dev) { static const uint32_t ptypes[] = { /* refers to rxq_cq_to_pkt_type() */ - RTE_PTYPE_L3_IPV4, - RTE_PTYPE_L3_IPV6, - RTE_PTYPE_INNER_L3_IPV4, - RTE_PTYPE_INNER_L3_IPV6, + RTE_PTYPE_L2_ETHER, + RTE_PTYPE_L3_IPV4_EXT_UNKNOWN, + RTE_PTYPE_L3_IPV6_EXT_UNKNOWN, + RTE_PTYPE_L4_NONFRAG, + RTE_PTYPE_L4_FRAG, + RTE_PTYPE_L4_TCP, + RTE_PTYPE_L4_UDP, + RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN, + RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN, + RTE_PTYPE_INNER_L4_NONFRAG, + RTE_PTYPE_INNER_L4_FRAG, + RTE_PTYPE_INNER_L4_TCP, + RTE_PTYPE_INNER_L4_UDP, RTE_PTYPE_UNKNOWN - }; - if (dev->rx_pkt_burst == mlx5_rx_burst) + if (dev->rx_pkt_burst == mlx5_rx_burst || + dev->rx_pkt_burst == mlx5_rx_burst_mprq || + dev->rx_pkt_burst == mlx5_rx_burst_vec) return ptypes; return NULL; } @@ -731,39 +599,48 @@ mlx5_dev_supported_ptypes_get(struct rte_eth_dev *dev) * * @param dev * Pointer to Ethernet device structure. - * @param wait_to_complete - * Wait for request completion (ignored). + * @param[out] link + * Storage for current link status. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. */ static int -mlx5_link_update_unlocked_gset(struct rte_eth_dev *dev, int wait_to_complete) +mlx5_link_update_unlocked_gset(struct rte_eth_dev *dev, + struct rte_eth_link *link) { - struct priv *priv = mlx5_get_priv(dev); + struct priv *priv = dev->data->dev_private; struct ethtool_cmd edata = { .cmd = ETHTOOL_GSET /* Deprecated since Linux v4.5. */ }; struct ifreq ifr; struct rte_eth_link dev_link; int link_speed = 0; + int ret; - /* priv_lock() is not taken to allow concurrent calls. */ - - (void)wait_to_complete; - if (priv_ifreq(priv, SIOCGIFFLAGS, &ifr)) { - WARN("ioctl(SIOCGIFFLAGS) failed: %s", strerror(errno)); - return -1; + ret = mlx5_ifreq(dev, SIOCGIFFLAGS, &ifr); + if (ret) { + DRV_LOG(WARNING, "port %u ioctl(SIOCGIFFLAGS) failed: %s", + dev->data->port_id, strerror(rte_errno)); + return ret; } - memset(&dev_link, 0, sizeof(dev_link)); - dev_link.link_status = ((ifr.ifr_flags & IFF_UP) && - (ifr.ifr_flags & IFF_RUNNING)); - ifr.ifr_data = (void *)&edata; - if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) { - WARN("ioctl(SIOCETHTOOL, ETHTOOL_GSET) failed: %s", - strerror(errno)); - return -1; + dev_link = (struct rte_eth_link) { + .link_status = ((ifr.ifr_flags & IFF_UP) && + (ifr.ifr_flags & IFF_RUNNING)), + }; + ifr = (struct ifreq) { + .ifr_data = (void *)&edata, + }; + ret = mlx5_ifreq(dev, SIOCETHTOOL, &ifr); + if (ret) { + DRV_LOG(WARNING, + "port %u ioctl(SIOCETHTOOL, ETHTOOL_GSET) failed: %s", + dev->data->port_id, strerror(rte_errno)); + return ret; } link_speed = ethtool_cmd_speed(&edata); if (link_speed == -1) - dev_link.link_speed = 0; + dev_link.link_speed = ETH_SPEED_NUM_NONE; else dev_link.link_speed = link_speed; priv->link_speed_capa = 0; @@ -783,13 +660,13 @@ mlx5_link_update_unlocked_gset(struct rte_eth_dev *dev, int wait_to_complete) ETH_LINK_HALF_DUPLEX : ETH_LINK_FULL_DUPLEX); dev_link.link_autoneg = !(dev->data->dev_conf.link_speeds & ETH_LINK_SPEED_FIXED); - if (memcmp(&dev_link, &dev->data->dev_link, sizeof(dev_link))) { - /* Link status changed. */ - dev->data->dev_link = dev_link; - return 0; + if (((dev_link.link_speed && !dev_link.link_status) || + (!dev_link.link_speed && dev_link.link_status))) { + rte_errno = EAGAIN; + return -rte_errno; } - /* Link status is still the same. */ - return -1; + *link = dev_link; + return 0; } /** @@ -797,83 +674,111 @@ mlx5_link_update_unlocked_gset(struct rte_eth_dev *dev, int wait_to_complete) * * @param dev * Pointer to Ethernet device structure. - * @param wait_to_complete - * Wait for request completion (ignored). + * @param[out] link + * Storage for current link status. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. */ static int -mlx5_link_update_unlocked_gs(struct rte_eth_dev *dev, int wait_to_complete) +mlx5_link_update_unlocked_gs(struct rte_eth_dev *dev, + struct rte_eth_link *link) + { - struct priv *priv = mlx5_get_priv(dev); - struct ethtool_link_settings edata = { - .cmd = ETHTOOL_GLINKSETTINGS, - }; + struct priv *priv = dev->data->dev_private; + struct ethtool_link_settings gcmd = { .cmd = ETHTOOL_GLINKSETTINGS }; struct ifreq ifr; struct rte_eth_link dev_link; uint64_t sc; + int ret; - (void)wait_to_complete; - if (priv_ifreq(priv, SIOCGIFFLAGS, &ifr)) { - WARN("ioctl(SIOCGIFFLAGS) failed: %s", strerror(errno)); - return -1; + ret = mlx5_ifreq(dev, SIOCGIFFLAGS, &ifr); + if (ret) { + DRV_LOG(WARNING, "port %u ioctl(SIOCGIFFLAGS) failed: %s", + dev->data->port_id, strerror(rte_errno)); + return ret; + } + dev_link = (struct rte_eth_link) { + .link_status = ((ifr.ifr_flags & IFF_UP) && + (ifr.ifr_flags & IFF_RUNNING)), + }; + ifr = (struct ifreq) { + .ifr_data = (void *)&gcmd, + }; + ret = mlx5_ifreq(dev, SIOCETHTOOL, &ifr); + if (ret) { + DRV_LOG(DEBUG, + "port %u ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS)" + " failed: %s", + dev->data->port_id, strerror(rte_errno)); + return ret; } - memset(&dev_link, 0, sizeof(dev_link)); - dev_link.link_status = ((ifr.ifr_flags & IFF_UP) && - (ifr.ifr_flags & IFF_RUNNING)); - ifr.ifr_data = (void *)&edata; - if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) { - DEBUG("ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS) failed: %s", - strerror(errno)); - return -1; + gcmd.link_mode_masks_nwords = -gcmd.link_mode_masks_nwords; + + alignas(struct ethtool_link_settings) + uint8_t data[offsetof(struct ethtool_link_settings, link_mode_masks) + + sizeof(uint32_t) * gcmd.link_mode_masks_nwords * 3]; + struct ethtool_link_settings *ecmd = (void *)data; + + *ecmd = gcmd; + ifr.ifr_data = (void *)ecmd; + ret = mlx5_ifreq(dev, SIOCETHTOOL, &ifr); + if (ret) { + DRV_LOG(DEBUG, + "port %u ioctl(SIOCETHTOOL, ETHTOOL_GLINKSETTINGS)" + " failed: %s", + dev->data->port_id, strerror(rte_errno)); + return ret; } - dev_link.link_speed = edata.speed; - sc = edata.link_mode_masks[0] | - ((uint64_t)edata.link_mode_masks[1] << 32); + dev_link.link_speed = ecmd->speed; + sc = ecmd->link_mode_masks[0] | + ((uint64_t)ecmd->link_mode_masks[1] << 32); priv->link_speed_capa = 0; - if (sc & ETHTOOL_LINK_MODE_Autoneg_BIT) + if (sc & MLX5_BITSHIFT(ETHTOOL_LINK_MODE_Autoneg_BIT)) priv->link_speed_capa |= ETH_LINK_SPEED_AUTONEG; - if (sc & (ETHTOOL_LINK_MODE_1000baseT_Full_BIT | - ETHTOOL_LINK_MODE_1000baseKX_Full_BIT)) + if (sc & (MLX5_BITSHIFT(ETHTOOL_LINK_MODE_1000baseT_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_1000baseKX_Full_BIT))) priv->link_speed_capa |= ETH_LINK_SPEED_1G; - if (sc & (ETHTOOL_LINK_MODE_10000baseKX4_Full_BIT | - ETHTOOL_LINK_MODE_10000baseKR_Full_BIT | - ETHTOOL_LINK_MODE_10000baseR_FEC_BIT)) + if (sc & (MLX5_BITSHIFT(ETHTOOL_LINK_MODE_10000baseKX4_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_10000baseKR_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_10000baseR_FEC_BIT))) priv->link_speed_capa |= ETH_LINK_SPEED_10G; - if (sc & (ETHTOOL_LINK_MODE_20000baseMLD2_Full_BIT | - ETHTOOL_LINK_MODE_20000baseKR2_Full_BIT)) + if (sc & (MLX5_BITSHIFT(ETHTOOL_LINK_MODE_20000baseMLD2_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_20000baseKR2_Full_BIT))) priv->link_speed_capa |= ETH_LINK_SPEED_20G; - if (sc & (ETHTOOL_LINK_MODE_40000baseKR4_Full_BIT | - ETHTOOL_LINK_MODE_40000baseCR4_Full_BIT | - ETHTOOL_LINK_MODE_40000baseSR4_Full_BIT | - ETHTOOL_LINK_MODE_40000baseLR4_Full_BIT)) + if (sc & (MLX5_BITSHIFT(ETHTOOL_LINK_MODE_40000baseKR4_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_40000baseCR4_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_40000baseSR4_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_40000baseLR4_Full_BIT))) priv->link_speed_capa |= ETH_LINK_SPEED_40G; - if (sc & (ETHTOOL_LINK_MODE_56000baseKR4_Full_BIT | - ETHTOOL_LINK_MODE_56000baseCR4_Full_BIT | - ETHTOOL_LINK_MODE_56000baseSR4_Full_BIT | - ETHTOOL_LINK_MODE_56000baseLR4_Full_BIT)) + if (sc & (MLX5_BITSHIFT(ETHTOOL_LINK_MODE_56000baseKR4_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_56000baseCR4_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_56000baseSR4_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_56000baseLR4_Full_BIT))) priv->link_speed_capa |= ETH_LINK_SPEED_56G; - if (sc & (ETHTOOL_LINK_MODE_25000baseCR_Full_BIT | - ETHTOOL_LINK_MODE_25000baseKR_Full_BIT | - ETHTOOL_LINK_MODE_25000baseSR_Full_BIT)) + if (sc & (MLX5_BITSHIFT(ETHTOOL_LINK_MODE_25000baseCR_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_25000baseKR_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_25000baseSR_Full_BIT))) priv->link_speed_capa |= ETH_LINK_SPEED_25G; - if (sc & (ETHTOOL_LINK_MODE_50000baseCR2_Full_BIT | - ETHTOOL_LINK_MODE_50000baseKR2_Full_BIT)) + if (sc & (MLX5_BITSHIFT(ETHTOOL_LINK_MODE_50000baseCR2_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_50000baseKR2_Full_BIT))) priv->link_speed_capa |= ETH_LINK_SPEED_50G; - if (sc & (ETHTOOL_LINK_MODE_100000baseKR4_Full_BIT | - ETHTOOL_LINK_MODE_100000baseSR4_Full_BIT | - ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT | - ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT)) + if (sc & (MLX5_BITSHIFT(ETHTOOL_LINK_MODE_100000baseKR4_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_100000baseSR4_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_100000baseCR4_Full_BIT) | + MLX5_BITSHIFT(ETHTOOL_LINK_MODE_100000baseLR4_ER4_Full_BIT))) priv->link_speed_capa |= ETH_LINK_SPEED_100G; - dev_link.link_duplex = ((edata.duplex == DUPLEX_HALF) ? + dev_link.link_duplex = ((ecmd->duplex == DUPLEX_HALF) ? ETH_LINK_HALF_DUPLEX : ETH_LINK_FULL_DUPLEX); dev_link.link_autoneg = !(dev->data->dev_conf.link_speeds & ETH_LINK_SPEED_FIXED); - if (memcmp(&dev_link, &dev->data->dev_link, sizeof(dev_link))) { - /* Link status changed. */ - dev->data->dev_link = dev_link; - return 0; + if (((dev_link.link_speed && !dev_link.link_status) || + (!dev_link.link_speed && dev_link.link_status))) { + rte_errno = EAGAIN; + return -rte_errno; } - /* Link status is still the same. */ - return -1; + *link = dev_link; + return 0; } /** @@ -882,165 +787,81 @@ mlx5_link_update_unlocked_gs(struct rte_eth_dev *dev, int wait_to_complete) * @param dev * Pointer to Ethernet device structure. * @param wait_to_complete - * Wait for request completion (ignored). + * Wait for request completion. + * + * @return + * 0 if link status was not updated, positive if it was, a negative errno + * value otherwise and rte_errno is set. */ int mlx5_link_update(struct rte_eth_dev *dev, int wait_to_complete) { - struct utsname utsname; - int ver[3]; - - if (uname(&utsname) == -1 || - sscanf(utsname.release, "%d.%d.%d", - &ver[0], &ver[1], &ver[2]) != 3 || - KERNEL_VERSION(ver[0], ver[1], ver[2]) < KERNEL_VERSION(4, 9, 0)) - return mlx5_link_update_unlocked_gset(dev, wait_to_complete); - return mlx5_link_update_unlocked_gs(dev, wait_to_complete); + int ret; + struct rte_eth_link dev_link; + time_t start_time = time(NULL); + + do { + ret = mlx5_link_update_unlocked_gs(dev, &dev_link); + if (ret) + ret = mlx5_link_update_unlocked_gset(dev, &dev_link); + if (ret == 0) + break; + /* Handle wait to complete situation. */ + if (wait_to_complete && ret == -EAGAIN) { + if (abs((int)difftime(time(NULL), start_time)) < + MLX5_LINK_STATUS_TIMEOUT) { + usleep(0); + continue; + } else { + rte_errno = EBUSY; + return -rte_errno; + } + } else if (ret < 0) { + return ret; + } + } while (wait_to_complete); + ret = !!memcmp(&dev->data->dev_link, &dev_link, + sizeof(struct rte_eth_link)); + dev->data->dev_link = dev_link; + return ret; } /** * DPDK callback to change the MTU. * - * Setting the MTU affects hardware MRU (packets larger than the MTU cannot be - * received). Use this as a hint to enable/disable scattered packets support - * and improve performance when not needed. - * Since failure is not an option, reconfiguring queues on the fly is not - * recommended. - * * @param dev * Pointer to Ethernet device structure. * @param in_mtu * New MTU. * * @return - * 0 on success, negative errno value on failure. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ int mlx5_dev_set_mtu(struct rte_eth_dev *dev, uint16_t mtu) { struct priv *priv = dev->data->dev_private; - int ret = 0; - unsigned int i; - uint16_t (*rx_func)(void *, struct rte_mbuf **, uint16_t) = - mlx5_rx_burst; - unsigned int max_frame_len; - int rehash; - int restart = priv->started; - - if (mlx5_is_secondary()) - return -E_RTE_SECONDARY; + uint16_t kern_mtu = 0; + int ret; - priv_lock(priv); + ret = mlx5_get_mtu(dev, &kern_mtu); + if (ret) + return ret; /* Set kernel interface MTU first. */ - if (priv_set_mtu(priv, mtu)) { - ret = errno; - WARN("cannot set port %u MTU to %u: %s", priv->port, mtu, - strerror(ret)); - goto out; - } else - DEBUG("adapter port %u MTU set to %u", priv->port, mtu); - /* Temporarily replace RX handler with a fake one, assuming it has not - * been copied elsewhere. */ - dev->rx_pkt_burst = removed_rx_burst; - /* Make sure everyone has left mlx5_rx_burst() and uses - * removed_rx_burst() instead. */ - rte_wmb(); - usleep(1000); - /* MTU does not include header and CRC. */ - max_frame_len = ETHER_HDR_LEN + mtu + ETHER_CRC_LEN; - /* Check if at least one queue is going to need a SGE update. */ - for (i = 0; i != priv->rxqs_n; ++i) { - struct rxq *rxq = (*priv->rxqs)[i]; - unsigned int mb_len; - unsigned int size = RTE_PKTMBUF_HEADROOM + max_frame_len; - unsigned int sges_n; - - if (rxq == NULL) - continue; - mb_len = rte_pktmbuf_data_room_size(rxq->mp); - assert(mb_len >= RTE_PKTMBUF_HEADROOM); - /* - * Determine the number of SGEs needed for a full packet - * and round it to the next power of two. - */ - sges_n = log2above((size / mb_len) + !!(size % mb_len)); - if (sges_n != rxq->sges_n) - break; - } - /* - * If all queues have the right number of SGEs, a simple rehash - * of their buffers is enough, otherwise SGE information can only - * be updated in a queue by recreating it. All resources that depend - * on queues (flows, indirection tables) must be recreated as well in - * that case. - */ - rehash = (i == priv->rxqs_n); - if (!rehash) { - /* Clean up everything as with mlx5_dev_stop(). */ - priv_special_flow_disable_all(priv); - priv_mac_addrs_disable(priv); - priv_destroy_hash_rxqs(priv); - priv_fdir_disable(priv); - priv_dev_interrupt_handler_uninstall(priv, dev); - } -recover: - /* Reconfigure each RX queue. */ - for (i = 0; (i != priv->rxqs_n); ++i) { - struct rxq *rxq = (*priv->rxqs)[i]; - struct rxq_ctrl *rxq_ctrl = - container_of(rxq, struct rxq_ctrl, rxq); - int sp; - unsigned int mb_len; - unsigned int tmp; - - if (rxq == NULL) - continue; - mb_len = rte_pktmbuf_data_room_size(rxq->mp); - assert(mb_len >= RTE_PKTMBUF_HEADROOM); - /* Toggle scattered support (sp) if necessary. */ - sp = (max_frame_len > (mb_len - RTE_PKTMBUF_HEADROOM)); - /* Provide new values to rxq_setup(). */ - dev->data->dev_conf.rxmode.jumbo_frame = sp; - dev->data->dev_conf.rxmode.max_rx_pkt_len = max_frame_len; - if (rehash) - ret = rxq_rehash(dev, rxq_ctrl); - else - ret = rxq_ctrl_setup(dev, rxq_ctrl, 1 << rxq->elts_n, - rxq_ctrl->socket, NULL, rxq->mp); - if (!ret) - continue; - /* Attempt to roll back in case of error. */ - tmp = (mb_len << rxq->sges_n) - RTE_PKTMBUF_HEADROOM; - if (max_frame_len != tmp) { - max_frame_len = tmp; - goto recover; - } - /* Double fault, disable RX. */ - break; - } - /* - * Use a safe RX burst function in case of error, otherwise mimic - * mlx5_dev_start(). - */ - if (ret) { - ERROR("unable to reconfigure RX queues, RX disabled"); - rx_func = removed_rx_burst; - } else if (restart && - !rehash && - !priv_create_hash_rxqs(priv) && - !priv_rehash_flows(priv)) { - if (dev->data->dev_conf.fdir_conf.mode == RTE_FDIR_MODE_NONE) - priv_fdir_enable(priv); - priv_dev_interrupt_handler_install(priv, dev); + ret = mlx5_set_mtu(dev, mtu); + if (ret) + return ret; + ret = mlx5_get_mtu(dev, &kern_mtu); + if (ret) + return ret; + if (kern_mtu == mtu) { + priv->mtu = mtu; + DRV_LOG(DEBUG, "port %u adapter MTU set to %u", + dev->data->port_id, mtu); + return 0; } - priv->mtu = mtu; - /* Burst functions can now be called again. */ - rte_wmb(); - dev->rx_pkt_burst = rx_func; -out: - priv_unlock(priv); - assert(ret >= 0); - return -ret; + rte_errno = EAGAIN; + return -rte_errno; } /** @@ -1052,31 +873,26 @@ out: * Flow control output buffer. * * @return - * 0 on success, negative errno value on failure. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ int mlx5_dev_get_flow_ctrl(struct rte_eth_dev *dev, struct rte_eth_fc_conf *fc_conf) { - struct priv *priv = dev->data->dev_private; struct ifreq ifr; struct ethtool_pauseparam ethpause = { .cmd = ETHTOOL_GPAUSEPARAM }; int ret; - if (mlx5_is_secondary()) - return -E_RTE_SECONDARY; - ifr.ifr_data = (void *)ðpause; - priv_lock(priv); - if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) { - ret = errno; - WARN("ioctl(SIOCETHTOOL, ETHTOOL_GPAUSEPARAM)" - " failed: %s", - strerror(ret)); - goto out; + ret = mlx5_ifreq(dev, SIOCETHTOOL, &ifr); + if (ret) { + DRV_LOG(WARNING, + "port %u ioctl(SIOCETHTOOL, ETHTOOL_GPAUSEPARAM) failed:" + " %s", + dev->data->port_id, strerror(rte_errno)); + return ret; } - fc_conf->autoneg = ethpause.autoneg; if (ethpause.rx_pause && ethpause.tx_pause) fc_conf->mode = RTE_FC_FULL; @@ -1086,12 +902,7 @@ mlx5_dev_get_flow_ctrl(struct rte_eth_dev *dev, struct rte_eth_fc_conf *fc_conf) fc_conf->mode = RTE_FC_TX_PAUSE; else fc_conf->mode = RTE_FC_NONE; - ret = 0; - -out: - priv_unlock(priv); - assert(ret >= 0); - return -ret; + return 0; } /** @@ -1103,21 +914,17 @@ out: * Flow control parameters. * * @return - * 0 on success, negative errno value on failure. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ int mlx5_dev_set_flow_ctrl(struct rte_eth_dev *dev, struct rte_eth_fc_conf *fc_conf) { - struct priv *priv = dev->data->dev_private; struct ifreq ifr; struct ethtool_pauseparam ethpause = { .cmd = ETHTOOL_SPAUSEPARAM }; int ret; - if (mlx5_is_secondary()) - return -E_RTE_SECONDARY; - ifr.ifr_data = (void *)ðpause; ethpause.autoneg = fc_conf->autoneg; if (((fc_conf->mode & RTE_FC_FULL) == RTE_FC_FULL) || @@ -1131,21 +938,15 @@ mlx5_dev_set_flow_ctrl(struct rte_eth_dev *dev, struct rte_eth_fc_conf *fc_conf) ethpause.tx_pause = 1; else ethpause.tx_pause = 0; - - priv_lock(priv); - if (priv_ifreq(priv, SIOCETHTOOL, &ifr)) { - ret = errno; - WARN("ioctl(SIOCETHTOOL, ETHTOOL_SPAUSEPARAM)" - " failed: %s", - strerror(ret)); - goto out; + ret = mlx5_ifreq(dev, SIOCETHTOOL, &ifr); + if (ret) { + DRV_LOG(WARNING, + "port %u ioctl(SIOCETHTOOL, ETHTOOL_SPAUSEPARAM)" + " failed: %s", + dev->data->port_id, strerror(rte_errno)); + return ret; } - ret = 0; - -out: - priv_unlock(priv); - assert(ret >= 0); - return -ret; + return 0; } /** @@ -1157,7 +958,7 @@ out: * PCI bus address output buffer. * * @return - * 0 on success, -1 on failure and errno is set. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ int mlx5_ibv_device_to_pci_addr(const struct ibv_device *device, @@ -1168,8 +969,10 @@ mlx5_ibv_device_to_pci_addr(const struct ibv_device *device, MKSTR(path, "%s/device/uevent", device->ibdev_path); file = fopen(path, "rb"); - if (file == NULL) - return -1; + if (file == NULL) { + rte_errno = errno; + return -rte_errno; + } while (fgets(line, sizeof(line), file) == line) { size_t len = strlen(line); int ret; @@ -1185,7 +988,7 @@ mlx5_ibv_device_to_pci_addr(const struct ibv_device *device, /* Extract information. */ if (sscanf(line, "PCI_SLOT_NAME=" - "%" SCNx16 ":%" SCNx8 ":%" SCNx8 ".%" SCNx8 "\n", + "%" SCNx32 ":%" SCNx8 ":%" SCNx8 ".%" SCNx8 "\n", &pci_addr->domain, &pci_addr->bus, &pci_addr->devid, @@ -1199,408 +1002,365 @@ mlx5_ibv_device_to_pci_addr(const struct ibv_device *device, } /** - * Link status handler. + * Device status handler. * - * @param priv - * Pointer to private structure. * @param dev - * Pointer to the rte_eth_dev structure. + * Pointer to Ethernet device. + * @param events + * Pointer to event flags holder. * * @return - * Nonzero if the callback process can be called immediately. + * Events bitmap of callback process which can be called immediately. */ -static int -priv_dev_link_status_handler(struct priv *priv, struct rte_eth_dev *dev) +static uint32_t +mlx5_dev_status_handler(struct rte_eth_dev *dev) { + struct priv *priv = dev->data->dev_private; struct ibv_async_event event; - struct rte_eth_link *link = &dev->data->dev_link; - int ret = 0; + uint32_t ret = 0; + if (mlx5_link_update(dev, 0) == -EAGAIN) { + usleep(0); + return 0; + } /* Read all message and acknowledge them. */ for (;;) { - if (ibv_get_async_event(priv->ctx, &event)) + if (mlx5_glue->get_async_event(priv->ctx, &event)) break; - - if (event.event_type != IBV_EVENT_PORT_ACTIVE && - event.event_type != IBV_EVENT_PORT_ERR) - DEBUG("event type %d on port %d not handled", - event.event_type, event.element.port_num); - ibv_ack_async_event(&event); - } - mlx5_link_update(dev, 0); - if (((link->link_speed == 0) && link->link_status) || - ((link->link_speed != 0) && !link->link_status)) { - if (!priv->pending_alarm) { - /* Inconsistent status, check again later. */ - priv->pending_alarm = 1; - rte_eal_alarm_set(MLX5_ALARM_TIMEOUT_US, - mlx5_dev_link_status_handler, - dev); - } - } else { - ret = 1; + if ((event.event_type == IBV_EVENT_PORT_ACTIVE || + event.event_type == IBV_EVENT_PORT_ERR) && + (dev->data->dev_conf.intr_conf.lsc == 1)) + ret |= (1 << RTE_ETH_EVENT_INTR_LSC); + else if (event.event_type == IBV_EVENT_DEVICE_FATAL && + dev->data->dev_conf.intr_conf.rmv == 1) + ret |= (1 << RTE_ETH_EVENT_INTR_RMV); + else + DRV_LOG(DEBUG, + "port %u event type %d on not handled", + dev->data->port_id, event.event_type); + mlx5_glue->ack_async_event(&event); } return ret; } /** - * Handle delayed link status event. + * Handle interrupts from the NIC. * - * @param arg - * Registered argument. + * @param[in] intr_handle + * Interrupt handler. + * @param cb_arg + * Callback argument. */ void -mlx5_dev_link_status_handler(void *arg) +mlx5_dev_interrupt_handler(void *cb_arg) { - struct rte_eth_dev *dev = arg; - struct priv *priv = dev->data->dev_private; - int ret; + struct rte_eth_dev *dev = cb_arg; + uint32_t events; - priv_lock(priv); - assert(priv->pending_alarm == 1); - priv->pending_alarm = 0; - ret = priv_dev_link_status_handler(priv, dev); - priv_unlock(priv); - if (ret) + events = mlx5_dev_status_handler(dev); + if (events & (1 << RTE_ETH_EVENT_INTR_LSC)) _rte_eth_dev_callback_process(dev, RTE_ETH_EVENT_INTR_LSC, NULL); + if (events & (1 << RTE_ETH_EVENT_INTR_RMV)) + _rte_eth_dev_callback_process(dev, RTE_ETH_EVENT_INTR_RMV, NULL); } /** - * Handle interrupts from the NIC. + * Handle interrupts from the socket. * - * @param[in] intr_handle - * Interrupt handler. * @param cb_arg * Callback argument. */ -void -mlx5_dev_interrupt_handler(struct rte_intr_handle *intr_handle, void *cb_arg) +static void +mlx5_dev_handler_socket(void *cb_arg) { struct rte_eth_dev *dev = cb_arg; - struct priv *priv = dev->data->dev_private; - int ret; - (void)intr_handle; - priv_lock(priv); - ret = priv_dev_link_status_handler(priv, dev); - priv_unlock(priv); - if (ret) - _rte_eth_dev_callback_process(dev, RTE_ETH_EVENT_INTR_LSC, NULL); + mlx5_socket_handle(dev); } /** * Uninstall interrupt handler. * - * @param priv - * Pointer to private structure. * @param dev - * Pointer to the rte_eth_dev structure. + * Pointer to Ethernet device. */ void -priv_dev_interrupt_handler_uninstall(struct priv *priv, struct rte_eth_dev *dev) +mlx5_dev_interrupt_handler_uninstall(struct rte_eth_dev *dev) { - if (!dev->data->dev_conf.intr_conf.lsc) - return; - rte_intr_callback_unregister(&priv->intr_handle, - mlx5_dev_interrupt_handler, - dev); - if (priv->pending_alarm) - rte_eal_alarm_cancel(mlx5_dev_link_status_handler, dev); - priv->pending_alarm = 0; + struct priv *priv = dev->data->dev_private; + + if (dev->data->dev_conf.intr_conf.lsc || + dev->data->dev_conf.intr_conf.rmv) + rte_intr_callback_unregister(&priv->intr_handle, + mlx5_dev_interrupt_handler, dev); + if (priv->primary_socket) + rte_intr_callback_unregister(&priv->intr_handle_socket, + mlx5_dev_handler_socket, dev); priv->intr_handle.fd = 0; priv->intr_handle.type = RTE_INTR_HANDLE_UNKNOWN; + priv->intr_handle_socket.fd = 0; + priv->intr_handle_socket.type = RTE_INTR_HANDLE_UNKNOWN; } /** * Install interrupt handler. * - * @param priv - * Pointer to private structure. * @param dev - * Pointer to the rte_eth_dev structure. + * Pointer to Ethernet device. */ void -priv_dev_interrupt_handler_install(struct priv *priv, struct rte_eth_dev *dev) +mlx5_dev_interrupt_handler_install(struct rte_eth_dev *dev) { - int rc, flags; + struct priv *priv = dev->data->dev_private; + int ret; + int flags; - if (!dev->data->dev_conf.intr_conf.lsc) - return; assert(priv->ctx->async_fd > 0); flags = fcntl(priv->ctx->async_fd, F_GETFL); - rc = fcntl(priv->ctx->async_fd, F_SETFL, flags | O_NONBLOCK); - if (rc < 0) { - INFO("failed to change file descriptor async event queue"); + ret = fcntl(priv->ctx->async_fd, F_SETFL, flags | O_NONBLOCK); + if (ret) { + DRV_LOG(INFO, + "port %u failed to change file descriptor async event" + " queue", + dev->data->port_id); dev->data->dev_conf.intr_conf.lsc = 0; - } else { + dev->data->dev_conf.intr_conf.rmv = 0; + } + if (dev->data->dev_conf.intr_conf.lsc || + dev->data->dev_conf.intr_conf.rmv) { priv->intr_handle.fd = priv->ctx->async_fd; priv->intr_handle.type = RTE_INTR_HANDLE_EXT; rte_intr_callback_register(&priv->intr_handle, - mlx5_dev_interrupt_handler, - dev); + mlx5_dev_interrupt_handler, dev); + } + ret = mlx5_socket_init(dev); + if (ret) + DRV_LOG(ERR, "port %u cannot initialise socket: %s", + dev->data->port_id, strerror(rte_errno)); + else if (priv->primary_socket) { + priv->intr_handle_socket.fd = priv->primary_socket; + priv->intr_handle_socket.type = RTE_INTR_HANDLE_EXT; + rte_intr_callback_register(&priv->intr_handle_socket, + mlx5_dev_handler_socket, dev); } } /** - * Change the link state (UP / DOWN). + * DPDK callback to bring the link DOWN. * - * @param priv + * @param dev * Pointer to Ethernet device structure. - * @param up - * Nonzero for link up, otherwise link down. * * @return - * 0 on success, errno value on failure. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ -static int -priv_set_link(struct priv *priv, int up) +int +mlx5_set_link_down(struct rte_eth_dev *dev) { - struct rte_eth_dev *dev = priv->dev; - int err; - - if (up) { - err = priv_set_flags(priv, ~IFF_UP, IFF_UP); - if (err) - return err; - priv_select_tx_function(priv); - priv_select_rx_function(priv); - } else { - err = priv_set_flags(priv, ~IFF_UP, ~IFF_UP); - if (err) - return err; - dev->rx_pkt_burst = removed_rx_burst; - dev->tx_pkt_burst = removed_tx_burst; - } - return 0; + return mlx5_set_flags(dev, ~IFF_UP, ~IFF_UP); } /** - * DPDK callback to bring the link DOWN. + * DPDK callback to bring the link UP. * * @param dev * Pointer to Ethernet device structure. * * @return - * 0 on success, errno value on failure. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ int -mlx5_set_link_down(struct rte_eth_dev *dev) +mlx5_set_link_up(struct rte_eth_dev *dev) +{ + return mlx5_set_flags(dev, ~IFF_UP, IFF_UP); +} + +/** + * Configure the TX function to use. + * + * @param dev + * Pointer to private data structure. + * + * @return + * Pointer to selected Tx burst function. + */ +eth_tx_burst_t +mlx5_select_tx_function(struct rte_eth_dev *dev) { struct priv *priv = dev->data->dev_private; - int err; + eth_tx_burst_t tx_pkt_burst = mlx5_tx_burst; + struct mlx5_dev_config *config = &priv->config; + uint64_t tx_offloads = dev->data->dev_conf.txmode.offloads; + int tso = !!(tx_offloads & (DEV_TX_OFFLOAD_TCP_TSO | + DEV_TX_OFFLOAD_VXLAN_TNL_TSO | + DEV_TX_OFFLOAD_GRE_TNL_TSO | + DEV_TX_OFFLOAD_IP_TNL_TSO | + DEV_TX_OFFLOAD_UDP_TNL_TSO)); + int swp = !!(tx_offloads & (DEV_TX_OFFLOAD_IP_TNL_TSO | + DEV_TX_OFFLOAD_UDP_TNL_TSO | + DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM)); + int vlan_insert = !!(tx_offloads & DEV_TX_OFFLOAD_VLAN_INSERT); - priv_lock(priv); - err = priv_set_link(priv, 0); - priv_unlock(priv); - return err; + assert(priv != NULL); + /* Select appropriate TX function. */ + if (vlan_insert || tso || swp) + return tx_pkt_burst; + if (config->mps == MLX5_MPW_ENHANCED) { + if (mlx5_check_vec_tx_support(dev) > 0) { + if (mlx5_check_raw_vec_tx_support(dev) > 0) + tx_pkt_burst = mlx5_tx_burst_raw_vec; + else + tx_pkt_burst = mlx5_tx_burst_vec; + DRV_LOG(DEBUG, + "port %u selected enhanced MPW Tx vectorized" + " function", + dev->data->port_id); + } else { + tx_pkt_burst = mlx5_tx_burst_empw; + DRV_LOG(DEBUG, + "port %u selected enhanced MPW Tx function", + dev->data->port_id); + } + } else if (config->mps && (config->txq_inline > 0)) { + tx_pkt_burst = mlx5_tx_burst_mpw_inline; + DRV_LOG(DEBUG, "port %u selected MPW inline Tx function", + dev->data->port_id); + } else if (config->mps) { + tx_pkt_burst = mlx5_tx_burst_mpw; + DRV_LOG(DEBUG, "port %u selected MPW Tx function", + dev->data->port_id); + } + return tx_pkt_burst; } /** - * DPDK callback to bring the link UP. + * Configure the RX function to use. + * + * @param dev + * Pointer to private data structure. + * + * @return + * Pointer to selected Rx burst function. + */ +eth_rx_burst_t +mlx5_select_rx_function(struct rte_eth_dev *dev) +{ + eth_rx_burst_t rx_pkt_burst = mlx5_rx_burst; + + assert(dev != NULL); + if (mlx5_check_vec_rx_support(dev) > 0) { + rx_pkt_burst = mlx5_rx_burst_vec; + DRV_LOG(DEBUG, "port %u selected Rx vectorized function", + dev->data->port_id); + } else if (mlx5_mprq_enabled(dev)) { + rx_pkt_burst = mlx5_rx_burst_mprq; + } + return rx_pkt_burst; +} + +/** + * Check if mlx5 device was removed. * * @param dev * Pointer to Ethernet device structure. * * @return - * 0 on success, errno value on failure. + * 1 when device is removed, otherwise 0. */ int -mlx5_set_link_up(struct rte_eth_dev *dev) +mlx5_is_removed(struct rte_eth_dev *dev) { + struct ibv_device_attr device_attr; struct priv *priv = dev->data->dev_private; - int err; - priv_lock(priv); - err = priv_set_link(priv, 1); - priv_unlock(priv); - return err; + if (mlx5_glue->query_device(priv->ctx, &device_attr) == EIO) + return 1; + return 0; } /** - * Configure secondary process queues from a private data pointer (primary - * or secondary) and update burst callbacks. Can take place only once. - * - * All queues must have been previously created by the primary process to - * avoid undefined behavior. + * Get port ID list of mlx5 instances sharing a common device. * - * @param priv - * Private data pointer from either primary or secondary process. + * @param[in] dev + * Device to look for. + * @param[out] port_list + * Result buffer for collected port IDs. + * @param port_list_n + * Maximum number of entries in result buffer. If 0, @p port_list can be + * NULL. * * @return - * Private data pointer from secondary process, NULL in case of error. + * Number of matching instances regardless of the @p port_list_n + * parameter, 0 if none were found. */ -struct priv * -mlx5_secondary_data_setup(struct priv *priv) +unsigned int +mlx5_dev_to_port_id(const struct rte_device *dev, uint16_t *port_list, + unsigned int port_list_n) { - unsigned int port_id = 0; - struct mlx5_secondary_data *sd; - void **tx_queues; - void **rx_queues; - unsigned int nb_tx_queues; - unsigned int nb_rx_queues; - unsigned int i; + uint16_t id; + unsigned int n = 0; - /* priv must be valid at this point. */ - assert(priv != NULL); - /* priv->dev must also be valid but may point to local memory from - * another process, possibly with the same address and must not - * be dereferenced yet. */ - assert(priv->dev != NULL); - /* Determine port ID by finding out where priv comes from. */ - while (1) { - sd = &mlx5_secondary_data[port_id]; - rte_spinlock_lock(&sd->lock); - /* Primary process? */ - if (sd->primary_priv == priv) - break; - /* Secondary process? */ - if (sd->data.dev_private == priv) - break; - rte_spinlock_unlock(&sd->lock); - if (++port_id == RTE_DIM(mlx5_secondary_data)) - port_id = 0; - } - /* Switch to secondary private structure. If private data has already - * been updated by another thread, there is nothing else to do. */ - priv = sd->data.dev_private; - if (priv->dev->data == &sd->data) - goto end; - /* Sanity checks. Secondary private structure is supposed to point - * to local eth_dev, itself still pointing to the shared device data - * structure allocated by the primary process. */ - assert(sd->shared_dev_data != &sd->data); - assert(sd->data.nb_tx_queues == 0); - assert(sd->data.tx_queues == NULL); - assert(sd->data.nb_rx_queues == 0); - assert(sd->data.rx_queues == NULL); - assert(priv != sd->primary_priv); - assert(priv->dev->data == sd->shared_dev_data); - assert(priv->txqs_n == 0); - assert(priv->txqs == NULL); - assert(priv->rxqs_n == 0); - assert(priv->rxqs == NULL); - nb_tx_queues = sd->shared_dev_data->nb_tx_queues; - nb_rx_queues = sd->shared_dev_data->nb_rx_queues; - /* Allocate local storage for queues. */ - tx_queues = rte_zmalloc("secondary ethdev->tx_queues", - sizeof(sd->data.tx_queues[0]) * nb_tx_queues, - RTE_CACHE_LINE_SIZE); - rx_queues = rte_zmalloc("secondary ethdev->rx_queues", - sizeof(sd->data.rx_queues[0]) * nb_rx_queues, - RTE_CACHE_LINE_SIZE); - if (tx_queues == NULL || rx_queues == NULL) - goto error; - /* Lock to prevent control operations during setup. */ - priv_lock(priv); - /* TX queues. */ - for (i = 0; i != nb_tx_queues; ++i) { - struct txq *primary_txq = (*sd->primary_priv->txqs)[i]; - struct txq_ctrl *primary_txq_ctrl; - struct txq_ctrl *txq_ctrl; - - if (primary_txq == NULL) - continue; - primary_txq_ctrl = container_of(primary_txq, - struct txq_ctrl, txq); - txq_ctrl = rte_calloc_socket("TXQ", 1, sizeof(*txq_ctrl) + - (1 << primary_txq->elts_n) * - sizeof(struct rte_mbuf *), 0, - primary_txq_ctrl->socket); - if (txq_ctrl != NULL) { - if (txq_ctrl_setup(priv->dev, - txq_ctrl, - 1 << primary_txq->elts_n, - primary_txq_ctrl->socket, - NULL) == 0) { - txq_ctrl->txq.stats.idx = - primary_txq->stats.idx; - tx_queues[i] = &txq_ctrl->txq; - continue; - } - rte_free(txq_ctrl); - } - while (i) { - txq_ctrl = tx_queues[--i]; - txq_cleanup(txq_ctrl); - rte_free(txq_ctrl); - } - goto error; - } - /* RX queues. */ - for (i = 0; i != nb_rx_queues; ++i) { - struct rxq_ctrl *primary_rxq = - container_of((*sd->primary_priv->rxqs)[i], - struct rxq_ctrl, rxq); + RTE_ETH_FOREACH_DEV(id) { + struct rte_eth_dev *ldev = &rte_eth_devices[id]; - if (primary_rxq == NULL) + if (ldev->device != dev) continue; - /* Not supported yet. */ - rx_queues[i] = NULL; + if (n < port_list_n) + port_list[n] = id; + n++; } - /* Update everything. */ - priv->txqs = (void *)tx_queues; - priv->txqs_n = nb_tx_queues; - priv->rxqs = (void *)rx_queues; - priv->rxqs_n = nb_rx_queues; - sd->data.rx_queues = rx_queues; - sd->data.tx_queues = tx_queues; - sd->data.nb_rx_queues = nb_rx_queues; - sd->data.nb_tx_queues = nb_tx_queues; - sd->data.dev_link = sd->shared_dev_data->dev_link; - sd->data.mtu = sd->shared_dev_data->mtu; - memcpy(sd->data.rx_queue_state, sd->shared_dev_data->rx_queue_state, - sizeof(sd->data.rx_queue_state)); - memcpy(sd->data.tx_queue_state, sd->shared_dev_data->tx_queue_state, - sizeof(sd->data.tx_queue_state)); - sd->data.dev_flags = sd->shared_dev_data->dev_flags; - /* Use local data from now on. */ - rte_mb(); - priv->dev->data = &sd->data; - rte_mb(); - priv_select_tx_function(priv); - priv_select_rx_function(priv); - priv_unlock(priv); -end: - /* More sanity checks. */ - assert(priv->dev->data == &sd->data); - rte_spinlock_unlock(&sd->lock); - return priv; -error: - priv_unlock(priv); - rte_free(tx_queues); - rte_free(rx_queues); - rte_spinlock_unlock(&sd->lock); - return NULL; + return n; } /** - * Configure the TX function to use. + * Get switch information associated with network interface. * - * @param priv - * Pointer to private structure. + * @param ifindex + * Network interface index. + * @param[out] info + * Switch information object, populated in case of success. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. */ -void -priv_select_tx_function(struct priv *priv) +int +mlx5_sysfs_switch_info(unsigned int ifindex, struct mlx5_switch_info *info) { - priv->dev->tx_pkt_burst = mlx5_tx_burst; - /* Select appropriate TX function. */ - if (priv->mps && priv->txq_inline) { - priv->dev->tx_pkt_burst = mlx5_tx_burst_mpw_inline; - DEBUG("selected MPW inline TX function"); - } else if (priv->mps) { - priv->dev->tx_pkt_burst = mlx5_tx_burst_mpw; - DEBUG("selected MPW TX function"); + char ifname[IF_NAMESIZE]; + FILE *file; + struct mlx5_switch_info data = { .master = 0, }; + bool port_name_set = false; + bool port_switch_id_set = false; + char c; + + if (!if_indextoname(ifindex, ifname)) { + rte_errno = errno; + return -rte_errno; } -} -/** - * Configure the RX function to use. - * - * @param priv - * Pointer to private structure. - */ -void -priv_select_rx_function(struct priv *priv) -{ - priv->dev->rx_pkt_burst = mlx5_rx_burst; + MKSTR(phys_port_name, "/sys/class/net/%s/phys_port_name", + ifname); + MKSTR(phys_switch_id, "/sys/class/net/%s/phys_switch_id", + ifname); + + file = fopen(phys_port_name, "rb"); + if (file != NULL) { + port_name_set = + fscanf(file, "%d%c", &data.port_name, &c) == 2 && + c == '\n'; + fclose(file); + } + file = fopen(phys_switch_id, "rb"); + if (file == NULL) { + rte_errno = errno; + return -rte_errno; + } + port_switch_id_set = + fscanf(file, "%" SCNx64 "%c", &data.switch_id, &c) == 2 && + c == '\n'; + fclose(file); + data.master = port_switch_id_set && !port_name_set; + data.representor = port_switch_id_set && port_name_set; + *info = data; + return 0; }