/*
* This function passes incoming fabric frames to libFC
*/
void fnic_handle_frame(struct work_struct *work)
{
struct fnic *fnic = container_of(work, struct fnic, frame_work);
struct fc_lport *lp = fnic->lport;
unsigned long flags;
struct sk_buff *skb;
struct fc_frame *fp;
while ((skb = skb_dequeue(&fnic->frame_queue))) {
spin_lock_irqsave(&fnic->fnic_lock, flags);
if (fnic->stop_rx_link_events) {
spin_unlock_irqrestore(&fnic->fnic_lock, flags);
dev_kfree_skb(skb);
return;
}
fp = (struct fc_frame *)skb;
/* if Flogi resp frame, register the address */
if (fr_flags(fp)) {
vnic_dev_add_addr(fnic->vdev,
fnic->data_src_addr);
fr_flags(fp) = 0;
}
spin_unlock_irqrestore(&fnic->fnic_lock, flags);
fc_exch_recv(lp, lp->emp, fp);
}
}
int enic_dev_add_addr(struct enic *enic, u8 *addr)
{
int err;
spin_lock(&enic->devcmd_lock);
err = vnic_dev_add_addr(enic->vdev, addr);
spin_unlock(&enic->devcmd_lock);
return err;
}
int enic_dev_add_station_addr(struct enic *enic)
{
int err;
if (!is_valid_ether_addr(enic->netdev->dev_addr))
return -EADDRNOTAVAIL;
spin_lock(&enic->devcmd_lock);
err = vnic_dev_add_addr(enic->vdev, enic->netdev->dev_addr);
spin_unlock(&enic->devcmd_lock);
return err;
}
void enic_set_mac_address(struct enic *enic, uint8_t *mac_addr)
{
int err;
if (!is_eth_addr_valid(mac_addr)) {
dev_err(enic, "invalid mac address\n");
return;
}
err = vnic_dev_del_addr(enic->vdev, mac_addr);
if (err) {
dev_err(enic, "del mac addr failed\n");
return;
}
ether_addr_copy((struct ether_addr *)mac_addr,
(struct ether_addr *)enic->mac_addr);
err = vnic_dev_add_addr(enic->vdev, mac_addr);
if (err) {
dev_err(enic, "add mac addr failed\n");
return;
}
}
static int fnic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
{
struct Scsi_Host *host;
struct fc_lport *lp;
struct fnic *fnic;
mempool_t *pool;
int err;
int i;
unsigned long flags;
/*
* Allocate SCSI Host and set up association between host,
* local port, and fnic
*/
lp = libfc_host_alloc(&fnic_host_template, sizeof(struct fnic));
if (!lp) {
printk(KERN_ERR PFX "Unable to alloc libfc local port\n");
err = -ENOMEM;
goto err_out;
}
host = lp->host;
fnic = lport_priv(lp);
fnic->lport = lp;
fnic->ctlr.lp = lp;
snprintf(fnic->name, sizeof(fnic->name) - 1, "%s%d", DRV_NAME,
host->host_no);
host->transportt = fnic_fc_transport;
err = fnic_stats_debugfs_init(fnic);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Failed to initialize debugfs for stats\n");
fnic_stats_debugfs_remove(fnic);
}
/* Setup PCI resources */
pci_set_drvdata(pdev, fnic);
fnic->pdev = pdev;
err = pci_enable_device(pdev);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Cannot enable PCI device, aborting.\n");
goto err_out_free_hba;
}
err = pci_request_regions(pdev, DRV_NAME);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Cannot enable PCI resources, aborting\n");
goto err_out_disable_device;
}
pci_set_master(pdev);
/* Query PCI controller on system for DMA addressing
* limitation for the device. Try 64-bit first, and
* fail to 32-bit.
*/
err = pci_set_dma_mask(pdev, DMA_BIT_MASK(64));
if (err) {
err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"No usable DMA configuration "
"aborting\n");
goto err_out_release_regions;
}
err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32));
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Unable to obtain 32-bit DMA "
"for consistent allocations, aborting.\n");
goto err_out_release_regions;
}
} else {
err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64));
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Unable to obtain 64-bit DMA "
"for consistent allocations, aborting.\n");
goto err_out_release_regions;
}
}
/* Map vNIC resources from BAR0 */
if (!(pci_resource_flags(pdev, 0) & IORESOURCE_MEM)) {
shost_printk(KERN_ERR, fnic->lport->host,
"BAR0 not memory-map'able, aborting.\n");
err = -ENODEV;
goto err_out_release_regions;
}
fnic->bar0.vaddr = pci_iomap(pdev, 0, 0);
fnic->bar0.bus_addr = pci_resource_start(pdev, 0);
fnic->bar0.len = pci_resource_len(pdev, 0);
if (!fnic->bar0.vaddr) {
shost_printk(KERN_ERR, fnic->lport->host,
"Cannot memory-map BAR0 res hdr, "
"aborting.\n");
err = -ENODEV;
goto err_out_release_regions;
}
fnic->vdev = vnic_dev_register(NULL, fnic, pdev, &fnic->bar0);
if (!fnic->vdev) {
shost_printk(KERN_ERR, fnic->lport->host,
"vNIC registration failed, "
"aborting.\n");
err = -ENODEV;
goto err_out_iounmap;
}
err = fnic_dev_wait(fnic->vdev, vnic_dev_open,
vnic_dev_open_done, 0);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"vNIC dev open failed, aborting.\n");
goto err_out_vnic_unregister;
}
err = vnic_dev_init(fnic->vdev, 0);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"vNIC dev init failed, aborting.\n");
goto err_out_dev_close;
}
err = vnic_dev_mac_addr(fnic->vdev, fnic->ctlr.ctl_src_addr);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"vNIC get MAC addr failed \n");
goto err_out_dev_close;
}
/* set data_src for point-to-point mode and to keep it non-zero */
memcpy(fnic->data_src_addr, fnic->ctlr.ctl_src_addr, ETH_ALEN);
/* Get vNIC configuration */
err = fnic_get_vnic_config(fnic);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Get vNIC configuration failed, "
"aborting.\n");
goto err_out_dev_close;
}
/* Configure Maximum Outstanding IO reqs*/
if (fnic->config.io_throttle_count != FNIC_UCSM_DFLT_THROTTLE_CNT_BLD) {
host->can_queue = min_t(u32, FNIC_MAX_IO_REQ,
max_t(u32, FNIC_MIN_IO_REQ,
fnic->config.io_throttle_count));
}
fnic->fnic_max_tag_id = host->can_queue;
err = scsi_init_shared_tag_map(host, fnic->fnic_max_tag_id);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Unable to alloc shared tag map\n");
goto err_out_dev_close;
}
host->max_lun = fnic->config.luns_per_tgt;
host->max_id = FNIC_MAX_FCP_TARGET;
host->max_cmd_len = FCOE_MAX_CMD_LEN;
fnic_get_res_counts(fnic);
err = fnic_set_intr_mode(fnic);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Failed to set intr mode, "
"aborting.\n");
goto err_out_dev_close;
}
err = fnic_alloc_vnic_resources(fnic);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Failed to alloc vNIC resources, "
"aborting.\n");
goto err_out_clear_intr;
}
/* initialize all fnic locks */
spin_lock_init(&fnic->fnic_lock);
for (i = 0; i < FNIC_WQ_MAX; i++)
spin_lock_init(&fnic->wq_lock[i]);
for (i = 0; i < FNIC_WQ_COPY_MAX; i++) {
spin_lock_init(&fnic->wq_copy_lock[i]);
fnic->wq_copy_desc_low[i] = DESC_CLEAN_LOW_WATERMARK;
fnic->fw_ack_recd[i] = 0;
fnic->fw_ack_index[i] = -1;
}
for (i = 0; i < FNIC_IO_LOCKS; i++)
spin_lock_init(&fnic->io_req_lock[i]);
fnic->io_req_pool = mempool_create_slab_pool(2, fnic_io_req_cache);
if (!fnic->io_req_pool)
goto err_out_free_resources;
pool = mempool_create_slab_pool(2, fnic_sgl_cache[FNIC_SGL_CACHE_DFLT]);
if (!pool)
goto err_out_free_ioreq_pool;
fnic->io_sgl_pool[FNIC_SGL_CACHE_DFLT] = pool;
pool = mempool_create_slab_pool(2, fnic_sgl_cache[FNIC_SGL_CACHE_MAX]);
if (!pool)
goto err_out_free_dflt_pool;
fnic->io_sgl_pool[FNIC_SGL_CACHE_MAX] = pool;
/* setup vlan config, hw inserts vlan header */
fnic->vlan_hw_insert = 1;
fnic->vlan_id = 0;
/* Initialize the FIP fcoe_ctrl struct */
fnic->ctlr.send = fnic_eth_send;
fnic->ctlr.update_mac = fnic_update_mac;
fnic->ctlr.get_src_addr = fnic_get_mac;
if (fnic->config.flags & VFCF_FIP_CAPABLE) {
shost_printk(KERN_INFO, fnic->lport->host,
"firmware supports FIP\n");
/* enable directed and multicast */
vnic_dev_packet_filter(fnic->vdev, 1, 1, 0, 0, 0);
vnic_dev_add_addr(fnic->vdev, FIP_ALL_ENODE_MACS);
vnic_dev_add_addr(fnic->vdev, fnic->ctlr.ctl_src_addr);
fnic->set_vlan = fnic_set_vlan;
fcoe_ctlr_init(&fnic->ctlr, FIP_MODE_AUTO);
setup_timer(&fnic->fip_timer, fnic_fip_notify_timer,
(unsigned long)fnic);
spin_lock_init(&fnic->vlans_lock);
INIT_WORK(&fnic->fip_frame_work, fnic_handle_fip_frame);
INIT_WORK(&fnic->event_work, fnic_handle_event);
skb_queue_head_init(&fnic->fip_frame_queue);
INIT_LIST_HEAD(&fnic->evlist);
INIT_LIST_HEAD(&fnic->vlans);
} else {
shost_printk(KERN_INFO, fnic->lport->host,
"firmware uses non-FIP mode\n");
fcoe_ctlr_init(&fnic->ctlr, FIP_MODE_NON_FIP);
}
fnic->state = FNIC_IN_FC_MODE;
atomic_set(&fnic->in_flight, 0);
fnic->state_flags = FNIC_FLAGS_NONE;
/* Enable hardware stripping of vlan header on ingress */
fnic_set_nic_config(fnic, 0, 0, 0, 0, 0, 0, 1);
/* Setup notification buffer area */
err = fnic_notify_set(fnic);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Failed to alloc notify buffer, aborting.\n");
goto err_out_free_max_pool;
}
/* Setup notify timer when using MSI interrupts */
if (vnic_dev_get_intr_mode(fnic->vdev) == VNIC_DEV_INTR_MODE_MSI)
setup_timer(&fnic->notify_timer,
fnic_notify_timer, (unsigned long)fnic);
/* allocate RQ buffers and post them to RQ*/
for (i = 0; i < fnic->rq_count; i++) {
err = vnic_rq_fill(&fnic->rq[i], fnic_alloc_rq_frame);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"fnic_alloc_rq_frame can't alloc "
"frame\n");
goto err_out_free_rq_buf;
}
}
/*
* Initialization done with PCI system, hardware, firmware.
* Add host to SCSI
*/
err = scsi_add_host(lp->host, &pdev->dev);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"fnic: scsi_add_host failed...exiting\n");
goto err_out_free_rq_buf;
}
/* Start local port initiatialization */
lp->link_up = 0;
lp->max_retry_count = fnic->config.flogi_retries;
lp->max_rport_retry_count = fnic->config.plogi_retries;
lp->service_params = (FCP_SPPF_INIT_FCN | FCP_SPPF_RD_XRDY_DIS |
FCP_SPPF_CONF_COMPL);
if (fnic->config.flags & VFCF_FCP_SEQ_LVL_ERR)
lp->service_params |= FCP_SPPF_RETRY;
lp->boot_time = jiffies;
lp->e_d_tov = fnic->config.ed_tov;
lp->r_a_tov = fnic->config.ra_tov;
lp->link_supported_speeds = FC_PORTSPEED_10GBIT;
fc_set_wwnn(lp, fnic->config.node_wwn);
fc_set_wwpn(lp, fnic->config.port_wwn);
fcoe_libfc_config(lp, &fnic->ctlr, &fnic_transport_template, 0);
if (!fc_exch_mgr_alloc(lp, FC_CLASS_3, FCPIO_HOST_EXCH_RANGE_START,
FCPIO_HOST_EXCH_RANGE_END, NULL)) {
err = -ENOMEM;
goto err_out_remove_scsi_host;
}
fc_lport_init_stats(lp);
fnic->stats_reset_time = jiffies;
fc_lport_config(lp);
if (fc_set_mfs(lp, fnic->config.maxdatafieldsize +
sizeof(struct fc_frame_header))) {
err = -EINVAL;
goto err_out_free_exch_mgr;
}
fc_host_maxframe_size(lp->host) = lp->mfs;
fc_host_dev_loss_tmo(lp->host) = fnic->config.port_down_timeout / 1000;
sprintf(fc_host_symbolic_name(lp->host),
DRV_NAME " v" DRV_VERSION " over %s", fnic->name);
spin_lock_irqsave(&fnic_list_lock, flags);
list_add_tail(&fnic->list, &fnic_list);
spin_unlock_irqrestore(&fnic_list_lock, flags);
INIT_WORK(&fnic->link_work, fnic_handle_link);
INIT_WORK(&fnic->frame_work, fnic_handle_frame);
skb_queue_head_init(&fnic->frame_queue);
skb_queue_head_init(&fnic->tx_queue);
/* Enable all queues */
for (i = 0; i < fnic->raw_wq_count; i++)
vnic_wq_enable(&fnic->wq[i]);
for (i = 0; i < fnic->rq_count; i++)
vnic_rq_enable(&fnic->rq[i]);
for (i = 0; i < fnic->wq_copy_count; i++)
vnic_wq_copy_enable(&fnic->wq_copy[i]);
fc_fabric_login(lp);
vnic_dev_enable(fnic->vdev);
err = fnic_request_intr(fnic);
if (err) {
shost_printk(KERN_ERR, fnic->lport->host,
"Unable to request irq.\n");
goto err_out_free_exch_mgr;
}
for (i = 0; i < fnic->intr_count; i++)
vnic_intr_unmask(&fnic->intr[i]);
fnic_notify_timer_start(fnic);
return 0;
err_out_free_exch_mgr:
fc_exch_mgr_free(lp);
err_out_remove_scsi_host:
fc_remove_host(lp->host);
scsi_remove_host(lp->host);
err_out_free_rq_buf:
for (i = 0; i < fnic->rq_count; i++)
vnic_rq_clean(&fnic->rq[i], fnic_free_rq_buf);
vnic_dev_notify_unset(fnic->vdev);
err_out_free_max_pool:
mempool_destroy(fnic->io_sgl_pool[FNIC_SGL_CACHE_MAX]);
err_out_free_dflt_pool:
mempool_destroy(fnic->io_sgl_pool[FNIC_SGL_CACHE_DFLT]);
err_out_free_ioreq_pool:
mempool_destroy(fnic->io_req_pool);
err_out_free_resources:
fnic_free_vnic_resources(fnic);
err_out_clear_intr:
fnic_clear_intr_mode(fnic);
err_out_dev_close:
vnic_dev_close(fnic->vdev);
err_out_vnic_unregister:
vnic_dev_unregister(fnic->vdev);
err_out_iounmap:
fnic_iounmap(fnic);
err_out_release_regions:
pci_release_regions(pdev);
err_out_disable_device:
pci_disable_device(pdev);
err_out_free_hba:
fnic_stats_debugfs_remove(fnic);
scsi_host_put(lp->host);
err_out:
return err;
}
static void enicpmd_dev_info_get(struct rte_eth_dev *eth_dev,
struct rte_eth_dev_info *device_info)
{
struct enic *enic = pmd_priv(eth_dev);
ENICPMD_FUNC_TRACE();
/* Scattered Rx uses two receive queues per rx queue exposed to dpdk */
device_info->max_rx_queues = enic->conf_rq_count / 2;
device_info->max_tx_queues = enic->conf_wq_count;
device_info->min_rx_bufsize = ENIC_MIN_MTU;
/* "Max" mtu is not a typo. HW receives packet sizes up to the
* max mtu regardless of the current mtu (vNIC's mtu). vNIC mtu is
* a hint to the driver to size receive buffers accordingly so that
* larger-than-vnic-mtu packets get truncated.. For DPDK, we let
* the user decide the buffer size via rxmode.max_rx_pkt_len, basically
* ignoring vNIC mtu.
*/
device_info->max_rx_pktlen = enic_mtu_to_max_rx_pktlen(enic->max_mtu);
device_info->max_mac_addrs = ENIC_UNICAST_PERFECT_FILTERS;
device_info->rx_offload_capa = enic->rx_offload_capa;
device_info->tx_offload_capa = enic->tx_offload_capa;
device_info->tx_queue_offload_capa = enic->tx_queue_offload_capa;
device_info->default_rxconf = (struct rte_eth_rxconf) {
.rx_free_thresh = ENIC_DEFAULT_RX_FREE_THRESH
};
device_info->reta_size = enic->reta_size;
device_info->hash_key_size = enic->hash_key_size;
device_info->flow_type_rss_offloads = enic->flow_type_rss_offloads;
device_info->rx_desc_lim = (struct rte_eth_desc_lim) {
.nb_max = enic->config.rq_desc_count,
.nb_min = ENIC_MIN_RQ_DESCS,
.nb_align = ENIC_ALIGN_DESCS,
};
device_info->tx_desc_lim = (struct rte_eth_desc_lim) {
.nb_max = enic->config.wq_desc_count,
.nb_min = ENIC_MIN_WQ_DESCS,
.nb_align = ENIC_ALIGN_DESCS,
.nb_seg_max = ENIC_TX_XMIT_MAX,
.nb_mtu_seg_max = ENIC_NON_TSO_MAX_DESC,
};
device_info->default_rxportconf = (struct rte_eth_dev_portconf) {
.burst_size = ENIC_DEFAULT_RX_BURST,
.ring_size = RTE_MIN(device_info->rx_desc_lim.nb_max,
ENIC_DEFAULT_RX_RING_SIZE),
.nb_queues = ENIC_DEFAULT_RX_RINGS,
};
device_info->default_txportconf = (struct rte_eth_dev_portconf) {
.burst_size = ENIC_DEFAULT_TX_BURST,
.ring_size = RTE_MIN(device_info->tx_desc_lim.nb_max,
ENIC_DEFAULT_TX_RING_SIZE),
.nb_queues = ENIC_DEFAULT_TX_RINGS,
};
}
static const uint32_t *enicpmd_dev_supported_ptypes_get(struct rte_eth_dev *dev)
{
static const uint32_t ptypes[] = {
RTE_PTYPE_L2_ETHER,
RTE_PTYPE_L2_ETHER_VLAN,
RTE_PTYPE_L3_IPV4_EXT_UNKNOWN,
RTE_PTYPE_L3_IPV6_EXT_UNKNOWN,
RTE_PTYPE_L4_TCP,
RTE_PTYPE_L4_UDP,
RTE_PTYPE_L4_FRAG,
RTE_PTYPE_L4_NONFRAG,
RTE_PTYPE_UNKNOWN
};
static const uint32_t ptypes_overlay[] = {
RTE_PTYPE_L2_ETHER,
RTE_PTYPE_L2_ETHER_VLAN,
RTE_PTYPE_L3_IPV4_EXT_UNKNOWN,
RTE_PTYPE_L3_IPV6_EXT_UNKNOWN,
RTE_PTYPE_L4_TCP,
RTE_PTYPE_L4_UDP,
RTE_PTYPE_L4_FRAG,
RTE_PTYPE_L4_NONFRAG,
RTE_PTYPE_TUNNEL_GRENAT,
RTE_PTYPE_INNER_L2_ETHER,
RTE_PTYPE_INNER_L3_IPV4_EXT_UNKNOWN,
RTE_PTYPE_INNER_L3_IPV6_EXT_UNKNOWN,
RTE_PTYPE_INNER_L4_TCP,
RTE_PTYPE_INNER_L4_UDP,
RTE_PTYPE_INNER_L4_FRAG,
RTE_PTYPE_INNER_L4_NONFRAG,
RTE_PTYPE_UNKNOWN
};
if (dev->rx_pkt_burst != enic_dummy_recv_pkts &&
dev->rx_pkt_burst != NULL) {
struct enic *enic = pmd_priv(dev);
if (enic->overlay_offload)
return ptypes_overlay;
else
return ptypes;
}
return NULL;
}
static void enicpmd_dev_promiscuous_enable(struct rte_eth_dev *eth_dev)
{
struct enic *enic = pmd_priv(eth_dev);
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return;
ENICPMD_FUNC_TRACE();
enic->promisc = 1;
enic_add_packet_filter(enic);
}
static void enicpmd_dev_promiscuous_disable(struct rte_eth_dev *eth_dev)
{
struct enic *enic = pmd_priv(eth_dev);
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return;
ENICPMD_FUNC_TRACE();
enic->promisc = 0;
enic_add_packet_filter(enic);
}
static void enicpmd_dev_allmulticast_enable(struct rte_eth_dev *eth_dev)
{
struct enic *enic = pmd_priv(eth_dev);
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return;
ENICPMD_FUNC_TRACE();
enic->allmulti = 1;
enic_add_packet_filter(enic);
}
static void enicpmd_dev_allmulticast_disable(struct rte_eth_dev *eth_dev)
{
struct enic *enic = pmd_priv(eth_dev);
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return;
ENICPMD_FUNC_TRACE();
enic->allmulti = 0;
enic_add_packet_filter(enic);
}
static int enicpmd_add_mac_addr(struct rte_eth_dev *eth_dev,
struct ether_addr *mac_addr,
__rte_unused uint32_t index, __rte_unused uint32_t pool)
{
struct enic *enic = pmd_priv(eth_dev);
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return -E_RTE_SECONDARY;
ENICPMD_FUNC_TRACE();
return enic_set_mac_address(enic, mac_addr->addr_bytes);
}
static void enicpmd_remove_mac_addr(struct rte_eth_dev *eth_dev, uint32_t index)
{
struct enic *enic = pmd_priv(eth_dev);
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return;
ENICPMD_FUNC_TRACE();
if (enic_del_mac_address(enic, index))
dev_err(enic, "del mac addr failed\n");
}
static int enicpmd_set_mac_addr(struct rte_eth_dev *eth_dev,
struct ether_addr *addr)
{
struct enic *enic = pmd_priv(eth_dev);
int ret;
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return -E_RTE_SECONDARY;
ENICPMD_FUNC_TRACE();
ret = enic_del_mac_address(enic, 0);
if (ret)
return ret;
return enic_set_mac_address(enic, addr->addr_bytes);
}
static void debug_log_add_del_addr(struct ether_addr *addr, bool add)
{
char mac_str[ETHER_ADDR_FMT_SIZE];
ether_format_addr(mac_str, ETHER_ADDR_FMT_SIZE, addr);
PMD_INIT_LOG(DEBUG, " %s address %s\n",
add ? "add" : "remove", mac_str);
}
static int enicpmd_set_mc_addr_list(struct rte_eth_dev *eth_dev,
struct ether_addr *mc_addr_set,
uint32_t nb_mc_addr)
{
struct enic *enic = pmd_priv(eth_dev);
char mac_str[ETHER_ADDR_FMT_SIZE];
struct ether_addr *addr;
uint32_t i, j;
int ret;
ENICPMD_FUNC_TRACE();
/* Validate the given addresses first */
for (i = 0; i < nb_mc_addr && mc_addr_set != NULL; i++) {
addr = &mc_addr_set[i];
if (!is_multicast_ether_addr(addr) ||
is_broadcast_ether_addr(addr)) {
ether_format_addr(mac_str, ETHER_ADDR_FMT_SIZE, addr);
PMD_INIT_LOG(ERR, " invalid multicast address %s\n",
mac_str);
return -EINVAL;
}
}
/* Flush all if requested */
if (nb_mc_addr == 0 || mc_addr_set == NULL) {
PMD_INIT_LOG(DEBUG, " flush multicast addresses\n");
for (i = 0; i < enic->mc_count; i++) {
addr = &enic->mc_addrs[i];
debug_log_add_del_addr(addr, false);
ret = vnic_dev_del_addr(enic->vdev, addr->addr_bytes);
if (ret)
return ret;
}
enic->mc_count = 0;
return 0;
}
if (nb_mc_addr > ENIC_MULTICAST_PERFECT_FILTERS) {
PMD_INIT_LOG(ERR, " too many multicast addresses: max=%d\n",
ENIC_MULTICAST_PERFECT_FILTERS);
return -ENOSPC;
}
/*
* devcmd is slow, so apply the difference instead of flushing and
* adding everything.
* 1. Delete addresses on the NIC but not on the host
*/
for (i = 0; i < enic->mc_count; i++) {
addr = &enic->mc_addrs[i];
for (j = 0; j < nb_mc_addr; j++) {
if (is_same_ether_addr(addr, &mc_addr_set[j]))
break;
}
if (j < nb_mc_addr)
continue;
debug_log_add_del_addr(addr, false);
ret = vnic_dev_del_addr(enic->vdev, addr->addr_bytes);
if (ret)
return ret;
}
/* 2. Add addresses on the host but not on the NIC */
for (i = 0; i < nb_mc_addr; i++) {
addr = &mc_addr_set[i];
for (j = 0; j < enic->mc_count; j++) {
if (is_same_ether_addr(addr, &enic->mc_addrs[j]))
break;
}
if (j < enic->mc_count)
continue;
debug_log_add_del_addr(addr, true);
ret = vnic_dev_add_addr(enic->vdev, addr->addr_bytes);
if (ret)
return ret;
}
/* Keep a copy so we can flush/apply later on.. */
memcpy(enic->mc_addrs, mc_addr_set,
nb_mc_addr * sizeof(struct ether_addr));
enic->mc_count = nb_mc_addr;
return 0;
}
static int enicpmd_mtu_set(struct rte_eth_dev *eth_dev, uint16_t mtu)
{
struct enic *enic = pmd_priv(eth_dev);
ENICPMD_FUNC_TRACE();
return enic_set_mtu(enic, mtu);
}
static int enicpmd_dev_rss_reta_query(struct rte_eth_dev *dev,
struct rte_eth_rss_reta_entry64
*reta_conf,
uint16_t reta_size)
{
struct enic *enic = pmd_priv(dev);
uint16_t i, idx, shift;
ENICPMD_FUNC_TRACE();
if (reta_size != ENIC_RSS_RETA_SIZE) {
dev_err(enic, "reta_query: wrong reta_size. given=%u expected=%u\n",
reta_size, ENIC_RSS_RETA_SIZE);
return -EINVAL;
}
for (i = 0; i < reta_size; i++) {
idx = i / RTE_RETA_GROUP_SIZE;
shift = i % RTE_RETA_GROUP_SIZE;
if (reta_conf[idx].mask & (1ULL << shift))
reta_conf[idx].reta[shift] = enic_sop_rq_idx_to_rte_idx(
enic->rss_cpu.cpu[i / 4].b[i % 4]);
}
return 0;
}
static int enicpmd_dev_rss_reta_update(struct rte_eth_dev *dev,
struct rte_eth_rss_reta_entry64
*reta_conf,
uint16_t reta_size)
{
struct enic *enic = pmd_priv(dev);
union vnic_rss_cpu rss_cpu;
uint16_t i, idx, shift;
ENICPMD_FUNC_TRACE();
if (reta_size != ENIC_RSS_RETA_SIZE) {
dev_err(enic, "reta_update: wrong reta_size. given=%u"
" expected=%u\n",
reta_size, ENIC_RSS_RETA_SIZE);
return -EINVAL;
}
/*
* Start with the current reta and modify it per reta_conf, as we
* need to push the entire reta even if we only modify one entry.
*/
rss_cpu = enic->rss_cpu;
for (i = 0; i < reta_size; i++) {
idx = i / RTE_RETA_GROUP_SIZE;
shift = i % RTE_RETA_GROUP_SIZE;
if (reta_conf[idx].mask & (1ULL << shift))
rss_cpu.cpu[i / 4].b[i % 4] =
enic_rte_rq_idx_to_sop_idx(
reta_conf[idx].reta[shift]);
}
return enic_set_rss_reta(enic, &rss_cpu);
}
static int enicpmd_dev_rss_hash_update(struct rte_eth_dev *dev,
struct rte_eth_rss_conf *rss_conf)
{
struct enic *enic = pmd_priv(dev);
ENICPMD_FUNC_TRACE();
return enic_set_rss_conf(enic, rss_conf);
}
static int enicpmd_dev_rss_hash_conf_get(struct rte_eth_dev *dev,
struct rte_eth_rss_conf *rss_conf)
{
struct enic *enic = pmd_priv(dev);
ENICPMD_FUNC_TRACE();
if (rss_conf == NULL)
return -EINVAL;
if (rss_conf->rss_key != NULL &&
rss_conf->rss_key_len < ENIC_RSS_HASH_KEY_SIZE) {
dev_err(enic, "rss_hash_conf_get: wrong rss_key_len. given=%u"
" expected=%u+\n",
rss_conf->rss_key_len, ENIC_RSS_HASH_KEY_SIZE);
return -EINVAL;
}
rss_conf->rss_hf = enic->rss_hf;
if (rss_conf->rss_key != NULL) {
int i;
for (i = 0; i < ENIC_RSS_HASH_KEY_SIZE; i++) {
rss_conf->rss_key[i] =
enic->rss_key.key[i / 10].b[i % 10];
}
rss_conf->rss_key_len = ENIC_RSS_HASH_KEY_SIZE;
}
return 0;
}
static void enicpmd_dev_rxq_info_get(struct rte_eth_dev *dev,
uint16_t rx_queue_id,
struct rte_eth_rxq_info *qinfo)
{
struct enic *enic = pmd_priv(dev);
struct vnic_rq *rq_sop;
struct vnic_rq *rq_data;
struct rte_eth_rxconf *conf;
uint16_t sop_queue_idx;
uint16_t data_queue_idx;
ENICPMD_FUNC_TRACE();
sop_queue_idx = enic_rte_rq_idx_to_sop_idx(rx_queue_id);
data_queue_idx = enic_rte_rq_idx_to_data_idx(rx_queue_id);
rq_sop = &enic->rq[sop_queue_idx];
rq_data = &enic->rq[data_queue_idx]; /* valid if data_queue_enable */
qinfo->mp = rq_sop->mp;
qinfo->scattered_rx = rq_sop->data_queue_enable;
qinfo->nb_desc = rq_sop->ring.desc_count;
if (qinfo->scattered_rx)
qinfo->nb_desc += rq_data->ring.desc_count;
conf = &qinfo->conf;
memset(conf, 0, sizeof(*conf));
conf->rx_free_thresh = rq_sop->rx_free_thresh;
conf->rx_drop_en = 1;
/*
* Except VLAN stripping (port setting), all the checksum offloads
* are always enabled.
*/
conf->offloads = enic->rx_offload_capa;
if (!enic->ig_vlan_strip_en)
conf->offloads &= ~DEV_RX_OFFLOAD_VLAN_STRIP;
/* rx_thresh and other fields are not applicable for enic */
}
static void enicpmd_dev_txq_info_get(struct rte_eth_dev *dev,
uint16_t tx_queue_id,
struct rte_eth_txq_info *qinfo)
{
struct enic *enic = pmd_priv(dev);
struct vnic_wq *wq = &enic->wq[tx_queue_id];
ENICPMD_FUNC_TRACE();
qinfo->nb_desc = wq->ring.desc_count;
memset(&qinfo->conf, 0, sizeof(qinfo->conf));
qinfo->conf.offloads = wq->offloads;
/* tx_thresh, and all the other fields are not applicable for enic */
}
static int enicpmd_dev_rx_queue_intr_enable(struct rte_eth_dev *eth_dev,
uint16_t rx_queue_id)
{
struct enic *enic = pmd_priv(eth_dev);
ENICPMD_FUNC_TRACE();
vnic_intr_unmask(&enic->intr[rx_queue_id + ENICPMD_RXQ_INTR_OFFSET]);
return 0;
}
static int enicpmd_dev_rx_queue_intr_disable(struct rte_eth_dev *eth_dev,
uint16_t rx_queue_id)
{
struct enic *enic = pmd_priv(eth_dev);
ENICPMD_FUNC_TRACE();
vnic_intr_mask(&enic->intr[rx_queue_id + ENICPMD_RXQ_INTR_OFFSET]);
return 0;
}
static int udp_tunnel_common_check(struct enic *enic,
struct rte_eth_udp_tunnel *tnl)
{
if (tnl->prot_type != RTE_TUNNEL_TYPE_VXLAN)
return -ENOTSUP;
if (!enic->overlay_offload) {
PMD_INIT_LOG(DEBUG, " vxlan (overlay offload) is not "
"supported\n");
return -ENOTSUP;
}
return 0;
}
static int update_vxlan_port(struct enic *enic, uint16_t port)
{
if (vnic_dev_overlay_offload_cfg(enic->vdev,
OVERLAY_CFG_VXLAN_PORT_UPDATE,
port)) {
PMD_INIT_LOG(DEBUG, " failed to update vxlan port\n");
return -EINVAL;
}
PMD_INIT_LOG(DEBUG, " updated vxlan port to %u\n", port);
enic->vxlan_port = port;
return 0;
}
static int enicpmd_dev_udp_tunnel_port_add(struct rte_eth_dev *eth_dev,
struct rte_eth_udp_tunnel *tnl)
{
struct enic *enic = pmd_priv(eth_dev);
int ret;
ENICPMD_FUNC_TRACE();
ret = udp_tunnel_common_check(enic, tnl);
if (ret)
return ret;
/*
* The NIC has 1 configurable VXLAN port number. "Adding" a new port
* number replaces it.
*/
if (tnl->udp_port == enic->vxlan_port || tnl->udp_port == 0) {
PMD_INIT_LOG(DEBUG, " %u is already configured or invalid\n",
tnl->udp_port);
return -EINVAL;
}
return update_vxlan_port(enic, tnl->udp_port);
}
static int enicpmd_dev_udp_tunnel_port_del(struct rte_eth_dev *eth_dev,
struct rte_eth_udp_tunnel *tnl)
{
struct enic *enic = pmd_priv(eth_dev);
int ret;
ENICPMD_FUNC_TRACE();
ret = udp_tunnel_common_check(enic, tnl);
if (ret)
return ret;
/*
* Clear the previously set port number and restore the
* hardware default port number. Some drivers disable VXLAN
* offloads when there are no configured port numbers. But
* enic does not do that as VXLAN is part of overlay offload,
* which is tied to inner RSS and TSO.
*/
if (tnl->udp_port != enic->vxlan_port) {
PMD_INIT_LOG(DEBUG, " %u is not a configured vxlan port\n",
tnl->udp_port);
return -EINVAL;
}
return update_vxlan_port(enic, ENIC_DEFAULT_VXLAN_PORT);
}
static int enicpmd_dev_fw_version_get(struct rte_eth_dev *eth_dev,
char *fw_version, size_t fw_size)
{
struct vnic_devcmd_fw_info *info;
struct enic *enic;
int ret;
ENICPMD_FUNC_TRACE();
if (fw_version == NULL || fw_size <= 0)
return -EINVAL;
enic = pmd_priv(eth_dev);
ret = vnic_dev_fw_info(enic->vdev, &info);
if (ret)
return ret;
snprintf(fw_version, fw_size, "%s %s",
info->fw_version, info->fw_build);
fw_version[fw_size - 1] = '\0';
return 0;
}
static const struct eth_dev_ops enicpmd_eth_dev_ops = {
.dev_configure = enicpmd_dev_configure,
.dev_start = enicpmd_dev_start,
.dev_stop = enicpmd_dev_stop,
.dev_set_link_up = NULL,
.dev_set_link_down = NULL,
.dev_close = enicpmd_dev_close,
.promiscuous_enable = enicpmd_dev_promiscuous_enable,
.promiscuous_disable = enicpmd_dev_promiscuous_disable,
.allmulticast_enable = enicpmd_dev_allmulticast_enable,
.allmulticast_disable = enicpmd_dev_allmulticast_disable,
.link_update = enicpmd_dev_link_update,
.stats_get = enicpmd_dev_stats_get,
.stats_reset = enicpmd_dev_stats_reset,
.queue_stats_mapping_set = NULL,
.dev_infos_get = enicpmd_dev_info_get,
.dev_supported_ptypes_get = enicpmd_dev_supported_ptypes_get,
.mtu_set = enicpmd_mtu_set,
.vlan_filter_set = NULL,
.vlan_tpid_set = NULL,
.vlan_offload_set = enicpmd_vlan_offload_set,
.vlan_strip_queue_set = NULL,
.rx_queue_start = enicpmd_dev_rx_queue_start,
.rx_queue_stop = enicpmd_dev_rx_queue_stop,
.tx_queue_start = enicpmd_dev_tx_queue_start,
.tx_queue_stop = enicpmd_dev_tx_queue_stop,
.rx_queue_setup = enicpmd_dev_rx_queue_setup,
.rx_queue_release = enicpmd_dev_rx_queue_release,
.rx_queue_count = enicpmd_dev_rx_queue_count,
.rx_descriptor_done = NULL,
.tx_queue_setup = enicpmd_dev_tx_queue_setup,
.tx_queue_release = enicpmd_dev_tx_queue_release,
.rx_queue_intr_enable = enicpmd_dev_rx_queue_intr_enable,
.rx_queue_intr_disable = enicpmd_dev_rx_queue_intr_disable,
.rxq_info_get = enicpmd_dev_rxq_info_get,
.txq_info_get = enicpmd_dev_txq_info_get,
.dev_led_on = NULL,
.dev_led_off = NULL,
.flow_ctrl_get = NULL,
.flow_ctrl_set = NULL,
.priority_flow_ctrl_set = NULL,
.mac_addr_add = enicpmd_add_mac_addr,
.mac_addr_remove = enicpmd_remove_mac_addr,
.mac_addr_set = enicpmd_set_mac_addr,
.set_mc_addr_list = enicpmd_set_mc_addr_list,
.filter_ctrl = enicpmd_dev_filter_ctrl,
.reta_query = enicpmd_dev_rss_reta_query,
.reta_update = enicpmd_dev_rss_reta_update,
.rss_hash_conf_get = enicpmd_dev_rss_hash_conf_get,
.rss_hash_update = enicpmd_dev_rss_hash_update,
.udp_tunnel_port_add = enicpmd_dev_udp_tunnel_port_add,
.udp_tunnel_port_del = enicpmd_dev_udp_tunnel_port_del,
.fw_version_get = enicpmd_dev_fw_version_get,
};
static int enic_parse_zero_one(const char *key,
const char *value,
void *opaque)
{
struct enic *enic;
bool b;
enic = (struct enic *)opaque;
if (strcmp(value, "0") == 0) {
b = false;
} else if (strcmp(value, "1") == 0) {
b = true;
} else {
dev_err(enic, "Invalid value for %s"
": expected=0|1 given=%s\n", key, value);
return -EINVAL;
}
if (strcmp(key, ENIC_DEVARG_DISABLE_OVERLAY) == 0)
enic->disable_overlay = b;
if (strcmp(key, ENIC_DEVARG_ENABLE_AVX2_RX) == 0)
enic->enable_avx2_rx = b;
return 0;
}
static int enic_parse_ig_vlan_rewrite(__rte_unused const char *key,
const char *value,
void *opaque)
{
struct enic *enic;
enic = (struct enic *)opaque;
if (strcmp(value, "trunk") == 0) {
/* Trunk mode: always tag */
enic->ig_vlan_rewrite_mode = IG_VLAN_REWRITE_MODE_DEFAULT_TRUNK;
} else if (strcmp(value, "untag") == 0) {
/* Untag default VLAN mode: untag if VLAN = default VLAN */
enic->ig_vlan_rewrite_mode =
IG_VLAN_REWRITE_MODE_UNTAG_DEFAULT_VLAN;
} else if (strcmp(value, "priority") == 0) {
/*
* Priority-tag default VLAN mode: priority tag (VLAN header
* with ID=0) if VLAN = default
*/
enic->ig_vlan_rewrite_mode =
IG_VLAN_REWRITE_MODE_PRIORITY_TAG_DEFAULT_VLAN;
} else if (strcmp(value, "pass") == 0) {
/* Pass through mode: do not touch tags */
enic->ig_vlan_rewrite_mode = IG_VLAN_REWRITE_MODE_PASS_THRU;
} else {
dev_err(enic, "Invalid value for " ENIC_DEVARG_IG_VLAN_REWRITE
": expected=trunk|untag|priority|pass given=%s\n",
value);
return -EINVAL;
}
return 0;
}
static int enic_check_devargs(struct rte_eth_dev *dev)
{
static const char *const valid_keys[] = {
ENIC_DEVARG_DISABLE_OVERLAY,
ENIC_DEVARG_ENABLE_AVX2_RX,
ENIC_DEVARG_IG_VLAN_REWRITE,
NULL};
struct enic *enic = pmd_priv(dev);
struct rte_kvargs *kvlist;
ENICPMD_FUNC_TRACE();
enic->disable_overlay = false;
enic->enable_avx2_rx = false;
enic->ig_vlan_rewrite_mode = IG_VLAN_REWRITE_MODE_PASS_THRU;
if (!dev->device->devargs)
return 0;
kvlist = rte_kvargs_parse(dev->device->devargs->args, valid_keys);
if (!kvlist)
return -EINVAL;
if (rte_kvargs_process(kvlist, ENIC_DEVARG_DISABLE_OVERLAY,
enic_parse_zero_one, enic) < 0 ||
rte_kvargs_process(kvlist, ENIC_DEVARG_ENABLE_AVX2_RX,
enic_parse_zero_one, enic) < 0 ||
rte_kvargs_process(kvlist, ENIC_DEVARG_IG_VLAN_REWRITE,
enic_parse_ig_vlan_rewrite, enic) < 0) {
rte_kvargs_free(kvlist);
return -EINVAL;
}
rte_kvargs_free(kvlist);
return 0;
}
/* Initialize the driver
* It returns 0 on success.
*/
static int eth_enicpmd_dev_init(struct rte_eth_dev *eth_dev)
{
struct rte_pci_device *pdev;
struct rte_pci_addr *addr;
struct enic *enic = pmd_priv(eth_dev);
int err;
ENICPMD_FUNC_TRACE();
enic->port_id = eth_dev->data->port_id;
enic->rte_dev = eth_dev;
eth_dev->dev_ops = &enicpmd_eth_dev_ops;
eth_dev->rx_pkt_burst = &enic_recv_pkts;
eth_dev->tx_pkt_burst = &enic_xmit_pkts;
eth_dev->tx_pkt_prepare = &enic_prep_pkts;
/* Let rte_eth_dev_close() release the port resources */
eth_dev->data->dev_flags |= RTE_ETH_DEV_CLOSE_REMOVE;
pdev = RTE_ETH_DEV_TO_PCI(eth_dev);
rte_eth_copy_pci_info(eth_dev, pdev);
enic->pdev = pdev;
addr = &pdev->addr;
snprintf(enic->bdf_name, ENICPMD_BDF_LENGTH, "%04x:%02x:%02x.%x",
addr->domain, addr->bus, addr->devid, addr->function);
err = enic_check_devargs(eth_dev);
if (err)
return err;
return enic_probe(enic);
}
static int eth_enic_pci_probe(struct rte_pci_driver *pci_drv __rte_unused,
struct rte_pci_device *pci_dev)
{
return rte_eth_dev_pci_generic_probe(pci_dev, sizeof(struct enic),
eth_enicpmd_dev_init);
}
static int eth_enic_pci_remove(struct rte_pci_device *pci_dev)
{
return rte_eth_dev_pci_generic_remove(pci_dev, NULL);
}
static struct rte_pci_driver rte_enic_pmd = {
.id_table = pci_id_enic_map,
.drv_flags = RTE_PCI_DRV_NEED_MAPPING | RTE_PCI_DRV_INTR_LSC |
RTE_PCI_DRV_IOVA_AS_VA,
.probe = eth_enic_pci_probe,
.remove = eth_enic_pci_remove,
};
RTE_PMD_REGISTER_PCI(net_enic, rte_enic_pmd);
RTE_PMD_REGISTER_PCI_TABLE(net_enic, pci_id_enic_map);
RTE_PMD_REGISTER_KMOD_DEP(net_enic, "* igb_uio | uio_pci_generic | vfio-pci");
RTE_PMD_REGISTER_PARAM_STRING(net_enic,
ENIC_DEVARG_DISABLE_OVERLAY "=0|1 "
ENIC_DEVARG_ENABLE_AVX2_RX "=0|1 "
ENIC_DEVARG_IG_VLAN_REWRITE "=trunk|untag|priority|pass");
void enic_add_multicast_addr(struct enic *enic, u8 *addr)
{
vnic_dev_add_addr(enic->vdev, addr);
}
void enic_add_station_addr(struct enic *enic)
{
vnic_dev_add_addr(enic->vdev, enic->mac_addr);
}
