/* SPDX-License-Identifier: BSD-3-Clause * Copyright (c) 2007-2013 Broadcom Corporation. * * Eric Davis * David Christensen * Gary Zambrano * * Copyright (c) 2013-2015 Brocade Communications Systems, Inc. * Copyright (c) 2015-2018 Cavium Inc. * All rights reserved. * www.cavium.com */ #include "bnx2x.h" #include "ecore_init.h" /**** Exe Queue interfaces ****/ /** * ecore_exe_queue_init - init the Exe Queue object * * @o: pointer to the object * @exe_len: length * @owner: pointer to the owner * @validate: validate function pointer * @optimize: optimize function pointer * @exec: execute function pointer * @get: get function pointer */ static void ecore_exe_queue_init(struct bnx2x_softc *sc __rte_unused, struct ecore_exe_queue_obj *o, int exe_len, union ecore_qable_obj *owner, exe_q_validate validate, exe_q_remove remove, exe_q_optimize optimize, exe_q_execute exec, exe_q_get get) { ECORE_MEMSET(o, 0, sizeof(*o)); ECORE_LIST_INIT(&o->exe_queue); ECORE_LIST_INIT(&o->pending_comp); ECORE_SPIN_LOCK_INIT(&o->lock, sc); o->exe_chunk_len = exe_len; o->owner = owner; /* Owner specific callbacks */ o->validate = validate; o->remove = remove; o->optimize = optimize; o->execute = exec; o->get = get; ECORE_MSG(sc, "Setup the execution queue with the chunk length of %d", exe_len); } static void ecore_exe_queue_free_elem(struct bnx2x_softc *sc __rte_unused, struct ecore_exeq_elem *elem) { ECORE_MSG(sc, "Deleting an exe_queue element"); ECORE_FREE(sc, elem, sizeof(*elem)); } static inline int ecore_exe_queue_length(struct ecore_exe_queue_obj *o) { struct ecore_exeq_elem *elem; int cnt = 0; ECORE_SPIN_LOCK_BH(&o->lock); ECORE_LIST_FOR_EACH_ENTRY(elem, &o->exe_queue, link, struct ecore_exeq_elem) cnt++; ECORE_SPIN_UNLOCK_BH(&o->lock); return cnt; } /** * ecore_exe_queue_add - add a new element to the execution queue * * @sc: driver handle * @o: queue * @cmd: new command to add * @restore: true - do not optimize the command * * If the element is optimized or is illegal, frees it. */ static int ecore_exe_queue_add(struct bnx2x_softc *sc, struct ecore_exe_queue_obj *o, struct ecore_exeq_elem *elem, int restore) { int rc; ECORE_SPIN_LOCK_BH(&o->lock); if (!restore) { /* Try to cancel this element queue */ rc = o->optimize(sc, o->owner, elem); if (rc) goto free_and_exit; /* Check if this request is ok */ rc = o->validate(sc, o->owner, elem); if (rc) { ECORE_MSG(sc, "Preamble failed: %d", rc); goto free_and_exit; } } /* If so, add it to the execution queue */ ECORE_LIST_PUSH_TAIL(&elem->link, &o->exe_queue); ECORE_SPIN_UNLOCK_BH(&o->lock); return ECORE_SUCCESS; free_and_exit: ecore_exe_queue_free_elem(sc, elem); ECORE_SPIN_UNLOCK_BH(&o->lock); return rc; } static void __ecore_exe_queue_reset_pending(struct bnx2x_softc *sc, struct ecore_exe_queue_obj *o) { struct ecore_exeq_elem *elem; while (!ECORE_LIST_IS_EMPTY(&o->pending_comp)) { elem = ECORE_LIST_FIRST_ENTRY(&o->pending_comp, struct ecore_exeq_elem, link); ECORE_LIST_REMOVE_ENTRY(&elem->link, &o->pending_comp); ecore_exe_queue_free_elem(sc, elem); } } static inline void ecore_exe_queue_reset_pending(struct bnx2x_softc *sc, struct ecore_exe_queue_obj *o) { ECORE_SPIN_LOCK_BH(&o->lock); __ecore_exe_queue_reset_pending(sc, o); ECORE_SPIN_UNLOCK_BH(&o->lock); } /** * ecore_exe_queue_step - execute one execution chunk atomically * * @sc: driver handle * @o: queue * @ramrod_flags: flags * * (Should be called while holding the exe_queue->lock). */ static int ecore_exe_queue_step(struct bnx2x_softc *sc, struct ecore_exe_queue_obj *o, unsigned long *ramrod_flags) { struct ecore_exeq_elem *elem, spacer; int cur_len = 0, rc; ECORE_MEMSET(&spacer, 0, sizeof(spacer)); /* Next step should not be performed until the current is finished, * unless a DRV_CLEAR_ONLY bit is set. In this case we just want to * properly clear object internals without sending any command to the FW * which also implies there won't be any completion to clear the * 'pending' list. */ if (!ECORE_LIST_IS_EMPTY(&o->pending_comp)) { if (ECORE_TEST_BIT(RAMROD_DRV_CLR_ONLY, ramrod_flags)) { ECORE_MSG(sc, "RAMROD_DRV_CLR_ONLY requested: resetting a pending_comp list"); __ecore_exe_queue_reset_pending(sc, o); } else { return ECORE_PENDING; } } /* Run through the pending commands list and create a next * execution chunk. */ while (!ECORE_LIST_IS_EMPTY(&o->exe_queue)) { elem = ECORE_LIST_FIRST_ENTRY(&o->exe_queue, struct ecore_exeq_elem, link); ECORE_DBG_BREAK_IF(!elem->cmd_len); if (cur_len + elem->cmd_len <= o->exe_chunk_len) { cur_len += elem->cmd_len; /* Prevent from both lists being empty when moving an * element. This will allow the call of * ecore_exe_queue_empty() without locking. */ ECORE_LIST_PUSH_TAIL(&spacer.link, &o->pending_comp); mb(); ECORE_LIST_REMOVE_ENTRY(&elem->link, &o->exe_queue); ECORE_LIST_PUSH_TAIL(&elem->link, &o->pending_comp); ECORE_LIST_REMOVE_ENTRY(&spacer.link, &o->pending_comp); } else break; } /* Sanity check */ if (!cur_len) return ECORE_SUCCESS; rc = o->execute(sc, o->owner, &o->pending_comp, ramrod_flags); if (rc < 0) /* In case of an error return the commands back to the queue * and reset the pending_comp. */ ECORE_LIST_SPLICE_INIT(&o->pending_comp, &o->exe_queue); else if (!rc) /* If zero is returned, means there are no outstanding pending * completions and we may dismiss the pending list. */ __ecore_exe_queue_reset_pending(sc, o); return rc; } static inline int ecore_exe_queue_empty(struct ecore_exe_queue_obj *o) { int empty = ECORE_LIST_IS_EMPTY(&o->exe_queue); /* Don't reorder!!! */ mb(); return empty && ECORE_LIST_IS_EMPTY(&o->pending_comp); } static struct ecore_exeq_elem *ecore_exe_queue_alloc_elem(struct bnx2x_softc *sc __rte_unused) { ECORE_MSG(sc, "Allocating a new exe_queue element"); return ECORE_ZALLOC(sizeof(struct ecore_exeq_elem), GFP_ATOMIC, sc); } /************************ raw_obj functions ***********************************/ static bool ecore_raw_check_pending(struct ecore_raw_obj *o) { /* * !! converts the value returned by ECORE_TEST_BIT such that it * is guaranteed not to be truncated regardless of int definition. * * Note we cannot simply define the function's return value type * to match the type returned by ECORE_TEST_BIT, as it varies by * platform/implementation. */ return ! !ECORE_TEST_BIT(o->state, o->pstate); } static void ecore_raw_clear_pending(struct ecore_raw_obj *o) { ECORE_SMP_MB_BEFORE_CLEAR_BIT(); ECORE_CLEAR_BIT(o->state, o->pstate); ECORE_SMP_MB_AFTER_CLEAR_BIT(); } static void ecore_raw_set_pending(struct ecore_raw_obj *o) { ECORE_SMP_MB_BEFORE_CLEAR_BIT(); ECORE_SET_BIT(o->state, o->pstate); ECORE_SMP_MB_AFTER_CLEAR_BIT(); } /** * ecore_state_wait - wait until the given bit(state) is cleared * * @sc: device handle * @state: state which is to be cleared * @state_p: state buffer * */ static int ecore_state_wait(struct bnx2x_softc *sc, int state, unsigned long *pstate) { /* can take a while if any port is running */ int cnt = 5000; if (CHIP_REV_IS_EMUL(sc)) cnt *= 20; ECORE_MSG(sc, "waiting for state to become %d", state); ECORE_MIGHT_SLEEP(); while (cnt--) { bnx2x_intr_legacy(sc); if (!ECORE_TEST_BIT(state, pstate)) { #ifdef ECORE_STOP_ON_ERROR ECORE_MSG(sc, "exit (cnt %d)", 5000 - cnt); #endif rte_atomic32_set(&sc->scan_fp, 0); return ECORE_SUCCESS; } ECORE_WAIT(sc, delay_us); if (sc->panic) { rte_atomic32_set(&sc->scan_fp, 0); return ECORE_IO; } } /* timeout! */ PMD_DRV_LOG(ERR, sc, "timeout waiting for state %d", state); rte_atomic32_set(&sc->scan_fp, 0); #ifdef ECORE_STOP_ON_ERROR ecore_panic(); #endif return ECORE_TIMEOUT; } static int ecore_raw_wait(struct bnx2x_softc *sc, struct ecore_raw_obj *raw) { return ecore_state_wait(sc, raw->state, raw->pstate); } /***************** Classification verbs: Set/Del MAC/VLAN/VLAN-MAC ************/ /* credit handling callbacks */ static bool ecore_get_cam_offset_mac(struct ecore_vlan_mac_obj *o, int *offset) { struct ecore_credit_pool_obj *mp = o->macs_pool; ECORE_DBG_BREAK_IF(!mp); return mp->get_entry(mp, offset); } static bool ecore_get_credit_mac(struct ecore_vlan_mac_obj *o) { struct ecore_credit_pool_obj *mp = o->macs_pool; ECORE_DBG_BREAK_IF(!mp); return mp->get(mp, 1); } static bool ecore_put_cam_offset_mac(struct ecore_vlan_mac_obj *o, int offset) { struct ecore_credit_pool_obj *mp = o->macs_pool; return mp->put_entry(mp, offset); } static bool ecore_put_credit_mac(struct ecore_vlan_mac_obj *o) { struct ecore_credit_pool_obj *mp = o->macs_pool; return mp->put(mp, 1); } /** * __ecore_vlan_mac_h_write_trylock - try getting the writer lock on vlan mac * head list. * * @sc: device handle * @o: vlan_mac object * * @details: Non-blocking implementation; should be called under execution * queue lock. */ static int __ecore_vlan_mac_h_write_trylock(struct bnx2x_softc *sc __rte_unused, struct ecore_vlan_mac_obj *o) { if (o->head_reader) { ECORE_MSG(sc, "vlan_mac_lock writer - There are readers; Busy"); return ECORE_BUSY; } ECORE_MSG(sc, "vlan_mac_lock writer - Taken"); return ECORE_SUCCESS; } /** * __ecore_vlan_mac_h_exec_pending - execute step instead of a previous step * which wasn't able to run due to a taken lock on vlan mac head list. * * @sc: device handle * @o: vlan_mac object * * @details Should be called under execution queue lock; notice it might release * and reclaim it during its run. */ static void __ecore_vlan_mac_h_exec_pending(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o) { int rc; unsigned long ramrod_flags = o->saved_ramrod_flags; ECORE_MSG(sc, "vlan_mac_lock execute pending command with ramrod flags %lu", ramrod_flags); o->head_exe_request = FALSE; o->saved_ramrod_flags = 0; rc = ecore_exe_queue_step(sc, &o->exe_queue, &ramrod_flags); if (rc != ECORE_SUCCESS) { PMD_DRV_LOG(ERR, sc, "execution of pending commands failed with rc %d", rc); #ifdef ECORE_STOP_ON_ERROR ecore_panic(); #endif } } /** * __ecore_vlan_mac_h_pend - Pend an execution step which couldn't have been * called due to vlan mac head list lock being taken. * * @sc: device handle * @o: vlan_mac object * @ramrod_flags: ramrod flags of missed execution * * @details Should be called under execution queue lock. */ static void __ecore_vlan_mac_h_pend(struct bnx2x_softc *sc __rte_unused, struct ecore_vlan_mac_obj *o, unsigned long ramrod_flags) { o->head_exe_request = TRUE; o->saved_ramrod_flags = ramrod_flags; ECORE_MSG(sc, "Placing pending execution with ramrod flags %lu", ramrod_flags); } /** * __ecore_vlan_mac_h_write_unlock - unlock the vlan mac head list writer lock * * @sc: device handle * @o: vlan_mac object * * @details Should be called under execution queue lock. Notice if a pending * execution exists, it would perform it - possibly releasing and * reclaiming the execution queue lock. */ static void __ecore_vlan_mac_h_write_unlock(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o) { /* It's possible a new pending execution was added since this writer * executed. If so, execute again. [Ad infinitum] */ while (o->head_exe_request) { ECORE_MSG(sc, "vlan_mac_lock - writer release encountered a pending request"); __ecore_vlan_mac_h_exec_pending(sc, o); } } /** * ecore_vlan_mac_h_write_unlock - unlock the vlan mac head list writer lock * * @sc: device handle * @o: vlan_mac object * * @details Notice if a pending execution exists, it would perform it - * possibly releasing and reclaiming the execution queue lock. */ void ecore_vlan_mac_h_write_unlock(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o) { ECORE_SPIN_LOCK_BH(&o->exe_queue.lock); __ecore_vlan_mac_h_write_unlock(sc, o); ECORE_SPIN_UNLOCK_BH(&o->exe_queue.lock); } /** * __ecore_vlan_mac_h_read_lock - lock the vlan mac head list reader lock * * @sc: device handle * @o: vlan_mac object * * @details Should be called under the execution queue lock. May sleep. May * release and reclaim execution queue lock during its run. */ static int __ecore_vlan_mac_h_read_lock(struct bnx2x_softc *sc __rte_unused, struct ecore_vlan_mac_obj *o) { /* If we got here, we're holding lock --> no WRITER exists */ o->head_reader++; ECORE_MSG(sc, "vlan_mac_lock - locked reader - number %d", o->head_reader); return ECORE_SUCCESS; } /** * ecore_vlan_mac_h_read_lock - lock the vlan mac head list reader lock * * @sc: device handle * @o: vlan_mac object * * @details May sleep. Claims and releases execution queue lock during its run. */ int ecore_vlan_mac_h_read_lock(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o) { int rc; ECORE_SPIN_LOCK_BH(&o->exe_queue.lock); rc = __ecore_vlan_mac_h_read_lock(sc, o); ECORE_SPIN_UNLOCK_BH(&o->exe_queue.lock); return rc; } /** * __ecore_vlan_mac_h_read_unlock - unlock the vlan mac head list reader lock * * @sc: device handle * @o: vlan_mac object * * @details Should be called under execution queue lock. Notice if a pending * execution exists, it would be performed if this was the last * reader. possibly releasing and reclaiming the execution queue lock. */ static void __ecore_vlan_mac_h_read_unlock(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o) { if (!o->head_reader) { PMD_DRV_LOG(ERR, sc, "Need to release vlan mac reader lock, but lock isn't taken"); #ifdef ECORE_STOP_ON_ERROR ecore_panic(); #endif } else { o->head_reader--; ECORE_MSG(sc, "vlan_mac_lock - decreased readers to %d", o->head_reader); } /* It's possible a new pending execution was added, and that this reader * was last - if so we need to execute the command. */ if (!o->head_reader && o->head_exe_request) { ECORE_MSG(sc, "vlan_mac_lock - reader release encountered a pending request"); /* Writer release will do the trick */ __ecore_vlan_mac_h_write_unlock(sc, o); } } /** * ecore_vlan_mac_h_read_unlock - unlock the vlan mac head list reader lock * * @sc: device handle * @o: vlan_mac object * * @details Notice if a pending execution exists, it would be performed if this * was the last reader. Claims and releases the execution queue lock * during its run. */ void ecore_vlan_mac_h_read_unlock(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o) { ECORE_SPIN_LOCK_BH(&o->exe_queue.lock); __ecore_vlan_mac_h_read_unlock(sc, o); ECORE_SPIN_UNLOCK_BH(&o->exe_queue.lock); } /** * ecore_vlan_mac_h_read_unlock - unlock the vlan mac head list reader lock * * @sc: device handle * @o: vlan_mac object * @n: number of elements to get * @base: base address for element placement * @stride: stride between elements (in bytes) */ static int ecore_get_n_elements(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o, int n, uint8_t * base, uint8_t stride, uint8_t size) { struct ecore_vlan_mac_registry_elem *pos; uint8_t *next = base; int counter = 0, read_lock; ECORE_MSG(sc, "get_n_elements - taking vlan_mac_lock (reader)"); read_lock = ecore_vlan_mac_h_read_lock(sc, o); if (read_lock != ECORE_SUCCESS) PMD_DRV_LOG(ERR, sc, "get_n_elements failed to get vlan mac reader lock; Access without lock"); /* traverse list */ ECORE_LIST_FOR_EACH_ENTRY(pos, &o->head, link, struct ecore_vlan_mac_registry_elem) { if (counter < n) { ECORE_MEMCPY(next, &pos->u, size); counter++; ECORE_MSG (sc, "copied element number %d to address %p element was:", counter, next); next += stride + size; } } if (read_lock == ECORE_SUCCESS) { ECORE_MSG(sc, "get_n_elements - releasing vlan_mac_lock (reader)"); ecore_vlan_mac_h_read_unlock(sc, o); } return counter * ETH_ALEN; } /* check_add() callbacks */ static int ecore_check_mac_add(struct bnx2x_softc *sc __rte_unused, struct ecore_vlan_mac_obj *o, union ecore_classification_ramrod_data *data) { struct ecore_vlan_mac_registry_elem *pos; ECORE_MSG(sc, "Checking MAC %02x:%02x:%02x:%02x:%02x:%02x for ADD command", data->mac.mac[0], data->mac.mac[1], data->mac.mac[2], data->mac.mac[3], data->mac.mac[4], data->mac.mac[5]); if (!ECORE_IS_VALID_ETHER_ADDR(data->mac.mac)) return ECORE_INVAL; /* Check if a requested MAC already exists */ ECORE_LIST_FOR_EACH_ENTRY(pos, &o->head, link, struct ecore_vlan_mac_registry_elem) if (!ECORE_MEMCMP(data->mac.mac, pos->u.mac.mac, ETH_ALEN) && (data->mac.is_inner_mac == pos->u.mac.is_inner_mac)) return ECORE_EXISTS; return ECORE_SUCCESS; } /* check_del() callbacks */ static struct ecore_vlan_mac_registry_elem *ecore_check_mac_del(struct bnx2x_softc *sc __rte_unused, struct ecore_vlan_mac_obj *o, union ecore_classification_ramrod_data *data) { struct ecore_vlan_mac_registry_elem *pos; ECORE_MSG(sc, "Checking MAC %02x:%02x:%02x:%02x:%02x:%02x for DEL command", data->mac.mac[0], data->mac.mac[1], data->mac.mac[2], data->mac.mac[3], data->mac.mac[4], data->mac.mac[5]); ECORE_LIST_FOR_EACH_ENTRY(pos, &o->head, link, struct ecore_vlan_mac_registry_elem) if ((!ECORE_MEMCMP(data->mac.mac, pos->u.mac.mac, ETH_ALEN)) && (data->mac.is_inner_mac == pos->u.mac.is_inner_mac)) return pos; return NULL; } /* check_move() callback */ static bool ecore_check_move(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *src_o, struct ecore_vlan_mac_obj *dst_o, union ecore_classification_ramrod_data *data) { struct ecore_vlan_mac_registry_elem *pos; int rc; /* Check if we can delete the requested configuration from the first * object. */ pos = src_o->check_del(sc, src_o, data); /* check if configuration can be added */ rc = dst_o->check_add(sc, dst_o, data); /* If this classification can not be added (is already set) * or can't be deleted - return an error. */ if (rc || !pos) return FALSE; return TRUE; } static bool ecore_check_move_always_err(__rte_unused struct bnx2x_softc *sc, __rte_unused struct ecore_vlan_mac_obj *src_o, __rte_unused struct ecore_vlan_mac_obj *dst_o, __rte_unused union ecore_classification_ramrod_data *data) { return FALSE; } static uint8_t ecore_vlan_mac_get_rx_tx_flag(struct ecore_vlan_mac_obj *o) { struct ecore_raw_obj *raw = &o->raw; uint8_t rx_tx_flag = 0; if ((raw->obj_type == ECORE_OBJ_TYPE_TX) || (raw->obj_type == ECORE_OBJ_TYPE_RX_TX)) rx_tx_flag |= ETH_CLASSIFY_CMD_HEADER_TX_CMD; if ((raw->obj_type == ECORE_OBJ_TYPE_RX) || (raw->obj_type == ECORE_OBJ_TYPE_RX_TX)) rx_tx_flag |= ETH_CLASSIFY_CMD_HEADER_RX_CMD; return rx_tx_flag; } void ecore_set_mac_in_nig(struct bnx2x_softc *sc, bool add, unsigned char *dev_addr, int index) { uint32_t wb_data[2]; uint32_t reg_offset = ECORE_PORT_ID(sc) ? NIG_REG_LLH1_FUNC_MEM : NIG_REG_LLH0_FUNC_MEM; if (!ECORE_IS_MF_SI_MODE(sc) && !IS_MF_AFEX(sc)) return; if (index > ECORE_LLH_CAM_MAX_PF_LINE) return; ECORE_MSG(sc, "Going to %s LLH configuration at entry %d", (add ? "ADD" : "DELETE"), index); if (add) { /* LLH_FUNC_MEM is a uint64_t WB register */ reg_offset += 8 * index; wb_data[0] = ((dev_addr[2] << 24) | (dev_addr[3] << 16) | (dev_addr[4] << 8) | dev_addr[5]); wb_data[1] = ((dev_addr[0] << 8) | dev_addr[1]); ECORE_REG_WR_DMAE_LEN(sc, reg_offset, wb_data, 2); } REG_WR(sc, (ECORE_PORT_ID(sc) ? NIG_REG_LLH1_FUNC_MEM_ENABLE : NIG_REG_LLH0_FUNC_MEM_ENABLE) + 4 * index, add); } /** * ecore_vlan_mac_set_cmd_hdr_e2 - set a header in a single classify ramrod * * @sc: device handle * @o: queue for which we want to configure this rule * @add: if TRUE the command is an ADD command, DEL otherwise * @opcode: CLASSIFY_RULE_OPCODE_XXX * @hdr: pointer to a header to setup * */ static void ecore_vlan_mac_set_cmd_hdr_e2(struct ecore_vlan_mac_obj *o, bool add, int opcode, struct eth_classify_cmd_header *hdr) { struct ecore_raw_obj *raw = &o->raw; hdr->client_id = raw->cl_id; hdr->func_id = raw->func_id; /* Rx or/and Tx (internal switching) configuration ? */ hdr->cmd_general_data |= ecore_vlan_mac_get_rx_tx_flag(o); if (add) hdr->cmd_general_data |= ETH_CLASSIFY_CMD_HEADER_IS_ADD; hdr->cmd_general_data |= (opcode << ETH_CLASSIFY_CMD_HEADER_OPCODE_SHIFT); } /** * ecore_vlan_mac_set_rdata_hdr_e2 - set the classify ramrod data header * * @cid: connection id * @type: ECORE_FILTER_XXX_PENDING * @hdr: pointer to header to setup * @rule_cnt: * * currently we always configure one rule and echo field to contain a CID and an * opcode type. */ static void ecore_vlan_mac_set_rdata_hdr_e2(uint32_t cid, int type, struct eth_classify_header *hdr, int rule_cnt) { hdr->echo = ECORE_CPU_TO_LE32((cid & ECORE_SWCID_MASK) | (type << ECORE_SWCID_SHIFT)); hdr->rule_cnt = (uint8_t) rule_cnt; } /* hw_config() callbacks */ static void ecore_set_one_mac_e2(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o, struct ecore_exeq_elem *elem, int rule_idx, __rte_unused int cam_offset) { struct ecore_raw_obj *raw = &o->raw; struct eth_classify_rules_ramrod_data *data = (struct eth_classify_rules_ramrod_data *)(raw->rdata); int rule_cnt = rule_idx + 1, cmd = elem->cmd_data.vlan_mac.cmd; union eth_classify_rule_cmd *rule_entry = &data->rules[rule_idx]; bool add = (cmd == ECORE_VLAN_MAC_ADD) ? TRUE : FALSE; unsigned long *vlan_mac_flags = &elem->cmd_data.vlan_mac.vlan_mac_flags; uint8_t *mac = elem->cmd_data.vlan_mac.u.mac.mac; /* Set LLH CAM entry: currently only iSCSI and ETH macs are * relevant. In addition, current implementation is tuned for a * single ETH MAC. * * When multiple unicast ETH MACs PF configuration in switch * independent mode is required (NetQ, multiple netdev MACs, * etc.), consider better utilisation of 8 per function MAC * entries in the LLH register. There is also * NIG_REG_P[01]_LLH_FUNC_MEM2 registers that complete the * total number of CAM entries to 16. * * Currently we won't configure NIG for MACs other than a primary ETH * MAC and iSCSI L2 MAC. * * If this MAC is moving from one Queue to another, no need to change * NIG configuration. */ if (cmd != ECORE_VLAN_MAC_MOVE) { if (ECORE_TEST_BIT(ECORE_ISCSI_ETH_MAC, vlan_mac_flags)) ecore_set_mac_in_nig(sc, add, mac, ECORE_LLH_CAM_ISCSI_ETH_LINE); else if (ECORE_TEST_BIT(ECORE_ETH_MAC, vlan_mac_flags)) ecore_set_mac_in_nig(sc, add, mac, ECORE_LLH_CAM_ETH_LINE); } /* Reset the ramrod data buffer for the first rule */ if (rule_idx == 0) ECORE_MEMSET(data, 0, sizeof(*data)); /* Setup a command header */ ecore_vlan_mac_set_cmd_hdr_e2(o, add, CLASSIFY_RULE_OPCODE_MAC, &rule_entry->mac.header); ECORE_MSG(sc, "About to %s MAC %02x:%02x:%02x:%02x:%02x:%02x for Queue %d", (add ? "add" : "delete"), mac[0], mac[1], mac[2], mac[3], mac[4], mac[5], raw->cl_id); /* Set a MAC itself */ ecore_set_fw_mac_addr(&rule_entry->mac.mac_msb, &rule_entry->mac.mac_mid, &rule_entry->mac.mac_lsb, mac); rule_entry->mac.inner_mac = elem->cmd_data.vlan_mac.u.mac.is_inner_mac; /* MOVE: Add a rule that will add this MAC to the target Queue */ if (cmd == ECORE_VLAN_MAC_MOVE) { rule_entry++; rule_cnt++; /* Setup ramrod data */ ecore_vlan_mac_set_cmd_hdr_e2(elem->cmd_data. vlan_mac.target_obj, TRUE, CLASSIFY_RULE_OPCODE_MAC, &rule_entry->mac.header); /* Set a MAC itself */ ecore_set_fw_mac_addr(&rule_entry->mac.mac_msb, &rule_entry->mac.mac_mid, &rule_entry->mac.mac_lsb, mac); rule_entry->mac.inner_mac = elem->cmd_data.vlan_mac.u.mac.is_inner_mac; } /* Set the ramrod data header */ ecore_vlan_mac_set_rdata_hdr_e2(raw->cid, raw->state, &data->header, rule_cnt); } /** * ecore_vlan_mac_set_rdata_hdr_e1x - set a header in a single classify ramrod * * @sc: device handle * @o: queue * @type: * @cam_offset: offset in cam memory * @hdr: pointer to a header to setup * * E1H */ static void ecore_vlan_mac_set_rdata_hdr_e1x(struct ecore_vlan_mac_obj *o, int type, int cam_offset, struct mac_configuration_hdr *hdr) { struct ecore_raw_obj *r = &o->raw; hdr->length = 1; hdr->offset = (uint8_t) cam_offset; hdr->client_id = ECORE_CPU_TO_LE16(0xff); hdr->echo = ECORE_CPU_TO_LE32((r->cid & ECORE_SWCID_MASK) | (type << ECORE_SWCID_SHIFT)); } static void ecore_vlan_mac_set_cfg_entry_e1x(struct ecore_vlan_mac_obj *o, int add, int opcode, uint8_t * mac, uint16_t vlan_id, struct mac_configuration_entry *cfg_entry) { struct ecore_raw_obj *r = &o->raw; uint32_t cl_bit_vec = (1 << r->cl_id); cfg_entry->clients_bit_vector = ECORE_CPU_TO_LE32(cl_bit_vec); cfg_entry->pf_id = r->func_id; cfg_entry->vlan_id = ECORE_CPU_TO_LE16(vlan_id); if (add) { ECORE_SET_FLAG(cfg_entry->flags, MAC_CONFIGURATION_ENTRY_ACTION_TYPE, T_ETH_MAC_COMMAND_SET); ECORE_SET_FLAG(cfg_entry->flags, MAC_CONFIGURATION_ENTRY_VLAN_FILTERING_MODE, opcode); /* Set a MAC in a ramrod data */ ecore_set_fw_mac_addr(&cfg_entry->msb_mac_addr, &cfg_entry->middle_mac_addr, &cfg_entry->lsb_mac_addr, mac); } else ECORE_SET_FLAG(cfg_entry->flags, MAC_CONFIGURATION_ENTRY_ACTION_TYPE, T_ETH_MAC_COMMAND_INVALIDATE); } static void ecore_vlan_mac_set_rdata_e1x(struct bnx2x_softc *sc __rte_unused, struct ecore_vlan_mac_obj *o, int type, int cam_offset, int add, uint8_t * mac, uint16_t vlan_id, int opcode, struct mac_configuration_cmd *config) { struct mac_configuration_entry *cfg_entry = &config->config_table[0]; ecore_vlan_mac_set_rdata_hdr_e1x(o, type, cam_offset, &config->hdr); ecore_vlan_mac_set_cfg_entry_e1x(o, add, opcode, mac, vlan_id, cfg_entry); ECORE_MSG(sc, "%s MAC %02x:%02x:%02x:%02x:%02x:%02x CLID %d CAM offset %d", (add ? "setting" : "clearing"), mac[0], mac[1], mac[2], mac[3], mac[4], mac[5], o->raw.cl_id, cam_offset); } /** * ecore_set_one_mac_e1x - fill a single MAC rule ramrod data * * @sc: device handle * @o: ecore_vlan_mac_obj * @elem: ecore_exeq_elem * @rule_idx: rule_idx * @cam_offset: cam_offset */ static void ecore_set_one_mac_e1x(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o, struct ecore_exeq_elem *elem, __rte_unused int rule_idx, int cam_offset) { struct ecore_raw_obj *raw = &o->raw; struct mac_configuration_cmd *config = (struct mac_configuration_cmd *)(raw->rdata); /* 57711 do not support MOVE command, * so it's either ADD or DEL */ int add = (elem->cmd_data.vlan_mac.cmd == ECORE_VLAN_MAC_ADD) ? TRUE : FALSE; /* Reset the ramrod data buffer */ ECORE_MEMSET(config, 0, sizeof(*config)); ecore_vlan_mac_set_rdata_e1x(sc, o, raw->state, cam_offset, add, elem->cmd_data.vlan_mac.u.mac.mac, 0, ETH_VLAN_FILTER_ANY_VLAN, config); } /** * ecore_vlan_mac_restore - reconfigure next MAC/VLAN/VLAN-MAC element * * @sc: device handle * @p: command parameters * @ppos: pointer to the cookie * * reconfigure next MAC/VLAN/VLAN-MAC element from the * previously configured elements list. * * from command parameters only RAMROD_COMP_WAIT bit in ramrod_flags is taken * into an account * * pointer to the cookie - that should be given back in the next call to make * function handle the next element. If *ppos is set to NULL it will restart the * iterator. If returned *ppos == NULL this means that the last element has been * handled. * */ static int ecore_vlan_mac_restore(struct bnx2x_softc *sc, struct ecore_vlan_mac_ramrod_params *p, struct ecore_vlan_mac_registry_elem **ppos) { struct ecore_vlan_mac_registry_elem *pos; struct ecore_vlan_mac_obj *o = p->vlan_mac_obj; /* If list is empty - there is nothing to do here */ if (ECORE_LIST_IS_EMPTY(&o->head)) { *ppos = NULL; return 0; } /* make a step... */ if (*ppos == NULL) *ppos = ECORE_LIST_FIRST_ENTRY(&o->head, struct ecore_vlan_mac_registry_elem, link); else *ppos = ECORE_LIST_NEXT(*ppos, link, struct ecore_vlan_mac_registry_elem); pos = *ppos; /* If it's the last step - return NULL */ if (ECORE_LIST_IS_LAST(&pos->link, &o->head)) *ppos = NULL; /* Prepare a 'user_req' */ ECORE_MEMCPY(&p->user_req.u, &pos->u, sizeof(pos->u)); /* Set the command */ p->user_req.cmd = ECORE_VLAN_MAC_ADD; /* Set vlan_mac_flags */ p->user_req.vlan_mac_flags = pos->vlan_mac_flags; /* Set a restore bit */ ECORE_SET_BIT_NA(RAMROD_RESTORE, &p->ramrod_flags); return ecore_config_vlan_mac(sc, p); } /* ecore_exeq_get_mac/ecore_exeq_get_vlan/ecore_exeq_get_vlan_mac return a * pointer to an element with a specific criteria and NULL if such an element * hasn't been found. */ static struct ecore_exeq_elem *ecore_exeq_get_mac(struct ecore_exe_queue_obj *o, struct ecore_exeq_elem *elem) { struct ecore_exeq_elem *pos; struct ecore_mac_ramrod_data *data = &elem->cmd_data.vlan_mac.u.mac; /* Check pending for execution commands */ ECORE_LIST_FOR_EACH_ENTRY(pos, &o->exe_queue, link, struct ecore_exeq_elem) if (!ECORE_MEMCMP(&pos->cmd_data.vlan_mac.u.mac, data, sizeof(*data)) && (pos->cmd_data.vlan_mac.cmd == elem->cmd_data.vlan_mac.cmd)) return pos; return NULL; } /** * ecore_validate_vlan_mac_add - check if an ADD command can be executed * * @sc: device handle * @qo: ecore_qable_obj * @elem: ecore_exeq_elem * * Checks that the requested configuration can be added. If yes and if * requested, consume CAM credit. * * The 'validate' is run after the 'optimize'. * */ static int ecore_validate_vlan_mac_add(struct bnx2x_softc *sc, union ecore_qable_obj *qo, struct ecore_exeq_elem *elem) { struct ecore_vlan_mac_obj *o = &qo->vlan_mac; struct ecore_exe_queue_obj *exeq = &o->exe_queue; int rc; /* Check the registry */ rc = o->check_add(sc, o, &elem->cmd_data.vlan_mac.u); if (rc) { ECORE_MSG(sc, "ADD command is not allowed considering current registry state."); return rc; } /* Check if there is a pending ADD command for this * MAC/VLAN/VLAN-MAC. Return an error if there is. */ if (exeq->get(exeq, elem)) { ECORE_MSG(sc, "There is a pending ADD command already"); return ECORE_EXISTS; } /* Consume the credit if not requested not to */ if (!(ECORE_TEST_BIT(ECORE_DONT_CONSUME_CAM_CREDIT, &elem->cmd_data.vlan_mac.vlan_mac_flags) || o->get_credit(o))) return ECORE_INVAL; return ECORE_SUCCESS; } /** * ecore_validate_vlan_mac_del - check if the DEL command can be executed * * @sc: device handle * @qo: quable object to check * @elem: element that needs to be deleted * * Checks that the requested configuration can be deleted. If yes and if * requested, returns a CAM credit. * * The 'validate' is run after the 'optimize'. */ static int ecore_validate_vlan_mac_del(struct bnx2x_softc *sc, union ecore_qable_obj *qo, struct ecore_exeq_elem *elem) { struct ecore_vlan_mac_obj *o = &qo->vlan_mac; struct ecore_vlan_mac_registry_elem *pos; struct ecore_exe_queue_obj *exeq = &o->exe_queue; struct ecore_exeq_elem query_elem; /* If this classification can not be deleted (doesn't exist) * - return a ECORE_EXIST. */ pos = o->check_del(sc, o, &elem->cmd_data.vlan_mac.u); if (!pos) { ECORE_MSG(sc, "DEL command is not allowed considering current registry state"); return ECORE_EXISTS; } /* Check if there are pending DEL or MOVE commands for this * MAC/VLAN/VLAN-MAC. Return an error if so. */ ECORE_MEMCPY(&query_elem, elem, sizeof(query_elem)); /* Check for MOVE commands */ query_elem.cmd_data.vlan_mac.cmd = ECORE_VLAN_MAC_MOVE; if (exeq->get(exeq, &query_elem)) { PMD_DRV_LOG(ERR, sc, "There is a pending MOVE command already"); return ECORE_INVAL; } /* Check for DEL commands */ if (exeq->get(exeq, elem)) { ECORE_MSG(sc, "There is a pending DEL command already"); return ECORE_EXISTS; } /* Return the credit to the credit pool if not requested not to */ if (!(ECORE_TEST_BIT(ECORE_DONT_CONSUME_CAM_CREDIT, &elem->cmd_data.vlan_mac.vlan_mac_flags) || o->put_credit(o))) { PMD_DRV_LOG(ERR, sc, "Failed to return a credit"); return ECORE_INVAL; } return ECORE_SUCCESS; } /** * ecore_validate_vlan_mac_move - check if the MOVE command can be executed * * @sc: device handle * @qo: quable object to check (source) * @elem: element that needs to be moved * * Checks that the requested configuration can be moved. If yes and if * requested, returns a CAM credit. * * The 'validate' is run after the 'optimize'. */ static int ecore_validate_vlan_mac_move(struct bnx2x_softc *sc, union ecore_qable_obj *qo, struct ecore_exeq_elem *elem) { struct ecore_vlan_mac_obj *src_o = &qo->vlan_mac; struct ecore_vlan_mac_obj *dest_o = elem->cmd_data.vlan_mac.target_obj; struct ecore_exeq_elem query_elem; struct ecore_exe_queue_obj *src_exeq = &src_o->exe_queue; struct ecore_exe_queue_obj *dest_exeq = &dest_o->exe_queue; /* Check if we can perform this operation based on the current registry * state. */ if (!src_o->check_move(sc, src_o, dest_o, &elem->cmd_data.vlan_mac.u)) { ECORE_MSG(sc, "MOVE command is not allowed considering current registry state"); return ECORE_INVAL; } /* Check if there is an already pending DEL or MOVE command for the * source object or ADD command for a destination object. Return an * error if so. */ ECORE_MEMCPY(&query_elem, elem, sizeof(query_elem)); /* Check DEL on source */ query_elem.cmd_data.vlan_mac.cmd = ECORE_VLAN_MAC_DEL; if (src_exeq->get(src_exeq, &query_elem)) { PMD_DRV_LOG(ERR, sc, "There is a pending DEL command on the source queue already"); return ECORE_INVAL; } /* Check MOVE on source */ if (src_exeq->get(src_exeq, elem)) { ECORE_MSG(sc, "There is a pending MOVE command already"); return ECORE_EXISTS; } /* Check ADD on destination */ query_elem.cmd_data.vlan_mac.cmd = ECORE_VLAN_MAC_ADD; if (dest_exeq->get(dest_exeq, &query_elem)) { PMD_DRV_LOG(ERR, sc, "There is a pending ADD command on the destination queue already"); return ECORE_INVAL; } /* Consume the credit if not requested not to */ if (!(ECORE_TEST_BIT(ECORE_DONT_CONSUME_CAM_CREDIT_DEST, &elem->cmd_data.vlan_mac.vlan_mac_flags) || dest_o->get_credit(dest_o))) return ECORE_INVAL; if (!(ECORE_TEST_BIT(ECORE_DONT_CONSUME_CAM_CREDIT, &elem->cmd_data.vlan_mac.vlan_mac_flags) || src_o->put_credit(src_o))) { /* return the credit taken from dest... */ dest_o->put_credit(dest_o); return ECORE_INVAL; } return ECORE_SUCCESS; } static int ecore_validate_vlan_mac(struct bnx2x_softc *sc, union ecore_qable_obj *qo, struct ecore_exeq_elem *elem) { switch (elem->cmd_data.vlan_mac.cmd) { case ECORE_VLAN_MAC_ADD: return ecore_validate_vlan_mac_add(sc, qo, elem); case ECORE_VLAN_MAC_DEL: return ecore_validate_vlan_mac_del(sc, qo, elem); case ECORE_VLAN_MAC_MOVE: return ecore_validate_vlan_mac_move(sc, qo, elem); default: return ECORE_INVAL; } } static int ecore_remove_vlan_mac(__rte_unused struct bnx2x_softc *sc, union ecore_qable_obj *qo, struct ecore_exeq_elem *elem) { int rc = 0; /* If consumption wasn't required, nothing to do */ if (ECORE_TEST_BIT(ECORE_DONT_CONSUME_CAM_CREDIT, &elem->cmd_data.vlan_mac.vlan_mac_flags)) return ECORE_SUCCESS; switch (elem->cmd_data.vlan_mac.cmd) { case ECORE_VLAN_MAC_ADD: case ECORE_VLAN_MAC_MOVE: rc = qo->vlan_mac.put_credit(&qo->vlan_mac); break; case ECORE_VLAN_MAC_DEL: rc = qo->vlan_mac.get_credit(&qo->vlan_mac); break; default: return ECORE_INVAL; } if (rc != TRUE) return ECORE_INVAL; return ECORE_SUCCESS; } /** * ecore_wait_vlan_mac - passively wait for 5 seconds until all work completes. * * @sc: device handle * @o: ecore_vlan_mac_obj * */ static int ecore_wait_vlan_mac(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o) { int cnt = 5000, rc; struct ecore_exe_queue_obj *exeq = &o->exe_queue; struct ecore_raw_obj *raw = &o->raw; while (cnt--) { /* Wait for the current command to complete */ rc = raw->wait_comp(sc, raw); if (rc) return rc; /* Wait until there are no pending commands */ if (!ecore_exe_queue_empty(exeq)) ECORE_WAIT(sc, 1000); else return ECORE_SUCCESS; } return ECORE_TIMEOUT; } static int __ecore_vlan_mac_execute_step(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o, unsigned long *ramrod_flags) { int rc = ECORE_SUCCESS; ECORE_SPIN_LOCK_BH(&o->exe_queue.lock); ECORE_MSG(sc, "vlan_mac_execute_step - trying to take writer lock"); rc = __ecore_vlan_mac_h_write_trylock(sc, o); if (rc != ECORE_SUCCESS) { __ecore_vlan_mac_h_pend(sc, o, *ramrod_flags); /** Calling function should not diffrentiate between this case * and the case in which there is already a pending ramrod */ rc = ECORE_PENDING; } else { rc = ecore_exe_queue_step(sc, &o->exe_queue, ramrod_flags); } ECORE_SPIN_UNLOCK_BH(&o->exe_queue.lock); return rc; } /** * ecore_complete_vlan_mac - complete one VLAN-MAC ramrod * * @sc: device handle * @o: ecore_vlan_mac_obj * @cqe: * @cont: if TRUE schedule next execution chunk * */ static int ecore_complete_vlan_mac(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o, union event_ring_elem *cqe, unsigned long *ramrod_flags) { struct ecore_raw_obj *r = &o->raw; int rc; /* Reset pending list */ ecore_exe_queue_reset_pending(sc, &o->exe_queue); /* Clear pending */ r->clear_pending(r); /* If ramrod failed this is most likely a SW bug */ if (cqe->message.error) return ECORE_INVAL; /* Run the next bulk of pending commands if requested */ if (ECORE_TEST_BIT(RAMROD_CONT, ramrod_flags)) { rc = __ecore_vlan_mac_execute_step(sc, o, ramrod_flags); if (rc < 0) return rc; } /* If there is more work to do return PENDING */ if (!ecore_exe_queue_empty(&o->exe_queue)) return ECORE_PENDING; return ECORE_SUCCESS; } /** * ecore_optimize_vlan_mac - optimize ADD and DEL commands. * * @sc: device handle * @o: ecore_qable_obj * @elem: ecore_exeq_elem */ static int ecore_optimize_vlan_mac(struct bnx2x_softc *sc, union ecore_qable_obj *qo, struct ecore_exeq_elem *elem) { struct ecore_exeq_elem query, *pos; struct ecore_vlan_mac_obj *o = &qo->vlan_mac; struct ecore_exe_queue_obj *exeq = &o->exe_queue; ECORE_MEMCPY(&query, elem, sizeof(query)); switch (elem->cmd_data.vlan_mac.cmd) { case ECORE_VLAN_MAC_ADD: query.cmd_data.vlan_mac.cmd = ECORE_VLAN_MAC_DEL; break; case ECORE_VLAN_MAC_DEL: query.cmd_data.vlan_mac.cmd = ECORE_VLAN_MAC_ADD; break; default: /* Don't handle anything other than ADD or DEL */ return 0; } /* If we found the appropriate element - delete it */ pos = exeq->get(exeq, &query); if (pos) { /* Return the credit of the optimized command */ if (!ECORE_TEST_BIT(ECORE_DONT_CONSUME_CAM_CREDIT, &pos->cmd_data.vlan_mac.vlan_mac_flags)) { if ((query.cmd_data.vlan_mac.cmd == ECORE_VLAN_MAC_ADD) && !o->put_credit(o)) { PMD_DRV_LOG(ERR, sc, "Failed to return the credit for the optimized ADD command"); return ECORE_INVAL; } else if (!o->get_credit(o)) { /* VLAN_MAC_DEL */ PMD_DRV_LOG(ERR, sc, "Failed to recover the credit from the optimized DEL command"); return ECORE_INVAL; } } ECORE_MSG(sc, "Optimizing %s command", (elem->cmd_data.vlan_mac.cmd == ECORE_VLAN_MAC_ADD) ? "ADD" : "DEL"); ECORE_LIST_REMOVE_ENTRY(&pos->link, &exeq->exe_queue); ecore_exe_queue_free_elem(sc, pos); return 1; } return 0; } /** * ecore_vlan_mac_get_registry_elem - prepare a registry element * * @sc: device handle * @o: * @elem: * @restore: * @re: * * prepare a registry element according to the current command request. */ static int ecore_vlan_mac_get_registry_elem(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o, struct ecore_exeq_elem *elem, int restore, struct ecore_vlan_mac_registry_elem **re) { enum ecore_vlan_mac_cmd cmd = elem->cmd_data.vlan_mac.cmd; struct ecore_vlan_mac_registry_elem *reg_elem; /* Allocate a new registry element if needed. */ if (!restore && ((cmd == ECORE_VLAN_MAC_ADD) || (cmd == ECORE_VLAN_MAC_MOVE))) { reg_elem = ECORE_ZALLOC(sizeof(*reg_elem), GFP_ATOMIC, sc); if (!reg_elem) return ECORE_NOMEM; /* Get a new CAM offset */ if (!o->get_cam_offset(o, ®_elem->cam_offset)) { /* This shall never happen, because we have checked the * CAM availability in the 'validate'. */ ECORE_DBG_BREAK_IF(1); ECORE_FREE(sc, reg_elem, sizeof(*reg_elem)); return ECORE_INVAL; } ECORE_MSG(sc, "Got cam offset %d", reg_elem->cam_offset); /* Set a VLAN-MAC data */ ECORE_MEMCPY(®_elem->u, &elem->cmd_data.vlan_mac.u, sizeof(reg_elem->u)); /* Copy the flags (needed for DEL and RESTORE flows) */ reg_elem->vlan_mac_flags = elem->cmd_data.vlan_mac.vlan_mac_flags; } else /* DEL, RESTORE */ reg_elem = o->check_del(sc, o, &elem->cmd_data.vlan_mac.u); *re = reg_elem; return ECORE_SUCCESS; } /** * ecore_execute_vlan_mac - execute vlan mac command * * @sc: device handle * @qo: * @exe_chunk: * @ramrod_flags: * * go and send a ramrod! */ static int ecore_execute_vlan_mac(struct bnx2x_softc *sc, union ecore_qable_obj *qo, ecore_list_t * exe_chunk, unsigned long *ramrod_flags) { struct ecore_exeq_elem *elem; struct ecore_vlan_mac_obj *o = &qo->vlan_mac, *cam_obj; struct ecore_raw_obj *r = &o->raw; int rc, idx = 0; int restore = ECORE_TEST_BIT(RAMROD_RESTORE, ramrod_flags); int drv_only = ECORE_TEST_BIT(RAMROD_DRV_CLR_ONLY, ramrod_flags); struct ecore_vlan_mac_registry_elem *reg_elem; enum ecore_vlan_mac_cmd cmd; /* If DRIVER_ONLY execution is requested, cleanup a registry * and exit. Otherwise send a ramrod to FW. */ if (!drv_only) { /* Set pending */ r->set_pending(r); /* Fill the ramrod data */ ECORE_LIST_FOR_EACH_ENTRY(elem, exe_chunk, link, struct ecore_exeq_elem) { cmd = elem->cmd_data.vlan_mac.cmd; /* We will add to the target object in MOVE command, so * change the object for a CAM search. */ if (cmd == ECORE_VLAN_MAC_MOVE) cam_obj = elem->cmd_data.vlan_mac.target_obj; else cam_obj = o; rc = ecore_vlan_mac_get_registry_elem(sc, cam_obj, elem, restore, ®_elem); if (rc) goto error_exit; ECORE_DBG_BREAK_IF(!reg_elem); /* Push a new entry into the registry */ if (!restore && ((cmd == ECORE_VLAN_MAC_ADD) || (cmd == ECORE_VLAN_MAC_MOVE))) ECORE_LIST_PUSH_HEAD(®_elem->link, &cam_obj->head); /* Configure a single command in a ramrod data buffer */ o->set_one_rule(sc, o, elem, idx, reg_elem->cam_offset); /* MOVE command consumes 2 entries in the ramrod data */ if (cmd == ECORE_VLAN_MAC_MOVE) idx += 2; else idx++; } /* * No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ rc = ecore_sp_post(sc, o->ramrod_cmd, r->cid, r->rdata_mapping, ETH_CONNECTION_TYPE); if (rc) goto error_exit; } /* Now, when we are done with the ramrod - clean up the registry */ ECORE_LIST_FOR_EACH_ENTRY(elem, exe_chunk, link, struct ecore_exeq_elem) { cmd = elem->cmd_data.vlan_mac.cmd; if ((cmd == ECORE_VLAN_MAC_DEL) || (cmd == ECORE_VLAN_MAC_MOVE)) { reg_elem = o->check_del(sc, o, &elem->cmd_data.vlan_mac.u); ECORE_DBG_BREAK_IF(!reg_elem); o->put_cam_offset(o, reg_elem->cam_offset); ECORE_LIST_REMOVE_ENTRY(®_elem->link, &o->head); ECORE_FREE(sc, reg_elem, sizeof(*reg_elem)); } } if (!drv_only) return ECORE_PENDING; else return ECORE_SUCCESS; error_exit: r->clear_pending(r); /* Cleanup a registry in case of a failure */ ECORE_LIST_FOR_EACH_ENTRY(elem, exe_chunk, link, struct ecore_exeq_elem) { cmd = elem->cmd_data.vlan_mac.cmd; if (cmd == ECORE_VLAN_MAC_MOVE) cam_obj = elem->cmd_data.vlan_mac.target_obj; else cam_obj = o; /* Delete all newly added above entries */ if (!restore && ((cmd == ECORE_VLAN_MAC_ADD) || (cmd == ECORE_VLAN_MAC_MOVE))) { reg_elem = o->check_del(sc, cam_obj, &elem->cmd_data.vlan_mac.u); if (reg_elem) { ECORE_LIST_REMOVE_ENTRY(®_elem->link, &cam_obj->head); ECORE_FREE(sc, reg_elem, sizeof(*reg_elem)); } } } return rc; } static int ecore_vlan_mac_push_new_cmd(struct bnx2x_softc *sc, struct ecore_vlan_mac_ramrod_params *p) { struct ecore_exeq_elem *elem; struct ecore_vlan_mac_obj *o = p->vlan_mac_obj; int restore = ECORE_TEST_BIT(RAMROD_RESTORE, &p->ramrod_flags); /* Allocate the execution queue element */ elem = ecore_exe_queue_alloc_elem(sc); if (!elem) return ECORE_NOMEM; /* Set the command 'length' */ switch (p->user_req.cmd) { case ECORE_VLAN_MAC_MOVE: elem->cmd_len = 2; break; default: elem->cmd_len = 1; } /* Fill the object specific info */ ECORE_MEMCPY(&elem->cmd_data.vlan_mac, &p->user_req, sizeof(p->user_req)); /* Try to add a new command to the pending list */ return ecore_exe_queue_add(sc, &o->exe_queue, elem, restore); } /** * ecore_config_vlan_mac - configure VLAN/MAC/VLAN_MAC filtering rules. * * @sc: device handle * @p: * */ int ecore_config_vlan_mac(struct bnx2x_softc *sc, struct ecore_vlan_mac_ramrod_params *p) { int rc = ECORE_SUCCESS; struct ecore_vlan_mac_obj *o = p->vlan_mac_obj; unsigned long *ramrod_flags = &p->ramrod_flags; int cont = ECORE_TEST_BIT(RAMROD_CONT, ramrod_flags); struct ecore_raw_obj *raw = &o->raw; /* * Add new elements to the execution list for commands that require it. */ if (!cont) { rc = ecore_vlan_mac_push_new_cmd(sc, p); if (rc) return rc; } /* If nothing will be executed further in this iteration we want to * return PENDING if there are pending commands */ if (!ecore_exe_queue_empty(&o->exe_queue)) rc = ECORE_PENDING; if (ECORE_TEST_BIT(RAMROD_DRV_CLR_ONLY, ramrod_flags)) { ECORE_MSG(sc, "RAMROD_DRV_CLR_ONLY requested: clearing a pending bit."); raw->clear_pending(raw); } /* Execute commands if required */ if (cont || ECORE_TEST_BIT(RAMROD_EXEC, ramrod_flags) || ECORE_TEST_BIT(RAMROD_COMP_WAIT, ramrod_flags)) { rc = __ecore_vlan_mac_execute_step(sc, p->vlan_mac_obj, &p->ramrod_flags); if (rc < 0) return rc; } /* RAMROD_COMP_WAIT is a superset of RAMROD_EXEC. If it was set * then user want to wait until the last command is done. */ if (ECORE_TEST_BIT(RAMROD_COMP_WAIT, &p->ramrod_flags)) { /* Wait maximum for the current exe_queue length iterations plus * one (for the current pending command). */ int max_iterations = ecore_exe_queue_length(&o->exe_queue) + 1; while (!ecore_exe_queue_empty(&o->exe_queue) && max_iterations--) { /* Wait for the current command to complete */ rc = raw->wait_comp(sc, raw); if (rc) return rc; /* Make a next step */ rc = __ecore_vlan_mac_execute_step(sc, p->vlan_mac_obj, &p->ramrod_flags); if (rc < 0) return rc; } return ECORE_SUCCESS; } return rc; } /** * ecore_vlan_mac_del_all - delete elements with given vlan_mac_flags spec * * @sc: device handle * @o: * @vlan_mac_flags: * @ramrod_flags: execution flags to be used for this deletion * * if the last operation has completed successfully and there are no * more elements left, positive value if the last operation has completed * successfully and there are more previously configured elements, negative * value is current operation has failed. */ static int ecore_vlan_mac_del_all(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *o, unsigned long *vlan_mac_flags, unsigned long *ramrod_flags) { struct ecore_vlan_mac_registry_elem *pos = NULL; int rc = 0, read_lock; struct ecore_vlan_mac_ramrod_params p; struct ecore_exe_queue_obj *exeq = &o->exe_queue; struct ecore_exeq_elem *exeq_pos, *exeq_pos_n; /* Clear pending commands first */ ECORE_SPIN_LOCK_BH(&exeq->lock); ECORE_LIST_FOR_EACH_ENTRY_SAFE(exeq_pos, exeq_pos_n, &exeq->exe_queue, link, struct ecore_exeq_elem) { if (exeq_pos->cmd_data.vlan_mac.vlan_mac_flags == *vlan_mac_flags) { rc = exeq->remove(sc, exeq->owner, exeq_pos); if (rc) { PMD_DRV_LOG(ERR, sc, "Failed to remove command"); ECORE_SPIN_UNLOCK_BH(&exeq->lock); return rc; } ECORE_LIST_REMOVE_ENTRY(&exeq_pos->link, &exeq->exe_queue); ecore_exe_queue_free_elem(sc, exeq_pos); } } ECORE_SPIN_UNLOCK_BH(&exeq->lock); /* Prepare a command request */ ECORE_MEMSET(&p, 0, sizeof(p)); p.vlan_mac_obj = o; p.ramrod_flags = *ramrod_flags; p.user_req.cmd = ECORE_VLAN_MAC_DEL; /* Add all but the last VLAN-MAC to the execution queue without actually * execution anything. */ ECORE_CLEAR_BIT_NA(RAMROD_COMP_WAIT, &p.ramrod_flags); ECORE_CLEAR_BIT_NA(RAMROD_EXEC, &p.ramrod_flags); ECORE_CLEAR_BIT_NA(RAMROD_CONT, &p.ramrod_flags); ECORE_MSG(sc, "vlan_mac_del_all -- taking vlan_mac_lock (reader)"); read_lock = ecore_vlan_mac_h_read_lock(sc, o); if (read_lock != ECORE_SUCCESS) return read_lock; ECORE_LIST_FOR_EACH_ENTRY(pos, &o->head, link, struct ecore_vlan_mac_registry_elem) { if (pos->vlan_mac_flags == *vlan_mac_flags) { p.user_req.vlan_mac_flags = pos->vlan_mac_flags; ECORE_MEMCPY(&p.user_req.u, &pos->u, sizeof(pos->u)); rc = ecore_config_vlan_mac(sc, &p); if (rc < 0) { PMD_DRV_LOG(ERR, sc, "Failed to add a new DEL command"); ecore_vlan_mac_h_read_unlock(sc, o); return rc; } } } ECORE_MSG(sc, "vlan_mac_del_all -- releasing vlan_mac_lock (reader)"); ecore_vlan_mac_h_read_unlock(sc, o); p.ramrod_flags = *ramrod_flags; ECORE_SET_BIT_NA(RAMROD_CONT, &p.ramrod_flags); return ecore_config_vlan_mac(sc, &p); } static void ecore_init_raw_obj(struct ecore_raw_obj *raw, uint8_t cl_id, uint32_t cid, uint8_t func_id, void *rdata, ecore_dma_addr_t rdata_mapping, int state, unsigned long *pstate, ecore_obj_type type) { raw->func_id = func_id; raw->cid = cid; raw->cl_id = cl_id; raw->rdata = rdata; raw->rdata_mapping = rdata_mapping; raw->state = state; raw->pstate = pstate; raw->obj_type = type; raw->check_pending = ecore_raw_check_pending; raw->clear_pending = ecore_raw_clear_pending; raw->set_pending = ecore_raw_set_pending; raw->wait_comp = ecore_raw_wait; } static void ecore_init_vlan_mac_common(struct ecore_vlan_mac_obj *o, uint8_t cl_id, uint32_t cid, uint8_t func_id, void *rdata, ecore_dma_addr_t rdata_mapping, int state, unsigned long *pstate, ecore_obj_type type, struct ecore_credit_pool_obj *macs_pool, struct ecore_credit_pool_obj *vlans_pool) { ECORE_LIST_INIT(&o->head); o->head_reader = 0; o->head_exe_request = FALSE; o->saved_ramrod_flags = 0; o->macs_pool = macs_pool; o->vlans_pool = vlans_pool; o->delete_all = ecore_vlan_mac_del_all; o->restore = ecore_vlan_mac_restore; o->complete = ecore_complete_vlan_mac; o->wait = ecore_wait_vlan_mac; ecore_init_raw_obj(&o->raw, cl_id, cid, func_id, rdata, rdata_mapping, state, pstate, type); } void ecore_init_mac_obj(struct bnx2x_softc *sc, struct ecore_vlan_mac_obj *mac_obj, uint8_t cl_id, uint32_t cid, uint8_t func_id, void *rdata, ecore_dma_addr_t rdata_mapping, int state, unsigned long *pstate, ecore_obj_type type, struct ecore_credit_pool_obj *macs_pool) { union ecore_qable_obj *qable_obj = (union ecore_qable_obj *)mac_obj; ecore_init_vlan_mac_common(mac_obj, cl_id, cid, func_id, rdata, rdata_mapping, state, pstate, type, macs_pool, NULL); /* CAM credit pool handling */ mac_obj->get_credit = ecore_get_credit_mac; mac_obj->put_credit = ecore_put_credit_mac; mac_obj->get_cam_offset = ecore_get_cam_offset_mac; mac_obj->put_cam_offset = ecore_put_cam_offset_mac; if (CHIP_IS_E1x(sc)) { mac_obj->set_one_rule = ecore_set_one_mac_e1x; mac_obj->check_del = ecore_check_mac_del; mac_obj->check_add = ecore_check_mac_add; mac_obj->check_move = ecore_check_move_always_err; mac_obj->ramrod_cmd = RAMROD_CMD_ID_ETH_SET_MAC; /* Exe Queue */ ecore_exe_queue_init(sc, &mac_obj->exe_queue, 1, qable_obj, ecore_validate_vlan_mac, ecore_remove_vlan_mac, ecore_optimize_vlan_mac, ecore_execute_vlan_mac, ecore_exeq_get_mac); } else { mac_obj->set_one_rule = ecore_set_one_mac_e2; mac_obj->check_del = ecore_check_mac_del; mac_obj->check_add = ecore_check_mac_add; mac_obj->check_move = ecore_check_move; mac_obj->ramrod_cmd = RAMROD_CMD_ID_ETH_CLASSIFICATION_RULES; mac_obj->get_n_elements = ecore_get_n_elements; /* Exe Queue */ ecore_exe_queue_init(sc, &mac_obj->exe_queue, CLASSIFY_RULES_COUNT, qable_obj, ecore_validate_vlan_mac, ecore_remove_vlan_mac, ecore_optimize_vlan_mac, ecore_execute_vlan_mac, ecore_exeq_get_mac); } } /* RX_MODE verbs: DROP_ALL/ACCEPT_ALL/ACCEPT_ALL_MULTI/ACCEPT_ALL_VLAN/NORMAL */ static void __storm_memset_mac_filters(struct bnx2x_softc *sc, struct tstorm_eth_mac_filter_config *mac_filters, uint16_t pf_id) { size_t size = sizeof(struct tstorm_eth_mac_filter_config); uint32_t addr = BAR_TSTRORM_INTMEM + TSTORM_MAC_FILTER_CONFIG_OFFSET(pf_id); ecore_storm_memset_struct(sc, addr, size, (uint32_t *) mac_filters); } static int ecore_set_rx_mode_e1x(struct bnx2x_softc *sc, struct ecore_rx_mode_ramrod_params *p) { /* update the sc MAC filter structure */ uint32_t mask = (1 << p->cl_id); struct tstorm_eth_mac_filter_config *mac_filters = (struct tstorm_eth_mac_filter_config *)p->rdata; /* initial setting is drop-all */ uint8_t drop_all_ucast = 1, drop_all_mcast = 1; uint8_t accp_all_ucast = 0, accp_all_bcast = 0, accp_all_mcast = 0; uint8_t unmatched_unicast = 0; /* In e1x there we only take into account rx accept flag since tx switching * isn't enabled. */ if (ECORE_TEST_BIT(ECORE_ACCEPT_UNICAST, &p->rx_accept_flags)) /* accept matched ucast */ drop_all_ucast = 0; if (ECORE_TEST_BIT(ECORE_ACCEPT_MULTICAST, &p->rx_accept_flags)) /* accept matched mcast */ drop_all_mcast = 0; if (ECORE_TEST_BIT(ECORE_ACCEPT_ALL_UNICAST, &p->rx_accept_flags)) { /* accept all mcast */ drop_all_ucast = 0; accp_all_ucast = 1; } if (ECORE_TEST_BIT(ECORE_ACCEPT_ALL_MULTICAST, &p->rx_accept_flags)) { /* accept all mcast */ drop_all_mcast = 0; accp_all_mcast = 1; } if (ECORE_TEST_BIT(ECORE_ACCEPT_BROADCAST, &p->rx_accept_flags)) /* accept (all) bcast */ accp_all_bcast = 1; if (ECORE_TEST_BIT(ECORE_ACCEPT_UNMATCHED, &p->rx_accept_flags)) /* accept unmatched unicasts */ unmatched_unicast = 1; mac_filters->ucast_drop_all = drop_all_ucast ? mac_filters->ucast_drop_all | mask : mac_filters->ucast_drop_all & ~mask; mac_filters->mcast_drop_all = drop_all_mcast ? mac_filters->mcast_drop_all | mask : mac_filters->mcast_drop_all & ~mask; mac_filters->ucast_accept_all = accp_all_ucast ? mac_filters->ucast_accept_all | mask : mac_filters->ucast_accept_all & ~mask; mac_filters->mcast_accept_all = accp_all_mcast ? mac_filters->mcast_accept_all | mask : mac_filters->mcast_accept_all & ~mask; mac_filters->bcast_accept_all = accp_all_bcast ? mac_filters->bcast_accept_all | mask : mac_filters->bcast_accept_all & ~mask; mac_filters->unmatched_unicast = unmatched_unicast ? mac_filters->unmatched_unicast | mask : mac_filters->unmatched_unicast & ~mask; ECORE_MSG(sc, "drop_ucast 0x%xdrop_mcast 0x%x accp_ucast 0x%x" "accp_mcast 0x%xaccp_bcast 0x%x", mac_filters->ucast_drop_all, mac_filters->mcast_drop_all, mac_filters->ucast_accept_all, mac_filters->mcast_accept_all, mac_filters->bcast_accept_all); /* write the MAC filter structure */ __storm_memset_mac_filters(sc, mac_filters, p->func_id); /* The operation is completed */ ECORE_CLEAR_BIT(p->state, p->pstate); ECORE_SMP_MB_AFTER_CLEAR_BIT(); return ECORE_SUCCESS; } /* Setup ramrod data */ static void ecore_rx_mode_set_rdata_hdr_e2(uint32_t cid, struct eth_classify_header *hdr, uint8_t rule_cnt) { hdr->echo = ECORE_CPU_TO_LE32(cid); hdr->rule_cnt = rule_cnt; } static void ecore_rx_mode_set_cmd_state_e2(unsigned long *accept_flags, struct eth_filter_rules_cmd *cmd, int clear_accept_all) { uint16_t state; /* start with 'drop-all' */ state = ETH_FILTER_RULES_CMD_UCAST_DROP_ALL | ETH_FILTER_RULES_CMD_MCAST_DROP_ALL; if (ECORE_TEST_BIT(ECORE_ACCEPT_UNICAST, accept_flags)) state &= ~ETH_FILTER_RULES_CMD_UCAST_DROP_ALL; if (ECORE_TEST_BIT(ECORE_ACCEPT_MULTICAST, accept_flags)) state &= ~ETH_FILTER_RULES_CMD_MCAST_DROP_ALL; if (ECORE_TEST_BIT(ECORE_ACCEPT_ALL_UNICAST, accept_flags)) { state &= ~ETH_FILTER_RULES_CMD_UCAST_DROP_ALL; state |= ETH_FILTER_RULES_CMD_UCAST_ACCEPT_ALL; } if (ECORE_TEST_BIT(ECORE_ACCEPT_ALL_MULTICAST, accept_flags)) { state |= ETH_FILTER_RULES_CMD_MCAST_ACCEPT_ALL; state &= ~ETH_FILTER_RULES_CMD_MCAST_DROP_ALL; } if (ECORE_TEST_BIT(ECORE_ACCEPT_BROADCAST, accept_flags)) state |= ETH_FILTER_RULES_CMD_BCAST_ACCEPT_ALL; if (ECORE_TEST_BIT(ECORE_ACCEPT_UNMATCHED, accept_flags)) { state &= ~ETH_FILTER_RULES_CMD_UCAST_DROP_ALL; state |= ETH_FILTER_RULES_CMD_UCAST_ACCEPT_UNMATCHED; } if (ECORE_TEST_BIT(ECORE_ACCEPT_ANY_VLAN, accept_flags)) state |= ETH_FILTER_RULES_CMD_ACCEPT_ANY_VLAN; /* Clear ACCEPT_ALL_XXX flags for FCoE L2 Queue */ if (clear_accept_all) { state &= ~ETH_FILTER_RULES_CMD_MCAST_ACCEPT_ALL; state &= ~ETH_FILTER_RULES_CMD_BCAST_ACCEPT_ALL; state &= ~ETH_FILTER_RULES_CMD_UCAST_ACCEPT_ALL; state &= ~ETH_FILTER_RULES_CMD_UCAST_ACCEPT_UNMATCHED; } cmd->state = ECORE_CPU_TO_LE16(state); } static int ecore_set_rx_mode_e2(struct bnx2x_softc *sc, struct ecore_rx_mode_ramrod_params *p) { struct eth_filter_rules_ramrod_data *data = p->rdata; int rc; uint8_t rule_idx = 0; /* Reset the ramrod data buffer */ ECORE_MEMSET(data, 0, sizeof(*data)); /* Setup ramrod data */ /* Tx (internal switching) */ if (ECORE_TEST_BIT(RAMROD_TX, &p->ramrod_flags)) { data->rules[rule_idx].client_id = p->cl_id; data->rules[rule_idx].func_id = p->func_id; data->rules[rule_idx].cmd_general_data = ETH_FILTER_RULES_CMD_TX_CMD; ecore_rx_mode_set_cmd_state_e2(&p->tx_accept_flags, &(data->rules[rule_idx++]), FALSE); } /* Rx */ if (ECORE_TEST_BIT(RAMROD_RX, &p->ramrod_flags)) { data->rules[rule_idx].client_id = p->cl_id; data->rules[rule_idx].func_id = p->func_id; data->rules[rule_idx].cmd_general_data = ETH_FILTER_RULES_CMD_RX_CMD; ecore_rx_mode_set_cmd_state_e2(&p->rx_accept_flags, &(data->rules[rule_idx++]), FALSE); } /* If FCoE Queue configuration has been requested configure the Rx and * internal switching modes for this queue in separate rules. * * FCoE queue shell never be set to ACCEPT_ALL packets of any sort: * MCAST_ALL, UCAST_ALL, BCAST_ALL and UNMATCHED. */ if (ECORE_TEST_BIT(ECORE_RX_MODE_FCOE_ETH, &p->rx_mode_flags)) { /* Tx (internal switching) */ if (ECORE_TEST_BIT(RAMROD_TX, &p->ramrod_flags)) { data->rules[rule_idx].client_id = ECORE_FCOE_CID(sc); data->rules[rule_idx].func_id = p->func_id; data->rules[rule_idx].cmd_general_data = ETH_FILTER_RULES_CMD_TX_CMD; ecore_rx_mode_set_cmd_state_e2(&p->tx_accept_flags, &(data->rules [rule_idx++]), TRUE); } /* Rx */ if (ECORE_TEST_BIT(RAMROD_RX, &p->ramrod_flags)) { data->rules[rule_idx].client_id = ECORE_FCOE_CID(sc); data->rules[rule_idx].func_id = p->func_id; data->rules[rule_idx].cmd_general_data = ETH_FILTER_RULES_CMD_RX_CMD; ecore_rx_mode_set_cmd_state_e2(&p->rx_accept_flags, &(data->rules [rule_idx++]), TRUE); } } /* Set the ramrod header (most importantly - number of rules to * configure). */ ecore_rx_mode_set_rdata_hdr_e2(p->cid, &data->header, rule_idx); ECORE_MSG (sc, "About to configure %d rules, rx_accept_flags 0x%lx, tx_accept_flags 0x%lx", data->header.rule_cnt, p->rx_accept_flags, p->tx_accept_flags); /* No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ /* Send a ramrod */ rc = ecore_sp_post(sc, RAMROD_CMD_ID_ETH_FILTER_RULES, p->cid, p->rdata_mapping, ETH_CONNECTION_TYPE); if (rc) return rc; /* Ramrod completion is pending */ return ECORE_PENDING; } static int ecore_wait_rx_mode_comp_e2(struct bnx2x_softc *sc, struct ecore_rx_mode_ramrod_params *p) { return ecore_state_wait(sc, p->state, p->pstate); } static int ecore_empty_rx_mode_wait(__rte_unused struct bnx2x_softc *sc, __rte_unused struct ecore_rx_mode_ramrod_params *p) { /* Do nothing */ return ECORE_SUCCESS; } int ecore_config_rx_mode(struct bnx2x_softc *sc, struct ecore_rx_mode_ramrod_params *p) { int rc; /* Configure the new classification in the chip */ if (p->rx_mode_obj->config_rx_mode) { rc = p->rx_mode_obj->config_rx_mode(sc, p); if (rc < 0) return rc; /* Wait for a ramrod completion if was requested */ if (ECORE_TEST_BIT(RAMROD_COMP_WAIT, &p->ramrod_flags)) { rc = p->rx_mode_obj->wait_comp(sc, p); if (rc) return rc; } } else { ECORE_MSG(sc, "ERROR: config_rx_mode is NULL"); return -1; } return rc; } void ecore_init_rx_mode_obj(struct bnx2x_softc *sc, struct ecore_rx_mode_obj *o) { if (CHIP_IS_E1x(sc)) { o->wait_comp = ecore_empty_rx_mode_wait; o->config_rx_mode = ecore_set_rx_mode_e1x; } else { o->wait_comp = ecore_wait_rx_mode_comp_e2; o->config_rx_mode = ecore_set_rx_mode_e2; } } /********************* Multicast verbs: SET, CLEAR ****************************/ static uint8_t ecore_mcast_bin_from_mac(uint8_t * mac) { return (ECORE_CRC32_LE(0, mac, ETH_ALEN) >> 24) & 0xff; } struct ecore_mcast_mac_elem { ecore_list_entry_t link; uint8_t mac[ETH_ALEN]; uint8_t pad[2]; /* For a natural alignment of the following buffer */ }; struct ecore_pending_mcast_cmd { ecore_list_entry_t link; int type; /* ECORE_MCAST_CMD_X */ union { ecore_list_t macs_head; uint32_t macs_num; /* Needed for DEL command */ int next_bin; /* Needed for RESTORE flow with aprox match */ } data; int done; /* set to TRUE, when the command has been handled, * practically used in 57712 handling only, where one pending * command may be handled in a few operations. As long as for * other chips every operation handling is completed in a * single ramrod, there is no need to utilize this field. */ }; static int ecore_mcast_wait(struct bnx2x_softc *sc, struct ecore_mcast_obj *o) { if (ecore_state_wait(sc, o->sched_state, o->raw.pstate) || o->raw.wait_comp(sc, &o->raw)) return ECORE_TIMEOUT; return ECORE_SUCCESS; } static int ecore_mcast_enqueue_cmd(struct bnx2x_softc *sc __rte_unused, struct ecore_mcast_obj *o, struct ecore_mcast_ramrod_params *p, enum ecore_mcast_cmd cmd) { int total_sz; struct ecore_pending_mcast_cmd *new_cmd; struct ecore_mcast_mac_elem *cur_mac = NULL; struct ecore_mcast_list_elem *pos; int macs_list_len = ((cmd == ECORE_MCAST_CMD_ADD) ? p->mcast_list_len : 0); /* If the command is empty ("handle pending commands only"), break */ if (!p->mcast_list_len) return ECORE_SUCCESS; total_sz = sizeof(*new_cmd) + macs_list_len * sizeof(struct ecore_mcast_mac_elem); /* Add mcast is called under spin_lock, thus calling with GFP_ATOMIC */ new_cmd = ECORE_ZALLOC(total_sz, GFP_ATOMIC, sc); if (!new_cmd) return ECORE_NOMEM; ECORE_MSG(sc, "About to enqueue a new %d command. macs_list_len=%d", cmd, macs_list_len); ECORE_LIST_INIT(&new_cmd->data.macs_head); new_cmd->type = cmd; new_cmd->done = FALSE; switch (cmd) { case ECORE_MCAST_CMD_ADD: cur_mac = (struct ecore_mcast_mac_elem *) ((uint8_t *) new_cmd + sizeof(*new_cmd)); /* Push the MACs of the current command into the pending command * MACs list: FIFO */ ECORE_LIST_FOR_EACH_ENTRY(pos, &p->mcast_list, link, struct ecore_mcast_list_elem) { ECORE_MEMCPY(cur_mac->mac, pos->mac, ETH_ALEN); ECORE_LIST_PUSH_TAIL(&cur_mac->link, &new_cmd->data.macs_head); cur_mac++; } break; case ECORE_MCAST_CMD_DEL: new_cmd->data.macs_num = p->mcast_list_len; break; case ECORE_MCAST_CMD_RESTORE: new_cmd->data.next_bin = 0; break; default: ECORE_FREE(sc, new_cmd, total_sz); PMD_DRV_LOG(ERR, sc, "Unknown command: %d", cmd); return ECORE_INVAL; } /* Push the new pending command to the tail of the pending list: FIFO */ ECORE_LIST_PUSH_TAIL(&new_cmd->link, &o->pending_cmds_head); o->set_sched(o); return ECORE_PENDING; } /** * ecore_mcast_get_next_bin - get the next set bin (index) * * @o: * @last: index to start looking from (including) * * returns the next found (set) bin or a negative value if none is found. */ static int ecore_mcast_get_next_bin(struct ecore_mcast_obj *o, int last) { int i, j, inner_start = last % BIT_VEC64_ELEM_SZ; for (i = last / BIT_VEC64_ELEM_SZ; i < ECORE_MCAST_VEC_SZ; i++) { if (o->registry.aprox_match.vec[i]) for (j = inner_start; j < BIT_VEC64_ELEM_SZ; j++) { int cur_bit = j + BIT_VEC64_ELEM_SZ * i; if (BIT_VEC64_TEST_BIT (o->registry.aprox_match.vec, cur_bit)) { return cur_bit; } } inner_start = 0; } /* None found */ return -1; } /** * ecore_mcast_clear_first_bin - find the first set bin and clear it * * @o: * * returns the index of the found bin or -1 if none is found */ static int ecore_mcast_clear_first_bin(struct ecore_mcast_obj *o) { int cur_bit = ecore_mcast_get_next_bin(o, 0); if (cur_bit >= 0) BIT_VEC64_CLEAR_BIT(o->registry.aprox_match.vec, cur_bit); return cur_bit; } static uint8_t ecore_mcast_get_rx_tx_flag(struct ecore_mcast_obj *o) { struct ecore_raw_obj *raw = &o->raw; uint8_t rx_tx_flag = 0; if ((raw->obj_type == ECORE_OBJ_TYPE_TX) || (raw->obj_type == ECORE_OBJ_TYPE_RX_TX)) rx_tx_flag |= ETH_MULTICAST_RULES_CMD_TX_CMD; if ((raw->obj_type == ECORE_OBJ_TYPE_RX) || (raw->obj_type == ECORE_OBJ_TYPE_RX_TX)) rx_tx_flag |= ETH_MULTICAST_RULES_CMD_RX_CMD; return rx_tx_flag; } static void ecore_mcast_set_one_rule_e2(struct bnx2x_softc *sc __rte_unused, struct ecore_mcast_obj *o, int idx, union ecore_mcast_config_data *cfg_data, enum ecore_mcast_cmd cmd) { struct ecore_raw_obj *r = &o->raw; struct eth_multicast_rules_ramrod_data *data = (struct eth_multicast_rules_ramrod_data *)(r->rdata); uint8_t func_id = r->func_id; uint8_t rx_tx_add_flag = ecore_mcast_get_rx_tx_flag(o); int bin; if ((cmd == ECORE_MCAST_CMD_ADD) || (cmd == ECORE_MCAST_CMD_RESTORE)) rx_tx_add_flag |= ETH_MULTICAST_RULES_CMD_IS_ADD; data->rules[idx].cmd_general_data |= rx_tx_add_flag; /* Get a bin and update a bins' vector */ switch (cmd) { case ECORE_MCAST_CMD_ADD: bin = ecore_mcast_bin_from_mac(cfg_data->mac); BIT_VEC64_SET_BIT(o->registry.aprox_match.vec, bin); break; case ECORE_MCAST_CMD_DEL: /* If there were no more bins to clear * (ecore_mcast_clear_first_bin() returns -1) then we would * clear any (0xff) bin. * See ecore_mcast_validate_e2() for explanation when it may * happen. */ bin = ecore_mcast_clear_first_bin(o); break; case ECORE_MCAST_CMD_RESTORE: bin = cfg_data->bin; break; default: PMD_DRV_LOG(ERR, sc, "Unknown command: %d", cmd); return; } ECORE_MSG(sc, "%s bin %d", ((rx_tx_add_flag & ETH_MULTICAST_RULES_CMD_IS_ADD) ? "Setting" : "Clearing"), bin); data->rules[idx].bin_id = (uint8_t) bin; data->rules[idx].func_id = func_id; data->rules[idx].engine_id = o->engine_id; } /** * ecore_mcast_handle_restore_cmd_e2 - restore configuration from the registry * * @sc: device handle * @o: * @start_bin: index in the registry to start from (including) * @rdata_idx: index in the ramrod data to start from * * returns last handled bin index or -1 if all bins have been handled */ static int ecore_mcast_handle_restore_cmd_e2(struct bnx2x_softc *sc, struct ecore_mcast_obj *o, int start_bin, int *rdata_idx) { int cur_bin, cnt = *rdata_idx; union ecore_mcast_config_data cfg_data = { NULL }; /* go through the registry and configure the bins from it */ for (cur_bin = ecore_mcast_get_next_bin(o, start_bin); cur_bin >= 0; cur_bin = ecore_mcast_get_next_bin(o, cur_bin + 1)) { cfg_data.bin = (uint8_t) cur_bin; o->set_one_rule(sc, o, cnt, &cfg_data, ECORE_MCAST_CMD_RESTORE); cnt++; ECORE_MSG(sc, "About to configure a bin %d", cur_bin); /* Break if we reached the maximum number * of rules. */ if (cnt >= o->max_cmd_len) break; } *rdata_idx = cnt; return cur_bin; } static void ecore_mcast_hdl_pending_add_e2(struct bnx2x_softc *sc, struct ecore_mcast_obj *o, struct ecore_pending_mcast_cmd *cmd_pos, int *line_idx) { struct ecore_mcast_mac_elem *pmac_pos, *pmac_pos_n; int cnt = *line_idx; union ecore_mcast_config_data cfg_data = { NULL }; ECORE_LIST_FOR_EACH_ENTRY_SAFE(pmac_pos, pmac_pos_n, &cmd_pos->data.macs_head, link, struct ecore_mcast_mac_elem) { cfg_data.mac = &pmac_pos->mac[0]; o->set_one_rule(sc, o, cnt, &cfg_data, cmd_pos->type); cnt++; ECORE_MSG (sc, "About to configure %02x:%02x:%02x:%02x:%02x:%02x mcast MAC", pmac_pos->mac[0], pmac_pos->mac[1], pmac_pos->mac[2], pmac_pos->mac[3], pmac_pos->mac[4], pmac_pos->mac[5]); ECORE_LIST_REMOVE_ENTRY(&pmac_pos->link, &cmd_pos->data.macs_head); /* Break if we reached the maximum number * of rules. */ if (cnt >= o->max_cmd_len) break; } *line_idx = cnt; /* if no more MACs to configure - we are done */ if (ECORE_LIST_IS_EMPTY(&cmd_pos->data.macs_head)) cmd_pos->done = TRUE; } static void ecore_mcast_hdl_pending_del_e2(struct bnx2x_softc *sc, struct ecore_mcast_obj *o, struct ecore_pending_mcast_cmd *cmd_pos, int *line_idx) { int cnt = *line_idx; while (cmd_pos->data.macs_num) { o->set_one_rule(sc, o, cnt, NULL, cmd_pos->type); cnt++; cmd_pos->data.macs_num--; ECORE_MSG(sc, "Deleting MAC. %d left,cnt is %d", cmd_pos->data.macs_num, cnt); /* Break if we reached the maximum * number of rules. */ if (cnt >= o->max_cmd_len) break; } *line_idx = cnt; /* If we cleared all bins - we are done */ if (!cmd_pos->data.macs_num) cmd_pos->done = TRUE; } static void ecore_mcast_hdl_pending_restore_e2(struct bnx2x_softc *sc, struct ecore_mcast_obj *o, struct ecore_pending_mcast_cmd *cmd_pos, int *line_idx) { cmd_pos->data.next_bin = o->hdl_restore(sc, o, cmd_pos->data.next_bin, line_idx); if (cmd_pos->data.next_bin < 0) /* If o->set_restore returned -1 we are done */ cmd_pos->done = TRUE; else /* Start from the next bin next time */ cmd_pos->data.next_bin++; } static int ecore_mcast_handle_pending_cmds_e2(struct bnx2x_softc *sc, struct ecore_mcast_ramrod_params *p) { struct ecore_pending_mcast_cmd *cmd_pos, *cmd_pos_n; int cnt = 0; struct ecore_mcast_obj *o = p->mcast_obj; ECORE_LIST_FOR_EACH_ENTRY_SAFE(cmd_pos, cmd_pos_n, &o->pending_cmds_head, link, struct ecore_pending_mcast_cmd) { switch (cmd_pos->type) { case ECORE_MCAST_CMD_ADD: ecore_mcast_hdl_pending_add_e2(sc, o, cmd_pos, &cnt); break; case ECORE_MCAST_CMD_DEL: ecore_mcast_hdl_pending_del_e2(sc, o, cmd_pos, &cnt); break; case ECORE_MCAST_CMD_RESTORE: ecore_mcast_hdl_pending_restore_e2(sc, o, cmd_pos, &cnt); break; default: PMD_DRV_LOG(ERR, sc, "Unknown command: %d", cmd_pos->type); return ECORE_INVAL; } /* If the command has been completed - remove it from the list * and free the memory */ if (cmd_pos->done) { ECORE_LIST_REMOVE_ENTRY(&cmd_pos->link, &o->pending_cmds_head); ECORE_FREE(sc, cmd_pos, cmd_pos->alloc_len); } /* Break if we reached the maximum number of rules */ if (cnt >= o->max_cmd_len) break; } return cnt; } static void ecore_mcast_hdl_add(struct bnx2x_softc *sc, struct ecore_mcast_obj *o, struct ecore_mcast_ramrod_params *p, int *line_idx) { struct ecore_mcast_list_elem *mlist_pos; union ecore_mcast_config_data cfg_data = { NULL }; int cnt = *line_idx; ECORE_LIST_FOR_EACH_ENTRY(mlist_pos, &p->mcast_list, link, struct ecore_mcast_list_elem) { cfg_data.mac = mlist_pos->mac; o->set_one_rule(sc, o, cnt, &cfg_data, ECORE_MCAST_CMD_ADD); cnt++; ECORE_MSG (sc, "About to configure %02x:%02x:%02x:%02x:%02x:%02x mcast MAC", mlist_pos->mac[0], mlist_pos->mac[1], mlist_pos->mac[2], mlist_pos->mac[3], mlist_pos->mac[4], mlist_pos->mac[5]); } *line_idx = cnt; } static void ecore_mcast_hdl_del(struct bnx2x_softc *sc, struct ecore_mcast_obj *o, struct ecore_mcast_ramrod_params *p, int *line_idx) { int cnt = *line_idx, i; for (i = 0; i < p->mcast_list_len; i++) { o->set_one_rule(sc, o, cnt, NULL, ECORE_MCAST_CMD_DEL); cnt++; ECORE_MSG(sc, "Deleting MAC. %d left", p->mcast_list_len - i - 1); } *line_idx = cnt; } /** * ecore_mcast_handle_current_cmd - * * @sc: device handle * @p: * @cmd: * @start_cnt: first line in the ramrod data that may be used * * This function is called if there is enough place for the current command in * the ramrod data. * Returns number of lines filled in the ramrod data in total. */ static int ecore_mcast_handle_current_cmd(struct bnx2x_softc *sc, struct ecore_mcast_ramrod_params *p, enum ecore_mcast_cmd cmd, int start_cnt) { struct ecore_mcast_obj *o = p->mcast_obj; int cnt = start_cnt; ECORE_MSG(sc, "p->mcast_list_len=%d", p->mcast_list_len); switch (cmd) { case ECORE_MCAST_CMD_ADD: ecore_mcast_hdl_add(sc, o, p, &cnt); break; case ECORE_MCAST_CMD_DEL: ecore_mcast_hdl_del(sc, o, p, &cnt); break; case ECORE_MCAST_CMD_RESTORE: o->hdl_restore(sc, o, 0, &cnt); break; default: PMD_DRV_LOG(ERR, sc, "Unknown command: %d", cmd); return ECORE_INVAL; } /* The current command has been handled */ p->mcast_list_len = 0; return cnt; } static int ecore_mcast_validate_e2(__rte_unused struct bnx2x_softc *sc, struct ecore_mcast_ramrod_params *p, enum ecore_mcast_cmd cmd) { struct ecore_mcast_obj *o = p->mcast_obj; int reg_sz = o->get_registry_size(o); switch (cmd) { /* DEL command deletes all currently configured MACs */ case ECORE_MCAST_CMD_DEL: o->set_registry_size(o, 0); /* fall-through */ /* RESTORE command will restore the entire multicast configuration */ case ECORE_MCAST_CMD_RESTORE: /* Here we set the approximate amount of work to do, which in * fact may be only less as some MACs in postponed ADD * command(s) scheduled before this command may fall into * the same bin and the actual number of bins set in the * registry would be less than we estimated here. See * ecore_mcast_set_one_rule_e2() for further details. */ p->mcast_list_len = reg_sz; break; case ECORE_MCAST_CMD_ADD: case ECORE_MCAST_CMD_CONT: /* Here we assume that all new MACs will fall into new bins. * However we will correct the real registry size after we * handle all pending commands. */ o->set_registry_size(o, reg_sz + p->mcast_list_len); break; default: PMD_DRV_LOG(ERR, sc, "Unknown command: %d", cmd); return ECORE_INVAL; } /* Increase the total number of MACs pending to be configured */ o->total_pending_num += p->mcast_list_len; return ECORE_SUCCESS; } static void ecore_mcast_revert_e2(__rte_unused struct bnx2x_softc *sc, struct ecore_mcast_ramrod_params *p, int old_num_bins, enum ecore_mcast_cmd cmd) { struct ecore_mcast_obj *o = p->mcast_obj; o->set_registry_size(o, old_num_bins); o->total_pending_num -= p->mcast_list_len; if (cmd == ECORE_MCAST_CMD_SET) o->total_pending_num -= o->max_cmd_len; } /** * ecore_mcast_set_rdata_hdr_e2 - sets a header values * * @sc: device handle * @p: * @len: number of rules to handle */ static void ecore_mcast_set_rdata_hdr_e2(__rte_unused struct bnx2x_softc *sc, struct ecore_mcast_ramrod_params *p, uint8_t len) { struct ecore_raw_obj *r = &p->mcast_obj->raw; struct eth_multicast_rules_ramrod_data *data = (struct eth_multicast_rules_ramrod_data *)(r->rdata); data->header.echo = ECORE_CPU_TO_LE32((r->cid & ECORE_SWCID_MASK) | (ECORE_FILTER_MCAST_PENDING << ECORE_SWCID_SHIFT)); data->header.rule_cnt = len; } /** * ecore_mcast_refresh_registry_e2 - recalculate the actual number of set bins * * @sc: device handle * @o: * * Recalculate the actual number of set bins in the registry using Brian * Kernighan's algorithm: it's execution complexity is as a number of set bins. */ static int ecore_mcast_refresh_registry_e2(struct ecore_mcast_obj *o) { int i, cnt = 0; uint64_t elem; for (i = 0; i < ECORE_MCAST_VEC_SZ; i++) { elem = o->registry.aprox_match.vec[i]; for (; elem; cnt++) elem &= elem - 1; } o->set_registry_size(o, cnt); return ECORE_SUCCESS; } static int ecore_mcast_setup_e2(struct bnx2x_softc *sc, struct ecore_mcast_ramrod_params *p, enum ecore_mcast_cmd cmd) { struct ecore_raw_obj *raw = &p->mcast_obj->raw; struct ecore_mcast_obj *o = p->mcast_obj; struct eth_multicast_rules_ramrod_data *data = (struct eth_multicast_rules_ramrod_data *)(raw->rdata); int cnt = 0, rc; /* Reset the ramrod data buffer */ ECORE_MEMSET(data, 0, sizeof(*data)); cnt = ecore_mcast_handle_pending_cmds_e2(sc, p); /* If there are no more pending commands - clear SCHEDULED state */ if (ECORE_LIST_IS_EMPTY(&o->pending_cmds_head)) o->clear_sched(o); /* The below may be TRUE if there was enough room in ramrod * data for all pending commands and for the current * command. Otherwise the current command would have been added * to the pending commands and p->mcast_list_len would have been * zeroed. */ if (p->mcast_list_len > 0) cnt = ecore_mcast_handle_current_cmd(sc, p, cmd, cnt); /* We've pulled out some MACs - update the total number of * outstanding. */ o->total_pending_num -= cnt; /* send a ramrod */ ECORE_DBG_BREAK_IF(o->total_pending_num < 0); ECORE_DBG_BREAK_IF(cnt > o->max_cmd_len); ecore_mcast_set_rdata_hdr_e2(sc, p, (uint8_t) cnt); /* Update a registry size if there are no more pending operations. * * We don't want to change the value of the registry size if there are * pending operations because we want it to always be equal to the * exact or the approximate number (see ecore_mcast_validate_e2()) of * set bins after the last requested operation in order to properly * evaluate the size of the next DEL/RESTORE operation. * * Note that we update the registry itself during command(s) handling * - see ecore_mcast_set_one_rule_e2(). That's because for 57712 we * aggregate multiple commands (ADD/DEL/RESTORE) into one ramrod but * with a limited amount of update commands (per MAC/bin) and we don't * know in this scope what the actual state of bins configuration is * going to be after this ramrod. */ if (!o->total_pending_num) ecore_mcast_refresh_registry_e2(o); /* If CLEAR_ONLY was requested - don't send a ramrod and clear * RAMROD_PENDING status immediately. */ if (ECORE_TEST_BIT(RAMROD_DRV_CLR_ONLY, &p->ramrod_flags)) { raw->clear_pending(raw); return ECORE_SUCCESS; } else { /* No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ /* Send a ramrod */ rc = ecore_sp_post(sc, RAMROD_CMD_ID_ETH_MULTICAST_RULES, raw->cid, raw->rdata_mapping, ETH_CONNECTION_TYPE); if (rc) return rc; /* Ramrod completion is pending */ return ECORE_PENDING; } } static int ecore_mcast_validate_e1h(__rte_unused struct bnx2x_softc *sc, struct ecore_mcast_ramrod_params *p, enum ecore_mcast_cmd cmd) { /* Mark, that there is a work to do */ if ((cmd == ECORE_MCAST_CMD_DEL) || (cmd == ECORE_MCAST_CMD_RESTORE)) p->mcast_list_len = 1; return ECORE_SUCCESS; } static void ecore_mcast_revert_e1h(__rte_unused struct bnx2x_softc *sc, __rte_unused struct ecore_mcast_ramrod_params *p, __rte_unused int old_num_bins, __rte_unused enum ecore_mcast_cmd cmd) { /* Do nothing */ } #define ECORE_57711_SET_MC_FILTER(filter, bit) \ do { \ (filter)[(bit) >> 5] |= (1 << ((bit) & 0x1f)); \ } while (0) static void ecore_mcast_hdl_add_e1h(struct bnx2x_softc *sc __rte_unused, struct ecore_mcast_obj *o, struct ecore_mcast_ramrod_params *p, uint32_t * mc_filter) { struct ecore_mcast_list_elem *mlist_pos; int bit; ECORE_LIST_FOR_EACH_ENTRY(mlist_pos, &p->mcast_list, link, struct ecore_mcast_list_elem) { bit = ecore_mcast_bin_from_mac(mlist_pos->mac); ECORE_57711_SET_MC_FILTER(mc_filter, bit); ECORE_MSG (sc, "About to configure %02x:%02x:%02x:%02x:%02x:%02x mcast MAC, bin %d", mlist_pos->mac[0], mlist_pos->mac[1], mlist_pos->mac[2], mlist_pos->mac[3], mlist_pos->mac[4], mlist_pos->mac[5], bit); /* bookkeeping... */ BIT_VEC64_SET_BIT(o->registry.aprox_match.vec, bit); } } static void ecore_mcast_hdl_restore_e1h(struct bnx2x_softc *sc __rte_unused, struct ecore_mcast_obj *o, uint32_t * mc_filter) { int bit; for (bit = ecore_mcast_get_next_bin(o, 0); bit >= 0; bit = ecore_mcast_get_next_bin(o, bit + 1)) { ECORE_57711_SET_MC_FILTER(mc_filter, bit); ECORE_MSG(sc, "About to set bin %d", bit); } } /* On 57711 we write the multicast MACs' approximate match * table by directly into the TSTORM's internal RAM. So we don't * really need to handle any tricks to make it work. */ static int ecore_mcast_setup_e1h(struct bnx2x_softc *sc, struct ecore_mcast_ramrod_params *p, enum ecore_mcast_cmd cmd) { int i; struct ecore_mcast_obj *o = p->mcast_obj; struct ecore_raw_obj *r = &o->raw; /* If CLEAR_ONLY has been requested - clear the registry * and clear a pending bit. */ if (!ECORE_TEST_BIT(RAMROD_DRV_CLR_ONLY, &p->ramrod_flags)) { uint32_t mc_filter[ECORE_MC_HASH_SIZE] = { 0 }; /* Set the multicast filter bits before writing it into * the internal memory. */ switch (cmd) { case ECORE_MCAST_CMD_ADD: ecore_mcast_hdl_add_e1h(sc, o, p, mc_filter); break; case ECORE_MCAST_CMD_DEL: ECORE_MSG(sc, "Invalidating multicast MACs configuration"); /* clear the registry */ ECORE_MEMSET(o->registry.aprox_match.vec, 0, sizeof(o->registry.aprox_match.vec)); break; case ECORE_MCAST_CMD_RESTORE: ecore_mcast_hdl_restore_e1h(sc, o, mc_filter); break; default: PMD_DRV_LOG(ERR, sc, "Unknown command: %d", cmd); return ECORE_INVAL; } /* Set the mcast filter in the internal memory */ for (i = 0; i < ECORE_MC_HASH_SIZE; i++) REG_WR(sc, ECORE_MC_HASH_OFFSET(sc, i), mc_filter[i]); } else /* clear the registry */ ECORE_MEMSET(o->registry.aprox_match.vec, 0, sizeof(o->registry.aprox_match.vec)); /* We are done */ r->clear_pending(r); return ECORE_SUCCESS; } static int ecore_mcast_get_registry_size_aprox(struct ecore_mcast_obj *o) { return o->registry.aprox_match.num_bins_set; } static void ecore_mcast_set_registry_size_aprox(struct ecore_mcast_obj *o, int n) { o->registry.aprox_match.num_bins_set = n; } int ecore_config_mcast(struct bnx2x_softc *sc, struct ecore_mcast_ramrod_params *p, enum ecore_mcast_cmd cmd) { struct ecore_mcast_obj *o = p->mcast_obj; struct ecore_raw_obj *r = &o->raw; int rc = 0, old_reg_size; /* This is needed to recover number of currently configured mcast macs * in case of failure. */ old_reg_size = o->get_registry_size(o); /* Do some calculations and checks */ rc = o->validate(sc, p, cmd); if (rc) return rc; /* Return if there is no work to do */ if ((!p->mcast_list_len) && (!o->check_sched(o))) return ECORE_SUCCESS; ECORE_MSG (sc, "o->total_pending_num=%d p->mcast_list_len=%d o->max_cmd_len=%d", o->total_pending_num, p->mcast_list_len, o->max_cmd_len); /* Enqueue the current command to the pending list if we can't complete * it in the current iteration */ if (r->check_pending(r) || ((o->max_cmd_len > 0) && (o->total_pending_num > o->max_cmd_len))) { rc = o->enqueue_cmd(sc, p->mcast_obj, p, cmd); if (rc < 0) goto error_exit1; /* As long as the current command is in a command list we * don't need to handle it separately. */ p->mcast_list_len = 0; } if (!r->check_pending(r)) { /* Set 'pending' state */ r->set_pending(r); /* Configure the new classification in the chip */ rc = o->config_mcast(sc, p, cmd); if (rc < 0) goto error_exit2; /* Wait for a ramrod completion if was requested */ if (ECORE_TEST_BIT(RAMROD_COMP_WAIT, &p->ramrod_flags)) rc = o->wait_comp(sc, o); } return rc; error_exit2: r->clear_pending(r); error_exit1: o->revert(sc, p, old_reg_size, cmd); return rc; } static void ecore_mcast_clear_sched(struct ecore_mcast_obj *o) { ECORE_SMP_MB_BEFORE_CLEAR_BIT(); ECORE_CLEAR_BIT(o->sched_state, o->raw.pstate); ECORE_SMP_MB_AFTER_CLEAR_BIT(); } static void ecore_mcast_set_sched(struct ecore_mcast_obj *o) { ECORE_SMP_MB_BEFORE_CLEAR_BIT(); ECORE_SET_BIT(o->sched_state, o->raw.pstate); ECORE_SMP_MB_AFTER_CLEAR_BIT(); } static bool ecore_mcast_check_sched(struct ecore_mcast_obj *o) { return ! !ECORE_TEST_BIT(o->sched_state, o->raw.pstate); } static bool ecore_mcast_check_pending(struct ecore_mcast_obj *o) { return o->raw.check_pending(&o->raw) || o->check_sched(o); } void ecore_init_mcast_obj(struct bnx2x_softc *sc, struct ecore_mcast_obj *mcast_obj, uint8_t mcast_cl_id, uint32_t mcast_cid, uint8_t func_id, uint8_t engine_id, void *rdata, ecore_dma_addr_t rdata_mapping, int state, unsigned long *pstate, ecore_obj_type type) { ECORE_MEMSET(mcast_obj, 0, sizeof(*mcast_obj)); ecore_init_raw_obj(&mcast_obj->raw, mcast_cl_id, mcast_cid, func_id, rdata, rdata_mapping, state, pstate, type); mcast_obj->engine_id = engine_id; ECORE_LIST_INIT(&mcast_obj->pending_cmds_head); mcast_obj->sched_state = ECORE_FILTER_MCAST_SCHED; mcast_obj->check_sched = ecore_mcast_check_sched; mcast_obj->set_sched = ecore_mcast_set_sched; mcast_obj->clear_sched = ecore_mcast_clear_sched; if (CHIP_IS_E1H(sc)) { mcast_obj->config_mcast = ecore_mcast_setup_e1h; mcast_obj->enqueue_cmd = NULL; mcast_obj->hdl_restore = NULL; mcast_obj->check_pending = ecore_mcast_check_pending; /* 57711 doesn't send a ramrod, so it has unlimited credit * for one command. */ mcast_obj->max_cmd_len = -1; mcast_obj->wait_comp = ecore_mcast_wait; mcast_obj->set_one_rule = NULL; mcast_obj->validate = ecore_mcast_validate_e1h; mcast_obj->revert = ecore_mcast_revert_e1h; mcast_obj->get_registry_size = ecore_mcast_get_registry_size_aprox; mcast_obj->set_registry_size = ecore_mcast_set_registry_size_aprox; } else { mcast_obj->config_mcast = ecore_mcast_setup_e2; mcast_obj->enqueue_cmd = ecore_mcast_enqueue_cmd; mcast_obj->hdl_restore = ecore_mcast_handle_restore_cmd_e2; mcast_obj->check_pending = ecore_mcast_check_pending; mcast_obj->max_cmd_len = 16; mcast_obj->wait_comp = ecore_mcast_wait; mcast_obj->set_one_rule = ecore_mcast_set_one_rule_e2; mcast_obj->validate = ecore_mcast_validate_e2; mcast_obj->revert = ecore_mcast_revert_e2; mcast_obj->get_registry_size = ecore_mcast_get_registry_size_aprox; mcast_obj->set_registry_size = ecore_mcast_set_registry_size_aprox; } } /*************************** Credit handling **********************************/ /** * atomic_add_ifless - add if the result is less than a given value. * * @v: pointer of type ecore_atomic_t * @a: the amount to add to v... * @u: ...if (v + a) is less than u. * * returns TRUE if (v + a) was less than u, and FALSE otherwise. * */ static bool __atomic_add_ifless(ecore_atomic_t *v, int a, int u) { int c, old; c = ECORE_ATOMIC_READ(v); for (;;) { if (ECORE_UNLIKELY(c + a >= u)) return FALSE; old = ECORE_ATOMIC_CMPXCHG((v), c, c + a); if (ECORE_LIKELY(old == c)) break; c = old; } return TRUE; } /** * atomic_dec_ifmoe - dec if the result is more or equal than a given value. * * @v: pointer of type ecore_atomic_t * @a: the amount to dec from v... * @u: ...if (v - a) is more or equal than u. * * returns TRUE if (v - a) was more or equal than u, and FALSE * otherwise. */ static bool __atomic_dec_ifmoe(ecore_atomic_t *v, int a, int u) { int c, old; c = ECORE_ATOMIC_READ(v); for (;;) { if (ECORE_UNLIKELY(c - a < u)) return FALSE; old = ECORE_ATOMIC_CMPXCHG((v), c, c - a); if (ECORE_LIKELY(old == c)) break; c = old; } return TRUE; } static bool ecore_credit_pool_get(struct ecore_credit_pool_obj *o, int cnt) { bool rc; ECORE_SMP_MB(); rc = __atomic_dec_ifmoe(&o->credit, cnt, 0); ECORE_SMP_MB(); return rc; } static bool ecore_credit_pool_put(struct ecore_credit_pool_obj *o, int cnt) { bool rc; ECORE_SMP_MB(); /* Don't let to refill if credit + cnt > pool_sz */ rc = __atomic_add_ifless(&o->credit, cnt, o->pool_sz + 1); ECORE_SMP_MB(); return rc; } static int ecore_credit_pool_check(struct ecore_credit_pool_obj *o) { int cur_credit; ECORE_SMP_MB(); cur_credit = ECORE_ATOMIC_READ(&o->credit); return cur_credit; } static bool ecore_credit_pool_always_TRUE(__rte_unused struct ecore_credit_pool_obj *o, __rte_unused int cnt) { return TRUE; } static bool ecore_credit_pool_get_entry(struct ecore_credit_pool_obj *o, int *offset) { int idx, vec, i; *offset = -1; /* Find "internal cam-offset" then add to base for this object... */ for (vec = 0; vec < ECORE_POOL_VEC_SIZE; vec++) { /* Skip the current vector if there are no free entries in it */ if (!o->pool_mirror[vec]) continue; /* If we've got here we are going to find a free entry */ for (idx = vec * BIT_VEC64_ELEM_SZ, i = 0; i < BIT_VEC64_ELEM_SZ; idx++, i++) if (BIT_VEC64_TEST_BIT(o->pool_mirror, idx)) { /* Got one!! */ BIT_VEC64_CLEAR_BIT(o->pool_mirror, idx); *offset = o->base_pool_offset + idx; return TRUE; } } return FALSE; } static bool ecore_credit_pool_put_entry(struct ecore_credit_pool_obj *o, int offset) { if (offset < o->base_pool_offset) return FALSE; offset -= o->base_pool_offset; if (offset >= o->pool_sz) return FALSE; /* Return the entry to the pool */ BIT_VEC64_SET_BIT(o->pool_mirror, offset); return TRUE; } static bool ecore_credit_pool_put_entry_always_TRUE(__rte_unused struct ecore_credit_pool_obj *o, __rte_unused int offset) { return TRUE; } static bool ecore_credit_pool_get_entry_always_TRUE(__rte_unused struct ecore_credit_pool_obj *o, __rte_unused int *offset) { *offset = -1; return TRUE; } /** * ecore_init_credit_pool - initialize credit pool internals. * * @p: * @base: Base entry in the CAM to use. * @credit: pool size. * * If base is negative no CAM entries handling will be performed. * If credit is negative pool operations will always succeed (unlimited pool). * */ void ecore_init_credit_pool(struct ecore_credit_pool_obj *p, int base, int credit) { /* Zero the object first */ ECORE_MEMSET(p, 0, sizeof(*p)); /* Set the table to all 1s */ ECORE_MEMSET(&p->pool_mirror, 0xff, sizeof(p->pool_mirror)); /* Init a pool as full */ ECORE_ATOMIC_SET(&p->credit, credit); /* The total poll size */ p->pool_sz = credit; p->base_pool_offset = base; /* Commit the change */ ECORE_SMP_MB(); p->check = ecore_credit_pool_check; /* if pool credit is negative - disable the checks */ if (credit >= 0) { p->put = ecore_credit_pool_put; p->get = ecore_credit_pool_get; p->put_entry = ecore_credit_pool_put_entry; p->get_entry = ecore_credit_pool_get_entry; } else { p->put = ecore_credit_pool_always_TRUE; p->get = ecore_credit_pool_always_TRUE; p->put_entry = ecore_credit_pool_put_entry_always_TRUE; p->get_entry = ecore_credit_pool_get_entry_always_TRUE; } /* If base is negative - disable entries handling */ if (base < 0) { p->put_entry = ecore_credit_pool_put_entry_always_TRUE; p->get_entry = ecore_credit_pool_get_entry_always_TRUE; } } void ecore_init_mac_credit_pool(struct bnx2x_softc *sc, struct ecore_credit_pool_obj *p, uint8_t func_id, uint8_t func_num) { #define ECORE_CAM_SIZE_EMUL 5 int cam_sz; if (CHIP_IS_E1H(sc)) { /* CAM credit is equally divided between all active functions * on the PORT!. */ if (func_num > 0) { if (!CHIP_REV_IS_SLOW(sc)) cam_sz = (MAX_MAC_CREDIT_E1H / (2 * func_num)); else cam_sz = ECORE_CAM_SIZE_EMUL; ecore_init_credit_pool(p, func_id * cam_sz, cam_sz); } else { /* this should never happen! Block MAC operations. */ ecore_init_credit_pool(p, 0, 0); } } else { /* * CAM credit is equaly divided between all active functions * on the PATH. */ if (func_num > 0) { if (!CHIP_REV_IS_SLOW(sc)) cam_sz = (MAX_MAC_CREDIT_E2 / func_num); else cam_sz = ECORE_CAM_SIZE_EMUL; /* No need for CAM entries handling for 57712 and * newer. */ ecore_init_credit_pool(p, -1, cam_sz); } else { /* this should never happen! Block MAC operations. */ ecore_init_credit_pool(p, 0, 0); } } } void ecore_init_vlan_credit_pool(struct bnx2x_softc *sc, struct ecore_credit_pool_obj *p, uint8_t func_id, uint8_t func_num) { if (CHIP_IS_E1x(sc)) { /* There is no VLAN credit in HW on 57711 only * MAC / MAC-VLAN can be set */ ecore_init_credit_pool(p, 0, -1); } else { /* CAM credit is equally divided between all active functions * on the PATH. */ if (func_num > 0) { int credit = MAX_VLAN_CREDIT_E2 / func_num; ecore_init_credit_pool(p, func_id * credit, credit); } else /* this should never happen! Block VLAN operations. */ ecore_init_credit_pool(p, 0, 0); } } /****************** RSS Configuration ******************/ /** * ecore_setup_rss - configure RSS * * @sc: device handle * @p: rss configuration * * sends on UPDATE ramrod for that matter. */ static int ecore_setup_rss(struct bnx2x_softc *sc, struct ecore_config_rss_params *p) { struct ecore_rss_config_obj *o = p->rss_obj; struct ecore_raw_obj *r = &o->raw; struct eth_rss_update_ramrod_data *data = (struct eth_rss_update_ramrod_data *)(r->rdata); uint8_t rss_mode = 0; int rc; ECORE_MEMSET(data, 0, sizeof(*data)); ECORE_MSG(sc, "Configuring RSS"); /* Set an echo field */ data->echo = ECORE_CPU_TO_LE32((r->cid & ECORE_SWCID_MASK) | (r->state << ECORE_SWCID_SHIFT)); /* RSS mode */ if (ECORE_TEST_BIT(ECORE_RSS_MODE_DISABLED, &p->rss_flags)) rss_mode = ETH_RSS_MODE_DISABLED; else if (ECORE_TEST_BIT(ECORE_RSS_MODE_REGULAR, &p->rss_flags)) rss_mode = ETH_RSS_MODE_REGULAR; data->rss_mode = rss_mode; ECORE_MSG(sc, "rss_mode=%d", rss_mode); /* RSS capabilities */ if (ECORE_TEST_BIT(ECORE_RSS_IPV4, &p->rss_flags)) data->capabilities |= ETH_RSS_UPDATE_RAMROD_DATA_IPV4_CAPABILITY; if (ECORE_TEST_BIT(ECORE_RSS_IPV4_TCP, &p->rss_flags)) data->capabilities |= ETH_RSS_UPDATE_RAMROD_DATA_IPV4_TCP_CAPABILITY; if (ECORE_TEST_BIT(ECORE_RSS_IPV4_UDP, &p->rss_flags)) data->capabilities |= ETH_RSS_UPDATE_RAMROD_DATA_IPV4_UDP_CAPABILITY; if (ECORE_TEST_BIT(ECORE_RSS_IPV6, &p->rss_flags)) data->capabilities |= ETH_RSS_UPDATE_RAMROD_DATA_IPV6_CAPABILITY; if (ECORE_TEST_BIT(ECORE_RSS_IPV6_TCP, &p->rss_flags)) data->capabilities |= ETH_RSS_UPDATE_RAMROD_DATA_IPV6_TCP_CAPABILITY; if (ECORE_TEST_BIT(ECORE_RSS_IPV6_UDP, &p->rss_flags)) data->capabilities |= ETH_RSS_UPDATE_RAMROD_DATA_IPV6_UDP_CAPABILITY; /* Hashing mask */ data->rss_result_mask = p->rss_result_mask; /* RSS engine ID */ data->rss_engine_id = o->engine_id; ECORE_MSG(sc, "rss_engine_id=%d", data->rss_engine_id); /* Indirection table */ ECORE_MEMCPY(data->indirection_table, p->ind_table, T_ETH_INDIRECTION_TABLE_SIZE); /* Remember the last configuration */ ECORE_MEMCPY(o->ind_table, p->ind_table, T_ETH_INDIRECTION_TABLE_SIZE); /* RSS keys */ if (ECORE_TEST_BIT(ECORE_RSS_SET_SRCH, &p->rss_flags)) { ECORE_MEMCPY(&data->rss_key[0], &p->rss_key[0], sizeof(data->rss_key)); data->capabilities |= ETH_RSS_UPDATE_RAMROD_DATA_UPDATE_RSS_KEY; } /* No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ /* Send a ramrod */ rc = ecore_sp_post(sc, RAMROD_CMD_ID_ETH_RSS_UPDATE, r->cid, r->rdata_mapping, ETH_CONNECTION_TYPE); if (rc < 0) return rc; return ECORE_PENDING; } int ecore_config_rss(struct bnx2x_softc *sc, struct ecore_config_rss_params *p) { int rc; struct ecore_rss_config_obj *o = p->rss_obj; struct ecore_raw_obj *r = &o->raw; /* Do nothing if only driver cleanup was requested */ if (ECORE_TEST_BIT(RAMROD_DRV_CLR_ONLY, &p->ramrod_flags)) return ECORE_SUCCESS; r->set_pending(r); rc = o->config_rss(sc, p); if (rc < 0) { r->clear_pending(r); return rc; } if (ECORE_TEST_BIT(RAMROD_COMP_WAIT, &p->ramrod_flags)) rc = r->wait_comp(sc, r); return rc; } void ecore_init_rss_config_obj(struct bnx2x_softc *sc __rte_unused, struct ecore_rss_config_obj *rss_obj, uint8_t cl_id, uint32_t cid, uint8_t func_id, uint8_t engine_id, void *rdata, ecore_dma_addr_t rdata_mapping, int state, unsigned long *pstate, ecore_obj_type type) { ecore_init_raw_obj(&rss_obj->raw, cl_id, cid, func_id, rdata, rdata_mapping, state, pstate, type); rss_obj->engine_id = engine_id; rss_obj->config_rss = ecore_setup_rss; } /********************** Queue state object ***********************************/ /** * ecore_queue_state_change - perform Queue state change transition * * @sc: device handle * @params: parameters to perform the transition * * returns 0 in case of successfully completed transition, negative error * code in case of failure, positive (EBUSY) value if there is a completion * to that is still pending (possible only if RAMROD_COMP_WAIT is * not set in params->ramrod_flags for asynchronous commands). * */ int ecore_queue_state_change(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_sp_obj *o = params->q_obj; int rc, pending_bit; unsigned long *pending = &o->pending; /* Check that the requested transition is legal */ rc = o->check_transition(sc, o, params); if (rc) { PMD_DRV_LOG(ERR, sc, "check transition returned an error. rc %d", rc); return ECORE_INVAL; } /* Set "pending" bit */ ECORE_MSG(sc, "pending bit was=%lx", o->pending); pending_bit = o->set_pending(o, params); ECORE_MSG(sc, "pending bit now=%lx", o->pending); /* Don't send a command if only driver cleanup was requested */ if (ECORE_TEST_BIT(RAMROD_DRV_CLR_ONLY, ¶ms->ramrod_flags)) o->complete_cmd(sc, o, pending_bit); else { /* Send a ramrod */ rc = o->send_cmd(sc, params); if (rc) { o->next_state = ECORE_Q_STATE_MAX; ECORE_CLEAR_BIT(pending_bit, pending); ECORE_SMP_MB_AFTER_CLEAR_BIT(); return rc; } if (ECORE_TEST_BIT(RAMROD_COMP_WAIT, ¶ms->ramrod_flags)) { rc = o->wait_comp(sc, o, pending_bit); if (rc) return rc; return ECORE_SUCCESS; } } return ECORE_RET_PENDING(pending_bit, pending); } static int ecore_queue_set_pending(struct ecore_queue_sp_obj *obj, struct ecore_queue_state_params *params) { enum ecore_queue_cmd cmd = params->cmd, bit; /* ACTIVATE and DEACTIVATE commands are implemented on top of * UPDATE command. */ if ((cmd == ECORE_Q_CMD_ACTIVATE) || (cmd == ECORE_Q_CMD_DEACTIVATE)) bit = ECORE_Q_CMD_UPDATE; else bit = cmd; ECORE_SET_BIT(bit, &obj->pending); return bit; } static int ecore_queue_wait_comp(struct bnx2x_softc *sc, struct ecore_queue_sp_obj *o, enum ecore_queue_cmd cmd) { return ecore_state_wait(sc, cmd, &o->pending); } /** * ecore_queue_comp_cmd - complete the state change command. * * @sc: device handle * @o: * @cmd: * * Checks that the arrived completion is expected. */ static int ecore_queue_comp_cmd(struct bnx2x_softc *sc __rte_unused, struct ecore_queue_sp_obj *o, enum ecore_queue_cmd cmd) { unsigned long cur_pending = o->pending; if (!ECORE_TEST_AND_CLEAR_BIT(cmd, &cur_pending)) { PMD_DRV_LOG(ERR, sc, "Bad MC reply %d for queue %d in state %d pending 0x%lx, next_state %d", cmd, o->cids[ECORE_PRIMARY_CID_INDEX], o->state, cur_pending, o->next_state); return ECORE_INVAL; } if (o->next_tx_only >= o->max_cos) /* >= because tx only must always be smaller than cos since the * primary connection supports COS 0 */ PMD_DRV_LOG(ERR, sc, "illegal value for next tx_only: %d. max cos was %d", o->next_tx_only, o->max_cos); ECORE_MSG(sc, "Completing command %d for queue %d, setting state to %d", cmd, o->cids[ECORE_PRIMARY_CID_INDEX], o->next_state); if (o->next_tx_only) /* print num tx-only if any exist */ ECORE_MSG(sc, "primary cid %d: num tx-only cons %d", o->cids[ECORE_PRIMARY_CID_INDEX], o->next_tx_only); o->state = o->next_state; o->num_tx_only = o->next_tx_only; o->next_state = ECORE_Q_STATE_MAX; /* It's important that o->state and o->next_state are * updated before o->pending. */ wmb(); ECORE_CLEAR_BIT(cmd, &o->pending); ECORE_SMP_MB_AFTER_CLEAR_BIT(); return ECORE_SUCCESS; } static void ecore_q_fill_setup_data_e2(struct ecore_queue_state_params *cmd_params, struct client_init_ramrod_data *data) { struct ecore_queue_setup_params *params = &cmd_params->params.setup; /* Rx data */ /* IPv6 TPA supported for E2 and above only */ data->rx.tpa_en |= ECORE_TEST_BIT(ECORE_Q_FLG_TPA_IPV6, ¶ms->flags) * CLIENT_INIT_RX_DATA_TPA_EN_IPV6; } static void ecore_q_fill_init_general_data(struct bnx2x_softc *sc __rte_unused, struct ecore_queue_sp_obj *o, struct ecore_general_setup_params *params, struct client_init_general_data *gen_data, unsigned long *flags) { gen_data->client_id = o->cl_id; if (ECORE_TEST_BIT(ECORE_Q_FLG_STATS, flags)) { gen_data->statistics_counter_id = params->stat_id; gen_data->statistics_en_flg = 1; gen_data->statistics_zero_flg = ECORE_TEST_BIT(ECORE_Q_FLG_ZERO_STATS, flags); } else gen_data->statistics_counter_id = DISABLE_STATISTIC_COUNTER_ID_VALUE; gen_data->is_fcoe_flg = ECORE_TEST_BIT(ECORE_Q_FLG_FCOE, flags); gen_data->activate_flg = ECORE_TEST_BIT(ECORE_Q_FLG_ACTIVE, flags); gen_data->sp_client_id = params->spcl_id; gen_data->mtu = ECORE_CPU_TO_LE16(params->mtu); gen_data->func_id = o->func_id; gen_data->cos = params->cos; gen_data->traffic_type = ECORE_TEST_BIT(ECORE_Q_FLG_FCOE, flags) ? LLFC_TRAFFIC_TYPE_FCOE : LLFC_TRAFFIC_TYPE_NW; ECORE_MSG(sc, "flags: active %d, cos %d, stats en %d", gen_data->activate_flg, gen_data->cos, gen_data->statistics_en_flg); } static void ecore_q_fill_init_tx_data(struct ecore_txq_setup_params *params, struct client_init_tx_data *tx_data, unsigned long *flags) { tx_data->enforce_security_flg = ECORE_TEST_BIT(ECORE_Q_FLG_TX_SEC, flags); tx_data->default_vlan = ECORE_CPU_TO_LE16(params->default_vlan); tx_data->default_vlan_flg = ECORE_TEST_BIT(ECORE_Q_FLG_DEF_VLAN, flags); tx_data->tx_switching_flg = ECORE_TEST_BIT(ECORE_Q_FLG_TX_SWITCH, flags); tx_data->anti_spoofing_flg = ECORE_TEST_BIT(ECORE_Q_FLG_ANTI_SPOOF, flags); tx_data->force_default_pri_flg = ECORE_TEST_BIT(ECORE_Q_FLG_FORCE_DEFAULT_PRI, flags); tx_data->refuse_outband_vlan_flg = ECORE_TEST_BIT(ECORE_Q_FLG_REFUSE_OUTBAND_VLAN, flags); tx_data->tunnel_non_lso_pcsum_location = ECORE_TEST_BIT(ECORE_Q_FLG_PCSUM_ON_PKT, flags) ? CSUM_ON_PKT : CSUM_ON_BD; tx_data->tx_status_block_id = params->fw_sb_id; tx_data->tx_sb_index_number = params->sb_cq_index; tx_data->tss_leading_client_id = params->tss_leading_cl_id; tx_data->tx_bd_page_base.lo = ECORE_CPU_TO_LE32(U64_LO(params->dscr_map)); tx_data->tx_bd_page_base.hi = ECORE_CPU_TO_LE32(U64_HI(params->dscr_map)); /* Don't configure any Tx switching mode during queue SETUP */ tx_data->state = 0; } static void ecore_q_fill_init_pause_data(struct rxq_pause_params *params, struct client_init_rx_data *rx_data) { /* flow control data */ rx_data->cqe_pause_thr_low = ECORE_CPU_TO_LE16(params->rcq_th_lo); rx_data->cqe_pause_thr_high = ECORE_CPU_TO_LE16(params->rcq_th_hi); rx_data->bd_pause_thr_low = ECORE_CPU_TO_LE16(params->bd_th_lo); rx_data->bd_pause_thr_high = ECORE_CPU_TO_LE16(params->bd_th_hi); rx_data->sge_pause_thr_low = ECORE_CPU_TO_LE16(params->sge_th_lo); rx_data->sge_pause_thr_high = ECORE_CPU_TO_LE16(params->sge_th_hi); rx_data->rx_cos_mask = ECORE_CPU_TO_LE16(params->pri_map); } static void ecore_q_fill_init_rx_data(struct ecore_rxq_setup_params *params, struct client_init_rx_data *rx_data, unsigned long *flags) { rx_data->tpa_en = ECORE_TEST_BIT(ECORE_Q_FLG_TPA, flags) * CLIENT_INIT_RX_DATA_TPA_EN_IPV4; rx_data->tpa_en |= ECORE_TEST_BIT(ECORE_Q_FLG_TPA_GRO, flags) * CLIENT_INIT_RX_DATA_TPA_MODE; rx_data->vmqueue_mode_en_flg = 0; rx_data->extra_data_over_sgl_en_flg = ECORE_TEST_BIT(ECORE_Q_FLG_OOO, flags); rx_data->cache_line_alignment_log_size = params->cache_line_log; rx_data->enable_dynamic_hc = ECORE_TEST_BIT(ECORE_Q_FLG_DHC, flags); rx_data->client_qzone_id = params->cl_qzone_id; rx_data->max_agg_size = ECORE_CPU_TO_LE16(params->tpa_agg_sz); /* Always start in DROP_ALL mode */ rx_data->state = ECORE_CPU_TO_LE16(CLIENT_INIT_RX_DATA_UCAST_DROP_ALL | CLIENT_INIT_RX_DATA_MCAST_DROP_ALL); /* We don't set drop flags */ rx_data->drop_ip_cs_err_flg = 0; rx_data->drop_tcp_cs_err_flg = 0; rx_data->drop_ttl0_flg = 0; rx_data->drop_udp_cs_err_flg = 0; rx_data->inner_vlan_removal_enable_flg = ECORE_TEST_BIT(ECORE_Q_FLG_VLAN, flags); rx_data->outer_vlan_removal_enable_flg = ECORE_TEST_BIT(ECORE_Q_FLG_OV, flags); rx_data->status_block_id = params->fw_sb_id; rx_data->rx_sb_index_number = params->sb_cq_index; rx_data->max_tpa_queues = params->max_tpa_queues; rx_data->max_bytes_on_bd = ECORE_CPU_TO_LE16(params->buf_sz); rx_data->bd_page_base.lo = ECORE_CPU_TO_LE32(U64_LO(params->dscr_map)); rx_data->bd_page_base.hi = ECORE_CPU_TO_LE32(U64_HI(params->dscr_map)); rx_data->cqe_page_base.lo = ECORE_CPU_TO_LE32(U64_LO(params->rcq_map)); rx_data->cqe_page_base.hi = ECORE_CPU_TO_LE32(U64_HI(params->rcq_map)); rx_data->is_leading_rss = ECORE_TEST_BIT(ECORE_Q_FLG_LEADING_RSS, flags); if (ECORE_TEST_BIT(ECORE_Q_FLG_MCAST, flags)) { rx_data->approx_mcast_engine_id = params->mcast_engine_id; rx_data->is_approx_mcast = 1; } rx_data->rss_engine_id = params->rss_engine_id; /* silent vlan removal */ rx_data->silent_vlan_removal_flg = ECORE_TEST_BIT(ECORE_Q_FLG_SILENT_VLAN_REM, flags); rx_data->silent_vlan_value = ECORE_CPU_TO_LE16(params->silent_removal_value); rx_data->silent_vlan_mask = ECORE_CPU_TO_LE16(params->silent_removal_mask); } /* initialize the general, tx and rx parts of a queue object */ static void ecore_q_fill_setup_data_cmn(struct bnx2x_softc *sc, struct ecore_queue_state_params *cmd_params, struct client_init_ramrod_data *data) { ecore_q_fill_init_general_data(sc, cmd_params->q_obj, &cmd_params->params.setup.gen_params, &data->general, &cmd_params->params.setup.flags); ecore_q_fill_init_tx_data(&cmd_params->params.setup.txq_params, &data->tx, &cmd_params->params.setup.flags); ecore_q_fill_init_rx_data(&cmd_params->params.setup.rxq_params, &data->rx, &cmd_params->params.setup.flags); ecore_q_fill_init_pause_data(&cmd_params->params.setup.pause_params, &data->rx); } /* initialize the general and tx parts of a tx-only queue object */ static void ecore_q_fill_setup_tx_only(struct bnx2x_softc *sc, struct ecore_queue_state_params *cmd_params, struct tx_queue_init_ramrod_data *data) { ecore_q_fill_init_general_data(sc, cmd_params->q_obj, &cmd_params->params.tx_only.gen_params, &data->general, &cmd_params->params.tx_only.flags); ecore_q_fill_init_tx_data(&cmd_params->params.tx_only.txq_params, &data->tx, &cmd_params->params.tx_only.flags); ECORE_MSG(sc, "cid %d, tx bd page lo %x hi %x", cmd_params->q_obj->cids[0], data->tx.tx_bd_page_base.lo, data->tx.tx_bd_page_base.hi); } /** * ecore_q_init - init HW/FW queue * * @sc: device handle * @params: * * HW/FW initial Queue configuration: * - HC: Rx and Tx * - CDU context validation * */ static int ecore_q_init(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_sp_obj *o = params->q_obj; struct ecore_queue_init_params *init = ¶ms->params.init; uint16_t hc_usec; uint8_t cos; /* Tx HC configuration */ if (ECORE_TEST_BIT(ECORE_Q_TYPE_HAS_TX, &o->type) && ECORE_TEST_BIT(ECORE_Q_FLG_HC, &init->tx.flags)) { hc_usec = init->tx.hc_rate ? 1000000 / init->tx.hc_rate : 0; ECORE_UPDATE_COALESCE_SB_INDEX(sc, init->tx.fw_sb_id, init->tx.sb_cq_index, !ECORE_TEST_BIT (ECORE_Q_FLG_HC_EN, &init->tx.flags), hc_usec); } /* Rx HC configuration */ if (ECORE_TEST_BIT(ECORE_Q_TYPE_HAS_RX, &o->type) && ECORE_TEST_BIT(ECORE_Q_FLG_HC, &init->rx.flags)) { hc_usec = init->rx.hc_rate ? 1000000 / init->rx.hc_rate : 0; ECORE_UPDATE_COALESCE_SB_INDEX(sc, init->rx.fw_sb_id, init->rx.sb_cq_index, !ECORE_TEST_BIT (ECORE_Q_FLG_HC_EN, &init->rx.flags), hc_usec); } /* Set CDU context validation values */ for (cos = 0; cos < o->max_cos; cos++) { ECORE_MSG(sc, "setting context validation. cid %d, cos %d", o->cids[cos], cos); ECORE_MSG(sc, "context pointer %p", init->cxts[cos]); ECORE_SET_CTX_VALIDATION(sc, init->cxts[cos], o->cids[cos]); } /* As no ramrod is sent, complete the command immediately */ o->complete_cmd(sc, o, ECORE_Q_CMD_INIT); ECORE_MMIOWB(); ECORE_SMP_MB(); return ECORE_SUCCESS; } static int ecore_q_send_setup_e1x(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_sp_obj *o = params->q_obj; struct client_init_ramrod_data *rdata = (struct client_init_ramrod_data *)o->rdata; ecore_dma_addr_t data_mapping = o->rdata_mapping; int ramrod = RAMROD_CMD_ID_ETH_CLIENT_SETUP; /* Clear the ramrod data */ ECORE_MEMSET(rdata, 0, sizeof(*rdata)); /* Fill the ramrod data */ ecore_q_fill_setup_data_cmn(sc, params, rdata); /* No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ return ecore_sp_post(sc, ramrod, o->cids[ECORE_PRIMARY_CID_INDEX], data_mapping, ETH_CONNECTION_TYPE); } static int ecore_q_send_setup_e2(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_sp_obj *o = params->q_obj; struct client_init_ramrod_data *rdata = (struct client_init_ramrod_data *)o->rdata; ecore_dma_addr_t data_mapping = o->rdata_mapping; int ramrod = RAMROD_CMD_ID_ETH_CLIENT_SETUP; /* Clear the ramrod data */ ECORE_MEMSET(rdata, 0, sizeof(*rdata)); /* Fill the ramrod data */ ecore_q_fill_setup_data_cmn(sc, params, rdata); ecore_q_fill_setup_data_e2(params, rdata); /* No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ return ecore_sp_post(sc, ramrod, o->cids[ECORE_PRIMARY_CID_INDEX], data_mapping, ETH_CONNECTION_TYPE); } static int ecore_q_send_setup_tx_only(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_sp_obj *o = params->q_obj; struct tx_queue_init_ramrod_data *rdata = (struct tx_queue_init_ramrod_data *)o->rdata; ecore_dma_addr_t data_mapping = o->rdata_mapping; int ramrod = RAMROD_CMD_ID_ETH_TX_QUEUE_SETUP; struct ecore_queue_setup_tx_only_params *tx_only_params = ¶ms->params.tx_only; uint8_t cid_index = tx_only_params->cid_index; if (ECORE_TEST_BIT(ECORE_Q_TYPE_FWD, &o->type)) ramrod = RAMROD_CMD_ID_ETH_FORWARD_SETUP; ECORE_MSG(sc, "sending forward tx-only ramrod"); if (cid_index >= o->max_cos) { PMD_DRV_LOG(ERR, sc, "queue[%d]: cid_index (%d) is out of range", o->cl_id, cid_index); return ECORE_INVAL; } ECORE_MSG(sc, "parameters received: cos: %d sp-id: %d", tx_only_params->gen_params.cos, tx_only_params->gen_params.spcl_id); /* Clear the ramrod data */ ECORE_MEMSET(rdata, 0, sizeof(*rdata)); /* Fill the ramrod data */ ecore_q_fill_setup_tx_only(sc, params, rdata); ECORE_MSG (sc, "sending tx-only ramrod: cid %d, client-id %d, sp-client id %d, cos %d", o->cids[cid_index], rdata->general.client_id, rdata->general.sp_client_id, rdata->general.cos); /* No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ return ecore_sp_post(sc, ramrod, o->cids[cid_index], data_mapping, ETH_CONNECTION_TYPE); } static void ecore_q_fill_update_data(struct ecore_queue_sp_obj *obj, struct ecore_queue_update_params *params, struct client_update_ramrod_data *data) { /* Client ID of the client to update */ data->client_id = obj->cl_id; /* Function ID of the client to update */ data->func_id = obj->func_id; /* Default VLAN value */ data->default_vlan = ECORE_CPU_TO_LE16(params->def_vlan); /* Inner VLAN stripping */ data->inner_vlan_removal_enable_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_IN_VLAN_REM, ¶ms->update_flags); data->inner_vlan_removal_change_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_IN_VLAN_REM_CHNG, ¶ms->update_flags); /* Outer VLAN stripping */ data->outer_vlan_removal_enable_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_OUT_VLAN_REM, ¶ms->update_flags); data->outer_vlan_removal_change_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_OUT_VLAN_REM_CHNG, ¶ms->update_flags); /* Drop packets that have source MAC that doesn't belong to this * Queue. */ data->anti_spoofing_enable_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_ANTI_SPOOF, ¶ms->update_flags); data->anti_spoofing_change_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_ANTI_SPOOF_CHNG, ¶ms->update_flags); /* Activate/Deactivate */ data->activate_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_ACTIVATE, ¶ms->update_flags); data->activate_change_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_ACTIVATE_CHNG, ¶ms->update_flags); /* Enable default VLAN */ data->default_vlan_enable_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_DEF_VLAN_EN, ¶ms->update_flags); data->default_vlan_change_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_DEF_VLAN_EN_CHNG, ¶ms->update_flags); /* silent vlan removal */ data->silent_vlan_change_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_SILENT_VLAN_REM_CHNG, ¶ms->update_flags); data->silent_vlan_removal_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_SILENT_VLAN_REM, ¶ms->update_flags); data->silent_vlan_value = ECORE_CPU_TO_LE16(params->silent_removal_value); data->silent_vlan_mask = ECORE_CPU_TO_LE16(params->silent_removal_mask); /* tx switching */ data->tx_switching_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_TX_SWITCHING, ¶ms->update_flags); data->tx_switching_change_flg = ECORE_TEST_BIT(ECORE_Q_UPDATE_TX_SWITCHING_CHNG, ¶ms->update_flags); } static int ecore_q_send_update(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_sp_obj *o = params->q_obj; struct client_update_ramrod_data *rdata = (struct client_update_ramrod_data *)o->rdata; ecore_dma_addr_t data_mapping = o->rdata_mapping; struct ecore_queue_update_params *update_params = ¶ms->params.update; uint8_t cid_index = update_params->cid_index; if (cid_index >= o->max_cos) { PMD_DRV_LOG(ERR, sc, "queue[%d]: cid_index (%d) is out of range", o->cl_id, cid_index); return ECORE_INVAL; } /* Clear the ramrod data */ ECORE_MEMSET(rdata, 0, sizeof(*rdata)); /* Fill the ramrod data */ ecore_q_fill_update_data(o, update_params, rdata); /* No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ return ecore_sp_post(sc, RAMROD_CMD_ID_ETH_CLIENT_UPDATE, o->cids[cid_index], data_mapping, ETH_CONNECTION_TYPE); } /** * ecore_q_send_deactivate - send DEACTIVATE command * * @sc: device handle * @params: * * implemented using the UPDATE command. */ static int ecore_q_send_deactivate(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_update_params *update = ¶ms->params.update; ECORE_MEMSET(update, 0, sizeof(*update)); ECORE_SET_BIT_NA(ECORE_Q_UPDATE_ACTIVATE_CHNG, &update->update_flags); return ecore_q_send_update(sc, params); } /** * ecore_q_send_activate - send ACTIVATE command * * @sc: device handle * @params: * * implemented using the UPDATE command. */ static int ecore_q_send_activate(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_update_params *update = ¶ms->params.update; ECORE_MEMSET(update, 0, sizeof(*update)); ECORE_SET_BIT_NA(ECORE_Q_UPDATE_ACTIVATE, &update->update_flags); ECORE_SET_BIT_NA(ECORE_Q_UPDATE_ACTIVATE_CHNG, &update->update_flags); return ecore_q_send_update(sc, params); } static int ecore_q_send_update_tpa(__rte_unused struct bnx2x_softc *sc, __rte_unused struct ecore_queue_state_params *params) { /* Not implemented yet. */ return -1; } static int ecore_q_send_halt(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_sp_obj *o = params->q_obj; /* build eth_halt_ramrod_data.client_id in a big-endian friendly way */ ecore_dma_addr_t data_mapping = 0; data_mapping = (ecore_dma_addr_t) o->cl_id; return ecore_sp_post(sc, RAMROD_CMD_ID_ETH_HALT, o->cids[ECORE_PRIMARY_CID_INDEX], data_mapping, ETH_CONNECTION_TYPE); } static int ecore_q_send_cfc_del(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_sp_obj *o = params->q_obj; uint8_t cid_idx = params->params.cfc_del.cid_index; if (cid_idx >= o->max_cos) { PMD_DRV_LOG(ERR, sc, "queue[%d]: cid_index (%d) is out of range", o->cl_id, cid_idx); return ECORE_INVAL; } return ecore_sp_post(sc, RAMROD_CMD_ID_COMMON_CFC_DEL, o->cids[cid_idx], 0, NONE_CONNECTION_TYPE); } static int ecore_q_send_terminate(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_sp_obj *o = params->q_obj; uint8_t cid_index = params->params.terminate.cid_index; if (cid_index >= o->max_cos) { PMD_DRV_LOG(ERR, sc, "queue[%d]: cid_index (%d) is out of range", o->cl_id, cid_index); return ECORE_INVAL; } return ecore_sp_post(sc, RAMROD_CMD_ID_ETH_TERMINATE, o->cids[cid_index], 0, ETH_CONNECTION_TYPE); } static int ecore_q_send_empty(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { struct ecore_queue_sp_obj *o = params->q_obj; return ecore_sp_post(sc, RAMROD_CMD_ID_ETH_EMPTY, o->cids[ECORE_PRIMARY_CID_INDEX], 0, ETH_CONNECTION_TYPE); } static int ecore_queue_send_cmd_cmn(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { switch (params->cmd) { case ECORE_Q_CMD_INIT: return ecore_q_init(sc, params); case ECORE_Q_CMD_SETUP_TX_ONLY: return ecore_q_send_setup_tx_only(sc, params); case ECORE_Q_CMD_DEACTIVATE: return ecore_q_send_deactivate(sc, params); case ECORE_Q_CMD_ACTIVATE: return ecore_q_send_activate(sc, params); case ECORE_Q_CMD_UPDATE: return ecore_q_send_update(sc, params); case ECORE_Q_CMD_UPDATE_TPA: return ecore_q_send_update_tpa(sc, params); case ECORE_Q_CMD_HALT: return ecore_q_send_halt(sc, params); case ECORE_Q_CMD_CFC_DEL: return ecore_q_send_cfc_del(sc, params); case ECORE_Q_CMD_TERMINATE: return ecore_q_send_terminate(sc, params); case ECORE_Q_CMD_EMPTY: return ecore_q_send_empty(sc, params); default: PMD_DRV_LOG(ERR, sc, "Unknown command: %d", params->cmd); return ECORE_INVAL; } } static int ecore_queue_send_cmd_e1x(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { switch (params->cmd) { case ECORE_Q_CMD_SETUP: return ecore_q_send_setup_e1x(sc, params); case ECORE_Q_CMD_INIT: case ECORE_Q_CMD_SETUP_TX_ONLY: case ECORE_Q_CMD_DEACTIVATE: case ECORE_Q_CMD_ACTIVATE: case ECORE_Q_CMD_UPDATE: case ECORE_Q_CMD_UPDATE_TPA: case ECORE_Q_CMD_HALT: case ECORE_Q_CMD_CFC_DEL: case ECORE_Q_CMD_TERMINATE: case ECORE_Q_CMD_EMPTY: return ecore_queue_send_cmd_cmn(sc, params); default: PMD_DRV_LOG(ERR, sc, "Unknown command: %d", params->cmd); return ECORE_INVAL; } } static int ecore_queue_send_cmd_e2(struct bnx2x_softc *sc, struct ecore_queue_state_params *params) { switch (params->cmd) { case ECORE_Q_CMD_SETUP: return ecore_q_send_setup_e2(sc, params); case ECORE_Q_CMD_INIT: case ECORE_Q_CMD_SETUP_TX_ONLY: case ECORE_Q_CMD_DEACTIVATE: case ECORE_Q_CMD_ACTIVATE: case ECORE_Q_CMD_UPDATE: case ECORE_Q_CMD_UPDATE_TPA: case ECORE_Q_CMD_HALT: case ECORE_Q_CMD_CFC_DEL: case ECORE_Q_CMD_TERMINATE: case ECORE_Q_CMD_EMPTY: return ecore_queue_send_cmd_cmn(sc, params); default: PMD_DRV_LOG(ERR, sc, "Unknown command: %d", params->cmd); return ECORE_INVAL; } } /** * ecore_queue_chk_transition - check state machine of a regular Queue * * @sc: device handle * @o: * @params: * * (not Forwarding) * It both checks if the requested command is legal in a current * state and, if it's legal, sets a `next_state' in the object * that will be used in the completion flow to set the `state' * of the object. * * returns 0 if a requested command is a legal transition, * ECORE_INVAL otherwise. */ static int ecore_queue_chk_transition(struct bnx2x_softc *sc __rte_unused, struct ecore_queue_sp_obj *o, struct ecore_queue_state_params *params) { enum ecore_q_state state = o->state, next_state = ECORE_Q_STATE_MAX; enum ecore_queue_cmd cmd = params->cmd; struct ecore_queue_update_params *update_params = ¶ms->params.update; uint8_t next_tx_only = o->num_tx_only; /* Forget all pending for completion commands if a driver only state * transition has been requested. */ if (ECORE_TEST_BIT(RAMROD_DRV_CLR_ONLY, ¶ms->ramrod_flags)) { o->pending = 0; o->next_state = ECORE_Q_STATE_MAX; } /* Don't allow a next state transition if we are in the middle of * the previous one. */ if (o->pending) { PMD_DRV_LOG(ERR, sc, "Blocking transition since pending was %lx", o->pending); return ECORE_BUSY; } switch (state) { case ECORE_Q_STATE_RESET: if (cmd == ECORE_Q_CMD_INIT) next_state = ECORE_Q_STATE_INITIALIZED; break; case ECORE_Q_STATE_INITIALIZED: if (cmd == ECORE_Q_CMD_SETUP) { if (ECORE_TEST_BIT(ECORE_Q_FLG_ACTIVE, ¶ms->params.setup.flags)) next_state = ECORE_Q_STATE_ACTIVE; else next_state = ECORE_Q_STATE_INACTIVE; } break; case ECORE_Q_STATE_ACTIVE: if (cmd == ECORE_Q_CMD_DEACTIVATE) next_state = ECORE_Q_STATE_INACTIVE; else if ((cmd == ECORE_Q_CMD_EMPTY) || (cmd == ECORE_Q_CMD_UPDATE_TPA)) next_state = ECORE_Q_STATE_ACTIVE; else if (cmd == ECORE_Q_CMD_SETUP_TX_ONLY) { next_state = ECORE_Q_STATE_MULTI_COS; next_tx_only = 1; } else if (cmd == ECORE_Q_CMD_HALT) next_state = ECORE_Q_STATE_STOPPED; else if (cmd == ECORE_Q_CMD_UPDATE) { /* If "active" state change is requested, update the * state accordingly. */ if (ECORE_TEST_BIT(ECORE_Q_UPDATE_ACTIVATE_CHNG, &update_params->update_flags) && !ECORE_TEST_BIT(ECORE_Q_UPDATE_ACTIVATE, &update_params->update_flags)) next_state = ECORE_Q_STATE_INACTIVE; else next_state = ECORE_Q_STATE_ACTIVE; } break; case ECORE_Q_STATE_MULTI_COS: if (cmd == ECORE_Q_CMD_TERMINATE) next_state = ECORE_Q_STATE_MCOS_TERMINATED; else if (cmd == ECORE_Q_CMD_SETUP_TX_ONLY) { next_state = ECORE_Q_STATE_MULTI_COS; next_tx_only = o->num_tx_only + 1; } else if ((cmd == ECORE_Q_CMD_EMPTY) || (cmd == ECORE_Q_CMD_UPDATE_TPA)) next_state = ECORE_Q_STATE_MULTI_COS; else if (cmd == ECORE_Q_CMD_UPDATE) { /* If "active" state change is requested, update the * state accordingly. */ if (ECORE_TEST_BIT(ECORE_Q_UPDATE_ACTIVATE_CHNG, &update_params->update_flags) && !ECORE_TEST_BIT(ECORE_Q_UPDATE_ACTIVATE, &update_params->update_flags)) next_state = ECORE_Q_STATE_INACTIVE; else next_state = ECORE_Q_STATE_MULTI_COS; } break; case ECORE_Q_STATE_MCOS_TERMINATED: if (cmd == ECORE_Q_CMD_CFC_DEL) { next_tx_only = o->num_tx_only - 1; if (next_tx_only == 0) next_state = ECORE_Q_STATE_ACTIVE; else next_state = ECORE_Q_STATE_MULTI_COS; } break; case ECORE_Q_STATE_INACTIVE: if (cmd == ECORE_Q_CMD_ACTIVATE) next_state = ECORE_Q_STATE_ACTIVE; else if ((cmd == ECORE_Q_CMD_EMPTY) || (cmd == ECORE_Q_CMD_UPDATE_TPA)) next_state = ECORE_Q_STATE_INACTIVE; else if (cmd == ECORE_Q_CMD_HALT) next_state = ECORE_Q_STATE_STOPPED; else if (cmd == ECORE_Q_CMD_UPDATE) { /* If "active" state change is requested, update the * state accordingly. */ if (ECORE_TEST_BIT(ECORE_Q_UPDATE_ACTIVATE_CHNG, &update_params->update_flags) && ECORE_TEST_BIT(ECORE_Q_UPDATE_ACTIVATE, &update_params->update_flags)) { if (o->num_tx_only == 0) next_state = ECORE_Q_STATE_ACTIVE; else /* tx only queues exist for this queue */ next_state = ECORE_Q_STATE_MULTI_COS; } else next_state = ECORE_Q_STATE_INACTIVE; } break; case ECORE_Q_STATE_STOPPED: if (cmd == ECORE_Q_CMD_TERMINATE) next_state = ECORE_Q_STATE_TERMINATED; break; case ECORE_Q_STATE_TERMINATED: if (cmd == ECORE_Q_CMD_CFC_DEL) next_state = ECORE_Q_STATE_RESET; break; default: PMD_DRV_LOG(ERR, sc, "Illegal state: %d", state); } /* Transition is assured */ if (next_state != ECORE_Q_STATE_MAX) { ECORE_MSG(sc, "Good state transition: %d(%d)->%d", state, cmd, next_state); o->next_state = next_state; o->next_tx_only = next_tx_only; return ECORE_SUCCESS; } ECORE_MSG(sc, "Bad state transition request: %d %d", state, cmd); return ECORE_INVAL; } /** * ecore_queue_chk_fwd_transition - check state machine of a Forwarding Queue. * * @sc: device handle * @o: * @params: * * It both checks if the requested command is legal in a current * state and, if it's legal, sets a `next_state' in the object * that will be used in the completion flow to set the `state' * of the object. * * returns 0 if a requested command is a legal transition, * ECORE_INVAL otherwise. */ static int ecore_queue_chk_fwd_transition(struct bnx2x_softc *sc __rte_unused, struct ecore_queue_sp_obj *o, struct ecore_queue_state_params *params) { enum ecore_q_state state = o->state, next_state = ECORE_Q_STATE_MAX; enum ecore_queue_cmd cmd = params->cmd; switch (state) { case ECORE_Q_STATE_RESET: if (cmd == ECORE_Q_CMD_INIT) next_state = ECORE_Q_STATE_INITIALIZED; break; case ECORE_Q_STATE_INITIALIZED: if (cmd == ECORE_Q_CMD_SETUP_TX_ONLY) { if (ECORE_TEST_BIT(ECORE_Q_FLG_ACTIVE, ¶ms->params.tx_only.flags)) next_state = ECORE_Q_STATE_ACTIVE; else next_state = ECORE_Q_STATE_INACTIVE; } break; case ECORE_Q_STATE_ACTIVE: case ECORE_Q_STATE_INACTIVE: if (cmd == ECORE_Q_CMD_CFC_DEL) next_state = ECORE_Q_STATE_RESET; break; default: PMD_DRV_LOG(ERR, sc, "Illegal state: %d", state); } /* Transition is assured */ if (next_state != ECORE_Q_STATE_MAX) { ECORE_MSG(sc, "Good state transition: %d(%d)->%d", state, cmd, next_state); o->next_state = next_state; return ECORE_SUCCESS; } ECORE_MSG(sc, "Bad state transition request: %d %d", state, cmd); return ECORE_INVAL; } void ecore_init_queue_obj(struct bnx2x_softc *sc, struct ecore_queue_sp_obj *obj, uint8_t cl_id, uint32_t * cids, uint8_t cid_cnt, uint8_t func_id, void *rdata, ecore_dma_addr_t rdata_mapping, unsigned long type) { ECORE_MEMSET(obj, 0, sizeof(*obj)); /* We support only ECORE_MULTI_TX_COS Tx CoS at the moment */ ECORE_BUG_ON(ECORE_MULTI_TX_COS < cid_cnt); rte_memcpy(obj->cids, cids, sizeof(obj->cids[0]) * cid_cnt); obj->max_cos = cid_cnt; obj->cl_id = cl_id; obj->func_id = func_id; obj->rdata = rdata; obj->rdata_mapping = rdata_mapping; obj->type = type; obj->next_state = ECORE_Q_STATE_MAX; if (CHIP_IS_E1x(sc)) obj->send_cmd = ecore_queue_send_cmd_e1x; else obj->send_cmd = ecore_queue_send_cmd_e2; if (ECORE_TEST_BIT(ECORE_Q_TYPE_FWD, &type)) obj->check_transition = ecore_queue_chk_fwd_transition; else obj->check_transition = ecore_queue_chk_transition; obj->complete_cmd = ecore_queue_comp_cmd; obj->wait_comp = ecore_queue_wait_comp; obj->set_pending = ecore_queue_set_pending; } /********************** Function state object *********************************/ enum ecore_func_state ecore_func_get_state(__rte_unused struct bnx2x_softc *sc, struct ecore_func_sp_obj *o) { /* in the middle of transaction - return INVALID state */ if (o->pending) return ECORE_F_STATE_MAX; /* unsure the order of reading of o->pending and o->state * o->pending should be read first */ rmb(); return o->state; } static int ecore_func_wait_comp(struct bnx2x_softc *sc, struct ecore_func_sp_obj *o, enum ecore_func_cmd cmd) { return ecore_state_wait(sc, cmd, &o->pending); } /** * ecore_func_state_change_comp - complete the state machine transition * * @sc: device handle * @o: * @cmd: * * Called on state change transition. Completes the state * machine transition only - no HW interaction. */ static int ecore_func_state_change_comp(struct bnx2x_softc *sc __rte_unused, struct ecore_func_sp_obj *o, enum ecore_func_cmd cmd) { unsigned long cur_pending = o->pending; if (!ECORE_TEST_AND_CLEAR_BIT(cmd, &cur_pending)) { PMD_DRV_LOG(ERR, sc, "Bad MC reply %d for func %d in state %d pending 0x%lx, next_state %d", cmd, ECORE_FUNC_ID(sc), o->state, cur_pending, o->next_state); return ECORE_INVAL; } ECORE_MSG(sc, "Completing command %d for func %d, setting state to %d", cmd, ECORE_FUNC_ID(sc), o->next_state); o->state = o->next_state; o->next_state = ECORE_F_STATE_MAX; /* It's important that o->state and o->next_state are * updated before o->pending. */ wmb(); ECORE_CLEAR_BIT(cmd, &o->pending); ECORE_SMP_MB_AFTER_CLEAR_BIT(); return ECORE_SUCCESS; } /** * ecore_func_comp_cmd - complete the state change command * * @sc: device handle * @o: * @cmd: * * Checks that the arrived completion is expected. */ static int ecore_func_comp_cmd(struct bnx2x_softc *sc, struct ecore_func_sp_obj *o, enum ecore_func_cmd cmd) { /* Complete the state machine part first, check if it's a * legal completion. */ int rc = ecore_func_state_change_comp(sc, o, cmd); return rc; } /** * ecore_func_chk_transition - perform function state machine transition * * @sc: device handle * @o: * @params: * * It both checks if the requested command is legal in a current * state and, if it's legal, sets a `next_state' in the object * that will be used in the completion flow to set the `state' * of the object. * * returns 0 if a requested command is a legal transition, * ECORE_INVAL otherwise. */ static int ecore_func_chk_transition(struct bnx2x_softc *sc __rte_unused, struct ecore_func_sp_obj *o, struct ecore_func_state_params *params) { enum ecore_func_state state = o->state, next_state = ECORE_F_STATE_MAX; enum ecore_func_cmd cmd = params->cmd; /* Forget all pending for completion commands if a driver only state * transition has been requested. */ if (ECORE_TEST_BIT(RAMROD_DRV_CLR_ONLY, ¶ms->ramrod_flags)) { o->pending = 0; o->next_state = ECORE_F_STATE_MAX; } /* Don't allow a next state transition if we are in the middle of * the previous one. */ if (o->pending) return ECORE_BUSY; switch (state) { case ECORE_F_STATE_RESET: if (cmd == ECORE_F_CMD_HW_INIT) next_state = ECORE_F_STATE_INITIALIZED; break; case ECORE_F_STATE_INITIALIZED: if (cmd == ECORE_F_CMD_START) next_state = ECORE_F_STATE_STARTED; else if (cmd == ECORE_F_CMD_HW_RESET) next_state = ECORE_F_STATE_RESET; break; case ECORE_F_STATE_STARTED: if (cmd == ECORE_F_CMD_STOP) next_state = ECORE_F_STATE_INITIALIZED; /* afex ramrods can be sent only in started mode, and only * if not pending for function_stop ramrod completion * for these events - next state remained STARTED. */ else if ((cmd == ECORE_F_CMD_AFEX_UPDATE) && (!ECORE_TEST_BIT(ECORE_F_CMD_STOP, &o->pending))) next_state = ECORE_F_STATE_STARTED; else if ((cmd == ECORE_F_CMD_AFEX_VIFLISTS) && (!ECORE_TEST_BIT(ECORE_F_CMD_STOP, &o->pending))) next_state = ECORE_F_STATE_STARTED; /* Switch_update ramrod can be sent in either started or * tx_stopped state, and it doesn't change the state. */ else if ((cmd == ECORE_F_CMD_SWITCH_UPDATE) && (!ECORE_TEST_BIT(ECORE_F_CMD_STOP, &o->pending))) next_state = ECORE_F_STATE_STARTED; else if (cmd == ECORE_F_CMD_TX_STOP) next_state = ECORE_F_STATE_TX_STOPPED; break; case ECORE_F_STATE_TX_STOPPED: if ((cmd == ECORE_F_CMD_SWITCH_UPDATE) && (!ECORE_TEST_BIT(ECORE_F_CMD_STOP, &o->pending))) next_state = ECORE_F_STATE_TX_STOPPED; else if (cmd == ECORE_F_CMD_TX_START) next_state = ECORE_F_STATE_STARTED; break; default: PMD_DRV_LOG(ERR, sc, "Unknown state: %d", state); } /* Transition is assured */ if (next_state != ECORE_F_STATE_MAX) { ECORE_MSG(sc, "Good function state transition: %d(%d)->%d", state, cmd, next_state); o->next_state = next_state; return ECORE_SUCCESS; } ECORE_MSG(sc, "Bad function state transition request: %d %d", state, cmd); return ECORE_INVAL; } /** * ecore_func_init_func - performs HW init at function stage * * @sc: device handle * @drv: * * Init HW when the current phase is * FW_MSG_CODE_DRV_LOAD_FUNCTION: initialize only FUNCTION-only * HW blocks. */ static int ecore_func_init_func(struct bnx2x_softc *sc, const struct ecore_func_sp_drv_ops *drv) { return drv->init_hw_func(sc); } /** * ecore_func_init_port - performs HW init at port stage * * @sc: device handle * @drv: * * Init HW when the current phase is * FW_MSG_CODE_DRV_LOAD_PORT: initialize PORT-only and * FUNCTION-only HW blocks. * */ static int ecore_func_init_port(struct bnx2x_softc *sc, const struct ecore_func_sp_drv_ops *drv) { int rc = drv->init_hw_port(sc); if (rc) return rc; return ecore_func_init_func(sc, drv); } /** * ecore_func_init_cmn_chip - performs HW init at chip-common stage * * @sc: device handle * @drv: * * Init HW when the current phase is * FW_MSG_CODE_DRV_LOAD_COMMON_CHIP: initialize COMMON_CHIP, * PORT-only and FUNCTION-only HW blocks. */ static int ecore_func_init_cmn_chip(struct bnx2x_softc *sc, const struct ecore_func_sp_drv_ops *drv) { int rc = drv->init_hw_cmn_chip(sc); if (rc) return rc; return ecore_func_init_port(sc, drv); } /** * ecore_func_init_cmn - performs HW init at common stage * * @sc: device handle * @drv: * * Init HW when the current phase is * FW_MSG_CODE_DRV_LOAD_COMMON_CHIP: initialize COMMON, * PORT-only and FUNCTION-only HW blocks. */ static int ecore_func_init_cmn(struct bnx2x_softc *sc, const struct ecore_func_sp_drv_ops *drv) { int rc = drv->init_hw_cmn(sc); if (rc) return rc; return ecore_func_init_port(sc, drv); } static int ecore_func_hw_init(struct bnx2x_softc *sc, struct ecore_func_state_params *params) { uint32_t load_code = params->params.hw_init.load_phase; struct ecore_func_sp_obj *o = params->f_obj; const struct ecore_func_sp_drv_ops *drv = o->drv; int rc = 0; ECORE_MSG(sc, "function %d load_code %x", ECORE_ABS_FUNC_ID(sc), load_code); /* Prepare FW */ rc = drv->init_fw(sc); if (rc) { PMD_DRV_LOG(ERR, sc, "Error loading firmware"); goto init_err; } /* Handle the beginning of COMMON_XXX pases separately... */ switch (load_code) { case FW_MSG_CODE_DRV_LOAD_COMMON_CHIP: rc = ecore_func_init_cmn_chip(sc, drv); if (rc) goto init_err; break; case FW_MSG_CODE_DRV_LOAD_COMMON: rc = ecore_func_init_cmn(sc, drv); if (rc) goto init_err; break; case FW_MSG_CODE_DRV_LOAD_PORT: rc = ecore_func_init_port(sc, drv); if (rc) goto init_err; break; case FW_MSG_CODE_DRV_LOAD_FUNCTION: rc = ecore_func_init_func(sc, drv); if (rc) goto init_err; break; default: PMD_DRV_LOG(ERR, sc, "Unknown load_code (0x%x) from MCP", load_code); rc = ECORE_INVAL; } init_err: /* In case of success, complete the command immediately: no ramrods * have been sent. */ if (!rc) o->complete_cmd(sc, o, ECORE_F_CMD_HW_INIT); return rc; } /** * ecore_func_reset_func - reset HW at function stage * * @sc: device handle * @drv: * * Reset HW at FW_MSG_CODE_DRV_UNLOAD_FUNCTION stage: reset only * FUNCTION-only HW blocks. */ static void ecore_func_reset_func(struct bnx2x_softc *sc, const struct ecore_func_sp_drv_ops *drv) { drv->reset_hw_func(sc); } /** * ecore_func_reset_port - reser HW at port stage * * @sc: device handle * @drv: * * Reset HW at FW_MSG_CODE_DRV_UNLOAD_PORT stage: reset * FUNCTION-only and PORT-only HW blocks. * * !!!IMPORTANT!!! * * It's important to call reset_port before reset_func() as the last thing * reset_func does is pf_disable() thus disabling PGLUE_B, which * makes impossible any DMAE transactions. */ static void ecore_func_reset_port(struct bnx2x_softc *sc, const struct ecore_func_sp_drv_ops *drv) { drv->reset_hw_port(sc); ecore_func_reset_func(sc, drv); } /** * ecore_func_reset_cmn - reser HW at common stage * * @sc: device handle * @drv: * * Reset HW at FW_MSG_CODE_DRV_UNLOAD_COMMON and * FW_MSG_CODE_DRV_UNLOAD_COMMON_CHIP stages: reset COMMON, * COMMON_CHIP, FUNCTION-only and PORT-only HW blocks. */ static void ecore_func_reset_cmn(struct bnx2x_softc *sc, const struct ecore_func_sp_drv_ops *drv) { ecore_func_reset_port(sc, drv); drv->reset_hw_cmn(sc); } static int ecore_func_hw_reset(struct bnx2x_softc *sc, struct ecore_func_state_params *params) { uint32_t reset_phase = params->params.hw_reset.reset_phase; struct ecore_func_sp_obj *o = params->f_obj; const struct ecore_func_sp_drv_ops *drv = o->drv; ECORE_MSG(sc, "function %d reset_phase %x", ECORE_ABS_FUNC_ID(sc), reset_phase); switch (reset_phase) { case FW_MSG_CODE_DRV_UNLOAD_COMMON: ecore_func_reset_cmn(sc, drv); break; case FW_MSG_CODE_DRV_UNLOAD_PORT: ecore_func_reset_port(sc, drv); break; case FW_MSG_CODE_DRV_UNLOAD_FUNCTION: ecore_func_reset_func(sc, drv); break; default: PMD_DRV_LOG(ERR, sc, "Unknown reset_phase (0x%x) from MCP", reset_phase); break; } /* Complete the command immediately: no ramrods have been sent. */ o->complete_cmd(sc, o, ECORE_F_CMD_HW_RESET); return ECORE_SUCCESS; } static int ecore_func_send_start(struct bnx2x_softc *sc, struct ecore_func_state_params *params) { struct ecore_func_sp_obj *o = params->f_obj; struct function_start_data *rdata = (struct function_start_data *)o->rdata; ecore_dma_addr_t data_mapping = o->rdata_mapping; struct ecore_func_start_params *start_params = ¶ms->params.start; ECORE_MEMSET(rdata, 0, sizeof(*rdata)); /* Fill the ramrod data with provided parameters */ rdata->function_mode = (uint8_t) start_params->mf_mode; rdata->sd_vlan_tag = ECORE_CPU_TO_LE16(start_params->sd_vlan_tag); rdata->path_id = ECORE_PATH_ID(sc); rdata->network_cos_mode = start_params->network_cos_mode; /* * No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ return ecore_sp_post(sc, RAMROD_CMD_ID_COMMON_FUNCTION_START, 0, data_mapping, NONE_CONNECTION_TYPE); } static int ecore_func_send_switch_update(struct bnx2x_softc *sc, struct ecore_func_state_params *params) { struct ecore_func_sp_obj *o = params->f_obj; struct function_update_data *rdata = (struct function_update_data *)o->rdata; ecore_dma_addr_t data_mapping = o->rdata_mapping; struct ecore_func_switch_update_params *switch_update_params = ¶ms->params.switch_update; ECORE_MEMSET(rdata, 0, sizeof(*rdata)); /* Fill the ramrod data with provided parameters */ if (ECORE_TEST_BIT(ECORE_F_UPDATE_TX_SWITCH_SUSPEND_CHNG, &switch_update_params->changes)) { rdata->tx_switch_suspend_change_flg = 1; rdata->tx_switch_suspend = ECORE_TEST_BIT(ECORE_F_UPDATE_TX_SWITCH_SUSPEND, &switch_update_params->changes); } rdata->echo = SWITCH_UPDATE; return ecore_sp_post(sc, RAMROD_CMD_ID_COMMON_FUNCTION_UPDATE, 0, data_mapping, NONE_CONNECTION_TYPE); } static int ecore_func_send_afex_update(struct bnx2x_softc *sc, struct ecore_func_state_params *params) { struct ecore_func_sp_obj *o = params->f_obj; struct function_update_data *rdata = (struct function_update_data *)o->afex_rdata; ecore_dma_addr_t data_mapping = o->afex_rdata_mapping; struct ecore_func_afex_update_params *afex_update_params = ¶ms->params.afex_update; ECORE_MEMSET(rdata, 0, sizeof(*rdata)); /* Fill the ramrod data with provided parameters */ rdata->vif_id_change_flg = 1; rdata->vif_id = ECORE_CPU_TO_LE16(afex_update_params->vif_id); rdata->afex_default_vlan_change_flg = 1; rdata->afex_default_vlan = ECORE_CPU_TO_LE16(afex_update_params->afex_default_vlan); rdata->allowed_priorities_change_flg = 1; rdata->allowed_priorities = afex_update_params->allowed_priorities; rdata->echo = AFEX_UPDATE; /* No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ ECORE_MSG(sc, "afex: sending func_update vif_id 0x%x dvlan 0x%x prio 0x%x", rdata->vif_id, rdata->afex_default_vlan, rdata->allowed_priorities); return ecore_sp_post(sc, RAMROD_CMD_ID_COMMON_FUNCTION_UPDATE, 0, data_mapping, NONE_CONNECTION_TYPE); } static inline int ecore_func_send_afex_viflists(struct bnx2x_softc *sc, struct ecore_func_state_params *params) { struct ecore_func_sp_obj *o = params->f_obj; struct afex_vif_list_ramrod_data *rdata = (struct afex_vif_list_ramrod_data *)o->afex_rdata; struct ecore_func_afex_viflists_params *afex_vif_params = ¶ms->params.afex_viflists; uint64_t *p_rdata = (uint64_t *) rdata; ECORE_MEMSET(rdata, 0, sizeof(*rdata)); /* Fill the ramrod data with provided parameters */ rdata->vif_list_index = ECORE_CPU_TO_LE16(afex_vif_params->vif_list_index); rdata->func_bit_map = afex_vif_params->func_bit_map; rdata->afex_vif_list_command = afex_vif_params->afex_vif_list_command; rdata->func_to_clear = afex_vif_params->func_to_clear; /* send in echo type of sub command */ rdata->echo = afex_vif_params->afex_vif_list_command; /* No need for an explicit memory barrier here as long we would * need to ensure the ordering of writing to the SPQ element * and updating of the SPQ producer which involves a memory * read and we will have to put a full memory barrier there * (inside ecore_sp_post()). */ ECORE_MSG (sc, "afex: ramrod lists, cmd 0x%x index 0x%x func_bit_map 0x%x func_to_clr 0x%x", rdata->afex_vif_list_command, rdata->vif_list_index, rdata->func_bit_map, rdata->func_to_clear); /* this ramrod sends data directly and not through DMA mapping */ return ecore_sp_post(sc, RAMROD_CMD_ID_COMMON_AFEX_VIF_LISTS, 0, *p_rdata, NONE_CONNECTION_TYPE); } static int ecore_func_send_stop(struct bnx2x_softc *sc, __rte_unused struct ecore_func_state_params *params) { return ecore_sp_post(sc, RAMROD_CMD_ID_COMMON_FUNCTION_STOP, 0, 0, NONE_CONNECTION_TYPE); } static int ecore_func_send_tx_stop(struct bnx2x_softc *sc, __rte_unused struct ecore_func_state_params *params) { return ecore_sp_post(sc, RAMROD_CMD_ID_COMMON_STOP_TRAFFIC, 0, 0, NONE_CONNECTION_TYPE); } static int ecore_func_send_tx_start(struct bnx2x_softc *sc, struct ecore_func_state_params *params) { struct ecore_func_sp_obj *o = params->f_obj; struct flow_control_configuration *rdata = (struct flow_control_configuration *)o->rdata; ecore_dma_addr_t data_mapping = o->rdata_mapping; struct ecore_func_tx_start_params *tx_start_params = ¶ms->params.tx_start; uint32_t i; ECORE_MEMSET(rdata, 0, sizeof(*rdata)); rdata->dcb_enabled = tx_start_params->dcb_enabled; rdata->dcb_version = tx_start_params->dcb_version; rdata->dont_add_pri_0_en = tx_start_params->dont_add_pri_0_en; for (i = 0; i < ARRAY_SIZE(rdata->traffic_type_to_priority_cos); i++) rdata->traffic_type_to_priority_cos[i] = tx_start_params->traffic_type_to_priority_cos[i]; return ecore_sp_post(sc, RAMROD_CMD_ID_COMMON_START_TRAFFIC, 0, data_mapping, NONE_CONNECTION_TYPE); } static int ecore_func_send_cmd(struct bnx2x_softc *sc, struct ecore_func_state_params *params) { switch (params->cmd) { case ECORE_F_CMD_HW_INIT: return ecore_func_hw_init(sc, params); case ECORE_F_CMD_START: return ecore_func_send_start(sc, params); case ECORE_F_CMD_STOP: return ecore_func_send_stop(sc, params); case ECORE_F_CMD_HW_RESET: return ecore_func_hw_reset(sc, params); case ECORE_F_CMD_AFEX_UPDATE: return ecore_func_send_afex_update(sc, params); case ECORE_F_CMD_AFEX_VIFLISTS: return ecore_func_send_afex_viflists(sc, params); case ECORE_F_CMD_TX_STOP: return ecore_func_send_tx_stop(sc, params); case ECORE_F_CMD_TX_START: return ecore_func_send_tx_start(sc, params); case ECORE_F_CMD_SWITCH_UPDATE: return ecore_func_send_switch_update(sc, params); default: PMD_DRV_LOG(ERR, sc, "Unknown command: %d", params->cmd); return ECORE_INVAL; } } void ecore_init_func_obj(__rte_unused struct bnx2x_softc *sc, struct ecore_func_sp_obj *obj, void *rdata, ecore_dma_addr_t rdata_mapping, void *afex_rdata, ecore_dma_addr_t afex_rdata_mapping, struct ecore_func_sp_drv_ops *drv_iface) { ECORE_MEMSET(obj, 0, sizeof(*obj)); ECORE_MUTEX_INIT(&obj->one_pending_mutex); obj->rdata = rdata; obj->rdata_mapping = rdata_mapping; obj->afex_rdata = afex_rdata; obj->afex_rdata_mapping = afex_rdata_mapping; obj->send_cmd = ecore_func_send_cmd; obj->check_transition = ecore_func_chk_transition; obj->complete_cmd = ecore_func_comp_cmd; obj->wait_comp = ecore_func_wait_comp; obj->drv = drv_iface; } /** * ecore_func_state_change - perform Function state change transition * * @sc: device handle * @params: parameters to perform the transaction * * returns 0 in case of successfully completed transition, * negative error code in case of failure, positive * (EBUSY) value if there is a completion to that is * still pending (possible only if RAMROD_COMP_WAIT is * not set in params->ramrod_flags for asynchronous * commands). */ int ecore_func_state_change(struct bnx2x_softc *sc, struct ecore_func_state_params *params) { struct ecore_func_sp_obj *o = params->f_obj; int rc, cnt = 300; enum ecore_func_cmd cmd = params->cmd; unsigned long *pending = &o->pending; ECORE_MUTEX_LOCK(&o->one_pending_mutex); /* Check that the requested transition is legal */ rc = o->check_transition(sc, o, params); if ((rc == ECORE_BUSY) && (ECORE_TEST_BIT(RAMROD_RETRY, ¶ms->ramrod_flags))) { while ((rc == ECORE_BUSY) && (--cnt > 0)) { ECORE_MUTEX_UNLOCK(&o->one_pending_mutex); ECORE_MSLEEP(10); ECORE_MUTEX_LOCK(&o->one_pending_mutex); rc = o->check_transition(sc, o, params); } if (rc == ECORE_BUSY) { ECORE_MUTEX_UNLOCK(&o->one_pending_mutex); PMD_DRV_LOG(ERR, sc, "timeout waiting for previous ramrod completion"); return rc; } } else if (rc) { ECORE_MUTEX_UNLOCK(&o->one_pending_mutex); return rc; } /* Set "pending" bit */ ECORE_SET_BIT(cmd, pending); /* Don't send a command if only driver cleanup was requested */ if (ECORE_TEST_BIT(RAMROD_DRV_CLR_ONLY, ¶ms->ramrod_flags)) { ecore_func_state_change_comp(sc, o, cmd); ECORE_MUTEX_UNLOCK(&o->one_pending_mutex); } else { /* Send a ramrod */ rc = o->send_cmd(sc, params); ECORE_MUTEX_UNLOCK(&o->one_pending_mutex); if (rc) { o->next_state = ECORE_F_STATE_MAX; ECORE_CLEAR_BIT(cmd, pending); ECORE_SMP_MB_AFTER_CLEAR_BIT(); return rc; } if (ECORE_TEST_BIT(RAMROD_COMP_WAIT, ¶ms->ramrod_flags)) { rc = o->wait_comp(sc, o, cmd); if (rc) return rc; return ECORE_SUCCESS; } } return ECORE_RET_PENDING(cmd, pending); } /****************************************************************************** * Description: * Calculates crc 8 on a word value: polynomial 0-1-2-8 * Code was translated from Verilog. * Return: *****************************************************************************/ uint8_t ecore_calc_crc8(uint32_t data, uint8_t crc) { uint8_t D[32]; uint8_t NewCRC[8]; uint8_t C[8]; uint8_t crc_res; uint8_t i; /* split the data into 31 bits */ for (i = 0; i < 32; i++) { D[i] = (uint8_t) (data & 1); data = data >> 1; } /* split the crc into 8 bits */ for (i = 0; i < 8; i++) { C[i] = crc & 1; crc = crc >> 1; } NewCRC[0] = D[31] ^ D[30] ^ D[28] ^ D[23] ^ D[21] ^ D[19] ^ D[18] ^ D[16] ^ D[14] ^ D[12] ^ D[8] ^ D[7] ^ D[6] ^ D[0] ^ C[4] ^ C[6] ^ C[7]; NewCRC[1] = D[30] ^ D[29] ^ D[28] ^ D[24] ^ D[23] ^ D[22] ^ D[21] ^ D[20] ^ D[18] ^ D[17] ^ D[16] ^ D[15] ^ D[14] ^ D[13] ^ D[12] ^ D[9] ^ D[6] ^ D[1] ^ D[0] ^ C[0] ^ C[4] ^ C[5] ^ C[6]; NewCRC[2] = D[29] ^ D[28] ^ D[25] ^ D[24] ^ D[22] ^ D[17] ^ D[15] ^ D[13] ^ D[12] ^ D[10] ^ D[8] ^ D[6] ^ D[2] ^ D[1] ^ D[0] ^ C[0] ^ C[1] ^ C[4] ^ C[5]; NewCRC[3] = D[30] ^ D[29] ^ D[26] ^ D[25] ^ D[23] ^ D[18] ^ D[16] ^ D[14] ^ D[13] ^ D[11] ^ D[9] ^ D[7] ^ D[3] ^ D[2] ^ D[1] ^ C[1] ^ C[2] ^ C[5] ^ C[6]; NewCRC[4] = D[31] ^ D[30] ^ D[27] ^ D[26] ^ D[24] ^ D[19] ^ D[17] ^ D[15] ^ D[14] ^ D[12] ^ D[10] ^ D[8] ^ D[4] ^ D[3] ^ D[2] ^ C[0] ^ C[2] ^ C[3] ^ C[6] ^ C[7]; NewCRC[5] = D[31] ^ D[28] ^ D[27] ^ D[25] ^ D[20] ^ D[18] ^ D[16] ^ D[15] ^ D[13] ^ D[11] ^ D[9] ^ D[5] ^ D[4] ^ D[3] ^ C[1] ^ C[3] ^ C[4] ^ C[7]; NewCRC[6] = D[29] ^ D[28] ^ D[26] ^ D[21] ^ D[19] ^ D[17] ^ D[16] ^ D[14] ^ D[12] ^ D[10] ^ D[6] ^ D[5] ^ D[4] ^ C[2] ^ C[4] ^ C[5]; NewCRC[7] = D[30] ^ D[29] ^ D[27] ^ D[22] ^ D[20] ^ D[18] ^ D[17] ^ D[15] ^ D[13] ^ D[11] ^ D[7] ^ D[6] ^ D[5] ^ C[3] ^ C[5] ^ C[6]; crc_res = 0; for (i = 0; i < 8; i++) { crc_res |= (NewCRC[i] << i); } return crc_res; } uint32_t ecore_calc_crc32(uint32_t crc, uint8_t const *p, uint32_t len, uint32_t magic) { int i; while (len--) { crc ^= *p++; for (i = 0; i < 8; i++) crc = (crc >> 1) ^ ((crc & 1) ? magic : 0); } return crc; }