diff --git a/sys/dev/mlx5/mlx5_en/en.h b/sys/dev/mlx5/mlx5_en/en.h index b990bb754962..2ec47fa96a6d 100644 --- a/sys/dev/mlx5/mlx5_en/en.h +++ b/sys/dev/mlx5/mlx5_en/en.h @@ -1,1193 +1,1192 @@ /*- * Copyright (c) 2015-2019 Mellanox Technologies. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS `AS IS' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * $FreeBSD$ */ #ifndef _MLX5_EN_H_ #define _MLX5_EN_H_ #include #include #include #include #include #include #include #include -#include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include "opt_rss.h" #ifdef RSS #include #include #endif #include #include #include #include #include #include #include #include #include #include #define MLX5E_MAX_PRIORITY 8 #define MLX5E_MAX_FEC_10X_25X 4 #define MLX5E_MAX_FEC_50X 4 /* IEEE 802.1Qaz standard supported values */ #define IEEE_8021QAZ_MAX_TCS 8 #define MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE 0x7 #define MLX5E_PARAMS_DEFAULT_LOG_SQ_SIZE 0xa #define MLX5E_PARAMS_MAXIMUM_LOG_SQ_SIZE 0xe #define MLX5E_PARAMS_MINIMUM_LOG_RQ_SIZE 0x7 #define MLX5E_PARAMS_DEFAULT_LOG_RQ_SIZE 0xa #define MLX5E_PARAMS_MAXIMUM_LOG_RQ_SIZE 0xe #define MLX5E_MAX_BUSDMA_RX_SEGS 15 #ifndef MLX5E_MAX_RX_BYTES #define MLX5E_MAX_RX_BYTES MCLBYTES #endif #define MLX5E_PARAMS_DEFAULT_LRO_WQE_SZ \ MIN(65535, 7 * MLX5E_MAX_RX_BYTES) #define MLX5E_DIM_DEFAULT_PROFILE 3 #define MLX5E_DIM_MAX_RX_CQ_MODERATION_PKTS_WITH_LRO 16 #define MLX5E_PARAMS_DEFAULT_RX_CQ_MODERATION_USEC 0x10 #define MLX5E_PARAMS_DEFAULT_RX_CQ_MODERATION_USEC_FROM_CQE 0x3 #define MLX5E_PARAMS_DEFAULT_RX_CQ_MODERATION_PKTS 0x20 #define MLX5E_PARAMS_DEFAULT_TX_CQ_MODERATION_USEC 0x10 #define MLX5E_PARAMS_DEFAULT_TX_CQ_MODERATION_PKTS 0x20 #define MLX5E_PARAMS_DEFAULT_MIN_RX_WQES 0x80 #define MLX5E_PARAMS_DEFAULT_RX_HASH_LOG_TBL_SZ 0x7 #define MLX5E_CACHELINE_SIZE CACHE_LINE_SIZE #define MLX5E_HW2SW_MTU(hwmtu) \ ((hwmtu) - (ETHER_HDR_LEN + ETHER_VLAN_ENCAP_LEN + ETHER_CRC_LEN)) #define MLX5E_SW2HW_MTU(swmtu) \ ((swmtu) + (ETHER_HDR_LEN + ETHER_VLAN_ENCAP_LEN + ETHER_CRC_LEN)) #define MLX5E_SW2MB_MTU(swmtu) \ (MLX5E_SW2HW_MTU(swmtu) + MLX5E_NET_IP_ALIGN) #define MLX5E_MTU_MIN 72 /* Min MTU allowed by the kernel */ #define MLX5E_MTU_MAX MIN(ETHERMTU_JUMBO, MJUM16BYTES) /* Max MTU of Ethernet * jumbo frames */ #define MLX5E_BUDGET_MAX 8192 /* RX and TX */ #define MLX5E_RX_BUDGET_MAX 256 #define MLX5E_SQ_BF_BUDGET 16 #define MLX5E_SQ_TX_QUEUE_SIZE 4096 /* SQ drbr queue size */ #define MLX5E_MAX_TX_NUM_TC 8 /* units */ #define MLX5E_MAX_TX_HEADER 128 /* bytes */ #define MLX5E_MAX_TX_PAYLOAD_SIZE 65536 /* bytes */ #define MLX5E_MAX_TX_MBUF_SIZE 65536 /* bytes */ #define MLX5E_MAX_TX_MBUF_FRAGS \ ((MLX5_SEND_WQE_MAX_WQEBBS * MLX5_SEND_WQEBB_NUM_DS) - \ (MLX5E_MAX_TX_HEADER / MLX5_SEND_WQE_DS) - \ 1 /* the maximum value of the DS counter is 0x3F and not 0x40 */) /* units */ #define MLX5E_MAX_TX_INLINE \ (MLX5E_MAX_TX_HEADER - sizeof(struct mlx5e_tx_wqe) + \ sizeof(((struct mlx5e_tx_wqe *)0)->eth.inline_hdr_start)) /* bytes */ #define MLX5E_100MB (100000) #define MLX5E_1GB (1000000) #define MLX5E_ZERO(ptr, field) \ memset(&(ptr)->field, 0, \ sizeof(*(ptr)) - __offsetof(__typeof(*(ptr)), field)) MALLOC_DECLARE(M_MLX5EN); struct mlx5_core_dev; struct mlx5e_cq; typedef void (mlx5e_cq_comp_t)(struct mlx5_core_cq *, struct mlx5_eqe *); #define mlx5_en_err(_dev, format, ...) \ if_printf(_dev, "ERR: ""%s:%d:(pid %d): " format, \ __func__, __LINE__, curthread->td_proc->p_pid, \ ##__VA_ARGS__) #define mlx5_en_warn(_dev, format, ...) \ if_printf(_dev, "WARN: ""%s:%d:(pid %d): " format, \ __func__, __LINE__, curthread->td_proc->p_pid, \ ##__VA_ARGS__) #define mlx5_en_info(_dev, format, ...) \ if_printf(_dev, "INFO: ""%s:%d:(pid %d): " format, \ __func__, __LINE__, curthread->td_proc->p_pid, \ ##__VA_ARGS__) #define MLX5E_STATS_COUNT(a, ...) a #define MLX5E_STATS_VAR(a, b, c, ...) b c; #define MLX5E_STATS_COUNTER(a, b, c, ...) counter_##b##_t c; #define MLX5E_STATS_DESC(a, b, c, d, e, ...) d, e, #define MLX5E_VPORT_STATS(m) \ /* HW counters */ \ m(+1, u64, rx_packets, "rx_packets", "Received packets") \ m(+1, u64, rx_bytes, "rx_bytes", "Received bytes") \ m(+1, u64, tx_packets, "tx_packets", "Transmitted packets") \ m(+1, u64, tx_bytes, "tx_bytes", "Transmitted bytes") \ m(+1, u64, rx_error_packets, "rx_error_packets", "Received error packets") \ m(+1, u64, rx_error_bytes, "rx_error_bytes", "Received error bytes") \ m(+1, u64, tx_error_packets, "tx_error_packets", "Transmitted error packets") \ m(+1, u64, tx_error_bytes, "tx_error_bytes", "Transmitted error bytes") \ m(+1, u64, rx_unicast_packets, "rx_unicast_packets", "Received unicast packets") \ m(+1, u64, rx_unicast_bytes, "rx_unicast_bytes", "Received unicast bytes") \ m(+1, u64, tx_unicast_packets, "tx_unicast_packets", "Transmitted unicast packets") \ m(+1, u64, tx_unicast_bytes, "tx_unicast_bytes", "Transmitted unicast bytes") \ m(+1, u64, rx_multicast_packets, "rx_multicast_packets", "Received multicast packets") \ m(+1, u64, rx_multicast_bytes, "rx_multicast_bytes", "Received multicast bytes") \ m(+1, u64, tx_multicast_packets, "tx_multicast_packets", "Transmitted multicast packets") \ m(+1, u64, tx_multicast_bytes, "tx_multicast_bytes", "Transmitted multicast bytes") \ m(+1, u64, rx_broadcast_packets, "rx_broadcast_packets", "Received broadcast packets") \ m(+1, u64, rx_broadcast_bytes, "rx_broadcast_bytes", "Received broadcast bytes") \ m(+1, u64, tx_broadcast_packets, "tx_broadcast_packets", "Transmitted broadcast packets") \ m(+1, u64, tx_broadcast_bytes, "tx_broadcast_bytes", "Transmitted broadcast bytes") \ m(+1, u64, rx_out_of_buffer, "rx_out_of_buffer", "Receive out of buffer, no recv wqes events") \ /* SW counters */ \ m(+1, u64, tso_packets, "tso_packets", "Transmitted TSO packets") \ m(+1, u64, tso_bytes, "tso_bytes", "Transmitted TSO bytes") \ m(+1, u64, lro_packets, "lro_packets", "Received LRO packets") \ m(+1, u64, lro_bytes, "lro_bytes", "Received LRO bytes") \ m(+1, u64, sw_lro_queued, "sw_lro_queued", "Packets queued for SW LRO") \ m(+1, u64, sw_lro_flushed, "sw_lro_flushed", "Packets flushed from SW LRO") \ m(+1, u64, rx_csum_good, "rx_csum_good", "Received checksum valid packets") \ m(+1, u64, rx_csum_none, "rx_csum_none", "Received no checksum packets") \ m(+1, u64, tx_csum_offload, "tx_csum_offload", "Transmit checksum offload packets") \ m(+1, u64, tx_queue_dropped, "tx_queue_dropped", "Transmit queue dropped") \ m(+1, u64, tx_defragged, "tx_defragged", "Transmit queue defragged") \ m(+1, u64, rx_wqe_err, "rx_wqe_err", "Receive WQE errors") \ m(+1, u64, tx_jumbo_packets, "tx_jumbo_packets", "TX packets greater than 1518 octets") \ m(+1, u64, rx_steer_missed_packets, "rx_steer_missed_packets", "RX packets dropped by steering rule(s)") #define MLX5E_VPORT_STATS_NUM (0 MLX5E_VPORT_STATS(MLX5E_STATS_COUNT)) struct mlx5e_vport_stats { struct sysctl_ctx_list ctx; u64 arg [0]; MLX5E_VPORT_STATS(MLX5E_STATS_VAR) }; #define MLX5E_PPORT_IEEE802_3_STATS(m) \ m(+1, u64, frames_tx, "frames_tx", "Frames transmitted") \ m(+1, u64, frames_rx, "frames_rx", "Frames received") \ m(+1, u64, check_seq_err, "check_seq_err", "Sequence errors") \ m(+1, u64, alignment_err, "alignment_err", "Alignment errors") \ m(+1, u64, octets_tx, "octets_tx", "Bytes transmitted") \ m(+1, u64, octets_received, "octets_received", "Bytes received") \ m(+1, u64, multicast_xmitted, "multicast_xmitted", "Multicast transmitted") \ m(+1, u64, broadcast_xmitted, "broadcast_xmitted", "Broadcast transmitted") \ m(+1, u64, multicast_rx, "multicast_rx", "Multicast received") \ m(+1, u64, broadcast_rx, "broadcast_rx", "Broadcast received") \ m(+1, u64, in_range_len_errors, "in_range_len_errors", "In range length errors") \ m(+1, u64, out_of_range_len, "out_of_range_len", "Out of range length errors") \ m(+1, u64, too_long_errors, "too_long_errors", "Too long errors") \ m(+1, u64, symbol_err, "symbol_err", "Symbol errors") \ m(+1, u64, mac_control_tx, "mac_control_tx", "MAC control transmitted") \ m(+1, u64, mac_control_rx, "mac_control_rx", "MAC control received") \ m(+1, u64, unsupported_op_rx, "unsupported_op_rx", "Unsupported operation received") \ m(+1, u64, pause_ctrl_rx, "pause_ctrl_rx", "Pause control received") \ m(+1, u64, pause_ctrl_tx, "pause_ctrl_tx", "Pause control transmitted") #define MLX5E_PPORT_RFC2819_STATS(m) \ m(+1, u64, drop_events, "drop_events", "Dropped events") \ m(+1, u64, octets, "octets", "Octets") \ m(+1, u64, pkts, "pkts", "Packets") \ m(+1, u64, broadcast_pkts, "broadcast_pkts", "Broadcast packets") \ m(+1, u64, multicast_pkts, "multicast_pkts", "Multicast packets") \ m(+1, u64, crc_align_errors, "crc_align_errors", "CRC alignment errors") \ m(+1, u64, undersize_pkts, "undersize_pkts", "Undersized packets") \ m(+1, u64, oversize_pkts, "oversize_pkts", "Oversized packets") \ m(+1, u64, fragments, "fragments", "Fragments") \ m(+1, u64, jabbers, "jabbers", "Jabbers") \ m(+1, u64, collisions, "collisions", "Collisions") #define MLX5E_PPORT_RFC2819_STATS_DEBUG(m) \ m(+1, u64, p64octets, "p64octets", "Bytes") \ m(+1, u64, p65to127octets, "p65to127octets", "Bytes") \ m(+1, u64, p128to255octets, "p128to255octets", "Bytes") \ m(+1, u64, p256to511octets, "p256to511octets", "Bytes") \ m(+1, u64, p512to1023octets, "p512to1023octets", "Bytes") \ m(+1, u64, p1024to1518octets, "p1024to1518octets", "Bytes") \ m(+1, u64, p1519to2047octets, "p1519to2047octets", "Bytes") \ m(+1, u64, p2048to4095octets, "p2048to4095octets", "Bytes") \ m(+1, u64, p4096to8191octets, "p4096to8191octets", "Bytes") \ m(+1, u64, p8192to10239octets, "p8192to10239octets", "Bytes") #define MLX5E_PPORT_RFC2863_STATS_DEBUG(m) \ m(+1, u64, in_octets, "in_octets", "In octets") \ m(+1, u64, in_ucast_pkts, "in_ucast_pkts", "In unicast packets") \ m(+1, u64, in_discards, "in_discards", "In discards") \ m(+1, u64, in_errors, "in_errors", "In errors") \ m(+1, u64, in_unknown_protos, "in_unknown_protos", "In unknown protocols") \ m(+1, u64, out_octets, "out_octets", "Out octets") \ m(+1, u64, out_ucast_pkts, "out_ucast_pkts", "Out unicast packets") \ m(+1, u64, out_discards, "out_discards", "Out discards") \ m(+1, u64, out_errors, "out_errors", "Out errors") \ m(+1, u64, in_multicast_pkts, "in_multicast_pkts", "In multicast packets") \ m(+1, u64, in_broadcast_pkts, "in_broadcast_pkts", "In broadcast packets") \ m(+1, u64, out_multicast_pkts, "out_multicast_pkts", "Out multicast packets") \ m(+1, u64, out_broadcast_pkts, "out_broadcast_pkts", "Out broadcast packets") #define MLX5E_PPORT_ETHERNET_EXTENDED_STATS_DEBUG(m) \ m(+1, u64, port_transmit_wait, "port_transmit_wait", "Port transmit wait") \ m(+1, u64, ecn_marked, "ecn_marked", "ECN marked") \ m(+1, u64, no_buffer_discard_mc, "no_buffer_discard_mc", "No buffer discard mc") \ m(+1, u64, rx_ebp, "rx_ebp", "RX EBP") \ m(+1, u64, tx_ebp, "tx_ebp", "TX EBP") \ m(+1, u64, rx_buffer_almost_full, "rx_buffer_almost_full", "RX buffer almost full") \ m(+1, u64, rx_buffer_full, "rx_buffer_full", "RX buffer full") \ m(+1, u64, rx_icrc_encapsulated, "rx_icrc_encapsulated", "RX ICRC encapsulated") \ m(+1, u64, ex_reserved_0, "ex_reserved_0", "Reserved") \ m(+1, u64, ex_reserved_1, "ex_reserved_1", "Reserved") \ m(+1, u64, tx_stat_p64octets, "tx_stat_p64octets", "Bytes") \ m(+1, u64, tx_stat_p65to127octets, "tx_stat_p65to127octets", "Bytes") \ m(+1, u64, tx_stat_p128to255octets, "tx_stat_p128to255octets", "Bytes") \ m(+1, u64, tx_stat_p256to511octets, "tx_stat_p256to511octets", "Bytes") \ m(+1, u64, tx_stat_p512to1023octets, "tx_stat_p512to1023octets", "Bytes") \ m(+1, u64, tx_stat_p1024to1518octets, "tx_stat_p1024to1518octets", "Bytes") \ m(+1, u64, tx_stat_p1519to2047octets, "tx_stat_p1519to2047octets", "Bytes") \ m(+1, u64, tx_stat_p2048to4095octets, "tx_stat_p2048to4095octets", "Bytes") \ m(+1, u64, tx_stat_p4096to8191octets, "tx_stat_p4096to8191octets", "Bytes") \ m(+1, u64, tx_stat_p8192to10239octets, "tx_stat_p8192to10239octets", "Bytes") #define MLX5E_PPORT_STATISTICAL_DEBUG(m) \ m(+1, u64, phy_time_since_last_clear, "phy_time_since_last_clear", \ "Time since last clear in milliseconds") \ m(+1, u64, phy_received_bits, "phy_received_bits", \ "Total amount of traffic received in bits before error correction") \ m(+1, u64, phy_symbol_errors, "phy_symbol_errors", \ "Total number of symbol errors before error correction") \ m(+1, u64, phy_corrected_bits, "phy_corrected_bits", \ "Total number of corrected bits ") \ m(+1, u64, phy_corrected_bits_lane0, "phy_corrected_bits_lane0", \ "Total number of corrected bits for lane 0") \ m(+1, u64, phy_corrected_bits_lane1, "phy_corrected_bits_lane1", \ "Total number of corrected bits for lane 1") \ m(+1, u64, phy_corrected_bits_lane2, "phy_corrected_bits_lane2", \ "Total number of corrected bits for lane 2") \ m(+1, u64, phy_corrected_bits_lane3, "phy_corrected_bits_lane3", \ "Total number of corrected bits for lane 3") #define MLX5E_PPORT_PHYSICAL_LAYER_STATS_DEBUG(m) \ m(+1, u64, time_since_last_clear, "time_since_last_clear", \ "Time since the last counters clear event (msec)") \ m(+1, u64, symbol_errors, "symbol_errors", "Symbol errors") \ m(+1, u64, sync_headers_errors, "sync_headers_errors", \ "Sync header error counter") \ m(+1, u64, bip_errors_lane0, "edpl_bip_errors_lane0", \ "Indicates the number of PRBS errors on lane 0") \ m(+1, u64, bip_errors_lane1, "edpl_bip_errors_lane1", \ "Indicates the number of PRBS errors on lane 1") \ m(+1, u64, bip_errors_lane2, "edpl_bip_errors_lane2", \ "Indicates the number of PRBS errors on lane 2") \ m(+1, u64, bip_errors_lane3, "edpl_bip_errors_lane3", \ "Indicates the number of PRBS errors on lane 3") \ m(+1, u64, fc_corrected_blocks_lane0, "fc_corrected_blocks_lane0", \ "FEC correctable block counter lane 0") \ m(+1, u64, fc_corrected_blocks_lane1, "fc_corrected_blocks_lane1", \ "FEC correctable block counter lane 1") \ m(+1, u64, fc_corrected_blocks_lane2, "fc_corrected_blocks_lane2", \ "FEC correctable block counter lane 2") \ m(+1, u64, fc_corrected_blocks_lane3, "fc_corrected_blocks_lane3", \ "FEC correctable block counter lane 3") \ m(+1, u64, rs_corrected_blocks, "rs_corrected_blocks", \ "FEC correcable block counter") \ m(+1, u64, rs_uncorrectable_blocks, "rs_uncorrectable_blocks", \ "FEC uncorrecable block counter") \ m(+1, u64, rs_no_errors_blocks, "rs_no_errors_blocks", \ "The number of RS-FEC blocks received that had no errors") \ m(+1, u64, rs_single_error_blocks, "rs_single_error_blocks", \ "The number of corrected RS-FEC blocks received that had" \ "exactly 1 error symbol") \ m(+1, u64, rs_corrected_symbols_total, "rs_corrected_symbols_total", \ "Port FEC corrected symbol counter") \ m(+1, u64, rs_corrected_symbols_lane0, "rs_corrected_symbols_lane0", \ "FEC corrected symbol counter lane 0") \ m(+1, u64, rs_corrected_symbols_lane1, "rs_corrected_symbols_lane1", \ "FEC corrected symbol counter lane 1") \ m(+1, u64, rs_corrected_symbols_lane2, "rs_corrected_symbols_lane2", \ "FEC corrected symbol counter lane 2") \ m(+1, u64, rs_corrected_symbols_lane3, "rs_corrected_symbols_lane3", \ "FEC corrected symbol counter lane 3") /* Per priority statistics for PFC */ #define MLX5E_PPORT_PER_PRIO_STATS_SUB(m,n,p) \ m(n, p, +1, u64, rx_octets, "rx_octets", "Received octets") \ m(n, p, +1, u64, rx_uc_frames, "rx_uc_frames", "Received unicast frames") \ m(n, p, +1, u64, rx_mc_frames, "rx_mc_frames", "Received multicast frames") \ m(n, p, +1, u64, rx_bc_frames, "rx_bc_frames", "Received broadcast frames") \ m(n, p, +1, u64, rx_frames, "rx_frames", "Received frames") \ m(n, p, +1, u64, tx_octets, "tx_octets", "Transmitted octets") \ m(n, p, +1, u64, tx_uc_frames, "tx_uc_frames", "Transmitted unicast frames") \ m(n, p, +1, u64, tx_mc_frames, "tx_mc_frames", "Transmitted multicast frames") \ m(n, p, +1, u64, tx_bc_frames, "tx_bc_frames", "Transmitted broadcast frames") \ m(n, p, +1, u64, tx_frames, "tx_frames", "Transmitted frames") \ m(n, p, +1, u64, rx_pause, "rx_pause", "Received pause frames") \ m(n, p, +1, u64, rx_pause_duration, "rx_pause_duration", \ "Received pause duration") \ m(n, p, +1, u64, tx_pause, "tx_pause", "Transmitted pause frames") \ m(n, p, +1, u64, tx_pause_duration, "tx_pause_duration", \ "Transmitted pause duration") \ m(n, p, +1, u64, rx_pause_transition, "rx_pause_transition", \ "Received pause transitions") \ m(n, p, +1, u64, rx_discards, "rx_discards", "Discarded received frames") \ m(n, p, +1, u64, device_stall_minor_watermark, \ "device_stall_minor_watermark", "Device stall minor watermark") \ m(n, p, +1, u64, device_stall_critical_watermark, \ "device_stall_critical_watermark", "Device stall critical watermark") #define MLX5E_PPORT_PER_PRIO_STATS_PREFIX(m,p,c,t,f,s,d) \ m(c, t, pri_##p##_##f, "prio" #p "_" s, "Priority " #p " - " d) #define MLX5E_PPORT_PER_PRIO_STATS_NUM_PRIO 8 #define MLX5E_PPORT_PER_PRIO_STATS(m) \ MLX5E_PPORT_PER_PRIO_STATS_SUB(MLX5E_PPORT_PER_PRIO_STATS_PREFIX,m,0) \ MLX5E_PPORT_PER_PRIO_STATS_SUB(MLX5E_PPORT_PER_PRIO_STATS_PREFIX,m,1) \ MLX5E_PPORT_PER_PRIO_STATS_SUB(MLX5E_PPORT_PER_PRIO_STATS_PREFIX,m,2) \ MLX5E_PPORT_PER_PRIO_STATS_SUB(MLX5E_PPORT_PER_PRIO_STATS_PREFIX,m,3) \ MLX5E_PPORT_PER_PRIO_STATS_SUB(MLX5E_PPORT_PER_PRIO_STATS_PREFIX,m,4) \ MLX5E_PPORT_PER_PRIO_STATS_SUB(MLX5E_PPORT_PER_PRIO_STATS_PREFIX,m,5) \ MLX5E_PPORT_PER_PRIO_STATS_SUB(MLX5E_PPORT_PER_PRIO_STATS_PREFIX,m,6) \ MLX5E_PPORT_PER_PRIO_STATS_SUB(MLX5E_PPORT_PER_PRIO_STATS_PREFIX,m,7) #define MLX5E_PCIE_PERFORMANCE_COUNTERS_64(m) \ m(+1, u64, life_time_counter_high, "life_time_counter", \ "Life time counter.", pcie_perf_counters) \ m(+1, u64, tx_overflow_buffer_pkt, "tx_overflow_buffer_pkt", \ "The number of packets dropped due to lack of PCIe buffers " \ "in receive path from NIC port toward the hosts.", \ pcie_perf_counters) \ m(+1, u64, tx_overflow_buffer_marked_pkt, \ "tx_overflow_buffer_marked_pkt", \ "The number of packets marked due to lack of PCIe buffers " \ "in receive path from NIC port toward the hosts.", \ pcie_perf_counters) #define MLX5E_PCIE_PERFORMANCE_COUNTERS_32(m) \ m(+1, u64, rx_errors, "rx_errors", \ "Number of transitions to recovery due to Framing " \ "errors and CRC errors.", pcie_perf_counters) \ m(+1, u64, tx_errors, "tx_errors", "Number of transitions " \ "to recovery due to EIEOS and TS errors.", pcie_perf_counters) \ m(+1, u64, l0_to_recovery_eieos, "l0_to_recovery_eieos", "Number of " \ "transitions to recovery due to getting EIEOS.", pcie_perf_counters)\ m(+1, u64, l0_to_recovery_ts, "l0_to_recovery_ts", "Number of " \ "transitions to recovery due to getting TS.", pcie_perf_counters) \ m(+1, u64, l0_to_recovery_framing, "l0_to_recovery_framing", "Number "\ "of transitions to recovery due to identifying framing " \ "errors at gen3/4.", pcie_perf_counters) \ m(+1, u64, l0_to_recovery_retrain, "l0_to_recovery_retrain", \ "Number of transitions to recovery due to link retrain request " \ "from data link.", pcie_perf_counters) \ m(+1, u64, crc_error_dllp, "crc_error_dllp", "Number of transitions " \ "to recovery due to identifying CRC DLLP errors.", \ pcie_perf_counters) \ m(+1, u64, crc_error_tlp, "crc_error_tlp", "Number of transitions to "\ "recovery due to identifying CRC TLP errors.", pcie_perf_counters) \ m(+1, u64, outbound_stalled_reads, "outbound_stalled_reads", \ "The percentage of time within the last second that the NIC had " \ "outbound non-posted read requests but could not perform the " \ "operation due to insufficient non-posted credits.", \ pcie_perf_counters) \ m(+1, u64, outbound_stalled_writes, "outbound_stalled_writes", \ "The percentage of time within the last second that the NIC had " \ "outbound posted writes requests but could not perform the " \ "operation due to insufficient posted credits.", \ pcie_perf_counters) \ m(+1, u64, outbound_stalled_reads_events, \ "outbound_stalled_reads_events", "The number of events where " \ "outbound_stalled_reads was above a threshold.", \ pcie_perf_counters) \ m(+1, u64, outbound_stalled_writes_events, \ "outbound_stalled_writes_events", \ "The number of events where outbound_stalled_writes was above " \ "a threshold.", pcie_perf_counters) #define MLX5E_PCIE_TIMERS_AND_STATES_COUNTERS_32(m) \ m(+1, u64, time_to_boot_image_start, "time_to_boot_image_start", \ "Time from start until FW boot image starts running in usec.", \ pcie_timers_states) \ m(+1, u64, time_to_link_image, "time_to_link_image", \ "Time from start until FW pci_link image starts running in usec.", \ pcie_timers_states) \ m(+1, u64, calibration_time, "calibration_time", \ "Time it took FW to do calibration in usec.", \ pcie_timers_states) \ m(+1, u64, time_to_first_perst, "time_to_first_perst", \ "Time form start until FW handle first perst. in usec.", \ pcie_timers_states) \ m(+1, u64, time_to_detect_state, "time_to_detect_state", \ "Time from start until first transition to LTSSM.Detect_Q in usec", \ pcie_timers_states) \ m(+1, u64, time_to_l0, "time_to_l0", \ "Time from start until first transition to LTSSM.L0 in usec", \ pcie_timers_states) \ m(+1, u64, time_to_crs_en, "time_to_crs_en", \ "Time from start until crs is enabled in usec", \ pcie_timers_states) \ m(+1, u64, time_to_plastic_image_start, "time_to_plastic_image_start",\ "Time form start until FW plastic image starts running in usec.", \ pcie_timers_states) \ m(+1, u64, time_to_iron_image_start, "time_to_iron_image_start", \ "Time form start until FW iron image starts running in usec.", \ pcie_timers_states) \ m(+1, u64, perst_handler, "perst_handler", \ "Number of persts arrived.", pcie_timers_states) \ m(+1, u64, times_in_l1, "times_in_l1", \ "Number of times LTSSM entered L1 flow.", pcie_timers_states) \ m(+1, u64, times_in_l23, "times_in_l23", \ "Number of times LTSSM entered L23 flow.", pcie_timers_states) \ m(+1, u64, dl_down, "dl_down", \ "Number of moves for DL_active to DL_down.", pcie_timers_states) \ m(+1, u64, config_cycle1usec, "config_cycle1usec", \ "Number of configuration requests that firmware " \ "handled in less than 1 usec.", pcie_timers_states) \ m(+1, u64, config_cycle2to7usec, "config_cycle2to7usec", \ "Number of configuration requests that firmware " \ "handled within 2 to 7 usec.", pcie_timers_states) \ m(+1, u64, config_cycle8to15usec, "config_cycle8to15usec", \ "Number of configuration requests that firmware " \ "handled within 8 to 15 usec.", pcie_timers_states) \ m(+1, u64, config_cycle16to63usec, "config_cycle16to63usec", \ "Number of configuration requests that firmware " \ "handled within 16 to 63 usec.", pcie_timers_states) \ m(+1, u64, config_cycle64usec, "config_cycle64usec", \ "Number of configuration requests that firmware " \ "handled took more than 64 usec.", pcie_timers_states) \ m(+1, u64, correctable_err_msg_sent, "correctable_err_msg_sent", \ "Number of correctable error messages sent.", pcie_timers_states) \ m(+1, u64, non_fatal_err_msg_sent, "non_fatal_err_msg_sent", \ "Number of non-Fatal error msg sent.", pcie_timers_states) \ m(+1, u64, fatal_err_msg_sent, "fatal_err_msg_sent", \ "Number of fatal error msg sent.", pcie_timers_states) #define MLX5E_PCIE_LANE_COUNTERS_32(m) \ m(+1, u64, error_counter_lane0, "error_counter_lane0", \ "Error counter for PCI lane 0", pcie_lanes_counters) \ m(+1, u64, error_counter_lane1, "error_counter_lane1", \ "Error counter for PCI lane 1", pcie_lanes_counters) \ m(+1, u64, error_counter_lane2, "error_counter_lane2", \ "Error counter for PCI lane 2", pcie_lanes_counters) \ m(+1, u64, error_counter_lane3, "error_counter_lane3", \ "Error counter for PCI lane 3", pcie_lanes_counters) \ m(+1, u64, error_counter_lane4, "error_counter_lane4", \ "Error counter for PCI lane 4", pcie_lanes_counters) \ m(+1, u64, error_counter_lane5, "error_counter_lane5", \ "Error counter for PCI lane 5", pcie_lanes_counters) \ m(+1, u64, error_counter_lane6, "error_counter_lane6", \ "Error counter for PCI lane 6", pcie_lanes_counters) \ m(+1, u64, error_counter_lane7, "error_counter_lane7", \ "Error counter for PCI lane 7", pcie_lanes_counters) \ m(+1, u64, error_counter_lane8, "error_counter_lane8", \ "Error counter for PCI lane 8", pcie_lanes_counters) \ m(+1, u64, error_counter_lane9, "error_counter_lane9", \ "Error counter for PCI lane 9", pcie_lanes_counters) \ m(+1, u64, error_counter_lane10, "error_counter_lane10", \ "Error counter for PCI lane 10", pcie_lanes_counters) \ m(+1, u64, error_counter_lane11, "error_counter_lane11", \ "Error counter for PCI lane 11", pcie_lanes_counters) \ m(+1, u64, error_counter_lane12, "error_counter_lane12", \ "Error counter for PCI lane 12", pcie_lanes_counters) \ m(+1, u64, error_counter_lane13, "error_counter_lane13", \ "Error counter for PCI lane 13", pcie_lanes_counters) \ m(+1, u64, error_counter_lane14, "error_counter_lane14", \ "Error counter for PCI lane 14", pcie_lanes_counters) \ m(+1, u64, error_counter_lane15, "error_counter_lane15", \ "Error counter for PCI lane 15", pcie_lanes_counters) /* * Make sure to update mlx5e_update_pport_counters() * when adding a new MLX5E_PPORT_STATS block */ #define MLX5E_PPORT_STATS(m) \ MLX5E_PPORT_PER_PRIO_STATS(m) \ MLX5E_PPORT_IEEE802_3_STATS(m) \ MLX5E_PPORT_RFC2819_STATS(m) #define MLX5E_PORT_STATS_DEBUG(m) \ MLX5E_PPORT_RFC2819_STATS_DEBUG(m) \ MLX5E_PPORT_RFC2863_STATS_DEBUG(m) \ MLX5E_PPORT_PHYSICAL_LAYER_STATS_DEBUG(m) \ MLX5E_PPORT_ETHERNET_EXTENDED_STATS_DEBUG(m) \ MLX5E_PPORT_STATISTICAL_DEBUG(m) \ MLX5E_PCIE_PERFORMANCE_COUNTERS_64(m) \ MLX5E_PCIE_PERFORMANCE_COUNTERS_32(m) \ MLX5E_PCIE_TIMERS_AND_STATES_COUNTERS_32(m) \ MLX5E_PCIE_LANE_COUNTERS_32(m) #define MLX5E_PPORT_IEEE802_3_STATS_NUM \ (0 MLX5E_PPORT_IEEE802_3_STATS(MLX5E_STATS_COUNT)) #define MLX5E_PPORT_RFC2819_STATS_NUM \ (0 MLX5E_PPORT_RFC2819_STATS(MLX5E_STATS_COUNT)) #define MLX5E_PPORT_STATS_NUM \ (0 MLX5E_PPORT_STATS(MLX5E_STATS_COUNT)) #define MLX5E_PPORT_PER_PRIO_STATS_NUM \ (0 MLX5E_PPORT_PER_PRIO_STATS(MLX5E_STATS_COUNT)) #define MLX5E_PPORT_RFC2819_STATS_DEBUG_NUM \ (0 MLX5E_PPORT_RFC2819_STATS_DEBUG(MLX5E_STATS_COUNT)) #define MLX5E_PPORT_RFC2863_STATS_DEBUG_NUM \ (0 MLX5E_PPORT_RFC2863_STATS_DEBUG(MLX5E_STATS_COUNT)) #define MLX5E_PPORT_PHYSICAL_LAYER_STATS_DEBUG_NUM \ (0 MLX5E_PPORT_PHYSICAL_LAYER_STATS_DEBUG(MLX5E_STATS_COUNT)) #define MLX5E_PPORT_ETHERNET_EXTENDED_STATS_DEBUG_NUM \ (0 MLX5E_PPORT_ETHERNET_EXTENDED_STATS_DEBUG(MLX5E_STATS_COUNT)) #define MLX5E_PPORT_STATISTICAL_DEBUG_NUM \ (0 MLX5E_PPORT_STATISTICAL_DEBUG(MLX5E_STATS_COUNT)) #define MLX5E_PORT_STATS_DEBUG_NUM \ (0 MLX5E_PORT_STATS_DEBUG(MLX5E_STATS_COUNT)) struct mlx5e_pport_stats { struct sysctl_ctx_list ctx; u64 arg [0]; MLX5E_PPORT_STATS(MLX5E_STATS_VAR) }; struct mlx5e_port_stats_debug { struct sysctl_ctx_list ctx; u64 arg [0]; MLX5E_PORT_STATS_DEBUG(MLX5E_STATS_VAR) }; #define MLX5E_RQ_STATS(m) \ m(+1, u64, packets, "packets", "Received packets") \ m(+1, u64, bytes, "bytes", "Received bytes") \ m(+1, u64, csum_none, "csum_none", "Received packets") \ m(+1, u64, lro_packets, "lro_packets", "Received LRO packets") \ m(+1, u64, lro_bytes, "lro_bytes", "Received LRO bytes") \ m(+1, u64, sw_lro_queued, "sw_lro_queued", "Packets queued for SW LRO") \ m(+1, u64, sw_lro_flushed, "sw_lro_flushed", "Packets flushed from SW LRO") \ m(+1, u64, wqe_err, "wqe_err", "Received packets") #define MLX5E_RQ_STATS_NUM (0 MLX5E_RQ_STATS(MLX5E_STATS_COUNT)) struct mlx5e_rq_stats { struct sysctl_ctx_list ctx; u64 arg [0]; MLX5E_RQ_STATS(MLX5E_STATS_VAR) }; #define MLX5E_SQ_STATS(m) \ m(+1, u64, packets, "packets", "Transmitted packets") \ m(+1, u64, bytes, "bytes", "Transmitted bytes") \ m(+1, u64, tso_packets, "tso_packets", "Transmitted packets") \ m(+1, u64, tso_bytes, "tso_bytes", "Transmitted bytes") \ m(+1, u64, csum_offload_none, "csum_offload_none", "Transmitted packets") \ m(+1, u64, defragged, "defragged", "Transmitted packets") \ m(+1, u64, dropped, "dropped", "Transmitted packets") \ m(+1, u64, enobuf, "enobuf", "Transmitted packets") \ m(+1, u64, nop, "nop", "Transmitted packets") #define MLX5E_SQ_STATS_NUM (0 MLX5E_SQ_STATS(MLX5E_STATS_COUNT)) struct mlx5e_sq_stats { struct sysctl_ctx_list ctx; u64 arg [0]; MLX5E_SQ_STATS(MLX5E_STATS_VAR) }; struct mlx5e_stats { struct mlx5e_vport_stats vport; struct mlx5e_pport_stats pport; struct mlx5e_port_stats_debug port_stats_debug; }; struct mlx5e_rq_param { u32 rqc [MLX5_ST_SZ_DW(rqc)]; struct mlx5_wq_param wq; }; struct mlx5e_sq_param { u32 sqc [MLX5_ST_SZ_DW(sqc)]; struct mlx5_wq_param wq; }; struct mlx5e_cq_param { u32 cqc [MLX5_ST_SZ_DW(cqc)]; struct mlx5_wq_param wq; }; struct mlx5e_params { u8 log_sq_size; u8 log_rq_size; u16 num_channels; u8 default_vlan_prio; u8 num_tc; u8 rx_cq_moderation_mode; u8 tx_cq_moderation_mode; u16 rx_cq_moderation_usec; u16 rx_cq_moderation_pkts; u16 tx_cq_moderation_usec; u16 tx_cq_moderation_pkts; u16 min_rx_wqes; bool hw_lro_en; bool cqe_zipping_en; u32 lro_wqe_sz; u16 rx_hash_log_tbl_sz; u32 tx_pauseframe_control __aligned(4); u32 rx_pauseframe_control __aligned(4); u16 tx_max_inline; u8 tx_min_inline_mode; u8 tx_priority_flow_control; u8 rx_priority_flow_control; u8 channels_rsss; }; #define MLX5E_PARAMS(m) \ m(+1, u64, tx_queue_size_max, "tx_queue_size_max", "Max send queue size") \ m(+1, u64, rx_queue_size_max, "rx_queue_size_max", "Max receive queue size") \ m(+1, u64, tx_queue_size, "tx_queue_size", "Default send queue size") \ m(+1, u64, rx_queue_size, "rx_queue_size", "Default receive queue size") \ m(+1, u64, channels, "channels", "Default number of channels") \ m(+1, u64, channels_rsss, "channels_rsss", "Default channels receive side scaling stride") \ m(+1, u64, coalesce_usecs_max, "coalesce_usecs_max", "Maximum usecs for joining packets") \ m(+1, u64, coalesce_pkts_max, "coalesce_pkts_max", "Maximum packets to join") \ m(+1, u64, rx_coalesce_usecs, "rx_coalesce_usecs", "Limit in usec for joining rx packets") \ m(+1, u64, rx_coalesce_pkts, "rx_coalesce_pkts", "Maximum number of rx packets to join") \ m(+1, u64, rx_coalesce_mode, "rx_coalesce_mode", "0: EQE fixed mode 1: CQE fixed mode 2: EQE auto mode 3: CQE auto mode") \ m(+1, u64, tx_coalesce_usecs, "tx_coalesce_usecs", "Limit in usec for joining tx packets") \ m(+1, u64, tx_coalesce_pkts, "tx_coalesce_pkts", "Maximum number of tx packets to join") \ m(+1, u64, tx_coalesce_mode, "tx_coalesce_mode", "0: EQE mode 1: CQE mode") \ m(+1, u64, tx_completion_fact, "tx_completion_fact", "1..MAX: Completion event ratio") \ m(+1, u64, tx_completion_fact_max, "tx_completion_fact_max", "Maximum completion event ratio") \ m(+1, u64, hw_lro, "hw_lro", "set to enable hw_lro") \ m(+1, u64, cqe_zipping, "cqe_zipping", "0 : CQE zipping disabled") \ m(+1, u64, modify_tx_dma, "modify_tx_dma", "0: Enable TX 1: Disable TX") \ m(+1, u64, modify_rx_dma, "modify_rx_dma", "0: Enable RX 1: Disable RX") \ m(+1, u64, diag_pci_enable, "diag_pci_enable", "0: Disabled 1: Enabled") \ m(+1, u64, diag_general_enable, "diag_general_enable", "0: Disabled 1: Enabled") \ m(+1, u64, hw_mtu, "hw_mtu", "Current hardware MTU value") \ m(+1, u64, mc_local_lb, "mc_local_lb", "0: Local multicast loopback enabled 1: Disabled") \ m(+1, u64, uc_local_lb, "uc_local_lb", "0: Local unicast loopback enabled 1: Disabled") #define MLX5E_PARAMS_NUM (0 MLX5E_PARAMS(MLX5E_STATS_COUNT)) struct mlx5e_params_ethtool { u64 arg [0]; MLX5E_PARAMS(MLX5E_STATS_VAR) u64 max_bw_value[IEEE_8021QAZ_MAX_TCS]; u8 max_bw_share[IEEE_8021QAZ_MAX_TCS]; u8 prio_tc[MLX5E_MAX_PRIORITY]; u8 dscp2prio[MLX5_MAX_SUPPORTED_DSCP]; u8 trust_state; u8 fec_mask_10x_25x[MLX5E_MAX_FEC_10X_25X]; u16 fec_mask_50x[MLX5E_MAX_FEC_50X]; u8 fec_avail_10x_25x[MLX5E_MAX_FEC_10X_25X]; u16 fec_avail_50x[MLX5E_MAX_FEC_50X]; u32 fec_mode_active; u32 hw_mtu_msb; s32 hw_val_temp[MLX5_MAX_TEMPERATURE]; u32 hw_num_temp; }; struct mlx5e_cq { /* data path - accessed per cqe */ struct mlx5_cqwq wq; /* data path - accessed per HW polling */ struct mlx5_core_cq mcq; /* control */ struct mlx5e_priv *priv; struct mlx5_wq_ctrl wq_ctrl; } __aligned(MLX5E_CACHELINE_SIZE); struct mlx5e_rq_mbuf { bus_dmamap_t dma_map; caddr_t data; struct mbuf *mbuf; }; struct mlx5e_rq { /* persistant fields */ struct mtx mtx; struct mlx5e_rq_stats stats; struct callout watchdog; /* data path */ #define mlx5e_rq_zero_start wq struct mlx5_wq_ll wq; bus_dma_tag_t dma_tag; u32 wqe_sz; u32 nsegs; struct mlx5e_rq_mbuf *mbuf; struct ifnet *ifp; struct mlx5e_cq cq; struct lro_ctrl lro; volatile int enabled; int ix; /* Dynamic Interrupt Moderation */ struct net_dim dim; /* control */ struct mlx5_wq_ctrl wq_ctrl; u32 rqn; struct mlx5e_channel *channel; } __aligned(MLX5E_CACHELINE_SIZE); struct mlx5e_sq_mbuf { bus_dmamap_t dma_map; struct mbuf *mbuf; volatile s32 *p_refcount; /* in use refcount, if any */ u32 num_bytes; u32 num_wqebbs; }; enum { MLX5E_SQ_READY, MLX5E_SQ_FULL }; struct mlx5e_sq { /* persistant fields */ struct mtx lock; struct mtx comp_lock; struct mlx5e_sq_stats stats; struct callout cev_callout; /* data path */ #define mlx5e_sq_zero_start dma_tag bus_dma_tag_t dma_tag; /* dirtied @completion */ u16 cc; /* dirtied @xmit */ u16 pc __aligned(MLX5E_CACHELINE_SIZE); u16 cev_counter; /* completion event counter */ u16 cev_factor; /* completion event factor */ u16 cev_next_state; /* next completion event state */ #define MLX5E_CEV_STATE_INITIAL 0 /* timer not started */ #define MLX5E_CEV_STATE_SEND_NOPS 1 /* send NOPs */ #define MLX5E_CEV_STATE_HOLD_NOPS 2 /* don't send NOPs yet */ u16 running; /* set if SQ is running */ union { u32 d32[2]; u64 d64; } doorbell; struct mlx5e_cq cq; /* pointers to per packet info: write@xmit, read@completion */ struct mlx5e_sq_mbuf *mbuf; /* read only */ struct mlx5_wq_cyc wq; void __iomem *uar_map; struct ifnet *ifp; u32 sqn; u32 mkey_be; u16 max_inline; u8 min_inline_mode; u8 min_insert_caps; #define MLX5E_INSERT_VLAN 1 #define MLX5E_INSERT_NON_VLAN 2 /* control path */ struct mlx5_wq_ctrl wq_ctrl; struct mlx5e_priv *priv; int tc; } __aligned(MLX5E_CACHELINE_SIZE); static inline bool mlx5e_sq_has_room_for(struct mlx5e_sq *sq, u16 n) { u16 cc = sq->cc; u16 pc = sq->pc; return ((sq->wq.sz_m1 & (cc - pc)) >= n || cc == pc); } static inline u32 mlx5e_sq_queue_level(struct mlx5e_sq *sq) { u16 cc; u16 pc; if (sq == NULL) return (0); cc = sq->cc; pc = sq->pc; return (((sq->wq.sz_m1 & (pc - cc)) * IF_SND_QUEUE_LEVEL_MAX) / sq->wq.sz_m1); } struct mlx5e_channel { struct mlx5e_rq rq; struct m_snd_tag tag; struct mlx5e_sq sq[MLX5E_MAX_TX_NUM_TC]; struct mlx5e_priv *priv; struct completion completion; int ix; } __aligned(MLX5E_CACHELINE_SIZE); enum mlx5e_traffic_types { MLX5E_TT_IPV4_TCP, MLX5E_TT_IPV6_TCP, MLX5E_TT_IPV4_UDP, MLX5E_TT_IPV6_UDP, MLX5E_TT_IPV4_IPSEC_AH, MLX5E_TT_IPV6_IPSEC_AH, MLX5E_TT_IPV4_IPSEC_ESP, MLX5E_TT_IPV6_IPSEC_ESP, MLX5E_TT_IPV4, MLX5E_TT_IPV6, MLX5E_TT_ANY, MLX5E_NUM_TT, }; enum { MLX5E_RQT_SPREADING = 0, MLX5E_RQT_DEFAULT_RQ = 1, MLX5E_NUM_RQT = 2, }; struct mlx5_flow_rule; struct mlx5e_eth_addr_info { u8 addr [ETH_ALEN + 2]; u32 tt_vec; /* flow table rule per traffic type */ struct mlx5_flow_rule *ft_rule[MLX5E_NUM_TT]; }; #define MLX5E_ETH_ADDR_HASH_SIZE (1 << BITS_PER_BYTE) struct mlx5e_eth_addr_hash_node; struct mlx5e_eth_addr_hash_head { struct mlx5e_eth_addr_hash_node *lh_first; }; struct mlx5e_eth_addr_db { struct mlx5e_eth_addr_hash_head if_uc[MLX5E_ETH_ADDR_HASH_SIZE]; struct mlx5e_eth_addr_hash_head if_mc[MLX5E_ETH_ADDR_HASH_SIZE]; struct mlx5e_eth_addr_info broadcast; struct mlx5e_eth_addr_info allmulti; struct mlx5e_eth_addr_info promisc; bool broadcast_enabled; bool allmulti_enabled; bool promisc_enabled; }; enum { MLX5E_STATE_ASYNC_EVENTS_ENABLE, MLX5E_STATE_OPENED, }; enum { MLX5_BW_NO_LIMIT = 0, MLX5_100_MBPS_UNIT = 3, MLX5_GBPS_UNIT = 4, }; struct mlx5e_vlan_db { unsigned long active_vlans[BITS_TO_LONGS(VLAN_N_VID)]; struct mlx5_flow_rule *active_vlans_ft_rule[VLAN_N_VID]; struct mlx5_flow_rule *untagged_ft_rule; struct mlx5_flow_rule *any_cvlan_ft_rule; struct mlx5_flow_rule *any_svlan_ft_rule; bool filter_disabled; }; struct mlx5e_flow_table { int num_groups; struct mlx5_flow_table *t; struct mlx5_flow_group **g; }; struct mlx5e_flow_tables { struct mlx5_flow_namespace *ns; struct mlx5e_flow_table vlan; struct mlx5e_flow_table main; struct mlx5e_flow_table inner_rss; }; struct mlx5e_xmit_args { volatile s32 *pref; u32 tisn; u16 ihs; }; #include "en_rl.h" #include "en_hw_tls.h" #define MLX5E_TSTMP_PREC 10 struct mlx5e_clbr_point { uint64_t base_curr; uint64_t base_prev; uint64_t clbr_hw_prev; uint64_t clbr_hw_curr; u_int clbr_gen; }; struct mlx5e_dcbx { u32 cable_len; u32 xoff; }; struct mlx5e_priv { struct mlx5_core_dev *mdev; /* must be first */ /* priv data path fields - start */ int order_base_2_num_channels; int queue_mapping_channel_mask; int num_tc; int default_vlan_prio; /* priv data path fields - end */ unsigned long state; int gone; #define PRIV_LOCK(priv) sx_xlock(&(priv)->state_lock) #define PRIV_UNLOCK(priv) sx_xunlock(&(priv)->state_lock) #define PRIV_LOCKED(priv) sx_xlocked(&(priv)->state_lock) #define PRIV_ASSERT_LOCKED(priv) sx_assert(&(priv)->state_lock, SA_XLOCKED) struct sx state_lock; /* Protects Interface state */ u32 pdn; u32 tdn; struct mlx5_core_mr mr; u32 tisn[MLX5E_MAX_TX_NUM_TC]; u32 rqtn; u32 tirn[MLX5E_NUM_TT]; struct mlx5e_flow_tables fts; struct mlx5e_eth_addr_db eth_addr; struct mlx5e_vlan_db vlan; struct mlx5e_params params; struct mlx5e_params_ethtool params_ethtool; union mlx5_core_pci_diagnostics params_pci; union mlx5_core_general_diagnostics params_general; struct mtx async_events_mtx; /* sync hw events */ struct work_struct update_stats_work; struct work_struct update_carrier_work; struct work_struct set_rx_mode_work; MLX5_DECLARE_DOORBELL_LOCK(doorbell_lock) struct ifnet *ifp; struct sysctl_ctx_list sysctl_ctx; struct sysctl_oid *sysctl_ifnet; struct sysctl_oid *sysctl_hw; int sysctl_debug; struct mlx5e_stats stats; int counter_set_id; struct workqueue_struct *wq; eventhandler_tag vlan_detach; eventhandler_tag vlan_attach; struct ifmedia media; int media_status_last; int media_active_last; struct callout watchdog; struct mlx5e_rl_priv_data rl; struct mlx5e_tls tls; struct callout tstmp_clbr; int clbr_done; int clbr_curr; struct mlx5e_clbr_point clbr_points[2]; u_int clbr_gen; struct mlx5e_dcbx dcbx; bool sw_is_port_buf_owner; struct mlx5_sq_bfreg bfreg; struct pfil_head *pfil; struct mlx5e_channel channel[]; }; #define MLX5E_NET_IP_ALIGN 2 struct mlx5e_tx_wqe { struct mlx5_wqe_ctrl_seg ctrl; struct mlx5_wqe_eth_seg eth; }; struct mlx5e_tx_umr_wqe { struct mlx5_wqe_ctrl_seg ctrl; struct mlx5_wqe_umr_ctrl_seg umr; uint8_t mkc[64]; }; struct mlx5e_tx_psv_wqe { struct mlx5_wqe_ctrl_seg ctrl; struct mlx5_seg_set_psv psv; }; struct mlx5e_rx_wqe { struct mlx5_wqe_srq_next_seg next; struct mlx5_wqe_data_seg data[]; }; /* the size of the structure above must be power of two */ CTASSERT(powerof2(sizeof(struct mlx5e_rx_wqe))); struct mlx5e_eeprom { int lock_bit; int i2c_addr; int page_num; int device_addr; int module_num; int len; int type; int page_valid; u32 *data; }; #define MLX5E_FLD_MAX(typ, fld) ((1ULL << __mlx5_bit_sz(typ, fld)) - 1ULL) bool mlx5e_do_send_cqe(struct mlx5e_sq *); int mlx5e_get_full_header_size(const struct mbuf *, const struct tcphdr **); int mlx5e_xmit(struct ifnet *, struct mbuf *); int mlx5e_open_locked(struct ifnet *); int mlx5e_close_locked(struct ifnet *); void mlx5e_cq_error_event(struct mlx5_core_cq *mcq, int event); mlx5e_cq_comp_t mlx5e_rx_cq_comp; mlx5e_cq_comp_t mlx5e_tx_cq_comp; struct mlx5_cqe64 *mlx5e_get_cqe(struct mlx5e_cq *cq); void mlx5e_dim_work(struct work_struct *); void mlx5e_dim_build_cq_param(struct mlx5e_priv *, struct mlx5e_cq_param *); int mlx5e_open_flow_table(struct mlx5e_priv *priv); void mlx5e_close_flow_table(struct mlx5e_priv *priv); void mlx5e_set_rx_mode_core(struct mlx5e_priv *priv); void mlx5e_set_rx_mode_work(struct work_struct *work); void mlx5e_vlan_rx_add_vid(void *, struct ifnet *, u16); void mlx5e_vlan_rx_kill_vid(void *, struct ifnet *, u16); void mlx5e_enable_vlan_filter(struct mlx5e_priv *priv); void mlx5e_disable_vlan_filter(struct mlx5e_priv *priv); int mlx5e_add_all_vlan_rules(struct mlx5e_priv *priv); void mlx5e_del_all_vlan_rules(struct mlx5e_priv *priv); static inline void mlx5e_tx_notify_hw(struct mlx5e_sq *sq, u32 *wqe) { /* ensure wqe is visible to device before updating doorbell record */ wmb(); *sq->wq.db = cpu_to_be32(sq->pc); /* * Ensure the doorbell record is visible to device before ringing * the doorbell: */ wmb(); mlx5_write64(wqe, sq->uar_map, MLX5_GET_DOORBELL_LOCK(&sq->priv->doorbell_lock)); } static inline void mlx5e_cq_arm(struct mlx5e_cq *cq, spinlock_t *dblock) { struct mlx5_core_cq *mcq; mcq = &cq->mcq; mlx5_cq_arm(mcq, MLX5_CQ_DB_REQ_NOT, mcq->uar->map, dblock, cq->wq.cc); } #define mlx5e_dbg(_IGN, _priv, ...) mlx5_core_dbg((_priv)->mdev, __VA_ARGS__) extern const struct ethtool_ops mlx5e_ethtool_ops; void mlx5e_create_ethtool(struct mlx5e_priv *); void mlx5e_create_stats(struct sysctl_ctx_list *, struct sysctl_oid_list *, const char *, const char **, unsigned, u64 *); void mlx5e_create_counter_stats(struct sysctl_ctx_list *, struct sysctl_oid_list *, const char *, const char **, unsigned, counter_u64_t *); void mlx5e_send_nop(struct mlx5e_sq *, u32); int mlx5e_sq_dump_xmit(struct mlx5e_sq *, struct mlx5e_xmit_args *, struct mbuf **); int mlx5e_sq_xmit(struct mlx5e_sq *, struct mbuf **); void mlx5e_sq_cev_timeout(void *); int mlx5e_refresh_channel_params(struct mlx5e_priv *); int mlx5e_open_cq(struct mlx5e_priv *, struct mlx5e_cq_param *, struct mlx5e_cq *, mlx5e_cq_comp_t *, int eq_ix); void mlx5e_close_cq(struct mlx5e_cq *); void mlx5e_free_sq_db(struct mlx5e_sq *); int mlx5e_alloc_sq_db(struct mlx5e_sq *); int mlx5e_enable_sq(struct mlx5e_sq *, struct mlx5e_sq_param *, int tis_num); int mlx5e_modify_sq(struct mlx5e_sq *, int curr_state, int next_state); void mlx5e_disable_sq(struct mlx5e_sq *); void mlx5e_drain_sq(struct mlx5e_sq *); void mlx5e_modify_tx_dma(struct mlx5e_priv *priv, uint8_t value); void mlx5e_modify_rx_dma(struct mlx5e_priv *priv, uint8_t value); void mlx5e_resume_sq(struct mlx5e_sq *sq); void mlx5e_update_sq_inline(struct mlx5e_sq *sq); void mlx5e_refresh_sq_inline(struct mlx5e_priv *priv); int mlx5e_update_buf_lossy(struct mlx5e_priv *priv); int mlx5e_fec_update(struct mlx5e_priv *priv); int mlx5e_hw_temperature_update(struct mlx5e_priv *priv); if_snd_tag_alloc_t mlx5e_ul_snd_tag_alloc; if_snd_tag_modify_t mlx5e_ul_snd_tag_modify; if_snd_tag_query_t mlx5e_ul_snd_tag_query; if_snd_tag_free_t mlx5e_ul_snd_tag_free; #endif /* _MLX5_EN_H_ */ diff --git a/sys/dev/mlx5/mlx5_ib/mlx5_ib.h b/sys/dev/mlx5/mlx5_ib/mlx5_ib.h index b9828f1da466..0420b294de80 100644 --- a/sys/dev/mlx5/mlx5_ib/mlx5_ib.h +++ b/sys/dev/mlx5/mlx5_ib/mlx5_ib.h @@ -1,1105 +1,1106 @@ /*- * Copyright (c) 2013-2015, Mellanox Technologies, Ltd. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS `AS IS' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * $FreeBSD$ */ #ifndef MLX5_IB_H #define MLX5_IB_H #include #include #include +#include #include #include #include #include #include #include #include #include #include #define mlx5_ib_dbg(dev, format, arg...) \ pr_debug("%s:%s:%d:(pid %d): " format, (dev)->ib_dev.name, __func__, \ __LINE__, current->pid, ##arg) #define mlx5_ib_err(dev, format, arg...) \ pr_err("%s: ERR: %s:%d:(pid %d): " format, (dev)->ib_dev.name, __func__, \ __LINE__, current->pid, ##arg) #define mlx5_ib_warn(dev, format, arg...) \ pr_warn("%s: WARN: %s:%d:(pid %d): " format, (dev)->ib_dev.name, __func__, \ __LINE__, current->pid, ##arg) #define field_avail(type, fld, sz) (offsetof(type, fld) + \ sizeof(((type *)0)->fld) <= (sz)) #define MLX5_IB_DEFAULT_UIDX 0xffffff #define MLX5_USER_ASSIGNED_UIDX_MASK __mlx5_mask(qpc, user_index) enum { MLX5_IB_MMAP_CMD_SHIFT = 8, MLX5_IB_MMAP_CMD_MASK = 0xff, }; enum mlx5_ib_mmap_cmd { MLX5_IB_MMAP_REGULAR_PAGE = 0, MLX5_IB_MMAP_GET_CONTIGUOUS_PAGES = 1, MLX5_IB_MMAP_WC_PAGE = 2, MLX5_IB_MMAP_NC_PAGE = 3, /* 5 is chosen in order to be compatible with old versions of libmlx5 */ MLX5_IB_MMAP_CORE_CLOCK = 5, }; enum { MLX5_RES_SCAT_DATA32_CQE = 0x1, MLX5_RES_SCAT_DATA64_CQE = 0x2, MLX5_REQ_SCAT_DATA32_CQE = 0x11, MLX5_REQ_SCAT_DATA64_CQE = 0x22, }; enum mlx5_ib_latency_class { MLX5_IB_LATENCY_CLASS_LOW, MLX5_IB_LATENCY_CLASS_MEDIUM, MLX5_IB_LATENCY_CLASS_HIGH, MLX5_IB_LATENCY_CLASS_FAST_PATH }; enum mlx5_ib_mad_ifc_flags { MLX5_MAD_IFC_IGNORE_MKEY = 1, MLX5_MAD_IFC_IGNORE_BKEY = 2, MLX5_MAD_IFC_NET_VIEW = 4, }; enum { MLX5_CROSS_CHANNEL_BFREG = 0, }; enum { MLX5_CQE_VERSION_V0, MLX5_CQE_VERSION_V1, }; enum { MLX5_IB_INVALID_UAR_INDEX = BIT(31), MLX5_IB_INVALID_BFREG = BIT(31), }; struct mlx5_ib_vma_private_data { struct list_head list; struct vm_area_struct *vma; }; struct mlx5_bfreg_info { u32 *sys_pages; int num_low_latency_bfregs; unsigned int *count; /* * protect bfreg allocation data structs */ struct mutex lock; u32 ver; u8 lib_uar_4k : 1; u8 lib_uar_dyn : 1; u32 num_sys_pages; u32 num_static_sys_pages; u32 total_num_bfregs; u32 num_dyn_bfregs; }; struct mlx5_ib_ucontext { struct ib_ucontext ibucontext; struct list_head db_page_list; /* protect doorbell record alloc/free */ struct mutex db_page_mutex; struct mlx5_bfreg_info bfregi; u8 cqe_version; /* Transport Domain number */ u32 tdn; struct list_head vma_private_list; }; static inline struct mlx5_ib_ucontext *to_mucontext(struct ib_ucontext *ibucontext) { return container_of(ibucontext, struct mlx5_ib_ucontext, ibucontext); } struct mlx5_ib_pd { struct ib_pd ibpd; u32 pdn; }; #define MLX5_IB_FLOW_MCAST_PRIO (MLX5_BY_PASS_NUM_PRIOS - 1) #define MLX5_IB_FLOW_LAST_PRIO (MLX5_BY_PASS_NUM_REGULAR_PRIOS - 1) #if (MLX5_IB_FLOW_LAST_PRIO <= 0) #error "Invalid number of bypass priorities" #endif #define MLX5_IB_FLOW_LEFTOVERS_PRIO (MLX5_IB_FLOW_MCAST_PRIO + 1) #define MLX5_IB_NUM_FLOW_FT (MLX5_IB_FLOW_LEFTOVERS_PRIO + 1) #define MLX5_IB_NUM_SNIFFER_FTS 2 struct mlx5_ib_flow_prio { struct mlx5_flow_table *flow_table; unsigned int refcount; }; struct mlx5_ib_flow_handler { struct list_head list; struct ib_flow ibflow; struct mlx5_ib_flow_prio *prio; struct mlx5_flow_rule *rule; }; struct mlx5_ib_flow_db { struct mlx5_ib_flow_prio prios[MLX5_IB_NUM_FLOW_FT]; struct mlx5_ib_flow_prio sniffer[MLX5_IB_NUM_SNIFFER_FTS]; struct mlx5_flow_table *lag_demux_ft; /* Protect flow steering bypass flow tables * when add/del flow rules. * only single add/removal of flow steering rule could be done * simultaneously. */ struct mutex lock; }; /* Use macros here so that don't have to duplicate * enum ib_send_flags and enum ib_qp_type for low-level driver */ #define MLX5_IB_SEND_UMR_UNREG IB_SEND_RESERVED_START #define MLX5_IB_SEND_UMR_FAIL_IF_FREE (IB_SEND_RESERVED_START << 1) #define MLX5_IB_SEND_UMR_UPDATE_MTT (IB_SEND_RESERVED_START << 2) #define MLX5_IB_SEND_UMR_UPDATE_TRANSLATION (IB_SEND_RESERVED_START << 3) #define MLX5_IB_SEND_UMR_UPDATE_PD (IB_SEND_RESERVED_START << 4) #define MLX5_IB_SEND_UMR_UPDATE_ACCESS IB_SEND_RESERVED_END #define MLX5_IB_QPT_REG_UMR IB_QPT_RESERVED1 /* * IB_QPT_GSI creates the software wrapper around GSI, and MLX5_IB_QPT_HW_GSI * creates the actual hardware QP. */ #define MLX5_IB_QPT_HW_GSI IB_QPT_RESERVED2 #define MLX5_IB_WR_UMR IB_WR_RESERVED1 /* Private QP creation flags to be passed in ib_qp_init_attr.create_flags. * * These flags are intended for internal use by the mlx5_ib driver, and they * rely on the range reserved for that use in the ib_qp_create_flags enum. */ #define MLX5_IB_QP_CREATE_SQPN_QP1 IB_QP_CREATE_RESERVED_START #define MLX5_IB_QP_CREATE_WC_TEST (IB_QP_CREATE_RESERVED_START << 1) struct wr_list { u16 opcode; u16 next; }; struct mlx5_ib_wq { u64 *wrid; u32 *wr_data; struct wr_list *w_list; unsigned *wqe_head; u16 unsig_count; /* serialize post to the work queue */ spinlock_t lock; int wqe_cnt; int max_post; int max_gs; int offset; int wqe_shift; unsigned head; unsigned tail; u16 cur_post; u16 last_poll; void *qend; }; struct mlx5_ib_rwq { struct ib_wq ibwq; struct mlx5_core_qp core_qp; u32 rq_num_pas; u32 log_rq_stride; u32 log_rq_size; u32 rq_page_offset; u32 log_page_size; struct ib_umem *umem; size_t buf_size; unsigned int page_shift; int create_type; struct mlx5_db db; u32 user_index; u32 wqe_count; u32 wqe_shift; int wq_sig; }; enum { MLX5_QP_USER, MLX5_QP_KERNEL, MLX5_QP_EMPTY }; enum { MLX5_WQ_USER, MLX5_WQ_KERNEL }; struct mlx5_ib_rwq_ind_table { struct ib_rwq_ind_table ib_rwq_ind_tbl; u32 rqtn; }; /* * Connect-IB can trigger up to four concurrent pagefaults * per-QP. */ enum mlx5_ib_pagefault_context { MLX5_IB_PAGEFAULT_RESPONDER_READ, MLX5_IB_PAGEFAULT_REQUESTOR_READ, MLX5_IB_PAGEFAULT_RESPONDER_WRITE, MLX5_IB_PAGEFAULT_REQUESTOR_WRITE, MLX5_IB_PAGEFAULT_CONTEXTS }; static inline enum mlx5_ib_pagefault_context mlx5_ib_get_pagefault_context(struct mlx5_pagefault *pagefault) { return pagefault->flags & (MLX5_PFAULT_REQUESTOR | MLX5_PFAULT_WRITE); } struct mlx5_ib_pfault { struct work_struct work; struct mlx5_pagefault mpfault; }; struct mlx5_ib_ubuffer { struct ib_umem *umem; int buf_size; u64 buf_addr; }; struct mlx5_ib_qp_base { struct mlx5_ib_qp *container_mibqp; struct mlx5_core_qp mqp; struct mlx5_ib_ubuffer ubuffer; }; struct mlx5_ib_qp_trans { struct mlx5_ib_qp_base base; u16 xrcdn; u8 alt_port; u8 atomic_rd_en; u8 resp_depth; }; struct mlx5_ib_rss_qp { u32 tirn; }; struct mlx5_ib_rq { struct mlx5_ib_qp_base base; struct mlx5_ib_wq *rq; struct mlx5_ib_ubuffer ubuffer; struct mlx5_db *doorbell; u32 tirn; u8 state; }; struct mlx5_ib_sq { struct mlx5_ib_qp_base base; struct mlx5_ib_wq *sq; struct mlx5_ib_ubuffer ubuffer; struct mlx5_db *doorbell; u32 tisn; u8 state; }; struct mlx5_ib_raw_packet_qp { struct mlx5_ib_sq sq; struct mlx5_ib_rq rq; }; struct mlx5_bf { int buf_size; unsigned long offset; struct mlx5_sq_bfreg *bfreg; spinlock_t lock32; }; struct mlx5_ib_qp { struct ib_qp ibqp; union { struct mlx5_ib_qp_trans trans_qp; struct mlx5_ib_raw_packet_qp raw_packet_qp; struct mlx5_ib_rss_qp rss_qp; }; struct mlx5_buf buf; struct mlx5_db db; struct mlx5_ib_wq rq; u8 sq_signal_bits; u8 fm_cache; struct mlx5_ib_wq sq; /* serialize qp state modifications */ struct mutex mutex; u32 flags; u8 port; u8 state; int wq_sig; int scat_cqe; int max_inline_data; struct mlx5_bf bf; int has_rq; /* only for user space QPs. For kernel * we have it from the bf object */ int bfregn; int create_type; /* Store signature errors */ bool signature_en; #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING /* * A flag that is true for QP's that are in a state that doesn't * allow page faults, and shouldn't schedule any more faults. */ int disable_page_faults; /* * The disable_page_faults_lock protects a QP's disable_page_faults * field, allowing for a thread to atomically check whether the QP * allows page faults, and if so schedule a page fault. */ spinlock_t disable_page_faults_lock; struct mlx5_ib_pfault pagefaults[MLX5_IB_PAGEFAULT_CONTEXTS]; #endif struct list_head qps_list; struct list_head cq_recv_list; struct list_head cq_send_list; }; struct mlx5_ib_cq_buf { struct mlx5_buf buf; struct ib_umem *umem; int cqe_size; int nent; }; enum mlx5_ib_qp_flags { MLX5_IB_QP_LSO = IB_QP_CREATE_IPOIB_UD_LSO, MLX5_IB_QP_BLOCK_MULTICAST_LOOPBACK = IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK, MLX5_IB_QP_CROSS_CHANNEL = IB_QP_CREATE_CROSS_CHANNEL, MLX5_IB_QP_MANAGED_SEND = IB_QP_CREATE_MANAGED_SEND, MLX5_IB_QP_MANAGED_RECV = IB_QP_CREATE_MANAGED_RECV, MLX5_IB_QP_SIGNATURE_HANDLING = 1 << 5, /* QP uses 1 as its source QP number */ MLX5_IB_QP_SQPN_QP1 = 1 << 6, MLX5_IB_QP_CAP_SCATTER_FCS = 1 << 7, MLX5_IB_QP_RSS = 1 << 8, }; struct mlx5_umr_wr { struct ib_send_wr wr; union { u64 virt_addr; u64 offset; } target; struct ib_pd *pd; unsigned int page_shift; unsigned int npages; u32 length; int access_flags; u32 mkey; }; static inline struct mlx5_umr_wr *umr_wr(struct ib_send_wr *wr) { return container_of(wr, struct mlx5_umr_wr, wr); } struct mlx5_shared_mr_info { int mr_id; struct ib_umem *umem; }; struct mlx5_ib_cq { struct ib_cq ibcq; struct mlx5_core_cq mcq; struct mlx5_ib_cq_buf buf; struct mlx5_db db; /* serialize access to the CQ */ spinlock_t lock; /* protect resize cq */ struct mutex resize_mutex; struct mlx5_ib_cq_buf *resize_buf; struct ib_umem *resize_umem; int cqe_size; struct list_head list_send_qp; struct list_head list_recv_qp; u32 create_flags; struct list_head wc_list; enum ib_cq_notify_flags notify_flags; struct work_struct notify_work; }; struct mlx5_ib_wc { struct ib_wc wc; struct list_head list; }; struct mlx5_ib_srq { struct ib_srq ibsrq; struct mlx5_core_srq msrq; struct mlx5_buf buf; struct mlx5_db db; u64 *wrid; /* protect SRQ hanlding */ spinlock_t lock; int head; int tail; u16 wqe_ctr; struct ib_umem *umem; /* serialize arming a SRQ */ struct mutex mutex; int wq_sig; }; struct mlx5_ib_xrcd { struct ib_xrcd ibxrcd; u32 xrcdn; }; enum mlx5_ib_mtt_access_flags { MLX5_IB_MTT_READ = (1 << 0), MLX5_IB_MTT_WRITE = (1 << 1), }; #define MLX5_IB_MTT_PRESENT (MLX5_IB_MTT_READ | MLX5_IB_MTT_WRITE) struct mlx5_ib_mr { struct ib_mr ibmr; void *descs; dma_addr_t desc_map; int ndescs; int max_descs; int desc_size; int access_mode; struct mlx5_core_mr mmkey; struct ib_umem *umem; struct mlx5_shared_mr_info *smr_info; struct list_head list; int order; int umred; int npages; struct mlx5_ib_dev *dev; u32 out[MLX5_ST_SZ_DW(create_mkey_out)]; struct mlx5_core_sig_ctx *sig; int live; void *descs_alloc; int access_flags; /* Needed for rereg MR */ struct mlx5_async_work cb_work; }; struct mlx5_ib_mw { struct ib_mw ibmw; struct mlx5_core_mr mmkey; }; struct mlx5_ib_umr_context { struct ib_cqe cqe; enum ib_wc_status status; struct completion done; }; struct umr_common { struct ib_pd *pd; struct ib_cq *cq; struct ib_qp *qp; /* control access to UMR QP */ struct semaphore sem; }; enum { MLX5_FMR_INVALID, MLX5_FMR_VALID, MLX5_FMR_BUSY, }; struct mlx5_cache_ent { struct list_head head; /* sync access to the cahce entry */ spinlock_t lock; struct dentry *dir; char name[4]; u32 order; u32 size; u32 cur; u32 miss; u32 limit; struct dentry *fsize; struct dentry *fcur; struct dentry *fmiss; struct dentry *flimit; struct mlx5_ib_dev *dev; struct work_struct work; struct delayed_work dwork; int pending; }; struct mlx5_mr_cache { struct workqueue_struct *wq; struct mlx5_cache_ent ent[MAX_MR_CACHE_ENTRIES]; int stopped; struct dentry *root; unsigned long last_add; }; struct mlx5_ib_gsi_qp; struct mlx5_ib_port_resources { struct mlx5_ib_resources *devr; struct mlx5_ib_gsi_qp *gsi; struct work_struct pkey_change_work; }; struct mlx5_ib_resources { struct ib_cq *c0; struct ib_xrcd *x0; struct ib_xrcd *x1; struct ib_pd *p0; struct ib_srq *s0; struct ib_srq *s1; struct mlx5_ib_port_resources ports[2]; /* Protects changes to the port resources */ struct mutex mutex; }; struct mlx5_ib_port { u16 q_cnt_id; }; struct mlx5_roce { /* Protect mlx5_ib_get_netdev from invoking dev_hold() with a NULL * netdev pointer */ rwlock_t netdev_lock; - struct net_device *netdev; + struct ifnet *netdev; struct notifier_block nb; atomic_t next_port; }; #define MLX5_IB_STATS_COUNT(a,b,c,d) a #define MLX5_IB_STATS_VAR(a,b,c,d) b; #define MLX5_IB_STATS_DESC(a,b,c,d) c, d, #define MLX5_IB_CONG_PARAMS(m) \ /* ECN RP */ \ m(+1, u64 rp_clamp_tgt_rate, "rp_clamp_tgt_rate", "If set, whenever a CNP is processed, the target rate is updated to be the current rate") \ m(+1, u64 rp_clamp_tgt_rate_ati, "rp_clamp_tgt_rate_ati", "If set, when receiving a CNP, the target rate should be updated if the transission rate was increased due to the timer, and not only due to the byte counter") \ m(+1, u64 rp_time_reset, "rp_time_reset", "Time in microseconds between rate increases if no CNPs are received") \ m(+1, u64 rp_byte_reset, "rp_byte_reset", "Transmitted data in bytes between rate increases if no CNP's are received. A value of zero means disabled.") \ m(+1, u64 rp_threshold, "rp_threshold", "The number of times rpByteStage or rpTimeStage can count before the RP rate control state machine advances states") \ m(+1, u64 rp_ai_rate, "rp_ai_rate", "The rate, in Mbits per second, used to increase rpTargetRate in the active increase state") \ m(+1, u64 rp_hai_rate, "rp_hai_rate", "The rate, in Mbits per second, used to increase rpTargetRate in the hyper increase state") \ m(+1, u64 rp_min_dec_fac, "rp_min_dec_fac", "The minimum factor by which the current transmit rate can be changed when processing a CNP. Value is given as a percentage, [1 .. 100]") \ m(+1, u64 rp_min_rate, "rp_min_rate", "The minimum value, in Mbps per second, for rate to limit") \ m(+1, u64 rp_rate_to_set_on_first_cnp, "rp_rate_to_set_on_first_cnp", "The rate that is set for the flow when a rate limiter is allocated to it upon first CNP received, in Mbps. A value of zero means use full port speed") \ m(+1, u64 rp_dce_tcp_g, "rp_dce_tcp_g", "Used to update the congestion estimator, alpha, once every dce_tcp_rtt once every dce_tcp_rtt microseconds") \ m(+1, u64 rp_dce_tcp_rtt, "rp_dce_tcp_rtt", "The time between updates of the aolpha value, in microseconds") \ m(+1, u64 rp_rate_reduce_monitor_period, "rp_rate_reduce_monitor_period", "The minimum time between two consecutive rate reductions for a single flow") \ m(+1, u64 rp_initial_alpha_value, "rp_initial_alpha_value", "The initial value of alpha to use when receiving the first CNP for a flow") \ m(+1, u64 rp_gd, "rp_gd", "If a CNP is received, the flow rate is reduced at the beginning of the next rate_reduce_monitor_period interval") \ /* ECN NP */ \ m(+1, u64 np_cnp_dscp, "np_cnp_dscp", "The DiffServ Code Point of the generated CNP for this port") \ m(+1, u64 np_cnp_prio_mode, "np_cnp_prio_mode", "The 802.1p priority value of the generated CNP for this port") \ m(+1, u64 np_cnp_prio, "np_cnp_prio", "The 802.1p priority value of the generated CNP for this port") #define MLX5_IB_CONG_PARAMS_NUM (0 MLX5_IB_CONG_PARAMS(MLX5_IB_STATS_COUNT)) #define MLX5_IB_CONG_STATS(m) \ m(+1, u64 syndrome, "syndrome", "Syndrome number") \ m(+1, u64 rp_cur_flows, "rp_cur_flows", "Number of flows limited") \ m(+1, u64 sum_flows, "sum_flows", "Sum of the number of flows limited over time") \ m(+1, u64 rp_cnp_ignored, "rp_cnp_ignored", "Number of CNPs and CNMs ignored") \ m(+1, u64 rp_cnp_handled, "rp_cnp_handled", "Number of CNPs and CNMs successfully handled") \ m(+1, u64 time_stamp, "time_stamp", "Time stamp in microseconds") \ m(+1, u64 accumulators_period, "accumulators_period", "The value of X variable for accumulating counters") \ m(+1, u64 np_ecn_marked_roce_packets, "np_ecn_marked_roce_packets", "Number of ECN marked packets seen") \ m(+1, u64 np_cnp_sent, "np_cnp_sent", "Number of CNPs sent") #define MLX5_IB_CONG_STATS_NUM (0 MLX5_IB_CONG_STATS(MLX5_IB_STATS_COUNT)) struct mlx5_ib_congestion { struct sysctl_ctx_list ctx; struct sx lock; struct delayed_work dwork; union { u64 arg[1]; struct { MLX5_IB_CONG_PARAMS(MLX5_IB_STATS_VAR) MLX5_IB_CONG_STATS(MLX5_IB_STATS_VAR) }; }; }; struct mlx5_ib_dev { struct ib_device ib_dev; struct mlx5_core_dev *mdev; struct mlx5_roce roce; MLX5_DECLARE_DOORBELL_LOCK(uar_lock); int num_ports; /* serialize update of capability mask */ struct mutex cap_mask_mutex; bool ib_active; struct umr_common umrc; /* sync used page count stats */ struct mlx5_ib_resources devr; struct mlx5_mr_cache cache; struct timer_list delay_timer; /* Prevents soft lock on massive reg MRs */ struct mutex slow_path_mutex; int fill_delay; #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING struct ib_odp_caps odp_caps; /* * Sleepable RCU that prevents destruction of MRs while they are still * being used by a page fault handler. */ struct srcu_struct mr_srcu; #endif struct mlx5_ib_flow_db flow_db; /* protect resources needed as part of reset flow */ spinlock_t reset_flow_resource_lock; struct list_head qp_list; /* Array with num_ports elements */ struct mlx5_ib_port *port; struct mlx5_sq_bfreg bfreg; struct mlx5_sq_bfreg wc_bfreg; struct mlx5_sq_bfreg fp_bfreg; struct mlx5_ib_congestion congestion; struct mlx5_async_ctx async_ctx; }; static inline struct mlx5_ib_cq *to_mibcq(struct mlx5_core_cq *mcq) { return container_of(mcq, struct mlx5_ib_cq, mcq); } static inline struct mlx5_ib_xrcd *to_mxrcd(struct ib_xrcd *ibxrcd) { return container_of(ibxrcd, struct mlx5_ib_xrcd, ibxrcd); } static inline struct mlx5_ib_dev *to_mdev(struct ib_device *ibdev) { return container_of(ibdev, struct mlx5_ib_dev, ib_dev); } static inline struct mlx5_ib_cq *to_mcq(struct ib_cq *ibcq) { return container_of(ibcq, struct mlx5_ib_cq, ibcq); } static inline struct mlx5_ib_qp *to_mibqp(struct mlx5_core_qp *mqp) { return container_of(mqp, struct mlx5_ib_qp_base, mqp)->container_mibqp; } static inline struct mlx5_ib_rwq *to_mibrwq(struct mlx5_core_qp *core_qp) { return container_of(core_qp, struct mlx5_ib_rwq, core_qp); } static inline struct mlx5_ib_mr *to_mibmr(struct mlx5_core_mr *mmkey) { return container_of(mmkey, struct mlx5_ib_mr, mmkey); } static inline struct mlx5_ib_pd *to_mpd(struct ib_pd *ibpd) { return container_of(ibpd, struct mlx5_ib_pd, ibpd); } static inline struct mlx5_ib_srq *to_msrq(struct ib_srq *ibsrq) { return container_of(ibsrq, struct mlx5_ib_srq, ibsrq); } static inline struct mlx5_ib_qp *to_mqp(struct ib_qp *ibqp) { return container_of(ibqp, struct mlx5_ib_qp, ibqp); } static inline struct mlx5_ib_rwq *to_mrwq(struct ib_wq *ibwq) { return container_of(ibwq, struct mlx5_ib_rwq, ibwq); } static inline struct mlx5_ib_rwq_ind_table *to_mrwq_ind_table(struct ib_rwq_ind_table *ib_rwq_ind_tbl) { return container_of(ib_rwq_ind_tbl, struct mlx5_ib_rwq_ind_table, ib_rwq_ind_tbl); } static inline struct mlx5_ib_srq *to_mibsrq(struct mlx5_core_srq *msrq) { return container_of(msrq, struct mlx5_ib_srq, msrq); } static inline struct mlx5_ib_mr *to_mmr(struct ib_mr *ibmr) { return container_of(ibmr, struct mlx5_ib_mr, ibmr); } static inline struct mlx5_ib_mw *to_mmw(struct ib_mw *ibmw) { return container_of(ibmw, struct mlx5_ib_mw, ibmw); } struct mlx5_ib_ah { struct ib_ah ibah; struct mlx5_av av; }; static inline struct mlx5_ib_ah *to_mah(struct ib_ah *ibah) { return container_of(ibah, struct mlx5_ib_ah, ibah); } int mlx5_ib_db_map_user(struct mlx5_ib_ucontext *context, unsigned long virt, struct mlx5_db *db); void mlx5_ib_db_unmap_user(struct mlx5_ib_ucontext *context, struct mlx5_db *db); void __mlx5_ib_cq_clean(struct mlx5_ib_cq *cq, u32 qpn, struct mlx5_ib_srq *srq); void mlx5_ib_cq_clean(struct mlx5_ib_cq *cq, u32 qpn, struct mlx5_ib_srq *srq); void mlx5_ib_free_srq_wqe(struct mlx5_ib_srq *srq, int wqe_index); int mlx5_MAD_IFC(struct mlx5_ib_dev *dev, int ignore_mkey, int ignore_bkey, u8 port, const struct ib_wc *in_wc, const struct ib_grh *in_grh, const void *in_mad, void *response_mad); struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, struct ib_udata *udata); int mlx5_ib_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr); int mlx5_ib_destroy_ah(struct ib_ah *ah); struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd, struct ib_srq_init_attr *init_attr, struct ib_udata *udata); int mlx5_ib_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr, enum ib_srq_attr_mask attr_mask, struct ib_udata *udata); int mlx5_ib_query_srq(struct ib_srq *ibsrq, struct ib_srq_attr *srq_attr); int mlx5_ib_destroy_srq(struct ib_srq *srq); int mlx5_ib_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, struct ib_recv_wr **bad_wr); struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *init_attr, struct ib_udata *udata); int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask, struct ib_udata *udata); int mlx5_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr); int mlx5_ib_destroy_qp(struct ib_qp *qp); int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, struct ib_send_wr **bad_wr); int mlx5_ib_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, struct ib_recv_wr **bad_wr); void *mlx5_get_send_wqe(struct mlx5_ib_qp *qp, int n); int mlx5_ib_read_user_wqe(struct mlx5_ib_qp *qp, int send, int wqe_index, void *buffer, u32 length, struct mlx5_ib_qp_base *base); struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, const struct ib_cq_init_attr *attr, struct ib_ucontext *context, struct ib_udata *udata); int mlx5_ib_destroy_cq(struct ib_cq *cq); int mlx5_ib_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc); int mlx5_ib_arm_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags); int mlx5_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period); int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata); struct ib_mr *mlx5_ib_get_dma_mr(struct ib_pd *pd, int acc); struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, u64 virt_addr, int access_flags, struct ib_udata *udata); struct ib_mw *mlx5_ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type, struct ib_udata *udata); int mlx5_ib_dealloc_mw(struct ib_mw *mw); int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index, int npages, int zap); int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start, u64 length, u64 virt_addr, int access_flags, struct ib_pd *pd, struct ib_udata *udata); int mlx5_ib_dereg_mr(struct ib_mr *ibmr); struct ib_mr *mlx5_ib_alloc_mr(struct ib_pd *pd, enum ib_mr_type mr_type, u32 max_num_sg); int mlx5_ib_map_mr_sg(struct ib_mr *ibmr, struct scatterlist *sg, int sg_nents, unsigned int *sg_offset); int mlx5_ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num, const struct ib_wc *in_wc, const struct ib_grh *in_grh, const struct ib_mad_hdr *in, size_t in_mad_size, struct ib_mad_hdr *out, size_t *out_mad_size, u16 *out_mad_pkey_index); struct ib_xrcd *mlx5_ib_alloc_xrcd(struct ib_device *ibdev, struct ib_ucontext *context, struct ib_udata *udata); int mlx5_ib_dealloc_xrcd(struct ib_xrcd *xrcd); int mlx5_ib_get_buf_offset(u64 addr, int page_shift, u32 *offset); int mlx5_query_ext_port_caps(struct mlx5_ib_dev *dev, u8 port); int mlx5_query_mad_ifc_smp_attr_node_info(struct ib_device *ibdev, struct ib_smp *out_mad); int mlx5_query_mad_ifc_system_image_guid(struct ib_device *ibdev, __be64 *sys_image_guid); int mlx5_query_mad_ifc_max_pkeys(struct ib_device *ibdev, u16 *max_pkeys); int mlx5_query_mad_ifc_vendor_id(struct ib_device *ibdev, u32 *vendor_id); int mlx5_query_mad_ifc_node_desc(struct mlx5_ib_dev *dev, char *node_desc); int mlx5_query_mad_ifc_node_guid(struct mlx5_ib_dev *dev, __be64 *node_guid); int mlx5_query_mad_ifc_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 *pkey); int mlx5_query_mad_ifc_gids(struct ib_device *ibdev, u8 port, int index, union ib_gid *gid); int mlx5_query_mad_ifc_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props); int mlx5_ib_query_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props); int mlx5_ib_init_fmr(struct mlx5_ib_dev *dev); void mlx5_ib_cleanup_fmr(struct mlx5_ib_dev *dev); void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift, int *ncont, int *order); void __mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem, int page_shift, size_t offset, size_t num_pages, __be64 *pas, int access_flags); void mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem, int page_shift, __be64 *pas, int access_flags); void mlx5_ib_copy_pas(u64 *old, u64 *new, int step, int num); int mlx5_ib_get_cqe_size(struct mlx5_ib_dev *dev, struct ib_cq *ibcq); int mlx5_mr_cache_init(struct mlx5_ib_dev *dev); int mlx5_mr_cache_cleanup(struct mlx5_ib_dev *dev); int mlx5_mr_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift); int mlx5_ib_check_mr_status(struct ib_mr *ibmr, u32 check_mask, struct ib_mr_status *mr_status); struct ib_wq *mlx5_ib_create_wq(struct ib_pd *pd, struct ib_wq_init_attr *init_attr, struct ib_udata *udata); int mlx5_ib_destroy_wq(struct ib_wq *wq); int mlx5_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr, u32 wq_attr_mask, struct ib_udata *udata); struct ib_rwq_ind_table *mlx5_ib_create_rwq_ind_table(struct ib_device *device, struct ib_rwq_ind_table_init_attr *init_attr, struct ib_udata *udata); int mlx5_ib_destroy_rwq_ind_table(struct ib_rwq_ind_table *wq_ind_table); #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING extern struct workqueue_struct *mlx5_ib_page_fault_wq; void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev); void mlx5_ib_mr_pfault_handler(struct mlx5_ib_qp *qp, struct mlx5_ib_pfault *pfault); void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp); int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev); void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *ibdev); int __init mlx5_ib_odp_init(void); void mlx5_ib_odp_cleanup(void); void mlx5_ib_qp_disable_pagefaults(struct mlx5_ib_qp *qp); void mlx5_ib_qp_enable_pagefaults(struct mlx5_ib_qp *qp); void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start, unsigned long end); #else /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */ static inline void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev) { return; } static inline void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp) {} static inline int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev) { return 0; } static inline void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *ibdev) {} static inline int mlx5_ib_odp_init(void) { return 0; } static inline void mlx5_ib_odp_cleanup(void) {} static inline void mlx5_ib_qp_disable_pagefaults(struct mlx5_ib_qp *qp) {} static inline void mlx5_ib_qp_enable_pagefaults(struct mlx5_ib_qp *qp) {} #endif /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */ int mlx5_ib_get_vf_config(struct ib_device *device, int vf, u8 port, struct ifla_vf_info *info); int mlx5_ib_set_vf_link_state(struct ib_device *device, int vf, u8 port, int state); int mlx5_ib_get_vf_stats(struct ib_device *device, int vf, u8 port, struct ifla_vf_stats *stats); int mlx5_ib_set_vf_guid(struct ib_device *device, int vf, u8 port, u64 guid, int type); __be16 mlx5_get_roce_udp_sport(struct mlx5_ib_dev *dev, u8 port_num, int index); int mlx5_get_roce_gid_type(struct mlx5_ib_dev *dev, u8 port_num, int index, enum ib_gid_type *gid_type); /* GSI QP helper functions */ struct ib_qp *mlx5_ib_gsi_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *init_attr); int mlx5_ib_gsi_destroy_qp(struct ib_qp *qp); int mlx5_ib_gsi_modify_qp(struct ib_qp *qp, struct ib_qp_attr *attr, int attr_mask); int mlx5_ib_gsi_query_qp(struct ib_qp *qp, struct ib_qp_attr *qp_attr, int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr); int mlx5_ib_gsi_post_send(struct ib_qp *qp, struct ib_send_wr *wr, struct ib_send_wr **bad_wr); int mlx5_ib_gsi_post_recv(struct ib_qp *qp, struct ib_recv_wr *wr, struct ib_recv_wr **bad_wr); void mlx5_ib_gsi_pkey_change(struct mlx5_ib_gsi_qp *gsi); int mlx5_ib_generate_wc(struct ib_cq *ibcq, struct ib_wc *wc); void mlx5_ib_free_bfreg(struct mlx5_ib_dev *dev, struct mlx5_bfreg_info *bfregi, int bfregn); static inline void init_query_mad(struct ib_smp *mad) { mad->base_version = 1; mad->mgmt_class = IB_MGMT_CLASS_SUBN_LID_ROUTED; mad->class_version = 1; mad->method = IB_MGMT_METHOD_GET; } static inline u8 convert_access(int acc) { return (acc & IB_ACCESS_REMOTE_ATOMIC ? MLX5_PERM_ATOMIC : 0) | (acc & IB_ACCESS_REMOTE_WRITE ? MLX5_PERM_REMOTE_WRITE : 0) | (acc & IB_ACCESS_REMOTE_READ ? MLX5_PERM_REMOTE_READ : 0) | (acc & IB_ACCESS_LOCAL_WRITE ? MLX5_PERM_LOCAL_WRITE : 0) | MLX5_PERM_LOCAL_READ; } static inline int is_qp1(enum ib_qp_type qp_type) { return qp_type == MLX5_IB_QPT_HW_GSI; } #define MLX5_MAX_UMR_SHIFT 16 #define MLX5_MAX_UMR_PAGES (1 << MLX5_MAX_UMR_SHIFT) static inline u32 check_cq_create_flags(u32 flags) { /* * It returns non-zero value for unsupported CQ * create flags, otherwise it returns zero. */ return (flags & ~(IB_CQ_FLAGS_IGNORE_OVERRUN | IB_CQ_FLAGS_TIMESTAMP_COMPLETION)); } static inline int verify_assign_uidx(u8 cqe_version, u32 cmd_uidx, u32 *user_index) { if (cqe_version) { if ((cmd_uidx == MLX5_IB_DEFAULT_UIDX) || (cmd_uidx & ~MLX5_USER_ASSIGNED_UIDX_MASK)) return -EINVAL; *user_index = cmd_uidx; } else { *user_index = MLX5_IB_DEFAULT_UIDX; } return 0; } static inline int get_qp_user_index(struct mlx5_ib_ucontext *ucontext, struct mlx5_ib_create_qp *ucmd, int inlen, u32 *user_index) { u8 cqe_version = ucontext->cqe_version; if (field_avail(struct mlx5_ib_create_qp, uidx, inlen) && !cqe_version && (ucmd->uidx == MLX5_IB_DEFAULT_UIDX)) return 0; if (!!(field_avail(struct mlx5_ib_create_qp, uidx, inlen) != !!cqe_version)) return -EINVAL; return verify_assign_uidx(cqe_version, ucmd->uidx, user_index); } static inline int get_srq_user_index(struct mlx5_ib_ucontext *ucontext, struct mlx5_ib_create_srq *ucmd, int inlen, u32 *user_index) { u8 cqe_version = ucontext->cqe_version; if (field_avail(struct mlx5_ib_create_srq, uidx, inlen) && !cqe_version && (ucmd->uidx == MLX5_IB_DEFAULT_UIDX)) return 0; if (!!(field_avail(struct mlx5_ib_create_srq, uidx, inlen) != !!cqe_version)) return -EINVAL; return verify_assign_uidx(cqe_version, ucmd->uidx, user_index); } void mlx5_ib_cleanup_congestion(struct mlx5_ib_dev *); int mlx5_ib_init_congestion(struct mlx5_ib_dev *); static inline int get_uars_per_sys_page(struct mlx5_ib_dev *dev, bool lib_support) { return lib_support && MLX5_CAP_GEN(dev->mdev, uar_4k) ? MLX5_UARS_IN_PAGE : 1; } static inline int get_num_static_uars(struct mlx5_ib_dev *dev, struct mlx5_bfreg_info *bfregi) { return get_uars_per_sys_page(dev, bfregi->lib_uar_4k) * bfregi->num_static_sys_pages; } int bfregn_to_uar_index(struct mlx5_ib_dev *dev, struct mlx5_bfreg_info *bfregi, u32 bfregn, bool dyn_bfreg); #endif /* MLX5_IB_H */ diff --git a/sys/dev/mlx5/mlx5_ib/mlx5_ib_main.c b/sys/dev/mlx5/mlx5_ib/mlx5_ib_main.c index f2485d13d791..c2a01b26de1c 100644 --- a/sys/dev/mlx5/mlx5_ib/mlx5_ib_main.c +++ b/sys/dev/mlx5/mlx5_ib/mlx5_ib_main.c @@ -1,3583 +1,3584 @@ /*- * Copyright (c) 2013-2021, Mellanox Technologies, Ltd. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS `AS IS' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * $FreeBSD$ */ #include #include #include #include #include #if defined(CONFIG_X86) #include #endif #include #include #include #undef inode #include #include #include #include #include #include #include #include #include #include #include #include "mlx5_ib.h" #define DRIVER_NAME "mlx5ib" #ifndef DRIVER_VERSION #define DRIVER_VERSION "3.6.0" #endif #define DRIVER_RELDATE "December 2020" MODULE_DESCRIPTION("Mellanox Connect-IB HCA IB driver"); MODULE_LICENSE("Dual BSD/GPL"); MODULE_DEPEND(mlx5ib, linuxkpi, 1, 1, 1); MODULE_DEPEND(mlx5ib, mlx5, 1, 1, 1); MODULE_DEPEND(mlx5ib, ibcore, 1, 1, 1); MODULE_VERSION(mlx5ib, 1); static const char mlx5_version[] = DRIVER_NAME ": Mellanox Connect-IB Infiniband driver " DRIVER_VERSION " (" DRIVER_RELDATE ")\n"; enum { MLX5_ATOMIC_SIZE_QP_8BYTES = 1 << 3, }; static enum rdma_link_layer mlx5_port_type_cap_to_rdma_ll(int port_type_cap) { switch (port_type_cap) { case MLX5_CAP_PORT_TYPE_IB: return IB_LINK_LAYER_INFINIBAND; case MLX5_CAP_PORT_TYPE_ETH: return IB_LINK_LAYER_ETHERNET; default: return IB_LINK_LAYER_UNSPECIFIED; } } static enum rdma_link_layer mlx5_ib_port_link_layer(struct ib_device *device, u8 port_num) { struct mlx5_ib_dev *dev = to_mdev(device); int port_type_cap = MLX5_CAP_GEN(dev->mdev, port_type); return mlx5_port_type_cap_to_rdma_ll(port_type_cap); } -static bool mlx5_netdev_match(struct net_device *ndev, +static bool mlx5_netdev_match(struct ifnet *ndev, struct mlx5_core_dev *mdev, const char *dname) { return ndev->if_type == IFT_ETHER && ndev->if_dname != NULL && strcmp(ndev->if_dname, dname) == 0 && ndev->if_softc != NULL && *(struct mlx5_core_dev **)ndev->if_softc == mdev; } static int mlx5_netdev_event(struct notifier_block *this, unsigned long event, void *ptr) { - struct net_device *ndev = netdev_notifier_info_to_dev(ptr); + struct ifnet *ndev = netdev_notifier_info_to_dev(ptr); struct mlx5_ib_dev *ibdev = container_of(this, struct mlx5_ib_dev, roce.nb); switch (event) { case NETDEV_REGISTER: case NETDEV_UNREGISTER: write_lock(&ibdev->roce.netdev_lock); /* check if network interface belongs to mlx5en */ if (mlx5_netdev_match(ndev, ibdev->mdev, "mce")) ibdev->roce.netdev = (event == NETDEV_UNREGISTER) ? NULL : ndev; write_unlock(&ibdev->roce.netdev_lock); break; case NETDEV_UP: case NETDEV_DOWN: { - struct net_device *upper = NULL; + struct ifnet *upper = NULL; if ((upper == ndev || (!upper && ndev == ibdev->roce.netdev)) && ibdev->ib_active) { struct ib_event ibev = {0}; ibev.device = &ibdev->ib_dev; ibev.event = (event == NETDEV_UP) ? IB_EVENT_PORT_ACTIVE : IB_EVENT_PORT_ERR; ibev.element.port_num = 1; ib_dispatch_event(&ibev); } break; } default: break; } return NOTIFY_DONE; } -static struct net_device *mlx5_ib_get_netdev(struct ib_device *device, +static struct ifnet *mlx5_ib_get_netdev(struct ib_device *device, u8 port_num) { struct mlx5_ib_dev *ibdev = to_mdev(device); - struct net_device *ndev; + struct ifnet *ndev; - /* Ensure ndev does not disappear before we invoke dev_hold() + /* Ensure ndev does not disappear before we invoke if_ref() */ read_lock(&ibdev->roce.netdev_lock); ndev = ibdev->roce.netdev; if (ndev) - dev_hold(ndev); + if_ref(ndev); read_unlock(&ibdev->roce.netdev_lock); return ndev; } static int translate_eth_proto_oper(u32 eth_proto_oper, u8 *active_speed, u8 *active_width) { switch (eth_proto_oper) { case MLX5E_PROT_MASK(MLX5E_1000BASE_CX_SGMII): case MLX5E_PROT_MASK(MLX5E_1000BASE_KX): case MLX5E_PROT_MASK(MLX5E_100BASE_TX): case MLX5E_PROT_MASK(MLX5E_1000BASE_T): *active_width = IB_WIDTH_1X; *active_speed = IB_SPEED_SDR; break; case MLX5E_PROT_MASK(MLX5E_10GBASE_T): case MLX5E_PROT_MASK(MLX5E_10GBASE_CX4): case MLX5E_PROT_MASK(MLX5E_10GBASE_KX4): case MLX5E_PROT_MASK(MLX5E_10GBASE_KR): case MLX5E_PROT_MASK(MLX5E_10GBASE_CR): case MLX5E_PROT_MASK(MLX5E_10GBASE_SR): case MLX5E_PROT_MASK(MLX5E_10GBASE_ER_LR): *active_width = IB_WIDTH_1X; *active_speed = IB_SPEED_QDR; break; case MLX5E_PROT_MASK(MLX5E_25GBASE_CR): case MLX5E_PROT_MASK(MLX5E_25GBASE_KR): case MLX5E_PROT_MASK(MLX5E_25GBASE_SR): *active_width = IB_WIDTH_1X; *active_speed = IB_SPEED_EDR; break; case MLX5E_PROT_MASK(MLX5E_40GBASE_CR4): case MLX5E_PROT_MASK(MLX5E_40GBASE_KR4): case MLX5E_PROT_MASK(MLX5E_40GBASE_SR4): case MLX5E_PROT_MASK(MLX5E_40GBASE_LR4_ER4): *active_width = IB_WIDTH_4X; *active_speed = IB_SPEED_QDR; break; case MLX5E_PROT_MASK(MLX5E_50GBASE_CR2): case MLX5E_PROT_MASK(MLX5E_50GBASE_KR2): case MLX5E_PROT_MASK(MLX5E_50GBASE_KR4): case MLX5E_PROT_MASK(MLX5E_50GBASE_SR2): *active_width = IB_WIDTH_1X; *active_speed = IB_SPEED_HDR; break; case MLX5E_PROT_MASK(MLX5E_56GBASE_R4): *active_width = IB_WIDTH_4X; *active_speed = IB_SPEED_FDR; break; case MLX5E_PROT_MASK(MLX5E_100GBASE_CR4): case MLX5E_PROT_MASK(MLX5E_100GBASE_SR4): case MLX5E_PROT_MASK(MLX5E_100GBASE_KR4): case MLX5E_PROT_MASK(MLX5E_100GBASE_LR4): *active_width = IB_WIDTH_4X; *active_speed = IB_SPEED_EDR; break; default: *active_width = IB_WIDTH_4X; *active_speed = IB_SPEED_QDR; return -EINVAL; } return 0; } static int translate_eth_ext_proto_oper(u32 eth_proto_oper, u8 *active_speed, u8 *active_width) { switch (eth_proto_oper) { case MLX5E_PROT_MASK(MLX5E_SGMII_100M): case MLX5E_PROT_MASK(MLX5E_1000BASE_X_SGMII): *active_width = IB_WIDTH_1X; *active_speed = IB_SPEED_SDR; break; case MLX5E_PROT_MASK(MLX5E_5GBASE_R): *active_width = IB_WIDTH_1X; *active_speed = IB_SPEED_DDR; break; case MLX5E_PROT_MASK(MLX5E_10GBASE_XFI_XAUI_1): *active_width = IB_WIDTH_1X; *active_speed = IB_SPEED_QDR; break; case MLX5E_PROT_MASK(MLX5E_40GBASE_XLAUI_4_XLPPI_4): *active_width = IB_WIDTH_4X; *active_speed = IB_SPEED_QDR; break; case MLX5E_PROT_MASK(MLX5E_25GAUI_1_25GBASE_CR_KR): *active_width = IB_WIDTH_1X; *active_speed = IB_SPEED_EDR; break; case MLX5E_PROT_MASK(MLX5E_50GAUI_2_LAUI_2_50GBASE_CR2_KR2): *active_width = IB_WIDTH_2X; *active_speed = IB_SPEED_EDR; break; case MLX5E_PROT_MASK(MLX5E_50GAUI_1_LAUI_1_50GBASE_CR_KR): *active_width = IB_WIDTH_1X; *active_speed = IB_SPEED_HDR; break; case MLX5E_PROT_MASK(MLX5E_CAUI_4_100GBASE_CR4_KR4): *active_width = IB_WIDTH_4X; *active_speed = IB_SPEED_EDR; break; case MLX5E_PROT_MASK(MLX5E_100GAUI_2_100GBASE_CR2_KR2): *active_width = IB_WIDTH_2X; *active_speed = IB_SPEED_HDR; break; case MLX5E_PROT_MASK(MLX5E_200GAUI_4_200GBASE_CR4_KR4): *active_width = IB_WIDTH_4X; *active_speed = IB_SPEED_HDR; break; default: *active_width = IB_WIDTH_4X; *active_speed = IB_SPEED_QDR; return -EINVAL; } return 0; } static int mlx5_query_port_roce(struct ib_device *device, u8 port_num, struct ib_port_attr *props) { struct mlx5_ib_dev *dev = to_mdev(device); u32 out[MLX5_ST_SZ_DW(ptys_reg)] = {}; - struct net_device *ndev; + struct ifnet *ndev; enum ib_mtu ndev_ib_mtu; u16 qkey_viol_cntr; u32 eth_prot_oper; bool ext; int err; memset(props, 0, sizeof(*props)); /* Possible bad flows are checked before filling out props so in case * of an error it will still be zeroed out. */ err = mlx5_query_port_ptys(dev->mdev, out, sizeof(out), MLX5_PTYS_EN, port_num); if (err) return err; ext = MLX5_CAP_PCAM_FEATURE(dev->mdev, ptys_extended_ethernet); eth_prot_oper = MLX5_GET_ETH_PROTO(ptys_reg, out, ext, eth_proto_oper); if (ext) translate_eth_ext_proto_oper(eth_prot_oper, &props->active_speed, &props->active_width); else translate_eth_proto_oper(eth_prot_oper, &props->active_speed, &props->active_width); props->port_cap_flags |= IB_PORT_CM_SUP; props->port_cap_flags |= IB_PORT_IP_BASED_GIDS; props->gid_tbl_len = MLX5_CAP_ROCE(dev->mdev, roce_address_table_size); props->max_mtu = IB_MTU_4096; props->max_msg_sz = 1 << MLX5_CAP_GEN(dev->mdev, log_max_msg); props->pkey_tbl_len = 1; props->state = IB_PORT_DOWN; props->phys_state = 3; mlx5_query_nic_vport_qkey_viol_cntr(dev->mdev, &qkey_viol_cntr); props->qkey_viol_cntr = qkey_viol_cntr; ndev = mlx5_ib_get_netdev(device, port_num); if (!ndev) return 0; - if (netif_running(ndev) && netif_carrier_ok(ndev)) { + if (ndev->if_drv_flags & IFF_DRV_RUNNING && + ndev->if_link_state == LINK_STATE_UP) { props->state = IB_PORT_ACTIVE; props->phys_state = 5; } ndev_ib_mtu = iboe_get_mtu(ndev->if_mtu); - dev_put(ndev); + if_rele(ndev); props->active_mtu = min(props->max_mtu, ndev_ib_mtu); return 0; } static void ib_gid_to_mlx5_roce_addr(const union ib_gid *gid, const struct ib_gid_attr *attr, void *mlx5_addr) { #define MLX5_SET_RA(p, f, v) MLX5_SET(roce_addr_layout, p, f, v) char *mlx5_addr_l3_addr = MLX5_ADDR_OF(roce_addr_layout, mlx5_addr, source_l3_address); void *mlx5_addr_mac = MLX5_ADDR_OF(roce_addr_layout, mlx5_addr, source_mac_47_32); u16 vlan_id; if (!gid) return; ether_addr_copy(mlx5_addr_mac, IF_LLADDR(attr->ndev)); vlan_id = rdma_vlan_dev_vlan_id(attr->ndev); if (vlan_id != 0xffff) { MLX5_SET_RA(mlx5_addr, vlan_valid, 1); MLX5_SET_RA(mlx5_addr, vlan_id, vlan_id); } switch (attr->gid_type) { case IB_GID_TYPE_IB: MLX5_SET_RA(mlx5_addr, roce_version, MLX5_ROCE_VERSION_1); break; case IB_GID_TYPE_ROCE_UDP_ENCAP: MLX5_SET_RA(mlx5_addr, roce_version, MLX5_ROCE_VERSION_2); break; default: WARN_ON(true); } if (attr->gid_type != IB_GID_TYPE_IB) { if (ipv6_addr_v4mapped((void *)gid)) MLX5_SET_RA(mlx5_addr, roce_l3_type, MLX5_ROCE_L3_TYPE_IPV4); else MLX5_SET_RA(mlx5_addr, roce_l3_type, MLX5_ROCE_L3_TYPE_IPV6); } if ((attr->gid_type == IB_GID_TYPE_IB) || !ipv6_addr_v4mapped((void *)gid)) memcpy(mlx5_addr_l3_addr, gid, sizeof(*gid)); else memcpy(&mlx5_addr_l3_addr[12], &gid->raw[12], 4); } static int set_roce_addr(struct ib_device *device, u8 port_num, unsigned int index, const union ib_gid *gid, const struct ib_gid_attr *attr) { struct mlx5_ib_dev *dev = to_mdev(device); u32 in[MLX5_ST_SZ_DW(set_roce_address_in)] = {0}; u32 out[MLX5_ST_SZ_DW(set_roce_address_out)] = {0}; void *in_addr = MLX5_ADDR_OF(set_roce_address_in, in, roce_address); enum rdma_link_layer ll = mlx5_ib_port_link_layer(device, port_num); if (ll != IB_LINK_LAYER_ETHERNET) return -EINVAL; ib_gid_to_mlx5_roce_addr(gid, attr, in_addr); MLX5_SET(set_roce_address_in, in, roce_address_index, index); MLX5_SET(set_roce_address_in, in, opcode, MLX5_CMD_OP_SET_ROCE_ADDRESS); return mlx5_cmd_exec(dev->mdev, in, sizeof(in), out, sizeof(out)); } static int mlx5_ib_add_gid(struct ib_device *device, u8 port_num, unsigned int index, const union ib_gid *gid, const struct ib_gid_attr *attr, __always_unused void **context) { return set_roce_addr(device, port_num, index, gid, attr); } static int mlx5_ib_del_gid(struct ib_device *device, u8 port_num, unsigned int index, __always_unused void **context) { return set_roce_addr(device, port_num, index, NULL, NULL); } __be16 mlx5_get_roce_udp_sport(struct mlx5_ib_dev *dev, u8 port_num, int index) { struct ib_gid_attr attr; union ib_gid gid; if (ib_get_cached_gid(&dev->ib_dev, port_num, index, &gid, &attr)) return 0; if (!attr.ndev) return 0; - dev_put(attr.ndev); + if_rele(attr.ndev); if (attr.gid_type != IB_GID_TYPE_ROCE_UDP_ENCAP) return 0; return cpu_to_be16(MLX5_CAP_ROCE(dev->mdev, r_roce_min_src_udp_port)); } int mlx5_get_roce_gid_type(struct mlx5_ib_dev *dev, u8 port_num, int index, enum ib_gid_type *gid_type) { struct ib_gid_attr attr; union ib_gid gid; int ret; ret = ib_get_cached_gid(&dev->ib_dev, port_num, index, &gid, &attr); if (ret) return ret; if (!attr.ndev) return -ENODEV; - dev_put(attr.ndev); + if_rele(attr.ndev); *gid_type = attr.gid_type; return 0; } static int mlx5_use_mad_ifc(struct mlx5_ib_dev *dev) { if (MLX5_CAP_GEN(dev->mdev, port_type) == MLX5_CAP_PORT_TYPE_IB) return !MLX5_CAP_GEN(dev->mdev, ib_virt); return 0; } enum { MLX5_VPORT_ACCESS_METHOD_MAD, MLX5_VPORT_ACCESS_METHOD_HCA, MLX5_VPORT_ACCESS_METHOD_NIC, }; static int mlx5_get_vport_access_method(struct ib_device *ibdev) { if (mlx5_use_mad_ifc(to_mdev(ibdev))) return MLX5_VPORT_ACCESS_METHOD_MAD; if (mlx5_ib_port_link_layer(ibdev, 1) == IB_LINK_LAYER_ETHERNET) return MLX5_VPORT_ACCESS_METHOD_NIC; return MLX5_VPORT_ACCESS_METHOD_HCA; } static void get_atomic_caps(struct mlx5_ib_dev *dev, struct ib_device_attr *props) { u8 tmp; u8 atomic_operations = MLX5_CAP_ATOMIC(dev->mdev, atomic_operations); u8 atomic_size_qp = MLX5_CAP_ATOMIC(dev->mdev, atomic_size_qp); u8 atomic_req_8B_endianness_mode = MLX5_CAP_ATOMIC(dev->mdev, atomic_req_8B_endianess_mode); /* Check if HW supports 8 bytes standard atomic operations and capable * of host endianness respond */ tmp = MLX5_ATOMIC_OPS_CMP_SWAP | MLX5_ATOMIC_OPS_FETCH_ADD; if (((atomic_operations & tmp) == tmp) && (atomic_size_qp & MLX5_ATOMIC_SIZE_QP_8BYTES) && (atomic_req_8B_endianness_mode)) { props->atomic_cap = IB_ATOMIC_HCA; } else { props->atomic_cap = IB_ATOMIC_NONE; } } static int mlx5_query_system_image_guid(struct ib_device *ibdev, __be64 *sys_image_guid) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; u64 tmp; int err; switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_system_image_guid(ibdev, sys_image_guid); case MLX5_VPORT_ACCESS_METHOD_HCA: err = mlx5_query_hca_vport_system_image_guid(mdev, &tmp); break; case MLX5_VPORT_ACCESS_METHOD_NIC: err = mlx5_query_nic_vport_system_image_guid(mdev, &tmp); break; default: return -EINVAL; } if (!err) *sys_image_guid = cpu_to_be64(tmp); return err; } static int mlx5_query_max_pkeys(struct ib_device *ibdev, u16 *max_pkeys) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_max_pkeys(ibdev, max_pkeys); case MLX5_VPORT_ACCESS_METHOD_HCA: case MLX5_VPORT_ACCESS_METHOD_NIC: *max_pkeys = mlx5_to_sw_pkey_sz(MLX5_CAP_GEN(mdev, pkey_table_size)); return 0; default: return -EINVAL; } } static int mlx5_query_vendor_id(struct ib_device *ibdev, u32 *vendor_id) { struct mlx5_ib_dev *dev = to_mdev(ibdev); switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_vendor_id(ibdev, vendor_id); case MLX5_VPORT_ACCESS_METHOD_HCA: case MLX5_VPORT_ACCESS_METHOD_NIC: return mlx5_core_query_vendor_id(dev->mdev, vendor_id); default: return -EINVAL; } } static int mlx5_query_node_guid(struct mlx5_ib_dev *dev, __be64 *node_guid) { u64 tmp; int err; switch (mlx5_get_vport_access_method(&dev->ib_dev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_node_guid(dev, node_guid); case MLX5_VPORT_ACCESS_METHOD_HCA: err = mlx5_query_hca_vport_node_guid(dev->mdev, &tmp); break; case MLX5_VPORT_ACCESS_METHOD_NIC: err = mlx5_query_nic_vport_node_guid(dev->mdev, &tmp); break; default: return -EINVAL; } if (!err) *node_guid = cpu_to_be64(tmp); return err; } struct mlx5_reg_node_desc { u8 desc[IB_DEVICE_NODE_DESC_MAX]; }; static int mlx5_query_node_desc(struct mlx5_ib_dev *dev, char *node_desc) { struct mlx5_reg_node_desc in; if (mlx5_use_mad_ifc(dev)) return mlx5_query_mad_ifc_node_desc(dev, node_desc); memset(&in, 0, sizeof(in)); return mlx5_core_access_reg(dev->mdev, &in, sizeof(in), node_desc, sizeof(struct mlx5_reg_node_desc), MLX5_REG_NODE_DESC, 0, 0); } static int mlx5_ib_query_device(struct ib_device *ibdev, struct ib_device_attr *props, struct ib_udata *uhw) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; int err = -ENOMEM; int max_sq_desc; int max_rq_sg; int max_sq_sg; u64 min_page_size = 1ull << MLX5_CAP_GEN(mdev, log_pg_sz); struct mlx5_ib_query_device_resp resp = {}; size_t resp_len; u64 max_tso; resp_len = sizeof(resp.comp_mask) + sizeof(resp.response_length); if (uhw->outlen && uhw->outlen < resp_len) return -EINVAL; else resp.response_length = resp_len; if (uhw->inlen && !ib_is_udata_cleared(uhw, 0, uhw->inlen)) return -EINVAL; memset(props, 0, sizeof(*props)); err = mlx5_query_system_image_guid(ibdev, &props->sys_image_guid); if (err) return err; err = mlx5_query_max_pkeys(ibdev, &props->max_pkeys); if (err) return err; err = mlx5_query_vendor_id(ibdev, &props->vendor_id); if (err) return err; props->fw_ver = ((u64)fw_rev_maj(dev->mdev) << 32) | ((u32)fw_rev_min(dev->mdev) << 16) | fw_rev_sub(dev->mdev); props->device_cap_flags = IB_DEVICE_CHANGE_PHY_PORT | IB_DEVICE_PORT_ACTIVE_EVENT | IB_DEVICE_SYS_IMAGE_GUID | IB_DEVICE_RC_RNR_NAK_GEN; if (MLX5_CAP_GEN(mdev, pkv)) props->device_cap_flags |= IB_DEVICE_BAD_PKEY_CNTR; if (MLX5_CAP_GEN(mdev, qkv)) props->device_cap_flags |= IB_DEVICE_BAD_QKEY_CNTR; if (MLX5_CAP_GEN(mdev, apm)) props->device_cap_flags |= IB_DEVICE_AUTO_PATH_MIG; if (MLX5_CAP_GEN(mdev, xrc)) props->device_cap_flags |= IB_DEVICE_XRC; if (MLX5_CAP_GEN(mdev, imaicl)) { props->device_cap_flags |= IB_DEVICE_MEM_WINDOW | IB_DEVICE_MEM_WINDOW_TYPE_2B; props->max_mw = 1 << MLX5_CAP_GEN(mdev, log_max_mkey); /* We support 'Gappy' memory registration too */ props->device_cap_flags |= IB_DEVICE_SG_GAPS_REG; } props->device_cap_flags |= IB_DEVICE_MEM_MGT_EXTENSIONS; if (MLX5_CAP_GEN(mdev, sho)) { props->device_cap_flags |= IB_DEVICE_SIGNATURE_HANDOVER; /* At this stage no support for signature handover */ props->sig_prot_cap = IB_PROT_T10DIF_TYPE_1 | IB_PROT_T10DIF_TYPE_2 | IB_PROT_T10DIF_TYPE_3; props->sig_guard_cap = IB_GUARD_T10DIF_CRC | IB_GUARD_T10DIF_CSUM; } if (MLX5_CAP_GEN(mdev, block_lb_mc)) props->device_cap_flags |= IB_DEVICE_BLOCK_MULTICAST_LOOPBACK; if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads)) { if (MLX5_CAP_ETH(mdev, csum_cap)) props->device_cap_flags |= IB_DEVICE_RAW_IP_CSUM; if (field_avail(typeof(resp), tso_caps, uhw->outlen)) { max_tso = MLX5_CAP_ETH(mdev, max_lso_cap); if (max_tso) { resp.tso_caps.max_tso = 1 << max_tso; resp.tso_caps.supported_qpts |= 1 << IB_QPT_RAW_PACKET; resp.response_length += sizeof(resp.tso_caps); } } if (field_avail(typeof(resp), rss_caps, uhw->outlen)) { resp.rss_caps.rx_hash_function = MLX5_RX_HASH_FUNC_TOEPLITZ; resp.rss_caps.rx_hash_fields_mask = MLX5_RX_HASH_SRC_IPV4 | MLX5_RX_HASH_DST_IPV4 | MLX5_RX_HASH_SRC_IPV6 | MLX5_RX_HASH_DST_IPV6 | MLX5_RX_HASH_SRC_PORT_TCP | MLX5_RX_HASH_DST_PORT_TCP | MLX5_RX_HASH_SRC_PORT_UDP | MLX5_RX_HASH_DST_PORT_UDP; resp.response_length += sizeof(resp.rss_caps); } } else { if (field_avail(typeof(resp), tso_caps, uhw->outlen)) resp.response_length += sizeof(resp.tso_caps); if (field_avail(typeof(resp), rss_caps, uhw->outlen)) resp.response_length += sizeof(resp.rss_caps); } if (MLX5_CAP_GEN(mdev, ipoib_ipoib_offloads)) { props->device_cap_flags |= IB_DEVICE_UD_IP_CSUM; props->device_cap_flags |= IB_DEVICE_UD_TSO; } if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads) && MLX5_CAP_ETH(dev->mdev, scatter_fcs)) props->device_cap_flags |= IB_DEVICE_RAW_SCATTER_FCS; if (mlx5_get_flow_namespace(dev->mdev, MLX5_FLOW_NAMESPACE_BYPASS)) props->device_cap_flags |= IB_DEVICE_MANAGED_FLOW_STEERING; props->vendor_part_id = mdev->pdev->device; props->hw_ver = mdev->pdev->revision; props->max_mr_size = ~0ull; props->page_size_cap = ~(min_page_size - 1); props->max_qp = 1 << MLX5_CAP_GEN(mdev, log_max_qp); props->max_qp_wr = 1 << MLX5_CAP_GEN(mdev, log_max_qp_sz); max_rq_sg = MLX5_CAP_GEN(mdev, max_wqe_sz_rq) / sizeof(struct mlx5_wqe_data_seg); max_sq_desc = min_t(int, MLX5_CAP_GEN(mdev, max_wqe_sz_sq), 512); max_sq_sg = (max_sq_desc - sizeof(struct mlx5_wqe_ctrl_seg) - sizeof(struct mlx5_wqe_raddr_seg)) / sizeof(struct mlx5_wqe_data_seg); props->max_sge = min(max_rq_sg, max_sq_sg); props->max_sge_rd = MLX5_MAX_SGE_RD; props->max_cq = 1 << MLX5_CAP_GEN(mdev, log_max_cq); props->max_cqe = (1 << MLX5_CAP_GEN(mdev, log_max_cq_sz)) - 1; props->max_mr = 1 << MLX5_CAP_GEN(mdev, log_max_mkey); props->max_pd = 1 << MLX5_CAP_GEN(mdev, log_max_pd); props->max_qp_rd_atom = 1 << MLX5_CAP_GEN(mdev, log_max_ra_req_qp); props->max_qp_init_rd_atom = 1 << MLX5_CAP_GEN(mdev, log_max_ra_res_qp); props->max_srq = 1 << MLX5_CAP_GEN(mdev, log_max_srq); props->max_srq_wr = (1 << MLX5_CAP_GEN(mdev, log_max_srq_sz)) - 1; props->local_ca_ack_delay = MLX5_CAP_GEN(mdev, local_ca_ack_delay); props->max_res_rd_atom = props->max_qp_rd_atom * props->max_qp; props->max_srq_sge = max_rq_sg - 1; props->max_fast_reg_page_list_len = 1 << MLX5_CAP_GEN(mdev, log_max_klm_list_size); get_atomic_caps(dev, props); props->masked_atomic_cap = IB_ATOMIC_NONE; props->max_mcast_grp = 1 << MLX5_CAP_GEN(mdev, log_max_mcg); props->max_mcast_qp_attach = MLX5_CAP_GEN(mdev, max_qp_mcg); props->max_total_mcast_qp_attach = props->max_mcast_qp_attach * props->max_mcast_grp; props->max_map_per_fmr = INT_MAX; /* no limit in ConnectIB */ props->hca_core_clock = MLX5_CAP_GEN(mdev, device_frequency_khz); props->timestamp_mask = 0x7FFFFFFFFFFFFFFFULL; #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING if (MLX5_CAP_GEN(mdev, pg)) props->device_cap_flags |= IB_DEVICE_ON_DEMAND_PAGING; props->odp_caps = dev->odp_caps; #endif if (MLX5_CAP_GEN(mdev, cd)) props->device_cap_flags |= IB_DEVICE_CROSS_CHANNEL; if (!mlx5_core_is_pf(mdev)) props->device_cap_flags |= IB_DEVICE_VIRTUAL_FUNCTION; if (mlx5_ib_port_link_layer(ibdev, 1) == IB_LINK_LAYER_ETHERNET) { props->rss_caps.max_rwq_indirection_tables = 1 << MLX5_CAP_GEN(dev->mdev, log_max_rqt); props->rss_caps.max_rwq_indirection_table_size = 1 << MLX5_CAP_GEN(dev->mdev, log_max_rqt_size); props->rss_caps.supported_qpts = 1 << IB_QPT_RAW_PACKET; props->max_wq_type_rq = 1 << MLX5_CAP_GEN(dev->mdev, log_max_rq); } if (uhw->outlen) { err = ib_copy_to_udata(uhw, &resp, resp.response_length); if (err) return err; } return 0; } enum mlx5_ib_width { MLX5_IB_WIDTH_1X = 1 << 0, MLX5_IB_WIDTH_2X = 1 << 1, MLX5_IB_WIDTH_4X = 1 << 2, MLX5_IB_WIDTH_8X = 1 << 3, MLX5_IB_WIDTH_12X = 1 << 4 }; static int translate_active_width(struct ib_device *ibdev, u8 active_width, u8 *ib_width) { struct mlx5_ib_dev *dev = to_mdev(ibdev); int err = 0; if (active_width & MLX5_IB_WIDTH_1X) { *ib_width = IB_WIDTH_1X; } else if (active_width & MLX5_IB_WIDTH_2X) { *ib_width = IB_WIDTH_2X; } else if (active_width & MLX5_IB_WIDTH_4X) { *ib_width = IB_WIDTH_4X; } else if (active_width & MLX5_IB_WIDTH_8X) { *ib_width = IB_WIDTH_8X; } else if (active_width & MLX5_IB_WIDTH_12X) { *ib_width = IB_WIDTH_12X; } else { mlx5_ib_dbg(dev, "Invalid active_width %d\n", (int)active_width); err = -EINVAL; } return err; } enum ib_max_vl_num { __IB_MAX_VL_0 = 1, __IB_MAX_VL_0_1 = 2, __IB_MAX_VL_0_3 = 3, __IB_MAX_VL_0_7 = 4, __IB_MAX_VL_0_14 = 5, }; enum mlx5_vl_hw_cap { MLX5_VL_HW_0 = 1, MLX5_VL_HW_0_1 = 2, MLX5_VL_HW_0_2 = 3, MLX5_VL_HW_0_3 = 4, MLX5_VL_HW_0_4 = 5, MLX5_VL_HW_0_5 = 6, MLX5_VL_HW_0_6 = 7, MLX5_VL_HW_0_7 = 8, MLX5_VL_HW_0_14 = 15 }; static int translate_max_vl_num(struct ib_device *ibdev, u8 vl_hw_cap, u8 *max_vl_num) { switch (vl_hw_cap) { case MLX5_VL_HW_0: *max_vl_num = __IB_MAX_VL_0; break; case MLX5_VL_HW_0_1: *max_vl_num = __IB_MAX_VL_0_1; break; case MLX5_VL_HW_0_3: *max_vl_num = __IB_MAX_VL_0_3; break; case MLX5_VL_HW_0_7: *max_vl_num = __IB_MAX_VL_0_7; break; case MLX5_VL_HW_0_14: *max_vl_num = __IB_MAX_VL_0_14; break; default: return -EINVAL; } return 0; } static int mlx5_query_hca_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; u32 *rep; int replen = MLX5_ST_SZ_BYTES(query_hca_vport_context_out); struct mlx5_ptys_reg *ptys; struct mlx5_pmtu_reg *pmtu; struct mlx5_pvlc_reg pvlc; void *ctx; int err; rep = mlx5_vzalloc(replen); ptys = kzalloc(sizeof(*ptys), GFP_KERNEL); pmtu = kzalloc(sizeof(*pmtu), GFP_KERNEL); if (!rep || !ptys || !pmtu) { err = -ENOMEM; goto out; } memset(props, 0, sizeof(*props)); err = mlx5_query_hca_vport_context(mdev, port, 0, rep, replen); if (err) goto out; ctx = MLX5_ADDR_OF(query_hca_vport_context_out, rep, hca_vport_context); props->lid = MLX5_GET(hca_vport_context, ctx, lid); props->lmc = MLX5_GET(hca_vport_context, ctx, lmc); props->sm_lid = MLX5_GET(hca_vport_context, ctx, sm_lid); props->sm_sl = MLX5_GET(hca_vport_context, ctx, sm_sl); props->state = MLX5_GET(hca_vport_context, ctx, vport_state); props->phys_state = MLX5_GET(hca_vport_context, ctx, port_physical_state); props->port_cap_flags = MLX5_GET(hca_vport_context, ctx, cap_mask1); props->gid_tbl_len = mlx5_get_gid_table_len(MLX5_CAP_GEN(mdev, gid_table_size)); props->max_msg_sz = 1 << MLX5_CAP_GEN(mdev, log_max_msg); props->pkey_tbl_len = mlx5_to_sw_pkey_sz(MLX5_CAP_GEN(mdev, pkey_table_size)); props->bad_pkey_cntr = MLX5_GET(hca_vport_context, ctx, pkey_violation_counter); props->qkey_viol_cntr = MLX5_GET(hca_vport_context, ctx, qkey_violation_counter); props->subnet_timeout = MLX5_GET(hca_vport_context, ctx, subnet_timeout); props->init_type_reply = MLX5_GET(hca_vport_context, ctx, init_type_reply); props->grh_required = MLX5_GET(hca_vport_context, ctx, grh_required); ptys->proto_mask |= MLX5_PTYS_IB; ptys->local_port = port; err = mlx5_core_access_ptys(mdev, ptys, 0); if (err) goto out; err = translate_active_width(ibdev, ptys->ib_link_width_oper, &props->active_width); if (err) goto out; props->active_speed = (u8)ptys->ib_proto_oper; pmtu->local_port = port; err = mlx5_core_access_pmtu(mdev, pmtu, 0); if (err) goto out; props->max_mtu = pmtu->max_mtu; props->active_mtu = pmtu->oper_mtu; memset(&pvlc, 0, sizeof(pvlc)); pvlc.local_port = port; err = mlx5_core_access_pvlc(mdev, &pvlc, 0); if (err) goto out; err = translate_max_vl_num(ibdev, pvlc.vl_hw_cap, &props->max_vl_num); out: kvfree(rep); kfree(ptys); kfree(pmtu); return err; } int mlx5_ib_query_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props) { switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_port(ibdev, port, props); case MLX5_VPORT_ACCESS_METHOD_HCA: return mlx5_query_hca_port(ibdev, port, props); case MLX5_VPORT_ACCESS_METHOD_NIC: return mlx5_query_port_roce(ibdev, port, props); default: return -EINVAL; } } static int mlx5_ib_query_gid(struct ib_device *ibdev, u8 port, int index, union ib_gid *gid) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_gids(ibdev, port, index, gid); case MLX5_VPORT_ACCESS_METHOD_HCA: return mlx5_query_hca_vport_gid(mdev, port, 0, index, gid); default: return -EINVAL; } } static int mlx5_ib_query_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 *pkey) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_pkey(ibdev, port, index, pkey); case MLX5_VPORT_ACCESS_METHOD_HCA: case MLX5_VPORT_ACCESS_METHOD_NIC: return mlx5_query_hca_vport_pkey(mdev, 0, port, 0, index, pkey); default: return -EINVAL; } } static int mlx5_ib_modify_device(struct ib_device *ibdev, int mask, struct ib_device_modify *props) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_reg_node_desc in; struct mlx5_reg_node_desc out; int err; if (mask & ~IB_DEVICE_MODIFY_NODE_DESC) return -EOPNOTSUPP; if (!(mask & IB_DEVICE_MODIFY_NODE_DESC)) return 0; /* * If possible, pass node desc to FW, so it can generate * a 144 trap. If cmd fails, just ignore. */ memcpy(&in, props->node_desc, IB_DEVICE_NODE_DESC_MAX); err = mlx5_core_access_reg(dev->mdev, &in, sizeof(in), &out, sizeof(out), MLX5_REG_NODE_DESC, 0, 1); if (err) return err; memcpy(ibdev->node_desc, props->node_desc, IB_DEVICE_NODE_DESC_MAX); return err; } static int mlx5_ib_modify_port(struct ib_device *ibdev, u8 port, int mask, struct ib_port_modify *props) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct ib_port_attr attr; u32 tmp; int err; /* * CM layer calls ib_modify_port() regardless of the link * layer. For Ethernet ports, qkey violation and Port * capabilities are meaningless. */ if (mlx5_ib_port_link_layer(ibdev, port) == IB_LINK_LAYER_ETHERNET) return 0; mutex_lock(&dev->cap_mask_mutex); err = mlx5_ib_query_port(ibdev, port, &attr); if (err) goto out; tmp = (attr.port_cap_flags | props->set_port_cap_mask) & ~props->clr_port_cap_mask; err = mlx5_set_port_caps(dev->mdev, port, tmp); out: mutex_unlock(&dev->cap_mask_mutex); return err; } static u16 calc_dynamic_bfregs(int uars_per_sys_page) { /* Large page with non 4k uar support might limit the dynamic size */ if (uars_per_sys_page == 1 && PAGE_SIZE > 4096) return MLX5_MIN_DYN_BFREGS; return MLX5_MAX_DYN_BFREGS; } static int calc_total_bfregs(struct mlx5_ib_dev *dev, bool lib_uar_4k, struct mlx5_ib_alloc_ucontext_req_v2 *req, struct mlx5_bfreg_info *bfregi) { int uars_per_sys_page; int bfregs_per_sys_page; int ref_bfregs = req->total_num_bfregs; if (req->total_num_bfregs == 0) return -EINVAL; BUILD_BUG_ON(MLX5_MAX_BFREGS % MLX5_NON_FP_BFREGS_IN_PAGE); BUILD_BUG_ON(MLX5_MAX_BFREGS < MLX5_NON_FP_BFREGS_IN_PAGE); if (req->total_num_bfregs > MLX5_MAX_BFREGS) return -ENOMEM; uars_per_sys_page = get_uars_per_sys_page(dev, lib_uar_4k); bfregs_per_sys_page = uars_per_sys_page * MLX5_NON_FP_BFREGS_PER_UAR; /* This holds the required static allocation asked by the user */ req->total_num_bfregs = ALIGN(req->total_num_bfregs, bfregs_per_sys_page); if (req->num_low_latency_bfregs > req->total_num_bfregs - 1) return -EINVAL; bfregi->num_static_sys_pages = req->total_num_bfregs / bfregs_per_sys_page; bfregi->num_dyn_bfregs = ALIGN(calc_dynamic_bfregs(uars_per_sys_page), bfregs_per_sys_page); bfregi->total_num_bfregs = req->total_num_bfregs + bfregi->num_dyn_bfregs; bfregi->num_sys_pages = bfregi->total_num_bfregs / bfregs_per_sys_page; mlx5_ib_dbg(dev, "uar_4k: fw support %s, lib support %s, user requested %d bfregs, allocated %d, total bfregs %d, using %d sys pages\n", MLX5_CAP_GEN(dev->mdev, uar_4k) ? "yes" : "no", lib_uar_4k ? "yes" : "no", ref_bfregs, req->total_num_bfregs, bfregi->total_num_bfregs, bfregi->num_sys_pages); return 0; } static int allocate_uars(struct mlx5_ib_dev *dev, struct mlx5_ib_ucontext *context) { struct mlx5_bfreg_info *bfregi; int err; int i; bfregi = &context->bfregi; for (i = 0; i < bfregi->num_static_sys_pages; i++) { err = mlx5_cmd_alloc_uar(dev->mdev, &bfregi->sys_pages[i]); if (err) goto error; mlx5_ib_dbg(dev, "allocated uar %d\n", bfregi->sys_pages[i]); } for (i = bfregi->num_static_sys_pages; i < bfregi->num_sys_pages; i++) bfregi->sys_pages[i] = MLX5_IB_INVALID_UAR_INDEX; return 0; error: for (--i; i >= 0; i--) if (mlx5_cmd_free_uar(dev->mdev, bfregi->sys_pages[i])) mlx5_ib_warn(dev, "failed to free uar %d\n", i); return err; } static void deallocate_uars(struct mlx5_ib_dev *dev, struct mlx5_ib_ucontext *context) { struct mlx5_bfreg_info *bfregi; int i; bfregi = &context->bfregi; for (i = 0; i < bfregi->num_sys_pages; i++) if (i < bfregi->num_static_sys_pages || bfregi->sys_pages[i] != MLX5_IB_INVALID_UAR_INDEX) mlx5_cmd_free_uar(dev->mdev, bfregi->sys_pages[i]); } static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, struct ib_udata *udata) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_ib_alloc_ucontext_req_v2 req = {}; struct mlx5_ib_alloc_ucontext_resp resp = {}; struct mlx5_ib_ucontext *context; struct mlx5_bfreg_info *bfregi; int ver; int err; size_t reqlen; size_t min_req_v2 = offsetof(struct mlx5_ib_alloc_ucontext_req_v2, max_cqe_version); bool lib_uar_4k; bool lib_uar_dyn; if (!dev->ib_active) return ERR_PTR(-EAGAIN); if (udata->inlen < sizeof(struct ib_uverbs_cmd_hdr)) return ERR_PTR(-EINVAL); reqlen = udata->inlen - sizeof(struct ib_uverbs_cmd_hdr); if (reqlen == sizeof(struct mlx5_ib_alloc_ucontext_req)) ver = 0; else if (reqlen >= min_req_v2) ver = 2; else return ERR_PTR(-EINVAL); err = ib_copy_from_udata(&req, udata, min(reqlen, sizeof(req))); if (err) return ERR_PTR(err); if (req.flags) return ERR_PTR(-EINVAL); if (req.comp_mask || req.reserved0 || req.reserved1 || req.reserved2) return ERR_PTR(-EOPNOTSUPP); req.total_num_bfregs = ALIGN(req.total_num_bfregs, MLX5_NON_FP_BFREGS_PER_UAR); if (req.num_low_latency_bfregs > req.total_num_bfregs - 1) return ERR_PTR(-EINVAL); if (reqlen > sizeof(req) && !ib_is_udata_cleared(udata, sizeof(req), reqlen - sizeof(req))) return ERR_PTR(-EOPNOTSUPP); resp.qp_tab_size = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp); if (mlx5_core_is_pf(dev->mdev) && MLX5_CAP_GEN(dev->mdev, bf)) resp.bf_reg_size = 1 << MLX5_CAP_GEN(dev->mdev, log_bf_reg_size); resp.cache_line_size = cache_line_size(); resp.max_sq_desc_sz = MLX5_CAP_GEN(dev->mdev, max_wqe_sz_sq); resp.max_rq_desc_sz = MLX5_CAP_GEN(dev->mdev, max_wqe_sz_rq); resp.max_send_wqebb = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz); resp.max_recv_wr = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz); resp.max_srq_recv_wr = 1 << MLX5_CAP_GEN(dev->mdev, log_max_srq_sz); resp.cqe_version = min_t(__u8, (__u8)MLX5_CAP_GEN(dev->mdev, cqe_version), req.max_cqe_version); resp.log_uar_size = MLX5_CAP_GEN(dev->mdev, uar_4k) ? MLX5_ADAPTER_PAGE_SHIFT : PAGE_SHIFT; resp.num_uars_per_page = MLX5_CAP_GEN(dev->mdev, uar_4k) ? MLX5_CAP_GEN(dev->mdev, num_of_uars_per_page) : 1; resp.response_length = min(offsetof(typeof(resp), response_length) + sizeof(resp.response_length), udata->outlen); context = kzalloc(sizeof(*context), GFP_KERNEL); if (!context) return ERR_PTR(-ENOMEM); lib_uar_4k = req.lib_caps & MLX5_LIB_CAP_4K_UAR; lib_uar_dyn = req.lib_caps & MLX5_LIB_CAP_DYN_UAR; bfregi = &context->bfregi; if (lib_uar_dyn) { bfregi->lib_uar_dyn = lib_uar_dyn; goto uar_done; } /* updates req->total_num_bfregs */ err = calc_total_bfregs(dev, lib_uar_4k, &req, bfregi); if (err) goto out_ctx; mutex_init(&bfregi->lock); bfregi->lib_uar_4k = lib_uar_4k; bfregi->count = kcalloc(bfregi->total_num_bfregs, sizeof(*bfregi->count), GFP_KERNEL); if (!bfregi->count) { err = -ENOMEM; goto out_ctx; } bfregi->sys_pages = kcalloc(bfregi->num_sys_pages, sizeof(*bfregi->sys_pages), GFP_KERNEL); if (!bfregi->sys_pages) { err = -ENOMEM; goto out_count; } err = allocate_uars(dev, context); if (err) goto out_sys_pages; uar_done: #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING context->ibucontext.invalidate_range = &mlx5_ib_invalidate_range; #endif if (MLX5_CAP_GEN(dev->mdev, log_max_transport_domain)) { err = mlx5_alloc_transport_domain(dev->mdev, &context->tdn); if (err) goto out_uars; } INIT_LIST_HEAD(&context->vma_private_list); INIT_LIST_HEAD(&context->db_page_list); mutex_init(&context->db_page_mutex); resp.tot_bfregs = lib_uar_dyn ? 0 : req.total_num_bfregs; resp.num_ports = MLX5_CAP_GEN(dev->mdev, num_ports); if (field_avail(typeof(resp), cqe_version, udata->outlen)) resp.response_length += sizeof(resp.cqe_version); if (field_avail(typeof(resp), cmds_supp_uhw, udata->outlen)) { resp.cmds_supp_uhw |= MLX5_USER_CMDS_SUPP_UHW_QUERY_DEVICE | MLX5_USER_CMDS_SUPP_UHW_CREATE_AH; resp.response_length += sizeof(resp.cmds_supp_uhw); } /* * We don't want to expose information from the PCI bar that is located * after 4096 bytes, so if the arch only supports larger pages, let's * pretend we don't support reading the HCA's core clock. This is also * forced by mmap function. */ if (offsetofend(typeof(resp), hca_core_clock_offset) <= udata->outlen) { if (PAGE_SIZE <= 4096) { resp.comp_mask |= MLX5_IB_ALLOC_UCONTEXT_RESP_MASK_CORE_CLOCK_OFFSET; resp.hca_core_clock_offset = offsetof(struct mlx5_init_seg, internal_timer_h) % PAGE_SIZE; } resp.response_length += sizeof(resp.hca_core_clock_offset); } if (offsetofend(typeof(resp), log_uar_size) <= udata->outlen) resp.response_length += sizeof(resp.log_uar_size); if (offsetofend(typeof(resp), num_uars_per_page) <= udata->outlen) resp.response_length += sizeof(resp.num_uars_per_page); if (offsetofend(typeof(resp), num_dyn_bfregs) <= udata->outlen) { resp.num_dyn_bfregs = bfregi->num_dyn_bfregs; resp.response_length += sizeof(resp.num_dyn_bfregs); } err = ib_copy_to_udata(udata, &resp, resp.response_length); if (err) goto out_td; bfregi->ver = ver; bfregi->num_low_latency_bfregs = req.num_low_latency_bfregs; context->cqe_version = resp.cqe_version; return &context->ibucontext; out_td: if (MLX5_CAP_GEN(dev->mdev, log_max_transport_domain)) mlx5_dealloc_transport_domain(dev->mdev, context->tdn); out_uars: deallocate_uars(dev, context); out_sys_pages: kfree(bfregi->sys_pages); out_count: kfree(bfregi->count); out_ctx: kfree(context); return ERR_PTR(err); } static int mlx5_ib_dealloc_ucontext(struct ib_ucontext *ibcontext) { struct mlx5_ib_ucontext *context = to_mucontext(ibcontext); struct mlx5_ib_dev *dev = to_mdev(ibcontext->device); struct mlx5_bfreg_info *bfregi; bfregi = &context->bfregi; if (MLX5_CAP_GEN(dev->mdev, log_max_transport_domain)) mlx5_dealloc_transport_domain(dev->mdev, context->tdn); deallocate_uars(dev, context); kfree(bfregi->sys_pages); kfree(bfregi->count); kfree(context); return 0; } static phys_addr_t uar_index2pfn(struct mlx5_ib_dev *dev, int uar_idx) { int fw_uars_per_page; fw_uars_per_page = MLX5_CAP_GEN(dev->mdev, uar_4k) ? MLX5_UARS_IN_PAGE : 1; return (pci_resource_start(dev->mdev->pdev, 0) >> PAGE_SHIFT) + uar_idx / fw_uars_per_page; } static int get_command(unsigned long offset) { return (offset >> MLX5_IB_MMAP_CMD_SHIFT) & MLX5_IB_MMAP_CMD_MASK; } static int get_arg(unsigned long offset) { return offset & ((1 << MLX5_IB_MMAP_CMD_SHIFT) - 1); } static int get_index(unsigned long offset) { return get_arg(offset); } /* Index resides in an extra byte to enable larger values than 255 */ static int get_extended_index(unsigned long offset) { return get_arg(offset) | ((offset >> 16) & 0xff) << 8; } static void mlx5_ib_vma_open(struct vm_area_struct *area) { /* vma_open is called when a new VMA is created on top of our VMA. This * is done through either mremap flow or split_vma (usually due to * mlock, madvise, munmap, etc.) We do not support a clone of the VMA, * as this VMA is strongly hardware related. Therefore we set the * vm_ops of the newly created/cloned VMA to NULL, to prevent it from * calling us again and trying to do incorrect actions. We assume that * the original VMA size is exactly a single page, and therefore all * "splitting" operation will not happen to it. */ area->vm_ops = NULL; } static void mlx5_ib_vma_close(struct vm_area_struct *area) { struct mlx5_ib_vma_private_data *mlx5_ib_vma_priv_data; /* It's guaranteed that all VMAs opened on a FD are closed before the * file itself is closed, therefore no sync is needed with the regular * closing flow. (e.g. mlx5 ib_dealloc_ucontext) * However need a sync with accessing the vma as part of * mlx5_ib_disassociate_ucontext. * The close operation is usually called under mm->mmap_sem except when * process is exiting. * The exiting case is handled explicitly as part of * mlx5_ib_disassociate_ucontext. */ mlx5_ib_vma_priv_data = (struct mlx5_ib_vma_private_data *)area->vm_private_data; /* setting the vma context pointer to null in the mlx5_ib driver's * private data, to protect a race condition in * mlx5_ib_disassociate_ucontext(). */ mlx5_ib_vma_priv_data->vma = NULL; list_del(&mlx5_ib_vma_priv_data->list); kfree(mlx5_ib_vma_priv_data); } static const struct vm_operations_struct mlx5_ib_vm_ops = { .open = mlx5_ib_vma_open, .close = mlx5_ib_vma_close }; static int mlx5_ib_set_vma_data(struct vm_area_struct *vma, struct mlx5_ib_ucontext *ctx) { struct mlx5_ib_vma_private_data *vma_prv; struct list_head *vma_head = &ctx->vma_private_list; vma_prv = kzalloc(sizeof(*vma_prv), GFP_KERNEL); if (!vma_prv) return -ENOMEM; vma_prv->vma = vma; vma->vm_private_data = vma_prv; vma->vm_ops = &mlx5_ib_vm_ops; list_add(&vma_prv->list, vma_head); return 0; } static void mlx5_ib_disassociate_ucontext(struct ib_ucontext *ibcontext) { int ret; struct vm_area_struct *vma; struct mlx5_ib_vma_private_data *vma_private, *n; struct mlx5_ib_ucontext *context = to_mucontext(ibcontext); struct task_struct *owning_process = NULL; struct mm_struct *owning_mm = NULL; owning_process = get_pid_task(ibcontext->tgid, PIDTYPE_PID); if (!owning_process) return; owning_mm = get_task_mm(owning_process); if (!owning_mm) { pr_info("no mm, disassociate ucontext is pending task termination\n"); while (1) { put_task_struct(owning_process); usleep_range(1000, 2000); owning_process = get_pid_task(ibcontext->tgid, PIDTYPE_PID); if (!owning_process || owning_process->task_thread-> td_proc->p_state == PRS_ZOMBIE) { pr_info("disassociate ucontext done, task was terminated\n"); /* in case task was dead need to release the * task struct. */ if (owning_process) put_task_struct(owning_process); return; } } } /* need to protect from a race on closing the vma as part of * mlx5_ib_vma_close. */ down_write(&owning_mm->mmap_sem); list_for_each_entry_safe(vma_private, n, &context->vma_private_list, list) { vma = vma_private->vma; ret = zap_vma_ptes(vma, vma->vm_start, PAGE_SIZE); if (ret == -ENOTSUP) { if (bootverbose) WARN_ONCE( "%s: zap_vma_ptes not implemented for unmanaged mappings", __func__); } else { WARN(ret, "%s: zap_vma_ptes failed, error %d", __func__, -ret); } /* context going to be destroyed, should * not access ops any more. */ /* XXXKIB vma->vm_flags &= ~(VM_SHARED | VM_MAYSHARE); */ vma->vm_ops = NULL; list_del(&vma_private->list); kfree(vma_private); } up_write(&owning_mm->mmap_sem); mmput(owning_mm); put_task_struct(owning_process); } static inline char *mmap_cmd2str(enum mlx5_ib_mmap_cmd cmd) { switch (cmd) { case MLX5_IB_MMAP_WC_PAGE: return "WC"; case MLX5_IB_MMAP_REGULAR_PAGE: return "best effort WC"; case MLX5_IB_MMAP_NC_PAGE: return "NC"; default: return NULL; } } static int uar_mmap(struct mlx5_ib_dev *dev, enum mlx5_ib_mmap_cmd cmd, struct vm_area_struct *vma, struct mlx5_ib_ucontext *context) { struct mlx5_bfreg_info *bfregi = &context->bfregi; int err; unsigned long idx; phys_addr_t pfn; pgprot_t prot; u32 bfreg_dyn_idx = 0; u32 uar_index; int dyn_uar = (cmd == MLX5_IB_MMAP_WC_PAGE); int max_valid_idx = dyn_uar ? bfregi->num_sys_pages : bfregi->num_static_sys_pages; if (bfregi->lib_uar_dyn) return -EINVAL; if (vma->vm_end - vma->vm_start != PAGE_SIZE) return -EINVAL; if (dyn_uar) idx = get_extended_index(vma->vm_pgoff) + bfregi->num_static_sys_pages; else idx = get_index(vma->vm_pgoff); if (idx >= max_valid_idx) { mlx5_ib_warn(dev, "invalid uar index %lu, max=%d\n", idx, max_valid_idx); return -EINVAL; } switch (cmd) { case MLX5_IB_MMAP_WC_PAGE: case MLX5_IB_MMAP_REGULAR_PAGE: /* For MLX5_IB_MMAP_REGULAR_PAGE do the best effort to get WC */ prot = pgprot_writecombine(vma->vm_page_prot); break; case MLX5_IB_MMAP_NC_PAGE: prot = pgprot_noncached(vma->vm_page_prot); break; default: return -EINVAL; } if (dyn_uar) { int uars_per_page; uars_per_page = get_uars_per_sys_page(dev, bfregi->lib_uar_4k); bfreg_dyn_idx = idx * (uars_per_page * MLX5_NON_FP_BFREGS_PER_UAR); if (bfreg_dyn_idx >= bfregi->total_num_bfregs) { mlx5_ib_warn(dev, "invalid bfreg_dyn_idx %u, max=%u\n", bfreg_dyn_idx, bfregi->total_num_bfregs); return -EINVAL; } mutex_lock(&bfregi->lock); /* Fail if uar already allocated, first bfreg index of each * page holds its count. */ if (bfregi->count[bfreg_dyn_idx]) { mlx5_ib_warn(dev, "wrong offset, idx %lu is busy, bfregn=%u\n", idx, bfreg_dyn_idx); mutex_unlock(&bfregi->lock); return -EINVAL; } bfregi->count[bfreg_dyn_idx]++; mutex_unlock(&bfregi->lock); err = mlx5_cmd_alloc_uar(dev->mdev, &uar_index); if (err) { mlx5_ib_warn(dev, "UAR alloc failed\n"); goto free_bfreg; } } else { uar_index = bfregi->sys_pages[idx]; } pfn = uar_index2pfn(dev, uar_index); mlx5_ib_dbg(dev, "uar idx 0x%lx, pfn %pa\n", idx, &pfn); vma->vm_page_prot = prot; err = io_remap_pfn_range(vma, vma->vm_start, pfn, PAGE_SIZE, vma->vm_page_prot); if (err) { mlx5_ib_err(dev, "io_remap_pfn_range failed with error=%d, vm_start=0x%llx, pfn=%pa, mmap_cmd=%s\n", err, (unsigned long long)vma->vm_start, &pfn, mmap_cmd2str(cmd)); goto err; } if (dyn_uar) bfregi->sys_pages[idx] = uar_index; return mlx5_ib_set_vma_data(vma, context); err: if (!dyn_uar) return err; mlx5_cmd_free_uar(dev->mdev, idx); free_bfreg: mlx5_ib_free_bfreg(dev, bfregi, bfreg_dyn_idx); return err; } static int mlx5_ib_mmap(struct ib_ucontext *ibcontext, struct vm_area_struct *vma) { struct mlx5_ib_ucontext *context = to_mucontext(ibcontext); struct mlx5_ib_dev *dev = to_mdev(ibcontext->device); unsigned long command; phys_addr_t pfn; command = get_command(vma->vm_pgoff); switch (command) { case MLX5_IB_MMAP_WC_PAGE: case MLX5_IB_MMAP_NC_PAGE: case MLX5_IB_MMAP_REGULAR_PAGE: return uar_mmap(dev, command, vma, context); case MLX5_IB_MMAP_GET_CONTIGUOUS_PAGES: return -ENOSYS; case MLX5_IB_MMAP_CORE_CLOCK: if (vma->vm_end - vma->vm_start != PAGE_SIZE) return -EINVAL; if (vma->vm_flags & VM_WRITE) return -EPERM; /* Don't expose to user-space information it shouldn't have */ if (PAGE_SIZE > 4096) return -EOPNOTSUPP; vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); pfn = (dev->mdev->iseg_base + offsetof(struct mlx5_init_seg, internal_timer_h)) >> PAGE_SHIFT; if (io_remap_pfn_range(vma, vma->vm_start, pfn, PAGE_SIZE, vma->vm_page_prot)) return -EAGAIN; mlx5_ib_dbg(dev, "mapped internal timer at 0x%llx, PA 0x%llx\n", (unsigned long long)vma->vm_start, (unsigned long long)pfn << PAGE_SHIFT); break; default: return -EINVAL; } return 0; } static struct ib_pd *mlx5_ib_alloc_pd(struct ib_device *ibdev, struct ib_ucontext *context, struct ib_udata *udata) { struct mlx5_ib_alloc_pd_resp resp; struct mlx5_ib_pd *pd; int err; pd = kmalloc(sizeof(*pd), GFP_KERNEL); if (!pd) return ERR_PTR(-ENOMEM); err = mlx5_core_alloc_pd(to_mdev(ibdev)->mdev, &pd->pdn); if (err) { kfree(pd); return ERR_PTR(err); } if (context) { resp.pdn = pd->pdn; if (ib_copy_to_udata(udata, &resp, sizeof(resp))) { mlx5_core_dealloc_pd(to_mdev(ibdev)->mdev, pd->pdn); kfree(pd); return ERR_PTR(-EFAULT); } } return &pd->ibpd; } static int mlx5_ib_dealloc_pd(struct ib_pd *pd) { struct mlx5_ib_dev *mdev = to_mdev(pd->device); struct mlx5_ib_pd *mpd = to_mpd(pd); mlx5_core_dealloc_pd(mdev->mdev, mpd->pdn); kfree(mpd); return 0; } enum { MATCH_CRITERIA_ENABLE_OUTER_BIT, MATCH_CRITERIA_ENABLE_MISC_BIT, MATCH_CRITERIA_ENABLE_INNER_BIT }; #define HEADER_IS_ZERO(match_criteria, headers) \ !(memchr_inv(MLX5_ADDR_OF(fte_match_param, match_criteria, headers), \ 0, MLX5_FLD_SZ_BYTES(fte_match_param, headers))) \ static u8 get_match_criteria_enable(u32 *match_criteria) { u8 match_criteria_enable; match_criteria_enable = (!HEADER_IS_ZERO(match_criteria, outer_headers)) << MATCH_CRITERIA_ENABLE_OUTER_BIT; match_criteria_enable |= (!HEADER_IS_ZERO(match_criteria, misc_parameters)) << MATCH_CRITERIA_ENABLE_MISC_BIT; match_criteria_enable |= (!HEADER_IS_ZERO(match_criteria, inner_headers)) << MATCH_CRITERIA_ENABLE_INNER_BIT; return match_criteria_enable; } static void set_proto(void *outer_c, void *outer_v, u8 mask, u8 val) { MLX5_SET(fte_match_set_lyr_2_4, outer_c, ip_protocol, mask); MLX5_SET(fte_match_set_lyr_2_4, outer_v, ip_protocol, val); } static void set_tos(void *outer_c, void *outer_v, u8 mask, u8 val) { MLX5_SET(fte_match_set_lyr_2_4, outer_c, ip_ecn, mask); MLX5_SET(fte_match_set_lyr_2_4, outer_v, ip_ecn, val); MLX5_SET(fte_match_set_lyr_2_4, outer_c, ip_dscp, mask >> 2); MLX5_SET(fte_match_set_lyr_2_4, outer_v, ip_dscp, val >> 2); } #define LAST_ETH_FIELD vlan_tag #define LAST_IB_FIELD sl #define LAST_IPV4_FIELD tos #define LAST_IPV6_FIELD traffic_class #define LAST_TCP_UDP_FIELD src_port /* Field is the last supported field */ #define FIELDS_NOT_SUPPORTED(filter, field)\ memchr_inv((void *)&filter.field +\ sizeof(filter.field), 0,\ sizeof(filter) -\ offsetof(typeof(filter), field) -\ sizeof(filter.field)) static int parse_flow_attr(u32 *match_c, u32 *match_v, const union ib_flow_spec *ib_spec) { void *outer_headers_c = MLX5_ADDR_OF(fte_match_param, match_c, outer_headers); void *outer_headers_v = MLX5_ADDR_OF(fte_match_param, match_v, outer_headers); void *misc_params_c = MLX5_ADDR_OF(fte_match_param, match_c, misc_parameters); void *misc_params_v = MLX5_ADDR_OF(fte_match_param, match_v, misc_parameters); switch (ib_spec->type) { case IB_FLOW_SPEC_ETH: if (FIELDS_NOT_SUPPORTED(ib_spec->eth.mask, LAST_ETH_FIELD)) return -ENOTSUPP; ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, dmac_47_16), ib_spec->eth.mask.dst_mac); ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, dmac_47_16), ib_spec->eth.val.dst_mac); ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, smac_47_16), ib_spec->eth.mask.src_mac); ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, smac_47_16), ib_spec->eth.val.src_mac); if (ib_spec->eth.mask.vlan_tag) { MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, cvlan_tag, 1); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, cvlan_tag, 1); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, first_vid, ntohs(ib_spec->eth.mask.vlan_tag)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, first_vid, ntohs(ib_spec->eth.val.vlan_tag)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, first_cfi, ntohs(ib_spec->eth.mask.vlan_tag) >> 12); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, first_cfi, ntohs(ib_spec->eth.val.vlan_tag) >> 12); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, first_prio, ntohs(ib_spec->eth.mask.vlan_tag) >> 13); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, first_prio, ntohs(ib_spec->eth.val.vlan_tag) >> 13); } MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ethertype, ntohs(ib_spec->eth.mask.ether_type)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ethertype, ntohs(ib_spec->eth.val.ether_type)); break; case IB_FLOW_SPEC_IPV4: if (FIELDS_NOT_SUPPORTED(ib_spec->ipv4.mask, LAST_IPV4_FIELD)) return -ENOTSUPP; MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ethertype, 0xffff); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ethertype, ETH_P_IP); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, src_ipv4_src_ipv6.ipv4_layout.ipv4), &ib_spec->ipv4.mask.src_ip, sizeof(ib_spec->ipv4.mask.src_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, src_ipv4_src_ipv6.ipv4_layout.ipv4), &ib_spec->ipv4.val.src_ip, sizeof(ib_spec->ipv4.val.src_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, dst_ipv4_dst_ipv6.ipv4_layout.ipv4), &ib_spec->ipv4.mask.dst_ip, sizeof(ib_spec->ipv4.mask.dst_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, dst_ipv4_dst_ipv6.ipv4_layout.ipv4), &ib_spec->ipv4.val.dst_ip, sizeof(ib_spec->ipv4.val.dst_ip)); set_tos(outer_headers_c, outer_headers_v, ib_spec->ipv4.mask.tos, ib_spec->ipv4.val.tos); set_proto(outer_headers_c, outer_headers_v, ib_spec->ipv4.mask.proto, ib_spec->ipv4.val.proto); break; case IB_FLOW_SPEC_IPV6: if (FIELDS_NOT_SUPPORTED(ib_spec->ipv6.mask, LAST_IPV6_FIELD)) return -ENOTSUPP; MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ethertype, 0xffff); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ethertype, IPPROTO_IPV6); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, src_ipv4_src_ipv6.ipv6_layout.ipv6), &ib_spec->ipv6.mask.src_ip, sizeof(ib_spec->ipv6.mask.src_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, src_ipv4_src_ipv6.ipv6_layout.ipv6), &ib_spec->ipv6.val.src_ip, sizeof(ib_spec->ipv6.val.src_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, dst_ipv4_dst_ipv6.ipv6_layout.ipv6), &ib_spec->ipv6.mask.dst_ip, sizeof(ib_spec->ipv6.mask.dst_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, dst_ipv4_dst_ipv6.ipv6_layout.ipv6), &ib_spec->ipv6.val.dst_ip, sizeof(ib_spec->ipv6.val.dst_ip)); set_tos(outer_headers_c, outer_headers_v, ib_spec->ipv6.mask.traffic_class, ib_spec->ipv6.val.traffic_class); set_proto(outer_headers_c, outer_headers_v, ib_spec->ipv6.mask.next_hdr, ib_spec->ipv6.val.next_hdr); MLX5_SET(fte_match_set_misc, misc_params_c, outer_ipv6_flow_label, ntohl(ib_spec->ipv6.mask.flow_label)); MLX5_SET(fte_match_set_misc, misc_params_v, outer_ipv6_flow_label, ntohl(ib_spec->ipv6.val.flow_label)); break; case IB_FLOW_SPEC_TCP: if (FIELDS_NOT_SUPPORTED(ib_spec->tcp_udp.mask, LAST_TCP_UDP_FIELD)) return -ENOTSUPP; MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ip_protocol, 0xff); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ip_protocol, IPPROTO_TCP); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, tcp_sport, ntohs(ib_spec->tcp_udp.mask.src_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, tcp_sport, ntohs(ib_spec->tcp_udp.val.src_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, tcp_dport, ntohs(ib_spec->tcp_udp.mask.dst_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, tcp_dport, ntohs(ib_spec->tcp_udp.val.dst_port)); break; case IB_FLOW_SPEC_UDP: if (FIELDS_NOT_SUPPORTED(ib_spec->tcp_udp.mask, LAST_TCP_UDP_FIELD)) return -ENOTSUPP; MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ip_protocol, 0xff); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ip_protocol, IPPROTO_UDP); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, udp_sport, ntohs(ib_spec->tcp_udp.mask.src_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, udp_sport, ntohs(ib_spec->tcp_udp.val.src_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, udp_dport, ntohs(ib_spec->tcp_udp.mask.dst_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, udp_dport, ntohs(ib_spec->tcp_udp.val.dst_port)); break; default: return -EINVAL; } return 0; } /* If a flow could catch both multicast and unicast packets, * it won't fall into the multicast flow steering table and this rule * could steal other multicast packets. */ static bool flow_is_multicast_only(struct ib_flow_attr *ib_attr) { struct ib_flow_spec_eth *eth_spec; if (ib_attr->type != IB_FLOW_ATTR_NORMAL || ib_attr->size < sizeof(struct ib_flow_attr) + sizeof(struct ib_flow_spec_eth) || ib_attr->num_of_specs < 1) return false; eth_spec = (struct ib_flow_spec_eth *)(ib_attr + 1); if (eth_spec->type != IB_FLOW_SPEC_ETH || eth_spec->size != sizeof(*eth_spec)) return false; return is_multicast_ether_addr(eth_spec->mask.dst_mac) && is_multicast_ether_addr(eth_spec->val.dst_mac); } static bool is_valid_attr(const struct ib_flow_attr *flow_attr) { union ib_flow_spec *ib_spec = (union ib_flow_spec *)(flow_attr + 1); bool has_ipv4_spec = false; bool eth_type_ipv4 = true; unsigned int spec_index; /* Validate that ethertype is correct */ for (spec_index = 0; spec_index < flow_attr->num_of_specs; spec_index++) { if (ib_spec->type == IB_FLOW_SPEC_ETH && ib_spec->eth.mask.ether_type) { if (!((ib_spec->eth.mask.ether_type == htons(0xffff)) && ib_spec->eth.val.ether_type == htons(ETH_P_IP))) eth_type_ipv4 = false; } else if (ib_spec->type == IB_FLOW_SPEC_IPV4) { has_ipv4_spec = true; } ib_spec = (void *)ib_spec + ib_spec->size; } return !has_ipv4_spec || eth_type_ipv4; } static void put_flow_table(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_prio *prio, bool ft_added) { prio->refcount -= !!ft_added; if (!prio->refcount) { mlx5_destroy_flow_table(prio->flow_table); prio->flow_table = NULL; } } static int mlx5_ib_destroy_flow(struct ib_flow *flow_id) { struct mlx5_ib_dev *dev = to_mdev(flow_id->qp->device); struct mlx5_ib_flow_handler *handler = container_of(flow_id, struct mlx5_ib_flow_handler, ibflow); struct mlx5_ib_flow_handler *iter, *tmp; mutex_lock(&dev->flow_db.lock); list_for_each_entry_safe(iter, tmp, &handler->list, list) { mlx5_del_flow_rule(iter->rule); put_flow_table(dev, iter->prio, true); list_del(&iter->list); kfree(iter); } mlx5_del_flow_rule(handler->rule); put_flow_table(dev, handler->prio, true); mutex_unlock(&dev->flow_db.lock); kfree(handler); return 0; } static int ib_prio_to_core_prio(unsigned int priority, bool dont_trap) { priority *= 2; if (!dont_trap) priority++; return priority; } enum flow_table_type { MLX5_IB_FT_RX, MLX5_IB_FT_TX }; #define MLX5_FS_MAX_TYPES 10 #define MLX5_FS_MAX_ENTRIES 32000UL static struct mlx5_ib_flow_prio *get_flow_table(struct mlx5_ib_dev *dev, struct ib_flow_attr *flow_attr, enum flow_table_type ft_type) { bool dont_trap = flow_attr->flags & IB_FLOW_ATTR_FLAGS_DONT_TRAP; struct mlx5_flow_namespace *ns = NULL; struct mlx5_ib_flow_prio *prio; struct mlx5_flow_table *ft; int num_entries; int num_groups; int priority; int err = 0; if (flow_attr->type == IB_FLOW_ATTR_NORMAL) { if (flow_is_multicast_only(flow_attr) && !dont_trap) priority = MLX5_IB_FLOW_MCAST_PRIO; else priority = ib_prio_to_core_prio(flow_attr->priority, dont_trap); ns = mlx5_get_flow_namespace(dev->mdev, MLX5_FLOW_NAMESPACE_BYPASS); num_entries = MLX5_FS_MAX_ENTRIES; num_groups = MLX5_FS_MAX_TYPES; prio = &dev->flow_db.prios[priority]; } else if (flow_attr->type == IB_FLOW_ATTR_ALL_DEFAULT || flow_attr->type == IB_FLOW_ATTR_MC_DEFAULT) { ns = mlx5_get_flow_namespace(dev->mdev, MLX5_FLOW_NAMESPACE_LEFTOVERS); build_leftovers_ft_param("bypass", &priority, &num_entries, &num_groups); prio = &dev->flow_db.prios[MLX5_IB_FLOW_LEFTOVERS_PRIO]; } else if (flow_attr->type == IB_FLOW_ATTR_SNIFFER) { if (!MLX5_CAP_FLOWTABLE(dev->mdev, allow_sniffer_and_nic_rx_shared_tir)) return ERR_PTR(-ENOTSUPP); ns = mlx5_get_flow_namespace(dev->mdev, ft_type == MLX5_IB_FT_RX ? MLX5_FLOW_NAMESPACE_SNIFFER_RX : MLX5_FLOW_NAMESPACE_SNIFFER_TX); prio = &dev->flow_db.sniffer[ft_type]; priority = 0; num_entries = 1; num_groups = 1; } if (!ns) return ERR_PTR(-ENOTSUPP); ft = prio->flow_table; if (!ft) { ft = mlx5_create_auto_grouped_flow_table(ns, priority, "bypass", num_entries, num_groups); if (!IS_ERR(ft)) { prio->refcount = 0; prio->flow_table = ft; } else { err = PTR_ERR(ft); } } return err ? ERR_PTR(err) : prio; } static struct mlx5_ib_flow_handler *create_flow_rule(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_prio *ft_prio, const struct ib_flow_attr *flow_attr, struct mlx5_flow_destination *dst) { struct mlx5_flow_table *ft = ft_prio->flow_table; struct mlx5_ib_flow_handler *handler; struct mlx5_flow_spec *spec; const void *ib_flow = (const void *)flow_attr + sizeof(*flow_attr); unsigned int spec_index; u32 action; int err = 0; if (!is_valid_attr(flow_attr)) return ERR_PTR(-EINVAL); spec = mlx5_vzalloc(sizeof(*spec)); handler = kzalloc(sizeof(*handler), GFP_KERNEL); if (!handler || !spec) { err = -ENOMEM; goto free; } INIT_LIST_HEAD(&handler->list); for (spec_index = 0; spec_index < flow_attr->num_of_specs; spec_index++) { err = parse_flow_attr(spec->match_criteria, spec->match_value, ib_flow); if (err < 0) goto free; ib_flow += ((union ib_flow_spec *)ib_flow)->size; } spec->match_criteria_enable = get_match_criteria_enable(spec->match_criteria); action = dst ? MLX5_FLOW_CONTEXT_ACTION_FWD_DEST : MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO; handler->rule = mlx5_add_flow_rule(ft, spec->match_criteria_enable, spec->match_criteria, spec->match_value, action, MLX5_FS_DEFAULT_FLOW_TAG, dst); if (IS_ERR(handler->rule)) { err = PTR_ERR(handler->rule); goto free; } ft_prio->refcount++; handler->prio = ft_prio; ft_prio->flow_table = ft; free: if (err) kfree(handler); kvfree(spec); return err ? ERR_PTR(err) : handler; } static struct mlx5_ib_flow_handler *create_dont_trap_rule(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_prio *ft_prio, struct ib_flow_attr *flow_attr, struct mlx5_flow_destination *dst) { struct mlx5_ib_flow_handler *handler_dst = NULL; struct mlx5_ib_flow_handler *handler = NULL; handler = create_flow_rule(dev, ft_prio, flow_attr, NULL); if (!IS_ERR(handler)) { handler_dst = create_flow_rule(dev, ft_prio, flow_attr, dst); if (IS_ERR(handler_dst)) { mlx5_del_flow_rule(handler->rule); ft_prio->refcount--; kfree(handler); handler = handler_dst; } else { list_add(&handler_dst->list, &handler->list); } } return handler; } enum { LEFTOVERS_MC, LEFTOVERS_UC, }; static struct mlx5_ib_flow_handler *create_leftovers_rule(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_prio *ft_prio, struct ib_flow_attr *flow_attr, struct mlx5_flow_destination *dst) { struct mlx5_ib_flow_handler *handler_ucast = NULL; struct mlx5_ib_flow_handler *handler = NULL; static struct { struct ib_flow_attr flow_attr; struct ib_flow_spec_eth eth_flow; } leftovers_specs[] = { [LEFTOVERS_MC] = { .flow_attr = { .num_of_specs = 1, .size = sizeof(leftovers_specs[0]) }, .eth_flow = { .type = IB_FLOW_SPEC_ETH, .size = sizeof(struct ib_flow_spec_eth), .mask = {.dst_mac = {0x1} }, .val = {.dst_mac = {0x1} } } }, [LEFTOVERS_UC] = { .flow_attr = { .num_of_specs = 1, .size = sizeof(leftovers_specs[0]) }, .eth_flow = { .type = IB_FLOW_SPEC_ETH, .size = sizeof(struct ib_flow_spec_eth), .mask = {.dst_mac = {0x1} }, .val = {.dst_mac = {} } } } }; handler = create_flow_rule(dev, ft_prio, &leftovers_specs[LEFTOVERS_MC].flow_attr, dst); if (!IS_ERR(handler) && flow_attr->type == IB_FLOW_ATTR_ALL_DEFAULT) { handler_ucast = create_flow_rule(dev, ft_prio, &leftovers_specs[LEFTOVERS_UC].flow_attr, dst); if (IS_ERR(handler_ucast)) { mlx5_del_flow_rule(handler->rule); ft_prio->refcount--; kfree(handler); handler = handler_ucast; } else { list_add(&handler_ucast->list, &handler->list); } } return handler; } static struct mlx5_ib_flow_handler *create_sniffer_rule(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_prio *ft_rx, struct mlx5_ib_flow_prio *ft_tx, struct mlx5_flow_destination *dst) { struct mlx5_ib_flow_handler *handler_rx; struct mlx5_ib_flow_handler *handler_tx; int err; static const struct ib_flow_attr flow_attr = { .num_of_specs = 0, .size = sizeof(flow_attr) }; handler_rx = create_flow_rule(dev, ft_rx, &flow_attr, dst); if (IS_ERR(handler_rx)) { err = PTR_ERR(handler_rx); goto err; } handler_tx = create_flow_rule(dev, ft_tx, &flow_attr, dst); if (IS_ERR(handler_tx)) { err = PTR_ERR(handler_tx); goto err_tx; } list_add(&handler_tx->list, &handler_rx->list); return handler_rx; err_tx: mlx5_del_flow_rule(handler_rx->rule); ft_rx->refcount--; kfree(handler_rx); err: return ERR_PTR(err); } static struct ib_flow *mlx5_ib_create_flow(struct ib_qp *qp, struct ib_flow_attr *flow_attr, int domain) { struct mlx5_ib_dev *dev = to_mdev(qp->device); struct mlx5_ib_qp *mqp = to_mqp(qp); struct mlx5_ib_flow_handler *handler = NULL; struct mlx5_flow_destination *dst = NULL; struct mlx5_ib_flow_prio *ft_prio_tx = NULL; struct mlx5_ib_flow_prio *ft_prio; int err; if (flow_attr->priority > MLX5_IB_FLOW_LAST_PRIO) return ERR_PTR(-ENOSPC); if (domain != IB_FLOW_DOMAIN_USER || flow_attr->port > MLX5_CAP_GEN(dev->mdev, num_ports) || (flow_attr->flags & ~IB_FLOW_ATTR_FLAGS_DONT_TRAP)) return ERR_PTR(-EINVAL); dst = kzalloc(sizeof(*dst), GFP_KERNEL); if (!dst) return ERR_PTR(-ENOMEM); mutex_lock(&dev->flow_db.lock); ft_prio = get_flow_table(dev, flow_attr, MLX5_IB_FT_RX); if (IS_ERR(ft_prio)) { err = PTR_ERR(ft_prio); goto unlock; } if (flow_attr->type == IB_FLOW_ATTR_SNIFFER) { ft_prio_tx = get_flow_table(dev, flow_attr, MLX5_IB_FT_TX); if (IS_ERR(ft_prio_tx)) { err = PTR_ERR(ft_prio_tx); ft_prio_tx = NULL; goto destroy_ft; } } dst->type = MLX5_FLOW_DESTINATION_TYPE_TIR; if (mqp->flags & MLX5_IB_QP_RSS) dst->tir_num = mqp->rss_qp.tirn; else dst->tir_num = mqp->raw_packet_qp.rq.tirn; if (flow_attr->type == IB_FLOW_ATTR_NORMAL) { if (flow_attr->flags & IB_FLOW_ATTR_FLAGS_DONT_TRAP) { handler = create_dont_trap_rule(dev, ft_prio, flow_attr, dst); } else { handler = create_flow_rule(dev, ft_prio, flow_attr, dst); } } else if (flow_attr->type == IB_FLOW_ATTR_ALL_DEFAULT || flow_attr->type == IB_FLOW_ATTR_MC_DEFAULT) { handler = create_leftovers_rule(dev, ft_prio, flow_attr, dst); } else if (flow_attr->type == IB_FLOW_ATTR_SNIFFER) { handler = create_sniffer_rule(dev, ft_prio, ft_prio_tx, dst); } else { err = -EINVAL; goto destroy_ft; } if (IS_ERR(handler)) { err = PTR_ERR(handler); handler = NULL; goto destroy_ft; } mutex_unlock(&dev->flow_db.lock); kfree(dst); return &handler->ibflow; destroy_ft: put_flow_table(dev, ft_prio, false); if (ft_prio_tx) put_flow_table(dev, ft_prio_tx, false); unlock: mutex_unlock(&dev->flow_db.lock); kfree(dst); kfree(handler); return ERR_PTR(err); } static int mlx5_ib_mcg_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) { struct mlx5_ib_dev *dev = to_mdev(ibqp->device); int err; err = mlx5_core_attach_mcg(dev->mdev, gid, ibqp->qp_num); if (err) mlx5_ib_warn(dev, "failed attaching QPN 0x%x, MGID %pI6\n", ibqp->qp_num, gid->raw); return err; } static int mlx5_ib_mcg_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) { struct mlx5_ib_dev *dev = to_mdev(ibqp->device); int err; err = mlx5_core_detach_mcg(dev->mdev, gid, ibqp->qp_num); if (err) mlx5_ib_warn(dev, "failed detaching QPN 0x%x, MGID %pI6\n", ibqp->qp_num, gid->raw); return err; } static int init_node_data(struct mlx5_ib_dev *dev) { int err; err = mlx5_query_node_desc(dev, dev->ib_dev.node_desc); if (err) return err; return mlx5_query_node_guid(dev, &dev->ib_dev.node_guid); } static ssize_t show_fw_pages(struct device *device, struct device_attribute *attr, char *buf) { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); return sprintf(buf, "%lld\n", (long long)dev->mdev->priv.fw_pages); } static ssize_t show_reg_pages(struct device *device, struct device_attribute *attr, char *buf) { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); return sprintf(buf, "%d\n", atomic_read(&dev->mdev->priv.reg_pages)); } static ssize_t show_hca(struct device *device, struct device_attribute *attr, char *buf) { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); return sprintf(buf, "MT%d\n", dev->mdev->pdev->device); } static ssize_t show_rev(struct device *device, struct device_attribute *attr, char *buf) { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); return sprintf(buf, "%x\n", dev->mdev->pdev->revision); } static ssize_t show_board(struct device *device, struct device_attribute *attr, char *buf) { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); return sprintf(buf, "%.*s\n", MLX5_BOARD_ID_LEN, dev->mdev->board_id); } static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL); static DEVICE_ATTR(hca_type, S_IRUGO, show_hca, NULL); static DEVICE_ATTR(board_id, S_IRUGO, show_board, NULL); static DEVICE_ATTR(fw_pages, S_IRUGO, show_fw_pages, NULL); static DEVICE_ATTR(reg_pages, S_IRUGO, show_reg_pages, NULL); static struct device_attribute *mlx5_class_attributes[] = { &dev_attr_hw_rev, &dev_attr_hca_type, &dev_attr_board_id, &dev_attr_fw_pages, &dev_attr_reg_pages, }; static void pkey_change_handler(struct work_struct *work) { struct mlx5_ib_port_resources *ports = container_of(work, struct mlx5_ib_port_resources, pkey_change_work); mutex_lock(&ports->devr->mutex); mlx5_ib_gsi_pkey_change(ports->gsi); mutex_unlock(&ports->devr->mutex); } static void mlx5_ib_handle_internal_error(struct mlx5_ib_dev *ibdev) { struct mlx5_ib_qp *mqp; struct mlx5_ib_cq *send_mcq, *recv_mcq; struct mlx5_core_cq *mcq; struct list_head cq_armed_list; unsigned long flags_qp; unsigned long flags_cq; unsigned long flags; INIT_LIST_HEAD(&cq_armed_list); /* Go over qp list reside on that ibdev, sync with create/destroy qp.*/ spin_lock_irqsave(&ibdev->reset_flow_resource_lock, flags); list_for_each_entry(mqp, &ibdev->qp_list, qps_list) { spin_lock_irqsave(&mqp->sq.lock, flags_qp); if (mqp->sq.tail != mqp->sq.head) { send_mcq = to_mcq(mqp->ibqp.send_cq); spin_lock_irqsave(&send_mcq->lock, flags_cq); if (send_mcq->mcq.comp && mqp->ibqp.send_cq->comp_handler) { if (!send_mcq->mcq.reset_notify_added) { send_mcq->mcq.reset_notify_added = 1; list_add_tail(&send_mcq->mcq.reset_notify, &cq_armed_list); } } spin_unlock_irqrestore(&send_mcq->lock, flags_cq); } spin_unlock_irqrestore(&mqp->sq.lock, flags_qp); spin_lock_irqsave(&mqp->rq.lock, flags_qp); /* no handling is needed for SRQ */ if (!mqp->ibqp.srq) { if (mqp->rq.tail != mqp->rq.head) { recv_mcq = to_mcq(mqp->ibqp.recv_cq); spin_lock_irqsave(&recv_mcq->lock, flags_cq); if (recv_mcq->mcq.comp && mqp->ibqp.recv_cq->comp_handler) { if (!recv_mcq->mcq.reset_notify_added) { recv_mcq->mcq.reset_notify_added = 1; list_add_tail(&recv_mcq->mcq.reset_notify, &cq_armed_list); } } spin_unlock_irqrestore(&recv_mcq->lock, flags_cq); } } spin_unlock_irqrestore(&mqp->rq.lock, flags_qp); } /*At that point all inflight post send were put to be executed as of we * lock/unlock above locks Now need to arm all involved CQs. */ list_for_each_entry(mcq, &cq_armed_list, reset_notify) { mcq->comp(mcq, NULL); } spin_unlock_irqrestore(&ibdev->reset_flow_resource_lock, flags); } static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context, enum mlx5_dev_event event, unsigned long param) { struct mlx5_ib_dev *ibdev = (struct mlx5_ib_dev *)context; struct ib_event ibev; bool fatal = false; u8 port = (u8)param; switch (event) { case MLX5_DEV_EVENT_SYS_ERROR: ibev.event = IB_EVENT_DEVICE_FATAL; mlx5_ib_handle_internal_error(ibdev); fatal = true; break; case MLX5_DEV_EVENT_PORT_UP: case MLX5_DEV_EVENT_PORT_DOWN: case MLX5_DEV_EVENT_PORT_INITIALIZED: /* In RoCE, port up/down events are handled in * mlx5_netdev_event(). */ if (mlx5_ib_port_link_layer(&ibdev->ib_dev, port) == IB_LINK_LAYER_ETHERNET) return; ibev.event = (event == MLX5_DEV_EVENT_PORT_UP) ? IB_EVENT_PORT_ACTIVE : IB_EVENT_PORT_ERR; break; case MLX5_DEV_EVENT_LID_CHANGE: ibev.event = IB_EVENT_LID_CHANGE; break; case MLX5_DEV_EVENT_PKEY_CHANGE: ibev.event = IB_EVENT_PKEY_CHANGE; schedule_work(&ibdev->devr.ports[port - 1].pkey_change_work); break; case MLX5_DEV_EVENT_GUID_CHANGE: ibev.event = IB_EVENT_GID_CHANGE; break; case MLX5_DEV_EVENT_CLIENT_REREG: ibev.event = IB_EVENT_CLIENT_REREGISTER; break; default: /* unsupported event */ return; } ibev.device = &ibdev->ib_dev; ibev.element.port_num = port; if (!rdma_is_port_valid(&ibdev->ib_dev, port)) { mlx5_ib_warn(ibdev, "warning: event(%d) on port %d\n", event, port); return; } if (ibdev->ib_active) ib_dispatch_event(&ibev); if (fatal) ibdev->ib_active = false; } static void get_ext_port_caps(struct mlx5_ib_dev *dev) { int port; for (port = 1; port <= MLX5_CAP_GEN(dev->mdev, num_ports); port++) mlx5_query_ext_port_caps(dev, port); } static int get_port_caps(struct mlx5_ib_dev *dev) { struct ib_device_attr *dprops = NULL; struct ib_port_attr *pprops = NULL; int err = -ENOMEM; int port; struct ib_udata uhw = {.inlen = 0, .outlen = 0}; pprops = kmalloc(sizeof(*pprops), GFP_KERNEL); if (!pprops) goto out; dprops = kmalloc(sizeof(*dprops), GFP_KERNEL); if (!dprops) goto out; err = mlx5_ib_query_device(&dev->ib_dev, dprops, &uhw); if (err) { mlx5_ib_warn(dev, "query_device failed %d\n", err); goto out; } for (port = 1; port <= MLX5_CAP_GEN(dev->mdev, num_ports); port++) { err = mlx5_ib_query_port(&dev->ib_dev, port, pprops); if (err) { mlx5_ib_warn(dev, "query_port %d failed %d\n", port, err); break; } dev->mdev->port_caps[port - 1].pkey_table_len = dprops->max_pkeys; dev->mdev->port_caps[port - 1].gid_table_len = pprops->gid_tbl_len; mlx5_ib_dbg(dev, "pkey_table_len %d, gid_table_len %d\n", dprops->max_pkeys, pprops->gid_tbl_len); } out: kfree(pprops); kfree(dprops); return err; } static void destroy_umrc_res(struct mlx5_ib_dev *dev) { int err; err = mlx5_mr_cache_cleanup(dev); if (err) mlx5_ib_warn(dev, "mr cache cleanup failed\n"); mlx5_ib_destroy_qp(dev->umrc.qp); ib_free_cq(dev->umrc.cq); ib_dealloc_pd(dev->umrc.pd); } enum { MAX_UMR_WR = 128, }; static int create_umr_res(struct mlx5_ib_dev *dev) { struct ib_qp_init_attr *init_attr = NULL; struct ib_qp_attr *attr = NULL; struct ib_pd *pd; struct ib_cq *cq; struct ib_qp *qp; int ret; attr = kzalloc(sizeof(*attr), GFP_KERNEL); init_attr = kzalloc(sizeof(*init_attr), GFP_KERNEL); if (!attr || !init_attr) { ret = -ENOMEM; goto error_0; } pd = ib_alloc_pd(&dev->ib_dev, 0); if (IS_ERR(pd)) { mlx5_ib_dbg(dev, "Couldn't create PD for sync UMR QP\n"); ret = PTR_ERR(pd); goto error_0; } cq = ib_alloc_cq(&dev->ib_dev, NULL, 128, 0, IB_POLL_SOFTIRQ); if (IS_ERR(cq)) { mlx5_ib_dbg(dev, "Couldn't create CQ for sync UMR QP\n"); ret = PTR_ERR(cq); goto error_2; } init_attr->send_cq = cq; init_attr->recv_cq = cq; init_attr->sq_sig_type = IB_SIGNAL_ALL_WR; init_attr->cap.max_send_wr = MAX_UMR_WR; init_attr->cap.max_send_sge = 1; init_attr->qp_type = MLX5_IB_QPT_REG_UMR; init_attr->port_num = 1; qp = mlx5_ib_create_qp(pd, init_attr, NULL); if (IS_ERR(qp)) { mlx5_ib_dbg(dev, "Couldn't create sync UMR QP\n"); ret = PTR_ERR(qp); goto error_3; } qp->device = &dev->ib_dev; qp->real_qp = qp; qp->uobject = NULL; qp->qp_type = MLX5_IB_QPT_REG_UMR; attr->qp_state = IB_QPS_INIT; attr->port_num = 1; ret = mlx5_ib_modify_qp(qp, attr, IB_QP_STATE | IB_QP_PKEY_INDEX | IB_QP_PORT, NULL); if (ret) { mlx5_ib_dbg(dev, "Couldn't modify UMR QP\n"); goto error_4; } memset(attr, 0, sizeof(*attr)); attr->qp_state = IB_QPS_RTR; attr->path_mtu = IB_MTU_256; ret = mlx5_ib_modify_qp(qp, attr, IB_QP_STATE, NULL); if (ret) { mlx5_ib_dbg(dev, "Couldn't modify umr QP to rtr\n"); goto error_4; } memset(attr, 0, sizeof(*attr)); attr->qp_state = IB_QPS_RTS; ret = mlx5_ib_modify_qp(qp, attr, IB_QP_STATE, NULL); if (ret) { mlx5_ib_dbg(dev, "Couldn't modify umr QP to rts\n"); goto error_4; } dev->umrc.qp = qp; dev->umrc.cq = cq; dev->umrc.pd = pd; sema_init(&dev->umrc.sem, MAX_UMR_WR); ret = mlx5_mr_cache_init(dev); if (ret) { mlx5_ib_warn(dev, "mr cache init failed %d\n", ret); goto error_4; } kfree(attr); kfree(init_attr); return 0; error_4: mlx5_ib_destroy_qp(qp); error_3: ib_free_cq(cq); error_2: ib_dealloc_pd(pd); error_0: kfree(attr); kfree(init_attr); return ret; } static int create_dev_resources(struct mlx5_ib_resources *devr) { struct ib_srq_init_attr attr; struct mlx5_ib_dev *dev; struct ib_cq_init_attr cq_attr = {.cqe = 1}; int port; int ret = 0; dev = container_of(devr, struct mlx5_ib_dev, devr); mutex_init(&devr->mutex); devr->p0 = mlx5_ib_alloc_pd(&dev->ib_dev, NULL, NULL); if (IS_ERR(devr->p0)) { ret = PTR_ERR(devr->p0); goto error0; } devr->p0->device = &dev->ib_dev; devr->p0->uobject = NULL; atomic_set(&devr->p0->usecnt, 0); devr->c0 = mlx5_ib_create_cq(&dev->ib_dev, &cq_attr, NULL, NULL); if (IS_ERR(devr->c0)) { ret = PTR_ERR(devr->c0); goto error1; } devr->c0->device = &dev->ib_dev; devr->c0->uobject = NULL; devr->c0->comp_handler = NULL; devr->c0->event_handler = NULL; devr->c0->cq_context = NULL; atomic_set(&devr->c0->usecnt, 0); devr->x0 = mlx5_ib_alloc_xrcd(&dev->ib_dev, NULL, NULL); if (IS_ERR(devr->x0)) { ret = PTR_ERR(devr->x0); goto error2; } devr->x0->device = &dev->ib_dev; devr->x0->inode = NULL; atomic_set(&devr->x0->usecnt, 0); mutex_init(&devr->x0->tgt_qp_mutex); INIT_LIST_HEAD(&devr->x0->tgt_qp_list); devr->x1 = mlx5_ib_alloc_xrcd(&dev->ib_dev, NULL, NULL); if (IS_ERR(devr->x1)) { ret = PTR_ERR(devr->x1); goto error3; } devr->x1->device = &dev->ib_dev; devr->x1->inode = NULL; atomic_set(&devr->x1->usecnt, 0); mutex_init(&devr->x1->tgt_qp_mutex); INIT_LIST_HEAD(&devr->x1->tgt_qp_list); memset(&attr, 0, sizeof(attr)); attr.attr.max_sge = 1; attr.attr.max_wr = 1; attr.srq_type = IB_SRQT_XRC; attr.ext.xrc.cq = devr->c0; attr.ext.xrc.xrcd = devr->x0; devr->s0 = mlx5_ib_create_srq(devr->p0, &attr, NULL); if (IS_ERR(devr->s0)) { ret = PTR_ERR(devr->s0); goto error4; } devr->s0->device = &dev->ib_dev; devr->s0->pd = devr->p0; devr->s0->uobject = NULL; devr->s0->event_handler = NULL; devr->s0->srq_context = NULL; devr->s0->srq_type = IB_SRQT_XRC; devr->s0->ext.xrc.xrcd = devr->x0; devr->s0->ext.xrc.cq = devr->c0; atomic_inc(&devr->s0->ext.xrc.xrcd->usecnt); atomic_inc(&devr->s0->ext.xrc.cq->usecnt); atomic_inc(&devr->p0->usecnt); atomic_set(&devr->s0->usecnt, 0); memset(&attr, 0, sizeof(attr)); attr.attr.max_sge = 1; attr.attr.max_wr = 1; attr.srq_type = IB_SRQT_BASIC; devr->s1 = mlx5_ib_create_srq(devr->p0, &attr, NULL); if (IS_ERR(devr->s1)) { ret = PTR_ERR(devr->s1); goto error5; } devr->s1->device = &dev->ib_dev; devr->s1->pd = devr->p0; devr->s1->uobject = NULL; devr->s1->event_handler = NULL; devr->s1->srq_context = NULL; devr->s1->srq_type = IB_SRQT_BASIC; devr->s1->ext.xrc.cq = devr->c0; atomic_inc(&devr->p0->usecnt); atomic_set(&devr->s0->usecnt, 0); for (port = 0; port < ARRAY_SIZE(devr->ports); ++port) { INIT_WORK(&devr->ports[port].pkey_change_work, pkey_change_handler); devr->ports[port].devr = devr; } return 0; error5: mlx5_ib_destroy_srq(devr->s0); error4: mlx5_ib_dealloc_xrcd(devr->x1); error3: mlx5_ib_dealloc_xrcd(devr->x0); error2: mlx5_ib_destroy_cq(devr->c0); error1: mlx5_ib_dealloc_pd(devr->p0); error0: return ret; } static void destroy_dev_resources(struct mlx5_ib_resources *devr) { struct mlx5_ib_dev *dev = container_of(devr, struct mlx5_ib_dev, devr); int port; mlx5_ib_destroy_srq(devr->s1); mlx5_ib_destroy_srq(devr->s0); mlx5_ib_dealloc_xrcd(devr->x0); mlx5_ib_dealloc_xrcd(devr->x1); mlx5_ib_destroy_cq(devr->c0); mlx5_ib_dealloc_pd(devr->p0); /* Make sure no change P_Key work items are still executing */ for (port = 0; port < dev->num_ports; ++port) cancel_work_sync(&devr->ports[port].pkey_change_work); } static u32 get_core_cap_flags(struct ib_device *ibdev) { struct mlx5_ib_dev *dev = to_mdev(ibdev); enum rdma_link_layer ll = mlx5_ib_port_link_layer(ibdev, 1); u8 l3_type_cap = MLX5_CAP_ROCE(dev->mdev, l3_type); u8 roce_version_cap = MLX5_CAP_ROCE(dev->mdev, roce_version); u32 ret = 0; if (ll == IB_LINK_LAYER_INFINIBAND) return RDMA_CORE_PORT_IBA_IB; if (!(l3_type_cap & MLX5_ROCE_L3_TYPE_IPV4_CAP)) return 0; if (!(l3_type_cap & MLX5_ROCE_L3_TYPE_IPV6_CAP)) return 0; if (roce_version_cap & MLX5_ROCE_VERSION_1_CAP) ret |= RDMA_CORE_PORT_IBA_ROCE; if (roce_version_cap & MLX5_ROCE_VERSION_2_CAP) ret |= RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP; return ret; } static int mlx5_port_immutable(struct ib_device *ibdev, u8 port_num, struct ib_port_immutable *immutable) { struct ib_port_attr attr; int err; err = mlx5_ib_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; immutable->core_cap_flags = get_core_cap_flags(ibdev); immutable->max_mad_size = IB_MGMT_MAD_SIZE; return 0; } static void get_dev_fw_str(struct ib_device *ibdev, char *str, size_t str_len) { struct mlx5_ib_dev *dev = container_of(ibdev, struct mlx5_ib_dev, ib_dev); snprintf(str, str_len, "%d.%d.%04d", fw_rev_maj(dev->mdev), fw_rev_min(dev->mdev), fw_rev_sub(dev->mdev)); } static int mlx5_roce_lag_init(struct mlx5_ib_dev *dev) { return 0; } static void mlx5_roce_lag_cleanup(struct mlx5_ib_dev *dev) { } static void mlx5_remove_roce_notifier(struct mlx5_ib_dev *dev) { if (dev->roce.nb.notifier_call) { unregister_netdevice_notifier(&dev->roce.nb); dev->roce.nb.notifier_call = NULL; } } static int mlx5_enable_roce(struct mlx5_ib_dev *dev) { VNET_ITERATOR_DECL(vnet_iter); - struct net_device *idev; + struct ifnet *idev; int err; /* Check if mlx5en net device already exists */ VNET_LIST_RLOCK(); VNET_FOREACH(vnet_iter) { IFNET_RLOCK(); CURVNET_SET_QUIET(vnet_iter); CK_STAILQ_FOREACH(idev, &V_ifnet, if_link) { /* check if network interface belongs to mlx5en */ if (!mlx5_netdev_match(idev, dev->mdev, "mce")) continue; write_lock(&dev->roce.netdev_lock); dev->roce.netdev = idev; write_unlock(&dev->roce.netdev_lock); } CURVNET_RESTORE(); IFNET_RUNLOCK(); } VNET_LIST_RUNLOCK(); dev->roce.nb.notifier_call = mlx5_netdev_event; err = register_netdevice_notifier(&dev->roce.nb); if (err) { dev->roce.nb.notifier_call = NULL; return err; } err = mlx5_nic_vport_enable_roce(dev->mdev); if (err) goto err_unregister_netdevice_notifier; err = mlx5_roce_lag_init(dev); if (err) goto err_disable_roce; return 0; err_disable_roce: mlx5_nic_vport_disable_roce(dev->mdev); err_unregister_netdevice_notifier: mlx5_remove_roce_notifier(dev); return err; } static void mlx5_disable_roce(struct mlx5_ib_dev *dev) { mlx5_roce_lag_cleanup(dev); mlx5_nic_vport_disable_roce(dev->mdev); } static void mlx5_ib_dealloc_q_port_counter(struct mlx5_ib_dev *dev, u8 port_num) { mlx5_vport_dealloc_q_counter(dev->mdev, MLX5_INTERFACE_PROTOCOL_IB, dev->port[port_num].q_cnt_id); dev->port[port_num].q_cnt_id = 0; } static void mlx5_ib_dealloc_q_counters(struct mlx5_ib_dev *dev) { unsigned int i; for (i = 0; i < dev->num_ports; i++) mlx5_ib_dealloc_q_port_counter(dev, i); } static int mlx5_ib_alloc_q_counters(struct mlx5_ib_dev *dev) { int i; int ret; for (i = 0; i < dev->num_ports; i++) { ret = mlx5_vport_alloc_q_counter(dev->mdev, MLX5_INTERFACE_PROTOCOL_IB, &dev->port[i].q_cnt_id); if (ret) { mlx5_ib_warn(dev, "couldn't allocate queue counter for port %d, err %d\n", i + 1, ret); goto dealloc_counters; } } return 0; dealloc_counters: while (--i >= 0) mlx5_ib_dealloc_q_port_counter(dev, i); return ret; } static const char * const names[] = { "rx_write_requests", "rx_read_requests", "rx_atomic_requests", "out_of_buffer", "out_of_sequence", "duplicate_request", "rnr_nak_retry_err", "packet_seq_err", "implied_nak_seq_err", "local_ack_timeout_err", }; static const size_t stats_offsets[] = { MLX5_BYTE_OFF(query_q_counter_out, rx_write_requests), MLX5_BYTE_OFF(query_q_counter_out, rx_read_requests), MLX5_BYTE_OFF(query_q_counter_out, rx_atomic_requests), MLX5_BYTE_OFF(query_q_counter_out, out_of_buffer), MLX5_BYTE_OFF(query_q_counter_out, out_of_sequence), MLX5_BYTE_OFF(query_q_counter_out, duplicate_request), MLX5_BYTE_OFF(query_q_counter_out, rnr_nak_retry_err), MLX5_BYTE_OFF(query_q_counter_out, packet_seq_err), MLX5_BYTE_OFF(query_q_counter_out, implied_nak_seq_err), MLX5_BYTE_OFF(query_q_counter_out, local_ack_timeout_err), }; static struct rdma_hw_stats *mlx5_ib_alloc_hw_stats(struct ib_device *ibdev, u8 port_num) { BUILD_BUG_ON(ARRAY_SIZE(names) != ARRAY_SIZE(stats_offsets)); /* We support only per port stats */ if (port_num == 0) return NULL; return rdma_alloc_hw_stats_struct(names, ARRAY_SIZE(names), RDMA_HW_STATS_DEFAULT_LIFESPAN); } static int mlx5_ib_get_hw_stats(struct ib_device *ibdev, struct rdma_hw_stats *stats, u8 port, int index) { struct mlx5_ib_dev *dev = to_mdev(ibdev); int outlen = MLX5_ST_SZ_BYTES(query_q_counter_out); void *out; __be32 val; int ret; int i; if (!port || !stats) return -ENOSYS; out = mlx5_vzalloc(outlen); if (!out) return -ENOMEM; ret = mlx5_vport_query_q_counter(dev->mdev, dev->port[port - 1].q_cnt_id, 0, out, outlen); if (ret) goto free; for (i = 0; i < ARRAY_SIZE(names); i++) { val = *(__be32 *)(out + stats_offsets[i]); stats->value[i] = (u64)be32_to_cpu(val); } free: kvfree(out); return ARRAY_SIZE(names); } static int mlx5_ib_stage_bfreg_init(struct mlx5_ib_dev *dev) { int err; err = mlx5_alloc_bfreg(dev->mdev, &dev->bfreg, false, false); if (err) return err; err = mlx5_alloc_bfreg(dev->mdev, &dev->fp_bfreg, false, true); if (err) { mlx5_free_bfreg(dev->mdev, &dev->bfreg); return err; } err = mlx5_alloc_bfreg(dev->mdev, &dev->wc_bfreg, true, false); if (err) { mlx5_free_bfreg(dev->mdev, &dev->fp_bfreg); mlx5_free_bfreg(dev->mdev, &dev->bfreg); } return err; } static void mlx5_ib_stage_bfreg_cleanup(struct mlx5_ib_dev *dev) { mlx5_free_bfreg(dev->mdev, &dev->wc_bfreg); mlx5_free_bfreg(dev->mdev, &dev->fp_bfreg); mlx5_free_bfreg(dev->mdev, &dev->bfreg); } static void *mlx5_ib_add(struct mlx5_core_dev *mdev) { struct mlx5_ib_dev *dev; enum rdma_link_layer ll; int port_type_cap; int err; int i; port_type_cap = MLX5_CAP_GEN(mdev, port_type); ll = mlx5_port_type_cap_to_rdma_ll(port_type_cap); if ((ll == IB_LINK_LAYER_ETHERNET) && !MLX5_CAP_GEN(mdev, roce)) return NULL; dev = (struct mlx5_ib_dev *)ib_alloc_device(sizeof(*dev)); if (!dev) return NULL; dev->mdev = mdev; dev->port = kcalloc(MLX5_CAP_GEN(mdev, num_ports), sizeof(*dev->port), GFP_KERNEL); if (!dev->port) goto err_dealloc; rwlock_init(&dev->roce.netdev_lock); err = get_port_caps(dev); if (err) goto err_free_port; if (mlx5_use_mad_ifc(dev)) get_ext_port_caps(dev); MLX5_INIT_DOORBELL_LOCK(&dev->uar_lock); snprintf(dev->ib_dev.name, IB_DEVICE_NAME_MAX, "mlx5_%d", device_get_unit(mdev->pdev->dev.bsddev)); dev->ib_dev.owner = THIS_MODULE; dev->ib_dev.node_type = RDMA_NODE_IB_CA; dev->ib_dev.local_dma_lkey = 0 /* not supported for now */; dev->num_ports = MLX5_CAP_GEN(mdev, num_ports); dev->ib_dev.phys_port_cnt = dev->num_ports; dev->ib_dev.num_comp_vectors = dev->mdev->priv.eq_table.num_comp_vectors; dev->ib_dev.dma_device = &mdev->pdev->dev; dev->ib_dev.uverbs_abi_ver = MLX5_IB_UVERBS_ABI_VERSION; dev->ib_dev.uverbs_cmd_mask = (1ull << IB_USER_VERBS_CMD_GET_CONTEXT) | (1ull << IB_USER_VERBS_CMD_QUERY_DEVICE) | (1ull << IB_USER_VERBS_CMD_QUERY_PORT) | (1ull << IB_USER_VERBS_CMD_ALLOC_PD) | (1ull << IB_USER_VERBS_CMD_DEALLOC_PD) | (1ull << IB_USER_VERBS_CMD_CREATE_AH) | (1ull << IB_USER_VERBS_CMD_DESTROY_AH) | (1ull << IB_USER_VERBS_CMD_REG_MR) | (1ull << IB_USER_VERBS_CMD_REREG_MR) | (1ull << IB_USER_VERBS_CMD_DEREG_MR) | (1ull << IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL) | (1ull << IB_USER_VERBS_CMD_CREATE_CQ) | (1ull << IB_USER_VERBS_CMD_RESIZE_CQ) | (1ull << IB_USER_VERBS_CMD_DESTROY_CQ) | (1ull << IB_USER_VERBS_CMD_CREATE_QP) | (1ull << IB_USER_VERBS_CMD_MODIFY_QP) | (1ull << IB_USER_VERBS_CMD_QUERY_QP) | (1ull << IB_USER_VERBS_CMD_DESTROY_QP) | (1ull << IB_USER_VERBS_CMD_ATTACH_MCAST) | (1ull << IB_USER_VERBS_CMD_DETACH_MCAST) | (1ull << IB_USER_VERBS_CMD_CREATE_SRQ) | (1ull << IB_USER_VERBS_CMD_MODIFY_SRQ) | (1ull << IB_USER_VERBS_CMD_QUERY_SRQ) | (1ull << IB_USER_VERBS_CMD_DESTROY_SRQ) | (1ull << IB_USER_VERBS_CMD_CREATE_XSRQ) | (1ull << IB_USER_VERBS_CMD_OPEN_QP); dev->ib_dev.uverbs_ex_cmd_mask = (1ull << IB_USER_VERBS_EX_CMD_QUERY_DEVICE) | (1ull << IB_USER_VERBS_EX_CMD_CREATE_CQ) | (1ull << IB_USER_VERBS_EX_CMD_CREATE_QP); dev->ib_dev.query_device = mlx5_ib_query_device; dev->ib_dev.query_port = mlx5_ib_query_port; dev->ib_dev.get_link_layer = mlx5_ib_port_link_layer; if (ll == IB_LINK_LAYER_ETHERNET) dev->ib_dev.get_netdev = mlx5_ib_get_netdev; dev->ib_dev.query_gid = mlx5_ib_query_gid; dev->ib_dev.add_gid = mlx5_ib_add_gid; dev->ib_dev.del_gid = mlx5_ib_del_gid; dev->ib_dev.query_pkey = mlx5_ib_query_pkey; dev->ib_dev.modify_device = mlx5_ib_modify_device; dev->ib_dev.modify_port = mlx5_ib_modify_port; dev->ib_dev.alloc_ucontext = mlx5_ib_alloc_ucontext; dev->ib_dev.dealloc_ucontext = mlx5_ib_dealloc_ucontext; dev->ib_dev.mmap = mlx5_ib_mmap; dev->ib_dev.alloc_pd = mlx5_ib_alloc_pd; dev->ib_dev.dealloc_pd = mlx5_ib_dealloc_pd; dev->ib_dev.create_ah = mlx5_ib_create_ah; dev->ib_dev.query_ah = mlx5_ib_query_ah; dev->ib_dev.destroy_ah = mlx5_ib_destroy_ah; dev->ib_dev.create_srq = mlx5_ib_create_srq; dev->ib_dev.modify_srq = mlx5_ib_modify_srq; dev->ib_dev.query_srq = mlx5_ib_query_srq; dev->ib_dev.destroy_srq = mlx5_ib_destroy_srq; dev->ib_dev.post_srq_recv = mlx5_ib_post_srq_recv; dev->ib_dev.create_qp = mlx5_ib_create_qp; dev->ib_dev.modify_qp = mlx5_ib_modify_qp; dev->ib_dev.query_qp = mlx5_ib_query_qp; dev->ib_dev.destroy_qp = mlx5_ib_destroy_qp; dev->ib_dev.post_send = mlx5_ib_post_send; dev->ib_dev.post_recv = mlx5_ib_post_recv; dev->ib_dev.create_cq = mlx5_ib_create_cq; dev->ib_dev.modify_cq = mlx5_ib_modify_cq; dev->ib_dev.resize_cq = mlx5_ib_resize_cq; dev->ib_dev.destroy_cq = mlx5_ib_destroy_cq; dev->ib_dev.poll_cq = mlx5_ib_poll_cq; dev->ib_dev.req_notify_cq = mlx5_ib_arm_cq; dev->ib_dev.get_dma_mr = mlx5_ib_get_dma_mr; dev->ib_dev.reg_user_mr = mlx5_ib_reg_user_mr; dev->ib_dev.rereg_user_mr = mlx5_ib_rereg_user_mr; dev->ib_dev.dereg_mr = mlx5_ib_dereg_mr; dev->ib_dev.attach_mcast = mlx5_ib_mcg_attach; dev->ib_dev.detach_mcast = mlx5_ib_mcg_detach; dev->ib_dev.process_mad = mlx5_ib_process_mad; dev->ib_dev.alloc_mr = mlx5_ib_alloc_mr; dev->ib_dev.map_mr_sg = mlx5_ib_map_mr_sg; dev->ib_dev.check_mr_status = mlx5_ib_check_mr_status; dev->ib_dev.get_port_immutable = mlx5_port_immutable; dev->ib_dev.get_dev_fw_str = get_dev_fw_str; if (mlx5_core_is_pf(mdev)) { dev->ib_dev.get_vf_config = mlx5_ib_get_vf_config; dev->ib_dev.set_vf_link_state = mlx5_ib_set_vf_link_state; dev->ib_dev.get_vf_stats = mlx5_ib_get_vf_stats; dev->ib_dev.set_vf_guid = mlx5_ib_set_vf_guid; } dev->ib_dev.disassociate_ucontext = mlx5_ib_disassociate_ucontext; mlx5_ib_internal_fill_odp_caps(dev); if (MLX5_CAP_GEN(mdev, imaicl)) { dev->ib_dev.alloc_mw = mlx5_ib_alloc_mw; dev->ib_dev.dealloc_mw = mlx5_ib_dealloc_mw; dev->ib_dev.uverbs_cmd_mask |= (1ull << IB_USER_VERBS_CMD_ALLOC_MW) | (1ull << IB_USER_VERBS_CMD_DEALLOC_MW); } if (MLX5_CAP_GEN(dev->mdev, out_of_seq_cnt) && MLX5_CAP_GEN(dev->mdev, retransmission_q_counters)) { dev->ib_dev.get_hw_stats = mlx5_ib_get_hw_stats; dev->ib_dev.alloc_hw_stats = mlx5_ib_alloc_hw_stats; } if (MLX5_CAP_GEN(mdev, xrc)) { dev->ib_dev.alloc_xrcd = mlx5_ib_alloc_xrcd; dev->ib_dev.dealloc_xrcd = mlx5_ib_dealloc_xrcd; dev->ib_dev.uverbs_cmd_mask |= (1ull << IB_USER_VERBS_CMD_OPEN_XRCD) | (1ull << IB_USER_VERBS_CMD_CLOSE_XRCD); } if (mlx5_ib_port_link_layer(&dev->ib_dev, 1) == IB_LINK_LAYER_ETHERNET) { dev->ib_dev.create_flow = mlx5_ib_create_flow; dev->ib_dev.destroy_flow = mlx5_ib_destroy_flow; dev->ib_dev.create_wq = mlx5_ib_create_wq; dev->ib_dev.modify_wq = mlx5_ib_modify_wq; dev->ib_dev.destroy_wq = mlx5_ib_destroy_wq; dev->ib_dev.create_rwq_ind_table = mlx5_ib_create_rwq_ind_table; dev->ib_dev.destroy_rwq_ind_table = mlx5_ib_destroy_rwq_ind_table; dev->ib_dev.uverbs_ex_cmd_mask |= (1ull << IB_USER_VERBS_EX_CMD_CREATE_FLOW) | (1ull << IB_USER_VERBS_EX_CMD_DESTROY_FLOW) | (1ull << IB_USER_VERBS_EX_CMD_CREATE_WQ) | (1ull << IB_USER_VERBS_EX_CMD_MODIFY_WQ) | (1ull << IB_USER_VERBS_EX_CMD_DESTROY_WQ) | (1ull << IB_USER_VERBS_EX_CMD_CREATE_RWQ_IND_TBL) | (1ull << IB_USER_VERBS_EX_CMD_DESTROY_RWQ_IND_TBL); } err = init_node_data(dev); if (err) goto err_free_port; mutex_init(&dev->flow_db.lock); mutex_init(&dev->cap_mask_mutex); INIT_LIST_HEAD(&dev->qp_list); spin_lock_init(&dev->reset_flow_resource_lock); if (ll == IB_LINK_LAYER_ETHERNET) { err = mlx5_enable_roce(dev); if (err) goto err_free_port; } err = create_dev_resources(&dev->devr); if (err) goto err_disable_roce; err = mlx5_ib_odp_init_one(dev); if (err) goto err_rsrc; err = mlx5_ib_alloc_q_counters(dev); if (err) goto err_odp; err = mlx5_ib_stage_bfreg_init(dev); if (err) goto err_q_cnt; err = ib_register_device(&dev->ib_dev, NULL); if (err) goto err_bfreg; err = create_umr_res(dev); if (err) goto err_dev; for (i = 0; i < ARRAY_SIZE(mlx5_class_attributes); i++) { err = device_create_file(&dev->ib_dev.dev, mlx5_class_attributes[i]); if (err) goto err_umrc; } err = mlx5_ib_init_congestion(dev); if (err) goto err_umrc; dev->ib_active = true; return dev; err_umrc: destroy_umrc_res(dev); err_dev: ib_unregister_device(&dev->ib_dev); err_bfreg: mlx5_ib_stage_bfreg_cleanup(dev); err_q_cnt: mlx5_ib_dealloc_q_counters(dev); err_odp: mlx5_ib_odp_remove_one(dev); err_rsrc: destroy_dev_resources(&dev->devr); err_disable_roce: if (ll == IB_LINK_LAYER_ETHERNET) { mlx5_disable_roce(dev); mlx5_remove_roce_notifier(dev); } err_free_port: kfree(dev->port); err_dealloc: ib_dealloc_device((struct ib_device *)dev); return NULL; } static void mlx5_ib_remove(struct mlx5_core_dev *mdev, void *context) { struct mlx5_ib_dev *dev = context; enum rdma_link_layer ll = mlx5_ib_port_link_layer(&dev->ib_dev, 1); mlx5_ib_cleanup_congestion(dev); mlx5_remove_roce_notifier(dev); ib_unregister_device(&dev->ib_dev); mlx5_ib_stage_bfreg_cleanup(dev); mlx5_ib_dealloc_q_counters(dev); destroy_umrc_res(dev); mlx5_ib_odp_remove_one(dev); destroy_dev_resources(&dev->devr); if (ll == IB_LINK_LAYER_ETHERNET) mlx5_disable_roce(dev); kfree(dev->port); ib_dealloc_device(&dev->ib_dev); } static struct mlx5_interface mlx5_ib_interface = { .add = mlx5_ib_add, .remove = mlx5_ib_remove, .event = mlx5_ib_event, .protocol = MLX5_INTERFACE_PROTOCOL_IB, }; static int __init mlx5_ib_init(void) { int err; err = mlx5_ib_odp_init(); if (err) return err; err = mlx5_register_interface(&mlx5_ib_interface); if (err) goto clean_odp; return err; clean_odp: mlx5_ib_odp_cleanup(); return err; } static void __exit mlx5_ib_cleanup(void) { mlx5_unregister_interface(&mlx5_ib_interface); mlx5_ib_odp_cleanup(); } static void mlx5_ib_show_version(void __unused *arg) { printf("%s", mlx5_version); } SYSINIT(mlx5_ib_show_version, SI_SUB_DRIVERS, SI_ORDER_ANY, mlx5_ib_show_version, NULL); module_init_order(mlx5_ib_init, SI_ORDER_SEVENTH); module_exit_order(mlx5_ib_cleanup, SI_ORDER_SEVENTH);