svn commit: r325760 - in projects/bsd_rdma_4_9/sys: conf dev/mlx4/mlx4_ib modules/mlx4ib
Hans Petter Selasky
hselasky at FreeBSD.org
Mon Nov 13 10:49:20 UTC 2017
Author: hselasky
Date: Mon Nov 13 10:49:18 2017
New Revision: 325760
URL: https://svnweb.freebsd.org/changeset/base/325760
Log:
Update mlx4ib(4) to Linux 4.9.
Sponsored by: Mellanox Technologies
Deleted:
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_exp.h
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_exp.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/user.h
Modified:
projects/bsd_rdma_4_9/sys/conf/files
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib.h
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_ah.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_alias_GUID.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_cm.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_cq.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_mad.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_main.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_mcg.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_mr.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_qp.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_srq.c
projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_sysfs.c
projects/bsd_rdma_4_9/sys/modules/mlx4ib/Makefile
Modified: projects/bsd_rdma_4_9/sys/conf/files
==============================================================================
--- projects/bsd_rdma_4_9/sys/conf/files Mon Nov 13 10:45:31 2017 (r325759)
+++ projects/bsd_rdma_4_9/sys/conf/files Mon Nov 13 10:49:18 2017 (r325760)
@@ -4550,8 +4550,6 @@ dev/mlx4/mlx4_ib/mlx4_ib_mad.c optional mlx4ib pci o
compile-with "${OFED_C}"
dev/mlx4/mlx4_ib/mlx4_ib_main.c optional mlx4ib pci ofed \
compile-with "${OFED_C}"
-dev/mlx4/mlx4_ib/mlx4_ib_exp.c optional mlx4ib pci ofed \
- compile-with "${OFED_C}"
dev/mlx4/mlx4_ib/mlx4_ib_mr.c optional mlx4ib pci ofed \
compile-with "${OFED_C}"
dev/mlx4/mlx4_ib/mlx4_ib_qp.c optional mlx4ib pci ofed \
Modified: projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib.h
==============================================================================
--- projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib.h Mon Nov 13 10:45:31 2017 (r325759)
+++ projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib.h Mon Nov 13 10:49:18 2017 (r325760)
@@ -41,6 +41,8 @@
#include <linux/rbtree.h>
#include <linux/notifier.h>
+#include <asm/atomic64.h>
+
#include <rdma/ib_verbs.h>
#include <rdma/ib_umem.h>
#include <rdma/ib_mad.h>
@@ -59,9 +61,6 @@
#define mlx4_ib_warn(ibdev, format, arg...) \
dev_warn((ibdev)->dma_device, MLX4_IB_DRV_NAME ": " format, ## arg)
-#define mlx4_ib_info(ibdev, format, arg...) \
- dev_info((ibdev)->dma_device, MLX4_IB_DRV_NAME ": " format, ## arg)
-
enum {
MLX4_IB_SQ_MIN_WQE_SHIFT = 6,
MLX4_IB_MAX_HEADROOM = 2048
@@ -75,17 +74,25 @@ extern int mlx4_ib_sm_guid_assign;
extern struct proc_dir_entry *mlx4_mrs_dir_entry;
#define MLX4_IB_UC_STEER_QPN_ALIGN 1
-#define MLX4_IB_UC_MAX_NUM_QPS (256 * 1024)
+#define MLX4_IB_UC_MAX_NUM_QPS 256
+enum hw_bar_type {
+ HW_BAR_BF,
+ HW_BAR_DB,
+ HW_BAR_CLOCK,
+ HW_BAR_COUNT
+};
-#define MLX4_IB_MMAP_CMD_MASK 0xFF
-#define MLX4_IB_MMAP_CMD_BITS 8
+struct mlx4_ib_vma_private_data {
+ struct vm_area_struct *vma;
+};
struct mlx4_ib_ucontext {
struct ib_ucontext ibucontext;
struct mlx4_uar uar;
struct list_head db_page_list;
struct mutex db_page_mutex;
+ struct mlx4_ib_vma_private_data hw_bar_info[HW_BAR_COUNT];
};
struct mlx4_ib_pd {
@@ -111,11 +118,6 @@ struct mlx4_ib_cq_resize {
int cqe;
};
-struct mlx4_shared_mr_info {
- int mr_id;
- struct ib_umem *umem;
-};
-
struct mlx4_ib_cq {
struct ib_cq ibcq;
struct mlx4_cq mcq;
@@ -127,15 +129,22 @@ struct mlx4_ib_cq {
struct ib_umem *umem;
struct ib_umem *resize_umem;
int create_flags;
+ /* List of qps that it serves.*/
+ struct list_head send_qp_list;
+ struct list_head recv_qp_list;
};
+#define MLX4_MR_PAGES_ALIGN 0x40
+
struct mlx4_ib_mr {
struct ib_mr ibmr;
+ __be64 *pages;
+ dma_addr_t page_map;
+ u32 npages;
+ u32 max_pages;
struct mlx4_mr mmr;
struct ib_umem *umem;
- struct mlx4_shared_mr_info *smr_info;
- atomic_t invalidated;
- struct completion invalidation_comp;
+ size_t page_map_size;
};
struct mlx4_ib_mw {
@@ -143,21 +152,22 @@ struct mlx4_ib_mw {
struct mlx4_mw mmw;
};
-struct mlx4_ib_fast_reg_page_list {
- struct ib_fast_reg_page_list ibfrpl;
- __be64 *mapped_page_list;
- dma_addr_t map;
-};
-
struct mlx4_ib_fmr {
struct ib_fmr ibfmr;
struct mlx4_fmr mfmr;
};
+#define MAX_REGS_PER_FLOW 2
+
+struct mlx4_flow_reg_id {
+ u64 id;
+ u64 mirror;
+};
+
struct mlx4_ib_flow {
struct ib_flow ibflow;
/* translating DMFS verbs sniffer rule to FW API requires two reg IDs */
- u64 reg_id[2];
+ struct mlx4_flow_reg_id reg_id[MAX_REGS_PER_FLOW];
};
struct mlx4_ib_wq {
@@ -172,13 +182,18 @@ struct mlx4_ib_wq {
unsigned tail;
};
+enum {
+ MLX4_IB_QP_CREATE_ROCE_V2_GSI = IB_QP_CREATE_RESERVED_START
+};
+
enum mlx4_ib_qp_flags {
MLX4_IB_QP_LSO = IB_QP_CREATE_IPOIB_UD_LSO,
MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK = IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK,
- MLX4_IB_QP_CAP_CROSS_CHANNEL = IB_QP_CREATE_CROSS_CHANNEL,
- MLX4_IB_QP_CAP_MANAGED_SEND = IB_QP_CREATE_MANAGED_SEND,
- MLX4_IB_QP_CAP_MANAGED_RECV = IB_QP_CREATE_MANAGED_RECV,
MLX4_IB_QP_NETIF = IB_QP_CREATE_NETIF_QP,
+ MLX4_IB_QP_CREATE_USE_GFP_NOIO = IB_QP_CREATE_USE_GFP_NOIO,
+
+ /* Mellanox specific flags start from IB_QP_CREATE_RESERVED_START */
+ MLX4_IB_ROCE_V2_GSI_QP = MLX4_IB_QP_CREATE_ROCE_V2_GSI,
MLX4_IB_SRIOV_TUNNEL_QP = 1 << 30,
MLX4_IB_SRIOV_SQP = 1 << 31,
};
@@ -190,13 +205,6 @@ struct mlx4_ib_gid_entry {
u8 port;
};
-enum mlx4_ib_mmap_cmd {
- MLX4_IB_MMAP_UAR_PAGE = 0,
- MLX4_IB_MMAP_BLUE_FLAME_PAGE = 1,
- MLX4_IB_MMAP_GET_CONTIGUOUS_PAGES = 2,
- MLX4_IB_MMAP_GET_HW_CLOCK = 3,
-};
-
enum mlx4_ib_qp_type {
/*
* IB_QPT_SMI and IB_QPT_GSI have to be the first two entries
@@ -287,17 +295,6 @@ struct mlx4_roce_smac_vlan_info {
int update_vid;
};
-struct mlx4_ib_qpg_data {
- unsigned long *tss_bitmap;
- unsigned long *rss_bitmap;
- struct mlx4_ib_qp *qpg_parent;
- int tss_qpn_base;
- int rss_qpn_base;
- u32 tss_child_count;
- u32 rss_child_count;
- u32 qpg_tss_mask_sz;
-};
-
struct mlx4_ib_qp {
struct ib_qp ibqp;
struct mlx4_qp mqp;
@@ -327,22 +324,16 @@ struct mlx4_ib_qp {
u8 sq_no_prefetch;
u8 state;
int mlx_type;
- enum ib_qpg_type qpg_type;
- struct mlx4_ib_qpg_data *qpg_data;
struct list_head gid_list;
struct list_head steering_rules;
struct mlx4_ib_buf *sqp_proxy_rcv;
struct mlx4_roce_smac_vlan_info pri;
struct mlx4_roce_smac_vlan_info alt;
- struct list_head rules_list;
u64 reg_id;
- int max_inline_data;
- struct mlx4_bf bf;
-
- /*
- * Experimental data
- */
- int max_inlr_data;
+ struct list_head qps_list;
+ struct list_head cq_recv_list;
+ struct list_head cq_send_list;
+ struct counter_index *counter_index;
};
struct mlx4_ib_srq {
@@ -376,23 +367,12 @@ struct mlx4_ib_ah {
#define MLX4_NOT_SET_GUID (0x00LL)
#define MLX4_GUID_FOR_DELETE_VAL (~(0x00LL))
-/****************************************/
-/* ioctl codes */
-/****************************************/
-#define MLX4_IOC_MAGIC 'm'
-#define MLX4_IOCHWCLOCKOFFSET _IOR(MLX4_IOC_MAGIC, 1, int)
-
enum mlx4_guid_alias_rec_status {
MLX4_GUID_INFO_STATUS_IDLE,
MLX4_GUID_INFO_STATUS_SET,
- MLX4_GUID_INFO_STATUS_PENDING,
};
-enum mlx4_guid_alias_rec_ownership {
- MLX4_GUID_DRIVER_ASSIGN,
- MLX4_GUID_SYSADMIN_ASSIGN,
- MLX4_GUID_NONE_ASSIGN, /*init state of each record*/
-};
+#define GUID_STATE_NEED_PORT_INIT 0x01
enum mlx4_guid_alias_rec_method {
MLX4_GUID_INFO_RECORD_SET = IB_MGMT_METHOD_SET,
@@ -403,8 +383,8 @@ struct mlx4_sriov_alias_guid_info_rec_det {
u8 all_recs[GUID_REC_SIZE * NUM_ALIAS_GUID_IN_REC];
ib_sa_comp_mask guid_indexes; /*indicates what from the 8 records are valid*/
enum mlx4_guid_alias_rec_status status; /*indicates the administraively status of the record.*/
- u8 method; /*set or delete*/
- enum mlx4_guid_alias_rec_ownership ownership; /*indicates who assign that alias_guid record*/
+ unsigned int guids_retry_schedule[NUM_ALIAS_GUID_IN_REC];
+ u64 time_to_run;
};
struct mlx4_sriov_alias_guid_port_rec_det {
@@ -412,6 +392,7 @@ struct mlx4_sriov_alias_guid_port_rec_det {
struct workqueue_struct *wq;
struct delayed_work alias_guid_work;
u8 port;
+ u32 state_flags;
struct mlx4_sriov_alias_guid *parent;
struct list_head cb_list;
};
@@ -461,7 +442,6 @@ struct mlx4_ib_demux_pv_ctx {
struct ib_device *ib_dev;
struct ib_cq *cq;
struct ib_pd *pd;
- struct ib_mr *mr;
struct work_struct work;
struct workqueue_struct *wq;
struct mlx4_ib_demux_pv_qp qp[2];
@@ -473,7 +453,7 @@ struct mlx4_ib_demux_ctx {
struct workqueue_struct *wq;
struct workqueue_struct *ud_wq;
spinlock_t ud_lock;
- __be64 subnet_prefix;
+ atomic64_t subnet_prefix;
__be64 guid_cache[128];
struct mlx4_ib_dev *dev;
/* the following lock protects both mcg_table and mcg_mgid0_list */
@@ -503,13 +483,27 @@ struct mlx4_ib_sriov {
struct idr pv_id_table;
};
+struct gid_cache_context {
+ int real_index;
+ int refcount;
+};
+
+struct gid_entry {
+ union ib_gid gid;
+ enum ib_gid_type gid_type;
+ struct gid_cache_context *ctx;
+};
+
+struct mlx4_port_gid_table {
+ struct gid_entry gids[MLX4_MAX_PORT_GIDS];
+};
+
struct mlx4_ib_iboe {
spinlock_t lock;
struct net_device *netdevs[MLX4_MAX_PORTS];
- struct net_device *masters[MLX4_MAX_PORTS];
- struct notifier_block nb;
- struct notifier_block nb_inet;
- union ib_gid gid_table[MLX4_MAX_PORTS][128];
+ atomic64_t mac[MLX4_MAX_PORTS];
+ struct notifier_block nb;
+ struct mlx4_port_gid_table gids[MLX4_MAX_PORTS];
};
struct pkey_mgt {
@@ -548,15 +542,32 @@ struct mlx4_ib_iov_port {
struct mlx4_ib_iov_sysfs_attr mcg_dentry;
};
-struct mlx4_ib_counter {
- int counter_index;
- int status;
+struct counter_index {
+ struct list_head list;
+ u32 index;
+ u8 allocated;
};
+struct mlx4_ib_counters {
+ struct list_head counters_list;
+ struct mutex mutex; /* mutex for accessing counters list */
+ u32 default_counter;
+};
+
+#define MLX4_DIAG_COUNTERS_TYPES 2
+
+struct mlx4_ib_diag_counters {
+ const char **name;
+ u32 *offset;
+ u32 num_counters;
+};
+
struct mlx4_ib_dev {
struct ib_device ib_dev;
struct mlx4_dev *dev;
int num_ports;
+ void __iomem *uar_map;
+
struct mlx4_uar priv_uar;
u32 priv_pdn;
MLX4_DECLARE_DOORBELL_LOCK(uar_lock);
@@ -564,14 +575,14 @@ struct mlx4_ib_dev {
struct ib_mad_agent *send_agent[MLX4_MAX_PORTS][2];
struct ib_ah *sm_ah[MLX4_MAX_PORTS];
spinlock_t sm_lock;
+ atomic64_t sl2vl[MLX4_MAX_PORTS];
struct mlx4_ib_sriov sriov;
struct mutex cap_mask_mutex;
bool ib_active;
struct mlx4_ib_iboe iboe;
- struct mlx4_ib_counter counters[MLX4_MAX_PORTS];
+ struct mlx4_ib_counters counters_table[MLX4_MAX_PORTS];
int *eq_table;
- int eq_added;
struct kobject *iov_parent;
struct kobject *ports_parent;
struct kobject *dev_ports_parent[MLX4_MFUNC_MAX];
@@ -580,12 +591,22 @@ struct mlx4_ib_dev {
unsigned long *ib_uc_qpns_bitmap;
int steer_qpn_count;
int steer_qpn_base;
+ int steering_support;
+ struct mlx4_ib_qp *qp1_proxy[MLX4_MAX_PORTS];
+ /* lock when destroying qp1_proxy and getting netdev events */
+ struct mutex qp1_proxy_lock[MLX4_MAX_PORTS];
+ u8 bond_next_port;
+ /* protect resources needed as part of reset flow */
+ spinlock_t reset_flow_resource_lock;
+ struct list_head qp_list;
+ struct mlx4_ib_diag_counters diag_counters[MLX4_DIAG_COUNTERS_TYPES];
};
struct ib_event_work {
struct work_struct work;
struct mlx4_ib_dev *ib_dev;
struct mlx4_eqe ib_eqe;
+ int port;
};
struct mlx4_ib_qp_tunnel_init_attr {
@@ -595,6 +616,21 @@ struct mlx4_ib_qp_tunnel_init_attr {
u8 port;
};
+struct mlx4_uverbs_ex_query_device {
+ __u32 comp_mask;
+ __u32 reserved;
+};
+
+enum query_device_resp_mask {
+ QUERY_DEVICE_RESP_MASK_TIMESTAMP = 1UL << 0,
+};
+
+struct mlx4_uverbs_ex_query_device_resp {
+ __u32 comp_mask;
+ __u32 response_length;
+ __u64 hca_core_clock_offset;
+};
+
static inline struct mlx4_ib_dev *to_mdev(struct ib_device *ibdev)
{
return container_of(ibdev, struct mlx4_ib_dev, ib_dev);
@@ -635,11 +671,6 @@ static inline struct mlx4_ib_mw *to_mmw(struct ib_mw *
return container_of(ibmw, struct mlx4_ib_mw, ibmw);
}
-static inline struct mlx4_ib_fast_reg_page_list *to_mfrpl(struct ib_fast_reg_page_list *ibfrpl)
-{
- return container_of(ibfrpl, struct mlx4_ib_fast_reg_page_list, ibfrpl);
-}
-
static inline struct mlx4_ib_fmr *to_mfmr(struct ib_fmr *ibfmr)
{
return container_of(ibfmr, struct mlx4_ib_fmr, ibfmr);
@@ -675,6 +706,13 @@ static inline struct mlx4_ib_ah *to_mah(struct ib_ah *
return container_of(ibah, struct mlx4_ib_ah, ibah);
}
+static inline u8 mlx4_ib_bond_next_port(struct mlx4_ib_dev *dev)
+{
+ dev->bond_next_port = (dev->bond_next_port + 1) % dev->num_ports;
+
+ return dev->bond_next_port + 1;
+}
+
int mlx4_ib_init_sriov(struct mlx4_ib_dev *dev);
void mlx4_ib_close_sriov(struct mlx4_ib_dev *dev);
@@ -685,30 +723,22 @@ void mlx4_ib_db_unmap_user(struct mlx4_ib_ucontext *co
struct ib_mr *mlx4_ib_get_dma_mr(struct ib_pd *pd, int acc);
int mlx4_ib_umem_write_mtt(struct mlx4_ib_dev *dev, struct mlx4_mtt *mtt,
struct ib_umem *umem);
-int mlx4_ib_umem_calc_optimal_mtt_size(struct ib_umem *umem,
- u64 start_va,
- int *num_of_mtts);
struct ib_mr *mlx4_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
u64 virt_addr, int access_flags,
- struct ib_udata *udata, int mr_id);
+ struct ib_udata *udata);
int mlx4_ib_dereg_mr(struct ib_mr *mr);
-struct ib_mw *mlx4_ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type);
-int mlx4_ib_bind_mw(struct ib_qp *qp, struct ib_mw *mw,
- struct ib_mw_bind *mw_bind);
+struct ib_mw *mlx4_ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type,
+ struct ib_udata *udata);
int mlx4_ib_dealloc_mw(struct ib_mw *mw);
-struct ib_mr *mlx4_ib_alloc_fast_reg_mr(struct ib_pd *pd,
- int max_page_list_len);
-struct ib_fast_reg_page_list *mlx4_ib_alloc_fast_reg_page_list(struct ib_device *ibdev,
- int page_list_len);
-void mlx4_ib_free_fast_reg_page_list(struct ib_fast_reg_page_list *page_list);
-
-int mlx4_ib_modify_cq(struct ib_cq *cq,
- struct ib_cq_attr *cq_attr,
- int cq_attr_mask);
+struct ib_mr *mlx4_ib_alloc_mr(struct ib_pd *pd,
+ enum ib_mr_type mr_type,
+ u32 max_num_sg);
+int mlx4_ib_map_mr_sg(struct ib_mr *ibmr, struct scatterlist *sg, int sg_nents,
+ unsigned int *sg_offset);
+int mlx4_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period);
int mlx4_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata);
-int mlx4_ib_ignore_overrun_cq(struct ib_cq *ibcq);
struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev,
- struct ib_cq_init_attr *attr,
+ const struct ib_cq_init_attr *attr,
struct ib_ucontext *context,
struct ib_udata *udata);
int mlx4_ib_destroy_cq(struct ib_cq *cq);
@@ -746,11 +776,13 @@ int mlx4_ib_post_recv(struct ib_qp *ibqp, struct ib_re
struct ib_recv_wr **bad_wr);
int mlx4_MAD_IFC(struct mlx4_ib_dev *dev, int mad_ifc_flags,
- int port, struct ib_wc *in_wc, struct ib_grh *in_grh,
- void *in_mad, void *response_mad);
+ int port, const struct ib_wc *in_wc, const struct ib_grh *in_grh,
+ const void *in_mad, void *response_mad);
int mlx4_ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
- struct ib_wc *in_wc, struct ib_grh *in_grh,
- struct ib_mad *in_mad, struct ib_mad *out_mad);
+ 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);
int mlx4_ib_mad_init(struct mlx4_ib_dev *dev);
void mlx4_ib_mad_cleanup(struct mlx4_ib_dev *dev);
@@ -768,29 +800,16 @@ int __mlx4_ib_query_pkey(struct ib_device *ibdev, u8 p
int __mlx4_ib_query_gid(struct ib_device *ibdev, u8 port, int index,
union ib_gid *gid, int netw_view);
-int mlx4_ib_resolve_grh(struct mlx4_ib_dev *dev, const struct ib_ah_attr *ah_attr,
- u8 *mac, int *is_mcast, u8 port);
-
-int mlx4_ib_query_if_stat(struct mlx4_ib_dev *dev, u32 counter_index,
- union mlx4_counter *counter, u8 clear);
-
-static inline int mlx4_ib_ah_grh_present(struct mlx4_ib_ah *ah)
+static inline bool mlx4_ib_ah_grh_present(struct mlx4_ib_ah *ah)
{
u8 port = be32_to_cpu(ah->av.ib.port_pd) >> 24 & 3;
if (rdma_port_get_link_layer(ah->ibah.device, port) == IB_LINK_LAYER_ETHERNET)
- return 1;
+ return true;
return !!(ah->av.ib.g_slid & 0x80);
}
-static inline int mlx4_ib_qp_has_rq(struct ib_qp_init_attr *attr)
-{
- if (attr->qp_type == IB_QPT_XRC_INI || attr->qp_type == IB_QPT_XRC_TGT)
- return 0;
- return !attr->srq;
-}
-
int mlx4_ib_mcg_port_init(struct mlx4_ib_demux_ctx *ctx);
void mlx4_ib_mcg_port_cleanup(struct mlx4_ib_demux_ctx *ctx, int destroy_wq);
void clean_vf_mcast(struct mlx4_ib_demux_ctx *ctx, int slave);
@@ -815,13 +834,16 @@ void mlx4_ib_tunnels_update_work(struct work_struct *w
int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port,
enum ib_qp_type qpt, struct ib_wc *wc,
struct ib_grh *grh, struct ib_mad *mad);
+
int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port,
enum ib_qp_type dest_qpt, u16 pkey_index, u32 remote_qpn,
- u32 qkey, struct ib_ah_attr *attr, u8 *s_mac, struct ib_mad *mad);
+ u32 qkey, struct ib_ah_attr *attr, u8 *s_mac,
+ u16 vlan_id, struct ib_mad *mad);
+
__be64 mlx4_ib_get_new_demux_tid(struct mlx4_ib_demux_ctx *ctx);
int mlx4_ib_demux_cm_handler(struct ib_device *ibdev, int port, int *slave,
- struct ib_mad *mad, int is_eth);
+ struct ib_mad *mad);
int mlx4_ib_multiplex_cm_handler(struct ib_device *ibdev, int port, int slave_id,
struct ib_mad *mad);
@@ -848,6 +870,8 @@ int add_sysfs_port_mcg_attr(struct mlx4_ib_dev *device
void del_sysfs_port_mcg_attr(struct mlx4_ib_dev *device, int port_num,
struct attribute *attr);
ib_sa_comp_mask mlx4_ib_get_aguid_comp_mask_from_ix(int index);
+void mlx4_ib_slave_alias_guid_event(struct mlx4_ib_dev *dev, int slave,
+ int port, int slave_init);
int mlx4_ib_device_register_sysfs(struct mlx4_ib_dev *device) ;
@@ -859,7 +883,16 @@ int mlx4_ib_steer_qp_alloc(struct mlx4_ib_dev *dev, in
void mlx4_ib_steer_qp_free(struct mlx4_ib_dev *dev, u32 qpn, int count);
int mlx4_ib_steer_qp_reg(struct mlx4_ib_dev *mdev, struct mlx4_ib_qp *mqp,
int is_attach);
-int mlx4_ib_query_device(struct ib_device *ibdev,
- struct ib_device_attr *props);
+int mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags,
+ u64 start, u64 length, u64 virt_addr,
+ int mr_access_flags, struct ib_pd *pd,
+ struct ib_udata *udata);
+int mlx4_ib_gid_index_to_real_index(struct mlx4_ib_dev *ibdev,
+ u8 port_num, int index);
+
+void mlx4_sched_ib_sl2vl_update_work(struct mlx4_ib_dev *ibdev,
+ int port);
+
+void mlx4_ib_sl2vl_update(struct mlx4_ib_dev *mdev, int port);
#endif /* MLX4_IB_H */
Modified: projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_ah.c
==============================================================================
--- projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_ah.c Mon Nov 13 10:45:31 2017 (r325759)
+++ projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_ah.c Mon Nov 13 10:49:18 2017 (r325760)
@@ -38,28 +38,10 @@
#include <linux/slab.h>
#include <linux/string.h>
+#include <linux/etherdevice.h>
#include "mlx4_ib.h"
-int mlx4_ib_resolve_grh(struct mlx4_ib_dev *dev, const struct ib_ah_attr *ah_attr,
- u8 *mac, int *is_mcast, u8 port)
-{
- struct in6_addr in6;
-
- *is_mcast = 0;
-
- memcpy(&in6, ah_attr->grh.dgid.raw, sizeof in6);
- if (rdma_link_local_addr(&in6))
- rdma_get_ll_mac(&in6, mac);
- else if (rdma_is_multicast_addr(&in6)) {
- rdma_get_mcast_mac(&in6, mac);
- *is_mcast = 1;
- } else
- return -EINVAL;
-
- return 0;
-}
-
static struct ib_ah *create_ib_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
struct mlx4_ib_ah *ah)
{
@@ -67,6 +49,7 @@ static struct ib_ah *create_ib_ah(struct ib_pd *pd, st
ah->av.ib.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24));
ah->av.ib.g_slid = ah_attr->src_path_bits;
+ ah->av.ib.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28);
if (ah_attr->ah_flags & IB_AH_GRH) {
ah->av.ib.g_slid |= 0x80;
ah->av.ib.gid_index = ah_attr->grh.sgid_index;
@@ -84,7 +67,6 @@ static struct ib_ah *create_ib_ah(struct ib_pd *pd, st
!(1 << ah->av.ib.stat_rate & dev->caps.stat_rate_support))
--ah->av.ib.stat_rate;
}
- ah->av.ib.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28);
return &ah->ibah;
}
@@ -96,21 +78,38 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd,
struct mlx4_dev *dev = ibdev->dev;
int is_mcast = 0;
struct in6_addr in6;
- u16 vlan_tag;
+ u16 vlan_tag = 0xffff;
+ union ib_gid sgid;
+ struct ib_gid_attr gid_attr;
+ int ret;
memcpy(&in6, ah_attr->grh.dgid.raw, sizeof(in6));
if (rdma_is_multicast_addr(&in6)) {
is_mcast = 1;
- resolve_mcast_mac(&in6, ah->av.eth.mac);
+ rdma_get_mcast_mac(&in6, ah->av.eth.mac);
} else {
- memcpy(ah->av.eth.mac, ah_attr->dmac, 6);
+ memcpy(ah->av.eth.mac, ah_attr->dmac, ETH_ALEN);
}
- vlan_tag = ah_attr->vlan_id;
+ ret = ib_get_cached_gid(pd->device, ah_attr->port_num,
+ ah_attr->grh.sgid_index, &sgid, &gid_attr);
+ if (ret)
+ return ERR_PTR(ret);
+ eth_zero_addr(ah->av.eth.s_mac);
+ if (gid_attr.ndev) {
+ if (is_vlan_dev(gid_attr.ndev))
+ vlan_tag = vlan_dev_vlan_id(gid_attr.ndev);
+ memcpy(ah->av.eth.s_mac, IF_LLADDR(gid_attr.ndev), ETH_ALEN);
+ dev_put(gid_attr.ndev);
+ }
if (vlan_tag < 0x1000)
vlan_tag |= (ah_attr->sl & 7) << 13;
ah->av.eth.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24));
- ah->av.eth.gid_index = ah_attr->grh.sgid_index;
+ ret = mlx4_ib_gid_index_to_real_index(ibdev, ah_attr->port_num, ah_attr->grh.sgid_index);
+ if (ret < 0)
+ return ERR_PTR(ret);
+ ah->av.eth.gid_index = ret;
ah->av.eth.vlan = cpu_to_be16(vlan_tag);
+ ah->av.eth.hop_limit = ah_attr->grh.hop_limit;
if (ah_attr->static_rate) {
ah->av.eth.stat_rate = ah_attr->static_rate + MLX4_STAT_RATE_OFFSET;
while (ah->av.eth.stat_rate > IB_RATE_2_5_GBPS + MLX4_STAT_RATE_OFFSET &&
@@ -168,9 +167,13 @@ int mlx4_ib_query_ah(struct ib_ah *ibah, struct ib_ah_
enum rdma_link_layer ll;
memset(ah_attr, 0, sizeof *ah_attr);
- ah_attr->sl = be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 28;
ah_attr->port_num = be32_to_cpu(ah->av.ib.port_pd) >> 24;
ll = rdma_port_get_link_layer(ibah->device, ah_attr->port_num);
+ if (ll == IB_LINK_LAYER_ETHERNET)
+ ah_attr->sl = be32_to_cpu(ah->av.eth.sl_tclass_flowlabel) >> 29;
+ else
+ ah_attr->sl = be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 28;
+
ah_attr->dlid = ll == IB_LINK_LAYER_INFINIBAND ? be16_to_cpu(ah->av.ib.dlid) : 0;
if (ah->av.ib.stat_rate)
ah_attr->static_rate = ah->av.ib.stat_rate - MLX4_STAT_RATE_OFFSET;
Modified: projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_alias_GUID.c
==============================================================================
--- projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_alias_GUID.c Mon Nov 13 10:45:31 2017 (r325759)
+++ projects/bsd_rdma_4_9/sys/dev/mlx4/mlx4_ib/mlx4_ib_alias_GUID.c Mon Nov 13 10:49:18 2017 (r325760)
@@ -42,6 +42,8 @@
#include <linux/errno.h>
#include <rdma/ib_user_verbs.h>
#include <linux/delay.h>
+#include <linux/math64.h>
+#include <linux/ktime.h>
#include "mlx4_ib.h"
/*
@@ -57,15 +59,19 @@ struct mlx4_alias_guid_work_context {
int query_id;
struct list_head list;
int block_num;
+ ib_sa_comp_mask guid_indexes;
u8 method;
};
struct mlx4_next_alias_guid_work {
u8 port;
u8 block_num;
+ u8 method;
struct mlx4_sriov_alias_guid_info_rec_det rec_det;
};
+static int get_low_record_time_index(struct mlx4_ib_dev *dev, u8 port,
+ int *resched_delay_sec);
void mlx4_ib_update_cache_on_guid_change(struct mlx4_ib_dev *dev, int block_num,
u8 port_num, u8 *p_data)
@@ -119,6 +125,57 @@ ib_sa_comp_mask mlx4_ib_get_aguid_comp_mask_from_ix(in
return IB_SA_COMP_MASK(4 + index);
}
+void mlx4_ib_slave_alias_guid_event(struct mlx4_ib_dev *dev, int slave,
+ int port, int slave_init)
+{
+ __be64 curr_guid, required_guid;
+ int record_num = slave / 8;
+ int index = slave % 8;
+ int port_index = port - 1;
+ unsigned long flags;
+ int do_work = 0;
+
+ spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags);
+ if (dev->sriov.alias_guid.ports_guid[port_index].state_flags &
+ GUID_STATE_NEED_PORT_INIT)
+ goto unlock;
+ if (!slave_init) {
+ curr_guid = *(__be64 *)&dev->sriov.
+ alias_guid.ports_guid[port_index].
+ all_rec_per_port[record_num].
+ all_recs[GUID_REC_SIZE * index];
+ if (curr_guid == cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL) ||
+ !curr_guid)
+ goto unlock;
+ required_guid = cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL);
+ } else {
+ required_guid = mlx4_get_admin_guid(dev->dev, slave, port);
+ if (required_guid == cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL))
+ goto unlock;
+ }
+ *(__be64 *)&dev->sriov.alias_guid.ports_guid[port_index].
+ all_rec_per_port[record_num].
+ all_recs[GUID_REC_SIZE * index] = required_guid;
+ dev->sriov.alias_guid.ports_guid[port_index].
+ all_rec_per_port[record_num].guid_indexes
+ |= mlx4_ib_get_aguid_comp_mask_from_ix(index);
+ dev->sriov.alias_guid.ports_guid[port_index].
+ all_rec_per_port[record_num].status
+ = MLX4_GUID_INFO_STATUS_IDLE;
+ /* set to run immediately */
+ dev->sriov.alias_guid.ports_guid[port_index].
+ all_rec_per_port[record_num].time_to_run = 0;
+ dev->sriov.alias_guid.ports_guid[port_index].
+ all_rec_per_port[record_num].
+ guids_retry_schedule[index] = 0;
+ do_work = 1;
+unlock:
+ spin_unlock_irqrestore(&dev->sriov.alias_guid.ag_work_lock, flags);
+
+ if (do_work)
+ mlx4_ib_init_alias_guid_work(dev, port_index);
+}
+
/*
* Whenever new GUID is set/unset (guid table change) create event and
* notify the relevant slave (master also should be notified).
@@ -134,15 +191,20 @@ void mlx4_ib_notify_slaves_on_guid_change(struct mlx4_
{
int i;
u64 guid_indexes;
- int slave_id;
+ int slave_id, slave_port;
enum slave_port_state new_state;
enum slave_port_state prev_state;
__be64 tmp_cur_ag, form_cache_ag;
enum slave_port_gen_event gen_event;
+ struct mlx4_sriov_alias_guid_info_rec_det *rec;
+ unsigned long flags;
+ __be64 required_value;
if (!mlx4_is_master(dev->dev))
return;
+ rec = &dev->sriov.alias_guid.ports_guid[port_num - 1].
+ all_rec_per_port[block_num];
guid_indexes = be64_to_cpu((__force __be64) dev->sriov.alias_guid.
ports_guid[port_num - 1].
all_rec_per_port[block_num].guid_indexes);
@@ -156,8 +218,13 @@ void mlx4_ib_notify_slaves_on_guid_change(struct mlx4_
continue;
slave_id = (block_num * NUM_ALIAS_GUID_IN_REC) + i ;
- if (slave_id >= dev->dev->num_slaves)
+ if (slave_id >= dev->dev->persist->num_vfs + 1)
return;
+
+ slave_port = mlx4_phys_to_slave_port(dev->dev, slave_id, port_num);
+ if (slave_port < 0) /* this port isn't available for the VF */
+ continue;
+
tmp_cur_ag = *(__be64 *)&p_data[i * GUID_REC_SIZE];
form_cache_ag = get_cached_alias_guid(dev, port_num,
(NUM_ALIAS_GUID_IN_REC * block_num) + i);
@@ -168,8 +235,27 @@ void mlx4_ib_notify_slaves_on_guid_change(struct mlx4_
*/
if (tmp_cur_ag != form_cache_ag)
continue;
- mlx4_gen_guid_change_eqe(dev->dev, slave_id, port_num);
+ spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags);
+ required_value = *(__be64 *)&rec->all_recs[i * GUID_REC_SIZE];
+
+ if (required_value == cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL))
+ required_value = 0;
+
+ if (tmp_cur_ag == required_value) {
+ rec->guid_indexes = rec->guid_indexes &
+ ~mlx4_ib_get_aguid_comp_mask_from_ix(i);
+ } else {
+ /* may notify port down if value is 0 */
+ if (tmp_cur_ag != MLX4_NOT_SET_GUID) {
+ spin_unlock_irqrestore(&dev->sriov.
+ alias_guid.ag_work_lock, flags);
+ continue;
+ }
+ }
+ spin_unlock_irqrestore(&dev->sriov.alias_guid.ag_work_lock,
+ flags);
+ mlx4_gen_guid_change_eqe(dev->dev, slave_id, port_num);
/*2 cases: Valid GUID, and Invalid Guid*/
if (tmp_cur_ag != MLX4_NOT_SET_GUID) { /*valid GUID*/
@@ -190,10 +276,14 @@ void mlx4_ib_notify_slaves_on_guid_change(struct mlx4_
set_and_calc_slave_port_state(dev->dev, slave_id, port_num,
MLX4_PORT_STATE_IB_EVENT_GID_INVALID,
&gen_event);
- pr_debug("sending PORT DOWN event to slave: %d, port: %d\n",
- slave_id, port_num);
- mlx4_gen_port_state_change_eqe(dev->dev, slave_id, port_num,
- MLX4_PORT_CHANGE_SUBTYPE_DOWN);
+ if (gen_event == SLAVE_PORT_GEN_EVENT_DOWN) {
+ pr_debug("sending PORT DOWN event to slave: %d, port: %d\n",
+ slave_id, port_num);
+ mlx4_gen_port_state_change_eqe(dev->dev,
+ slave_id,
+ port_num,
+ MLX4_PORT_CHANGE_SUBTYPE_DOWN);
+ }
}
}
}
@@ -208,6 +298,9 @@ static void aliasguid_query_handler(int status,
int i;
struct mlx4_sriov_alias_guid_info_rec_det *rec;
unsigned long flags, flags1;
+ ib_sa_comp_mask declined_guid_indexes = 0;
+ ib_sa_comp_mask applied_guid_indexes = 0;
+ unsigned int resched_delay_sec = 0;
if (!context)
return;
@@ -218,9 +311,9 @@ static void aliasguid_query_handler(int status,
all_rec_per_port[cb_ctx->block_num];
if (status) {
- rec->status = MLX4_GUID_INFO_STATUS_IDLE;
pr_debug("(port: %d) failed: status = %d\n",
cb_ctx->port, status);
+ rec->time_to_run = ktime_get_ns() + 1 * NSEC_PER_SEC;
goto out;
}
@@ -237,69 +330,101 @@ static void aliasguid_query_handler(int status,
rec = &dev->sriov.alias_guid.ports_guid[port_index].
all_rec_per_port[guid_rec->block_num];
- rec->status = MLX4_GUID_INFO_STATUS_SET;
- rec->method = MLX4_GUID_INFO_RECORD_SET;
-
+ spin_lock_irqsave(&dev->sriov.alias_guid.ag_work_lock, flags);
for (i = 0 ; i < NUM_ALIAS_GUID_IN_REC; i++) {
- __be64 tmp_cur_ag;
- tmp_cur_ag = *(__be64 *)&guid_rec->guid_info_list[i * GUID_REC_SIZE];
- if ((cb_ctx->method == MLX4_GUID_INFO_RECORD_DELETE)
- && (MLX4_NOT_SET_GUID == tmp_cur_ag)) {
- pr_debug("%s:Record num %d in block_num:%d "
- "was deleted by SM,ownership by %d "
- "(0 = driver, 1=sysAdmin, 2=None)\n",
- __func__, i, guid_rec->block_num,
- rec->ownership);
- rec->guid_indexes = rec->guid_indexes &
- ~mlx4_ib_get_aguid_comp_mask_from_ix(i);
+ __be64 sm_response, required_val;
+
+ if (!(cb_ctx->guid_indexes &
+ mlx4_ib_get_aguid_comp_mask_from_ix(i)))
continue;
+ sm_response = *(__be64 *)&guid_rec->guid_info_list
+ [i * GUID_REC_SIZE];
+ required_val = *(__be64 *)&rec->all_recs[i * GUID_REC_SIZE];
+ if (cb_ctx->method == MLX4_GUID_INFO_RECORD_DELETE) {
+ if (required_val ==
+ cpu_to_be64(MLX4_GUID_FOR_DELETE_VAL))
+ goto next_entry;
+
+ /* A new value was set till we got the response */
+ pr_debug("need to set new value %llx, record num %d, block_num:%d\n",
+ (long long)be64_to_cpu(required_val),
+ i, guid_rec->block_num);
+ goto entry_declined;
}
/* check if the SM didn't assign one of the records.
- * if it didn't, if it was not sysadmin request:
- * ask the SM to give a new GUID, (instead of the driver request).
+ * if it didn't, re-ask for.
*/
- if (tmp_cur_ag == MLX4_NOT_SET_GUID) {
- mlx4_ib_warn(&dev->ib_dev, "%s:Record num %d in "
- "block_num: %d was declined by SM, "
- "ownership by %d (0 = driver, 1=sysAdmin,"
- " 2=None)\n", __func__, i,
- guid_rec->block_num, rec->ownership);
- if (rec->ownership == MLX4_GUID_DRIVER_ASSIGN) {
- /* if it is driver assign, asks for new GUID from SM*/
- *(__be64 *)&rec->all_recs[i * GUID_REC_SIZE] =
- MLX4_NOT_SET_GUID;
-
- /* Mark the record as not assigned, and let it
- * be sent again in the next work sched.*/
- rec->status = MLX4_GUID_INFO_STATUS_IDLE;
- rec->guid_indexes |= mlx4_ib_get_aguid_comp_mask_from_ix(i);
- }
+ if (sm_response == MLX4_NOT_SET_GUID) {
+ if (rec->guids_retry_schedule[i] == 0)
+ mlx4_ib_warn(&dev->ib_dev,
+ "%s:Record num %d in block_num: %d was declined by SM\n",
+ __func__, i,
+ guid_rec->block_num);
+ goto entry_declined;
} else {
/* properly assigned record. */
/* We save the GUID we just got from the SM in the
* admin_guid in order to be persistent, and in the
* request from the sm the process will ask for the same GUID */
- if (rec->ownership == MLX4_GUID_SYSADMIN_ASSIGN &&
- tmp_cur_ag != *(__be64 *)&rec->all_recs[i * GUID_REC_SIZE]) {
- /* the sysadmin assignment failed.*/
- mlx4_ib_warn(&dev->ib_dev, "%s: Failed to set"
- " admin guid after SysAdmin "
- "configuration. "
- "Record num %d in block_num:%d "
- "was declined by SM, "
- "new val(0x%llx) was kept\n",
- __func__, i,
- guid_rec->block_num,
- (long long)be64_to_cpu(*(__be64 *) &
- rec->all_recs[i * GUID_REC_SIZE]));
+ if (required_val &&
+ sm_response != required_val) {
+ /* Warn only on first retry */
+ if (rec->guids_retry_schedule[i] == 0)
+ mlx4_ib_warn(&dev->ib_dev, "%s: Failed to set"
+ " admin guid after SysAdmin "
+ "configuration. "
+ "Record num %d in block_num:%d "
+ "was declined by SM, "
+ "new val(0x%llx) was kept, SM returned (0x%llx)\n",
+ __func__, i,
+ guid_rec->block_num,
+ (long long)be64_to_cpu(required_val),
+ (long long)be64_to_cpu(sm_response));
+ goto entry_declined;
} else {
- memcpy(&rec->all_recs[i * GUID_REC_SIZE],
- &guid_rec->guid_info_list[i * GUID_REC_SIZE],
- GUID_REC_SIZE);
+ *(__be64 *)&rec->all_recs[i * GUID_REC_SIZE] =
+ sm_response;
+ if (required_val == 0)
+ mlx4_set_admin_guid(dev->dev,
+ sm_response,
+ (guid_rec->block_num
+ * NUM_ALIAS_GUID_IN_REC) + i,
+ cb_ctx->port);
+ goto next_entry;
}
}
+entry_declined:
+ declined_guid_indexes |= mlx4_ib_get_aguid_comp_mask_from_ix(i);
+ rec->guids_retry_schedule[i] =
+ (rec->guids_retry_schedule[i] == 0) ? 1 :
+ min((unsigned int)60,
+ rec->guids_retry_schedule[i] * 2);
+ /* using the minimum value among all entries in that record */
+ resched_delay_sec = (resched_delay_sec == 0) ?
+ rec->guids_retry_schedule[i] :
+ min(resched_delay_sec,
+ rec->guids_retry_schedule[i]);
+ continue;
+
+next_entry:
+ rec->guids_retry_schedule[i] = 0;
}
+
*** DIFF OUTPUT TRUNCATED AT 1000 LINES ***
More information about the svn-src-projects
mailing list