libibverbs: Add support for XRC QPs

XRC queue pairs: xrc defines two new types of QPs.  The
initiator, or send-side, xrc qp behaves similar to a send-
only RC qp.  xrc send qp's are managed through the existing
QP functions.  The send_wr structure is extended in a back-
wards compatible way to support posting sends on a send xrc
qp, which require specifying the remote xrc srq.

The target, or receive-side, xrc qp behaves differently
than other implemented qp's.  A recv xrc qp can be created,
modified, and destroyed like other qp's through the existing
calls.  The qp_init_attr structure is extended for xrc qp's.

Because xrc recv qp's are bound to an xrcd, rather than a pd,
it is intended to be used among multiple processes.  Any process
with access to an xrcd may allocate and connect an xrc recv qp.
The actual xrc recv qp is allocated and managed by the kernel.
If the owning process explicit destroys the xrc recv qp, it is
destroyed.  However, if the xrc recv qp is left open when the
user process exits or closes its device, then the lifetime of
the xrc recv qp is bound with the lifetime of the xrcd.

Signed-off-by: Sean Hefty <sean.hefty@intel.com>

diff --git a/include/infiniband/driver.h b/include/infiniband/driver.h
index cac48ab40b43aa394347849b0983446bb0384a5a..c86109df3b61bdc32a76d890cbbfb7aa64ee4beb 100644 (file)
--- a/include/infiniband/driver.h
+++ b/include/infiniband/driver.h
@@ -122,6 +122,10 @@ int ibv_cmd_create_qp(struct ibv_pd *pd,
                      struct ibv_qp *qp, struct ibv_qp_init_attr *attr,
                      struct ibv_create_qp *cmd, size_t cmd_size,
                      struct ibv_create_qp_resp *resp, size_t resp_size);
+int ibv_cmd_create_qp_ex(struct ibv_pd *pd,
+                        struct ibv_qp *qp, struct ibv_qp_init_attr_ex *attr_ex,
+                        struct ibv_create_qp *cmd, size_t cmd_size,
+                        struct ibv_create_qp_resp *resp, size_t resp_size);
 int ibv_cmd_query_qp(struct ibv_qp *qp, struct ibv_qp_attr *qp_attr,
                     int attr_mask,
                     struct ibv_qp_init_attr *qp_init_attr,

diff --git a/include/infiniband/kern-abi.h b/include/infiniband/kern-abi.h
index 3d72fa738c964c1cedee6a3f26cfb921b9c03031..b6d5ce9f776f3830c17a17c4bb05ef46636ca31d 100644 (file)
--- a/include/infiniband/kern-abi.h
+++ b/include/infiniband/kern-abi.h
@@ -617,6 +617,11 @@ struct ibv_kern_send_wr {
                        __u32 remote_qkey;
                        __u32 reserved;
                } ud;
+               struct {
+                       __u64 reserved[3];
+                       __u32 reserved2;
+                       __u32 remote_srqn;
+               } xrc;
        } wr;
 };
 

diff --git a/include/infiniband/verbs.h b/include/infiniband/verbs.h
index 58d6927dc6687c533da206ff700ac3fea36854f1..62b1ff64014b218ac309703126f531e7c6faefff 100644 (file)
--- a/include/infiniband/verbs.h
+++ b/include/infiniband/verbs.h
@@ -106,7 +106,8 @@ enum ibv_device_cap_flags {
        IBV_DEVICE_SYS_IMAGE_GUID       = 1 << 11,
        IBV_DEVICE_RC_RNR_NAK_GEN       = 1 << 12,
        IBV_DEVICE_SRQ_RESIZE           = 1 << 13,
-       IBV_DEVICE_N_NOTIFY_CQ          = 1 << 14
+       IBV_DEVICE_N_NOTIFY_CQ          = 1 << 14,
+       IBV_DEVICE_XRC                  = 1 << 20
 };
 
 enum ibv_atomic_cap {
@@ -462,7 +463,9 @@ struct ibv_srq_init_attr_ex {
 enum ibv_qp_type {
        IBV_QPT_RC = 2,
        IBV_QPT_UC,
-       IBV_QPT_UD
+       IBV_QPT_UD,
+       IBV_QPT_XRC_SEND = 9,
+       IBV_QPT_XRC_RECV
 };
 
 struct ibv_qp_cap {
@@ -483,6 +486,24 @@ struct ibv_qp_init_attr {
        int                     sq_sig_all;
 };
 
+enum ibv_qp_init_attr_mask {
+       IBV_QP_INIT_ATTR_XRCD           = 1 << 0,
+       IBV_QP_INIT_ATTR_RESERVED       = 1 << 1
+};
+
+struct ibv_qp_init_attr_ex {
+       void                   *qp_context;
+       struct ibv_cq          *send_cq;
+       struct ibv_cq          *recv_cq;
+       struct ibv_srq         *srq;
+       struct ibv_qp_cap       cap;
+       enum ibv_qp_type        qp_type;
+       int                     sq_sig_all;
+
+       uint64_t                comp_mask;
+       struct ibv_xrcd        *xrcd;
+};
+
 enum ibv_qp_attr_mask {
        IBV_QP_STATE                    = 1 <<  0,
        IBV_QP_CUR_STATE                = 1 <<  1,
@@ -598,6 +619,11 @@ struct ibv_send_wr {
                        uint32_t        remote_qpn;
                        uint32_t        remote_qkey;
                } ud;
+               struct {
+                       uint64_t        reserved[3];
+                       uint32_t        reserved2;
+                       uint32_t        remote_srqn;
+               } xrc;
        } wr;
 };
 
@@ -643,7 +669,8 @@ struct ibv_srq {
 };
 
 enum ibv_qp_mask {
-       IBV_QP_RESERVED                 = 1 << 0
+       IBV_QP_XRCD                     = 1 << 0,
+       IBV_QP_RESERVED                 = 1 << 1
 };
 
 struct ibv_qp {
@@ -663,6 +690,7 @@ struct ibv_qp {
        uint32_t                events_completed;
 
        uint32_t                comp_mask;
+       struct ibv_xrcd        *xrcd;
 };
 
 enum ibv_comp_channel_mask {
@@ -820,6 +848,8 @@ struct verbs_context {
        int (*drv_new_func1) ();        new corresponding provider call of func1
        int (*lib_new_func1) ();        New library call func1
        */
+       struct ibv_qp *         (*create_qp_ex)(struct ibv_pd *pd,
+                                               struct ibv_qp_init_attr_ex *qp_init_attr_ex);
        struct ibv_srq *        (*create_srq_ex)(struct ibv_pd *pd,
                                                 struct ibv_srq_init_attr_ex *srq_init_attr_ex);
        struct ibv_xrcd *       (*open_xrcd)(struct ibv_context *context,
@@ -1137,6 +1167,9 @@ static inline int ibv_post_srq_recv(struct ibv_srq *srq,
 struct ibv_qp *ibv_create_qp(struct ibv_pd *pd,
                             struct ibv_qp_init_attr *qp_init_attr);
 
+struct ibv_qp *ibv_create_qp_ex(struct ibv_pd *pd,
+                               struct ibv_qp_init_attr_ex *qp_init_attr_ex);
+
 /**
  * ibv_modify_qp - Modify a queue pair.
  */

diff --git a/src/cmd.c b/src/cmd.c
index 2a64a2b677ef071f79e5e49b03e3ea957014b7a5..b9d96f9095b70c6d985e0c56aaf3fc44ed5e9f61 100644 (file)
--- a/src/cmd.c
+++ b/src/cmd.c
@@ -628,43 +628,56 @@ int ibv_cmd_destroy_srq(struct ibv_srq *srq)
        return 0;
 }
 
-int ibv_cmd_create_qp(struct ibv_pd *pd,
-                     struct ibv_qp *qp, struct ibv_qp_init_attr *attr,
-                     struct ibv_create_qp *cmd, size_t cmd_size,
-                     struct ibv_create_qp_resp *resp, size_t resp_size)
+int ibv_cmd_create_qp_ex(struct ibv_pd *pd,
+                        struct ibv_qp *qp, struct ibv_qp_init_attr_ex *attr_ex,
+                        struct ibv_create_qp *cmd, size_t cmd_size,
+                        struct ibv_create_qp_resp *resp, size_t resp_size)
 {
+       struct ibv_context *context;
+
        IBV_INIT_CMD_RESP(cmd, cmd_size, CREATE_QP, resp, resp_size);
 
        cmd->user_handle     = (uintptr_t) qp;
-       cmd->pd_handle       = pd->handle;
-       cmd->send_cq_handle  = attr->send_cq->handle;
-       cmd->recv_cq_handle  = attr->recv_cq->handle;
-       cmd->srq_handle      = attr->srq ? attr->srq->handle : 0;
-       cmd->max_send_wr     = attr->cap.max_send_wr;
-       cmd->max_recv_wr     = attr->cap.max_recv_wr;
-       cmd->max_send_sge    = attr->cap.max_send_sge;
-       cmd->max_recv_sge    = attr->cap.max_recv_sge;
-       cmd->max_inline_data = attr->cap.max_inline_data;
-       cmd->sq_sig_all      = attr->sq_sig_all;
-       cmd->qp_type         = attr->qp_type;
-       cmd->is_srq          = !!attr->srq;
+
+       if (attr_ex->comp_mask & IBV_QP_INIT_ATTR_XRCD) {
+               context         = attr_ex->xrcd->context;
+               cmd->pd_handle  = attr_ex->xrcd->handle;
+       } else {
+               context         = pd->context;
+               cmd->pd_handle  = pd->handle;
+               cmd->send_cq_handle = attr_ex->send_cq->handle;
+
+               if (attr_ex->qp_type != IBV_QPT_XRC_SEND) {
+                       cmd->recv_cq_handle = attr_ex->recv_cq->handle;
+                       cmd->srq_handle = attr_ex->srq ? attr_ex->srq->handle : 0;
+               }
+       }
+
+       cmd->max_send_wr     = attr_ex->cap.max_send_wr;
+       cmd->max_recv_wr     = attr_ex->cap.max_recv_wr;
+       cmd->max_send_sge    = attr_ex->cap.max_send_sge;
+       cmd->max_recv_sge    = attr_ex->cap.max_recv_sge;
+       cmd->max_inline_data = attr_ex->cap.max_inline_data;
+       cmd->sq_sig_all      = attr_ex->sq_sig_all;
+       cmd->qp_type         = attr_ex->qp_type;
+       cmd->is_srq          = !!attr_ex->srq;
        cmd->reserved        = 0;
 
-       if (write(pd->context->cmd_fd, cmd, cmd_size) != cmd_size)
+       if (write(context->cmd_fd, cmd, cmd_size) != cmd_size)
                return errno;
 
        (void) VALGRIND_MAKE_MEM_DEFINED(resp, resp_size);
 
        qp->handle                = resp->qp_handle;
        qp->qp_num                = resp->qpn;
-       qp->context               = pd->context;
+       qp->context               = context;
 
        if (abi_ver > 3) {
-               attr->cap.max_recv_sge    = resp->max_recv_sge;
-               attr->cap.max_send_sge    = resp->max_send_sge;
-               attr->cap.max_recv_wr     = resp->max_recv_wr;
-               attr->cap.max_send_wr     = resp->max_send_wr;
-               attr->cap.max_inline_data = resp->max_inline_data;
+               attr_ex->cap.max_recv_sge    = resp->max_recv_sge;
+               attr_ex->cap.max_send_sge    = resp->max_send_sge;
+               attr_ex->cap.max_recv_wr     = resp->max_recv_wr;
+               attr_ex->cap.max_send_wr     = resp->max_send_wr;
+               attr_ex->cap.max_inline_data = resp->max_inline_data;
        }
 
        if (abi_ver == 4) {
@@ -686,6 +699,23 @@ int ibv_cmd_create_qp(struct ibv_pd *pd,
        return 0;
 }
 
+int ibv_cmd_create_qp(struct ibv_pd *pd,
+                     struct ibv_qp *qp, struct ibv_qp_init_attr *attr,
+                     struct ibv_create_qp *cmd, size_t cmd_size,
+                     struct ibv_create_qp_resp *resp, size_t resp_size)
+{
+       struct ibv_qp_init_attr_ex attr_ex;
+       int ret;
+
+       memcpy(&attr_ex, attr, sizeof *attr);
+       attr_ex.comp_mask = 0;
+       ret = ibv_cmd_create_qp_ex(pd, qp, &attr_ex, cmd, cmd_size, resp, resp_size);
+       if (!ret)
+               memcpy(attr, &attr_ex, sizeof *attr);
+
+       return ret;
+}
+
 int ibv_cmd_query_qp(struct ibv_qp *qp, struct ibv_qp_attr *attr,
                     int attr_mask,
                     struct ibv_qp_init_attr *init_attr,

diff --git a/src/libibverbs.map b/src/libibverbs.map
index 92bfabd37790f5dc8c9ccaa7520fa3cd142c4eac..9ee1cf21ae8f5577e833b24438c5add4ac4f9607 100644 (file)
--- a/src/libibverbs.map
+++ b/src/libibverbs.map
@@ -104,5 +104,7 @@ IBVERBS_1.1 {
                ibv_cmd_close_xrcd;
                ibv_create_srq_ex;
                ibv_cmd_create_srq_ex;
+               ibv_create_qp_ex;
+               ibv_cmd_create_qp_ex;
                
 } IBVERBS_1.0;

diff --git a/src/verbs.c b/src/verbs.c
index 2690b2904f07d9bb8ca3d0f18426823b0575f7c9..9c6391715627e3cf397970fcb504d3da4fc71801 100644 (file)
--- a/src/verbs.c
+++ b/src/verbs.c
@@ -462,19 +462,45 @@ int __ibv_destroy_srq(struct ibv_srq *srq)
 }
 default_symver(__ibv_destroy_srq, ibv_destroy_srq);
 
-struct ibv_qp *__ibv_create_qp(struct ibv_pd *pd,
-                              struct ibv_qp_init_attr *qp_init_attr)
+struct ibv_qp *__ibv_create_qp_ex(struct ibv_pd *pd,
+                                 struct ibv_qp_init_attr_ex *qp_init_attr_ex)
 {
-       struct ibv_qp *qp = pd->context->ops.create_qp(pd, qp_init_attr);
+       struct verbs_context *context_ex;
+       struct ibv_context *context = NULL;
+       struct ibv_qp *qp;
+
+       if (pd) {
+               context = pd->context;
+       } else {
+               if (qp_init_attr_ex->comp_mask & IBV_QP_INIT_ATTR_XRCD)
+                       context = qp_init_attr_ex->xrcd->context;
+       }
+       if (!context) {
+               errno = EINVAL;
+               return NULL;
+       }
+
+       context_ex = verbs_get_ctx(context);
+       if (qp_init_attr_ex->comp_mask) {
+               if (!context_ex->create_qp_ex ||
+                   qp_init_attr_ex->comp_mask >= IBV_QP_INIT_ATTR_RESERVED) {
+                       errno = ENOSYS;
+                       return NULL;
+               }
+               qp = context_ex->create_qp_ex(pd, qp_init_attr_ex);
+       } else {
+               qp = context->ops.create_qp(pd, (struct ibv_qp_init_attr *)
+                                               qp_init_attr_ex);
+       }
 
        if (qp) {
-               qp->context          = pd->context;
-               qp->qp_context       = qp_init_attr->qp_context;
+               qp->context          = context;
+               qp->qp_context       = qp_init_attr_ex->qp_context;
                qp->pd               = pd;
-               qp->send_cq          = qp_init_attr->send_cq;
-               qp->recv_cq          = qp_init_attr->recv_cq;
-               qp->srq              = qp_init_attr->srq;
-               qp->qp_type          = qp_init_attr->qp_type;
+               qp->send_cq          = qp_init_attr_ex->send_cq;
+               qp->recv_cq          = qp_init_attr_ex->recv_cq;
+               qp->srq              = qp_init_attr_ex->srq;
+               qp->qp_type          = qp_init_attr_ex->qp_type;
                qp->state            = IBV_QPS_RESET;
                qp->events_completed = 0;
                pthread_mutex_init(&qp->mutex, NULL);
@@ -483,6 +509,22 @@ struct ibv_qp *__ibv_create_qp(struct ibv_pd *pd,
 
        return qp;
 }
+default_symver(__ibv_create_qp_ex, ibv_create_qp_ex);
+
+struct ibv_qp *__ibv_create_qp(struct ibv_pd *pd,
+                              struct ibv_qp_init_attr *qp_init_attr)
+{
+       struct ibv_qp_init_attr_ex qp_init_attr_ex;
+       struct ibv_qp *qp;
+
+       memcpy(&qp_init_attr_ex, qp_init_attr, sizeof *qp_init_attr);
+       qp_init_attr_ex.comp_mask = 0;
+       qp = ibv_create_qp_ex(pd, &qp_init_attr_ex);
+       if (qp)
+               memcpy(qp_init_attr, &qp_init_attr_ex, sizeof *qp_init_attr);
+
+       return qp;
+}
 default_symver(__ibv_create_qp, ibv_create_qp);
 
 int __ibv_query_qp(struct ibv_qp *qp, struct ibv_qp_attr *attr,