[IB] uverbs: have kernel return QP capabilities

Move the computation of QP capabilities (max scatter/gather entries,
max inline data, etc) into the kernel, and have the uverbs module
return the values as part of the create QP response.  This keeps
precise knowledge of device limits in the low-level kernel driver.

This requires an ABI bump, so while we're making changes, get rid of
the max_sge parameter for the modify SRQ command -- it's not used and
shouldn't be there.

Signed-off-by: Jack Morgenstein <jackm@mellanox.co.il>
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

6 files changed, 84 insertions, 9 deletions

diff --git a/drivers/infiniband/hw/mthca/mthca_cmd.c b/drivers/infiniband/hw/mthca/mthca_cmd.c
index 49f211d55df7..9ed34587fc5c 100644
--- a/drivers/infiniband/hw/mthca/mthca_cmd.c
+++ b/drivers/infiniband/hw/mthca/mthca_cmd.c
@@ -1060,6 +1060,8 @@ int mthca_QUERY_DEV_LIM(struct mthca_dev *dev,
                 dev_lim->hca.arbel.resize_srq = field & 1;
                 MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_SG_RQ_OFFSET);
                 dev_lim->max_sg = min_t(int, field, dev_lim->max_sg);
+                MTHCA_GET(size, outbox, QUERY_DEV_LIM_MAX_DESC_SZ_RQ_OFFSET);
+                dev_lim->max_desc_sz = min_t(int, size, dev_lim->max_desc_sz);
                 MTHCA_GET(size, outbox, QUERY_DEV_LIM_MPT_ENTRY_SZ_OFFSET);
                 dev_lim->mpt_entry_sz = size;
                 MTHCA_GET(field, outbox, QUERY_DEV_LIM_PBL_SZ_OFFSET);
diff --git a/drivers/infiniband/hw/mthca/mthca_dev.h b/drivers/infiniband/hw/mthca/mthca_dev.h
index 808037f25c78..497ff794ef6a 100644
--- a/drivers/infiniband/hw/mthca/mthca_dev.h
+++ b/drivers/infiniband/hw/mthca/mthca_dev.h
@@ -131,6 +131,7 @@ struct mthca_limits {
         int      max_sg;
         int      num_qps;
         int      max_wqes;
+        int      max_desc_sz;
         int      max_qp_init_rdma;
         int      reserved_qps;
         int      num_srqs;
diff --git a/drivers/infiniband/hw/mthca/mthca_main.c b/drivers/infiniband/hw/mthca/mthca_main.c
index 16594d1342df..147f248a8073 100644
--- a/drivers/infiniband/hw/mthca/mthca_main.c
+++ b/drivers/infiniband/hw/mthca/mthca_main.c
@@ -168,6 +168,7 @@ static int __devinit mthca_dev_lim(struct mthca_dev *mdev, struct mthca_dev_lim
         mdev->limits.max_srq_wqes       = dev_lim->max_srq_sz;
         mdev->limits.reserved_srqs      = dev_lim->reserved_srqs;
         mdev->limits.reserved_eecs      = dev_lim->reserved_eecs;
+        mdev->limits.max_desc_sz        = dev_lim->max_desc_sz;
         /*
          * Subtract 1 from the limit because we need to allocate a
          * spare CQE so the HCA HW can tell the difference between an
diff --git a/drivers/infiniband/hw/mthca/mthca_provider.c b/drivers/infiniband/hw/mthca/mthca_provider.c
index e78259b2664b..4cc7e2846df1 100644
--- a/drivers/infiniband/hw/mthca/mthca_provider.c
+++ b/drivers/infiniband/hw/mthca/mthca_provider.c
@@ -616,11 +616,11 @@ static struct ib_qp *mthca_create_qp(struct ib_pd *pd,
                 return ERR_PTR(err);
         }
 
-        init_attr->cap.max_inline_data = 0;
         init_attr->cap.max_send_wr     = qp->sq.max;
         init_attr->cap.max_recv_wr     = qp->rq.max;
         init_attr->cap.max_send_sge    = qp->sq.max_gs;
         init_attr->cap.max_recv_sge    = qp->rq.max_gs;
+        init_attr->cap.max_inline_data = qp->max_inline_data;
 
         return &qp->ibqp;
 }
diff --git a/drivers/infiniband/hw/mthca/mthca_provider.h b/drivers/infiniband/hw/mthca/mthca_provider.h
index bcd4b01a339c..1e73947b4702 100644
--- a/drivers/infiniband/hw/mthca/mthca_provider.h
+++ b/drivers/infiniband/hw/mthca/mthca_provider.h
@@ -251,6 +251,7 @@ struct mthca_qp {
         struct mthca_wq        sq;
         enum ib_sig_type       sq_policy;
         int                    send_wqe_offset;
+        int                    max_inline_data;
 
         u64                   *wrid;
         union mthca_buf        queue;
diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c
index 8852ea477c21..7f39af44b274 100644
--- a/drivers/infiniband/hw/mthca/mthca_qp.c
+++ b/drivers/infiniband/hw/mthca/mthca_qp.c
@@ -885,6 +885,48 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
         return err;
 }
 
+static void mthca_adjust_qp_caps(struct mthca_dev *dev,
+                                 struct mthca_pd *pd,
+                                 struct mthca_qp *qp)
+{
+        int max_data_size;
+
+        /*
+         * Calculate the maximum size of WQE s/g segments, excluding
+         * the next segment and other non-data segments.
+         */
+        max_data_size = min(dev->limits.max_desc_sz, 1 << qp->sq.wqe_shift) -
+                sizeof (struct mthca_next_seg);
+
+        switch (qp->transport) {
+        case MLX:
+                max_data_size -= 2 * sizeof (struct mthca_data_seg);
+                break;
+
+        case UD:
+                if (mthca_is_memfree(dev))
+                        max_data_size -= sizeof (struct mthca_arbel_ud_seg);
+                else
+                        max_data_size -= sizeof (struct mthca_tavor_ud_seg);
+                break;
+
+        default:
+                max_data_size -= sizeof (struct mthca_raddr_seg);
+                break;
+        }
+
+        /* We don't support inline data for kernel QPs (yet). */
+        if (!pd->ibpd.uobject)
+                qp->max_inline_data = 0;
+        else
+                qp->max_inline_data = max_data_size - MTHCA_INLINE_HEADER_SIZE;
+
+        qp->sq.max_gs = max_data_size / sizeof (struct mthca_data_seg);
+        qp->rq.max_gs = (min(dev->limits.max_desc_sz, 1 << qp->rq.wqe_shift) -
+                        sizeof (struct mthca_next_seg)) /
+                        sizeof (struct mthca_data_seg);
+}
+
 /*
  * Allocate and register buffer for WQEs.  qp->rq.max, sq.max,
  * rq.max_gs and sq.max_gs must all be assigned.
@@ -902,27 +944,53 @@ static int mthca_alloc_wqe_buf(struct mthca_dev *dev,
         size = sizeof (struct mthca_next_seg) +
                 qp->rq.max_gs * sizeof (struct mthca_data_seg);
 
+        if (size > dev->limits.max_desc_sz)
+                return -EINVAL;
+
         for (qp->rq.wqe_shift = 6; 1 << qp->rq.wqe_shift < size;
              qp->rq.wqe_shift++)
                 ; /* nothing */
 
-        size = sizeof (struct mthca_next_seg) +
-                qp->sq.max_gs * sizeof (struct mthca_data_seg);
+        size = qp->sq.max_gs * sizeof (struct mthca_data_seg);
         switch (qp->transport) {
         case MLX:
                 size += 2 * sizeof (struct mthca_data_seg);
                 break;
+
         case UD:
-                if (mthca_is_memfree(dev))
-                        size += sizeof (struct mthca_arbel_ud_seg);
-                else
-                        size += sizeof (struct mthca_tavor_ud_seg);
+                size += mthca_is_memfree(dev) ?
+                        sizeof (struct mthca_arbel_ud_seg) :
+                        sizeof (struct mthca_tavor_ud_seg);
                 break;
+
+        case UC:
+                size += sizeof (struct mthca_raddr_seg);
+                break;
+
+        case RC:
+                size += sizeof (struct mthca_raddr_seg);
+                /*
+                 * An atomic op will require an atomic segment, a
+                 * remote address segment and one scatter entry.
+                 */
+                size = max_t(int, size,
+                             sizeof (struct mthca_atomic_seg) +
+                             sizeof (struct mthca_raddr_seg) +
+                             sizeof (struct mthca_data_seg));
+                break;
+
         default:
-                /* bind seg is as big as atomic + raddr segs */
-                size += sizeof (struct mthca_bind_seg);
+                break;
         }
 
+        /* Make sure that we have enough space for a bind request */
+        size = max_t(int, size, sizeof (struct mthca_bind_seg));
+
+        size += sizeof (struct mthca_next_seg);
+
+        if (size > dev->limits.max_desc_sz)
+                return -EINVAL;
+
         for (qp->sq.wqe_shift = 6; 1 << qp->sq.wqe_shift < size;
              qp->sq.wqe_shift++)
                 ; /* nothing */
@@ -1066,6 +1134,8 @@ static int mthca_alloc_qp_common(struct mthca_dev *dev,
                 return ret;
         }
 
+        mthca_adjust_qp_caps(dev, pd, qp);
+
         /*
          * If this is a userspace QP, we're done now.  The doorbells
          * will be allocated and buffers will be initialized in