aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/infiniband/hw
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/infiniband/hw')
-rw-r--r--drivers/infiniband/hw/amso1100/c2_qp.c14
-rw-r--r--drivers/infiniband/hw/cxgb3/iwch_qp.c32
-rw-r--r--drivers/infiniband/hw/ehca/ehca_reqs.c67
-rw-r--r--drivers/infiniband/hw/ipath/ipath_driver.c10
-rw-r--r--drivers/infiniband/hw/mlx4/main.c2
-rw-r--r--drivers/infiniband/hw/mlx4/qp.c25
6 files changed, 77 insertions, 73 deletions
diff --git a/drivers/infiniband/hw/amso1100/c2_qp.c b/drivers/infiniband/hw/amso1100/c2_qp.c
index a6d8944..ad51886 100644
--- a/drivers/infiniband/hw/amso1100/c2_qp.c
+++ b/drivers/infiniband/hw/amso1100/c2_qp.c
@@ -798,8 +798,10 @@ int c2_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr,
u8 actual_sge_count;
u32 msg_size;
- if (qp->state > IB_QPS_RTS)
- return -EINVAL;
+ if (qp->state > IB_QPS_RTS) {
+ err = -EINVAL;
+ goto out;
+ }
while (ib_wr) {
@@ -930,6 +932,7 @@ int c2_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr,
ib_wr = ib_wr->next;
}
+out:
if (err)
*bad_wr = ib_wr;
return err;
@@ -944,8 +947,10 @@ int c2_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *ib_wr,
unsigned long lock_flags;
int err = 0;
- if (qp->state > IB_QPS_RTS)
- return -EINVAL;
+ if (qp->state > IB_QPS_RTS) {
+ err = -EINVAL;
+ goto out;
+ }
/*
* Try and post each work request
@@ -998,6 +1003,7 @@ int c2_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *ib_wr,
ib_wr = ib_wr->next;
}
+out:
if (err)
*bad_wr = ib_wr;
return err;
diff --git a/drivers/infiniband/hw/cxgb3/iwch_qp.c b/drivers/infiniband/hw/cxgb3/iwch_qp.c
index 1cecf98..3eb8cec 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_qp.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_qp.c
@@ -365,18 +365,19 @@ int iwch_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
spin_lock_irqsave(&qhp->lock, flag);
if (qhp->attr.state > IWCH_QP_STATE_RTS) {
spin_unlock_irqrestore(&qhp->lock, flag);
- return -EINVAL;
+ err = -EINVAL;
+ goto out;
}
num_wrs = Q_FREECNT(qhp->wq.sq_rptr, qhp->wq.sq_wptr,
qhp->wq.sq_size_log2);
if (num_wrs <= 0) {
spin_unlock_irqrestore(&qhp->lock, flag);
- return -ENOMEM;
+ err = -ENOMEM;
+ goto out;
}
while (wr) {
if (num_wrs == 0) {
err = -ENOMEM;
- *bad_wr = wr;
break;
}
idx = Q_PTR2IDX(qhp->wq.wptr, qhp->wq.size_log2);
@@ -428,10 +429,8 @@ int iwch_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
wr->opcode);
err = -EINVAL;
}
- if (err) {
- *bad_wr = wr;
+ if (err)
break;
- }
wqe->send.wrid.id0.hi = qhp->wq.sq_wptr;
sqp->wr_id = wr->wr_id;
sqp->opcode = wr2opcode(t3_wr_opcode);
@@ -454,6 +453,10 @@ int iwch_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
}
spin_unlock_irqrestore(&qhp->lock, flag);
ring_doorbell(qhp->wq.doorbell, qhp->wq.qpid);
+
+out:
+ if (err)
+ *bad_wr = wr;
return err;
}
@@ -471,18 +474,19 @@ int iwch_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
spin_lock_irqsave(&qhp->lock, flag);
if (qhp->attr.state > IWCH_QP_STATE_RTS) {
spin_unlock_irqrestore(&qhp->lock, flag);
- return -EINVAL;
+ err = -EINVAL;
+ goto out;
}
num_wrs = Q_FREECNT(qhp->wq.rq_rptr, qhp->wq.rq_wptr,
qhp->wq.rq_size_log2) - 1;
if (!wr) {
spin_unlock_irqrestore(&qhp->lock, flag);
- return -EINVAL;
+ err = -ENOMEM;
+ goto out;
}
while (wr) {
if (wr->num_sge > T3_MAX_SGE) {
err = -EINVAL;
- *bad_wr = wr;
break;
}
idx = Q_PTR2IDX(qhp->wq.wptr, qhp->wq.size_log2);
@@ -494,10 +498,10 @@ int iwch_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
err = build_zero_stag_recv(qhp, wqe, wr);
else
err = -ENOMEM;
- if (err) {
- *bad_wr = wr;
+
+ if (err)
break;
- }
+
build_fw_riwrh((void *) wqe, T3_WR_RCV, T3_COMPLETION_FLAG,
Q_GENBIT(qhp->wq.wptr, qhp->wq.size_log2),
0, sizeof(struct t3_receive_wr) >> 3, T3_SOPEOP);
@@ -511,6 +515,10 @@ int iwch_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
}
spin_unlock_irqrestore(&qhp->lock, flag);
ring_doorbell(qhp->wq.doorbell, qhp->wq.qpid);
+
+out:
+ if (err)
+ *bad_wr = wr;
return err;
}
diff --git a/drivers/infiniband/hw/ehca/ehca_reqs.c b/drivers/infiniband/hw/ehca/ehca_reqs.c
index 8fd88cd..e3ec7fd 100644
--- a/drivers/infiniband/hw/ehca/ehca_reqs.c
+++ b/drivers/infiniband/hw/ehca/ehca_reqs.c
@@ -400,7 +400,6 @@ static inline void map_ib_wc_status(u32 cqe_status,
static inline int post_one_send(struct ehca_qp *my_qp,
struct ib_send_wr *cur_send_wr,
- struct ib_send_wr **bad_send_wr,
int hidden)
{
struct ehca_wqe *wqe_p;
@@ -412,8 +411,6 @@ static inline int post_one_send(struct ehca_qp *my_qp,
wqe_p = ipz_qeit_get_inc(&my_qp->ipz_squeue);
if (unlikely(!wqe_p)) {
/* too many posted work requests: queue overflow */
- if (bad_send_wr)
- *bad_send_wr = cur_send_wr;
ehca_err(my_qp->ib_qp.device, "Too many posted WQEs "
"qp_num=%x", my_qp->ib_qp.qp_num);
return -ENOMEM;
@@ -433,8 +430,6 @@ static inline int post_one_send(struct ehca_qp *my_qp,
*/
if (unlikely(ret)) {
my_qp->ipz_squeue.current_q_offset = start_offset;
- if (bad_send_wr)
- *bad_send_wr = cur_send_wr;
ehca_err(my_qp->ib_qp.device, "Could not write WQE "
"qp_num=%x", my_qp->ib_qp.qp_num);
return -EINVAL;
@@ -448,7 +443,6 @@ int ehca_post_send(struct ib_qp *qp,
struct ib_send_wr **bad_send_wr)
{
struct ehca_qp *my_qp = container_of(qp, struct ehca_qp, ib_qp);
- struct ib_send_wr *cur_send_wr;
int wqe_cnt = 0;
int ret = 0;
unsigned long flags;
@@ -457,7 +451,8 @@ int ehca_post_send(struct ib_qp *qp,
if (unlikely(my_qp->state < IB_QPS_RTS)) {
ehca_err(qp->device, "Invalid QP state qp_state=%d qpn=%x",
my_qp->state, qp->qp_num);
- return -EINVAL;
+ ret = -EINVAL;
+ goto out;
}
/* LOCK the QUEUE */
@@ -476,24 +471,21 @@ int ehca_post_send(struct ib_qp *qp,
struct ib_send_wr circ_wr;
memset(&circ_wr, 0, sizeof(circ_wr));
circ_wr.opcode = IB_WR_RDMA_READ;
- post_one_send(my_qp, &circ_wr, NULL, 1); /* ignore retcode */
+ post_one_send(my_qp, &circ_wr, 1); /* ignore retcode */
wqe_cnt++;
ehca_dbg(qp->device, "posted circ wr qp_num=%x", qp->qp_num);
my_qp->message_count = my_qp->packet_count = 0;
}
/* loop processes list of send reqs */
- for (cur_send_wr = send_wr; cur_send_wr != NULL;
- cur_send_wr = cur_send_wr->next) {
- ret = post_one_send(my_qp, cur_send_wr, bad_send_wr, 0);
+ while (send_wr) {
+ ret = post_one_send(my_qp, send_wr, 0);
if (unlikely(ret)) {
- /* if one or more WQEs were successful, don't fail */
- if (wqe_cnt)
- ret = 0;
goto post_send_exit0;
}
wqe_cnt++;
- } /* eof for cur_send_wr */
+ send_wr = send_wr->next;
+ }
post_send_exit0:
iosync(); /* serialize GAL register access */
@@ -503,6 +495,10 @@ post_send_exit0:
my_qp, qp->qp_num, wqe_cnt, ret);
my_qp->message_count += wqe_cnt;
spin_unlock_irqrestore(&my_qp->spinlock_s, flags);
+
+out:
+ if (ret)
+ *bad_send_wr = send_wr;
return ret;
}
@@ -511,7 +507,6 @@ static int internal_post_recv(struct ehca_qp *my_qp,
struct ib_recv_wr *recv_wr,
struct ib_recv_wr **bad_recv_wr)
{
- struct ib_recv_wr *cur_recv_wr;
struct ehca_wqe *wqe_p;
int wqe_cnt = 0;
int ret = 0;
@@ -522,27 +517,23 @@ static int internal_post_recv(struct ehca_qp *my_qp,
if (unlikely(!HAS_RQ(my_qp))) {
ehca_err(dev, "QP has no RQ ehca_qp=%p qp_num=%x ext_type=%d",
my_qp, my_qp->real_qp_num, my_qp->ext_type);
- return -ENODEV;
+ ret = -ENODEV;
+ goto out;
}
/* LOCK the QUEUE */
spin_lock_irqsave(&my_qp->spinlock_r, flags);
- /* loop processes list of send reqs */
- for (cur_recv_wr = recv_wr; cur_recv_wr != NULL;
- cur_recv_wr = cur_recv_wr->next) {
+ /* loop processes list of recv reqs */
+ while (recv_wr) {
u64 start_offset = my_qp->ipz_rqueue.current_q_offset;
/* get pointer next to free WQE */
wqe_p = ipz_qeit_get_inc(&my_qp->ipz_rqueue);
if (unlikely(!wqe_p)) {
/* too many posted work requests: queue overflow */
- if (bad_recv_wr)
- *bad_recv_wr = cur_recv_wr;
- if (wqe_cnt == 0) {
- ret = -ENOMEM;
- ehca_err(dev, "Too many posted WQEs "
- "qp_num=%x", my_qp->real_qp_num);
- }
+ ret = -ENOMEM;
+ ehca_err(dev, "Too many posted WQEs "
+ "qp_num=%x", my_qp->real_qp_num);
goto post_recv_exit0;
}
/*
@@ -552,7 +543,7 @@ static int internal_post_recv(struct ehca_qp *my_qp,
rq_map_idx = start_offset / my_qp->ipz_rqueue.qe_size;
/* write a RECV WQE into the QUEUE */
- ret = ehca_write_rwqe(&my_qp->ipz_rqueue, wqe_p, cur_recv_wr,
+ ret = ehca_write_rwqe(&my_qp->ipz_rqueue, wqe_p, recv_wr,
rq_map_idx);
/*
* if something failed,
@@ -560,22 +551,20 @@ static int internal_post_recv(struct ehca_qp *my_qp,
*/
if (unlikely(ret)) {
my_qp->ipz_rqueue.current_q_offset = start_offset;
- *bad_recv_wr = cur_recv_wr;
- if (wqe_cnt == 0) {
- ret = -EINVAL;
- ehca_err(dev, "Could not write WQE "
- "qp_num=%x", my_qp->real_qp_num);
- }
+ ret = -EINVAL;
+ ehca_err(dev, "Could not write WQE "
+ "qp_num=%x", my_qp->real_qp_num);
goto post_recv_exit0;
}
qmap_entry = &my_qp->rq_map.map[rq_map_idx];
- qmap_entry->app_wr_id = get_app_wr_id(cur_recv_wr->wr_id);
+ qmap_entry->app_wr_id = get_app_wr_id(recv_wr->wr_id);
qmap_entry->reported = 0;
qmap_entry->cqe_req = 1;
wqe_cnt++;
- } /* eof for cur_recv_wr */
+ recv_wr = recv_wr->next;
+ } /* eof for recv_wr */
post_recv_exit0:
iosync(); /* serialize GAL register access */
@@ -584,6 +573,11 @@ post_recv_exit0:
ehca_dbg(dev, "ehca_qp=%p qp_num=%x wqe_cnt=%d ret=%i",
my_qp, my_qp->real_qp_num, wqe_cnt, ret);
spin_unlock_irqrestore(&my_qp->spinlock_r, flags);
+
+out:
+ if (ret)
+ *bad_recv_wr = recv_wr;
+
return ret;
}
@@ -597,6 +591,7 @@ int ehca_post_recv(struct ib_qp *qp,
if (unlikely(my_qp->state == IB_QPS_RESET)) {
ehca_err(qp->device, "Invalid QP state qp_state=%d qpn=%x",
my_qp->state, qp->qp_num);
+ *bad_recv_wr = recv_wr;
return -EINVAL;
}
diff --git a/drivers/infiniband/hw/ipath/ipath_driver.c b/drivers/infiniband/hw/ipath/ipath_driver.c
index 013d138..d2787fe 100644
--- a/drivers/infiniband/hw/ipath/ipath_driver.c
+++ b/drivers/infiniband/hw/ipath/ipath_driver.c
@@ -39,6 +39,7 @@
#include <linux/delay.h>
#include <linux/netdevice.h>
#include <linux/vmalloc.h>
+#include <linux/bitmap.h>
#include "ipath_kernel.h"
#include "ipath_verbs.h"
@@ -1697,7 +1698,7 @@ void ipath_chg_pioavailkernel(struct ipath_devdata *dd, unsigned start,
unsigned len, int avail)
{
unsigned long flags;
- unsigned end, cnt = 0, next;
+ unsigned end, cnt = 0;
/* There are two bits per send buffer (busy and generation) */
start *= 2;
@@ -1748,12 +1749,7 @@ void ipath_chg_pioavailkernel(struct ipath_devdata *dd, unsigned start,
if (dd->ipath_pioupd_thresh) {
end = 2 * (dd->ipath_piobcnt2k + dd->ipath_piobcnt4k);
- next = find_first_bit(dd->ipath_pioavailkernel, end);
- while (next < end) {
- cnt++;
- next = find_next_bit(dd->ipath_pioavailkernel, end,
- next + 1);
- }
+ cnt = bitmap_weight(dd->ipath_pioavailkernel, end);
}
spin_unlock_irqrestore(&ipath_pioavail_lock, flags);
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 3cb3f47..e596537 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -103,7 +103,7 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
props->device_cap_flags |= IB_DEVICE_UD_AV_PORT_ENFORCE;
if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_IPOIB_CSUM)
props->device_cap_flags |= IB_DEVICE_UD_IP_CSUM;
- if (dev->dev->caps.max_gso_sz)
+ if (dev->dev->caps.max_gso_sz && dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_BLH)
props->device_cap_flags |= IB_DEVICE_UD_TSO;
if (dev->dev->caps.bmme_flags & MLX4_BMME_FLAG_RESERVED_LKEY)
props->device_cap_flags |= IB_DEVICE_LOCAL_DMA_LKEY;
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 219b103..847030c 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -54,7 +54,8 @@ enum {
/*
* Largest possible UD header: send with GRH and immediate data.
*/
- MLX4_IB_UD_HEADER_SIZE = 72
+ MLX4_IB_UD_HEADER_SIZE = 72,
+ MLX4_IB_LSO_HEADER_SPARE = 128,
};
struct mlx4_ib_sqp {
@@ -67,7 +68,8 @@ struct mlx4_ib_sqp {
};
enum {
- MLX4_IB_MIN_SQ_STRIDE = 6
+ MLX4_IB_MIN_SQ_STRIDE = 6,
+ MLX4_IB_CACHE_LINE_SIZE = 64,
};
static const __be32 mlx4_ib_opcode[] = {
@@ -261,7 +263,7 @@ static int send_wqe_overhead(enum ib_qp_type type, u32 flags)
case IB_QPT_UD:
return sizeof (struct mlx4_wqe_ctrl_seg) +
sizeof (struct mlx4_wqe_datagram_seg) +
- ((flags & MLX4_IB_QP_LSO) ? 64 : 0);
+ ((flags & MLX4_IB_QP_LSO) ? MLX4_IB_LSO_HEADER_SPARE : 0);
case IB_QPT_UC:
return sizeof (struct mlx4_wqe_ctrl_seg) +
sizeof (struct mlx4_wqe_raddr_seg);
@@ -897,7 +899,6 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
context->flags = cpu_to_be32((to_mlx4_state(new_state) << 28) |
(to_mlx4_st(ibqp->qp_type) << 16));
- context->flags |= cpu_to_be32(1 << 8); /* DE? */
if (!(attr_mask & IB_QP_PATH_MIG_STATE))
context->flags |= cpu_to_be32(MLX4_QP_PM_MIGRATED << 11);
@@ -1467,16 +1468,12 @@ static void __set_data_seg(struct mlx4_wqe_data_seg *dseg, struct ib_sge *sg)
static int build_lso_seg(struct mlx4_wqe_lso_seg *wqe, struct ib_send_wr *wr,
struct mlx4_ib_qp *qp, unsigned *lso_seg_len,
- __be32 *lso_hdr_sz)
+ __be32 *lso_hdr_sz, __be32 *blh)
{
unsigned halign = ALIGN(sizeof *wqe + wr->wr.ud.hlen, 16);
- /*
- * This is a temporary limitation and will be removed in
- * a forthcoming FW release:
- */
- if (unlikely(halign > 64))
- return -EINVAL;
+ if (unlikely(halign > MLX4_IB_CACHE_LINE_SIZE))
+ *blh = cpu_to_be32(1 << 6);
if (unlikely(!(qp->flags & MLX4_IB_QP_LSO) &&
wr->num_sge > qp->sq.max_gs - (halign >> 4)))
@@ -1522,6 +1519,7 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
__be32 dummy;
__be32 *lso_wqe;
__be32 uninitialized_var(lso_hdr_sz);
+ __be32 blh;
int i;
spin_lock_irqsave(&qp->sq.lock, flags);
@@ -1530,6 +1528,7 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
for (nreq = 0; wr; ++nreq, wr = wr->next) {
lso_wqe = &dummy;
+ blh = 0;
if (mlx4_wq_overflow(&qp->sq, nreq, qp->ibqp.send_cq)) {
err = -ENOMEM;
@@ -1616,7 +1615,7 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
size += sizeof (struct mlx4_wqe_datagram_seg) / 16;
if (wr->opcode == IB_WR_LSO) {
- err = build_lso_seg(wqe, wr, qp, &seglen, &lso_hdr_sz);
+ err = build_lso_seg(wqe, wr, qp, &seglen, &lso_hdr_sz, &blh);
if (unlikely(err)) {
*bad_wr = wr;
goto out;
@@ -1687,7 +1686,7 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
}
ctrl->owner_opcode = mlx4_ib_opcode[wr->opcode] |
- (ind & qp->sq.wqe_cnt ? cpu_to_be32(1 << 31) : 0);
+ (ind & qp->sq.wqe_cnt ? cpu_to_be32(1 << 31) : 0) | blh;
stamp = ind + qp->sq_spare_wqes;
ind += DIV_ROUND_UP(size * 16, 1U << qp->sq.wqe_shift);
generated by cgit v1.1 at 2025-08-21 05:12:57 +0000