From cd80ec6f81db89d109187a673470c04af4c09a63 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Thu, 15 May 2008 15:28:55 -0700 Subject: IB/ipath: Fix printk format for ipath_sdma_status Commit f018c7e1 ("IB/ipath: Change ipath_devdata.ipath_sdma_status to be unsigned long") changed ipath_sdma_status to be unsigned long, but left a few debug messages that printed it out with a %016llx format, which generates the warnings drivers/infiniband/hw/ipath/ipath_sdma.c:348: warning: format '%016llx' expects type 'long long unsigned int', but argument 3 has type 'long unsigned int' drivers/infiniband/hw/ipath/ipath_sdma.c:618: warning: format '%016llx' expects type 'long long unsigned int', but argument 3 has type 'long unsigned int' Fix this by changing the format used to print out the value to %08lx (8 hex digits are now sufficient, because the highest bit used is 31). Warnings reported by Randy Dunlap . Signed-off-by: Roland Dreier --- drivers/infiniband/hw/ipath/ipath_sdma.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers') diff --git a/drivers/infiniband/hw/ipath/ipath_sdma.c b/drivers/infiniband/hw/ipath/ipath_sdma.c index 3697449c1ba..0a8c1b8091a 100644 --- a/drivers/infiniband/hw/ipath/ipath_sdma.c +++ b/drivers/infiniband/hw/ipath/ipath_sdma.c @@ -345,7 +345,7 @@ resched: * state change */ if (jiffies > dd->ipath_sdma_abort_jiffies) { - ipath_dbg("looping with status 0x%016llx\n", + ipath_dbg("looping with status 0x%08lx\n", dd->ipath_sdma_status); dd->ipath_sdma_abort_jiffies = jiffies + 5 * HZ; } @@ -615,7 +615,7 @@ void ipath_restart_sdma(struct ipath_devdata *dd) } spin_unlock_irqrestore(&dd->ipath_sdma_lock, flags); if (!needed) { - ipath_dbg("invalid attempt to restart SDMA, status 0x%016llx\n", + ipath_dbg("invalid attempt to restart SDMA, status 0x%08lx\n", dd->ipath_sdma_status); goto bail; } -- cgit v1.2.3-18-g5258 From df3f0da8db6b5e7e8f0585221c8b1cd8ff806d35 Mon Sep 17 00:00:00 2001 From: Ralph Campbell Date: Thu, 15 May 2008 16:37:25 -0700 Subject: IB/ipath: Fix UC receive completion opcode for RDMA WRITE with immediate When I fixed the RC receive completion opcode in 2bfc8e9e ("IB/ipath: Return the correct opcode for RDMA WRITE with immediate"), I forgot to fix UC, which had the same problem for RDMA write with immediate returning the wrong opcode. Signed-off-by: Ralph Campbell Signed-off-by: Roland Dreier --- drivers/infiniband/hw/ipath/ipath_uc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers') diff --git a/drivers/infiniband/hw/ipath/ipath_uc.c b/drivers/infiniband/hw/ipath/ipath_uc.c index 7fd18e83390..0596ec16fcb 100644 --- a/drivers/infiniband/hw/ipath/ipath_uc.c +++ b/drivers/infiniband/hw/ipath/ipath_uc.c @@ -407,12 +407,11 @@ void ipath_uc_rcv(struct ipath_ibdev *dev, struct ipath_ib_header *hdr, dev->n_pkt_drops++; goto done; } - /* XXX Need to free SGEs */ + wc.opcode = IB_WC_RECV; last_imm: ipath_copy_sge(&qp->r_sge, data, tlen); wc.wr_id = qp->r_wr_id; wc.status = IB_WC_SUCCESS; - wc.opcode = IB_WC_RECV; wc.qp = &qp->ibqp; wc.src_qp = qp->remote_qpn; wc.slid = qp->remote_ah_attr.dlid; @@ -514,6 +513,7 @@ void ipath_uc_rcv(struct ipath_ibdev *dev, struct ipath_ib_header *hdr, goto done; } wc.byte_len = qp->r_len; + wc.opcode = IB_WC_RECV_RDMA_WITH_IMM; goto last_imm; case OP(RDMA_WRITE_LAST): -- cgit v1.2.3-18-g5258 From a3d8e1591dc90d359d444c759dfda2c6fc605251 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Fri, 16 May 2008 14:28:30 -0700 Subject: IB/mlx4: Fix uninitialized-var warning in mlx4_ib_post_send() drivers/infiniband/hw/mlx4/qp.c: In function 'mlx4_ib_post_send': drivers/infiniband/hw/mlx4/qp.c:1460: warning: 'seglen' may be used uninitialized in this function This is the dopey gcc-doesn't-know-that-foo(&var)-writes-to-var problem. Signed-off-by: Andrew Morton Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mlx4/qp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index 8e02ecfec18..cec030e118d 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -1457,7 +1457,7 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, unsigned ind; int uninitialized_var(stamp); int uninitialized_var(size); - unsigned seglen; + unsigned uninitialized_var(seglen); int i; spin_lock_irqsave(&qp->sq.lock, flags); -- cgit v1.2.3-18-g5258 From 21609ae3efa42f4118ce741f7e55d66d716cb17c Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Fri, 16 May 2008 14:58:40 -0700 Subject: RDMA/cxgb3: Fix uninitialized variable warning in iwch_post_send() drivers/infiniband/hw/cxgb3/iwch_qp.c: In function 'iwch_post_send': drivers/infiniband/hw/cxgb3/iwch_qp.c:232: warning: 't3_wr_flit_cnt' may be used uninitialized in this function This is what akpm describes as "the dopey gcc-doesn't-know-that-foo(&var)-writes-to-var problem." Signed-off-by: Roland Dreier Acked-by: Steve Wise --- drivers/infiniband/hw/cxgb3/iwch_qp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/infiniband/hw/cxgb3/iwch_qp.c b/drivers/infiniband/hw/cxgb3/iwch_qp.c index 79dbe5beae5..99261379922 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_qp.c +++ b/drivers/infiniband/hw/cxgb3/iwch_qp.c @@ -229,7 +229,7 @@ int iwch_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, struct ib_send_wr **bad_wr) { int err = 0; - u8 t3_wr_flit_cnt; + u8 uninitialized_var(t3_wr_flit_cnt); enum t3_wr_opcode t3_wr_opcode = 0; enum t3_wr_flags t3_wr_flags; struct iwch_qp *qhp; -- cgit v1.2.3-18-g5258 From 12103dca52e79e23afe2fbcaf3d9e7fc9ceb6b18 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Fri, 16 May 2008 14:58:44 -0700 Subject: IB/mthca: Fix max_sge value returned by query_device The mthca driver returns the maximum number of scatter/gather entries returned by the firmware as the max_sge value when device properties are queried. However, the firmware also reports a limit on the maximum descriptor size allowed, and because mthca takes into account the worst case send request overhead when checking whether to allow a QP to be created, the largest number of scatter/gather entries that can be used with mthca may be limited by the maximum descriptor size rather than just by the actual s/g entry limit. This means that applications cannot actually create QPs with max_send_sge equal to the limit returned by ib_query_device(). Fix this by checking if the maximum descriptor size imposes a lower limit and if so returning that lower limit. Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_main.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/infiniband/hw/mthca/mthca_main.c b/drivers/infiniband/hw/mthca/mthca_main.c index 9ebadd6e0cf..200cf13fc9b 100644 --- a/drivers/infiniband/hw/mthca/mthca_main.c +++ b/drivers/infiniband/hw/mthca/mthca_main.c @@ -45,6 +45,7 @@ #include "mthca_cmd.h" #include "mthca_profile.h" #include "mthca_memfree.h" +#include "mthca_wqe.h" MODULE_AUTHOR("Roland Dreier"); MODULE_DESCRIPTION("Mellanox InfiniBand HCA low-level driver"); @@ -200,7 +201,18 @@ static int mthca_dev_lim(struct mthca_dev *mdev, struct mthca_dev_lim *dev_lim) mdev->limits.gid_table_len = dev_lim->max_gids; mdev->limits.pkey_table_len = dev_lim->max_pkeys; mdev->limits.local_ca_ack_delay = dev_lim->local_ca_ack_delay; - mdev->limits.max_sg = dev_lim->max_sg; + /* + * Need to allow for worst case send WQE overhead and check + * whether max_desc_sz imposes a lower limit than max_sg; UD + * send has the biggest overhead. + */ + mdev->limits.max_sg = min_t(int, dev_lim->max_sg, + (dev_lim->max_desc_sz - + sizeof (struct mthca_next_seg) - + (mthca_is_memfree(mdev) ? + sizeof (struct mthca_arbel_ud_seg) : + sizeof (struct mthca_tavor_ud_seg))) / + sizeof (struct mthca_data_seg)); mdev->limits.max_wqes = dev_lim->max_qp_sz; mdev->limits.max_qp_init_rdma = dev_lim->max_requester_per_qp; mdev->limits.reserved_qps = dev_lim->reserved_qps; -- cgit v1.2.3-18-g5258 From cd155c1c7c9e64df6afb5504d292fef7cb783a4f Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Tue, 20 May 2008 14:00:02 -0700 Subject: IB/mlx4: Fix creation of kernel QP with max number of send s/g entries When creating a kernel QP where the consumer asked for a send queue with lots of scatter/gater entries, set_kernel_sq_size() incorrectly returned an error if the send queue stride is larger than the hardware's maximum send work request descriptor size. This is not a problem; the only issue is to make sure that the actual descriptors used do not overflow the maximum descriptor size, so check this instead. Clamp the returned max_send_sge value to be no bigger than what query_device returns for the max_sge to avoid confusing hapless users, even if the hardware is capable of handling a few more s/g entries. This bug caused NFS/RDMA mounts to fail when the server adapter used the mlx4 driver. Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mlx4/qp.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) (limited to 'drivers') diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index cec030e118d..a80df22deae 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -333,6 +333,9 @@ static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap, cap->max_inline_data + sizeof (struct mlx4_wqe_inline_seg)) + send_wqe_overhead(type, qp->flags); + if (s > dev->dev->caps.max_sq_desc_sz) + return -EINVAL; + /* * Hermon supports shrinking WQEs, such that a single work * request can include multiple units of 1 << wqe_shift. This @@ -372,9 +375,6 @@ static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap, qp->sq.wqe_shift = ilog2(roundup_pow_of_two(s)); for (;;) { - if (1 << qp->sq.wqe_shift > dev->dev->caps.max_sq_desc_sz) - return -EINVAL; - qp->sq_max_wqes_per_wr = DIV_ROUND_UP(s, 1U << qp->sq.wqe_shift); /* @@ -395,7 +395,8 @@ static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap, ++qp->sq.wqe_shift; } - qp->sq.max_gs = ((qp->sq_max_wqes_per_wr << qp->sq.wqe_shift) - + qp->sq.max_gs = (min(dev->dev->caps.max_sq_desc_sz, + (qp->sq_max_wqes_per_wr << qp->sq.wqe_shift)) - send_wqe_overhead(type, qp->flags)) / sizeof (struct mlx4_wqe_data_seg); @@ -411,7 +412,9 @@ static int set_kernel_sq_size(struct mlx4_ib_dev *dev, struct ib_qp_cap *cap, cap->max_send_wr = qp->sq.max_post = (qp->sq.wqe_cnt - qp->sq_spare_wqes) / qp->sq_max_wqes_per_wr; - cap->max_send_sge = qp->sq.max_gs; + cap->max_send_sge = min(qp->sq.max_gs, + min(dev->dev->caps.max_sq_sg, + dev->dev->caps.max_rq_sg)); /* We don't support inline sends for kernel QPs (yet) */ cap->max_inline_data = 0; -- cgit v1.2.3-18-g5258 From e1d50dce5af77cb6d33555af70e2b8748dd84009 Mon Sep 17 00:00:00 2001 From: Jack Morgenstein Date: Tue, 20 May 2008 15:41:09 -0700 Subject: IPoIB: Test for NULL broadcast object in ipiob_mcast_join_finish() We saw a kernel oops in our regression testing when a multicast "join finish" occurred just after the interface was -- this is . The test randomly causes the HCA physical port to go down then up. The cause of this is that ipoib_mcast_join_finish() processing happen just after ipoib_mcast_dev_flush() was invoked (in which case the broadcast pointer is NULL). This patch tests for and handles the case where priv->broadcast is NULL. Cc: Signed-off-by: Jack Morgenstein Signed-off-by: Roland Dreier --- drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'drivers') diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index d00a2c174ae..3f663fb852c 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -194,7 +194,13 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, /* Set the cached Q_Key before we attach if it's the broadcast group */ if (!memcmp(mcast->mcmember.mgid.raw, priv->dev->broadcast + 4, sizeof (union ib_gid))) { + spin_lock_irq(&priv->lock); + if (!priv->broadcast) { + spin_unlock_irq(&priv->lock); + return -EAGAIN; + } priv->qkey = be32_to_cpu(priv->broadcast->mcmember.qkey); + spin_unlock_irq(&priv->lock); priv->tx_wr.wr.ud.remote_qkey = priv->qkey; } -- cgit v1.2.3-18-g5258 From 5a4f2b675210718aceb4abf41617a3af31bba718 Mon Sep 17 00:00:00 2001 From: Dave Olson Date: Fri, 23 May 2008 10:52:59 -0700 Subject: IB/mad: Fix kernel crash when .process_mad() returns SUCCESS|CONSUMED If a low-level driver returns IB_MAD_RESULT_SUCCESS | IB_MAD_RESULT_CONSUMED, handle_outgoing_dr_smp() doesn't clean up properly. The fix is to kfree the local data and break, rather than falling through. This was observed with the ipath driver, but could happen with any driver. This fixes . Signed-off-by: Dave Olson Signed-off-by: Roland Dreier --- drivers/infiniband/core/mad.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'drivers') diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index fbe16d5250a..1adf2efd3cb 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -747,7 +747,9 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv, break; case IB_MAD_RESULT_SUCCESS | IB_MAD_RESULT_CONSUMED: kmem_cache_free(ib_mad_cache, mad_priv); - break; + kfree(local); + ret = 1; + goto out; case IB_MAD_RESULT_SUCCESS: /* Treat like an incoming receive MAD */ port_priv = ib_get_mad_port(mad_agent_priv->agent.device, -- cgit v1.2.3-18-g5258