Add the rt linux 4.1.3-rt3 as base
[kvmfornfv.git] / kernel / drivers / infiniband / hw / cxgb4 / ev.c
diff --git a/kernel/drivers/infiniband/hw/cxgb4/ev.c b/kernel/drivers/infiniband/hw/cxgb4/ev.c
new file mode 100644 (file)
index 0000000..bdfac2c
--- /dev/null
@@ -0,0 +1,244 @@
+/*
+ * Copyright (c) 2009-2010 Chelsio, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include <linux/slab.h>
+#include <linux/mman.h>
+#include <net/sock.h>
+
+#include "iw_cxgb4.h"
+
+static void print_tpte(struct c4iw_dev *dev, u32 stag)
+{
+       int ret;
+       struct fw_ri_tpte tpte;
+
+       ret = cxgb4_read_tpte(dev->rdev.lldi.ports[0], stag,
+                             (__be32 *)&tpte);
+       if (ret) {
+               dev_err(&dev->rdev.lldi.pdev->dev,
+                       "%s cxgb4_read_tpte err %d\n", __func__, ret);
+               return;
+       }
+       PDBG("stag idx 0x%x valid %d key 0x%x state %d pdid %d "
+              "perm 0x%x ps %d len 0x%llx va 0x%llx\n",
+              stag & 0xffffff00,
+              FW_RI_TPTE_VALID_G(ntohl(tpte.valid_to_pdid)),
+              FW_RI_TPTE_STAGKEY_G(ntohl(tpte.valid_to_pdid)),
+              FW_RI_TPTE_STAGSTATE_G(ntohl(tpte.valid_to_pdid)),
+              FW_RI_TPTE_PDID_G(ntohl(tpte.valid_to_pdid)),
+              FW_RI_TPTE_PERM_G(ntohl(tpte.locread_to_qpid)),
+              FW_RI_TPTE_PS_G(ntohl(tpte.locread_to_qpid)),
+              ((u64)ntohl(tpte.len_hi) << 32) | ntohl(tpte.len_lo),
+              ((u64)ntohl(tpte.va_hi) << 32) | ntohl(tpte.va_lo_fbo));
+}
+
+static void dump_err_cqe(struct c4iw_dev *dev, struct t4_cqe *err_cqe)
+{
+       __be64 *p = (void *)err_cqe;
+
+       dev_err(&dev->rdev.lldi.pdev->dev,
+               "AE qpid %d opcode %d status 0x%x "
+               "type %d len 0x%x wrid.hi 0x%x wrid.lo 0x%x\n",
+               CQE_QPID(err_cqe), CQE_OPCODE(err_cqe),
+               CQE_STATUS(err_cqe), CQE_TYPE(err_cqe), ntohl(err_cqe->len),
+               CQE_WRID_HI(err_cqe), CQE_WRID_LOW(err_cqe));
+
+       PDBG("%016llx %016llx %016llx %016llx\n",
+            be64_to_cpu(p[0]), be64_to_cpu(p[1]), be64_to_cpu(p[2]),
+            be64_to_cpu(p[3]));
+
+       /*
+        * Ingress WRITE and READ_RESP errors provide
+        * the offending stag, so parse and log it.
+        */
+       if (RQ_TYPE(err_cqe) && (CQE_OPCODE(err_cqe) == FW_RI_RDMA_WRITE ||
+                                CQE_OPCODE(err_cqe) == FW_RI_READ_RESP))
+               print_tpte(dev, CQE_WRID_STAG(err_cqe));
+}
+
+static void post_qp_event(struct c4iw_dev *dev, struct c4iw_cq *chp,
+                         struct c4iw_qp *qhp,
+                         struct t4_cqe *err_cqe,
+                         enum ib_event_type ib_event)
+{
+       struct ib_event event;
+       struct c4iw_qp_attributes attrs;
+       unsigned long flag;
+
+       dump_err_cqe(dev, err_cqe);
+
+       if (qhp->attr.state == C4IW_QP_STATE_RTS) {
+               attrs.next_state = C4IW_QP_STATE_TERMINATE;
+               c4iw_modify_qp(qhp->rhp, qhp, C4IW_QP_ATTR_NEXT_STATE,
+                              &attrs, 0);
+       }
+
+       event.event = ib_event;
+       event.device = chp->ibcq.device;
+       if (ib_event == IB_EVENT_CQ_ERR)
+               event.element.cq = &chp->ibcq;
+       else
+               event.element.qp = &qhp->ibqp;
+       if (qhp->ibqp.event_handler)
+               (*qhp->ibqp.event_handler)(&event, qhp->ibqp.qp_context);
+
+       spin_lock_irqsave(&chp->comp_handler_lock, flag);
+       (*chp->ibcq.comp_handler)(&chp->ibcq, chp->ibcq.cq_context);
+       spin_unlock_irqrestore(&chp->comp_handler_lock, flag);
+}
+
+void c4iw_ev_dispatch(struct c4iw_dev *dev, struct t4_cqe *err_cqe)
+{
+       struct c4iw_cq *chp;
+       struct c4iw_qp *qhp;
+       u32 cqid;
+
+       spin_lock_irq(&dev->lock);
+       qhp = get_qhp(dev, CQE_QPID(err_cqe));
+       if (!qhp) {
+               printk(KERN_ERR MOD "BAD AE qpid 0x%x opcode %d "
+                      "status 0x%x type %d wrid.hi 0x%x wrid.lo 0x%x\n",
+                      CQE_QPID(err_cqe),
+                      CQE_OPCODE(err_cqe), CQE_STATUS(err_cqe),
+                      CQE_TYPE(err_cqe), CQE_WRID_HI(err_cqe),
+                      CQE_WRID_LOW(err_cqe));
+               spin_unlock_irq(&dev->lock);
+               goto out;
+       }
+
+       if (SQ_TYPE(err_cqe))
+               cqid = qhp->attr.scq;
+       else
+               cqid = qhp->attr.rcq;
+       chp = get_chp(dev, cqid);
+       if (!chp) {
+               printk(KERN_ERR MOD "BAD AE cqid 0x%x qpid 0x%x opcode %d "
+                      "status 0x%x type %d wrid.hi 0x%x wrid.lo 0x%x\n",
+                      cqid, CQE_QPID(err_cqe),
+                      CQE_OPCODE(err_cqe), CQE_STATUS(err_cqe),
+                      CQE_TYPE(err_cqe), CQE_WRID_HI(err_cqe),
+                      CQE_WRID_LOW(err_cqe));
+               spin_unlock_irq(&dev->lock);
+               goto out;
+       }
+
+       c4iw_qp_add_ref(&qhp->ibqp);
+       atomic_inc(&chp->refcnt);
+       spin_unlock_irq(&dev->lock);
+
+       /* Bad incoming write */
+       if (RQ_TYPE(err_cqe) &&
+           (CQE_OPCODE(err_cqe) == FW_RI_RDMA_WRITE)) {
+               post_qp_event(dev, chp, qhp, err_cqe, IB_EVENT_QP_REQ_ERR);
+               goto done;
+       }
+
+       switch (CQE_STATUS(err_cqe)) {
+
+       /* Completion Events */
+       case T4_ERR_SUCCESS:
+               printk(KERN_ERR MOD "AE with status 0!\n");
+               break;
+
+       case T4_ERR_STAG:
+       case T4_ERR_PDID:
+       case T4_ERR_QPID:
+       case T4_ERR_ACCESS:
+       case T4_ERR_WRAP:
+       case T4_ERR_BOUND:
+       case T4_ERR_INVALIDATE_SHARED_MR:
+       case T4_ERR_INVALIDATE_MR_WITH_MW_BOUND:
+               post_qp_event(dev, chp, qhp, err_cqe, IB_EVENT_QP_ACCESS_ERR);
+               break;
+
+       /* Device Fatal Errors */
+       case T4_ERR_ECC:
+       case T4_ERR_ECC_PSTAG:
+       case T4_ERR_INTERNAL_ERR:
+               post_qp_event(dev, chp, qhp, err_cqe, IB_EVENT_DEVICE_FATAL);
+               break;
+
+       /* QP Fatal Errors */
+       case T4_ERR_OUT_OF_RQE:
+       case T4_ERR_PBL_ADDR_BOUND:
+       case T4_ERR_CRC:
+       case T4_ERR_MARKER:
+       case T4_ERR_PDU_LEN_ERR:
+       case T4_ERR_DDP_VERSION:
+       case T4_ERR_RDMA_VERSION:
+       case T4_ERR_OPCODE:
+       case T4_ERR_DDP_QUEUE_NUM:
+       case T4_ERR_MSN:
+       case T4_ERR_TBIT:
+       case T4_ERR_MO:
+       case T4_ERR_MSN_GAP:
+       case T4_ERR_MSN_RANGE:
+       case T4_ERR_RQE_ADDR_BOUND:
+       case T4_ERR_IRD_OVERFLOW:
+               post_qp_event(dev, chp, qhp, err_cqe, IB_EVENT_QP_FATAL);
+               break;
+
+       default:
+               printk(KERN_ERR MOD "Unknown T4 status 0x%x QPID 0x%x\n",
+                      CQE_STATUS(err_cqe), qhp->wq.sq.qid);
+               post_qp_event(dev, chp, qhp, err_cqe, IB_EVENT_QP_FATAL);
+               break;
+       }
+done:
+       if (atomic_dec_and_test(&chp->refcnt))
+               wake_up(&chp->wait);
+       c4iw_qp_rem_ref(&qhp->ibqp);
+out:
+       return;
+}
+
+int c4iw_ev_handler(struct c4iw_dev *dev, u32 qid)
+{
+       struct c4iw_cq *chp;
+       unsigned long flag;
+
+       spin_lock_irqsave(&dev->lock, flag);
+       chp = get_chp(dev, qid);
+       if (chp) {
+               atomic_inc(&chp->refcnt);
+               spin_unlock_irqrestore(&dev->lock, flag);
+               t4_clear_cq_armed(&chp->cq);
+               spin_lock_irqsave(&chp->comp_handler_lock, flag);
+               (*chp->ibcq.comp_handler)(&chp->ibcq, chp->ibcq.cq_context);
+               spin_unlock_irqrestore(&chp->comp_handler_lock, flag);
+               if (atomic_dec_and_test(&chp->refcnt))
+                       wake_up(&chp->wait);
+       } else {
+               PDBG("%s unknown cqid 0x%x\n", __func__, qid);
+               spin_unlock_irqrestore(&dev->lock, flag);
+       }
+       return 0;
+}