Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/gi…

…t/rdma/rdma

Pull rdma fixes from Doug Ledford:

 - Various driver bug fixes in mlx5, mlx4, bnxt_re and qedr, ranging
   from bugs under load to bad error case handling

 - There in one largish patch fixing the locking in bnxt_re to avoid a
   machine hard lock situation

 - A few core bugs on error paths

 - A patch to reduce stack usage in the new CQ API

 - One mlx5 regression introduced in this merge window

 - There were new syzkaller scripts written for the RDMA subsystem and
   we are fixing issues found by the bot

 - One of the commits (aa0de36 “RDMA/mlx5: Fix integer overflow
   while resizing CQ”) is missing part of the commit log message and one
   of the SOB lines. The original patch was from Leon Romanovsky, and a
   cut-n-paste separator in the commit message confused patchworks which
   then put the end of message separator in the wrong place in the
   downloaded patch, and I didn’t notice in time. The patch made it into
   the official branch, and the only way to fix it in-place was to
   rebase. Given the pain that a rebase causes, and the fact that the
   patch has relevant tags for stable and syzkaller, a revert of the
   munged patch and a reapplication of the original patch with the log
   message intact was done.

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (25 commits)
  RDMA/mlx5: Fix integer overflow while resizing CQ
  Revert "RDMA/mlx5: Fix integer overflow while resizing CQ"
  RDMA/ucma: Check that user doesn't overflow QP state
  RDMA/mlx5: Fix integer overflow while resizing CQ
  RDMA/ucma: Limit possible option size
  IB/core: Fix possible crash to access NULL netdev
  RDMA/bnxt_re: Avoid Hard lockup during error CQE processing
  RDMA/core: Reduce poll batch for direct cq polling
  IB/mlx5: Fix an error code in __mlx5_ib_modify_qp()
  IB/mlx5: When not in dual port RoCE mode, use provided port as native
  IB/mlx4: Include GID type when deleting GIDs from HW table under RoCE
  IB/mlx4: Fix corruption of RoCEv2 IPv4 GIDs
  RDMA/qedr: Fix iWARP write and send with immediate
  RDMA/qedr: Fix kernel panic when running fio over NFSoRDMA
  RDMA/qedr: Fix iWARP connect with port mapper
  RDMA/qedr: Fix ipv6 destination address resolution
  IB/core : Add null pointer check in addr_resolve
  RDMA/bnxt_re: Fix the ib_reg failure cleanup
  RDMA/bnxt_re: Fix incorrect DB offset calculation
  RDMA/bnxt_re: Unconditionly fence non wire memory operations
  ...

drivers/infiniband/core/addr.c

@@ -550,18 +550,13 @@ static int addr_resolve(struct sockaddr *src_in,
 		dst_release(dst);
 	}
 
-	if (ndev->flags & IFF_LOOPBACK) {
-		ret = rdma_translate_ip(dst_in, addr);
-		/*
-		 * Put the loopback device and get the translated
-		 * device instead.
-		 */
+	if (ndev) {
+		if (ndev->flags & IFF_LOOPBACK)
+			ret = rdma_translate_ip(dst_in, addr);
+		else
+			addr->bound_dev_if = ndev->ifindex;
 		dev_put(ndev);
-		ndev = dev_get_by_index(addr->net, addr->bound_dev_if);
-	} else {
-		addr->bound_dev_if = ndev->ifindex;
 	}
-	dev_put(ndev);
 
 	return ret;
 }

drivers/infiniband/core/cq.c

@@ -17,6 +17,7 @@
 
 /* # of WCs to poll for with a single call to ib_poll_cq */
 #define IB_POLL_BATCH			16
+#define IB_POLL_BATCH_DIRECT		8
 
 /* # of WCs to iterate over before yielding */
 #define IB_POLL_BUDGET_IRQ		256
@@ -25,18 +26,18 @@
 #define IB_POLL_FLAGS \
 	(IB_CQ_NEXT_COMP | IB_CQ_REPORT_MISSED_EVENTS)
 
-static int __ib_process_cq(struct ib_cq *cq, int budget, struct ib_wc *poll_wc)
+static int __ib_process_cq(struct ib_cq *cq, int budget, struct ib_wc *wcs,
+			   int batch)
 {
 	int i, n, completed = 0;
-	struct ib_wc *wcs = poll_wc ? : cq->wc;
 
 	/*
 	 * budget might be (-1) if the caller does not
 	 * want to bound this call, thus we need unsigned
 	 * minimum here.
 	 */
-	while ((n = ib_poll_cq(cq, min_t(u32, IB_POLL_BATCH,
-			budget - completed), wcs)) > 0) {
+	while ((n = ib_poll_cq(cq, min_t(u32, batch,
+					 budget - completed), wcs)) > 0) {
 		for (i = 0; i < n; i++) {
 			struct ib_wc *wc = &wcs[i];
 
@@ -48,8 +49,7 @@ static int __ib_process_cq(struct ib_cq *cq, int budget, struct ib_wc *poll_wc)
 
 		completed += n;
 
-		if (n != IB_POLL_BATCH ||
-		    (budget != -1 && completed >= budget))
+		if (n != batch || (budget != -1 && completed >= budget))
 			break;
 	}
 
@@ -72,9 +72,9 @@ static int __ib_process_cq(struct ib_cq *cq, int budget, struct ib_wc *poll_wc)
  */
 int ib_process_cq_direct(struct ib_cq *cq, int budget)
 {
-	struct ib_wc wcs[IB_POLL_BATCH];
+	struct ib_wc wcs[IB_POLL_BATCH_DIRECT];
 
-	return __ib_process_cq(cq, budget, wcs);
+	return __ib_process_cq(cq, budget, wcs, IB_POLL_BATCH_DIRECT);
 }
 EXPORT_SYMBOL(ib_process_cq_direct);
 
@@ -88,7 +88,7 @@ static int ib_poll_handler(struct irq_poll *iop, int budget)
 	struct ib_cq *cq = container_of(iop, struct ib_cq, iop);
 	int completed;
 
-	completed = __ib_process_cq(cq, budget, NULL);
+	completed = __ib_process_cq(cq, budget, cq->wc, IB_POLL_BATCH);
 	if (completed < budget) {
 		irq_poll_complete(&cq->iop);
 		if (ib_req_notify_cq(cq, IB_POLL_FLAGS) > 0)
@@ -108,7 +108,8 @@ static void ib_cq_poll_work(struct work_struct *work)
 	struct ib_cq *cq = container_of(work, struct ib_cq, work);
 	int completed;
 
-	completed = __ib_process_cq(cq, IB_POLL_BUDGET_WORKQUEUE, NULL);
+	completed = __ib_process_cq(cq, IB_POLL_BUDGET_WORKQUEUE, cq->wc,
+				    IB_POLL_BATCH);
 	if (completed >= IB_POLL_BUDGET_WORKQUEUE ||
 	    ib_req_notify_cq(cq, IB_POLL_FLAGS) > 0)
 		queue_work(ib_comp_wq, &cq->work);

drivers/infiniband/core/device.c

@@ -536,14 +536,14 @@ int ib_register_device(struct ib_device *device,
 	ret = device->query_device(device, &device->attrs, &uhw);
 	if (ret) {
 		pr_warn("Couldn't query the device attributes\n");
-		goto cache_cleanup;
+		goto cg_cleanup;
 	}
 
 	ret = ib_device_register_sysfs(device, port_callback);
 	if (ret) {
 		pr_warn("Couldn't register device %s with driver model\n",
 			device->name);
-		goto cache_cleanup;
+		goto cg_cleanup;
 	}
 
 	device->reg_state = IB_DEV_REGISTERED;
@@ -559,6 +559,8 @@ int ib_register_device(struct ib_device *device,
 	mutex_unlock(&device_mutex);
 	return 0;
 
+cg_cleanup:
+	ib_device_unregister_rdmacg(device);
 cache_cleanup:
 	ib_cache_cleanup_one(device);
 	ib_cache_release_one(device);

drivers/infiniband/core/sa_query.c

@@ -1291,10 +1291,9 @@ int ib_init_ah_attr_from_path(struct ib_device *device, u8 port_num,
 
 		resolved_dev = dev_get_by_index(dev_addr.net,
 						dev_addr.bound_dev_if);
-		if (resolved_dev->flags & IFF_LOOPBACK) {
-			dev_put(resolved_dev);
-			resolved_dev = idev;
-			dev_hold(resolved_dev);
+		if (!resolved_dev) {
+			dev_put(idev);
+			return -ENODEV;
 		}
 		ndev = ib_get_ndev_from_path(rec);
 		rcu_read_lock();

drivers/infiniband/core/ucma.c

@@ -1149,6 +1149,9 @@ static ssize_t ucma_init_qp_attr(struct ucma_file *file,
 	if (copy_from_user(&cmd, inbuf, sizeof(cmd)))
 		return -EFAULT;
 
+	if (cmd.qp_state > IB_QPS_ERR)
+		return -EINVAL;
+
 	ctx = ucma_get_ctx(file, cmd.id);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
@@ -1294,6 +1297,9 @@ static ssize_t ucma_set_option(struct ucma_file *file, const char __user *inbuf,
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 
+	if (unlikely(cmd.optval > KMALLOC_MAX_SIZE))
+		return -EINVAL;
+
 	optval = memdup_user((void __user *) (unsigned long) cmd.optval,
 			     cmd.optlen);
 	if (IS_ERR(optval)) {

drivers/infiniband/hw/bnxt_re/ib_verbs.c

@@ -785,7 +785,7 @@ int bnxt_re_query_ah(struct ib_ah *ib_ah, struct rdma_ah_attr *ah_attr)
 	return 0;
 }
 
-static unsigned long bnxt_re_lock_cqs(struct bnxt_re_qp *qp)
+unsigned long bnxt_re_lock_cqs(struct bnxt_re_qp *qp)
 	__acquires(&qp->scq->cq_lock) __acquires(&qp->rcq->cq_lock)
 {
 	unsigned long flags;
@@ -799,8 +799,8 @@ static unsigned long bnxt_re_lock_cqs(struct bnxt_re_qp *qp)
 	return flags;
 }
 
-static void bnxt_re_unlock_cqs(struct bnxt_re_qp *qp,
-			       unsigned long flags)
+void bnxt_re_unlock_cqs(struct bnxt_re_qp *qp,
+			unsigned long flags)
 	__releases(&qp->scq->cq_lock) __releases(&qp->rcq->cq_lock)
 {
 	if (qp->rcq != qp->scq)
@@ -1606,6 +1606,7 @@ int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr,
 	int status;
 	union ib_gid sgid;
 	struct ib_gid_attr sgid_attr;
+	unsigned int flags;
 	u8 nw_type;
 
 	qp->qplib_qp.modify_flags = 0;
@@ -1634,14 +1635,18 @@ int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr,
 			dev_dbg(rdev_to_dev(rdev),
 				"Move QP = %p to flush list\n",
 				qp);
+			flags = bnxt_re_lock_cqs(qp);
 			bnxt_qplib_add_flush_qp(&qp->qplib_qp);
+			bnxt_re_unlock_cqs(qp, flags);
 		}
 		if (!qp->sumem &&
 		    qp->qplib_qp.state == CMDQ_MODIFY_QP_NEW_STATE_RESET) {
 			dev_dbg(rdev_to_dev(rdev),
 				"Move QP = %p out of flush list\n",
 				qp);
+			flags = bnxt_re_lock_cqs(qp);
 			bnxt_qplib_clean_qp(&qp->qplib_qp);
+			bnxt_re_unlock_cqs(qp, flags);
 		}
 	}
 	if (qp_attr_mask & IB_QP_EN_SQD_ASYNC_NOTIFY) {
@@ -2227,10 +2232,13 @@ static int bnxt_re_build_inv_wqe(struct ib_send_wr *wr,
 	wqe->type = BNXT_QPLIB_SWQE_TYPE_LOCAL_INV;
 	wqe->local_inv.inv_l_key = wr->ex.invalidate_rkey;
 
+	/* Need unconditional fence for local invalidate
+	 * opcode to work as expected.
+	 */
+	wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE;
+
 	if (wr->send_flags & IB_SEND_SIGNALED)
 		wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP;
-	if (wr->send_flags & IB_SEND_FENCE)
-		wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE;
 	if (wr->send_flags & IB_SEND_SOLICITED)
 		wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT;
 
@@ -2251,8 +2259,12 @@ static int bnxt_re_build_reg_wqe(struct ib_reg_wr *wr,
 	wqe->frmr.levels = qplib_frpl->hwq.level + 1;
 	wqe->type = BNXT_QPLIB_SWQE_TYPE_REG_MR;
 
-	if (wr->wr.send_flags & IB_SEND_FENCE)
-		wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE;
+	/* Need unconditional fence for reg_mr
+	 * opcode to function as expected.
+	 */
+
+	wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE;
+
 	if (wr->wr.send_flags & IB_SEND_SIGNALED)
 		wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP;
 

drivers/infiniband/hw/bnxt_re/ib_verbs.h

@@ -222,4 +222,7 @@ struct ib_ucontext *bnxt_re_alloc_ucontext(struct ib_device *ibdev,
 					   struct ib_udata *udata);
 int bnxt_re_dealloc_ucontext(struct ib_ucontext *context);
 int bnxt_re_mmap(struct ib_ucontext *context, struct vm_area_struct *vma);
+
+unsigned long bnxt_re_lock_cqs(struct bnxt_re_qp *qp);
+void bnxt_re_unlock_cqs(struct bnxt_re_qp *qp, unsigned long flags);
 #endif /* __BNXT_RE_IB_VERBS_H__ */

drivers/infiniband/hw/bnxt_re/main.c

@@ -730,6 +730,13 @@ static int bnxt_re_handle_qp_async_event(struct creq_qp_event *qp_event,
 					 struct bnxt_re_qp *qp)
 {
 	struct ib_event event;
+	unsigned int flags;
+
+	if (qp->qplib_qp.state == CMDQ_MODIFY_QP_NEW_STATE_ERR) {
+		flags = bnxt_re_lock_cqs(qp);
+		bnxt_qplib_add_flush_qp(&qp->qplib_qp);
+		bnxt_re_unlock_cqs(qp, flags);
+	}
 
 	memset(&event, 0, sizeof(event));
 	if (qp->qplib_qp.srq) {
@@ -1416,9 +1423,12 @@ static void bnxt_re_task(struct work_struct *work)
 	switch (re_work->event) {
 	case NETDEV_REGISTER:
 		rc = bnxt_re_ib_reg(rdev);
-		if (rc)
+		if (rc) {
 			dev_err(rdev_to_dev(rdev),
 				"Failed to register with IB: %#x", rc);
+			bnxt_re_remove_one(rdev);
+			bnxt_re_dev_unreg(rdev);
+		}
 		break;
 	case NETDEV_UP:
 		bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1,