IB/hfi1: Fix an Oops on pci device force remove
This patch fixes an Oops on device unbind, when the device is used
by a PSM user process. PSM processes access device resources which
are freed on device removal. Similar protection exists in uverbs
in ib_core for Verbs clients, but PSM doesn't use ib_uverbs hence
a separate protection is required for PSM clients.
Cc: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Dean Luick <dean.luick@intel.com>
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Tadeusz Struk <tadeusz.struk@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
diff --git a/drivers/infiniband/hw/hfi1/init.c b/drivers/infiniband/hw/hfi1/init.c
index 60db615..e28a6b6 100644
--- a/drivers/infiniband/hw/hfi1/init.c
+++ b/drivers/infiniband/hw/hfi1/init.c
@@ -1538,12 +1538,31 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
return ret;
}
+static void wait_for_clients(struct hfi1_devdata *dd)
+{
+ /*
+ * Remove the device init value and complete the device if there is
+ * no clients or wait for active clients to finish.
+ */
+ if (atomic_dec_and_test(&dd->user_refcount))
+ complete(&dd->user_comp);
+
+ wait_for_completion(&dd->user_comp);
+}
+
static void remove_one(struct pci_dev *pdev)
{
struct hfi1_devdata *dd = pci_get_drvdata(pdev);
/* close debugfs files before ib unregister */
hfi1_dbg_ibdev_exit(&dd->verbs_dev);
+
+ /* remove the /dev hfi1 interface */
+ hfi1_device_remove(dd);
+
+ /* wait for existing user space clients to finish */
+ wait_for_clients(dd);
+
/* unregister from IB core */
hfi1_unregister_ib_device(dd);
@@ -1558,8 +1577,6 @@ static void remove_one(struct pci_dev *pdev)
/* wait until all of our (qsfp) queue_work() calls complete */
flush_workqueue(ib_wq);
- hfi1_device_remove(dd);
-
postinit_cleanup(dd);
}