Thread (25 messages) 25 messages, 4 authors, 2023-06-15

Re: [PATCH v3 1/6] PCI: hv: Fix a race condition bug in hv_pci_query_relations()

From: Lorenzo Pieralisi <lpieralisi@kernel.org>
Date: 2023-05-25 08:14:54
Also in: linux-hyperv, linux-pci, linux-rdma, lkml, stable

On Wed, Apr 19, 2023 at 07:40:32PM -0700, Dexuan Cui wrote:
Fix the longstanding race between hv_pci_query_relations() and
survey_child_resources() by flushing the workqueue before we exit from
hv_pci_query_relations().
"Fix the longstanding race" is vague. Please describe the race
succinctly at least to give an idea of what the problem is.
quoted hunk ↗ jump to hunk
Fixes: 4daace0d8ce8 ("PCI: hv: Add paravirtual PCI front-end for Microsoft Hyper-V VMs")
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Reviewed-by: Michael Kelley <redacted>
Cc: stable@vger.kernel.org
---

v2:
  Removed the "debug code".
  No change to the patch body.
  Added Cc:stable

v3:
  Added Michael's Reviewed-by.

 drivers/pci/controller/pci-hyperv.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
diff --git a/drivers/pci/controller/pci-hyperv.c b/drivers/pci/controller/pci-hyperv.c
index f33370b756283..b82c7cde19e66 100644
--- a/drivers/pci/controller/pci-hyperv.c
+++ b/drivers/pci/controller/pci-hyperv.c
@@ -3308,6 +3308,19 @@ static int hv_pci_query_relations(struct hv_device *hdev)
 	if (!ret)
 		ret = wait_for_response(hdev, &comp);
 
+	/*
+	 * In the case of fast device addition/removal, it's possible that
+	 * vmbus_sendpacket() or wait_for_response() returns -ENODEV but we
+	 * already got a PCI_BUS_RELATIONS* message from the host and the
+	 * channel callback already scheduled a work to hbus->wq, which can be
+	 * running survey_child_resources() -> complete(&hbus->survey_event),
+	 * even after hv_pci_query_relations() exits and the stack variable
+	 * 'comp' is no longer valid. This can cause a strange hang issue
"A strange hang" sounds like we don't understand what's happening, it
does not feel like it is a solid understanding of the issue.

I would remove it - given that you already explain that comp is no
longer valid - that is already a bug that needs fixing.

Acked-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
+	 * or sometimes a page fault. Flush hbus->wq before we exit from
+	 * hv_pci_query_relations() to avoid the issues.
+	 */
+	flush_workqueue(hbus->wq);
+
 	return ret;
 }
 
-- 
2.25.1
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help