Thread (6 messages) 6 messages, 4 authors, 2025-05-30

Re: [Bug] "possible deadlock in rtnl_newlink" in Linux kernel v6.13

From: Joe Damato <hidden>
Date: 2025-05-30 01:12:28
Also in: lkml

On Thu, May 29, 2025 at 05:16:40PM -0700, Jakub Kicinski wrote:
On Thu, 29 May 2025 16:50:17 -0700 Joe Damato wrote:
quoted
@@ -1262,6 +1258,11 @@ static void e1000_remove(struct pci_dev *pdev)
        bool disable_dev;

        e1000_down_and_stop(adapter);
+
+       /* Only kill reset task if adapter is not resetting */
+       if (!test_bit(__E1000_RESETTING, &adapter->flags))
+               cancel_work_sync(&adapter->reset_task);
+
        e1000_release_manageability(adapter);

        unregister_netdev(netdev);
LGTM, FWIW.
For extra points you can move it after the unregister_netdev(),
the existing code cancels the work but netdev may still be up
and kick it back in..
Good idea, thanks.

I'll post something to the list, but I don't have a reproducer to
test. I'm a noob with syzbot, but maybe there's a way to trigger it
to re-run with a posted patch?
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help