Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.37 as permitted sender) client-ip=23.128.96.37;
Message-ID: <8db5b21fee0f0e547d89b2797b0f4cc40062339b.camel@sipsolutions.net>
Subject: Re: Thinkpad P17 keep hanging in ipv6_addrconf addrconf_verify_work
 / netlink in 6.4 and 6.6
From:   Johannes Berg <johannes@sipsolutions.net>
To:     Marc MERLIN <marc@merlins.org>
Cc:     Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        netdev@vger.kernel.org, linux-wireless@vger.kernel.org,
        ilw@linux.intel.com, "Paul E. McKenney" <paulmck@kernel.org>
Date:   Mon, 04 Dec 2023 19:34:49 +0100
In-Reply-To: <20231204165239.GB9208@merlins.org>
References: <20231202171326.GB24486@merlins.org>
         <20231204004003.GB29484@merlins.org> <20231204073515.GA9208@merlins.org>
         <69717129398d05b18df1c1300bfb41da268c52a0.camel@sipsolutions.net>
         <20231204165239.GB9208@merlins.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.50.1 (3.50.1-1.fc39) 
MIME-Version: 1.0
Precedence: bulk

On Mon, 2023-12-04 at 08:52 -0800, Marc MERLIN wrote:
> >=20
> > So looks like bug in the 'igc' driver wrt. runtime PM locking.
>=20
> Thank you for the analysis Johannes, I appreciate it. I'm at a spot
> where I have to return this laptop very soon if I can't get it working
> (or will get stuck with it, it's a temp one until I get my original P73
> fixed, and from the little I've seen the 3 year newer P17 with I 11th
> gen (10nm) instead of I 9th gen (14nm), there isn't a huge difference
> between the 2, especially as my I9 had 8 cores and the I11 only 6 cores.
> So I may go back to the P73 as soon as I can get it fixed).
>=20
> In the meantime, anything I can try to get around that bug and confirm
> that assumption? I assume that driver isn't something I can turn off
> since everything depends on it, correct?
>=20

Well I assume your ethernet connectivity depends on it :) If you don't
need that, I guess you can just blacklist it.

You could probably remove the runtime PM ops:

--- a/drivers/net/ethernet/intel/igc/igc_main.c
+++ b/drivers/net/ethernet/intel/igc/igc_main.c
@@ -7321,8 +7321,6 @@ static const struct pci_error_handlers igc_err_handle=
r =3D {
 #ifdef CONFIG_PM
 static const struct dev_pm_ops igc_pm_ops =3D {
 	SET_SYSTEM_SLEEP_PM_OPS(igc_suspend, igc_resume)
-	SET_RUNTIME_PM_OPS(igc_runtime_suspend, igc_runtime_resume,
-			   igc_runtime_idle)
 };
 #endif
=20

But I really have no idea about this driver, this is purely from reading
the stack dump and code.

johannes