netdev - Re: [PATCH] igc: Avoid possible deadlock during suspend/resume

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <d1867ea068e57702bdd953668a3c992f3e205ded.camel@gmx.de>
Date:   Tue, 14 Dec 2021 07:39:52 +0100
From:   Stefan Dietrich <roots@....de>
To:     Vinicius Costa Gomes <vinicius.gomes@...el.com>,
        Thorsten Leemhuis <regressions@...mhuis.info>
Cc:     kuba@...nel.org, greg@...ah.com, netdev@...r.kernel.org,
        intel-wired-lan@...ts.osuosl.org, regressions@...ts.linux.dev
Subject: Re: [PATCH] igc: Avoid possible deadlock during suspend/resume

Hi Vinicius,

thanks for the info and you work on that issue.


Stefan

On Mon, 2021-12-13 at 10:32 -0800, Vinicius Costa Gomes wrote:
> Hi Stefan,
>
> Stefan Dietrich <roots@....de> writes:
>
> > Hi Vinicius,
> >
> > thanks a lot - that patch fixed it! Both "normal" shutdown as well
> > as
> > ifdown/ifup are working without issues now if CONFIG_PCIE_PTM is
> > enabled in the kernel config.
>
> Great!
>
> This patch is mostly to give us time to investigate, this seems to be
> an
> issue related to that specific i225 model. I have to track one down
> and
> perhaps talk to the hardware folks and see what I am doing wrong.
>
> > I've done a DSL download/upload speed comparison against my current
> > 5.14.0-19.2 and did not see any performance differences outside
> > margin
> > of error. I currently have no other Linux machine I could use for
> > iperf
> > but I will report if I encounter any issues.
> >
>
> I wasn't expecting any changes in performance, I was more asking if
> you
> had some use case for PCIe PTM, and something stopped working. It
> seems
> that the answer is no. That's good.
>
> > As I am not familiar with the kernel development procedure: can you
> > give a rough estimate when we may expect this patch in the stable
> > branch?
>
> I will write a useful commit message, take another closer look to see
> if
> I am still missing something and propose the patch upstream. From
> there
> until it's accepted in a stable tree, I guess it could take a few
> days,
> a week, perhaps.
>
> >
> > Thanks again,
> > Stefan
> >
> >
> >
> > On Fri, 2021-12-10 at 16:41 -0800, Vinicius Costa Gomes wrote:
> > > Hi Stefan,
> > >
> > > Stefan Dietrich <roots@....de> writes:
> > >
> > > > Agreed and thanks for the pointers; please see the log files
> > > > and
> > > > .config attached as requested.
> > > >
> > >
> > > Thanks for the logs.
> > >
> > > Very interesting that the initialization of the device is fine,
> > > so
> > > it's
> > > something that happens later.
> > >
> > > Can you test the attached patch?
> > >
> > > If the patch works, I would also be interested if you notice any
> > > loss
> > > of
> > > functionality with your NIC. (I wouldn't think so, as far as I
> > > know,
> > > i225-V models have PTM support but don't have any PTP support).
> > >
> > > > Cheers,
> > > > Stefan
> > > >
> > > >
> > > > On Fri, 2021-12-10 at 15:01 +0100, Thorsten Leemhuis wrote:
> > > > > On 10.12.21 14:45, Stefan Dietrich wrote:
> > > > > > thanks for keeping an eye on the issue. I've sent the files
> > > > > > in
> > > > > > private
> > > > > > because I did not want to spam the mailing lists with them.
> > > > > > Please
> > > > > > let
> > > > > > me know if this is the correct procedure.
> > >
> > > Cheers,
>
> Cheers,