[prev in list] [next in list] [prev in thread] [next in thread]
List: qubes-devel
Subject: Re: [qubes-devel] Potential ethernet bug
From: Outback Dingo <outbackdingo () gmail ! com>
Date: 2016-02-24 23:20:14
Message-ID: CAKYr3zyj6xS8J5fsA3P8ty0-FR_AfZyMDqse-oVs1eGyuLYBxg () mail ! gmail ! com
[Download RAW message or body]
On Wed, Feb 24, 2016 at 5:10 AM, <pagedw@gmail.com> wrote:
>
>
> On Tuesday, February 9, 2016 at 6:43:54 PM UTC-5, Marek
> Marczykowski-Górecki wrote:
> >
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA256
> >
> > On Sun, Jan 31, 2016 at 06:54:09PM -0800, Eric Shelton wrote:
> > > On Sunday, January 31, 2016 at 9:19:50 PM UTC-5, Marek
> > Marczykowski-Górecki
> > > wrote:
> > > >
> > > > -----BEGIN PGP SIGNED MESSAGE-----
> > > > Hash: SHA256
> > > >
> > > > On Sun, Jan 31, 2016 at 05:25:22PM -0800, pag...@gmail.com
> > <javascript:>
> > > > wrote:
> > > > >
> > > > >
> > > > > On Sunday, January 31, 2016 at 5:52:52 PM UTC-5, Marek
> > > > Marczykowski-Górecki
> > > > > wrote:
> > > > > >
> > > > > > -----BEGIN PGP SIGNED MESSAGE-----
> > > > > > Hash: SHA256
> > > > > >
> > > > > > On Thu, Jan 28, 2016 at 05:33:41PM -0800, pag...@gmail.com
> > > > <javascript:>
> > > > > > wrote:
> > > > > > >
> > > > > > >
> > > > > > > On Thursday, January 28, 2016 at 8:29:49 PM UTC-5, Marek
> > > > > > > Marczykowski-Górecki wrote:
> > > > > > > > On Thu, Jan 28, 2016 at 05:20:25PM -0800, pag...@gmail.com
> > > > > > <javascript:>
> > > > > > > > wrote:
> > > > > > > > > I have a supermicro motherboard with 2 onboard intel I210
> > nics.
> > > > > > > > >
> > > > > > > > > Using 4.1.13-8.pvops.qubes.x86_64 and
> > > > 4.1.13-7.pvops.qubes.x86_64
> > > > > > the
> > > > > > > > > ethernet in sys-net will never get an ip address.
> > > > > > > > > It will constantly be resetting like the bug report here:
> > > > > > > > > https://bugs.centos.org/view.php?id=7034
> > > > > > > > >
> > > > > > > > > However, if I use the 4.1.13-6.pvops.qubes.x86_64 kernel,
> > it
> > > > > > connects
> > > > > > > > > without errors and no NIC resets.
> > > > > > > >
> > > > > > > > Are you sure it is the only difference? There is no change in
> > any
> > > > > > > > network driver between those versions. The only difference is
> > a
> > > > > > security
> > > > > > > > fix[1].
> > > > > > > >
> > > > > > > >
> > > > > >
> > > >
> > https://github.com/QubesOS/qubes-secpack/blob/master/QSBs/qsb-023-2015.txt
> > > > > > > >
> > > > > > >
> > > > > > > Yes, that is the only difference. It just depends on which
> > kernel I
> > > > use
> > > > > > in
> > > > > > > the Qubes grub list. Using the same sys-net vm.
> > > > > >
> > > > > > "grub list" - so it's about kernel in dom0? Just to make it
> > clear, can
> > > > > > you send output of `uname -r` from working version and not
> > working,
> > > > from
> > > > > > *both* dom0 and sys-net? And preferably also kernel messages
> > (both
> > > > dom0
> > > > > > and sys-net, both versions) - so it can be compared.
> > > > > >
> > > > >
> > > > > Here is the information you requested.
> > > > >
> > > > > Working
> > > > >
> > > > > dom0:
> > > > >
> > > > > 4.1.13-6.pvops.qubes.x86_64
> > > > >
> > > > > sys-net:
> > > > >
> > > > > 4.1.13-8.pvops.qubes.x86_64
> > > > >
> > > > >
> > > > > Non Working
> > > > >
> > > > > dom0:
> > > > >
> > > > > 4.1.13-8.pvops.qubes.x86_64
> > > > >
> > > > > sys-net:
> > > > >
> > > > > 4.1.13-8.pvops.qubes.x86_64
> > > >
> > > >
> > > > Ok, this is really interesting. I see those errors in sys-net:
> > > > IPv6: ADDRCONF(NETDEV_UP): enp0s0: link is not ready
> > > > xen:events: Failed to obtain physical IRQ 109
> > > > xen:events: Failed to obtain physical IRQ 110
> > > > xen:events: Failed to obtain physical IRQ 111
> > > > xen:events: Failed to obtain physical IRQ 112
> > > > xen:events: Failed to obtain physical IRQ 113
> > > > IPv6: ADDRCONF(NETDEV_UP): enp0s0: link is not ready
> > > > IPv6: ADDRCONF(NETDEV_UP): enp0s1: link is not ready
> > > > xen:events: Failed to obtain physical IRQ 115
> > > > xen:events: Failed to obtain physical IRQ 116
> > > > xen:events: Failed to obtain physical IRQ 117
> > > > xen:events: Failed to obtain physical IRQ 118
> > > > xen:events: Failed to obtain physical IRQ 119
> > > > IPv6: ADDRCONF(NETDEV_UP): enp0s1: link is not ready
> > > > IPv6: ADDRCONF(NETDEV_UP): vif2.0: link is not ready
> > > >
> > > > Which is very similar to errors got with USB 3.0 controller in this
> > > > thread:
> > > >
> > > >
> > https://groups.google.com/d/msgid/qubes-users/20160129225226.GG1702%40mail-itl
> > > >
> > > > And also here (network adapter):
> > > > http://lists.xen.org/archives/html/xen-devel/2016-01/msg03553.html
> > > > <
> > http://www.google.com/url?q=http%3A%2F%2Flists.xen.org%2Farchives%2Fhtml%2Fxen-devel%2F2016-01%2Fmsg03553.html&sa=D&sntz=1&usg=AFQjCNFqW62V4ZOdaeBDZocS1BJzgRC82w>
> >
> > > > And according to the next message there, downgrade of dom0 kernel
> > also
> > > > fixes the problem.
> > > >
> > > > I have no idea what dom0 kernel version have to do here, because
> > those
> > > > IRQ fails are returned directly from Xen, but apparently there is
> > some
> > > > link between them. Since kernel changes between those versions are
> > not
> > > > that big, it's feasible to just review them and try to understand
> > what
> > > > is going on.
> > > >
> > > > @Eric maybe you've already figured this out? Anyway letting you know
> > > > about this one, since you track Xen related issues (Big Thanks!).
> > > >
> > >
> > > No. But it sounds like there may be a problem in the XSA-155 related
> > > patches (the official Xen ones or the additional ones Qubes did), since
> > > that seems to be the only difference between the working and
> > non-working
> > > kernels discussed above.
> >
> > Just an update also in this thread: there (hopefully) fixed
> > kernel-4.1.13-8.3 package in qubes-dom0-update repo for R3.1.
> > Some details/links:
> > https://github.com/QubesOS/qubes-issues/issues/1734
> >
> > - --
> > Best Regards,
> > Marek Marczykowski-Górecki
> > Invisible Things Lab
> > A: Because it messes up the order in which people normally read text.
> > Q: Why is top-posting such a bad thing?
> > -----BEGIN PGP SIGNATURE-----
> > Version: GnuPG v2
> >
> > iQEcBAEBCAAGBQJWunmxAAoJENuP0xzK19csOAMH/jE59o00DYgdhNCP3HZVg3EF
> > ny8UbG+NKt7FgSB+YT4mVRnVVsFMGK+dOz6Zw8PIgrawsgTCtVlW0eA7XSHQ41+P
> > +/OU/ufZwxXcr2lHLBkwz0zj3lyh+fFUXVVlTRId1SPma92kBnG4/pGTZ8kDv1Ou
> > udTtf5mPvfNhSq4mDFhja1xhR9baKHa823Kakje5pJzRccTtscVDn722TxIYP34Y
> > RysgALW3RpBdRUEla8Yyw9AU1zNf26LRE+RD4++XAO1IjcRavESkGY8t61IxOd6I
> > oedldR+KjRVoWM1rcHYVjKn7vJg2X8tvR4U7BIgoDrSxSK/YEDTVUEqgdn/q8II=
> > =QW/s
> > -----END PGP SIGNATURE-----
> >
>
>
> Just wanted to let you know that updating dom0, I am now successfully
> running 4.1.13-9 kernel and the ethernet is working properly. ]
>
Boy if it were only that easy to fix the pci passthru issue im still
having, to the best i can tell the issue is the vm kernel crashed on r8169
nic, with shared pci bus.
again, it works fine on fedora 23 / kernel Linux localhost.localdomain
4.3.5-300.fc23.x86_64 #1 SMP Mon Feb 1 03:18:41 UTC 2016 x86_64 x86_64
x86_64 GNU/Linux
with XEN xen_version : 4.5.2
Im close to pretty convinced it is the kernel in the vm itself... always
segfaulting, preventing networking from coming alive.
> --
> You received this message because you are subscribed to the Google Groups
> "qubes-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to qubes-devel+unsubscribe@googlegroups.com.
> To post to this group, send email to qubes-devel@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/qubes-devel/64d02b96-d386-4a10-a8da-80a64ac89d84%40googlegroups.com
> <https://groups.google.com/d/msgid/qubes-devel/64d02b96-d386-4a10-a8da-80a64ac89d84%40googlegroups.com?utm_medium=email&utm_source=footer>
>
> .
>
> For more options, visit https://groups.google.com/d/optout.
>
--
You received this message because you are subscribed to the Google Groups \
"qubes-devel" group. To unsubscribe from this group and stop receiving emails from \
it, send an email to qubes-devel+unsubscribe@googlegroups.com. To post to this group, \
send email to qubes-devel@googlegroups.com. To view this discussion on the web visit \
https://groups.google.com/d/msgid/qubes-devel/CAKYr3zyj6xS8J5fsA3P8ty0-FR_AfZyMDqse-oVs1eGyuLYBxg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.
[Attachment #3 (text/html)]
<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Feb \
24, 2016 at 5:10 AM, <span dir="ltr"><<a href="mailto:pagedw@gmail.com" \
target="_blank">pagedw@gmail.com</a>></span> wrote:<br><blockquote \
class="gmail_quote" style="margin:0px 0px 0px \
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><div \
dir="ltr"><div><div class="h5"><br><br>On Tuesday, February 9, 2016 at 6:43:54 PM \
UTC-5, Marek Marczykowski-Górecki wrote:<blockquote class="gmail_quote" \
style="margin:0px 0px 0px \
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">-----BEGIN \
PGP SIGNED MESSAGE----- <br>Hash: SHA256
<br>
<br>On Sun, Jan 31, 2016 at 06:54:09PM -0800, Eric Shelton wrote:
<br>> On Sunday, January 31, 2016 at 9:19:50 PM UTC-5, Marek Marczykowski-Górecki \
<br>> wrote:
<br>> >
<br>> > -----BEGIN PGP SIGNED MESSAGE-----
<br>> > Hash: SHA256
<br>> >
<br>> > On Sun, Jan 31, 2016 at 05:25:22PM -0800, <a>pag...@gmail.com</a> \
<javascript:> <br>> > wrote:
<br>> > >
<br>> > >
<br>> > > On Sunday, January 31, 2016 at 5:52:52 PM UTC-5, Marek
<br>> > Marczykowski-Górecki
<br>> > > wrote:
<br>> > > >
<br>> > > > -----BEGIN PGP SIGNED MESSAGE-----
<br>> > > > Hash: SHA256
<br>> > > >
<br>> > > > On Thu, Jan 28, 2016 at 05:33:41PM -0800, \
<a>pag...@gmail.com</a> <br>> > <javascript:>
<br>> > > > wrote:
<br>> > > > >
<br>> > > > >
<br>> > > > > On Thursday, January 28, 2016 at 8:29:49 PM UTC-5, Marek \
<br>> > > > > Marczykowski-Górecki wrote:
<br>> > > > > > On Thu, Jan 28, 2016 at 05:20:25PM -0800, \
<a>pag...@gmail.com</a> <br>> > > > <javascript:>
<br>> > > > > > wrote:
<br>> > > > > > > I have a supermicro motherboard with 2 onboard \
intel I210 nics. <br>> > > > > > >
<br>> > > > > > > Using 4.1.13-8.pvops.qubes.x86_64 and
<br>> > 4.1.13-7.pvops.qubes.x86_64
<br>> > > > the
<br>> > > > > > > ethernet in sys-net will never get an ip \
address. <br>> > > > > > > It will constantly be resetting like \
the bug report here: <br>> > > > > > > <a \
href="https://bugs.centos.org/view.php?id=7034" rel="nofollow" \
target="_blank">https://bugs.centos.org/view.php?id=7034</a> <br>> > > > \
> > > <br>> > > > > > > However, if I use the \
4.1.13-6.pvops.qubes.x86_64 kernel, it <br>> > > > connects
<br>> > > > > > > without errors and no NIC resets.
<br>> > > > > >
<br>> > > > > > Are you sure it is the only difference? There is no \
change in any <br>> > > > > > network driver between those \
versions. The only difference is a <br>> > > > security
<br>> > > > > > fix[1].
<br>> > > > > >
<br>> > > > > >
<br>> > > >
<br>> > <a href="https://github.com/QubesOS/qubes-secpack/blob/master/QSBs/qsb-023-2015.txt" \
rel="nofollow" target="_blank">https://github.com/QubesOS/qubes-secpack/blob/master/QSBs/qsb-023-2015.txt</a> \
<br>> > > > > >
<br>> > > > >
<br>> > > > > Yes, that is the only difference. It just depends on \
which kernel I <br>> > use
<br>> > > > in
<br>> > > > > the Qubes grub list. Using the same sys-net vm.
<br>> > > >
<br>> > > > "grub list" - so it's about kernel in dom0? \
Just to make it clear, can <br>> > > > you send output of `uname -r` \
from working version and not working, <br>> > from
<br>> > > > *both* dom0 and sys-net? And preferably also kernel messages \
(both <br>> > dom0
<br>> > > > and sys-net, both versions) - so it can be compared.
<br>> > > >
<br>> > >
<br>> > > Here is the information you requested.
<br>> > >
<br>> > > Working
<br>> > >
<br>> > > dom0:
<br>> > >
<br>> > > 4.1.13-6.pvops.qubes.x86_64
<br>> > >
<br>> > > sys-net:
<br>> > >
<br>> > > 4.1.13-8.pvops.qubes.x86_64
<br>> > >
<br>> > >
<br>> > > Non Working
<br>> > >
<br>> > > dom0:
<br>> > >
<br>> > > 4.1.13-8.pvops.qubes.x86_64
<br>> > >
<br>> > > sys-net:
<br>> > >
<br>> > > 4.1.13-8.pvops.qubes.x86_64
<br>> >
<br>> >
<br>> > Ok, this is really interesting. I see those errors in sys-net:
<br>> > IPv6: ADDRCONF(NETDEV_UP): enp0s0: link is not ready
<br>> > xen:events: Failed to obtain physical IRQ 109
<br>> > xen:events: Failed to obtain physical IRQ 110
<br>> > xen:events: Failed to obtain physical IRQ 111
<br>> > xen:events: Failed to obtain physical IRQ 112
<br>> > xen:events: Failed to obtain physical IRQ 113
<br>> > IPv6: ADDRCONF(NETDEV_UP): enp0s0: link is not ready
<br>> > IPv6: ADDRCONF(NETDEV_UP): enp0s1: link is not ready
<br>> > xen:events: Failed to obtain physical IRQ 115
<br>> > xen:events: Failed to obtain physical IRQ 116
<br>> > xen:events: Failed to obtain physical IRQ 117
<br>> > xen:events: Failed to obtain physical IRQ 118
<br>> > xen:events: Failed to obtain physical IRQ 119
<br>> > IPv6: ADDRCONF(NETDEV_UP): enp0s1: link is not ready
<br>> > IPv6: ADDRCONF(NETDEV_UP): vif2.0: link is not ready
<br>> >
<br>> > Which is very similar to errors got with USB 3.0 controller in this
<br>> > thread:
<br>> >
<br>> > <a href="https://groups.google.com/d/msgid/qubes-users/20160129225226.GG1702%40mail-itl" \
rel="nofollow" target="_blank">https://groups.google.com/d/msgid/qubes-users/20160129225226.GG1702%40mail-itl</a> \
<br>> >
<br>> > And also here (network adapter):
<br>> > <a href="http://lists.xen.org/archives/html/xen-devel/2016-01/msg03553.html" \
rel="nofollow" target="_blank">http://lists.xen.org/archives/html/xen-devel/2016-01/msg03553.html</a> \
<br>> > <<a \
href="http://www.google.com/url?q=http%3A%2F%2Flists.xen.org%2Farchives%2Fhtml%2Fxen-d \
evel%2F2016-01%2Fmsg03553.html&sa=D&sntz=1&usg=AFQjCNFqW62V4ZOdaeBDZocS1BJzgRC82w" \
rel="nofollow" target="_blank">http://www.google.com/url?q=http%3A%2F%2Flists.xen.org% \
2Farchives%2Fhtml%2Fxen-devel%2F2016-01%2Fmsg03553.html&sa=D&sntz=1&usg=AFQjCNFqW62V4ZOdaeBDZocS1BJzgRC82w</a>> \
<br>> > And according to the next message there, downgrade of dom0 kernel also \
<br>> > fixes the problem.
<br>> >
<br>> > I have no idea what dom0 kernel version have to do here, because those
<br>> > IRQ fails are returned directly from Xen, but apparently there is some
<br>> > link between them. Since kernel changes between those versions are not
<br>> > that big, it's feasible to just review them and try to understand \
what <br>> > is going on.
<br>> >
<br>> > @Eric maybe you've already figured this out? Anyway letting you \
know <br>> > about this one, since you track Xen related issues (Big Thanks!). \
<br>> >
<br>>
<br>> No. But it sounds like there may be a problem in the XSA-155 related
<br>> patches (the official Xen ones or the additional ones Qubes did), since
<br>> that seems to be the only difference between the working and non-working
<br>> kernels discussed above.
<br>
<br>Just an update also in this thread: there (hopefully) fixed
<br>kernel-4.1.13-8.3 package in qubes-dom0-update repo for R3.1.
<br>Some details/links:
<br><a href="https://github.com/QubesOS/qubes-issues/issues/1734" rel="nofollow" \
target="_blank">https://github.com/QubesOS/qubes-issues/issues/1734</a> <br>
<br>- --
<br>Best Regards,
<br>Marek Marczykowski-Górecki
<br>Invisible Things Lab
<br>A: Because it messes up the order in which people normally read text.
<br>Q: Why is top-posting such a bad thing?
<br>-----BEGIN PGP SIGNATURE-----
<br>Version: GnuPG v2
<br>
<br>iQEcBAEBCAAGBQJWunmxAAoJENuP0xzK19csOAMH/jE59o00DYgdhNCP3HZVg3EF
<br>ny8UbG+NKt7FgSB+YT4mVRnVVsFMGK+dOz6Zw8PIgrawsgTCtVlW0eA7XSHQ41+P
<br>+/OU/ufZwxXcr2lHLBkwz0zj3lyh+fFUXVVlTRId1SPma92kBnG4/pGTZ8kDv1Ou
<br>udTtf5mPvfNhSq4mDFhja1xhR9baKHa823Kakje5pJzRccTtscVDn722TxIYP34Y
<br>RysgALW3RpBdRUEla8Yyw9AU1zNf26LRE+RD4++XAO1IjcRavESkGY8t61IxOd6I
<br>oedldR+KjRVoWM1rcHYVjKn7vJg2X8tvR4U7BIgoDrSxSK/YEDTVUEqgdn/q8II=
<br>=QW/s
<br>-----END PGP SIGNATURE-----
<br></blockquote></div></div><div><br><br>Just wanted to let you know that updating \
dom0, I am now successfully running 4.1.13-9 kernel and the ethernet is working \
properly. ]<br></div></div></blockquote><div><br></div><div><br></div><div><br></div><div>Boy \
if it were only that easy to fix the pci passthru issue im still having, to the best \
i can tell the issue is the vm kernel crashed on r8169 nic, with shared pci \
bus.</div><div>again, it works fine on fedora 23 / kernel <span \
style="color:rgb(0,0,0);font-family:monospace">Linux localhost.localdomain \
4.3.5-300.fc23.x86_64 #1 SMP Mon Feb 1 03:18:41 UTC 2016 x86_64 x86_64 x86_64 \
GNU/Linux</span></div><div><span style="color:rgb(0,0,0);font-family:monospace">with \
XEN </span><span style="color:rgb(0,0,0);font-family:monospace">xen_version \
: 4.5.2</span></div><span style="font-family:monospace"> <br></span><div>Im close to \
pretty convinced it is the kernel in the vm itself... always segfaulting, preventing \
networking from coming alive. </div><blockquote class="gmail_quote" \
style="margin:0px 0px 0px \
0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><div \
dir="ltr"><div></div></div><span class="">
<p></p>
-- <br>
You received this message because you are subscribed to the Google Groups \
"qubes-devel" group.<br> To unsubscribe from this group and stop receiving \
emails from it, send an email to <a \
href="mailto:qubes-devel+unsubscribe@googlegroups.com" \
target="_blank">qubes-devel+unsubscribe@googlegroups.com</a>.<br> To post to this \
group, send email to <a href="mailto:qubes-devel@googlegroups.com" \
target="_blank">qubes-devel@googlegroups.com</a>.<br></span> To view this discussion \
on the web visit <a href="https://groups.google.com/d/msgid/qubes-devel/64d02b96-d386- \
4a10-a8da-80a64ac89d84%40googlegroups.com?utm_medium=email&utm_source=footer" \
target="_blank">https://groups.google.com/d/msgid/qubes-devel/64d02b96-d386-4a10-a8da-80a64ac89d84%40googlegroups.com</a>.<div \
class=""><div class="h5"><br> For more options, visit <a \
href="https://groups.google.com/d/optout" \
target="_blank">https://groups.google.com/d/optout</a>.<br> \
</div></div></blockquote></div><br></div></div>
<p></p>
-- <br />
You received this message because you are subscribed to the Google Groups \
"qubes-devel" group.<br /> To unsubscribe from this group and stop \
receiving emails from it, send an email to <a \
href="mailto:qubes-devel+unsubscribe@googlegroups.com">qubes-devel+unsubscribe@googlegroups.com</a>.<br \
/> To post to this group, send email to <a \
href="mailto:qubes-devel@googlegroups.com">qubes-devel@googlegroups.com</a>.<br /> To \
view this discussion on the web visit <a \
href="https://groups.google.com/d/msgid/qubes-devel/CAKYr3zyj6xS8J5fsA3P8ty0-FR_AfZyMD \
qse-oVs1eGyuLYBxg%40mail.gmail.com?utm_medium=email&utm_source=footer">https://groups. \
google.com/d/msgid/qubes-devel/CAKYr3zyj6xS8J5fsA3P8ty0-FR_AfZyMDqse-oVs1eGyuLYBxg%40mail.gmail.com</a>.<br \
/> For more options, visit <a \
href="https://groups.google.com/d/optout">https://groups.google.com/d/optout</a>.<br \
/>
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic