NB: Have been informed that this will trickle down from upstream into
the 5.10 kernels for Raspberry Pi OS as well.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
The upstream fix "b160c28548bc tcp: do not mess with cloned skbs in
tcp_add_backlog()" has been applied to focal/linux as part of the fixes
for bug 1915195 (Focal update: v5.4.93 upstream stable release). It will
be included for the next focal/linux-raspi, however this bug report will
not be
Mentioned to Raspberry Pi people at:
https://github.com/raspberrypi/linux/issues/3850
If anyone knows of a more current bug report for this issue, let me
know.
** Bug watch added: github.com/raspberrypi/linux/issues #3850
https://github.com/raspberrypi/linux/issues/3850
--
You received this
** Changed in: linux-raspi (Ubuntu Focal)
Status: In Progress => Fix Committed
** Changed in: linux-raspi (Ubuntu Groovy)
Status: In Progress => Fix Committed
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in
** Changed in: linux-raspi (Ubuntu Focal)
Status: Confirmed => In Progress
** Changed in: linux-raspi (Ubuntu Groovy)
Status: Confirmed => In Progress
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
** Description changed:
+ [Impact]
+
+ A mix of incoming TCP traffic, forwarding packets and saturating the
+ link results in lan78xx TX queue timeouts which brings the network
+ interface down.
+
+ [Test Case]
+
+ Se comment #99 below.
+
+ [Where Problems Could Occur]
+
+ Probably lost or
+1 -- I'm stuck on an ancient version of the Pi OS because of this
issue...
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network dies shortly
Nice :)
Does anyone know if this is getting into Raspberry Pi OS?
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network dies shortly after a burst
Upstream fix:
https://lore.kernel.org/netdev/20210119164900.766957-1-eric.duma...@gmail.com/
** Also affects: linux-raspi2 (Ubuntu Groovy)
Importance: Undecided
Status: New
** Also affects: linux-raspi (Ubuntu Groovy)
Importance: Undecided
Status: New
** No longer affects:
I finally managed to reproduce this problem consistently and within 30
secs. Bisecting the kernel points at [1] to be problematic. Whether that
commit introduces a real regression or merely exposes a different
underlying issue is unclear at the moment. I wasn't able to reproduce
the problem with a
Thanks for the added information. I'll take another crack at trying to
reproduce the problem.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network
I can deterministically reproduce this crash with Wireguard on the RPI3
running 20.10 and hence using the lan78xx driver on eth0. A few minutes
of browsing public web sites on a client connected through Wireguard to
the RPI3 crashes the RPI. Changing nothing except for running
Wireguard's
I found a way to trigger the problem fast (within one minute).
On Raspberry Pi 3 B+:
* Enable IPv4 forwarding: sysctl net.ipv4.ip_forward=1
* Enable masquerading: iptables -t nat -A POSTROUTING -o eth0 -j MASQUERADE
* Start a netcat listening process: sudo nc -l -p 1234 > /dev/null
On another
I was finally able to reproduce the problem by pushing data through the
Pi, i.e. sender -> Pi -> wireguard tunnel -> receiver. Takes about 1.5
hrs to trigger the timeout.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in
Well, it took more than 24 hours for my Pi to crash again, but it did
happen. I wasn't able to capture much of what triggered it as i wasn't
home at the time.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
Ok yup this is almost certainly the issue (when the tunnel is pi is
tunnelling traffic for something other than itself)
My tunnel has been up for 12 hours, no crashes at all.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2
Ahhh. I do have 3 tunnels up as well. Good find!
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network dies shortly after a burst of IPv6 tunnel
@juergh I may have (somewhat) found the trigger.
As previously noted, my internal clients have a static route to the remote
wireguard networks via my raspberry pi. This is defined in my internal router
(Unifi)
ie:
client > router > static route (10.241.0.0/24) to pi eth0 IP > pi > tunnel > VPS
RE IPv6, there is some manual config.
I have each vlan/network assigned a /64. On my pi's they get an address
in that /64 (different networks for eth0 and wlan0) but they do have a
static assignment as well (should have included that earlier:
interface eth0
metric 300
static
So exactly what type of payload triggers the issues, that i cant work
out.
I know the traffic profile, vast majority of it is monitoring related as my VPS
is running grafana/influx and my internal clients are sending data over this
tunnel using telegraf.
There is also a fair bit of ICMP etc
@seamoose Thanks. That's pretty much the config that I had. I've now setup an
external VM and created a wireguard tunnel from my Pi 3B+ but it still won't
fail running iperf3. Is there a specific payload that you're transferring?
Also, your IPv6 config, is that just the default or did you do
And the most recent crash:
Oct 22 11:44:51 dns1 kernel: [ 3096.793118] [ cut here ]
Oct 22 11:44:51 dns1 kernel: [ 3096.793183] NETDEV WATCHDOG: eth0 (lan78xx):
transmit queue 0 timed out
Oct 22 11:44:51 dns1 kernel: [ 3096.793282] WARNING: CPU: 2 PID: 0 at
Forgot to include:
Raspberry PI:
ubuntu@dns1:~$ uname -a
Linux dns1 5.4.0-1022-raspi #25-Ubuntu SMP PREEMPT Thu Oct 15 13:31:49 UTC 2020
aarch64 aarch64 aarch64 GNU/Linux
ubuntu@dns1:~$
VPS:
root@m21:~# uname -a
Linux m21 5.4.0-52-generic #57-Ubuntu SMP Thu Oct 15 10:57:00 UTC 2020 x86_64
Absolutely!
--
My Raspberry Pi sits inside my home network, its config:
eth0:
IP: 192.168.200.11
GW: 192.168.200.1
IPv6: enabled (prefix delegation from ISP)
Metric: 300
wlan0:
IP: 192.168.209.11
GW: 192.168.209.1
IPv6: enabled (prefix delegation from ISP)
Metric: 400
wg0: flags=209
@seamoose So you're just setting up a wireguard tunnel and push data
through it to make it fail? That's what I did but in my case between two
Pis on the same network. Works just fine for me. Can you share some of
your relevant (non-standard) configs?
--
You received this bug notification because
I am also on a 3B+
ubuntu@dns1:~$ cat /proc/device-tree/model
Raspberry Pi 3 Model B Plus Rev 1.3
I dont believe you need anything fancy with IPv6, im not tunneling any v6 and
its occurring with me.
I do have IPv6 enabled on the interface used for tuneling.
Hui Wang, im happy to provide a
I only have the 3B+ board - the issue was encountered, and is still
reproducible on that board. OS has always been Ubuntu.
I'll try to reduce the repro steps/conditions when I have time; my setup
is the v6 tunnel +v6 router + ip6tables + bridge reproduces the issue
immediately, so hopefully
Oh, the board I have is also a 3B+, maybe that is the reason I can't
reproduce the issue.
the silkscreen on my board is:
Raspberry Pi 3 Model B+
@ Raspberry Pi 2017.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in
I'm unable to reproduce this issue. I have not fiddled with IPv6 yet due
to my lack of knowledge and (IMO) unclear instructions in previous
comments. I've pushed data through a wireguard tunnel while loading the
Pi (because this sounds similar:
https://github.com/raspberrypi/linux/issues/3782 but
I want to add that in my case it's not related to WLAN. I'm not using
WLAN on the Pi and also totally disabled it in the config ( I was hoping
that helped, but it didn't ). So the issue is related to the ETH0
--
You received this bug notification because you are a member of Kernel
Packages,
@mlx -- dunno. What I do know is reverting to prior firmware solves the
issue for me and at least @Tim.
I had the same issue in May as there was another release. I reverted
and it became stable.
This summer, I tried to sudo apt upgrade and hit the problem again. So
I reverted and am stable.
@Chrostopher: if it's a driver problem, wouldn't your issue with WLAN be
distinct from mine, where I'm seeing the wired Ethernet break?
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
@Juerg:
My config for my network setup, which is an AP via wlan0, not bridged
and station via USB Wifi dongle:
-
#!/bin/bash
# Error management
set -o pipefail
set -o nounset
set -x
apt install -y hostapd dnsmasq
systemctl unmask hostapd
Ok, I'm utterly confused about the reported issue. Simple steps to
reproduce, please?
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network dies
FWIW, I tried the dwc2 overlay again with kernel 5.4.0-1021-raspi - as
before in #21 this changes nothing - network still dies during first
major (>100MB) download, while other USB ports continue working.
On the plus side, the Pi can now actually boot with an external NIC
plugged in. I'm not sure
Have you tried the new overlay (dwc2)? It worked really well for me,
thread and instructions here:
https://github.com/raspberrypi/linux/issues/3843
** Bug watch added: github.com/raspberrypi/linux/issues #3843
https://github.com/raspberrypi/linux/issues/3843
--
You received this bug
I'm having the exact same issue on my RPi 3. I've been using the Pi for
a very long time now and never had an issue. But since last week I
updated the OS to the latest Raspberry OS and the Pi is now running as a
K3s master node. Since then the Pi keeps crashing like this. I've having
a lot of
Bummers.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network dies shortly after a burst of IPv6 tunnel
network load ((lan78xx): transmit queue
Cancel that. As expected, issue still present.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network dies shortly after a burst of IPv6 tunnel
I *may* be jumping the gun here, but having just installed
5.4.0-1019-raspi this morning, things appear to be good!
Did an upgrade ~ 30 minutes ago (usually i find the issue is triggered
after around 10 minutes or so).
Will update if/when it crashes.
--
You received this bug notification
No idea. I didn't march forward on the commits until it breaks, and I
don't know when the bug occurs or I'd scream about it.
What I do know is this old version is stable and I've had demonstrably
no problems with it.
--
You received this bug notification because you are a member of Kernel
@Christopher,
It is devicetree, kernel and bootloader binary update, do you know what
change on source code level fix this issue?
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
4.19.75, https://github.com/Hexxeh/rpi-
firmware/commit/86b202d127ca3d413d0779d870cce2169afaacd7
Ancient. And functional. ;)
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
@scubachristopher what kernel version did that put you on?
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network dies shortly after a burst of IPv6
Also happens on Linux 5.4 on Raspbian 10.
Okay, I'll keep that in mind. At the moment I'm trying a different
overlay (driver) for the USB chip to see if that helps (not sure yet,
though it hasn't helped with the HDD).
--
You received this bug notification because you are a member of Kernel
@hamishbm -- I solved this by reverting to a prior fw release:
sudo rpi-update 86b202d127ca3d413d0779d870cce2169afaacd7
Zero issues since. JFYI
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
I'm having this issue as well, but on Raspbian 10. Symptoms seems the
same, generally occurs with high load. However, I don't seem to need
IPv6 packets to cause this - my network is IPv4 only.
I thought I had a power problem because I am also using an external HDD
(it's also working as a NAS),
NB: On Linux 4.9. I'll post here again if it happens on the new kernel
(5.4).
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network dies shortly
RPI 3B+. My configuration is as follows:
Raspian (buster)
5.4.51-v7+#1333 SMP Mon Aug 10 16:45:19 BST 2020
wlan0 AP, no bridge, fixed IP.
wlan1 USB Wifi adapter
ppp0 Hologram Nova Modem
My application is in the field, so Wifi HotSpots can be unreliable (I
can get an IP but can't connect). So I
@Matt I didn't find any special way to easily select between kernels on
raspberry - installing the linux-{headers,image}--raspi2 would
overwrite whatever was on the FAT partition, making it the kernel loaded
on next boot. To revert to previous kernel, its package would need to be
reinstalled (apt
@mlx I'm running eth0 in bridge, and also using a tunnel broker ipv6
tunnel like the original poster. How would I go about running the
testing kernel?
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
The Eoan Ermine has reached end of life, so this bug will not be fixed
for that release
** Changed in: linux-raspi2 (Ubuntu Eoan)
Status: Confirmed => Won't Fix
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in
This has probably already been noticed, but in an attempt to simply get
my PI working somewhat the way i wanted, i changed my tunnel to run over
wlan0 instead of eth0 (they both connect back to the same network so
doesn't make a huge difference).
Tunnel and interfaces have been up for days, no
Thanks. I hope the installation worked - I only have Focal now, and the
headers package in that folder was asking for a dependency from Eoan
which I don't have.
Still, I managed to boot the system, and unfortunately the behaviour is
the same.
Aug 1 12:04:56 rpi3 kernel: [ 117.985005]
Please test this kernel, I enabled the preempt in the testing kernel.
https://people.canonical.com/~hwang4/rpi-preempt/
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
@Hui, is there a guide/checklist how to test the Pi OS kernel on Ubuntu?
I trie replacing the contents of the FAT partition, and the kernel did
boot, but that's still missing the modules, possibly something else..?
--
You received this bug notification because you are a member of Kernel
Recently we found the CONFIG_PREEMPT is not enabled in the ubuntu
kernel, but it is enabled in the Pi OS's kernel. We plan to enable the
CONFIG_PREEMPT in the ubuntu kernel too. So please wait for that kernel
and then test a kernel with PREEMPT enabled.
Or you could test Pi OS's kernel first.
** Attachment added: "kernel_5.4.0-1015.log"
https://bugs.launchpad.net/ubuntu/+source/linux-raspi2/+bug/1861936/+attachment/5397093/+files/kernel_5.4.0-1015.log
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
I got myself the Ruideng TC66 USB tester and played a bit with it - it
appears that after the network dies on Raspberry, the system's stable
power consumption ends up higher than when it's functional and idling.
Some stray endless loop, perhaps?
** Attachment added: "RPi3B+ power consumption
FWIW, i actually thought this was a bug with the latest raspbian image
and decided to move forward with using 20.04 on my raspberry pi's, turns
out that wasn't the case.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in
Forgot to include:
Jul 28 18:18:25 dns1 kernel: [ 1289.970276] [ cut here ]
Jul 28 18:18:25 dns1 kernel: [ 1289.970354] NETDEV WATCHDOG: eth0 (lan78xx):
transmit queue 0 timed out
Jul 28 18:18:25 dns1 kernel: [ 1289.970463] WARNING: CPU: 1 PID: 0 at
I can confirm I am also seeing a very similar set of circumstances:
-Hardware name: Raspberry Pi 3 Model B Plus Rev 1.3 (DT)
-5.4.0-1015-raspi #15-Ubuntu SMP Fri Jul 10 05:34:24 UTC 2020 aarch64 aarch64
aarch64 GNU/Linux
-The "tunnel" in my case is using Wireguard to a server hosted externally.
@hui.wang - yes, I have seen this a few times in Focal since upgrading
about a month ago. I didn't look at the logs closely, but the message
that appears repeatedly after the interface fails is "eth0: kevent 0 may
have been dropped" I think, and bug #1647397 may be related/identical.
User from
@mlx and Matt,
We have ubuntu-5.4 kernel for raspi (20.04), could you please test if
this issue still happen with 5.4 kernel?
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
I can confirm having the same issue, including on multiple raspberry pi
3s, one with an IPv6 tunnel from tunnelbroker.net, and one without. It
can be triggered by a lot of IPv4 traffic for me as well.
--
You received this bug notification because you are a member of Kernel
Packages, which is
Status changed to 'Confirmed' because the bug affects multiple users.
** Changed in: linux-raspi2 (Ubuntu Focal)
Status: New => Confirmed
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
Still observed with 5.3.0-1023 on eoan. Will be switching to focal,
seeing as it should now boot.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3
** Also affects: linux-raspi2 (Ubuntu Focal)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network dies
Status changed to 'Confirmed' because the bug affects multiple users.
** Changed in: linux-raspi2 (Ubuntu Eoan)
Status: New => Confirmed
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
Hi all,
>From my tests, it doesn't seem like a problem related to ipv6, but with
high load in 3B+ version. In my case its easy to reproduce if I use
ipforward and connection state over 20.04 server pi image, more less
from scratch, static ip, pihole, tun0 and no ipv6.
root@pin:/home/janzun#
Minor update - tried to check what's going on with 5.4 kernel by
listening on the UART - there's no output. Setting standard priority on
the focal-proposed repo, and reinstalling the linux-raspi2 package
(along with its dependencies) from focal-proposed makes no difference.
--
You received this
** Also affects: linux-raspi2 (Ubuntu Eoan)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux-raspi2 in Ubuntu.
https://bugs.launchpad.net/bugs/1861936
Title:
Raspberry Pi 3 network dies
For reference, recovered the installation with:
cd <1st partition, FAT>
for FILE in *.bak ;do mv -i ${FILE%%.bak} ${FILE%%.bak}.broken; mv -i ${FILE}
${FILE%%.bak}; done
So yeah, kernel 5.4.0.1004.4 is unbootable.
--
You received this bug notification because you are a member of Kernel
@juergh - I'm using origin/priority/pinning, so I installed that kernel
with
apt install linux-raspi2=5.4.0.1004.4 linux-image-raspi2=5.4.0.1004.4
linux-headers=5.4.0.1004.4 # (or something like that)
Now my RPi is stuck at the color palette screen. As there is no grub2
AFAICT, what do I need to
Can you try the Focal 5.4 kernel to see if the problem still exists
there as well?
Add the following line to /etc/apt/sources.list and run (I believe) 'apt
install linux-raspi2':
deb http://ports.ubuntu.com/ubuntu-ports focal-proposed universe
--
You received this bug notification because you
** Description changed:
- Raspberry Pi 3 network dies shortly after a burst of SD card IO (writes) and
network load. Happened a few times (every time) when I tried to download a
gparted zip release (~350 MiB) directly to the SD card, and then finally when I
tried to scp the finished download
76 matches
Mail list logo
