Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.kernel > #71742 > unrolled thread

Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38

Back to article view | Back to linux.debian.kernel

Contents

  Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-05-23 12:10 +0200
    Processed: Re: Bug#989010: linux-image-5.10.0-7-amd64: No display  (post, grub, boot messages and desktop) after the upgrade to 5.10.38 "Debian Bug Tracking System" <owner@bugs.debian.org> - 2021-05-23 13:40 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 Salvatore Bonaccorso <carnil@debian.org> - 2021-05-23 13:40 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-05-25 10:30 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-05-29 17:00 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-05-29 21:00 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-05-30 07:30 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-05-30 09:40 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-05-30 18:10 +0200
    Processed: Severity raise "Debian Bug Tracking System" <owner@bugs.debian.org> - 2021-05-31 17:40 +0200
    Bug#989010: Severity raise jim_p <pitsiorisj@gmail.com> - 2021-05-31 17:40 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 Salvatore Bonaccorso <carnil@debian.org> - 2021-05-31 21:00 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-06-01 06:40 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-06-02 10:20 +0200
    Processed: Re: linux-image-5.10.0-7-amd64: No display (post,  grub, boot messages and desktop) after the upgrade to 5.10.38 "Debian Bug Tracking System" <owner@bugs.debian.org> - 2021-06-04 09:20 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-06-04 09:20 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 Laurențiu Păncescu <lpancescu@gmail.com> - 2021-06-04 12:00 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-06-05 19:40 +0200
      Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 Ben Hutchings <ben@decadent.org.uk> - 2021-06-05 21:10 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-06-06 07:10 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-06-13 10:30 +0200
      Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 Bernhard Übelacker <bernhardu@mailbox.org> - 2021-06-13 17:00 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-06-14 07:10 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-06-26 09:50 +0200
      Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 Steve McIntyre <93sam@debian.org> - 2021-07-03 03:50 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-07-05 13:00 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@gmail.com> - 2021-07-19 11:40 +0200
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@outlook.com> - 2021-12-29 17:00 +0100
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@outlook.com> - 2021-12-30 06:00 +0100
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@outlook.com> - 2022-02-03 18:10 +0100
      Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 Diederik de Haas <didi.debian@cknow.org> - 2022-02-03 19:10 +0100
    Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 jim_p <pitsiorisj@outlook.com> - 2022-04-30 14:40 +0200
      Bug#989010: linux-image-5.10.0-7-amd64: No display (post, grub, boot messages and desktop) after the upgrade to 5.10.38 Bernhard Übelacker <bernhardu@mailbox.org> - 2022-05-17 11:30 +0200

Page 2 of 2 — ← Prev page 1 [2]

#72174

Source: linux
Followup-For: Bug #989010
X-Debbugs-Cc: pitsiorisj@gmail.com

One more week passed with no issues. That's how things are when a kernel works
as it should and the gpu driver is flawless. Stability that matches the one of
a really stable distro or the one of windows 7.

As I expected, there were no replies on my issue, even now that you have full
logs of what was going on. Maybe it was because I did not raise the severity to
critical, so as to "force" you look at it.
Or, one simple word, incompetence. You just want to blame others'
configurations for your faulty builds and hide your irresponsibilities behind
words like "tainted", admit it.

As for the cards, now that they were tested by a different person, the ati is
totally dead and the nvidia one fails to build on 8 out of 10 times. So it is
not my hardware that makes them "seem dead". They ARE dead, thanks to that
garbage kernel update you made.

I read today that debian 11 will be released by the end of July, so I hope
there will be at least one more kernel upgrade until then. I will check if any
of the kernel parameters have changed by then and I will consider testing it
again.

[toc] | [prev] | [next] | [standalone]

#72176

Hello Jim,
I am not involved in packaging, but came to this report by chance.


> The attached file contains all the changes you devs have made in the kernel
> configs from 5.10.28 (-6 package) to 5.10.38/.40 (-7 package). It was made with
> meld.
> ~10 kernel parameters have changed and led to this mess, so I assume it would
> be trivial for you to find the faulty one.

These parameters are not all that changed - there are around 1400 patches
added upstream to the kernel between v5.10.28 and v5.10.38.

https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/log/?h=v5.10.38


Kind regards,
Bernhard

[toc] | [prev] | [next] | [standalone]

#72183

Source: linux
Followup-For: Bug #989010
X-Debbugs-Cc: pitsiorisj@gmail.com

Finally, a sane reply that does not blame nvidia's driver.

I have written above that something must have changed, at least regarding intel
hw, because I was missing the coretemp readings on my conky. However, from the
page you mention, only 2 commits refer to intel and I think both of them are
not related to my hardware.

If it is a kernel issue, which means I will come accross it on any distro I
choose, I should report it on kernel's github page... if it had issue reports.
Besides that, I think I am not smart enough to do it. I reported an similar
issue on openssh (bug 912087) some years ago, but I just could not follow the
conversation until it was resolved, because it was way too technical for me.
Shouldn't a maintainer do that for me like they did for openssh? Now they have
all the logs they need.

The other solution, hopefully, is to wait for the freeze to end, debian to move
to 5.12 or newer and me praying it won't have the same issue.

[toc] | [prev] | [next] | [standalone]

#72301

Source: linux
Followup-For: Bug #989010
X-Debbugs-Cc: pitsiorisj@gmail.com

I noticed that 5.10.46 reached unstable yesterday. If it was not for this
issue, I would have already upgraded, like I udually do when a kernel upgrade
comes out.
However, this time I had to check the config file for that -8 upgrade. And the
only difference between -7 and -8 is this version number. The build flags, and
I mean ALL of them, are EXACTLY the same.

So, since I have no other gpus to sacrifice, I won't risk it again. If anyone
with more experience can inform me about the worthy changes, please do. All I
can see in the changelog that might be related is this bit

    - [x86] nmi_watchdog: Fix old-style NMI watchdog regression on old Intel
CPUs

If there is any more info on that change, please show me.

In other news, the gpu that half-survived (gt210) was tested on a last year's
ryzen pc (with windows though) and it boots and works fine. It was tested for
~1 hour with no issues.

[toc] | [prev] | [next] | [standalone]

#72340

I came to this bug report when searching for other things; I've read
through the history here and I feel compelled to comment.

Jim: given your appalling lack of respect you've shown for Debian
developers here, it's hardly surprising that nobody seems to want to
engage with you. Your tone and wording in several of your messages
here are utterly unacceptable. You've used abusive language about
people who are volunteers working to improve Debian in their own
time. I would strongly suggest that you revisit what you've written
and consider apologising.

-- 
Steve McIntyre                                          93sam@debian.org
Debian Community Team                               community@debian.org

[toc] | [prev] | [next] | [standalone]

#72354

Source: linux
Followup-For: Bug #989010
X-Debbugs-Cc: pitsiorisj@gmail.com

There are 2 words that can completely justify my "appaling lack of respect for
debian developers": hardware failure. And it happened not once, but twice.
Plus, the support I received from here is non existant in every troubleshooting
step I made.

Do you see any response now that the requested logs are at anyone's disposal?
Did you see anyone replying on how will I get those logs when I asked before?
Did you see any meaningful response to why did a card that had nothing to do
with the nvidia driver (the ati one) died after a few boots with 5.10.38/40?
No to all I guess. So, do you still call all that "support"? Because I
definitely don't. All I see is an "nvidia is bad" attitude which has no
foundations.

On the other hand, the only good thing all this... condition has caused is for
me to see how bad nouveau is compared to nvidia. I already knew it was bad
before that, but now I have experienced it and I can judge it from every
aspect.

Tomorrow marks 30 full days and ~100 boots without a single issue and that is
because I stayed on 5.10.28. I would try a newer kernel if I could afford a new
card. Unfortunately I don't. I will stick to it as much as I can, but, as I see
it, in the very end, I will probably upgrade straight to 5.13+ or move to arch.
The latter seems more probable btw.

Last but not least, for me, this issue remains of critical severity despite the
downgrade it got. And I am sure if anyone of you faced the same you would have
a similar attitude.
Have a nice day.

p.s. In case anyone is wondering why arch and not some other distro like
opensuse or fedora or anything, the reason is simple. Arch, via aur, can still
provide me with nvidia 340, the other distros can't.

[toc] | [prev] | [next] | [standalone]

#72392

Source: linux
Followup-For: Bug #989010
X-Debbugs-Cc: pitsiorisj@gmail.com

One thing I noticed this weekend.
Time it took for 5.10.38/40 to reach testing from unstable: 7 days, despite
this awful bug here.
Time it took for 5.10.46 to reach testing from unstable: 20+ days and still
counting.

Please do not mention freeze as an excuse. Debian has been in hard freeze since
March, so I doubt that some "full freeze policy" has to do anything with it.
Imho, one more release critical bug like this one would delay the release of
bullseye even more, so you just marked it as important and let it be until it
is "magically" solved one day.
"Quality" testing...

In other news, I asked an experienced linux user to have a look at my logs and
he said there is nothing unusual. He also asked if there is a copy of
/var/log/kern.log from those problematic boots, but unfortunately there is none
right now. If I knew, I would have kept a copy.
Once again, a random user offered me more support that the debian kernel
team... on an issue that has to do with the kernel which is built from that
team.

[toc] | [prev] | [next] | [standalone]

#74054

Source: linux
Followup-For: Bug #989010
X-Debbugs-Cc: pitsiorisj@outlook.com

Time to close this bug.

I had been on 5.10.28 since May and I had zero issues for 6+ months. The nvidia
340 legacy driver was working flawlessly since its -11 update. Today, the -12
update reached testing and although the driver is built with no issues for
5.10.28, lightdm (or the user with startx) can not start xorg.
I instantly assumed that the driver is not compatible after the -12 update with
that "old" 5.10.28 I had, so I upgraded to 5.15.x and xorg started with no
issues. Now I am just waiting for the forementioned issue to reappear.
I hope it won't though, because I am out of gpus and I have no money to buy a
used one.


-- System Information:
Debian Release: bookworm/sid
  APT prefers testing
  APT policy: (990, 'testing'), (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)

Kernel: Linux 5.15.0-2-amd64 (SMP w/2 CPU threads)
Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

[toc] | [prev] | [next] | [standalone]

#74067

Source: linux
Followup-For: Bug #989010
X-Debbugs-Cc: pitsiorisj@outlook.com

And it just happened again, on the very first boot of the day, as if nothing
has changed since May. I was right for not upgrading past 5.10.28 all those
months.


-- System Information:
Debian Release: bookworm/sid
  APT prefers testing
  APT policy: (990, 'testing'), (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)

Kernel: Linux 5.15.0-2-amd64 (SMP w/2 CPU threads)
Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

[toc] | [prev] | [next] | [standalone]

#74299

Source: linux
Followup-For: Bug #989010
X-Debbugs-Cc: pitsiorisj@outlook.com

Preface: In case I did not mention it above, I had a suspicion that the issue
might be related to cpu frequency scaling.

1+ month has passed with 5.15. The blank screen issue appeared a few more times
the following days. A couple of days later, I decided to disable cpufrequtils,
which set my frequency scheduler to schedutil (from ondemand). And the issue
has not happened since then!

Then I checked debian wiki, which says that cpufrequtils is replaced in favor
of cpupower, so I removed it completely and installed cpupower (linux-cpupower
package).
However, cpupower in debian does not come with a systemd service and it does
not have /etc/default/cpupower so as to set a desired governor.
I really need such functionality, because as I have said, I set my governor to
powersave from May to September every year and now I have no way to.


-- System Information:
Debian Release: bookworm/sid
  APT prefers testing
  APT policy: (990, 'testing'), (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)

Kernel: Linux 5.15.0-3-amd64 (SMP w/2 CPU threads)
Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

[toc] | [prev] | [next] | [standalone]

#74300

[Multipart message — attachments visible in raw view] — view raw

On Thursday, 3 February 2022 18:04:45 CET jim_p wrote:
> However, cpupower in debian does not come with a systemd service and it does
> not have /etc/default/cpupower so as to set a desired governor.
> I really need such functionality, because as I have said, I set my governor
> to powersave from May to September every year and now I have no way to.

Yes you can. You can create your own 'boot-up script' or make your own systemd 
unit file and enable that.

# cpupower frequency-info
This tells you the frequencies/governor available and currently selected. The 
governor is likely 'schedutil'.
In the 'available cpufreq governors' you probably don't see 'powersave', but 
you can modprobe that with `modprobe cpufreq-powersave` and after that lsmod 
should tell you it's loaded and `cpupower frequency-info` will now also list 
it as an available governor.

# cpupower frequency-set --governor powersave
This will switch the governor to 'powersave'

So if you put the following lines in the 'Service' section of a systemd unit 
file and install/enable it, you should have the functionality that you want.
You'll have to figure out the rest of the fields yourself, but that shouldn't be 
too hard as there are plenty .service files on your system and the internet.

ExecStart=/sbin/modprobe  cpufreq-powersave
ExecStart=/usr/bin/cpupower frequency-set --governor powersave

HTH,
  Diederik

[toc] | [prev] | [next] | [standalone]

#75184

Source: linux
Followup-For: Bug #989010
X-Debbugs-Cc: pitsiorisj@outlook.com

After 4 relatively quiet months with 5.15 and 5.16, and with May and its hot
temperatures just around the corner, I got the upgrade to 5.17(.3) four days
ago and a (probably) new issue has emerged.

Since the upgrade to 5.17, I get lockups in just minutes, if not seconds, after
xorg starts. The system keeps running, I can connect to it via ssh and reboot
it, but there is no input with the keyboard or the mouse. Once it reboots, it
runs fine until it is completely shut down again. If it is left off for hours,
which is quite common on my end, the issue reappears on the next boot. I then
reboot it via ssh or via reisub and pray it works on next boot.

This line in dmesg however shows that something more critical is happening

NVRM: GPU at 0000:01:00.0 has fallen off the bus.

After a quick search, I found out that it happens when the gpu gets less power
than what it needs. However, my 9500gt just has 50w of tdp and is powered only
by the pcie slot.
Assuming that it is an issue with 5.17, because it did not happen at all with
5.15 or 5.16 those 4 months, and as usual I will troubleshoot it on my own. So,
how can I install 5.16 now that it is gone from the repo?
Yes, when I install a new kernel, I remove the old one to make space because
the size of the amd64 kernels has become huge since a few years ago.


-- System Information:
Debian Release: bookworm/sid
  APT prefers testing
  APT policy: (990, 'testing'), (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)

Kernel: Linux 5.17.0-1-amd64 (SMP w/2 CPU threads; PREEMPT)
Kernel taint flags: TAINT_PROPRIETARY_MODULE, TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

[toc] | [prev] | [next] | [standalone]

#75253

On Sat, 30 Apr 2022 15:26:48 +0300 jim_p <pitsiorisj@outlook.com> wrote:

> ... So,
> how can I install 5.16 now that it is gone from the repo?

Old packages are still available via snapshot.debian.org.

E.g. here are all versions:
https://snapshot.debian.org/package/linux/

I guess these 4 packages are needed at least to download and install manually.

https://snapshot.debian.org/package/linux/5.16.18-1/#linux-image-5.16.0-6-amd64-unsigned_5.16.18-1
https://snapshot.debian.org/package/linux/5.16.18-1/#linux-headers-5.16.0-6-amd64_5.16.18-1
https://snapshot.debian.org/package/linux/5.16.18-1/#linux-headers-5.16.0-6-common_5.16.18-1
https://snapshot.debian.org/package/linux/5.16.18-1/#linux-kbuild-5.16_5.16.18-1

But this might show some package dependency conflicts that
you need to be able to handle too.

Kind regards,
Bernhard

[toc] | [prev] | [standalone]

Page 2 of 2 — ← Prev page 1 [2]

Back to top | Article view | linux.debian.kernel

csiph-web