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


Groups > linux.debian.kernel > #63694 > unrolled thread

Bug#924895: No problems since updating ucodes from upstream git

Started byDiederik de Haas <didi.debian@cknow.org>
First post2019-03-26 16:20 +0100
Last post2019-05-19 01:40 +0200
Articles 3 — 2 participants

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

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Bug#924895: No problems since updating ucodes from upstream git Diederik de Haas <didi.debian@cknow.org> - 2019-03-26 16:20 +0100
    Bug#924895: Updating microcode files didn't help, maybe a kernel problem after all Diederik de Haas <didi.debian@cknow.org> - 2019-05-18 15:20 +0200
      Bug#924895: Updating microcode files didn't help, maybe a kernel problem after all Jose R R <Jose.r.r@metztli-it.com> - 2019-05-19 01:40 +0200

#63694 — Bug#924895: No problems since updating ucodes from upstream git

FromDiederik de Haas <didi.debian@cknow.org>
Date2019-03-26 16:20 +0100
SubjectBug#924895: No problems since updating ucodes from upstream git
Message-ID<xFW1P-51X-1@gated-at.bofh.it>

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

I updated /lib/firmware/amd-ucode/*.bin from upstream git (commit 
7bc246451318b3536d9bfd3c4e46d541a9831b33) and updated initramfs (I'm not sure 
that's the right place) and since then I haven't had the problem anymore.

I realize that we're in freeze mode, but the 18.50 release looks to become 
important for proper functioning hardware pretty quickly.

CC-ing Kernel Team to make sure they're aware too.

[toc] | [next] | [standalone]


#64126 — Bug#924895: Updating microcode files didn't help, maybe a kernel problem after all

FromDiederik de Haas <didi.debian@cknow.org>
Date2019-05-18 15:20 +0200
SubjectBug#924895: Updating microcode files didn't help, maybe a kernel problem after all
Message-ID<xZ7pL-5zj-3@gated-at.bofh.it>
In reply to#63694

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

On dinsdag 26 maart 2019 16:11:47 CEST Diederik de Haas wrote:
> I updated /lib/firmware/amd-ucode/*.bin from upstream git (commit
> 7bc246451318b3536d9bfd3c4e46d541a9831b33) and updated initramfs (I'm not
> sure that's the right place) and since then I haven't had the problem
> anymore.

Correction: it still does happen :-/

I reported this bug hoping that a fix for this would fix an issue I'm having 
since I build this AMD Ryzen system in October 2017, namely freezes and 
spontaneous reboots.

I have replaced basically every component, except CPU and RAM.
I have installed every BIOS update hoping that it would fix my issues, but no.
I've had this issue with every kernel version I tried since Oct 2017 (I run 
Debian Sid).
I _think_ the chipset driver 18.50.x 'issue' I mentioned in OP is now fixed 
with updated firmware-nonfree package (although I got a different issue with 
that, see bug #928631)

Doing another search for "microcode 8001137", I found https://
community.amd.com/thread/225795 and through that I found
https://bugzilla.kernel.org/show_bug.cgi?id=196683

And that sounds a LOT like the issue I'm having for 1.5 years now.

So maybe this bug should be re-assigned to the Linux kernel instead (and 
retitled and set Forwarded to that bugzilla report)?
OTOH it very much sounds like a hardware/CPU bug which may be able to be fixed 
with an(other) microcode update?

I'm now trying to figure out how to compile my own kernel so I can enable 
CONFIG_RCU_NOCB_CPU and see whether that makes a difference


Full system spec:
MB: Asus ROG CROSSHAIR VII HERO (BIOS: 2203)
CPU: AMD Ryzen 7 1800X Eight-Core Processor (family: 0x17, model: 0x1, 
stepping: 0x1)
RAM: 2x F4-3200C16D-32GTZ (making for 4x16GB)
PSU: Corsair AX1600i 1600W
GPU: Asus ROG Strix RX VEGA64 OC edition 8GB
OS: Debian GNU/Linux Sid with currently kernel 4.19.37-3

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


#64129 — Bug#924895: Updating microcode files didn't help, maybe a kernel problem after all

FromJose R R <Jose.r.r@metztli-it.com>
Date2019-05-19 01:40 +0200
SubjectBug#924895: Updating microcode files didn't help, maybe a kernel problem after all
Message-ID<xZh5M-3cn-9@gated-at.bofh.it>
In reply to#64126
Niltze [Hello] Diederik-


On Sat, May 18, 2019 at 6:11 AM Diederik de Haas <didi.debian@cknow.org> wrote:
>
> On dinsdag 26 maart 2019 16:11:47 CEST Diederik de Haas wrote:
> > I updated /lib/firmware/amd-ucode/*.bin from upstream git (commit
> > 7bc246451318b3536d9bfd3c4e46d541a9831b33) and updated initramfs (I'm not
> > sure that's the right place) and since then I haven't had the problem
> > anymore.
>
> Correction: it still does happen :-/
>
> I reported this bug hoping that a fix for this would fix an issue I'm having
> since I build this AMD Ryzen system in October 2017, namely freezes and
> spontaneous reboots.
>
> I have replaced basically every component, except CPU and RAM.
> I have installed every BIOS update hoping that it would fix my issues, but no.
> I've had this issue with every kernel version I tried since Oct 2017 (I run
> Debian Sid).
> I _think_ the chipset driver 18.50.x 'issue' I mentioned in OP is now fixed
> with updated firmware-nonfree package (although I got a different issue with
> that, see bug #928631)
>
> Doing another search for "microcode 8001137", I found https://
> community.amd.com/thread/225795 and through that I found
> https://bugzilla.kernel.org/show_bug.cgi?id=196683
>
> And that sounds a LOT like the issue I'm having for 1.5 years now.
>
> So maybe this bug should be re-assigned to the Linux kernel instead (and
> retitled and set Forwarded to that bugzilla report)?
> OTOH it very much sounds like a hardware/CPU bug which may be able to be fixed
> with an(other) microcode update?
>
> I'm now trying to figure out how to compile my own kernel so I can enable
> CONFIG_RCU_NOCB_CPU and see whether that makes a difference
I will likely be doing a remote installation onto an AMD Ryzen 7 2700X
3.7/4.3GHz in a couple of months.
Accordingly, I took interest in your post.

Although the kernels that I build are Reiser4 -enabled and I have not
created yet the netboot for my project, I have done a preliminary
build of a kernel 4.14.21.
< https://metztli.it/readOnlyEphemeral/ryzen-reiser4.0.2-for_Debian_4.14.21-1.png
>
You are welcomed to give it a spin as it will also support the other
usual filesystems. In your case, you do not have to do anything
special as you are not running reiser4. Hence just
[sudo] dpkg -i linux-image-4.14.0-3+reiser4.0.2-amd64_4.14.21-1+reiser4.0.2_amd64.deb

from (URL link on a single line)
< https://sourceforge.net/projects/metztli-reiser4/files/Ryzen_Reiser4-SFRN-4.0.2_Linux-4.14.21-1_Stretch-bp/
>
might suffice.

If it works and you like the 'Debian way' of building kernel packages,
I will provide you some guidelines -- based on your feedback for your
own customization -- so that you can build your own kernels for Ryzen.

> Full system spec:
> MB: Asus ROG CROSSHAIR VII HERO (BIOS: 2203)
> CPU: AMD Ryzen 7 1800X Eight-Core Processor (family: 0x17, model: 0x1,
> stepping: 0x1)
> RAM: 2x F4-3200C16D-32GTZ (making for 4x16GB)
> PSU: Corsair AX1600i 1600W
> GPU: Asus ROG Strix RX VEGA64 OC edition 8GB
> OS: Debian GNU/Linux Sid with currently kernel 4.19.37-3

Best Professional Regards.

-- 
Jose R R
http://metztli.it
---------------------------------------------------------------------------------------------
Download Metztli Reiser4: Debian Stretch w/ Linux 5.0.10 AMD64
---------------------------------------------------------------------------------------------
feats ZSTD compression https://sf.net/projects/metztli-reiser4/
-------------------------------------------------------------------------------------------
Official current Reiser4 resources: https://reiser4.wiki.kernel.org/

[toc] | [prev] | [standalone]


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


csiph-web