Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
Kernel 5.10.37 doesnt boot
View unanswered posts
View posts from last 24 hours

Goto page 1, 2  Next  
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware
View previous topic :: View next topic  
Author Message
pietinger
Moderator
Moderator


Joined: 17 Oct 2006
Posts: 5372
Location: Bavaria

PostPosted: Sat May 15, 2021 9:27 am    Post subject: Kernel 5.10.37 doesnt boot Reply with quote

Today I tried 5.10.37 coming from 5.10.36 (with "make oldconfig") on a Gigabyte Z270X-K5. It compiled with no errors or warnings.

Before I had an Update of the Intel Microcode also (I have intel-ucode/06-5e-03). So it is also possible this is the reason.

Anybody else with this problem ?


P.S.: It is a stub kernel which will be loaded from UEFI directly. No errors or screen output; only black screen and BIOS trying to load again.
Back to top
View user's profile Send private message
CaptainBlood
Advocate
Advocate


Joined: 24 Jan 2010
Posts: 3999

PostPosted: Sat May 15, 2021 10:48 am    Post subject: Reply with quote

Fails to boot here, as non UEFI.
µcode updated here, not causing any issue with 5.12.4.

Didn't fallback to previous firmware yet.

Thks 4 ur attention, interest & support.
_________________
USE="-* ..." in /etc/portage/make.conf here, i.e. a countermeasure to portage implicit braces, belt & diaper paradigm
LT: "I've been doing a passable imitation of the Fontana di Trevi, except my medium is mucus. Sooo much mucus. "
Back to top
View user's profile Send private message
CaptainBlood
Advocate
Advocate


Joined: 24 Jan 2010
Posts: 3999

PostPosted: Sat May 15, 2021 11:00 am    Post subject: Reply with quote

firmware fallback didn't help.
Thks 4 spotting it out.
Code:
gcc-11 installed versions:  5.10.37(5.10.37)^bs(12:37:13 15/05/2021)(experimental -build -symlink)
Thks 4 ur attention, interest & support.
_________________
USE="-* ..." in /etc/portage/make.conf here, i.e. a countermeasure to portage implicit braces, belt & diaper paradigm
LT: "I've been doing a passable imitation of the Fontana di Trevi, except my medium is mucus. Sooo much mucus. "


Last edited by CaptainBlood on Sat May 15, 2021 11:21 am; edited 2 times in total
Back to top
View user's profile Send private message
pietinger
Moderator
Moderator


Joined: 17 Oct 2006
Posts: 5372
Location: Bavaria

PostPosted: Sat May 15, 2021 11:15 am    Post subject: Reply with quote

CaptainBlood wrote:
µcode updated here, not causing any issue with 5.12.4.

Thank you very much for your feedback. So, I dont need to test the new microcode with 5.10.36

I looked into https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/log/?h=linux-5.10.y
Such a huge amount of patches in one minor version I only saw a long time before ...
Back to top
View user's profile Send private message
web.user
n00b
n00b


Joined: 14 May 2021
Posts: 17

PostPosted: Sat May 15, 2021 11:16 am    Post subject: Reply with quote

See https://forums.gentoo.org/viewtopic-t-1135239.html
That person has same problem as you if I understand correctly
Back to top
View user's profile Send private message
Goverp
Advocate
Advocate


Joined: 07 Mar 2007
Posts: 2202

PostPosted: Sat May 15, 2021 5:08 pm    Post subject: Reply with quote

This will be of neither use nor comfort, but that's why I settled on installing grub as my boot manager, and having a suitable grub.conf to select new, current and old kernels.
The trouble with the EFI stub is that it tells BIOS (or rather EFI) that your linux kernel is a bootloader, but that's really a lie - it's just a kernel that implements the EFI bootloader ABI. It's great until it breaks :-( .
_________________
Greybeard
Back to top
View user's profile Send private message
Tony0945
Watchman
Watchman


Joined: 25 Jul 2006
Posts: 5127
Location: Illinois, USA

PostPosted: Sat May 15, 2021 6:13 pm    Post subject: Reply with quote

Easy to switch kernels with sys-boot/refind using keyboard
Back to top
View user's profile Send private message
CaptainBlood
Advocate
Advocate


Joined: 24 Jan 2010
Posts: 3999

PostPosted: Sat May 15, 2021 7:16 pm    Post subject: Reply with quote

It would be nice if buddies with easy kernel switch capabilities could effectivly test 5.10.37 bootability.
Thks 4 ur attention, interest & support.
_________________
USE="-* ..." in /etc/portage/make.conf here, i.e. a countermeasure to portage implicit braces, belt & diaper paradigm
LT: "I've been doing a passable imitation of the Fontana di Trevi, except my medium is mucus. Sooo much mucus. "
Back to top
View user's profile Send private message
Tony0945
Watchman
Watchman


Joined: 25 Jul 2006
Posts: 5127
Location: Illinois, USA

PostPosted: Sat May 15, 2021 7:23 pm    Post subject: Reply with quote

CaptainBlood wrote:
It would be nice if buddies with easy kernel switch capabilities could effectivly test 5.10.37 bootability.

I would but I don't want to unmask gcc and glibc and I'm running all AMD systems anyway so I can't test 5the microcode.
I think the microcode is the problem. Sometimes you get a bad BIOS. Sometimes you get bad microcode.
Back to top
View user's profile Send private message
CaptainBlood
Advocate
Advocate


Joined: 24 Jan 2010
Posts: 3999

PostPosted: Sat May 15, 2021 8:03 pm    Post subject: Reply with quote

Tony0945 wrote:
I think the microcode is the problem.
Both former and updated intel µcode failed to boot 5.10.37.

5.10.36 boots fine here.

gcc-11 isn't required. OP doesn't say anything about it.

Thks 4 ur attention, interest & support.
_________________
USE="-* ..." in /etc/portage/make.conf here, i.e. a countermeasure to portage implicit braces, belt & diaper paradigm
LT: "I've been doing a passable imitation of the Fontana di Trevi, except my medium is mucus. Sooo much mucus. "
Back to top
View user's profile Send private message
Tony0945
Watchman
Watchman


Joined: 25 Jul 2006
Posts: 5127
Location: Illinois, USA

PostPosted: Sat May 15, 2021 8:34 pm    Post subject: Reply with quote

Updating portage to get 5.10.37
First I have 218 updates to build. i notice that they are amending ebuilds without revbump. Example: guchar unconditionally requires vala even if -vala flag is set. if a flag dopes nothing, why have it? and why not revbump the ebuild?

When I'm all through, I can report if 5.10.37 boots on ryzen. I still don't think that it proves anything for intel, but it's a data point.

Still encourage OP to install refind instead of stub kernel. For one thing when you build a new kernel it becomes default without any extra commands.
Back to top
View user's profile Send private message
pietinger
Moderator
Moderator


Joined: 17 Oct 2006
Posts: 5372
Location: Bavaria

PostPosted: Sat May 15, 2021 10:01 pm    Post subject: Reply with quote

Goverp wrote:
This will be of neither use nor comfort, but that's why I settled on installing grub as my boot manager, and having a suitable grub.conf to select new, current and old kernels.

Maybe I wasnt exact when writing my first post. I have no problem with going back to 5.10.36 because I have two other entries in my UEFI: One is my old grub2, the other is a stub kernel with a command line doing no IMA (ima_appraise=fix). So I have no problems when a new kernel has problems - I can go back immediately.

My real intention was only to see if it is a very special problem only I have (because I have e.g.: SecureBoot and IMA), or if there are more users concerned - and if yes it should/could be a little warning. So please let me apologize if I was not concrete.
Back to top
View user's profile Send private message
Tony0945
Watchman
Watchman


Joined: 25 Jul 2006
Posts: 5127
Location: Illinois, USA

PostPosted: Sun May 16, 2021 2:26 am    Post subject: Reply with quote

Updated 217 packages (ugh another python change), keyworded gentoo-sources-5.10.27 and built it. Interestingly, the only config change was breaking GCC-NATIVE into two listings, one for Intel and one for AMD (why? gcc can determine amd or intel!).
Will test tomorrow

EDIT"
Wait a minute! You are taling about 5.10.37 but only up to 5.10.32 is in the tree! As of this afternoon.

EDIT2:
Forget it. i was looking at the wrong machine. on the right machine:
Code:
MSI ~ # ls -ltr /boot/vml*
-rw-r--r-- 1 root root 5517968 Mar 25  2020 /boot/vmlinuz-4.17.19-gentoo.old
-rw-r--r-- 1 root root 8364784 Aug 24  2020 /boot/vmlinuz-4.17.19-gentoo
-rw-r--r-- 1 root root 5579328 Aug 26  2020 /boot/vmlinuz-4.14.194-gentoo
-rw-r--r-- 1 root root 5669536 Aug 26  2020 /boot/vmlinuz-4.19.141-gentoo
-rw-r--r-- 1 root root 8610976 Oct 27  2020 /boot/vmlinuz-5.4.72-gentoo.old
-rw-r--r-- 1 root root 8610976 Jan  8 14:04 /boot/vmlinuz-5.4.72-gentoo
-rw-r--r-- 1 root root 8619168 Feb  7 10:39 /boot/vmlinuz-5.4.92-gentoo
-rw-r--r-- 1 root root 8623200 Mar 25 21:35 /boot/vmlinuz-5.4.97-gentoo.old
-rw-r--r-- 1 root root 8627296 Apr  3 11:43 /boot/vmlinuz-5.4.97-gentoo
-rw-r--r-- 1 root root 8727168 Apr 28 18:36 /boot/vmlinuz-5.10.27-gentoo.old
-rw-r--r-- 1 root root 8727104 Apr 30 09:04 /boot/vmlinuz-5.10.27-gentoo
-rw-r--r-- 1 root root 8727328 May 15 21:14 /boot/vmlinuz-5.10.37-gentoo
Not going to test until I have a night's sleep!
Back to top
View user's profile Send private message
Tony0945
Watchman
Watchman


Joined: 25 Jul 2006
Posts: 5127
Location: Illinois, USA

PostPosted: Sun May 16, 2021 3:27 pm    Post subject: Reply with quote

Code:
tony@MSI ~ $ uname -a
Linux MSI 5.10.37-gentoo #1 SMP Sat May 15 21:11:36 CDT 2021 x86_64 AMD Ryzen 7 2700X Eight-Core Processor AuthenticAMD GNU/Linux

No problem at all booting. I still think it's an Intel microcode issue.

Here's my config ... Oops. wgetpaste stopped working!
https://dpaste.com/3KFPJC5LK
Problem was latest "stable" wgetpaste. I re-emerged the previous version and it works. Masking latest stable.
Back to top
View user's profile Send private message
CaptainBlood
Advocate
Advocate


Joined: 24 Jan 2010
Posts: 3999

PostPosted: Sun May 16, 2021 5:07 pm    Post subject: Reply with quote

Finally got it working as in:
Code:
uname -r
5.10.37-gentoo-try
make mrproper
make x86_64_defconfig
make
which I should have tried in the first place.

gcc-11 + latest intel µcode here.

Thks 4 ur attention, interest & support.
_________________
USE="-* ..." in /etc/portage/make.conf here, i.e. a countermeasure to portage implicit braces, belt & diaper paradigm
LT: "I've been doing a passable imitation of the Fontana di Trevi, except my medium is mucus. Sooo much mucus. "
Back to top
View user's profile Send private message
pietinger
Moderator
Moderator


Joined: 17 Oct 2006
Posts: 5372
Location: Bavaria

PostPosted: Sun May 16, 2021 5:26 pm    Post subject: Reply with quote

CaptainBlood wrote:
gcc-11 + latest intel µcode here.

Thanks again. Yes, I forgot to mention gcc; I have gcc 10.2.0-r5 (actual stable).
Back to top
View user's profile Send private message
Tony0945
Watchman
Watchman


Joined: 25 Jul 2006
Posts: 5127
Location: Illinois, USA

PostPosted: Sun May 16, 2021 11:46 pm    Post subject: Reply with quote

Ah? I used gcc-9.3
Reading lots of problems with 11 on the forum.
Back to top
View user's profile Send private message
CaptainBlood
Advocate
Advocate


Joined: 24 Jan 2010
Posts: 3999

PostPosted: Wed May 19, 2021 6:51 am    Post subject: Reply with quote

Code:
diff /usr/src/linux-5.10.36-gentoo-classic/.config  /usr/src/linux-5.10.37-gentoo-classic/.config
3c3
< # Linux/x86 5.10.36-gentoo-classic Kernel Configuration
---
> # Linux/x86 5.10.37-gentoo-classic Kernel Configuration
3389c3389
< CONFIG_INTEL_IOMMU_DEFAULT_ON=y
---
> # CONFIG_INTEL_IOMMU_DEFAULT_ON is not set
seems to be the culprit here.

Context:
Code:
[    0.313747] DMAR: [Firmware Bug]: No firmware reserved region can cover this RMRR [0x0000000088800000-0x000000008affffff], contact BIOS vendor for fixes
[    0.313750] DMAR: [Firmware Bug]: Your BIOS is broken; bad RMRR [0x0000000088800000-0x000000008affffff]
               BIOS vendor: American Megatrends Inc.; Ver: R01-A2
persistently reported here, whatever kernel major version.
Could be the source of the issue when activating INTEL_IOMMU.

Kernel command line activation untested here yet.

Will see if issue keeps raising for forthcoming version.

Thks 4 ur attention, interest & support.
_________________
USE="-* ..." in /etc/portage/make.conf here, i.e. a countermeasure to portage implicit braces, belt & diaper paradigm
LT: "I've been doing a passable imitation of the Fontana di Trevi, except my medium is mucus. Sooo much mucus. "
Back to top
View user's profile Send private message
CaptainBlood
Advocate
Advocate


Joined: 24 Jan 2010
Posts: 3999

PostPosted: Wed May 19, 2021 8:59 am    Post subject: Reply with quote

CaptainBlood wrote:

Will see if issue keeps raising for forthcoming version.
5.10.38 seems to revert some former commits...
Wait & see...
Thks 4 ur attention, interest & support.
_________________
USE="-* ..." in /etc/portage/make.conf here, i.e. a countermeasure to portage implicit braces, belt & diaper paradigm
LT: "I've been doing a passable imitation of the Fontana di Trevi, except my medium is mucus. Sooo much mucus. "
Back to top
View user's profile Send private message
Zucca
Moderator
Moderator


Joined: 14 Jun 2007
Posts: 3930
Location: Rasi, Finland

PostPosted: Wed May 19, 2021 9:15 am    Post subject: Reply with quote

Thanks for the heads up guys. I'll refrain from updating my intel laptop's kernel then.
On my old AMD desktop (FX-8350) 5.10 was unusable (wayland crashes).

I'm starting to dislike 5.10 -series...
_________________
..: Zucca :..

My gentoo installs:
init=/sbin/openrc-init
-systemd -logind -elogind seatd

Quote:
I am NaN! I am a man!
Back to top
View user's profile Send private message
pietinger
Moderator
Moderator


Joined: 17 Oct 2006
Posts: 5372
Location: Bavaria

PostPosted: Wed May 19, 2021 9:23 am    Post subject: Reply with quote

CaptainBlood wrote:
5.10.38 seems to revert some former commits...
Wait & see...

I have just looked into the content of 5.10.38. Not quite as huge as 37 but also very many patches. I will it test as soon as possible.

@ Zucca: For me I had no problems with 5.10 (I had almost every minor version of it) :-)
Back to top
View user's profile Send private message
Zucca
Moderator
Moderator


Joined: 14 Jun 2007
Posts: 3930
Location: Rasi, Finland

PostPosted: Wed May 19, 2021 9:40 am    Post subject: Reply with quote

pietinger wrote:
Zucca: For me I had no problems with 5.10 (I had almost every minor version of it) :-)
What's your CPU and GPU?
I have a hunch which tell me that the problems are because of certain hardware combinations. :(
_________________
..: Zucca :..

My gentoo installs:
init=/sbin/openrc-init
-systemd -logind -elogind seatd

Quote:
I am NaN! I am a man!
Back to top
View user's profile Send private message
pietinger
Moderator
Moderator


Joined: 17 Oct 2006
Posts: 5372
Location: Bavaria

PostPosted: Wed May 19, 2021 9:46 am    Post subject: Reply with quote

Zucca wrote:
What's your CPU and GPU?

I have an Intel i7-6700 and using it also as gpu (no extra graphics card) on a Gigabyte Z270X-K5 with 16 GB RAM. At the moment I am on 5.10.36 as a monolithic signed stub kernel (doing SecureBoot). I am using IMA and AppArmor as security solutions.
Back to top
View user's profile Send private message
Zucca
Moderator
Moderator


Joined: 14 Jun 2007
Posts: 3930
Location: Rasi, Finland

PostPosted: Wed May 19, 2021 10:00 am    Post subject: Reply with quote

Yup. I suspect its either my GPU, R9 Nano, or the CPU FX-8350.
Gotta suffer with that old CPU until chip makers' supply chain starts working again.

As for intel side of things - I'll wait and see how people manage with later updates to 5.10...
_________________
..: Zucca :..

My gentoo installs:
init=/sbin/openrc-init
-systemd -logind -elogind seatd

Quote:
I am NaN! I am a man!
Back to top
View user's profile Send private message
pietinger
Moderator
Moderator


Joined: 17 Oct 2006
Posts: 5372
Location: Bavaria

PostPosted: Thu May 20, 2021 12:21 am    Post subject: Reply with quote

I have just booted Kernel 5.10.38 without any problem.

A diff of "dmesg -t" from 5.10.36 and this one showed me:

1. I have a new microcode; from old "e2" switched now to "ea" (2021-01-25). (This was the microcode update between 5.10.36 and ..37). So, I know the microcde was NOT the problem.
2. I have one new log-entry:
Quote:
IP idents hash table entries: 262144 (order: 9, 2097152 bytes, linear)


My suspicion is there was a problem with IOMMU in 5.10.37 ... because I saw two reverts in 5.10.38 ... but I dont know.
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware All times are GMT
Goto page 1, 2  Next
Page 1 of 2

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum