system starts, but display seems to deactivate

loqs · 2024-10-27 01:19:24

Are either of https://pkgbuild.com//~gromit/linux-bis … kg.tar.zst and https://pkgbuild.com//~gromit/linux-bis … kg.tar.zst affected by the issue?

sukolyn · 2024-10-27 09:23:04

v6.10.r8045 is bad
6.10-1 is good

seth · 2024-10-27 13:14:51

https://pkgbuild.com//~gromit/linux-bis … kg.tar.zst ?

Edit:

do I `bisect reset' ?

No.

==> ERROR: A failure occurred in build().
Aborting.

Specifically what error?

Last edited by seth (2024-10-27 13:15:09)

sukolyn · 2024-10-27 13:29:53

nothing more was said.

sukolyn · 2024-10-27 13:42:25

linux-mainline-v6.10.r4972.gf0d17d6-1-x86_64 is good

seth · 2024-10-27 13:54:29

https://pkgbuild.com//~gromit/linux-bis … kg.tar.zst
If that's bad, https://pkgbuild.com//~gromit/linux-bis … kg.tar.zst

sukolyn · 2024-10-27 14:30:00

linux-mainline-v6.10.r4977.gf64c1a4-1-x86_64 is good.

seth · 2024-10-27 14:54:20

That's it for the precompiled kernels in that repository then.

nothing more was said.

I'm pretty sure makepkg printed more than just that.

loqs · 2024-10-27 15:44:46

$ git bisect bad
Bisecting: 2238 revisions left to test after this (roughly 11 steps)
[e3950967f6e6b74a3606739ec50ed19f3398c7d8] Merge tag 'soc-dt-6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc

linux-mainline-6.10.r4472.g51835949dda3-1-x86_64.pkg.tar.zst/linux-mainline-headers-6.10.r4472.g51835949dda3-1-x86_64.pkg.tar.zst

Have you tried building linux 6.10 locally and does that have the issue? I may have misread but it appears all the bisection kernels you built were bad?
Edit:
If linux-mainline-6.10.r4472.g51835949dda3 is good:

$ git bisect good
Bisecting: 1119 revisions left to test after this (roughly 10 steps)
[67e9847b0208ae02efbe42dbcda4c1a9a1a853cd] wifi: mt76: mt7925: remove the unused mt7925_mcu_set_chan_info
$ git describe 
v6.10-rc5-1275-g67e9847b0208

linux-mainline-6.10rc5.r1275.g67e9847b0208-1-x86_64.pkg.tar.zst/linux-mainline-headers-6.10rc5.r1275.g67e9847b0208-1-x86_64.pkg.tar.zst

Last edited by loqs (2024-10-27 16:08:47)

sukolyn · 2024-10-27 16:06:22

seth wrote:

That's it for the precompiled kernels in that repository then.
nothing more was said.
I'm pretty sure makepkg printed more than just that.

no, there was things about processes going well.
the "ERROR...Aborting" was the last message.

I tried to reproduce so I can copy the whole output, but, now it's compiling correctly.

sukolyn · 2024-10-27 16:13:35

loqs wrote:

Have you tried building linux 6.10 locally and does that have the issue?

probably no, when kernel updated to 6.11 during system update, and didn't boot as usual, I first downgraded to 6.10.9 (which is the latest kernel that worked)

I may have misread but it appears all the bisection kernels you built were bad?

that's it, all were bad.

loqs · 2024-10-27 21:55:06

You can either rely on gromit's or my kernel builds or whatever issue is causing your system to build bad kernels needs to be resolved.

sukolyn · 2024-10-28 00:42:23

sukolyn wrote:

I tried to reproduce so I can copy the whole output, but, now it's compiling correctly.

oops! of course it compiled : I was not in linux-stable!

now I can reproduce in linux-stable :

~/build/linux-stable $ makepkg --noprepare
==> Making package: linux 6.11.3-1 (Mon Oct 28 01:38:03 2024)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
  -> Updating linux git repo...
remote: Enumerating objects: 1531, done.
remote: Counting objects: 100% (984/984), done.
remote: Compressing objects: 100% (49/49), done.
remote: Total 1531 (delta 954), reused 936 (delta 935), pack-reused 547
Receiving objects: 100% (1531/1531), 1.01 MiB | 21.57 MiB/s, done.
Resolving deltas: 100% (1145/1145), completed with 456 local objects.
From https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux
   c2ee9f594da8..819837584309  master     -> master
 * [new tag]                   v6.12-rc5  -> v6.12-rc5
  -> Found config
==> Validating source files with sha256sums...
    linux ... Passed
    config ... Passed
==> Validating source files with b2sums...
    linux ... Passed
    config ... Passed
==> Verifying source file signatures with gpg...
    linux git repo ... Passed
==> Extracting sources...
  -> Creating working copy of linux git repo...
From /home/sukolyn/build/linux-stable/linux
   c2ee9f594da8..819837584309  master     -> origin/master
 * [new tag]                   v6.12-rc5  -> v6.12-rc5
Reset branch 'makepkg'
==> Starting pkgver()...
==> Removing existing $pkgdir/ directory...
==> Starting build()...
==> ERROR: A failure occurred in build().
    Aborting...

loqs · 2024-10-28 01:03:17

linux-mainline-6.10rc5.r1275.g67e9847b0208-1.src.tar.gz extract the contents the linked PKGBUILD src tarball. Copy or bind mount the linux mainline kernel tree into a directory called linux-mainline in the extracted directory. See if that produces the same error.

Last edited by loqs (2024-10-28 01:03:54)

seth · 2024-10-28 06:35:22

It's because the sources still get updated what likely sullies the git tree?

makepkg --noprepare --holdver

loqs · 2024-10-28 07:11:25

The build is failing because prepare has never been executed once so the 'version' file was never created so `test -s version` fails.
linux-mainline-6.10rc5.r1275.g67e9847b0208-1.src.tar.gz removes use of the 'version' file.
Edit:
Locally I have a copy of the kernel tree tracking the bisection and clean chroot build each bisection so prepare will always run for me.
Edit2:
prepare() still has to be run once or Arch's config will not be copied into the build tree which might explain all the broken builds.

Last edited by loqs (2024-10-28 07:27:08)

sukolyn · 2024-10-28 14:33:57

here we go again:

~ $ rm -rf build/linux-stable/
~ $ mkdir -p build/linux-stable
~ $ cd build/linux-stable
~/build/linux-stable $ wget -qO PKGBUILD http://0x0.st/XUOW.txt
~/build/linux-stable $ wget -qO config https://gitlab.archlinux.org/archlinux/packaging/packages/linux/-/raw/main/config
~/build/linux-stable $ makepkg
==> Création du paquet linux 6.11.3-1 (lun. 28 oct. 2024 11:18:22)
==> Vérification des dépendances pour l’exécution…
==> Vérification des dépendances pour la compilation…
==> Récupération des sources…
  -> Clonage du dépôt linux git…
Clonage dans le dépôt nu '/home/sukolyn/build/linux-stable/linux'
remote: Enumerating objects: 12902401, done.
remote: Counting objects: 100% (55/55), done.
remote: Compressing objects: 100% (49/49), done.
remote: Total 12902401 (delta 25), reused 10 (delta 6), pack-reused 12902346
Réception d'objets: 100% (12902401/12902401), 5.11 Gio | 23.95 Mio/s, fait.
Résolution des deltas: 100% (10311733/10311733), fait.
  -> config trouvé
==> Validation des fichiers source avec sha256sums…
    linux ... Réussite
    config ... ÉCHEC
==> ERREUR : Un ou plusieurs fichiers ne sont pas valides !
~/build/linux-stable $

sorry, I forgot to set LC to have messages in english.

sukolyn · 2024-10-28 14:39:48

loqs wrote:

linux-mainline-6.10rc5.r1275.g67e9847b0208-1.src.tar.gz extract the contents the linked PKGBUILD src tarball. Copy or bind mount the linux mainline kernel tree into a directory called linux-mainline in the extracted directory. See if that produces the same error.

work in progress.

Copy or bind mount the linux mainline kernel tree

just, where do I get this ?

I'm not used to all this.

Last edited by sukolyn (2024-10-28 14:44:39)

seth · 2024-10-28 15:10:42

config ... ÉCHEC

The config hash is wrong again, "makepkg -g"…

sukolyn · 2024-10-28 16:59:55

~/build/linux-stable $ LC_ALL=C makepkg -g
==> Retrieving sources...
  -> Updating linux git repo...
  -> Found config
==> Generating checksums for source files...
sha256sums=('63b8c1628eb14c4a0d65b8b601d928ff673fbaa92f4ecf2d08da8edeabea072a'
            'fb5245563b6f320007b1dac7e2e2b415e8b5901e4a93337491ddb3134a05e46b')
b2sums=('027058e2970c767fd5af5cbf885d61a5e269d9a5a41b3eaf25f58192b618052660bc78a30e9b7932a085aab7d627fc5924fe09be9f1fbf4a531efaf1821b1a39'
        '2454f23cf2ebfad6ca78d5955b982837351f0c77d5583c806462a62186c7701069981184307d06b038170fed5157b9e1f3270da557c7c1e74bed2f04a7a79044')

and then ?

loqs · 2024-10-28 17:21:28

sukolyn wrote:

Copy or bind mount the linux mainline kernel tree
just, where do I get this ?
I'm not used to all this.

If the linux git clone is in ~/build/linux-stable/linux and you have extracted the src.tar.gz to ~/build/linux-mainline then:

mkdir ~/build/linux-mainline/linux-mainline
# mount -o bind ~/build/linux-stable/linux ~/build/linux-mainline/linux-mainline

sukolyn · 2024-10-28 17:35:59

$ wget -O linux-mainline-6.10rc5.r1275.g67e9847b0208-1.src.tar.gz 'https://drive.google.com/file/d/17eCJ4SdfWeye6JYbuNtI9swY1RPLjal1/view?usp=sharing'
--2024-10-28 18:33:04--  https://drive.google.com/file/d/17eCJ4SdfWeye6JYbuNtI9swY1RPLjal1/view?usp=sharing
Loaded CA certificate '/etc/ssl/certs/ca-certificates.crt'
Resolving drive.google.com (drive.google.com)... 2a00:1450:4007:818::200e, 216.58.214.174
Connecting to drive.google.com (drive.google.com)|2a00:1450:4007:818::200e|:443... connected.
HTTP request sent, awaiting response... 404 Not Found
2024-10-28 18:33:04 ERROR 404: Not Found.

loqs · 2024-10-28 17:57:00

I changed the the archive in post #116 to remove the use of the version file Note you need to use a browser to download it as google drive uses jscript and dynamic URLs.
Edit:
You can also get the source using:

# pacman -S devtools # needed for pkgctl
pkgctl repo clone --protocol https linux
mv linux linux-mainline
cd linux-mainline
git checkout 6.11.5.arch1-1
curl -o PKGBUILD.diff  http://0x0.st/X08_.diff
git apply -v PKGBUILD.diff
mkdir linux-mainline
# mount -o bind ~/build/linux-stable/linux linux-mainline
pkgctl build # this will build the package in a clean chroot

Last edited by loqs (2024-10-28 18:09:56)

sukolyn · 2024-10-28 20:28:30

pkgctl repo clone --protocol https linux
==> Cloning linux ...
Cloning into 'linux'...
remote: Enumerating objects: 6395, done.
remote: Counting objects: 100% (24/24), done.
remote: Compressing objects: 100% (24/24), done.
remote: Total 6395 (delta 8), reused 0 (delta 0), pack-reused 6371 (from 1)
Receiving objects: 100% (6395/6395), 1.50 MiB | 12.52 MiB/s, done.
Resolving deltas: 100% (3156/3156), done.
==> Collecting packager identity from makepkg.conf
  -> name    : undefined
  -> email   : undefined
  -> gpg-key : undefined
  -> protocol: https
==> Configuring linux
$ ls -A linux
keys  .git  config  PKGBUILD  .SRCINFO
$ mv linux linux-mainline
$ cd linux-mainline

$ LC_ALL=C git checkout 6.11.5.arch1-1
fatal: not a git repository (or any parent up to mount point /)
Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set).

loqs · 2024-10-28 20:47:05

$ cd linux-mainline
$ LC_ALL=C git checkout 6.11.5.arch1-1

Did you do anything between these two commands? What is the contents of linux-mainline? If there already was a linux-mainline directory the linux directory would be copied into that directory rather than being renamed to linux-mainline.

Last edited by loqs (2024-10-28 20:55:24)

Arch Linux

#101 2024-10-27 01:19:24

Re: system starts, but display seems to deactivate

#102 2024-10-27 09:23:04

Re: system starts, but display seems to deactivate

#103 2024-10-27 13:14:51

Re: system starts, but display seems to deactivate

#104 2024-10-27 13:29:53

Re: system starts, but display seems to deactivate

#105 2024-10-27 13:42:25

Re: system starts, but display seems to deactivate

#106 2024-10-27 13:54:29

Re: system starts, but display seems to deactivate

#107 2024-10-27 14:30:00

Re: system starts, but display seems to deactivate

#108 2024-10-27 14:54:20

Re: system starts, but display seems to deactivate

#109 2024-10-27 15:44:46

Re: system starts, but display seems to deactivate

#110 2024-10-27 16:06:22

Re: system starts, but display seems to deactivate

#111 2024-10-27 16:13:35

Re: system starts, but display seems to deactivate

#112 2024-10-27 21:55:06

Re: system starts, but display seems to deactivate

#113 2024-10-28 00:42:23

Re: system starts, but display seems to deactivate

#114 2024-10-28 01:03:17

Re: system starts, but display seems to deactivate

#115 2024-10-28 06:35:22

Re: system starts, but display seems to deactivate

#116 2024-10-28 07:11:25

Re: system starts, but display seems to deactivate

#117 2024-10-28 14:33:57

Re: system starts, but display seems to deactivate

#118 2024-10-28 14:39:48

Re: system starts, but display seems to deactivate

#119 2024-10-28 15:10:42

Re: system starts, but display seems to deactivate

#120 2024-10-28 16:59:55

Re: system starts, but display seems to deactivate

#121 2024-10-28 17:21:28

Re: system starts, but display seems to deactivate

#122 2024-10-28 17:35:59

Re: system starts, but display seems to deactivate

#123 2024-10-28 17:57:00

Re: system starts, but display seems to deactivate

#124 2024-10-28 20:28:30

Re: system starts, but display seems to deactivate

#125 2024-10-28 20:47:05

Re: system starts, but display seems to deactivate

Board footer