Closed SergiiDmytruk closed 2 days ago
@SergiiDmytruk which command have you used? Normal build: kas-container build meta-trenchboot/kas-generic-tb.yml
?
From which branch/commit?
kas-container build meta-trenchboot/kas-generic-tb.yml
on master
(827552ed0ca9d47137b7da834c1f144fb148a0c3).
Do you use kas-container
version 3.0.2?
Can you nmap -p 9418 git.sv.gnu.org
from outside and inside kas-container
?
kas-container shell meta-trenchboot/kas-generic-tb.yml
sudo apt update
sudo apt install nmap
nmap -p 9418 git.sv.gnu.org
[~]$ kas-container --version
kas-container 3.0.2
Outside container:
[~]$ nmap -p 9418 git.sv.gnu.org
Starting Nmap 7.95 ( https://nmap.org ) at 2024-07-03 17:42 EEST
Nmap scan report for git.sv.gnu.org (209.51.188.168)
Host is up (0.14s latency).
Other addresses for git.sv.gnu.org (not scanned): 2001:470:142::168
rDNS record for 209.51.188.168: vcs2.savannah.gnu.org
PORT STATE SERVICE
9418/tcp open git
Nmap done: 1 IP address (1 host up) scanned in 2.78 seconds
Inside container:
builder@093ec895f562:/build$ nmap -p 9418 git.sv.gnu.org
Starting Nmap 7.80 ( https://nmap.org ) at 2024-07-03 14:43 UTC
Nmap scan report for git.sv.gnu.org (209.51.188.168)
Host is up (0.14s latency).
Other addresses for git.sv.gnu.org (not scanned): 2001:470:142::168
rDNS record for 209.51.188.168: vcs2.savannah.gnu.org
PORT STATE SERVICE
9418/tcp open git
Nmap done: 1 IP address (1 host up) scanned in 2.80 second
That's interesting, just now I cleaned grub-native
recipe and tried to rebuild and got different error.
From what I can find it can be fixed if you have access to remote repository you are pulling from but as it's official GRUB repo that's impossible.
| DEBUG: Executing shell function do_configure
| /build/tmp/work/x86_64-linux/grub-native/2.06/git/bootstrap: Bootstrapping from checked-out grub sources...
| /build/tmp/work/x86_64-linux/grub-native/2.06/git/bootstrap: consider installing git-merge-changelog from gnulib
| /build/tmp/work/x86_64-linux/grub-native/2.06/git/bootstrap: getting gnulib files...
| Cloning into 'gnulib'...
| remote: warning: suboptimal pack - out of memory
| remote: fatal: Out of memory, malloc failed (tried to allocate 5977375 bytes)
| remote: aborting due to possible repository corruption on the remote side.
| fatal: early EOF
| fatal: index-pack failed
| WARNING: exit code 128 from a shell command.
ERROR: Task (virtual:native:/work/poky/meta/recipes-bsp/grub/grub_2.12.bb:do_configure) failed with exit code '1'
NOTE: Tasks Summary: Attempted 4444 tasks of which 4439 didn't need to be rerun and 1 failed.
Summary: 1 task failed:
virtual:native:/work/poky/meta/recipes-bsp/grub/grub_2.12.bb:do_configure
Summary: There was 1 WARNING message.
Summary: There was 1 ERROR message, returning a non-zero exit code.
You can temporarily fix this by cloning gnulib repo:
cd build/tmp/work/x86_64-linux/grub-native/2.06/git
git clone git://git.sv.gnu.org/gnulib
It worked for me when I did this outside of kas-container
. Tomorrow I'll check more.
And can you try to build again to see If you get same/different error?
It was still fatal: unable to look up git.sv.gnu.org (port 9418) (Temporary failure in name resolution)
on a retry.
Cloning gnulib
to build/tmp/work/core2-64-tb-linux/grub/2.06/git/gnulib
and build/tmp/work/core2-64-tb-linux/grub-efi/2.06/git/gnulib
(not grub-native
) resulted in a different error:
[...]
| gnulib/gnulib-tool: *** file /build/tmp/work/core2-64-tb-linux/grub/2.06/git/gnulib/m4/lib-ld.m4 not found
| gnulib/gnulib-tool: *** Stop.
| /build/tmp/work/core2-64-tb-linux/grub/2.06/git/bootstrap: gnulib-tool failed
That file and some others in/m4
actually get removed during build, but I guess they need to be there initially.
After checking out d271f868a8df9bbec29049d01e056481b7a1a263
(as in bootstrap.conf
), retrying several times and doing reset --hard
in kas-container shell
I was able to finish the build. Thanks for the pointers with path and shell.
It could be one of weird issues of docker, seems quite common: https://stackoverflow.com/questions/44761246/temporary-failure-in-name-resolution-errno-3-with-docker. That said I did try to hard-code address in /etc/hosts
and use host network and then got an issue connecting to that address, so it wasn't just DNS resolution. Next time I'll try to restart docker first (there was no reboot since yesterday, so no restart between retries).
If you won't find anything obvious quickly, just close this as it's probably unrelated to the contents of the repository. Could even be some kernel container issue.
@SergiiDmytruk This new error you got after cloning to grub
and grub-efi
is most likely related to https://github.com/3mdeb/meta-trenchboot/issues/31 issue.
Does the problem still persist? I just build TrenchBoot from zero and it completed successfully.
This new error you got after cloning to
grub
andgrub-efi
is most likely related to #31 issue.
That sounds like it and there is a better workaround there.
Does the problem still persist? I just build TrenchBoot from zero and it completed successfully.
I still get the error and restarting docker doesn't help, but it's probably some local thing.
Following instructions in the README (keeping caching on), I get this error:
It seems like network is just unavailable at that point of build process, but it's required for build. Any idea why it happens and how can be addressed?