Open viferga opened 1 month ago
I’ve attempted several strategies to resolve the multi-architecture build issues, including:
Guix=jobs=1
)GUIX_JOBS
environment variable to limit the number of jobs to 1:
ENV GUIX_JOBS=1
sudo fallocate -l 16G /swapfile
sudo chmod 600 /swapfile
sudo mkswap /swapfile
sudo swapon /swapfile
swapon --show
free -h
docker run --memory=4g ...
RUN
commands in the Dockerfile to isolate steps and reduce the chance of failures:
RUN guix pull
RUN guix package --fallback -i nss-certs
arm64/v8
, I commented out the other platforms in the Docker build command:
docker buildx build --platform linux/arm64/v8 --push ...
debian:bookworm
and debian:bookworm-slim
by modifying the FROM
line in the Dockerfile:
FROM debian:bookworm
or
FROM debian:bookworm-slim
wget https://git.savannah.gnu.org/cgit/guix.git/plain/etc/guix-install.sh
sudo bash guix-install.sh
Despite trying all these solutions, the problem is still not resolved. Any further advice or suggestions would be really helpful.
@Bcoderx6 thank you so much for the investigation.
linux/arm/v7
It seems this issue depends on the kernel image, we should investigate it further:
Reference: https://lists.gnu.org/archive/html/help-guix/2017-12/msg00023.html
linux/arm64/v8, linux/ppc64le
For OOM, we should find a way to tweak QEMU options and truly investigate if the OOM is reached due to QEMU limits or the VM from GitHub itself. Here's the reference:
Reference: https://github.com/docker/setup-qemu-action https://hub.docker.com/r/tonistiigi/binfmt https://github.com/tonistiigi/binfmt