Closed jean-paullizotte closed 1 year ago
The entrypoint.sh
file should already have execute permission set when checked out of the repo and that should carry over when copied into the container.
It shouldn't need to be explicitly set
It's working as expected when I run the docker-demian.sh
script on my Fedora machine.
What environment are you running the build script?
(In future please fill out the full template you are presented with when opening an issue, it collects all the information we need to work on problems.)
I'm running it on raspberry pi OS. I'm using a NFS remotely shared folder, for the root of the build, if that has any incidence. Is there any harm in adding it just in case?
Do the files in the docker-custom/scripts
directory have the execute bits set when checked out?
I tried a git clone from a different machine and the execute bits are there! I have no clue, why.
Please look at the nfs mount options in the /etc/fstab
for the mount in question on the pi with the build problem, looking for the noexec
option
Will. Do, I'll close the issue in the meanwhile, since there's a straightfoward way to avoid this. Thanks for the help!
Fairly straightforward, when I try to run a built container with the "docker-custom\docker-debian.sh" script. The "docker-custom\Dockerfile.debian" seems to be missing a
RUN chmod +x ./entrypoint.sh
in it.