New package: radm-0.6.4

[Goorzhel]

• I tested the changes in this PR: YES
• This new package (however self-serving) conforms to the package requirements: YES

[Goorzhel]

Both CI failures had to do with the chvt crate. I'll try to reproduce this locally (

❯ for arch in {x86_64,armv{6,7}l,i686,aarch64}{,-musl}; do
  ./xbps-src pkg -a $arch $(git b) &> $arch.log
done

) and submit any necessary patches to the crate's upstream.

[Goorzhel]

Also, a majority of the architectures on this run returned success, but actually failed with

=> ERROR: pandoc-2.17.1.1_1: cannot be cross compiled...
=> ERROR: pandoc-2.17.1.1_1: yes

because I mistakenly added pandoc to makedepends rather than hostmakedepends. I should investigate that later, too.

[Goorzhel]

Patches sent upstream to unbreak non-Intel builds:
jdonszelmann/chvt-rs#2
1wilkens/pam#36

I still need to test these changes at runtime so I'll leave the branch unsquashed for now.

[Goorzhel]

Also, a majority of the architectures on this run returned success, but actually failed

Turns out build.sh was eating the exit code. #38841 I'm an idiot.

[q66]

pandoc is not portable, put the manpage behind a build option that is only enabled if the host architecture is one that has it (x86_64*|i686|ppc64le*|ppc64)

[Goorzhel]

This is as done as it's ever going to be. Squashing.