chipsec: Python2 to Python3 conversion/refactoring

[net-wayfarer]

This discussion is a continuation from a PR that I previously made.

Initially I was misunderstanding that chipsec package depended on python2* and so I decided to go modifying the PKGBUILD further to have makepkg do the conversion/refactoring during the prepare() {} process,

$ grep -A 8 prepare PKGBUILD
prepare() {
  cd "$pkgname"
  
  2to3 -w .

  cd "drivers/linux"

  patch -Np0 < "$srcdir/Makefile.patch"
}

What actually ended up happening is most probably that the Python files got refactored (warning, huge paste ahead). I haven't experimented with chipsec_main.py yet, but am intending to canvass some opinions on this. Would there be any point in pushing a PR with the aforementioned changes even if it may not be useful? I am also partly thinking aloud here.

Full log: Deleted https://paste.rs/jar (Warning: extremely long log)

[aancw]

Hey, can you just paste link for full log? It's so very long scroll.

[net-wayfarer]

Sorry, I just went and fixed that.

[aancw]

Thank for your understanding 🙂

[Edu4rdSHL]

Hello, @net-wayfarer, @aancw, hope that you're fine. I see that chipsec is currently being compiled using Python3, what are the advantages of using 2to3 in this case? I think it will probably bring more problems than improvements.

[net-wayfarer]

Hi @Edu4rdSHL. Yeah it is compiled in Python3 as I later realised after checking on older prebuilt package that I did; chipsec-4:1070.7966175. Hence brought me to the discussion of questioning would there be any merits in using 2to3 -w . when there appears to be some changes being applied to the Python files.

In either case, I have not tested chipsec fully. Regardless if I refactored it or not, as I was primarily concerned with the error status whenever pacman invokes dkms during system updates, and notably kernel updates.

I think at this point, there appears to be no real benefit (if any) and am going to leave the PKGBUILD unchanged in the Blackarch repo. One good thing that I didn't push any changes and PR 😄 with regards to this.

Thanks for inputs!

[Edu4rdSHL]

Hey @net-wayfarer, I agree, we should keep the PKGBUILD without any change, it's working fine currently (at least the module installation) so there isn't a real reason to include the change mentioned above.

Thank you too!