stdenv: switch default from gcc10 to gcc11

[fabianhjr]

Description of changes

This will mainly affect i686 and ARM which is still using gcc10;

x86_64 (darwin and linux), RISC-V (linux), and aarch64 (darwin) are
already using gcc11 so this will remove their version override.

aarch64 (linux), vc4, and relibc will remain pinned to previous versions
due to unresolved issues.

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandbox = true set in nix.conf? (See Nix manual)
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 22.05 Release Notes (or backporting 21.11 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
  • (Release notes changes) Ran nixos/doc/manual/md-to-db.sh to update generated release notes
• Fits CONTRIBUTING.md.

[fabianhjr]

The platforms that changed version would be ARM and i686 (need testing)

I tested x86_64 linux and had to rebuild a couple hundred or so packages and stdenv.

Would appreciate testing on other platforms like aarch64-linux (gcc9), x86_64-darwin, aarch64-darwin, riscv-linux, etc.

[vcunat]

I tested x86_64 linux and had to rebuild a couple hundred or so packages and stdenv.

I don't understand this line. Oh you meant you rebuilt those because you based on staging? (as this commit doesn't switch that)

i686-linux couldn't make this recently, but I have a vague memory that the issue has been fixed in the meantime. I'm already rebuilding wine which I used earlier to run into the problem.

[fabianhjr]

I don't understand this line. Oh you meant you rebuilt those because you based on staging? (as this commit doesn't switch that)

This commit is on top of master (not staging) unsure why I had a stdenv rebuild and a couple hundred rebuilds though.

[vcunat]

No (stdenv) rebuild on x86_64-linux for me (or OfBorg).

[fabianhjr]

Ah maybe I catched the merge of staging-next into master, its good to know that this doesn't cause a mass-rebuild/stdenv-rebuild.

[vcunat]

That had binaries in cache.nixos.org long before the merge. (except for aarch64-linux)

[SuperSandro2000]

If we stay in not red rebuilds we could easily merge this into master if you want. I don't think there is much testing for platforms other the two darwins and linux x86_64 and aarch64 on staging anyway.

[vcunat]

wine built fine for me (over 700 builds, uh), so i686-linux is probably good to go. I don't know if someone can test 32-bit ARM.

[fabianhjr]

Maybe someone from @NixOS/mobile-nixos ?

[misuzu]

boehm-gc tests are broken on armv7l with gcc11: https://gist.github.com/misuzu/46c4eb4387676124054f7e20a252554b
Maybe related: ivmai/bdwgc#428 (comment)

[misuzu]

Tested building nix and git with this change on top of 74a1793 for armv7l-linux, both native & cross.
The issue with boehm-gc is gone now.