qemu: enable debug info

[alyssais]

Motivation for this change

I dropped autoPatchelfHook because it got confused with the debug objects, and I can't reproduce the pixman linker errors that caused it to be introduced in #106988 ("qemu: 5.1.0 -> 5.2.0").

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/)
• 21.11 Release Notes (or backporting 21.05 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.

[alyssais]

Drafted because I reproduced the autoPatchelfHook issue when testing with the QEMU rc.

[alyssais]

654d6b0453aa6eb19af0d75b0f087a97a5776da7 is the first bad commit

[alyssais]

In Nixpkgs we only have Meson 0.57.1, so that might explain why this isn't working.

[alyssais]

Yeah, that's because of the Meson version being too old, and QEMU using its internal version instead of our patched version. I've opened #147210 to make sure this never happens by accident, and I'll make sure we have the right version of Meson before we upgrade QEMU.

So with that all sorted out, we can go ahead with this, and we won't be at risk of rpath regressions. So this is ready for review.

[Artturin]

https://hydra.nixos.org/build/162289307
Probably caused by this pr

[vcunat]

This blocks the nixos-unstable channel now.

The 2G limit is only crossed on aarch64-linux. Stats from the previous master build (successful):

$ nix path-info -s ./result ./result-ga ./result-debug
/nix/store/fjashkx22bw4zrl9phl91acw6jgn43i6-qemu-6.1.0           1068005248
/nix/store/9spayjr081j7jlvvdassypsv0a6dk3yw-qemu-6.1.0-ga            655960

and the current build:

$ nix path-info -s ./result ./result-ga ./result-debug
/nix/store/fw893ccw8rnnfg066l2b91ypd63wx58v-qemu-6.1.0           1122797000
/nix/store/051rn39nhw9ayqh7sfx8yavysfmb113r-qemu-6.1.0-ga            776864
/nix/store/cq8cj8jd3x96kyhy2gq9f6r7xrss4yf3-qemu-6.1.0-debug     1114494472

[vcunat]

Well, x86_64-linux got to 96% of the limit, so that perhaps also isn't safe over long term.

$ nix path-info -s ./result*
/nix/store/45j5s96ivk8k4d61cayr0hla67s7bi7i-qemu-6.1.0-ga            762520
/nix/store/h6381cwzaw6fs9dzv9bgz0mgjnxk3vfx-qemu-6.1.0           1132452408
/nix/store/lyvqx84m4w02q3p2qxi3vh4c70dybi8a-qemu-6.1.0-debug      924265248

[vcunat]

Quick workaround: #152218