Implement current_exe for AIX

[bzEq]

Added current_exe implementation for AIX. This implementation doesn't rely on system specific syscalls, I guess some other targets can also use it.

[rustbot]

r? @Mark-Simulacrum

(rustbot has picked a reviewer for you, use r? to override)

[rustbot]

Hey! It looks like you've submitted a new PR for the library teams!

If this PR contains changes to any rust-lang/rust public library APIs then please comment with @rustbot label +T-libs-api -T-libs to tag it appropriately. If this PR contains changes to any unstable APIs please edit the PR description to add a link to the relevant API Change Proposal or create one if you haven't already. If you're unsure where your change falls no worries, just leave it as is and the reviewer will take a look and make a decision to forward on if necessary.

Examples of T-libs-api changes:

• Stabilizing library features
• Introducing insta-stable changes such as new implementations of existing stable traits on existing stable types
• Introducing new or changing existing unstable library APIs (excluding permanently unstable features / features without a tracking issue)
• Changing public documentation in ways that create new stability guarantees
• Changing observable runtime behavior of library APIs

[Mark-Simulacrum]

Are we searching PATH and PWD on other platforms today? I'm not sure we should do that here -- it increases the amount of work done and if it's not done on other platforms is somewhat inconsistent (and could have false positives much more easily).

[bzEq]

Are we searching PATH and PWD on other platforms today?

I don't know much about other rare seen platforms. On AIX, we don't have information like /proc/self/exe in /proc filesystem. We are using similar method in LLVM on AIX right now, see https://github.com/llvm/llvm-project/blob/42b21ddaadd3545945f29a8ccdcc89779542c30e/llvm/lib/Support/Unix/Path.inc#L251. (Actually, we don't have /proc/curproc/file on AIX).

[Mark-Simulacrum]

Right, but you're using argv[0] just above which is what we do on other platforms (AFAICT, looking in this file, for example).

I am inclined to leave it there unless there's strong motivation for doing this cwd/path traversal. current_exe is documented as fallible for a reason.

[bzEq]

I extended the one for fuchsia. Correct me if I still failed to get your point. :)

[Mark-Simulacrum]

Apologies if I wasn't clear. I think we should remove the PATH searching from this function, just returning the passed argument. IOW, I think if just adding aix to the fuchsia function works, then that seems like the right initial implementation here. (Not adding any other code to it).

[bzEq]

I think we should remove the PATH searching from this function, just returning the passed argument.

Could you please elaborate more about your concerns of searching in PATH here? Cuz on AIX, we don't have other methods to find the executable file of current process if argv0 is not an absolute path.

[rust-log-analyzer]

The job x86_64-gnu-llvm-13 failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

   Compiling addr2line v0.17.0
error: unexpected `cfg` condition value
   --> library/std/src/sys/unix/os.rs:454:34
    |
454 | #[cfg(any(target_os = "fuchsia", target_os = "aix"))]
    |
    |
    = note: expected values for `target_os` are: android, cuda, dragonfly, emscripten, espidf, freebsd, fuchsia, haiku, hermit, horizon, illumos, ios, l4re, linux, macos, netbsd, none, openbsd, psp, redox, solaris, solid_asp3, tvos, uefi, unknown, vxworks, wasi, watchos, windows, xous
    = note: `-D unexpected-cfgs` implied by `-D warnings`
error: could not compile `std` due to previous error
Build completed unsuccessfully in 0:00:23

[bzEq]

Hi @Mark-Simulacrum , this PR is subsumed by #109882, could you please help review that PR?