Make -betterC not disable EH, let -nothrow handle that

[ryuukk]

See #16174 for context

-nothrow was added recently, wich is what the user should use to disable EH, otherwise APIs like backtrace can't be used to generate stack traces, wich is useful for debugging

I added a change log entry, but my english might be poor, so please don't hesitate to rephrase it

[dlang-bot]

Thanks for your pull request and interest in making D better, @ryuukk! We are looking forward to reviewing it, and you should be hearing from a maintainer soon.
Please verify that your PR follows this checklist:

• My PR is fully covered with tests (you can see the coverage diff by visiting the details link of the codecov check)
• My PR is as minimal as possible (smaller, focused PRs are easier to review than big ones)
• I have provided a detailed rationale explaining my changes
• New or modified functions have Ddoc comments (with Params: and Returns:)

Please see CONTRIBUTING.md for more information.

---

If you have addressed all reviews or aren't sure how to proceed, don't hesitate to ping us with a simple comment.

Bugzilla references

Your PR doesn't reference any Bugzilla issue.

If your PR contains non-trivial changes, please reference a Bugzilla issue or create a manual changelog.

Testing this PR locally

If you don't have a local development environment setup, you can use Digger to test this PR:

dub run digger -- build "master + dmd#16177"

[ryuukk]

Can someone review this please?

[ryuukk]

To help understand what this PR does:

For cases where one uses -betterC or uses a custom runtime, it makes it possible to still use debugging APIs like backtrace

Example code:

extern(C) void main()
{
    rt_register_crash_handler();
    crash_me();
}


void crash_me()
{
    int* a;
    *a = 42;
}

(full code + rt_register_crash_handler impl: https://gist.github.com/ryuukk/)e2d5260b4c4acec1b972ad9ce9228753)

The code should crash and print:

-------------------------------------------------------------------+
Received signal 'SIGSEGV' (11)
-------------------------------------------------------------------+
executable: /tmp/dmd_runXn753y
backtrace: 7
    ??:? _start+0x25
    ??:? __libc_start_main+0x8a
    ??:? +0x25cd0
    /run/media/ryuukk/E0C0C01FC0BFFA3C/dev/kdom/./_.d:4 main+0xe
    /run/media/ryuukk/E0C0C01FC0BFFA3C/dev/kdom/./_.d:11 void _.crash_me()+0xe

But if i compile with -betterC or with a custom runtime, i get:

-------------------------------------------------------------------+
Received signal 'SIGSEGV' (11)
-------------------------------------------------------------------+
executable: /tmp/dmd_run1UPD7W
backtrace: 1

Not helpful at all..

This PR solves that

ping @WalterBright

[dkorpel]

The idea sounds good, but tests are currently failing and the log is expired so I can't see why.

It's also curious that there are no changes for Windows, does that already generate the necessary code for backtraces with -betterC?

[dkorpel]

When you say 'working', it might give the impression that a failing assert now gives a stack trace out of the box. Perhaps clarify that it 'works' on a more technical level, that the generated code contains the unwind info needed for e.g. libunwind.

[WalterBright]

The idea of -betterC is to not require the use of the D runtime library - only the C runtime library is needed. This PR requires linking in druntime, defeating the entire purpose of betterC.

To get backtraces in betterC, do one of:

1. do it like C compilers do
2. turn off -betterC, debug the code, then turn it back on

[WalterBright]

See my other comment

[ryuukk]

The idea of -betterC is to not require the use of the D runtime library - only the C runtime library is needed. This PR requires linking in druntime, defeating the entire purpose of betterC.

To get backtraces in betterC, do one of:

1. do it like C compilers do

2. turn off -betterC, debug the code, then turn it back on

I didn't know it would require linking druntime, in that case, this is a bad PR

1. i'll investigate, if you have pointers/ideas please share
2. that's what i am doing currently, but would prefer not to, since i rely on -betterC to make sure i don't do things that require TypeInfo

I'll close it until i can find a better solution

[rikkimax]

I've had to do another look into ldc, looks like I've been wrong, it does turn off the unwinding tables for -betterC too.

Which is unfortunate, this certainly could be templated and turned on, oh well.