[WebAssembly] Use absolute pointers

[MaxDesiatov]

WebAssembly target uses absolute pointers, as explained by @kateinoigakukun on Swift Forums:

It's due to the design of WebAssembly and LLVM support.
If LLVM supports substruction relocation type on WebAssembly like R_X86_64_PC32 or X86_64_RELOC_SUBTRACTOR, this issue can be solved easily.

But LLVM doesn't support this relocation type because of the design of WebAssembly.

First, R_X86_64_PC32 and X86_64_RELOC_SUBTRACTOR are mainly used to generate PIC but WebAssembly doesn't require PIC because it doesn't support dynamic linking. In addition, the memory space also begins at 0, so it's unnecessary to relocate at load time. All absolute addresses can be embedded in wasm binary file directly.

As confirmed by @jckarter:

I agree using an absolute pointer in place of relative is the way to go.

Related to SR-9307.

[MaxDesiatov]

@compnerd could you have a look at this please? Thanks!

[compnerd]

This seems fine, but doesn’t IRGen need to be updated as well?

[MaxDesiatov]

@compnerd it does, would you like me to include that in this PR or a separate PR?

[compnerd]

Yes please, could you make those changes as part of this PR?

[jckarter]

Similar to #31811, it'd be nice if we could make this behavior conditional on the Runtime template argument, instead of using conditional compilation. That would make it easier to support cross-platform reflection tools where a non-wasm host tool wants to use MetadataReader to inspect a wasm binary or running process.

[MaxDesiatov]

@swift-ci please smoke test

[MaxDesiatov]

@jckarter I'm happy to make this conditional based on the Runtime template argument, but I'm not sure how that argument should be passed. detail::applyRelativeOffset takes only two template arguments, should I add a third one with typename Runtime? Maybe applyRelativeOffset should become a member function of Runtime?

Also, even with that approach the point of use of detail::applyRelativeOffset doesn't take any template arguments for runtime. How would I pass it at the point of use then?

[kateinoigakukun]

From our forum discussion, we are going to use absolute pointer only for function pointer #42094