[PATCH] D96035: [WIP][dsymutil][DWARFlinker] implement separate multi-thread processing for compile units.

[David Blaikie via llvm-commits]

On Fri, Feb 5, 2021 at 7:38 AM Alexey Lapshin via Phabricator
<reviews at reviews.llvm.org> wrote:

> 1. In multi-thread mode, some canonical dies might become not seen by some threads.
>
>   If the request for setting canonical dies come from several threads at the same time - some threads might receive information that there is not canonical die(though it is actually, it would be set by another thread). As the result, some type would not be de-duplicated. But the number of such cases is small. For the clang, there is no noticeable difference in the size of the binary.

That sounds concerning - does this mean the result would be
nondeterministic? Determinism is an important feature.

> >> e) .debug_names is not supported by this patch.
> >
> > Any reason why this can't be done? I believe that dsymutil just generates the .debug_names (or Apple accelerator tables) after you have written all of the DWARF out by traversing it no? If this doesn't happen, it can be added with the --update option to dsymutil, but again, this takes time.
>
> There is no fundamental reason. I think it might be supported.
> I just need more time to check that code and understand how to implement it properly.
> I would update this patch with implementation later.

Yeah, if it's not already supported (don't think it is), it'd be good
to do it separately/follow-up patch. Smallest patches making
incremental progress (without having to go through too many
convolutions/rewrites/etc) make it easier to review, etc.