[Lldb-commits] [PATCH] D39727: Make TestTopLevelExprs more robust in face of linker GC

[Zachary Turner via lldb-commits]

Fair enough, originally I thought this was the “traditional” kind of GC,
which would have made this approach simpler
On Tue, Nov 7, 2017 at 6:23 AM Pavel Labath <labath at google.com> wrote:

> On 7 November 2017 at 13:51, Pavel Labath <labath at google.com> wrote:
> > On 7 November 2017 at 13:29, Zachary Turner <zturner at google.com> wrote:
> >> If it works, is easy, and doesn’t regress anything I’d honestly rather
> just
> >> disable linker gc.
> > Ok, I'll try that.
> I'm starting not to to like this. Adding --whole-archive produces
> zillions of "multiple definitions of a symbol" errors on android.
> These can be squelched by telling the linker to ignore these errors
> (-z muldefs), but that's not really a supported configuration, and at
> that point this starts looking like a quite ugly hack.
>
> I think we should just stick to the original proposal of ensuring the
> code is pulled in by actually introducing a dependency. If it turns
> out we need to do this in a lot of places, we can revisit this
> approach.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/lldb-commits/attachments/20171107/6c4ca932/attachment.html>