[PATCH] D24595: [lib/LTO] Remove now unneded hack for undefined symbols

[Teresa Johnson via llvm-commits]

On Thu, Sep 15, 2016 at 10:40 AM, Rafael Espíndola <
rafael.espindola at gmail.com> wrote:

>
>> Not sure I understand the question. We were previously calling
>> UpdateCompilerUsed before invoking llvm::internalizeModule, presumably to
>> prevent internalization. Now we simply use the callback to query this
>> information directly.
>>
>>
> Note that in the case of assembly uses, we don't want to prevent
> internalization, we want to prevent the internalized symbol from being
> deleted.
>

I see, so we need to call that routine to update the llvm.compiler.used
regardless of internalization. Which means I should put it back in
lto::thinBackend, not in thinLTOInternalizeModule.

And presumably lto::backend() should get an instance of it too?

Teresa


> Cheers,
> Rafael
>
>



-- 
Teresa Johnson |  Software Engineer |  tejohnson at google.com |  408-460-2413
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20160915/9657cafc/attachment.html>