[PATCH] D62055: [ARM][AArch64] Revert Android Bionic PT_TLS overaligning hack

[Rui Ueyama via llvm-commits]

I didn't mean that I wanted to re-submit it for 8.0.1, but I made another
comment for another thread. As you originally pointed out, --android-tls
seems like the way to go.

On Mon, May 20, 2019 at 4:12 PM Fangrui Song via Phabricator <
reviews at reviews.llvm.org> wrote:

> MaskRay added a comment.
>
> In D62055#1508038 <https://reviews.llvm.org/D62055#1508038>, @ruiu wrote:
>
> > I'd vote for re-submitting the original patch with a larger alignment to
> avoid introducing Android-specific ABI to the output, as I said in a
> different review thread. Is there any problem to do that?
>
>
> @ruiu To fix the hack isn't impossible, but that would make the code more
> hacky (D61824 <https://reviews.llvm.org/D61824> and D61825 <
> https://reviews.llvm.org/D61825>), and some ARM/AArch users (e.g. musl)
> are not happy. As I described in https://reviews.llvm.org/D62055#1507426,
> the relevant support is all in place, the nearly zero-overhead workaround
> in Bionic `crtbegin_{dynamic,static}.o` should be straightforward now. I
> don't know why we would like to re-introduce the hack and cause churn in
> the 8.0.1 release.
>
>
> Repository:
>   rL LLVM
>
> CHANGES SINCE LAST ACTION
>   https://reviews.llvm.org/D62055/new/
>
> https://reviews.llvm.org/D62055
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190520/87594eb5/attachment.html>