Hi,

I mentioned this a while back, and don't recall any other response except from Martin on this.  I'd like to drop the ucontext implementation in the Fiber implementation and move the inline asm into threadasm.S so everyone benefits.

This is one of the blockers for gdc on Mac/OSX for instance, and a slimmed down size of Fiber for gdc on GNU/Linux would be ideal - along with removing the current modifications we make to thread.d/thread.di to make the static asserts pass.

Regards
Iain
_______________________________________________
dmd-internals mailing list
dmd-internals@puremagic.com
http://lists.puremagic.com/mailman/listinfo/dmd-internals

On 03/23/2014 01:17 PM, Iain Buclaw wrote:
> Hi,
>
> I mentioned this a while back, and don't recall any other response
> except from Martin on this.  I'd like to drop the ucontext
> implementation in the Fiber implementation and move the inline asm
> into threadasm.S so everyone benefits.
>
> This is one of the blockers for gdc on Mac/OSX for instance, and a
> slimmed down size of Fiber for gdc on GNU/Linux would be ideal - along
> with removing the current modifications we make to thread.d/thread.di
> to make the static asserts pass.
>
> Regards
> Iain
> _______________________________________________
> dmd-internals mailing list
> dmd-internals@puremagic.com
> http://lists.puremagic.com/mailman/listinfo/dmd-internals
>
Absolutely fine for me, ucontex_t is a terrible / slow default implementation.
Discussion was here https://github.com/D-Programming-Language/druntime/pull/710.
_______________________________________________
dmd-internals mailing list
dmd-internals@puremagic.com
http://lists.puremagic.com/mailman/listinfo/dmd-internals