On Wed, 15 May 2019 at 10:48, Stefan Hajnoczi <stefa...@gmail.com> wrote: > > On Thu, May 09, 2019 at 02:44:39PM +0100, Peter Maydell wrote: > > On Sun, 5 May 2019 at 16:41, Alex Bennée <alex.ben...@linaro.org> wrote: > > > > > > > > > Paolo Bonzini <pbonz...@redhat.com> writes: > > > > > > > *** BLURB HERE *** > > > > > > I assume there was going to be a bit more background here? > > > > Mmm, could we have the rationale, please ? > > Paolo can add more if necessary, but my understanding is: > > 1. It's required for Intel Control-flow Enforcement Technology (CET). > The existing ucontext backend doesn't work with CET. > 2. It's faster than the existing ucontext implementation.
Mmm, I think we've talked about 1 before, but I think it would be useful to clearly state why we need to do things here. It's also useful for identifying whether we need an asm backend for every host, or only some hosts (and if so which). I'm unconvinced by 2 as a rationale for adding more host asm. Coroutines were already bad enough when they were at least vaguely portable C code. thanks -- PMM
