Re: sigsp on Hurd/x86_64

[Sergey Bugaev]

Hello,

On Fri, May 12, 2023 at 10:15 PM Bruno Haible <bruno@clisp.org> wrote:
> make no sense to me. After this statement, sigsp is
> either 0x0000000000000000
> or     0x0000000000000010.
>
> If aligning SP is intended, the line should read
>
>   sigsp = (void *) ((uintptr_t) sigsp & -16UL);

Indeed -- thank you for spotting it! I'll include the fix in my next
patch series.

As you can see we have not tested signals on x86_64 at all yet --
though exactly as of today there are now enough things working (the
proc server, the signal thread) to actually attempt it :)

Sergey