Sergey Bugaev, le mer. 09 août 2023 11:48:29 +0300, a ecrit:
> On Wed, Aug 9, 2023 at 4:10 AM Samuel Thibault <samuel.thibault@gnu.org> wrote:
> > So, is anybody against making this change?
>
> I trust Flávio to understand RPC ABI much better than I do.
>
> That being said, I don't quite understand the point of this change: we
> cannot get rid of mach_msg_type_long_t completely on 32-bit platforms
> anyway, so what advantages does this bring?
It makes mach_msg_type_long_t shorter, down to 8 from 16 bytes.
Also, and perhaps more important, it removes all the existing holes in mach_msg_type_t and mach_msg_type_long_t, which prevents possible leakages to user space.
I think we will need to patch some code in the dde repository because of the way we build mach_msg_type_t (we need to use designated initializers).
Somewhat unrelated question: what is the future of DDE? It seems like the dde repo has a lot of forked code from the main hurd repo which makes it easy to miss some changes.
Samuel