[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 64bit GNU Mach

From: Samuel Thibault
Subject: Re: 64bit GNU Mach
Date: Mon, 2 Apr 2012 14:18:34 +0200
User-agent: Mutt/1.5.21+34 (58baf7c9f32f) (2010-12-30)

Richard Braun, le Mon 02 Apr 2012 14:12:27 +0200, a écrit :
> On Mon, Apr 02, 2012 at 01:57:31PM +0200, Samuel Thibault wrote:
> > Richard Braun, le Mon 02 Apr 2012 13:46:08 +0200, a écrit :
> > > How do they convert names from/to pointers ? Regular IPC space lookups ?
> > 
> > Well, yes, just like GNU Mach does, in ipc_kmsg_copyin_header etc.
> So we'd loose the optimization, which doesn't seem to be a problem
> considering it doesn't concern our main performance issues, and we'd
> gain a cleaner interface between ports and names. From what I've seen,
> the Hurd doesn't rely on the size of port names either, as it always
> uses some lookup mechanism (actually ihash only if I'm right). Now, is
> there a point having two types for kernel/user port names, or can't we
> just stick to mach_port_t having the right size (the one which matters
> for user tasks) ?

But then it needs some trick to convert to pointer everywhere needed.
Trying to manage low-4GiB allocation to avoid the trick would make
things nasty: remember that x86_64 does not have segmentation any more,
only flat addressing space, so we'd still need a 4GiB User/Kernel
separation, which was precisely what I wanted to avoid.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]