Re: Getting Started with Hurd-L4

[Bas Wijnen]

Sam Mason wrote:
> Why aren't the rootservers going to use it?  is it that glibc is just
> too big and general purpose and the rootservers are generally going to
> be doing such specalised stuff that it doesn't make it worthwhile, or
> is it something else?

Libc will be implementing its services for a big part by calling 
physmem, task and perhaps some other servers.  The rootserver (and 
physmem and task) cannot use that, of course.  As for the other parts, 
the parts they need from it are currently in libc-parts.  It's just the 
implementation from libc AFAIK, but it takes much less space, indeed.

>>[capability stuff by Marcus]
> I'll try and keep my eyes out!  I'm assuming this is going to be the
> basis for security in the system?

The capability system can be used by any server to provide secure access 
by clients.  Physmem uses it to let tasks get and manipulate their 
mappings, task uses it to allow task creation and destruction, to name a 
few things.  It's not all security-related, but if it has bugs, then 
they are likely a big security problem, because the tasks with access to 
the root server use it.

> > The only other part that has some merit is the wortel/startup.c and
> > wortel/wortel.c:start_task() code which shows how to spawn a new
> > program (ie, the exec protocol).  That is still under development,
> > though.
>
>
> I think I found that already and tied myself in knots trying to figure
> out how it worked.

Please tie a little more.  There seems to be a bug somewhere there, and 
I've spend days trying to find it, so far without success.  Currently I 
have the code from startup.c calling physmem, physmem picks up the call, 
replies and receives a timeout, meaning that the thread isn't in a 
receive ipc phase.  That's not possible of course, as it just made a 
call (with infinite receive timeout.)  Anyway, this isn't any stranger 
than what I saw before :-(  If you could spot that problem, it would be 
great :-)

> > The rest (except for the other support libraries, like libhurd-slab
> > and libhurd-ihash, and libpthread) is basically bootstrap junk.
>
>
> You you think they're going to be fleshed out more, or will they get
> dumped as things progress?

Bootstrap stuff will always be needed of course.  I think (but note I 
don't decide things here) that they'll get a bit more configurable, but 
stay roughly as they are.  Or we may want to change the bootstrapping 
process, but I don't expect that to happen before the system can at 
least boot.  Actually I don't expect it to happen after that either, as 
there isn't anything wrong with the current approach AFAIK.

Thanks,
Bas

signature.asc
Description: OpenPGP digital signature