Re: Pthread assumption + starting real port

[Gernot Heiser]

On Wed, 15 Nov 2000 02:56:44 +0100, Farid Hajji <address@hidden> wrote,

farid.hajji> gernot> We're now working on our own libc. SawMill is doing 
something similar,
farid.hajji> gernot> and we hope to merge/share code. Doing a complete libc 
which is as
farid.hajji> gernot> independent from the underlying OS as possible (obviously 
parts are
farid.hajji> gernot> inherently tied to the OS) would be beneficial for lots of 
people.
farid.hajji> Indeed, that's absolutely true. It would be wonderful to use a 
truly
farid.hajji> portable libc in this case. This is yet another reason to avoid as 
much
farid.hajji> glibc specialities in the new Hurd as possible. Let us restrict 
ourselves
farid.hajji> to what POSIX specifies (without the Unix dependencies?) in the 
Hurd,
farid.hajji> and we'll get a system that (hopefully) compiles everywhere.
farid.hajji> 
farid.hajji> How do you plan to add support for underlying kernels (be they 
Unix or
farid.hajji> ukernels)? Through hooks?

We're primarily interested in having a library for our own use, so
prime emphasis is on clearly separating OS dependent from OS
independent parts. How to keep as much as possible of the OS-dependent
bits portable is an issue we'll look at when we are facing the
problem.

Good news was that the port of the first part from Unix to Mungi
involved mostly debugging Mungi rather than the library ;-)

Gernot


Gernot Heiser                ,--_|\   School of Computer Sci. & Engin.
Phone:  +61 2 9385 5156     /      \  The University of NSW
Fax:    +61 2 9385 5533     \_,--._*  UNSW SYDNEY NSW 2052, Australia
E-mail: address@hidden        v   http://www.cse.unsw.edu.au/~gernot