[FreeVMS] Re: split with 2.4.18 or 2.6.x?


Subject: [FreeVMS] Re: split with 2.4.18 or 2.6.x?
From: BERTRAND Joël (joel.k.bertrand@free.fr)
Date: Sun Jan 18 2004 - 16:22:47 CET


Roar Thronæs wrote:
> Hi

        Hello Roar,

> I have been wondering whether to make the split/fork with 2.4.18
> or do it with a 2.6.x instead.
> (Meaning a definitive move away from Linux in such a way that it takes
> its own path)
>
> With 2.4.18:
> Negative sides:
> Some effort needed to be able to use gcc 3.3 and newer.
> Might get some less functionality and older drivers.
>
> Positive sides:
> Can have uml with backtracing.
> We are already using it.
>
> With 2.6.x:
> Positive sides:
> Might get a better uml.
> We get some newer drivers and more functionality.
>
> Negative sides:
> We might have to cut away those drivers and functionality anyway, or
> be slowed down in the process of changing more stuff than with 2.4.18.
> 2.6.0 is not in sync regarding uml; need a patch from test9.
> Can not use gcc 2.96 to compile uml. (And looses backtracing)
> It might demand too much effort.
>
> I have begun to look at 2.6.0, but have gotten some problems.
> So far:
> Uml w/o MM and I/O boots, 386 on PC and bochs w/o MM and I/O locks
> before starting sh, and have not gotten any kernel with MM and I/O to compile.
> (The MM and I/O changes since 2.4.18 might have some tricky stuff.)

        I have tried to compare the official linux kernel 2.4.x and the new 2.6
branch. For me, the 2.6 kernel is not stable enough (some drivers I have
tested are broken and the new build process is not mature and give me
some 'unresolved references'). I have tested it only on i386
(uniprocessor) because other ports are known broken.

> Right now I am inclined to stick with 2.4.18 and do a split/fork with it.

        Yes. I think it's a good idea.

        Best regards,

        JKB

-- 
Liste de diffusion FreeVMS
Pour se désinscrire : mailto:freevms-request@ml.free.fr?subject=unsubscribe



This archive was generated by hypermail 2b25 : Sun Jan 18 2004 - 16:22:17 CET