Re: Open letter to the UDI folks?
Jim Freeman (jfree@sovereign.org)
Thu, 24 Sep 1998 22:52:45 -0600 (MDT)
Alan writes:
> > If the Linux community would concentrate on _just_ making the UDI
> > support layer robust, Linux could focus on capably exploiting
> > whatever drivers come from whatever sources.
> >
> > GPL the UDI support layer module, but leave it out of the
> > official kernel sources, so that vendors aren't left to think
> > that just providing a UDI driver gets them of the hook for
> > "real" Linux support.
>
> Having been through the entire current UDI documentation there are two
> problems with your great suggestion
>
> 1. The documentation is currently so tattered and incomplete you
> couldnt do it. It only claims to be a draft so thats quite
> understandable.
>
> 2. There is probably 6 to 8 weeks work to get it even going. You
> would need to hack up the linux memory allocator to make it work
> at all. The way descriptor blocks are passed is going to make
> networking using such devices crawl.
>
> Its too big a job to be worth doing. I 6 to 8 weeks solid work I can
> probably write 3 or 4 drivers
With Linus's official "wait and see" attitude, let the Project UDI
folk do (or persuade/pay whoever to do) the support module implementation,
consuming exported ksyms. For things that need to be re-worked as
you say, let the onus be on the UDI implementors to convince the
Kernel Gods (SM - registration applied for) of any kernel facilities
that need to added or changed, BUT on the basis of how/why such
changes are good for Linux, NOT on the basis of them being required
for UDI. If the kernel doesn't get bettered by accommodating UDI,
UDI doesn't get accomodated.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/