[lkml]   [2000]   [Jul]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: a joint letter on low latency and Linux
Albert D. Cahalan wrote:
> >> No, you can use a pure PentiumIII-SMP kernel as long as it provides
> >> the code-modification services for generic modules. Compiled-in
> >> drivers simply don't call (or trap to) the self-modification engine.
> >
> > What if i want a pure PIII-UP kernel? How are the different data
> > structures handled?
> You get it, but you may need to give up generic module support.
> This isn't any worse than what we have today.

But it _is_ worse, unless you expect the generic module support
not to be widely adopted, in which case there is no point in
having it...
Having a generic driver api and being able to use one binary driver
(with or w/o src) with all kernel configuration would certainly
be a good thing, but pushing the support cost into the core
kernel isn't. Not when you can build "generic" modules on top
of another layer which hides all the differences from the driver.
Yes, the runtime cost is much higher, but it is localized -- except
for the generic modules nothing else is affected.
Hmm, you could also use the previously mentioned linking tricks
to significantly reduce the overhead added by the extra layer.
(eg by having a smarter "insmod" loader)
Now the kernel and all "native" drivers aren't slowed down, and
the binary drivers are reasonably efficient.

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at

 \ /
  Last update: 2005-03-22 13:57    [W:0.078 / U:3.904 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site