Provided these modern fast CPUs serve old slow peripheral devices
or devices, which are insensitive to latency. Do we really leave dumb,
but high speed, serial devices to RTlinux or to that poor netbsd? 8)
At a certain point, it might not be worth trying to give up performance
for common CPU's just to support really ancient hardware. For example,
Linux won't work well on a PDP-11 or a VAX. Should we make radical
changes just to support it, especially if it lowers performance on
commonly available CPU's?
By the way, the system used by code, working at BHs, does not differ
of single-level spls essentially, to be honest.
When I introduced BH's into Linux (yes they were my fault --- so shoot
me), I did so with an eye to the fact that we weren't going to be
getting SPL's into Linux, but that it was the minimal requirement that
would allow the serial driver to go fast (even 8250's on an i386).
Just because it's a simple implementation of BSD's software interrupts
doesn't mean that the more complicated scheme is a good thing.
- Ted
-
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/