[ Thanks to Anon for
this link. ]
“Some oppose a preemptible kernel because of throughput
concerns. Others oppose preemptibility because of concerns about
growing complexity in the kernel. This argument is specious,
because the preemption approach takes advantage of already required
and in place SMP locking. No additional complexity is created. All
Linux kernel engineering must already take into account SMP
requirements. Some oppose continued refinement of SMP locking to
achieve better SMP scaling (on higher way SMP systems); such
refinement has the beneficial side effect of also reducing
preemption off periods in a preemptible kernel.Preemptibility on 2.4 already provides dramatic improvements in
user process responsiveness, and while further improvement would be
beneficial, the current level of improvement is already of
tremendous value. Hence, the pro’s and con’s of improving SMP
scaling in Linux can be debated relatively independently of
preemptibility improvement opportunities.Embedded Linux companies have responsibilities to the open
source and Linux communities, as well as to the embedded system
product development communities. They have a responsibility to
innovate and release innovations early and often, for public
comment and contribution. They have a corporate responsibility to
do their best to enable Linux to be a viable operating system
platform for embedded system design and implementation. Their
customers also will find significant value in the exercise of that
responsibility, through the delivery of such product technologies
as a preemptible Linux kernel.”