I added the basic plumbing to have the the processor group framework query the CPU Power manager about power manageable CPU groupings, and then instantiate those groupings so that the dispatcher can implement scheduling policies against them.
The CPU power manager just has stubbing routines at this point...just enough to create some power processor groups. Looking ahead, I believe the CPU power manager should query the CPU PM platform driver (which will in turn ask ACPI) to enumerate the groupings and power states. Also, the CPU power manager needs interfaces from the platform driver to implement the state changes for CPU resources. That should be enough mechanism for us to sprinkle some policy over to get a basic functional prototype, I think. :) Mark, Anup, do you guys have some code we could populate into cpu_pm.c and cpu_pm.h? Thanks, -Eric Eric.Saxe at Sun.COM wrote: > Author: Eric Saxe <eric.saxe at sun.com> > Repository: /hg/tesla/pad-gate > Latest revision: d4ce43d203c81ad377d81847efa12687b620e51a > Total changesets: 1 > Log message: > PAD: Basic DVFS domain instantiation > > Files: > create: usr/src/uts/common/os/cpu_pm.c > create: usr/src/uts/common/sys/cpu_pm.h > update: usr/src/uts/common/Makefile.files > update: usr/src/uts/common/disp/cmt.c > update: usr/src/uts/common/os/pghw.c > update: usr/src/uts/common/sys/Makefile > update: usr/src/uts/common/sys/pghw.h > update: usr/src/uts/i86pc/os/mp_machdep.c > _______________________________________________ > tesla-dev mailing list > tesla-dev at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/tesla-dev >
