[PATCH 1/2] 476: Set CCR2[DSTI] to prevent isync from flushing shadow TLB
shaggy at linux.vnet.ibm.com
Tue Sep 28 01:26:45 EST 2010
On Mon, 2010-09-27 at 11:04 -0400, Josh Boyer wrote:
> On Fri, Sep 24, 2010 at 01:01:36PM -0500, Dave Kleikamp wrote:
> >When the DSTI (Disable Shadow TLB Invalidate) bit is set in the CCR2
> >register, the isync command does not flush the shadow TLB (iTLB & dTLB).
> >However, since the shadow TLB does not contain context information, we
> >want the shadow TLB flushed in situations where we are switching context.
> >In those situations, we explicitly clear the DSTI bit before performing
> >isync, and set it again afterward. We also need to do the same when we
> >perform isync after explicitly flushing the TLB.
> >Th setting of the DSTI bit is dependent on
> >CONFIG_PPC_47x_DISABLE_SHADOW_TLB_INVALIDATE. When we are confident that
> >the feature works as expected, the option can probably be removed.
> You're defaulting it to 'y' in the Kconfig. Technically someone could
> turn it off I guess, but practice mostly shows that nobody mucks with
> the defaults. Do you want it to default 'n' for now if you aren't
> confident in it just quite yet?
I think I made it a config option at Ben's request when I first started
this work last year, before being sidetracked by other priorities. I
could either remove the option, or default it to 'n'. It might be best
to just hard-code the behavior to make sure it's exercised, since
there's no 47x hardware in production yet, but we can give Ben a chance
to weigh in with his opinion.
> (Linus also has some kind of gripe with new options being default 'y',
> but I don't recall all the details and I doubt he'd care about something
> in low-level PPC code.)
IBM Linux Technology Center
More information about the Linuxppc-dev