On Mon, Sep 8, 2014 at 3:16 PM, Arnd Bergmann [off-list ref] wrote:
[Me]
quoted
I can add a comment explaining that ARMs terminology does
not match the academic terminology or something, and say that
the thing we poke into "way-size" is actually "set size", if we agree
that is what we're seeing here.
Let's see again: 16KB way-size * 8 ways = 128KB, that seems correct.
16KB way-size / 32B line-size = 512 sets, that also seems realistic.
128KB cache-size / 512 sets = 256B set-size. 256B / 32B = 8 ways,
so everything fits together.
I've sent a v2 version of the patch using the terminology and calculations
from this thread, and it yields the right numbers, for PB1176, I will check
with the values from other RealView platforms to see if they also give the
same result back.
I had to squash Florians & my patch, it didn't make sense to have them
artificially separate.
Yours,
Linus Walleij