Le 02/04/2019 à 17:42, Aneesh Kumar K.V a écrit :
On 4/2/19 9:06 PM, Christophe Leroy wrote:
Le 02/04/2019 à 16:34, Aneesh Kumar K.V a écrit :
Currently, our mm_context_t on book3s64 include all hash specific
context details like slice mask, subpage protection details. We
can skip allocating those on radix. This will help us to save
8K per mm_context with radix translation.
With the patch applied we have
sizeof(mm_context_t) = 136
sizeof(struct hash_mm_context) = 8288
Signed-off-by: Aneesh Kumar K.V <[email protected]>
---
NOTE:
If we want to do this, I am still trying to figure out how best we
can do this
without all the #ifdef and other overhead for 8xx book3e
Did you have a look at my series
https://patchwork.ozlabs.org/project/linuxppc-dev/list/?series=98170 ?
It tries to reduce as much as feasible the #ifdefs and stuff.
Not yet. But a cursory look tell me introducing hash_mm_context
complicates this further unless I introduce something similar for nohash
32? Are you ok with that?
Have a look at my review in the other mail, I think we can limit the
changes and avoid introducing the hash_mm_context for 8xx.
Otherwise, we should call it something else, for instance
extended_mm_context, but that looks unnecessary from my point of view.
Christophe