On  27 Aug, this message from Adrian Cox echoed through cyberspace:

I'm currently trying out the pre-zeroing of pages in the idle task, and
I'm a little worried about its effect on cache-coherency. I can't prove
to
my satisfaction that none of the pages being cleared by zero_paged() is
in the cache, and I'm not sure that turning off caching in the PTE will
have any effect on cache lines already allocated in L1 or L2.

Cort Dougan had a paper about memory management in Linux on the PowerPC
online at NMT, but at this time the link is broken
(http://www.cs.nmt.edu/~cort/papers/linuxppc-mm/). If I remeber right,
he also spoke about zeroing pegaes in the idle task.

Cort, are you among us? What's up with your personal pages?

Obviously, any incoherency would be extremely rare, but ensuring that
the page is invalidated from the caches before it is zeroed has the
added bonus of removing the need for a flush_page_to_ram() in my
modified anonymous memory allocator.

I can't remeber whether Cort also treated this subject.

Michel

-------------------------------------------------------------------------
Michel Lanners                 |  " Read Philosophy.  Study Art.
23, Rue Paul Henkes            |    Ask Questions.  Make Mistakes.
L-1710 Luxembourg              |
email   mlan@cpu.lu            |
http://www.cpu.lu/~mlan        |                     Learn Always. "


[[ This message was sent via the linuxppc-dev mailing list.  Replies are ]]
[[ not  forced  back  to the list, so be sure to Cc linuxppc-dev if your ]]
[[ reply is of general interest. Please check http://lists.linuxppc.org/ ]]
[[ and http://www.linuxppc.org/ for useful information before posting.   ]]