Re: [PATCH] Flushing cache at exit_mmap

From: David S. Miller (davem@redhat.com)
Date: Wed Jul 26 2000 - 23:15:19 EST


   From: NIIBE Yutaka <gniibe@chroot.org>
   Date: Fri, 21 Jul 2000 12:57:26 +0900

   SH-4 has virtually indexed and physically tagged cache, which has
   alias problem to be handled by kernel. When process exits, we need
   to flush the caches used by the process. If we don't do flush
   here, something wrong will be happened later, because of cache
   aliases, when the pages will be reused. The page will be polluted
   by old cache data in unrelated way.

Can you show me a call path to exit_mmap which does not satisfy one of
the following two cases?

1) Some function higher in the call path to exit_mmap performs
   a flush_cache_mm/flush_tlb_mm sequence.

2) It is known that the mm whose address space is being torn down by
   exit_mmap has never executed ever on a cpu (ie. how kernel/fork.c
   invokes mmput() for error recovery).

If you cannot, I really prefer that this change not go in, because it
means that you honestly have a bug somewhere else, perhaps in your
flush_cache_mm implementation.

Later,
David S. Miller
davem@redhat.com

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Mon Jul 31 2000 - 21:00:23 EST