mm: fix races between swapoff and flush dcache
authorHuang Ying <ying.huang@intel.com>
Thu, 5 Apr 2018 23:24:39 +0000 (16:24 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Fri, 6 Apr 2018 04:36:26 +0000 (21:36 -0700)
commitcb9f753a3731f7fe16447bea45cb6f8e8bb432fb
treef9c9d1ec34be344b46c0a26675fbc7b8830933b5
parent1c0ff0f1bdeb183608ac9a345996d4d1ebfed632
mm: fix races between swapoff and flush dcache

Thanks to commit 4b3ef9daa4fc ("mm/swap: split swap cache into 64MB
trunks"), after swapoff the address_space associated with the swap
device will be freed.  So page_mapping() users which may touch the
address_space need some kind of mechanism to prevent the address_space
from being freed during accessing.

The dcache flushing functions (flush_dcache_page(), etc) in architecture
specific code may access the address_space of swap device for anonymous
pages in swap cache via page_mapping() function.  But in some cases
there are no mechanisms to prevent the swap device from being swapoff,
for example,

  CPU1 CPU2
  __get_user_pages() swapoff()
    flush_dcache_page()
      mapping = page_mapping()
        ...   exit_swap_address_space()
        ...     kvfree(spaces)
        mapping_mapped(mapping)

The address space may be accessed after being freed.

But from cachetlb.txt and Russell King, flush_dcache_page() only care
about file cache pages, for anonymous pages, flush_anon_page() should be
used.  The implementation of flush_dcache_page() in all architectures
follows this too.  They will check whether page_mapping() is NULL and
whether mapping_mapped() is true to determine whether to flush the
dcache immediately.  And they will use interval tree (mapping->i_mmap)
to find all user space mappings.  While mapping_mapped() and
mapping->i_mmap isn't used by anonymous pages in swap cache at all.

So, to fix the race between swapoff and flush dcache, __page_mapping()
is add to return the address_space for file cache pages and NULL
otherwise.  All page_mapping() invoking in flush dcache functions are
replaced with page_mapping_file().

[akpm@linux-foundation.org: simplify page_mapping_file(), per Mike]
Link: http://lkml.kernel.org/r/20180305083634.15174-1-ying.huang@intel.com
Signed-off-by: "Huang, Ying" <ying.huang@intel.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Chen Liqin <liqin.linux@gmail.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: "James E.J. Bottomley" <jejb@parisc-linux.org>
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Chris Zankel <chris@zankel.net>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Ley Foon Tan <lftan@altera.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Mike Rapoport <rppt@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
19 files changed:
arch/arc/mm/cache.c
arch/arm/mm/copypage-v4mc.c
arch/arm/mm/copypage-v6.c
arch/arm/mm/copypage-xscale.c
arch/arm/mm/fault-armv.c
arch/arm/mm/flush.c
arch/mips/mm/cache.c
arch/nios2/mm/cacheflush.c
arch/parisc/kernel/cache.c
arch/sh/mm/cache-sh4.c
arch/sh/mm/cache-sh7705.c
arch/sparc/kernel/smp_64.c
arch/sparc/mm/init_64.c
arch/sparc/mm/tlb.c
arch/unicore32/mm/flush.c
arch/unicore32/mm/mmu.c
arch/xtensa/mm/cache.c
include/linux/mm.h
mm/util.c