Thanks to commit 4b3ef9daa4 ("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>
Replace #ifdef'fed/commented out debug printk statements with pr_debug.
Replace printk statements with pr_* equivalents.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Functions clear_user_highpage, copy_user_highpage, flush_dcache_page,
local_flush_cache_range and local_flush_cache_page may be used from
modules. Export them.
Cc: stable@vger.kernel.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Currently building kernel for xtensa core with aliasing WT cache fails
with the following messages:
mm/memory.c:2152: undefined reference to `flush_dcache_page'
mm/memory.c:2332: undefined reference to `local_flush_cache_page'
mm/memory.c:1919: undefined reference to `local_flush_cache_range'
mm/memory.c:4179: undefined reference to `copy_to_user_page'
mm/memory.c:4183: undefined reference to `copy_from_user_page'
This happens because implementation of these functions is only compiled
when data cache is WB, which looks wrong: even when data cache doesn't
need flushing it still needs invalidation. The functions like
__flush_[invalidate_]dcache_* are correctly defined for both WB and WT
caches (and even if they weren't that'd still be ok, just slower).
Fix this by providing the same implementation of the above functions for
both WB and WT cache.
Cc: stable@vger.kernel.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Disabling pagefault makes little sense there, preemption disabling is
what was meant.
Cc: stable@vger.kernel.org
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Use __flush_invalidate_dcache_page_alias with alias set to color of the
page physical address instead of __flush_invalidate_dcache_page: this
works for high memory pages and mapping/unmapping to the TLBTEMP area is
virtually free.
Allow building configurations with aliasing cache and highmem enabled.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Existing clear_user_page and copy_user_page cannot be used with highmem
because they calculate physical page address from its virtual address
and do it incorrectly in case of high memory page mapped with
kmap_atomic. Also kmap is not needed, as most likely userspace mapping
color would be different from the kmapped color.
Provide clear_user_highpage and copy_user_highpage functions that
determine if temporary mapping is needed for the pages. Move most of the
logic of the former clear_user_page and copy_user_page to
xtensa/mm/cache.c only leaving temporary mapping setup, invalidation and
clearing/copying in the xtensa/mm/misc.S. Rename these functions to
clear_page_alias and copy_page_alias.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Introduce fixmap area just below the vmalloc region. Use it for atomic
mapping of high memory pages.
High memory on cores with cache aliasing is not supported and is still
to be implemented. Fail build for such configurations for now.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
This is largely based on SMP code from the xtensa-2.6.29-smp tree by
Piet Delaney, Marc Gauthier, Joe Taylor, Christian Zankel (and possibly
other Tensilica folks).
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Chris Zankel <chris@zankel.net>
Remove heading and trailing spaces, trim trailing lines, and wrap lines
that are longer than 80 characters.
Signed-off-by: Chris Zankel <chris@zankel.net>
On VIVT ARM, when we have multiple shared mappings of the same file
in the same MM, we need to ensure that we have coherency across all
copies. We do this via make_coherent() by making the pages
uncacheable.
This used to work fine, until we allowed highmem with highpte - we
now have a page table which is mapped as required, and is not available
for modification via update_mmu_cache().
Ralf Beache suggested getting rid of the PTE value passed to
update_mmu_cache():
On MIPS update_mmu_cache() calls __update_tlb() which walks pagetables
to construct a pointer to the pte again. Passing a pte_t * is much
more elegant. Maybe we might even replace the pte argument with the
pte_t?
Ben Herrenschmidt would also like the pte pointer for PowerPC:
Passing the ptep in there is exactly what I want. I want that
-instead- of the PTE value, because I have issue on some ppc cases,
for I$/D$ coherency, where set_pte_at() may decide to mask out the
_PAGE_EXEC.
So, pass in the mapped page table pointer into update_mmu_cache(), and
remove the PTE value, updating all implementations and call sites to
suit.
Includes a fix from Stephen Rothwell:
sparc: fix fallout from update_mmu_cache API change
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
The TLB entry for the user address doesn't exist at the time we
want to flush the caches, so use the page address. Note that processor
configurations with cache-aliasing issues are treated separately.
Signed-off-by: Chris Zankel <chris@zankel.net>
Signed-off-by: Lucas Woods <woodzy@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Christian Zankel <chris@zankel.net>
Add support for processors that have cache-aliasing issues, such as
the Stretch S5000 processor. Cache-aliasing means that the size of
the cache (for one way) is larger than the page size, thus, a page
can end up in several places in cache depending on the virtual to
physical translation. The method used here is to map a user page
temporarily through the auto-refill way 0 and of of the DTLB.
We probably will want to revisit this issue and use a better
approach with kmap/kunmap.
Signed-off-by: Chris Zankel <chris@zankel.net>
Huang Ying
Max Filippov
Chris Zankel
Russell King