diff options
| author | David S. Miller <davem@davemloft.net> | 2016-03-08 12:34:12 -0500 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2016-03-08 12:34:12 -0500 |
| commit | 810813c47a564416f6306ae214e2661366c987a7 (patch) | |
| tree | f0d1f856d4b0024324f642fe519963248828b83f /arch/powerpc/mm/hugepage-hash64.c | |
| parent | d66ab51442211158b677c2f12310c314d9587f74 (diff) | |
| parent | e2857b8f11a289ed2b61d18d0665e05c1053c446 (diff) | |
Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
Several cases of overlapping changes, as well as one instance
(vxlan) of a bug fix in 'net' overlapping with code movement
in 'net-next'.
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'arch/powerpc/mm/hugepage-hash64.c')
| -rw-r--r-- | arch/powerpc/mm/hugepage-hash64.c | 12 |
1 files changed, 11 insertions, 1 deletions
diff --git a/arch/powerpc/mm/hugepage-hash64.c b/arch/powerpc/mm/hugepage-hash64.c index 49b152b0f926..eb2accdd76fd 100644 --- a/arch/powerpc/mm/hugepage-hash64.c +++ b/arch/powerpc/mm/hugepage-hash64.c @@ -78,9 +78,19 @@ int __hash_page_thp(unsigned long ea, unsigned long access, unsigned long vsid, * base page size. This is because demote_segment won't flush * hash page table entries. */ - if ((old_pmd & _PAGE_HASHPTE) && !(old_pmd & _PAGE_COMBO)) + if ((old_pmd & _PAGE_HASHPTE) && !(old_pmd & _PAGE_COMBO)) { flush_hash_hugepage(vsid, ea, pmdp, MMU_PAGE_64K, ssize, flags); + /* + * With THP, we also clear the slot information with + * respect to all the 64K hash pte mapping the 16MB + * page. They are all invalid now. This make sure we + * don't find the slot valid when we fault with 4k + * base page size. + * + */ + memset(hpte_slot_array, 0, PTE_FRAG_SIZE); + } } valid = hpte_valid(hpte_slot_array, index); |