diff options
| author | Benjamin Herrenschmidt <benh@kernel.crashing.org> | 2005-09-27 21:45:45 -0700 |
|---|---|---|
| committer | Linus Torvalds <torvalds@g5.osdl.org> | 2005-09-28 07:46:42 -0700 |
| commit | 0f9578b70a9f112bfb541e1d5ab486a376e64503 (patch) | |
| tree | 00e93df9f9920c43ace34e28298255dc8a0f9263 /arch/ppc64/mm/tlb.c | |
| parent | 485ef69edefd7fc7f351c94d0d77b3ed8a242f7b (diff) | |
[PATCH] ppc64: More hugepage fixes
My previous patch fixing invalidation of huge PTEs wasn't good enough, we
still had an issue if a PTE invalidation batch contained both small and
large pages. This patch fixes this by making sure the batch is flushed if
the page size fed to it changes.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Diffstat (limited to 'arch/ppc64/mm/tlb.c')
| -rw-r--r-- | arch/ppc64/mm/tlb.c | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/arch/ppc64/mm/tlb.c b/arch/ppc64/mm/tlb.c index d8a6593a13f..21fbffb23a4 100644 --- a/arch/ppc64/mm/tlb.c +++ b/arch/ppc64/mm/tlb.c @@ -143,7 +143,8 @@ void hpte_update(struct mm_struct *mm, unsigned long addr, * up scanning and resetting referenced bits then our batch context * will change mid stream. */ - if (unlikely(i != 0 && context != batch->context)) { + if (i != 0 && (context != batch->context || + batch->large != pte_huge(pte))) { flush_tlb_pending(); i = 0; } @@ -151,6 +152,7 @@ void hpte_update(struct mm_struct *mm, unsigned long addr, if (i == 0) { batch->context = context; batch->mm = mm; + batch->large = pte_huge(pte); } batch->pte[i] = __pte(pte); batch->addr[i] = addr; |