[PATCH] mm: arches skip ptlock

Convert those few architectures which are calling pud_alloc, pmd_alloc,
pte_alloc_map on a user mm, not to take the page_table_lock first, nor drop it
after.  Each of these can continue to use pte_alloc_map, no need to change
over to pte_alloc_map_lock, they're neither racy nor swappable.

In the sparc64 io_remap_pfn_range, flush_tlb_range then falls outside of the
page_table_lock: that's okay, on sparc64 it's like flush_tlb_mm, and that has
always been called from outside of page_table_lock in dup_mmap.

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

1 files changed, 0 insertions, 15 deletions

diff --git a/arch/arm26/mm/memc.c b/arch/arm26/mm/memc.c
index d6b008b8db76..34def6397c3c 100644
--- a/arch/arm26/mm/memc.c
+++ b/arch/arm26/mm/memc.c
@@ -79,12 +79,6 @@ pgd_t *get_pgd_slow(struct mm_struct *mm)
                 goto no_pgd;
 
         /*
-         * This lock is here just to satisfy pmd_alloc and pte_lock
-         * FIXME: I bet we could avoid taking it pretty much altogether
-         */
-        spin_lock(&mm->page_table_lock);
-
-        /*
          * On ARM, first page must always be allocated since it contains
          * the machine vectors.
          */
@@ -113,23 +107,14 @@ pgd_t *get_pgd_slow(struct mm_struct *mm)
         memcpy(new_pgd + FIRST_KERNEL_PGD_NR, init_pgd + FIRST_KERNEL_PGD_NR,
                 (PTRS_PER_PGD - FIRST_KERNEL_PGD_NR) * sizeof(pgd_t));
 
-        spin_unlock(&mm->page_table_lock);
-
         /* update MEMC tables */
         cpu_memc_update_all(new_pgd);
         return new_pgd;
 
 no_pte:
-        spin_unlock(&mm->page_table_lock);
         pmd_free(new_pmd);
-        free_pgd_slow(new_pgd);
-        return NULL;
-
 no_pmd:
-        spin_unlock(&mm->page_table_lock);
         free_pgd_slow(new_pgd);
-        return NULL;
-
 no_pgd:
         return NULL;
 }