xen/mmu: Cleanup xen_pagetable_p2m_copy a bit.
authorKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Fri, 3 Jan 2014 19:08:39 +0000 (14:08 -0500)
committerKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Mon, 6 Jan 2014 15:44:04 +0000 (10:44 -0500)
Stefano noticed that the code runs only under 64-bit so
the comments about 32-bit are pointless.

Also we change the condition for xen_revector_p2m_tree
returning the same value (because it could not allocate
a swath of space to put the new P2M in) or it had been
called once already. In such we return early from the
function.

Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Acked-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
arch/x86/xen/mmu.c

index c140eff..9d74249 100644 (file)
@@ -1209,29 +1209,29 @@ static void __init xen_pagetable_p2m_copy(void)
 
        size = PAGE_ALIGN(xen_start_info->nr_pages * sizeof(unsigned long));
 
-       /* On 32-bit, we get zero so this never gets executed. */
        new_mfn_list = xen_revector_p2m_tree();
-       if (new_mfn_list && new_mfn_list != xen_start_info->mfn_list) {
-               /* using __ka address and sticking INVALID_P2M_ENTRY! */
-               memset((void *)xen_start_info->mfn_list, 0xff, size);
-
-               /* We should be in __ka space. */
-               BUG_ON(xen_start_info->mfn_list < __START_KERNEL_map);
-               addr = xen_start_info->mfn_list;
-               /* We roundup to the PMD, which means that if anybody at this stage is
-                * using the __ka address of xen_start_info or xen_start_info->shared_info
-                * they are in going to crash. Fortunatly we have already revectored
-                * in xen_setup_kernel_pagetable and in xen_setup_shared_info. */
-               size = roundup(size, PMD_SIZE);
-               xen_cleanhighmap(addr, addr + size);
-
-               size = PAGE_ALIGN(xen_start_info->nr_pages * sizeof(unsigned long));
-               memblock_free(__pa(xen_start_info->mfn_list), size);
-               /* And revector! Bye bye old array */
-               xen_start_info->mfn_list = new_mfn_list;
-       } else
+       /* No memory or already called. */
+       if (!new_mfn_list || new_mfn_list == xen_start_info->mfn_list)
                return;
 
+       /* using __ka address and sticking INVALID_P2M_ENTRY! */
+       memset((void *)xen_start_info->mfn_list, 0xff, size);
+
+       /* We should be in __ka space. */
+       BUG_ON(xen_start_info->mfn_list < __START_KERNEL_map);
+       addr = xen_start_info->mfn_list;
+       /* We roundup to the PMD, which means that if anybody at this stage is
+        * using the __ka address of xen_start_info or xen_start_info->shared_info
+        * they are in going to crash. Fortunatly we have already revectored
+        * in xen_setup_kernel_pagetable and in xen_setup_shared_info. */
+       size = roundup(size, PMD_SIZE);
+       xen_cleanhighmap(addr, addr + size);
+
+       size = PAGE_ALIGN(xen_start_info->nr_pages * sizeof(unsigned long));
+       memblock_free(__pa(xen_start_info->mfn_list), size);
+       /* And revector! Bye bye old array */
+       xen_start_info->mfn_list = new_mfn_list;
+
        /* At this stage, cleanup_highmap has already cleaned __ka space
         * from _brk_limit way up to the max_pfn_mapped (which is the end of
         * the ramdisk). We continue on, erasing PMD entries that point to page