x86: introduce max_low_pfn_mapped for 64-bit

when more than 4g memory is installed, don't map the big hole below 4g.

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

4 files changed, 21 insertions, 3 deletions

diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c
index b5a0fd5f4c5..029e8cffca9 100644
--- a/arch/x86/mm/init_32.c
+++ b/arch/x86/mm/init_32.c
@@ -50,6 +50,7 @@
 
 unsigned int __VMALLOC_RESERVE = 128 << 20;
 
+unsigned long max_low_pfn_mapped;
 unsigned long max_pfn_mapped;
 
 DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
index 48548ef7ddf..122bcef222f 100644
--- a/arch/x86/mm/init_64.c
+++ b/arch/x86/mm/init_64.c
@@ -53,6 +53,7 @@
  * The direct mapping extends to max_pfn_mapped, so that we can directly access
  * apertures, ACPI and other tables without having to play with fixmaps.
  */
+unsigned long max_low_pfn_mapped;
 unsigned long max_pfn_mapped;
 
 static unsigned long dma_reserve __initdata;
diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c
index afd40054d15..0389cb8f6b1 100644
--- a/arch/x86/mm/pageattr.c
+++ b/arch/x86/mm/pageattr.c
@@ -536,8 +536,14 @@ static int split_large_page(pte_t *kpte, unsigned long address)
                 set_pte(&pbase[i], pfn_pte(pfn, ref_prot));
 
         if (address >= (unsigned long)__va(0) &&
+                address < (unsigned long)__va(max_low_pfn_mapped << PAGE_SHIFT))
+                split_page_count(level);
+
+#ifdef CONFIG_X86_64
+        if (address >= (unsigned long)__va(1UL<<32) &&
                 address < (unsigned long)__va(max_pfn_mapped << PAGE_SHIFT))
                 split_page_count(level);
+#endif
 
         /*
          * Install the new, split up pagetable. Important details here:
@@ -655,12 +661,21 @@ static int cpa_process_alias(struct cpa_data *cpa)
         if (cpa->pfn > max_pfn_mapped)
                 return 0;
 
+#ifdef CONFIG_X86_64
+        if (cpa->pfn > max_low_pfn_mapped && cpa->pfn < (1UL<<(32-PAGE_SHIFT)))
+                return 0;
+#endif
         /*
          * No need to redo, when the primary call touched the direct
          * mapping already:
          */
-        if (!within(cpa->vaddr, PAGE_OFFSET,
-                    PAGE_OFFSET + (max_pfn_mapped << PAGE_SHIFT))) {
+        if (!(within(cpa->vaddr, PAGE_OFFSET,
+                    PAGE_OFFSET + (max_low_pfn_mapped << PAGE_SHIFT))
+#ifdef CONFIG_X86_64
+                || within(cpa->vaddr, PAGE_OFFSET + (1UL<<32),
+                    PAGE_OFFSET + (max_pfn_mapped << PAGE_SHIFT))
+#endif
+        )) {
 
                 alias_cpa = *cpa;
                 alias_cpa.vaddr = (unsigned long) __va(cpa->pfn << PAGE_SHIFT);
diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c
index a885a1019b8..749766c3c5c 100644
--- a/arch/x86/mm/pat.c
+++ b/arch/x86/mm/pat.c
@@ -449,7 +449,8 @@ int phys_mem_access_prot_allowed(struct file *file, unsigned long pfn,
         if (retval < 0)
                 return 0;
 
-        if (pfn <= max_pfn_mapped &&
+        if (((pfn <= max_low_pfn_mapped) ||
+             (pfn >= (1UL<<(32 - PAGE_SHIFT)) && pfn <= max_pfn_mapped)) &&
             ioremap_change_attr((unsigned long)__va(offset), size, flags) < 0) {
                 free_memtype(offset, offset + size);
                 printk(KERN_INFO