CONFIG_HIGHPTE vs. sub-page page tables.

Background: I've implemented 1K/2K page tables for s390.  These sub-page
page tables are required to properly support the s390 virtualization
instruction with KVM.  The SIE instruction requires that the page tables
have 256 page table entries (pte) followed by 256 page status table entries
(pgste).  The pgstes are only required if the process is using the SIE
instruction.  The pgstes are updated by the hardware and by the hypervisor
for a number of reasons, one of them is dirty and reference bit tracking.
To avoid wasting memory the standard pte table allocation should return
1K/2K (31/64 bit) and 2K/4K if the process is using SIE.

Problem: Page size on s390 is 4K, page table size is 1K or 2K.  That means
the s390 version for pte_alloc_one cannot return a pointer to a struct
page.  Trouble is that with the CONFIG_HIGHPTE feature on x86 pte_alloc_one
cannot return a pointer to a pte either, since that would require more than
32 bit for the return value of pte_alloc_one (and the pte * would not be
accessible since its not kmapped).

Solution: The only solution I found to this dilemma is a new typedef: a
pgtable_t.  For s390 pgtable_t will be a (pte *) - to be introduced with a
later patch.  For everybody else it will be a (struct page *).  The
additional problem with the initialization of the ptl lock and the
NR_PAGETABLE accounting is solved with a constructor pgtable_page_ctor and
a destructor pgtable_page_dtor.  The page table allocation and free
functions need to call these two whenever a page table page is allocated or
freed.  pmd_populate will get a pgtable_t instead of a struct page pointer.
 To get the pgtable_t back from a pmd entry that has been installed with
pmd_populate a new function pmd_pgtable is added.  It replaces the pmd_page
call in free_pte_range and apply_to_pte_range.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: <linux-arch@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 files changed, 45 insertions, 21 deletions

diff --git a/arch/frv/mm/pgalloc.c b/arch/frv/mm/pgalloc.c
index 1a2e5c8d03a9..66f616fb4860 100644
--- a/arch/frv/mm/pgalloc.c
+++ b/arch/frv/mm/pgalloc.c
@@ -28,7 +28,7 @@ pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
         return pte;
 }
 
-struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
+pgtable_t pte_alloc_one(struct mm_struct *mm, unsigned long address)
 {
         struct page *page;
 
@@ -37,9 +37,11 @@ struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
 #else
         page = alloc_pages(GFP_KERNEL|__GFP_REPEAT, 0);
 #endif
-        if (page)
+        if (page) {
                 clear_highpage(page);
-        flush_dcache_page(page);
+                pgtable_page_ctor(page);
+                flush_dcache_page(page);
+        }
         return page;
 }
 
diff --git a/arch/powerpc/mm/pgtable_32.c b/arch/powerpc/mm/pgtable_32.c
index f80f90c4d58b..ac3390f81900 100644
--- a/arch/powerpc/mm/pgtable_32.c
+++ b/arch/powerpc/mm/pgtable_32.c
@@ -107,19 +107,20 @@ __init_refok pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long add
         return pte;
 }
 
-struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
+pgtable_t pte_alloc_one(struct mm_struct *mm, unsigned long address)
 {
         struct page *ptepage;
 
 #ifdef CONFIG_HIGHPTE
-        gfp_t flags = GFP_KERNEL | __GFP_HIGHMEM | __GFP_REPEAT;
+        gfp_t flags = GFP_KERNEL | __GFP_HIGHMEM | __GFP_REPEAT | __GFP_ZERO;
 #else
-        gfp_t flags = GFP_KERNEL | __GFP_REPEAT;
+        gfp_t flags = GFP_KERNEL | __GFP_REPEAT | __GFP_ZERO;
 #endif
 
         ptepage = alloc_pages(flags, 0);
-        if (ptepage)
-                clear_highpage(ptepage);
+        if (!ptepage)
+                return NULL;
+        pgtable_page_ctor(ptepage);
         return ptepage;
 }
 
@@ -131,11 +132,12 @@ void pte_free_kernel(struct mm_struct *mm, pte_t *pte)
         free_page((unsigned long)pte);
 }
 
-void pte_free(struct mm_struct *mm, struct page *ptepage)
+void pte_free(struct mm_struct *mm, pgtable_t ptepage)
 {
 #ifdef CONFIG_SMP
         hash_page_sync();
 #endif
+        pgtable_page_dtor(ptepage);
         __free_page(ptepage);
 }
 
diff --git a/arch/ppc/mm/pgtable.c b/arch/ppc/mm/pgtable.c
index 409fcaa4994a..03a79bff1271 100644
--- a/arch/ppc/mm/pgtable.c
+++ b/arch/ppc/mm/pgtable.c
@@ -95,7 +95,7 @@ __init_refok pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long add
         return pte;
 }
 
-struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
+pgtable_t pte_alloc_one(struct mm_struct *mm, unsigned long address)
 {
         struct page *ptepage;
 
@@ -106,8 +106,10 @@ struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
 #endif
 
         ptepage = alloc_pages(flags, 0);
-        if (ptepage)
+        if (ptepage) {
                 clear_highpage(ptepage);
+                pgtable_page_ctor(ptepage);
+        }
         return ptepage;
 }
 
@@ -119,11 +121,12 @@ void pte_free_kernel(struct mm_struct *mm, pte_t *pte)
         free_page((unsigned long)pte);
 }
 
-void pte_free(struct mm_struct *mm, struct page *ptepage)
+void pte_free(struct mm_struct *mm, pgtable_t ptepage)
 {
 #ifdef CONFIG_SMP
         hash_page_sync();
 #endif
+        pgtable_page_dtor(ptepage);
         __free_page(ptepage);
 }
 
diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c
index e60e0ae13402..019f518cd5a0 100644
--- a/arch/s390/mm/pgtable.c
+++ b/arch/s390/mm/pgtable.c
@@ -78,6 +78,7 @@ unsigned long *page_table_alloc(int noexec)
                 clear_table(table, _PAGE_TYPE_EMPTY, PAGE_SIZE);
                 page->index = (addr_t) table;
         }
+        pgtable_page_ctor(page);
         table = (unsigned long *) page_to_phys(page);
         clear_table(table, _PAGE_TYPE_EMPTY, PAGE_SIZE);
         return table;
@@ -87,6 +88,7 @@ void page_table_free(unsigned long *table)
 {
         unsigned long *shadow = get_shadow_pte(table);
 
+        pgtable_page_dtor(virt_to_page(table));
         if (shadow)
                 free_page((unsigned long) shadow);
         free_page((unsigned long) table);
diff --git a/arch/sparc/mm/srmmu.c b/arch/sparc/mm/srmmu.c
index dc98e3844a0a..23d3291a3e81 100644
--- a/arch/sparc/mm/srmmu.c
+++ b/arch/sparc/mm/srmmu.c
@@ -489,14 +489,17 @@ srmmu_pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
         return (pte_t *)srmmu_get_nocache(PTE_SIZE, PTE_SIZE);
 }
 
-static struct page *
+static pgtable_t
 srmmu_pte_alloc_one(struct mm_struct *mm, unsigned long address)
 {
         unsigned long pte;
+        struct page *page;
 
         if ((pte = (unsigned long)srmmu_pte_alloc_one_kernel(mm, address)) == 0)
                 return NULL;
-        return pfn_to_page( __nocache_pa(pte) >> PAGE_SHIFT );
+        page = pfn_to_page( __nocache_pa(pte) >> PAGE_SHIFT );
+        pgtable_page_ctor(page);
+        return page;
 }
 
 static void srmmu_free_pte_fast(pte_t *pte)
@@ -504,10 +507,11 @@ static void srmmu_free_pte_fast(pte_t *pte)
         srmmu_free_nocache((unsigned long)pte, PTE_SIZE);
 }
 
-static void srmmu_pte_free(struct page *pte)
+static void srmmu_pte_free(pgtable_t pte)
 {
         unsigned long p;
 
+        pgtable_page_dtor(pte);
         p = (unsigned long)page_address(pte);   /* Cached address (for test) */
         if (p == 0)
                 BUG();
diff --git a/arch/sparc/mm/sun4c.c b/arch/sparc/mm/sun4c.c
index 0729305f2f59..c0442e8c4b15 100644
--- a/arch/sparc/mm/sun4c.c
+++ b/arch/sparc/mm/sun4c.c
@@ -1947,12 +1947,17 @@ static pte_t *sun4c_pte_alloc_one_kernel(struct mm_struct *mm, unsigned long add
         return pte;
 }
 
-static struct page *sun4c_pte_alloc_one(struct mm_struct *mm, unsigned long address)
+static pgtable_t sun4c_pte_alloc_one(struct mm_struct *mm, unsigned long address)
 {
-        pte_t *pte = sun4c_pte_alloc_one_kernel(mm, address);
+        pte_t *pte;
+        struct page *page;
+
+        pte = sun4c_pte_alloc_one_kernel(mm, address);
         if (pte == NULL)
                 return NULL;
-        return virt_to_page(pte);
+        page = virt_to_page(pte);
+        pgtable_page_ctor(page);
+        return page;
 }
 
 static inline void sun4c_free_pte_fast(pte_t *pte)
@@ -1962,8 +1967,9 @@ static inline void sun4c_free_pte_fast(pte_t *pte)
         pgtable_cache_size++;
 }
 
-static void sun4c_pte_free(struct page *pte)
+static void sun4c_pte_free(pgtable_t pte)
 {
+        pgtable_page_dtor(pte);
         sun4c_free_pte_fast(page_address(pte));
 }
 
diff --git a/arch/um/kernel/mem.c b/arch/um/kernel/mem.c
index d872fdce1d7e..2627ce82e918 100644
--- a/arch/um/kernel/mem.c
+++ b/arch/um/kernel/mem.c
@@ -354,11 +354,13 @@ pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
         return pte;
 }
 
-struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
+pgtable_t pte_alloc_one(struct mm_struct *mm, unsigned long address)
 {
         struct page *pte;
 
         pte = alloc_page(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO);
+        if (pte)
+                pgtable_page_ctor(pte);
         return pte;
 }
 
diff --git a/arch/x86/mm/pgtable_32.c b/arch/x86/mm/pgtable_32.c
index 6c1914622a88..73aba7125203 100644
--- a/arch/x86/mm/pgtable_32.c
+++ b/arch/x86/mm/pgtable_32.c
@@ -183,7 +183,7 @@ pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
         return (pte_t *)__get_free_page(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO);
 }
 
-struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
+pgtable_t pte_alloc_one(struct mm_struct *mm, unsigned long address)
 {
         struct page *pte;
 
@@ -192,6 +192,8 @@ struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
 #else
         pte = alloc_pages(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO, 0);
 #endif
+        if (pte)
+                pgtable_page_ctor(pte);
         return pte;
 }
 
@@ -365,6 +367,7 @@ void check_pgt_cache(void)
 
 void __pte_free_tlb(struct mmu_gather *tlb, struct page *pte)
 {
+        pgtable_page_dtor(pte);
         paravirt_release_pt(page_to_pfn(pte));
         tlb_remove_page(tlb, pte);
 }