ARM: pgtable: switch order of Linux vs hardware page tables

This switches the ordering of the Linux vs hardware page tables in
each page, thereby eliminating some of the arithmetic in the page
table walks.  As we now place the Linux page table at the beginning
of the page, we can deal with the offset in the pgt by simply masking
it away, along with the other control bits.

This also makes the arithmetic all be positive, rather than a mixture.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

5 files changed, 41 insertions, 49 deletions

diff --git a/arch/arm/include/asm/pgalloc.h b/arch/arm/include/asm/pgalloc.h
index 1f1064b519c0..9763be04f77e 100644
--- a/arch/arm/include/asm/pgalloc.h
+++ b/arch/arm/include/asm/pgalloc.h
@@ -35,6 +35,11 @@ extern void pgd_free(struct mm_struct *mm, pgd_t *pgd);
 
 #define PGALLOC_GFP     (GFP_KERNEL | __GFP_NOTRACK | __GFP_REPEAT | __GFP_ZERO)
 
+static inline void clean_pte_table(pte_t *pte)
+{
+        clean_dcache_area(pte + PTE_HWTABLE_PTRS, PTE_HWTABLE_SIZE);
+}
+
 /*
  * Allocate one PTE table.
  *
@@ -42,14 +47,14 @@ extern void pgd_free(struct mm_struct *mm, pgd_t *pgd);
  * into one table thus:
  *
  *  +------------+
- *  |  h/w pt 0  |
- *  +------------+
- *  |  h/w pt 1  |
- *  +------------+
  *  | Linux pt 0 |
  *  +------------+
  *  | Linux pt 1 |
  *  +------------+
+ *  |  h/w pt 0  |
+ *  +------------+
+ *  |  h/w pt 1  |
+ *  +------------+
  */
 static inline pte_t *
 pte_alloc_one_kernel(struct mm_struct *mm, unsigned long addr)
@@ -57,10 +62,8 @@ pte_alloc_one_kernel(struct mm_struct *mm, unsigned long addr)
         pte_t *pte;
 
         pte = (pte_t *)__get_free_page(PGALLOC_GFP);
-        if (pte) {
-                clean_dcache_area(pte, sizeof(pte_t) * PTRS_PER_PTE);
-                pte += PTRS_PER_PTE;
-        }
+        if (pte)
+                clean_pte_table(pte);
 
         return pte;
 }
@@ -76,10 +79,8 @@ pte_alloc_one(struct mm_struct *mm, unsigned long addr)
         pte = alloc_pages(PGALLOC_GFP, 0);
 #endif
         if (pte) {
-                if (!PageHighMem(pte)) {
-                        void *page = page_address(pte);
-                        clean_dcache_area(page, sizeof(pte_t) * PTRS_PER_PTE);
-                }
+                if (!PageHighMem(pte))
+                        clean_pte_table(page_address(pte));
                 pgtable_page_ctor(pte);
         }
 
@@ -91,10 +92,8 @@ pte_alloc_one(struct mm_struct *mm, unsigned long addr)
  */
 static inline void pte_free_kernel(struct mm_struct *mm, pte_t *pte)
 {
-        if (pte) {
-                pte -= PTRS_PER_PTE;
+        if (pte)
                 free_page((unsigned long)pte);
-        }
 }
 
 static inline void pte_free(struct mm_struct *mm, pgtable_t pte)
@@ -106,7 +105,7 @@ static inline void pte_free(struct mm_struct *mm, pgtable_t pte)
 static inline void __pmd_populate(pmd_t *pmdp, phys_addr_t pte,
         unsigned long prot)
 {
-        unsigned long pmdval = pte | prot;
+        unsigned long pmdval = (pte + PTE_HWTABLE_OFF) | prot;
         pmdp[0] = __pmd(pmdval);
         pmdp[1] = __pmd(pmdval + 256 * sizeof(pte_t));
         flush_pmd_entry(pmdp);
@@ -121,14 +120,10 @@ static inline void __pmd_populate(pmd_t *pmdp, phys_addr_t pte,
 static inline void
 pmd_populate_kernel(struct mm_struct *mm, pmd_t *pmdp, pte_t *ptep)
 {
-        unsigned long pte_ptr = (unsigned long)ptep;
-
         /*
-         * The pmd must be loaded with the physical
-         * address of the PTE table
+         * The pmd must be loaded with the physical address of the PTE table
          */
-        pte_ptr -= PTRS_PER_PTE * sizeof(void *);
-        __pmd_populate(pmdp, __pa(pte_ptr), _PAGE_KERNEL_TABLE);
+        __pmd_populate(pmdp, __pa(ptep), _PAGE_KERNEL_TABLE);
 }
 
 static inline void
diff --git a/arch/arm/include/asm/pgtable.h b/arch/arm/include/asm/pgtable.h
index 50eb0b4278ec..e582214b00df 100644
--- a/arch/arm/include/asm/pgtable.h
+++ b/arch/arm/include/asm/pgtable.h
@@ -55,7 +55,7 @@
  * Therefore, we tweak the implementation slightly - we tell Linux that we
  * have 2048 entries in the first level, each of which is 8 bytes (iow, two
  * hardware pointers to the second level.)  The second level contains two
- * hardware PTE tables arranged contiguously, followed by Linux versions
+ * hardware PTE tables arranged contiguously, preceded by Linux versions
  * which contain the state information Linux needs.  We, therefore, end up
  * with 512 entries in the "PTE" level.
  *
@@ -63,15 +63,15 @@
  *
  *    pgd             pte
  * |        |
- * +--------+ +0
- * |        |-----> +------------+ +0
+ * +--------+
+ * |        |       +------------+ +0
+ * +- - - - +       | Linux pt 0 |
+ * |        |       +------------+ +1024
+ * +--------+ +0    | Linux pt 1 |
+ * |        |-----> +------------+ +2048
  * +- - - - + +4    |  h/w pt 0  |
- * |        |-----> +------------+ +1024
+ * |        |-----> +------------+ +3072
  * +--------+ +8    |  h/w pt 1  |
- * |        |       +------------+ +2048
- * +- - - - +       | Linux pt 0 |
- * |        |       +------------+ +3072
- * +--------+       | Linux pt 1 |
  * |        |       +------------+ +4096
  *
  * See L_PTE_xxx below for definitions of bits in the "Linux pt", and
@@ -103,6 +103,10 @@
 #define PTRS_PER_PMD            1
 #define PTRS_PER_PGD            2048
 
+#define PTE_HWTABLE_PTRS        (PTRS_PER_PTE)
+#define PTE_HWTABLE_OFF         (PTE_HWTABLE_PTRS * sizeof(pte_t))
+#define PTE_HWTABLE_SIZE        (PTRS_PER_PTE * sizeof(u32))
+
 /*
  * PMD_SHIFT determines the size of the area a second-level page table can map
  * PGDIR_SHIFT determines what a third-level page table entry can map
@@ -323,12 +327,7 @@ extern pgd_t swapper_pg_dir[PTRS_PER_PGD];
 
 static inline pte_t *pmd_page_vaddr(pmd_t pmd)
 {
-        phys_addr_t ptr;
-
-        ptr = pmd_val(pmd) & ~(PTRS_PER_PTE * sizeof(void *) - 1);
-        ptr += PTRS_PER_PTE * sizeof(void *);
-
-        return __va(ptr);
+        return __va(pmd_val(pmd) & PAGE_MASK);
 }
 
 #define pmd_page(pmd)           pfn_to_page(__phys_to_pfn(pmd_val(pmd)))
@@ -341,8 +340,8 @@ static inline pte_t *pmd_page_vaddr(pmd_t pmd)
 #define __pte_map(pmd)          pmd_page_vaddr(*(pmd))
 #define __pte_unmap(pte)        do { } while (0)
 #else
-#define __pte_map(pmd)          ((pte_t *)kmap_atomic(pmd_page(*(pmd))) + PTRS_PER_PTE)
-#define __pte_unmap(pte)        kunmap_atomic((pte - PTRS_PER_PTE))
+#define __pte_map(pmd)          (pte_t *)kmap_atomic(pmd_page(*(pmd)))
+#define __pte_unmap(pte)        kunmap_atomic(pte)
 #endif
 
 #define pte_index(addr)         (((addr) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1))
diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c
index 1e21e125fe3a..f10f9bac2206 100644
--- a/arch/arm/mm/fault.c
+++ b/arch/arm/mm/fault.c
@@ -108,7 +108,7 @@ void show_pte(struct mm_struct *mm, unsigned long addr)
 
                 pte = pte_offset_map(pmd, addr);
                 printk(", *pte=%08lx", pte_val(*pte));
-                printk(", *ppte=%08lx", pte_val(pte[-PTRS_PER_PTE]));
+                printk(", *ppte=%08lx", pte_val(pte[PTE_HWTABLE_PTRS]));
                 pte_unmap(pte);
         } while(0);
 
diff --git a/arch/arm/mm/proc-macros.S b/arch/arm/mm/proc-macros.S
index 7d63beaf9745..cbedf9c46b9d 100644
--- a/arch/arm/mm/proc-macros.S
+++ b/arch/arm/mm/proc-macros.S
@@ -121,7 +121,7 @@
         .endm
 
         .macro  armv6_set_pte_ext pfx
-        str     r1, [r0], #-2048                @ linux version
+        str     r1, [r0], #2048                 @ linux version
 
         bic     r3, r1, #0x000003fc
         bic     r3, r3, #PTE_TYPE_MASK
@@ -170,7 +170,7 @@
  *  1111  0xff  r/w     r/w
  */
         .macro  armv3_set_pte_ext wc_disable=1
-        str     r1, [r0], #-2048                @ linux version
+        str     r1, [r0], #2048                 @ linux version
 
         eor     r3, r1, #L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_WRITE | L_PTE_DIRTY
 
@@ -193,7 +193,7 @@
         bicne   r2, r2, #PTE_BUFFERABLE
 #endif
         .endif
-        str     r2, [r0]                        @ hardware version
+        str     r2, [r0]                @ hardware version
         .endm
 
 
@@ -213,7 +213,7 @@
  *  1111  11    r/w     r/w
  */
         .macro  xscale_set_pte_ext_prologue
-        str     r1, [r0], #-2048                @ linux version
+        str     r1, [r0]                        @ linux version
 
         eor     r3, r1, #L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_WRITE | L_PTE_DIRTY
 
@@ -232,7 +232,7 @@
         tst     r3, #L_PTE_PRESENT | L_PTE_YOUNG        @ present and young?
         movne   r2, #0                          @ no -> fault
 
-        str     r2, [r0]                        @ hardware version
+        str     r2, [r0, #2048]!                @ hardware version
         mov     ip, #0
         mcr     p15, 0, r0, c7, c10, 1          @ clean L1 D line
         mcr     p15, 0, ip, c7, c10, 4          @ data write barrier
diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S
index 53cbe2225153..89c31a6dae5c 100644
--- a/arch/arm/mm/proc-v7.S
+++ b/arch/arm/mm/proc-v7.S
@@ -124,15 +124,13 @@ ENDPROC(cpu_v7_switch_mm)
  *      Set a level 2 translation table entry.
  *
  *      - ptep  - pointer to level 2 translation table entry
- *                (hardware version is stored at -1024 bytes)
+ *                (hardware version is stored at +2048 bytes)
  *      - pte   - PTE value to store
  *      - ext   - value for extended PTE bits
  */
 ENTRY(cpu_v7_set_pte_ext)
 #ifdef CONFIG_MMU
- ARM(   str     r1, [r0], #-2048        )       @ linux version
- THUMB( str     r1, [r0]                )       @ linux version
- THUMB( sub     r0, r0, #2048           )
+        str     r1, [r0]                        @ linux version
 
         bic     r3, r1, #0x000003f0
         bic     r3, r3, #PTE_TYPE_MASK
@@ -158,7 +156,7 @@ ENTRY(cpu_v7_set_pte_ext)
         tstne   r1, #L_PTE_PRESENT
         moveq   r3, #0
 
-        str     r3, [r0]
+        str     r3, [r0, #2048]!
         mcr     p15, 0, r0, c7, c10, 1          @ flush_pte
 #endif
         mov     pc, lr