[patch 24/52] [PATCH] s390: use change recording override for kernel mapping

From: Martin Schwidefsky
Date: Fri Nov 13 2009 - 10:20:14 EST


From: Christian Borntraeger <borntraeger@xxxxxxxxxx>

We dont need the dirty bit if a write access is done via the kernel
mapping. In that case SetPageDirty and friends are used anyway, no
need to do that a second time. We can use the change-recording
overide function for the kernel mapping, if available.

Signed-off-by: Christian Borntraeger <borntraeger@xxxxxxxxxx>
Signed-off-by: Martin Schwidefsky <schwidefsky@xxxxxxxxxx>
---

arch/s390/include/asm/pgtable.h | 4 +++-
arch/s390/mm/vmem.c | 11 ++++++++---
2 files changed, 11 insertions(+), 4 deletions(-)

Index: quilt-2.6/arch/s390/include/asm/pgtable.h
===================================================================
--- quilt-2.6.orig/arch/s390/include/asm/pgtable.h 2009-11-13 15:48:32.000000000 +0100
+++ quilt-2.6/arch/s390/include/asm/pgtable.h 2009-11-13 16:08:17.000000000 +0100
@@ -169,12 +169,13 @@
* STL Segment-Table-Length: Segment-table length (STL+1*16 entries -> up to 2048)
*
* A 64 bit pagetable entry of S390 has following format:
- * | PFRA |0IP0| OS |
+ * | PFRA |0IPC| OS |
* 0000000000111111111122222222223333333333444444444455555555556666
* 0123456789012345678901234567890123456789012345678901234567890123
*
* I Page-Invalid Bit: Page is not available for address-translation
* P Page-Protection Bit: Store access not possible for page
+ * C Change-bit override: HW is not required to set change bit
*
* A 64 bit segmenttable entry of S390 has following format:
* | P-table origin | TT
@@ -218,6 +219,7 @@
*/

/* Hardware bits in the page table entry */
+#define _PAGE_CO 0x100 /* HW Change-bit override */
#define _PAGE_RO 0x200 /* HW read-only bit */
#define _PAGE_INVALID 0x400 /* HW invalid bit */

Index: quilt-2.6/arch/s390/mm/vmem.c
===================================================================
--- quilt-2.6.orig/arch/s390/mm/vmem.c 2009-11-13 15:48:32.000000000 +0100
+++ quilt-2.6/arch/s390/mm/vmem.c 2009-11-13 16:08:17.000000000 +0100
@@ -70,8 +70,12 @@
pte = alloc_bootmem(PTRS_PER_PTE * sizeof(pte_t));
if (!pte)
return NULL;
- clear_table((unsigned long *) pte, _PAGE_TYPE_EMPTY,
- PTRS_PER_PTE * sizeof(pte_t));
+ if (MACHINE_HAS_HPAGE)
+ clear_table((unsigned long *) pte, _PAGE_TYPE_EMPTY | _PAGE_CO,
+ PTRS_PER_PTE * sizeof(pte_t));
+ else
+ clear_table((unsigned long *) pte, _PAGE_TYPE_EMPTY,
+ PTRS_PER_PTE * sizeof(pte_t));
return pte;
}

@@ -112,7 +116,8 @@
if (MACHINE_HAS_HPAGE && !(address & ~HPAGE_MASK) &&
(address + HPAGE_SIZE <= start + size) &&
(address >= HPAGE_SIZE)) {
- pte_val(pte) |= _SEGMENT_ENTRY_LARGE;
+ pte_val(pte) |= _SEGMENT_ENTRY_LARGE |
+ _SEGMENT_ENTRY_CO;
pmd_val(*pm_dir) = pte_val(pte);
address += HPAGE_SIZE - PAGE_SIZE;
continue;

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/