Merge bk://kernel.bkbits.net/davem/net-2.6

into ppc970.osdl.org:/home/torvalds/v2.6/linux

arch/ppc64/kernel/xics.c

@@ -620,7 +620,7 @@ static void xics_set_affinity(unsigned int virq, cpumask_t cpumask)
 	cpumask_t tmp = CPU_MASK_NONE;
 
 	irq = virt_irq_to_real(irq_offset_down(virq));
-	if (irq == XICS_IPI)
+	if (irq == XICS_IPI || irq == NO_IRQ)
 		return;
 
 	status = rtas_call(ibm_get_xive, 1, 3, (void *)&xics_status, irq);

drivers/ide/ide-disk.c

@@ -358,13 +358,6 @@ ide_startstop_t __ide_do_rw_disk (ide_drive_t *drive, struct request *rq, sector
 
 	nsectors.all		= (u16) rq->nr_sectors;
 
-	if (drive->using_tcq && idedisk_start_tag(drive, rq)) {
-		if (!ata_pending_commands(drive))
-			BUG();
-
-		return ide_started;
-	}
-
 	if (IDE_CONTROL_REG)
 		hwif->OUTB(drive->ctl, IDE_CONTROL_REG);
 
@@ -482,7 +475,7 @@ ide_startstop_t __ide_do_rw_disk (ide_drive_t *drive, struct request *rq, sector
 			   ((lba48) ? WIN_READ_EXT : WIN_READ));
 		ide_execute_command(drive, command, &read_intr, WAIT_CMD, NULL);
 		return ide_started;
-	} else if (rq_data_dir(rq) == WRITE) {
+	} else {
 		ide_startstop_t startstop;
 #ifdef CONFIG_BLK_DEV_IDE_TCQ
 		if (blk_rq_tagged(rq))
@@ -520,9 +513,6 @@ ide_startstop_t __ide_do_rw_disk (ide_drive_t *drive, struct request *rq, sector
 		}
 		return ide_started;
 	}
-	blk_dump_rq_flags(rq, "__ide_do_rw_disk - bad command");
-	ide_end_request(drive, 0, 0);
-	return ide_stopped;
 }
 EXPORT_SYMBOL_GPL(__ide_do_rw_disk);
 
@@ -539,26 +529,11 @@ static ide_startstop_t lba_48_rw_disk(ide_drive_t *, struct request *, unsigned
  */
 ide_startstop_t __ide_do_rw_disk (ide_drive_t *drive, struct request *rq, sector_t block)
 {
-	BUG_ON(drive->blocked);
-	if (!blk_fs_request(rq)) {
-		blk_dump_rq_flags(rq, "__ide_do_rw_disk - bad command");
-		ide_end_request(drive, 0, 0);
-		return ide_stopped;
-	}
-
 	/*
 	 * 268435455  == 137439 MB or 28bit limit
 	 *
 	 * need to add split taskfile operations based on 28bit threshold.
 	 */
-
-	if (drive->using_tcq && idedisk_start_tag(drive, rq)) {
-		if (!ata_pending_commands(drive))
-			BUG();
-
-		return ide_started;
-	}
-
 	if (drive->addressing == 1)		/* 48-bit LBA */
 		return lba_48_rw_disk(drive, rq, (unsigned long long) block);
 	if (drive->select.b.lba)		/* 28-bit LBA */
@@ -734,6 +709,21 @@ static ide_startstop_t ide_do_rw_disk (ide_drive_t *drive, struct request *rq, s
 {
 	ide_hwif_t *hwif = HWIF(drive);
 
+	BUG_ON(drive->blocked);
+
+	if (!blk_fs_request(rq)) {
+		blk_dump_rq_flags(rq, "ide_do_rw_disk - bad command");
+		ide_end_request(drive, 0, 0);
+		return ide_stopped;
+	}
+
+	if (drive->using_tcq && idedisk_start_tag(drive, rq)) {
+		if (!ata_pending_commands(drive))
+			BUG();
+
+		return ide_started;
+	}
+
 	if (hwif->rw_disk)
 		return hwif->rw_disk(drive, rq, block);
 	else

drivers/ide/ide-iops.c

@@ -767,7 +767,7 @@ int ide_driveid_update (ide_drive_t *drive)
 	SELECT_MASK(drive, 1);
 	if (IDE_CONTROL_REG)
 		hwif->OUTB(drive->ctl,IDE_CONTROL_REG);
-	ide_delay_50ms();
+	msleep(50);
 	hwif->OUTB(WIN_IDENTIFY, IDE_COMMAND_REG);
 	timeout = jiffies + WAIT_WORSTCASE;
 	do {
@@ -775,9 +775,9 @@ int ide_driveid_update (ide_drive_t *drive)
 			SELECT_MASK(drive, 0);
 			return 0;	/* drive timed-out */
 		}
-		ide_delay_50ms();	/* give drive a breather */
+		msleep(50);	/* give drive a breather */
 	} while (hwif->INB(IDE_ALTSTATUS_REG) & BUSY_STAT);
-	ide_delay_50ms();	/* wait for IRQ and DRQ_STAT */
+	msleep(50);	/* wait for IRQ and DRQ_STAT */
 	if (!OK_STAT(hwif->INB(IDE_STATUS_REG),DRQ_STAT,BAD_R_STAT)) {
 		SELECT_MASK(drive, 0);
 		printk("%s: CHECK for good STATUS\n", drive->name);
@@ -827,7 +827,7 @@ int ide_config_drive_speed (ide_drive_t *drive, u8 speed)
 	u8 stat;
 
 //	while (HWGROUP(drive)->busy)
-//		ide_delay_50ms();
+//		msleep(50);
 
 #ifdef CONFIG_BLK_DEV_IDEDMA
 	if (hwif->ide_dma_check)	 /* check if host supports DMA */

drivers/ide/ide-probe.c

@@ -283,9 +283,10 @@ static int actual_try_to_identify (ide_drive_t *drive, u8 cmd)
 	unsigned long timeout;
 	u8 s = 0, a = 0;
 
-	if (IDE_CONTROL_REG) {
 	/* take a deep breath */
-		ide_delay_50ms();
+	msleep(50);
+
+	if (IDE_CONTROL_REG) {
 		a = hwif->INB(IDE_ALTSTATUS_REG);
 		s = hwif->INB(IDE_STATUS_REG);
 		if ((a ^ s) & ~INDEX_STAT) {
@@ -297,10 +298,8 @@ static int actual_try_to_identify (ide_drive_t *drive, u8 cmd)
 			/* use non-intrusive polling */
 			hd_status = IDE_ALTSTATUS_REG;
 		}
-	} else {
-		ide_delay_50ms();
+	} else
 		hd_status = IDE_STATUS_REG;
-	}
 
 	/* set features register for atapi
 	 * identify command to be sure of reply
@@ -324,11 +323,11 @@ static int actual_try_to_identify (ide_drive_t *drive, u8 cmd)
 			return 1;
 		}
 		/* give drive a breather */
-		ide_delay_50ms();
+		msleep(50);
 	} while ((hwif->INB(hd_status)) & BUSY_STAT);
 
 	/* wait for IRQ and DRQ_STAT */
-	ide_delay_50ms();
+	msleep(50);
 	if (OK_STAT((hwif->INB(IDE_STATUS_REG)), DRQ_STAT, BAD_R_STAT)) {
 		unsigned long flags;
 
@@ -457,15 +456,15 @@ static int do_probe (ide_drive_t *drive, u8 cmd)
 	/* needed for some systems
 	 * (e.g. crw9624 as drive0 with disk as slave)
 	 */
-	ide_delay_50ms();
+	msleep(50);
 	SELECT_DRIVE(drive);
-	ide_delay_50ms();
+	msleep(50);
 	if (hwif->INB(IDE_SELECT_REG) != drive->select.all && !drive->present) {
 		if (drive->select.b.unit != 0) {
 			/* exit with drive0 selected */
 			SELECT_DRIVE(&hwif->drives[0]);
 			/* allow BUSY_STAT to assert & clear */
-			ide_delay_50ms();
+			msleep(50);
 		}
 		/* no i/f present: mmm.. this should be a 4 -ml */
 		return 3;
@@ -488,14 +487,14 @@ static int do_probe (ide_drive_t *drive, u8 cmd)
 			printk("%s: no response (status = 0x%02x), "
 				"resetting drive\n", drive->name,
 				hwif->INB(IDE_STATUS_REG));
-			ide_delay_50ms();
+			msleep(50);
 			hwif->OUTB(drive->select.all, IDE_SELECT_REG);
-			ide_delay_50ms();
+			msleep(50);
 			hwif->OUTB(WIN_SRST, IDE_COMMAND_REG);
 			timeout = jiffies;
 			while (((hwif->INB(IDE_STATUS_REG)) & BUSY_STAT) &&
 			       time_before(jiffies, timeout + WAIT_WORSTCASE))
-				ide_delay_50ms();
+				msleep(50);
 			rc = try_to_identify(drive, cmd);
 		}
 		if (rc == 1)
@@ -510,7 +509,7 @@ static int do_probe (ide_drive_t *drive, u8 cmd)
 	if (drive->select.b.unit != 0) {
 		/* exit with drive0 selected */
 		SELECT_DRIVE(&hwif->drives[0]);
-		ide_delay_50ms();
+		msleep(50);
 		/* ensure drive irq is clear */
 		(void) hwif->INB(IDE_STATUS_REG);
 	}
@@ -527,7 +526,7 @@ static void enable_nest (ide_drive_t *drive)
 
 	printk("%s: enabling %s -- ", hwif->name, drive->id->model);
 	SELECT_DRIVE(drive);
-	ide_delay_50ms();
+	msleep(50);
 	hwif->OUTB(EXABYTE_ENABLE_NEST, IDE_COMMAND_REG);
 	timeout = jiffies + WAIT_WORSTCASE;
 	do {
@@ -535,10 +534,10 @@ static void enable_nest (ide_drive_t *drive)
 			printk("failed (timeout)\n");
 			return;
 		}
-		ide_delay_50ms();
+		msleep(50);
 	} while ((hwif->INB(IDE_STATUS_REG)) & BUSY_STAT);
 
-	ide_delay_50ms();
+	msleep(50);
 
 	if (!OK_STAT((hwif->INB(IDE_STATUS_REG)), 0, BAD_STAT)) {
 		printk("failed (status = 0x%02x)\n", hwif->INB(IDE_STATUS_REG));
@@ -781,7 +780,7 @@ void probe_hwif (ide_hwif_t *hwif)
 		udelay(10);
 		hwif->OUTB(8, hwif->io_ports[IDE_CONTROL_OFFSET]);
 		do {
-			ide_delay_50ms();
+			msleep(50);
 			stat = hwif->INB(hwif->io_ports[IDE_STATUS_OFFSET]);
 		} while ((stat & BUSY_STAT) && time_after(timeout, jiffies));
 

drivers/ide/ide.c

@@ -1387,26 +1387,6 @@ void ide_add_generic_settings (ide_drive_t *drive)
 		ide_add_setting(drive,	"ide-scsi",		SETTING_RW,					-1,		HDIO_SET_IDE_SCSI,		TYPE_BYTE,	0,	1,				1,		1,		&drive->scsi,			ide_atapi_to_scsi);
 }
 
-/*
- * Delay for *at least* 50ms.  As we don't know how much time is left
- * until the next tick occurs, we wait an extra tick to be safe.
- * This is used only during the probing/polling for drives at boot time.
- *
- * However, its usefullness may be needed in other places, thus we export it now.
- * The future may change this to a millisecond setable delay.
- */
-void ide_delay_50ms (void)
-{
-#ifndef CONFIG_BLK_DEV_IDECS
-	mdelay(50);
-#else
-	__set_current_state(TASK_UNINTERRUPTIBLE);
-	schedule_timeout(1+HZ/20);
-#endif /* CONFIG_BLK_DEV_IDECS */
-}
-
-EXPORT_SYMBOL(ide_delay_50ms);
-
 int system_bus_clock (void)
 {
 	return((int) ((!system_bus_speed) ? ide_system_bus_speed() : system_bus_speed ));

drivers/ide/legacy/pdc4030.c

@@ -283,7 +283,7 @@ int __init detect_pdc4030(ide_hwif_t *hwif)
 	hwif->OUTB(0x14, IDE_SELECT_REG);
 	hwif->OUTB(PROMISE_EXTENDED_COMMAND, IDE_COMMAND_REG);
 
-	ide_delay_50ms();
+	msleep(50);
 
 	if (hwif->INB(IDE_ERROR_REG) == 'P' &&
 	    hwif->INB(IDE_NSECTOR_REG) == 'T' &&
@@ -756,12 +756,6 @@ static ide_startstop_t promise_rw_disk (ide_drive_t *drive, struct request *rq,
 
 	BUG_ON(rq->nr_sectors > 127);
 
-	if (!blk_fs_request(rq)) {
-		blk_dump_rq_flags(rq, "promise_rw_disk - bad command");
-		DRIVER(drive)->end_request(drive, 0, 0);
-		return ide_stopped;
-	}
-
 #ifdef DEBUG
 	printk(KERN_DEBUG "%s: %sing: LBAsect=%lu, sectors=%lu\n",
 			  drive->name, rq_data_dir(rq) ? "writ" : "read",

include/asm-generic/pgtable.h

@@ -2,22 +2,34 @@
 #define _ASM_GENERIC_PGTABLE_H
 
 #ifndef __HAVE_ARCH_PTEP_ESTABLISH
-
-#ifndef ptep_update_dirty_accessed
-#define ptep_update_dirty_accessed(__ptep, __entry, __dirty) set_pte(__ptep, __entry)
-#endif
-
 /*
  * Establish a new mapping:
  *  - flush the old one
  *  - update the page tables
  *  - inform the TLB about the new one
  *
- * We hold the mm semaphore for reading and vma->vm_mm->page_table_lock
+ * We hold the mm semaphore for reading and vma->vm_mm->page_table_lock.
+ *
+ * Note: the old pte is known to not be writable, so we don't need to
+ * worry about dirty bits etc getting lost.
+ */
+#define ptep_establish(__vma, __address, __ptep, __entry)		\
+do {				  					\
+	set_pte(__ptep, __entry);					\
+	flush_tlb_page(__vma, __address);				\
+} while (0)
+#endif
+
+#ifndef __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
+/*
+ * Largely same as above, but only sets the access flags (dirty,
+ * accessed, and writable). Furthermore, we know it always gets set
+ * to a "more permissive" setting, which allows most architectures
+ * to optimize this.
  */
-#define ptep_establish(__vma, __address, __ptep, __entry, __dirty)	\
+#define ptep_set_access_flags(__vma, __address, __ptep, __entry, __dirty) \
 do {				  					  \
-	ptep_update_dirty_accessed(__ptep, __entry, __dirty);		\
+	set_pte(__ptep, __entry);					  \
 	flush_tlb_page(__vma, __address);				  \
 } while (0)
 #endif

include/asm-i386/pgtable.h

@@ -325,9 +325,13 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
  * bit at the same time.
  */
 #define update_mmu_cache(vma,address,pte) do { } while (0)
-#define ptep_update_dirty_accessed(__ptep, __entry, __dirty)	\
+#define  __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
+#define ptep_set_access_flags(__vma, __address, __ptep, __entry, __dirty) \
 	do {								  \
-		if (__dirty) set_pte(__ptep, __entry);		\
+		if (__dirty) {						  \
+			(__ptep)->pte_low = (__entry).pte_low;	  	  \
+			flush_tlb_page(__vma, __address);		  \
+		}							  \
 	} while (0)
 
 /* Encode and de-code a swap entry */

include/asm-ppc/pgtable.h

@@ -548,6 +548,16 @@ static inline void ptep_mkdirty(pte_t *ptep)
 	pte_update(ptep, 0, _PAGE_DIRTY);
 }
 
+#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
+static inline void __ptep_set_access_flags(pte_t *ptep, pte_t entry, int dirty)
+{
+	unsigned long bits = pte_val(entry) &
+		(_PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_RW);
+	pte_update(ptep, 0, bits);
+}
+#define  ptep_set_access_flags(__vma, __address, __ptep, __entry, __dirty) \
+        __ptep_set_access_flags(__ptep, __entry, __dirty)
+
 /*
  * Macro to mark a page protection value as "uncacheable".
  */

include/asm-ppc64/pgtable.h

@@ -306,7 +306,10 @@ static inline unsigned long pte_update(pte_t *p, unsigned long clr)
 	return old;
 }
 
-/* PTE updating functions */
+/* PTE updating functions, this function puts the PTE in the
+ * batch, doesn't actually triggers the hash flush immediately,
+ * you need to call flush_tlb_pending() to do that.
+ */
 extern void hpte_update(pte_t *ptep, unsigned long pte, int wrprot);
 
 static inline int ptep_test_and_clear_young(pte_t *ptep)
@@ -318,7 +321,7 @@ static inline int ptep_test_and_clear_young(pte_t *ptep)
 	old = pte_update(ptep, _PAGE_ACCESSED);
 	if (old & _PAGE_HASHPTE) {
 		hpte_update(ptep, old, 0);
-		flush_tlb_pending();	/* XXX generic code doesn't flush */
+		flush_tlb_pending();
 	}
 	return (old & _PAGE_ACCESSED) != 0;
 }
@@ -396,11 +399,37 @@ static inline void pte_clear(pte_t * ptep)
  */
 static inline void set_pte(pte_t *ptep, pte_t pte)
 {
-	if (pte_present(*ptep))
+	if (pte_present(*ptep)) {
 		pte_clear(ptep);
+		flush_tlb_pending();
+	}
 	*ptep = __pte(pte_val(pte)) & ~_PAGE_HPTEFLAGS;
 }
 
+/* Set the dirty and/or accessed bits atomically in a linux PTE, this
+ * function doesn't need to flush the hash entry
+ */
+#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
+static inline void __ptep_set_access_flags(pte_t *ptep, pte_t entry, int dirty)
+{
+	unsigned long bits = pte_val(entry) &
+		(_PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_RW);
+	unsigned long old, tmp;
+
+	__asm__ __volatile__(
+	"1:	ldarx	%0,0,%4\n\
+		andi.	%1,%0,%6\n\
+		bne-	1b \n\
+		or	%0,%3,%0\n\
+		stdcx.	%0,0,%4\n\
+		bne-	1b"
+	:"=&r" (old), "=&r" (tmp), "=m" (*ptep)
+	:"r" (bits), "r" (ptep), "m" (ptep), "i" (_PAGE_BUSY)
+	:"cc");
+}
+#define  ptep_set_access_flags(__vma, __address, __ptep, __entry, __dirty) \
+        __ptep_set_access_flags(__ptep, __entry, __dirty)
+
 /*
  * Macro to mark a page protection value as "uncacheable".
  */

include/asm-s390/pgtable.h

@@ -581,7 +581,7 @@ static inline void ptep_mkdirty(pte_t *ptep)
 static inline void
 ptep_establish(struct vm_area_struct *vma, 
 	       unsigned long address, pte_t *ptep,
-	       pte_t entry, int dirty)
+	       pte_t entry)
 {
 	ptep_clear_flush(vma, address, ptep);
 	set_pte(ptep, entry);

include/linux/ide.h

@@ -1477,7 +1477,6 @@ int ide_taskfile_ioctl(ide_drive_t *, unsigned int, unsigned long);
 int ide_cmd_ioctl(ide_drive_t *, unsigned int, unsigned long);
 int ide_task_ioctl(ide_drive_t *, unsigned int, unsigned long);
 
-extern void ide_delay_50ms(void);
 extern int system_bus_clock(void);
 
 extern u8 ide_auto_reduce_xfer(ide_drive_t *);

mm/memory.c

@@ -1004,7 +1004,7 @@ static inline void break_cow(struct vm_area_struct * vma, struct page * new_page
 	flush_cache_page(vma, address);
 	entry = maybe_mkwrite(pte_mkdirty(mk_pte(new_page, vma->vm_page_prot)),
 			      vma);
-	ptep_establish(vma, address, page_table, entry, 1);
+	ptep_establish(vma, address, page_table, entry);
 	update_mmu_cache(vma, address, entry);
 }
 
@@ -1056,7 +1056,7 @@ static int do_wp_page(struct mm_struct *mm, struct vm_area_struct * vma,
 			flush_cache_page(vma, address);
 			entry = maybe_mkwrite(pte_mkyoung(pte_mkdirty(pte)),
 					      vma);
-			ptep_establish(vma, address, page_table, entry, 1);
+			ptep_set_access_flags(vma, address, page_table, entry, 1);
 			update_mmu_cache(vma, address, entry);
 			pte_unmap(page_table);
 			spin_unlock(&mm->page_table_lock);
@@ -1646,7 +1646,7 @@ static inline int handle_pte_fault(struct mm_struct *mm,
 		entry = pte_mkdirty(entry);
 	}
 	entry = pte_mkyoung(entry);
-	ptep_establish(vma, address, pte, entry, write_access);
+	ptep_set_access_flags(vma, address, pte, entry, write_access);
 	update_mmu_cache(vma, address, entry);
 	pte_unmap(pte);
 	spin_unlock(&mm->page_table_lock);