Merge

Documentation/md.txt

 Tools that manage md devices can be found at
-   http://www.<country>.kernel.org/pub/linux/daemons/raid/....
+   http://www.<country>.kernel.org/pub/linux/utils/raid/....
 
 
 
-You can boot (if you selected boot support in the configuration) with your md 
-device with the following kernel command lines:
+You can boot with your md device with the following kernel command
+lines:
 
 for old raid arrays without persistent superblocks:
   md=<md device no.>,<raid level>,<chunk size factor>,<fault level>,dev0,dev1,...,devn
@@ -34,3 +34,63 @@ A possible loadlin line (Harald Hoyer <HarryH@Royal.Net>)  looks like this:
 
 e:\loadlin\loadlin e:\zimage root=/dev/md0 md=0,0,4,0,/dev/hdb2,/dev/hdc3 ro
 
+-------------------------------
+The md driver can support a variety of different superblock formats.
+(It doesn't yet, but it can)
+
+The kernel does *NOT* autodetect which format superblock is being
+used. It must be told.
+
+Superblock format '0' is treated differently to others for legacy
+reasons.
+
+
+General Rules - apply for all superblock formats
+------------------------------------------------
+
+An array is 'created' by writing appropriate superblocks to all
+devices.
+It is 'assembled' by associating each of these devices with an
+particular md virtual device.  Once it is completely assembled, it can
+be accessed.
+
+An array should be created by a user-space tool.  This will write
+superblocks to all devices.  It will usually mark the array as
+'unclean', or with some devices missing so that the kernel md driver
+can create approrpriate redundancy (copying in raid1, parity
+calculation in raid4/5).
+
+When an array is assembled, it is first initialised with the
+SET_ARRAY_INFO ioctl.  This contains, in particular, a major and minor
+version number.  The major version number selects which superblock
+format is to be used.  The minor number might be used to tune handling
+of the format, such as suggesting where on each device to look for the
+superblock.
+
+Then each device is added using the ADD_NEW_DISK ioctl.  This
+provides, in particular, a major and minor number identifying the
+device to add.
+
+The array is started with the RUN_ARRAY ioctl.
+
+Once started, new devices can be added.  They should have an
+appropriate superblock written to them, and then passed be in with
+ADD_NEW_DISK.
+
+Devices that have failed or are not yet active can be detached from an
+array using HOT_REMOVE_DISK.
+
+
+Specific Rules that apply to format-0 super block arrays, and
+       arrays with no superblock (non-presistant).
+-------------------------------------------------------------
+
+An array can be 'created' by describing the array (level, chunksize
+etc) in a SET_ARRAY_INFO ioctl.  This must has major_version==0 and
+raid_disks != 0.
+Then uninitialised devices can be added with ADD_NEW_DISK.  The
+structure passed to ADD_NEW_DISK must specify the state of the device
+and it's role in the array.
+
+One started with RUN_ARRAY, uninitialised spares can be added with
+HOT_ADD_DISK.

arch/i386/kernel/cpu/centaur.c

@@ -248,6 +248,37 @@ static void __init winchip2_protect_mcr(void)
 }
 #endif
 
+static void __init init_c3(struct cpuinfo_x86 *c)
+{
+	u32  lo, hi;
+
+	/* Test for Centaur Extended Feature Flags presence */
+	if (cpuid_eax(0xC0000000) >= 0xC0000001) {
+		/* store Centaur Extended Feature Flags as
+		 * word 5 of the CPU capability bit array
+		 */
+		c->x86_capability[5] = cpuid_edx(0xC0000001);
+	}
+
+	switch (c->x86_model) {
+		case 6 ... 8:		/* Cyrix III family */
+			rdmsr (MSR_VIA_FCR, lo, hi);
+			lo |= (1<<1 | 1<<7);	/* Report CX8 & enable PGE */
+			wrmsr (MSR_VIA_FCR, lo, hi);
+
+			set_bit(X86_FEATURE_CX8, c->x86_capability);
+			set_bit(X86_FEATURE_3DNOW, c->x86_capability);
+
+			/* fall through */
+
+		case 9:	/* Nehemiah */
+		default:
+			get_model_name(c);
+			display_cacheinfo(c);
+			break;
+	}
+}
+
 static void __init init_centaur(struct cpuinfo_x86 *c)
 {
 	enum {
@@ -386,21 +417,7 @@ static void __init init_centaur(struct cpuinfo_x86 *c)
 			break;
 
 		case 6:
-			switch (c->x86_model) {
-				case 6 ... 8:		/* Cyrix III family */
-					rdmsr (MSR_VIA_FCR, lo, hi);
-					lo |= (1<<1 | 1<<7);	/* Report CX8 & enable PGE */
-					wrmsr (MSR_VIA_FCR, lo, hi);
-
-					set_bit(X86_FEATURE_CX8, c->x86_capability);
-					set_bit(X86_FEATURE_3DNOW, c->x86_capability);
-
-				case 9:	/* Nehemiah */
-				default:
-					get_model_name(c);
-					display_cacheinfo(c);
-					break;
-			}
+			init_c3(c);
 			break;
 	}
 }

arch/i386/kernel/cpu/common.c

@@ -211,9 +211,10 @@ void __init generic_identify(struct cpuinfo_x86 * c)
 	
 		/* Intel-defined flags: level 0x00000001 */
 		if ( c->cpuid_level >= 0x00000001 ) {
-			u32 capability;
-			cpuid(0x00000001, &tfms, &junk, &junk, &capability);
+			u32 capability, excap;
+			cpuid(0x00000001, &tfms, &junk, &excap, &capability);
 			c->x86_capability[0] = capability;
+			c->x86_capability[4] = excap;
 			c->x86 = (tfms >> 8) & 15;
 			c->x86_model = (tfms >> 4) & 15;
 			c->x86_mask = tfms & 15;

arch/i386/kernel/cpu/proc.c

@@ -37,7 +37,20 @@ static int show_cpuinfo(struct seq_file *m, void *v)
 		NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
 
 		/* Other (Linux-defined) */
-		"cxmmx", "k6_mtrr", "cyrix_arr", "centaur_mcr", NULL, NULL, NULL, NULL,
+		"cxmmx", "k6_mtrr", "cyrix_arr", "centaur_mcr",
+		NULL, NULL, NULL, NULL,
+		NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
+		NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
+		NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
+
+		/* Intel-defined (#2) */
+		"pni", NULL, NULL, "monitor", "ds_cpl", NULL, NULL, NULL,
+		"tm2", NULL, "cnxt_id", NULL, NULL, NULL, NULL, NULL,
+		NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
+		NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
+
+		/* VIA/Cyrix/Centaur-defined */
+		NULL, NULL, "xstore", NULL, NULL, NULL, NULL, NULL,
 		NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
 		NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
 		NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,

drivers/base/platform.c

@@ -76,5 +76,6 @@ int __init platform_bus_init(void)
 }
 
 EXPORT_SYMBOL(legacy_bus);
+EXPORT_SYMBOL(platform_bus_type);
 EXPORT_SYMBOL(platform_device_register);
 EXPORT_SYMBOL(platform_device_unregister);

drivers/char/Kconfig

@@ -709,39 +709,20 @@ config NWFLASH
 
 	  If you're not sure, say N.
 
-config INTEL_RNG
-	tristate "Intel i8x0 Random Number Generator support"
+config HW_RANDOM
+	tristate "Intel/AMD/VIA HW Random Number Generator support"
 	depends on (X86 || IA64) && PCI
 	---help---
 	  This driver provides kernel-side support for the Random Number
-	  Generator hardware found on Intel i8xx-based motherboards.
+	  Generator hardware found on Intel i8xx-based motherboards,
+	  AMD 76x-based motherboards, and Via Nehemiah CPUs.
 
-	  Both a character driver, used to read() entropy data, and a timer
-	  function which automatically adds entropy directly into the
-	  kernel pool, are exported by this driver.
+	  Provides a character driver, used to read() entropy data.
 
 	  To compile this driver as a module ( = code which can be inserted in
 	  and removed from the running kernel whenever you want), say M here
 	  and read <file:Documentation/modules.txt>. The module will be called
-	  i810_rng.
-
-	  If unsure, say N.
-
-config AMD_RNG
-	tristate "AMD 768 Random Number Generator support"
-	depends on X86 && PCI
-	---help---
-	  This driver provides kernel-side support for the Random Number
-	  Generator hardware found on AMD 76x based motherboards.
-
-	  Both a character driver, used to read() entropy data, and a timer
-	  function which automatically adds entropy directly into the
-	  kernel pool, are exported by this driver.
-
-	  To compile this driver as a module ( = code which can be inserted in
-	  and removed from the running kernel whenever you want), say M here
-	  and read <file:Documentation/modules.txt>. The module will be called
-	  amd768_rng.
+	  hw_random.
 
 	  If unsure, say N.
 

drivers/char/Makefile

@@ -59,8 +59,7 @@ endif
 obj-$(CONFIG_TOSHIBA) += toshiba.o
 obj-$(CONFIG_I8K) += i8k.o
 obj-$(CONFIG_DS1620) += ds1620.o
-obj-$(CONFIG_INTEL_RNG) += i810_rng.o
-obj-$(CONFIG_AMD_RNG) += amd768_rng.o
+obj-$(CONFIG_HW_RANDOM) += hw_random.o
 obj-$(CONFIG_QIC02_TAPE) += tpqic02.o
 obj-$(CONFIG_FTAPE) += ftape/
 obj-$(CONFIG_H8) += h8.o

drivers/char/amd768_rng.c

-/*
- 	Hardware driver for the AMD 768 Random Number Generator (RNG)
-	(c) Copyright 2001 Red Hat Inc <alan@redhat.com>
- 
- 	derived from
- 
-	Hardware driver for Intel i810 Random Number Generator (RNG)
-	Copyright 2000,2001 Jeff Garzik <jgarzik@pobox.com>
-	Copyright 2000,2001 Philipp Rumpf <prumpf@mandrakesoft.com>
-
-	Please read Documentation/i810_rng.txt for details on use.
-
-	----------------------------------------------------------
-	This software may be used and distributed according to the terms
-        of the GNU General Public License, incorporated herein by reference.
-
- */
-
-
-#include <linux/module.h>
-#include <linux/kernel.h>
-#include <linux/fs.h>
-#include <linux/init.h>
-#include <linux/pci.h>
-#include <linux/interrupt.h>
-#include <linux/spinlock.h>
-#include <linux/random.h>
-#include <linux/miscdevice.h>
-#include <linux/smp_lock.h>
-#include <linux/mm.h>
-#include <linux/delay.h>
-
-#include <asm/io.h>
-#include <asm/uaccess.h>
-
-
-/*
- * core module and version information
- */
-#define RNG_VERSION "0.1.0"
-#define RNG_MODULE_NAME "amd768_rng"
-#define RNG_DRIVER_NAME   RNG_MODULE_NAME " hardware driver " RNG_VERSION
-#define PFX RNG_MODULE_NAME ": "
-
-
-/*
- * debugging macros
- */
-#undef RNG_DEBUG /* define to enable copious debugging info */
-
-#ifdef RNG_DEBUG
-/* note: prints function name for you */
-#define DPRINTK(fmt, args...) printk(KERN_DEBUG "%s: " fmt, __FUNCTION__ , ## args)
-#else
-#define DPRINTK(fmt, args...)
-#endif
-
-#undef RNG_NDEBUG        /* define to disable lightweight runtime checks */
-#ifdef RNG_NDEBUG
-#define assert(expr)
-#else
-#define assert(expr) \
-        if(!(expr)) {                                   \
-        printk( "Assertion failed! %s,%s,%s,line=%d\n", \
-        #expr,__FILE__,__FUNCTION__,__LINE__);          \
-        }
-#endif
-
-#define RNG_MISCDEV_MINOR		183 /* official */
-
-/*
- * various RNG status variables.  they are globals
- * as we only support a single RNG device
- */
-
-static u32 pmbase;			/* PMxx I/O base */
-static struct semaphore rng_open_sem;	/* Semaphore for serializing rng_open/release */
-
-
-/*
- * inlined helper functions for accessing RNG registers
- */
-
-static inline int rng_data_present (void)
-{
-      	return inl(pmbase+0xF4) & 1;
-}
-
-
-static inline int rng_data_read (void)
-{
-	return inl(pmbase+0xF0);
-}
-
-static int rng_dev_open (struct inode *inode, struct file *filp)
-{
-	if ((filp->f_mode & FMODE_READ) == 0)
-		return -EINVAL;
-	if (filp->f_mode & FMODE_WRITE)
-		return -EINVAL;
-
-	/* wait for device to become free */
-	if (filp->f_flags & O_NONBLOCK) {
-		if (down_trylock (&rng_open_sem))
-			return -EAGAIN;
-	} else {
-		if (down_interruptible (&rng_open_sem))
-			return -ERESTARTSYS;
-	}
-	return 0;
-}
-
-
-static int rng_dev_release (struct inode *inode, struct file *filp)
-{
-	up(&rng_open_sem);
-	return 0;
-}
-
-
-static ssize_t rng_dev_read (struct file *filp, char *buf, size_t size,
-			     loff_t * offp)
-{
-	static spinlock_t rng_lock = SPIN_LOCK_UNLOCKED;
-	int have_data;
-	u32 data = 0;
-	ssize_t ret = 0;
-
-	while (size) {
-		spin_lock(&rng_lock);
-
-		have_data = 0;
-		if (rng_data_present()) {
-			data = rng_data_read();
-			have_data = 4;
-		}
-
-		spin_unlock (&rng_lock);
-
-		while (have_data > 0) {
-			if (put_user((u8)data, buf++)) {
-				ret = ret ? : -EFAULT;
-				break;
-			}
-			size--;
-			ret++;
-			have_data--;
-			data>>=8;
-		}
-
-		if (filp->f_flags & O_NONBLOCK)
-			return ret ? : -EAGAIN;
-
-		if(need_resched())
-		{
-			current->state = TASK_INTERRUPTIBLE;
-			schedule_timeout(1);
-		}
-		else
-			udelay(200);	/* FIXME: We could poll for 250uS ?? */
-
-		if (signal_pending (current))
-			return ret ? : -ERESTARTSYS;
-	}
-	return ret;
-}
-
-
-static struct file_operations rng_chrdev_ops = {
-	.owner		= THIS_MODULE,
-	.open		= rng_dev_open,
-	.release	= rng_dev_release,
-	.read		= rng_dev_read,
-};
-
-
-static struct miscdevice rng_miscdev = {
-	RNG_MISCDEV_MINOR,
-	RNG_MODULE_NAME,
-	&rng_chrdev_ops,
-};
-
-
-/*
- * rng_init_one - look for and attempt to init a single RNG
- */
-static int __init rng_init_one (struct pci_dev *dev)
-{
-	int rc;
-	u8 rnen;
-
-	DPRINTK ("ENTER\n");
-
-	rc = misc_register (&rng_miscdev);
-	if (rc) {
-		printk (KERN_ERR PFX "cannot register misc device\n");
-		DPRINTK ("EXIT, returning %d\n", rc);
-		goto err_out;
-	}
-
-	pci_read_config_dword(dev, 0x58, &pmbase);
-
-	pmbase&=0x0000FF00;
-
-	if(pmbase == 0)
-	{
-		printk (KERN_ERR PFX "power management base not set\n");
-		DPRINTK ("EXIT, returning %d\n", rc);
-		goto err_out_free_miscdev;
-	}
-
-	pci_read_config_byte(dev, 0x40, &rnen);
-	rnen|=(1<<7);	/* RNG on */
-	pci_write_config_byte(dev, 0x40, rnen);
-
-	pci_read_config_byte(dev, 0x41, &rnen);
-	rnen|=(1<<7);	/* PMIO enable */
-	pci_write_config_byte(dev, 0x41, rnen);
-
-	printk(KERN_INFO PFX "AMD768 system management I/O registers at 0x%X.\n", pmbase);
-	DPRINTK ("EXIT, returning 0\n");
-	return 0;
-
-err_out_free_miscdev:
-	misc_deregister (&rng_miscdev);
-err_out:
-	return rc;
-}
-
-
-/*
- * Data for PCI driver interface
- *
- * This data only exists for exporting the supported
- * PCI ids via MODULE_DEVICE_TABLE.  We do not actually
- * register a pci_driver, because someone else might one day
- * want to register another driver on the same PCI id.
- */
-static struct pci_device_id rng_pci_tbl[] __initdata = {
-	{ 0x1022, 0x7443, PCI_ANY_ID, PCI_ANY_ID, },
-	{ 0, },
-};
-MODULE_DEVICE_TABLE (pci, rng_pci_tbl);
-
-
-MODULE_AUTHOR("Alan Cox, Jeff Garzik, Philipp Rumpf, Matt Sottek");
-MODULE_DESCRIPTION("AMD 768 Random Number Generator (RNG) driver");
-MODULE_LICENSE("GPL");
-
-
-/*
- * rng_init - initialize RNG module
- */
-static int __init rng_init (void)
-{
-	int rc;
-	struct pci_dev *pdev;
-
-	DPRINTK ("ENTER\n");
-
-	init_MUTEX (&rng_open_sem);
-
-	pci_for_each_dev(pdev) {
-		if (pci_match_device (rng_pci_tbl, pdev) != NULL)
-			goto match;
-	}
-
-	DPRINTK ("EXIT, returning -ENODEV\n");
-	return -ENODEV;
-
-match:
-	rc = rng_init_one (pdev);
-	if (rc)
-		return rc;
-
-	printk (KERN_INFO RNG_DRIVER_NAME " loaded\n");
-
-	DPRINTK ("EXIT, returning 0\n");
-	return 0;
-}
-
-
-/*
- * rng_init - shutdown RNG module
- */
-static void __exit rng_cleanup (void)
-{
-	DPRINTK ("ENTER\n");
-	misc_deregister (&rng_miscdev);
-	DPRINTK ("EXIT\n");
-}
-
-
-module_init (rng_init);
-module_exit (rng_cleanup);

drivers/char/i810_rng.c → drivers/char/hw_random.c

 /*
+ 	Hardware driver for the Intel/AMD/VIA Random Number Generators (RNG)
+	(c) Copyright 2003 Red Hat Inc <jgarzik@redhat.com>
+ 
+ 	derived from
+ 
+        Hardware driver for the AMD 768 Random Number Generator (RNG)
+        (c) Copyright 2001 Red Hat Inc <alan@redhat.com>
+
+ 	derived from
  
 	Hardware driver for Intel i810 Random Number Generator (RNG)
 	Copyright 2000,2001 Jeff Garzik <jgarzik@pobox.com>
 	Copyright 2000,2001 Philipp Rumpf <prumpf@mandrakesoft.com>
 
-	Driver Web site:  http://sourceforge.net/projects/gkernel/
-
-	Please read Documentation/i810_rng.txt for details on use.
+	Please read Documentation/hw_random.txt for details on use.
 
 	----------------------------------------------------------
-
 	This software may be used and distributed according to the terms
         of the GNU General Public License, incorporated herein by reference.
 
@@ -29,6 +35,11 @@
 #include <linux/mm.h>
 #include <linux/delay.h>
 
+#ifdef __i386__
+#include <asm/msr.h>
+#include <asm/cpufeature.h>
+#endif
+
 #include <asm/io.h>
 #include <asm/uaccess.h>
 
@@ -36,8 +47,8 @@
 /*
  * core module and version information
  */
-#define RNG_VERSION "0.9.8"
-#define RNG_MODULE_NAME "i810_rng"
+#define RNG_VERSION "0.9.0"
+#define RNG_MODULE_NAME "hw_random"
 #define RNG_DRIVER_NAME   RNG_MODULE_NAME " hardware driver " RNG_VERSION
 #define PFX RNG_MODULE_NAME ": "
 
@@ -54,7 +65,7 @@
 #define DPRINTK(fmt, args...)
 #endif
 
-#undef RNG_NDEBUG        /* define to disable lightweight runtime checks */
+#define RNG_NDEBUG        /* define to disable lightweight runtime checks */
 #ifdef RNG_NDEBUG
 #define assert(expr)
 #else
@@ -65,150 +76,405 @@
         }
 #endif
 
+#define RNG_MISCDEV_MINOR		183 /* official */
 
-/*
- * RNG registers (offsets from rng_mem)
- */
-#define RNG_HW_STATUS			0
-#define		RNG_PRESENT		0x40
-#define		RNG_ENABLED		0x01
-#define RNG_STATUS			1
-#define		RNG_DATA_PRESENT	0x01
-#define RNG_DATA			2
+static int rng_dev_open (struct inode *inode, struct file *filp);
+static ssize_t rng_dev_read (struct file *filp, char *buf, size_t size,
+			     loff_t * offp);
+
+static int __init intel_init (struct pci_dev *dev);
+static void intel_cleanup(void);
+static unsigned int intel_data_present (void);
+static u32 intel_data_read (void);
+
+static int __init amd_init (struct pci_dev *dev);
+static void amd_cleanup(void);
+static unsigned int amd_data_present (void);
+static u32 amd_data_read (void);
+
+static int __init via_init(struct pci_dev *dev);
+static void via_cleanup(void);
+static unsigned int via_data_present (void);
+static u32 via_data_read (void);
+
+struct rng_operations {
+	int (*init) (struct pci_dev *dev);
+	void (*cleanup) (void);
+	unsigned int (*data_present) (void);
+	u32 (*data_read) (void);
+	unsigned int n_bytes; /* number of bytes per ->data_read */
+};
+static struct rng_operations *rng_ops;
+
+static struct file_operations rng_chrdev_ops = {
+	.owner		= THIS_MODULE,
+	.open		= rng_dev_open,
+	.read		= rng_dev_read,
+};
+
+
+static struct miscdevice rng_miscdev = {
+	RNG_MISCDEV_MINOR,
+	RNG_MODULE_NAME,
+	&rng_chrdev_ops,
+};
+
+enum {
+	rng_hw_none,
+	rng_hw_intel,
+	rng_hw_amd,
+	rng_hw_via,
+};
+
+static struct rng_operations rng_vendor_ops[] __initdata = {
+	/* rng_hw_none */
+	{ },
+
+	/* rng_hw_intel */
+	{ intel_init, intel_cleanup, intel_data_present,
+	  intel_data_read, 1 },
+
+	/* rng_hw_amd */
+	{ amd_init, amd_cleanup, amd_data_present, amd_data_read, 4 },
+
+	/* rng_hw_via */
+	{ via_init, via_cleanup, via_data_present, via_data_read, 1 },
+};
 
 /*
- * Magic address at which Intel PCI bridges locate the RNG
+ * Data for PCI driver interface
+ *
+ * This data only exists for exporting the supported
+ * PCI ids via MODULE_DEVICE_TABLE.  We do not actually
+ * register a pci_driver, because someone else might one day
+ * want to register another driver on the same PCI id.
  */
-#define RNG_ADDR			0xFFBC015F
-#define RNG_ADDR_LEN			3
+static struct pci_device_id rng_pci_tbl[] __initdata = {
+	{ 0x1022, 0x7443, PCI_ANY_ID, PCI_ANY_ID, 0, 0, rng_hw_amd },
 
-#define RNG_MISCDEV_MINOR		183 /* official */
+	{ 0x8086, 0x2418, PCI_ANY_ID, PCI_ANY_ID, 0, 0, rng_hw_intel },
+	{ 0x8086, 0x2428, PCI_ANY_ID, PCI_ANY_ID, 0, 0, rng_hw_intel },
+	{ 0x8086, 0x2448, PCI_ANY_ID, PCI_ANY_ID, 0, 0, rng_hw_intel },
+	{ 0x8086, 0x244e, PCI_ANY_ID, PCI_ANY_ID, 0, 0, rng_hw_intel },
+	{ 0x8086, 0x245e, PCI_ANY_ID, PCI_ANY_ID, 0, 0, rng_hw_intel },
 
-/*
- * various RNG status variables.  they are globals
- * as we only support a single RNG device
+	{ 0, },	/* terminate list */
+};
+MODULE_DEVICE_TABLE (pci, rng_pci_tbl);
+
+
+/***********************************************************************
+ *
+ * Intel RNG operations
+ *
  */
-static void *rng_mem;			/* token to our ioremap'd RNG register area */
-static struct semaphore rng_open_sem;	/* Semaphore for serializing rng_open/release */
 
+/*
+ * RNG registers (offsets from rng_mem)
+ */
+#define INTEL_RNG_HW_STATUS			0
+#define         INTEL_RNG_PRESENT		0x40
+#define         INTEL_RNG_ENABLED		0x01
+#define INTEL_RNG_STATUS			1
+#define         INTEL_RNG_DATA_PRESENT		0x01
+#define INTEL_RNG_DATA				2
 
 /*
- * inlined helper functions for accessing RNG registers
+ * Magic address at which Intel PCI bridges locate the RNG
  */
-static inline u8 rng_hwstatus (void)
+#define INTEL_RNG_ADDR				0xFFBC015F
+#define INTEL_RNG_ADDR_LEN			3
+
+/* token to our ioremap'd RNG register area */
+static void *rng_mem;
+
+static inline u8 intel_hwstatus (void)
 {
 	assert (rng_mem != NULL);
-	return readb (rng_mem + RNG_HW_STATUS);
+	return readb (rng_mem + INTEL_RNG_HW_STATUS);
 }
 
-static inline u8 rng_hwstatus_set (u8 hw_status)
+static inline u8 intel_hwstatus_set (u8 hw_status)
 {
 	assert (rng_mem != NULL);
-	writeb (hw_status, rng_mem + RNG_HW_STATUS);
-	return rng_hwstatus ();
+	writeb (hw_status, rng_mem + INTEL_RNG_HW_STATUS);
+	return intel_hwstatus ();
 }
 
+static unsigned int intel_data_present(void)
+{
+	assert (rng_mem != NULL);
 
-static inline int rng_data_present (void)
+	return (readb (rng_mem + INTEL_RNG_STATUS) & INTEL_RNG_DATA_PRESENT) ?
+		1 : 0;
+}
+
+static u32 intel_data_read(void)
 {
 	assert (rng_mem != NULL);
 
-	return (readb (rng_mem + RNG_STATUS) & RNG_DATA_PRESENT) ? 1 : 0;
+	return readb (rng_mem + INTEL_RNG_DATA);
 }
 
+static int __init intel_init (struct pci_dev *dev)
+{
+	int rc;
+	u8 hw_status;
+
+	DPRINTK ("ENTER\n");
+
+	rng_mem = ioremap (INTEL_RNG_ADDR, INTEL_RNG_ADDR_LEN);
+	if (rng_mem == NULL) {
+		printk (KERN_ERR PFX "cannot ioremap RNG Memory\n");
+		rc = -EBUSY;
+		goto err_out;
+	}
+
+	/* Check for Intel 82802 */
+	hw_status = intel_hwstatus ();
+	if ((hw_status & INTEL_RNG_PRESENT) == 0) {
+		printk (KERN_ERR PFX "RNG not detected\n");
+		rc = -ENODEV;
+		goto err_out_free_map;
+	}
+
+	/* turn RNG h/w on, if it's off */
+	if ((hw_status & INTEL_RNG_ENABLED) == 0)
+		hw_status = intel_hwstatus_set (hw_status | INTEL_RNG_ENABLED);
+	if ((hw_status & INTEL_RNG_ENABLED) == 0) {
+		printk (KERN_ERR PFX "cannot enable RNG, aborting\n");
+		rc = -EIO;
+		goto err_out_free_map;
+	}
+
+	DPRINTK ("EXIT, returning 0\n");
+	return 0;
+
+err_out_free_map:
+	iounmap (rng_mem);
+	rng_mem = NULL;
+err_out:
+	DPRINTK ("EXIT, returning %d\n", rc);
+	return rc;
+}
 
-static inline int rng_data_read (void)
+static void intel_cleanup(void)
 {
-	assert (rng_mem != NULL);
+	u8 hw_status;
 
-	return readb (rng_mem + RNG_DATA);
+	hw_status = intel_hwstatus ();
+	if (hw_status & INTEL_RNG_ENABLED)
+		intel_hwstatus_set (hw_status & ~INTEL_RNG_ENABLED);
+	else
+		printk(KERN_WARNING PFX "unusual: RNG already disabled\n");
+	iounmap(rng_mem);
+	rng_mem = NULL;
 }
 
-/*
- * rng_enable - enable the RNG hardware
+/***********************************************************************
+ *
+ * AMD RNG operations
+ *
  */
 
-static int rng_enable (void)
+static u32 pmbase;			/* PMxx I/O base */
+static struct pci_dev *amd_dev;
+
+static unsigned int amd_data_present (void)
+{
+      	return inl(pmbase + 0xF4) & 1;
+}
+
+
+static u32 amd_data_read (void)
 {
-	int rc = 0;
-	u8 hw_status, new_status;
+	return inl(pmbase + 0xF0);
+}
+
+static int __init amd_init (struct pci_dev *dev)
+{
+	int rc;
+	u8 rnen;
 
 	DPRINTK ("ENTER\n");
 
-	hw_status = rng_hwstatus ();
+	pci_read_config_dword(dev, 0x58, &pmbase);
 
-	if ((hw_status & RNG_ENABLED) == 0) {
-		new_status = rng_hwstatus_set (hw_status | RNG_ENABLED);
+	pmbase &= 0x0000FF00;
 
-		if (new_status & RNG_ENABLED)
-			printk (KERN_INFO PFX "RNG h/w enabled\n");
-		else {
-			printk (KERN_ERR PFX "Unable to enable the RNG\n");
+	if (pmbase == 0)
+	{
+		printk (KERN_ERR PFX "power management base not set\n");
 		rc = -EIO;
+		goto err_out;
 	}
-	}
 
+	pci_read_config_byte(dev, 0x40, &rnen);
+	rnen |= (1 << 7);	/* RNG on */
+	pci_write_config_byte(dev, 0x40, rnen);
+
+	pci_read_config_byte(dev, 0x41, &rnen);
+	rnen |= (1 << 7);	/* PMIO enable */
+	pci_write_config_byte(dev, 0x41, rnen);
+
+	printk(KERN_INFO PFX "AMD768 system management I/O registers at 0x%X.\n", pmbase);
+
+	amd_dev = dev;
+
+	DPRINTK ("EXIT, returning 0\n");
+	return 0;
+
+err_out:
 	DPRINTK ("EXIT, returning %d\n", rc);
 	return rc;
 }
 
+static void amd_cleanup(void)
+{
+	u8 rnen;
+
+	pci_read_config_byte(amd_dev, 0x40, &rnen);
+	rnen &= ~(1 << 7);	/* RNG off */
+	pci_write_config_byte(amd_dev, 0x40, rnen);
+
+	/* FIXME: twiddle pmio, also? */
+}
+
+/***********************************************************************
+ *
+ * VIA RNG operations
+ *
+ */
+
+enum {
+	VIA_STRFILT_CNT_SHIFT	= 16,
+	VIA_STRFILT_FAIL	= (1 << 15),
+	VIA_STRFILT_ENABLE	= (1 << 14),
+	VIA_RAWBITS_ENABLE	= (1 << 13),
+	VIA_RNG_ENABLE		= (1 << 6),
+	VIA_XSTORE_CNT_MASK	= 0x0F,
+
+	VIA_RNG_CHUNK_8		= 0x00,	/* 64 rand bits, 64 stored bits */
+	VIA_RNG_CHUNK_4		= 0x01,	/* 32 rand bits, 32 stored bits */
+	VIA_RNG_CHUNK_4_MASK	= 0xFFFFFFFF,
+	VIA_RNG_CHUNK_2		= 0x02,	/* 16 rand bits, 32 stored bits */
+	VIA_RNG_CHUNK_2_MASK	= 0xFFFF,
+	VIA_RNG_CHUNK_1		= 0x03,	/* 8 rand bits, 32 stored bits */
+	VIA_RNG_CHUNK_1_MASK	= 0xFF,
+};
+
+u32 via_rng_datum;
+
 /*
- * rng_disable - disable the RNG hardware
+ * Investigate using the 'rep' prefix to obtain 32 bits of random data
+ * in one insn.  The upside is potentially better performance.  The
+ * downside is that the instruction becomes no longer atomic.  Due to
+ * this, just like familiar issues with /dev/random itself, the worst
+ * case of a 'rep xstore' could potentially pause a cpu for an
+ * unreasonably long time.  In practice, this condition would likely
+ * only occur when the hardware is failing.  (or so we hope :))
+ *
+ * Another possible performance boost may come from simply buffering
+ * until we have 4 bytes, thus returning a u32 at a time,
+ * instead of the current u8-at-a-time.
  */
 
-static void rng_disable(void)
+static inline u32 xstore(u32 *addr, u32 edx_in)
 {
-	u8 hw_status, new_status;
+	u32 eax_out;
 
-	DPRINTK ("ENTER\n");
+	asm(".byte 0x0F,0xA7,0xC0 /* xstore %%edi (addr=%0) */"
+		:"=m"(*addr), "=a"(eax_out)
+		:"D"(addr), "d"(edx_in));
 
-	hw_status = rng_hwstatus ();
+	return eax_out;
+}
 
-	if (hw_status & RNG_ENABLED) {
-		new_status = rng_hwstatus_set (hw_status & ~RNG_ENABLED);
+static unsigned int via_data_present(void)
+{
+	u32 bytes_out;
 
-		if ((new_status & RNG_ENABLED) == 0)
-			printk (KERN_INFO PFX "RNG h/w disabled\n");
-		else {
-			printk (KERN_ERR PFX "Unable to disable the RNG\n");
-		}
-	}
+	/* We choose the recommended 1-byte-per-instruction RNG rate,
+	 * for greater randomness at the expense of speed.  Larger
+	 * values 2, 4, or 8 bytes-per-instruction yield greater
+	 * speed at lesser randomness.
+	 *
+	 * If you change this to another VIA_CHUNK_n, you must also
+	 * change the ->n_bytes values in rng_vendor_ops[] tables.
+	 * VIA_CHUNK_8 requires further code changes.
+	 *
+	 * A copy of MSR_VIA_RNG is placed in eax_out when xstore
+	 * completes.
+	 */
+	via_rng_datum = 0; /* paranoia, not really necessary */
+	bytes_out = xstore(&via_rng_datum, VIA_RNG_CHUNK_1) & VIA_XSTORE_CNT_MASK;
+	if (bytes_out == 0)
+		return 0;
 
-	DPRINTK ("EXIT\n");
+	return 1;
 }
 
-static int rng_dev_open (struct inode *inode, struct file *filp)
+static u32 via_data_read(void)
 {
-	int rc;
-
-	if ((filp->f_mode & FMODE_READ) == 0)
-		return -EINVAL;
-	if (filp->f_mode & FMODE_WRITE)
-		return -EINVAL;
+	return via_rng_datum;
+}
 
-	/* wait for device to become free */
-	if (filp->f_flags & O_NONBLOCK) {
-		if (down_trylock (&rng_open_sem))
-			return -EAGAIN;
-	} else {
-		if (down_interruptible (&rng_open_sem))
-			return -ERESTARTSYS;
-	}
+static int __init via_init(struct pci_dev *dev)
+{
+	u32 lo, hi, old_lo;
 
-	rc = rng_enable ();
-	if (rc) {
-		up (&rng_open_sem);
-		return rc;
+	/* Control the RNG via MSR.  Tread lightly and pay very close
+	 * close attention to values written, as the reserved fields
+	 * are documented to be "undefined and unpredictable"; but it
+	 * does not say to write them as zero, so I make a guess that
+	 * we restore the values we find in the register.
+	 */
+	rdmsr(MSR_VIA_RNG, lo, hi);
+
+	old_lo = lo;
+	lo &= ~(0x7f << VIA_STRFILT_CNT_SHIFT);
+	lo &= ~VIA_XSTORE_CNT_MASK;
+	lo &= ~(VIA_STRFILT_ENABLE | VIA_STRFILT_FAIL | VIA_RAWBITS_ENABLE);
+	lo |= VIA_RNG_ENABLE;
+
+	if (lo != old_lo)
+		wrmsr(MSR_VIA_RNG, lo, hi);
+
+	/* perhaps-unnecessary sanity check; remove after testing if
+	   unneeded */
+	rdmsr(MSR_VIA_RNG, lo, hi);
+	if ((lo & VIA_RNG_ENABLE) == 0) {
+		printk(KERN_ERR PFX "cannot enable VIA C3 RNG, aborting\n");
+		return -ENODEV;
 	}
 
 	return 0;
 }
 
+static void via_cleanup(void)
+{
+	u32 lo, hi;
+
+	rdmsr(MSR_VIA_RNG, lo, hi);
+	lo &= ~VIA_RNG_ENABLE;
+	wrmsr(MSR_VIA_RNG, lo, hi);
+}
+
 
-static int rng_dev_release (struct inode *inode, struct file *filp)
+/***********************************************************************
+ *
+ * /dev/hwrandom character device handling (major 10, minor 183)
+ *
+ */
+
+static int rng_dev_open (struct inode *inode, struct file *filp)
 {
-	rng_disable ();
-	up (&rng_open_sem);
+	/* enforce read-only access to this chrdev */
+	if ((filp->f_mode & FMODE_READ) == 0)
+		return -EINVAL;
+	if (filp->f_mode & FMODE_WRITE)
+		return -EINVAL;
+
 	return 0;
 }
 
@@ -217,63 +483,50 @@ static ssize_t rng_dev_read (struct file *filp, char *buf, size_t size,
 			     loff_t * offp)
 {
 	static spinlock_t rng_lock = SPIN_LOCK_UNLOCKED;
-	int have_data;
-	u8 data = 0;
+	unsigned int have_data;
+	u32 data = 0;
 	ssize_t ret = 0;
 
 	while (size) {
-		spin_lock (&rng_lock);
+		spin_lock(&rng_lock);
 
 		have_data = 0;
-		if (rng_data_present ()) {
-			data = rng_data_read ();
-			have_data = 1;
+		if (rng_ops->data_present()) {
+			data = rng_ops->data_read();
+			have_data = rng_ops->n_bytes;
 		}
 
 		spin_unlock (&rng_lock);
 
-		if (have_data) {
-			if (put_user (data, buf++)) {
+		while (have_data > 0) {
+			if (put_user((u8)data, buf++)) {
 				ret = ret ? : -EFAULT;
 				break;
 			}
 			size--;
 			ret++;
+			have_data--;
+			data>>=8;
 		}
 
 		if (filp->f_flags & O_NONBLOCK)
 			return ret ? : -EAGAIN;
 
-		if (need_resched())
+		if(need_resched())
 		{
 			current->state = TASK_INTERRUPTIBLE;
 			schedule_timeout(1);
 		}
 		else
-			udelay(200);
+			udelay(200);	/* FIXME: We could poll for 250uS ?? */
 
 		if (signal_pending (current))
 			return ret ? : -ERESTARTSYS;
 	}
-
 	return ret;
 }
 
 
-static struct file_operations rng_chrdev_ops = {
-	.owner		= THIS_MODULE,
-	.open		= rng_dev_open,
-	.release	= rng_dev_release,
-	.read		= rng_dev_read,
-};
-
-
-static struct miscdevice rng_miscdev = {
-	RNG_MISCDEV_MINOR,
-	RNG_MODULE_NAME,
-	&rng_chrdev_ops,
-};
-
 
 /*
  * rng_init_one - look for and attempt to init a single RNG
@@ -281,75 +534,35 @@ static struct miscdevice rng_miscdev = {
 static int __init rng_init_one (struct pci_dev *dev)
 {
 	int rc;
-	u8 hw_status;
 
 	DPRINTK ("ENTER\n");
 
-	rc = misc_register (&rng_miscdev);
-	if (rc) {
-		printk (KERN_ERR PFX "cannot register misc device\n");
-		DPRINTK ("EXIT, returning %d\n", rc);
-		goto err_out;
-	}
-
-	rng_mem = ioremap (RNG_ADDR, RNG_ADDR_LEN);
-	if (rng_mem == NULL) {
-		printk (KERN_ERR PFX "cannot ioremap RNG Memory\n");
-		DPRINTK ("EXIT, returning -EBUSY\n");
-		rc = -EBUSY;
-		goto err_out_free_miscdev;
-	}
+	assert(rng_ops != NULL);
 
-	/* Check for Intel 82802 */
-	hw_status = rng_hwstatus ();
-	if ((hw_status & RNG_PRESENT) == 0) {
-		printk (KERN_ERR PFX "RNG not detected\n");
-		DPRINTK ("EXIT, returning -ENODEV\n");
-		rc = -ENODEV;
-		goto err_out_free_map;
-	}
+	rc = rng_ops->init(dev);
+	if (rc)
+		goto err_out;
 
-	/* turn RNG h/w off, if it's on */
-	if (hw_status & RNG_ENABLED)
-		hw_status = rng_hwstatus_set (hw_status & ~RNG_ENABLED);
-	if (hw_status & RNG_ENABLED) {
-		printk (KERN_ERR PFX "cannot disable RNG, aborting\n");
-		goto err_out_free_map;
+	rc = misc_register (&rng_miscdev);
+	if (rc) {
+		printk (KERN_ERR PFX "misc device register failed\n");
+		goto err_out_cleanup_hw;
 	}
 
 	DPRINTK ("EXIT, returning 0\n");
 	return 0;
 
-err_out_free_map:
-	iounmap (rng_mem);
-err_out_free_miscdev:
-	misc_deregister (&rng_miscdev);
+err_out_cleanup_hw:
+	rng_ops->cleanup();
 err_out:
+	DPRINTK ("EXIT, returning %d\n", rc);
 	return rc;
 }
 
 
-/*
- * Data for PCI driver interface
- *
- * This data only exists for exporting the supported
- * PCI ids via MODULE_DEVICE_TABLE.  We do not actually
- * register a pci_driver, because someone else might one day
- * want to register another driver on the same PCI id.
- */
-static struct pci_device_id rng_pci_tbl[] __initdata = {
-	{ 0x8086, 0x2418, PCI_ANY_ID, PCI_ANY_ID, },
-	{ 0x8086, 0x2428, PCI_ANY_ID, PCI_ANY_ID, },
-	{ 0x8086, 0x2448, PCI_ANY_ID, PCI_ANY_ID, },
-	{ 0x8086, 0x244e, PCI_ANY_ID, PCI_ANY_ID, },
-	{ 0x8086, 0x245e, PCI_ANY_ID, PCI_ANY_ID, },
-	{ 0, },
-};
-MODULE_DEVICE_TABLE (pci, rng_pci_tbl);
-
 
-MODULE_AUTHOR("Jeff Garzik, Philipp Rumpf, Matt Sottek");
-MODULE_DESCRIPTION("Intel i8xx chipset Random Number Generator (RNG) driver");
+MODULE_AUTHOR("The Linux Kernel team");
+MODULE_DESCRIPTION("H/W Random Number Generator (RNG) driver");
 MODULE_LICENSE("GPL");
 
 
@@ -360,15 +573,27 @@ static int __init rng_init (void)
 {
 	int rc;
 	struct pci_dev *pdev;
+	const struct pci_device_id *ent;
 
 	DPRINTK ("ENTER\n");
 
-	init_MUTEX (&rng_open_sem);
-
+	/* Probe for Intel, AMD RNGs */
 	pci_for_each_dev(pdev) {
-		if (pci_match_device (rng_pci_tbl, pdev) != NULL)
+		ent = pci_match_device (rng_pci_tbl, pdev);
+		if (ent) {
+			rng_ops = &rng_vendor_ops[ent->driver_data];
 			goto match;
 		}
+	}
+
+#ifdef __i386__
+	/* Probe for VIA RNG */
+	if (cpu_has_xstore) {
+		rng_ops = &rng_vendor_ops[rng_hw_via];
+		pdev = NULL;
+		goto match;
+	}
+#endif
 
 	DPRINTK ("EXIT, returning -ENODEV\n");
 	return -ENODEV;
@@ -394,7 +619,8 @@ static void __exit rng_cleanup (void)
 
 	misc_deregister (&rng_miscdev);
 
-	iounmap (rng_mem);
+	if (rng_ops->cleanup)
+		rng_ops->cleanup();
 
 	DPRINTK ("EXIT\n");
 }

drivers/i2c/i2c-core.c

@@ -30,6 +30,7 @@
 #include <linux/proc_fs.h>
 #include <linux/i2c.h>
 #include <linux/init.h>
+#include <linux/seq_file.h>
 #include <asm/uaccess.h>
 
 
@@ -326,42 +327,45 @@ int i2c_check_addr(struct i2c_adapter *adapter, int addr)
 int i2c_attach_client(struct i2c_client *client)
 {
 	struct i2c_adapter *adapter = client->adapter;
-	int res =  -EBUSY, i;
+	int i;
 
 	down(&adapter->list);
-	if (__i2c_check_addr(client->adapter,client->addr))
+	if (__i2c_check_addr(client->adapter, client->addr))
 		goto out_unlock_list;
 
-	for (i = 0; i < I2C_CLIENT_MAX; i++)
-		if (NULL == adapter->clients[i])
-			break;
-	if (I2C_CLIENT_MAX == i) {
+	for (i = 0; i < I2C_CLIENT_MAX; i++) {
+		if (!adapter->clients[i])
+			goto free_slot;
+	}
+
 	printk(KERN_WARNING 
 	       " i2c-core.o: attach_client(%s) - enlarge I2C_CLIENT_MAX.\n",
 	       client->name);
-		res = -ENOMEM;
-		goto out_unlock_list;
-	}
 
+ out_unlock_list:
+	up(&adapter->list);
+	return -EBUSY;
+
+ free_slot:
 	adapter->clients[i] = client;
 	up(&adapter->list);
 	
-	if (adapter->client_register) 
-		if (adapter->client_register(client)) 
-			printk(KERN_DEBUG "i2c-core.o: warning: client_register seems "
+	if (adapter->client_register)  {
+		if (adapter->client_register(client))  {
+			printk(KERN_DEBUG
+			       "i2c-core.o: warning: client_register seems "
 			       "to have failed for client %02x at adapter %s\n",
-			       client->addr,adapter->name);
-	DEB(printk(KERN_DEBUG "i2c-core.o: client [%s] registered to adapter [%s](pos. %d).\n",
-		client->name, adapter->name,i));
+			       client->addr, adapter->name);
+		}
+	}
 
-	if(client->flags & I2C_CLIENT_ALLOW_USE)
-		client->usage_count = 0;
+	DEB(printk(KERN_DEBUG
+		   "i2c-core.o: client [%s] registered to adapter [%s] "
+		   "(pos. %d).\n", client->name, adapter->name, i));
 
+	if (client->flags & I2C_CLIENT_ALLOW_USE)
+		client->usage_count = 0;
 	return 0;
-
- out_unlock_list:
-	up(&adapter->list);
-	return res;
 }
 
 
@@ -376,28 +380,30 @@ int i2c_detach_client(struct i2c_client *client)
 	if (adapter->client_unregister)  {
 		res = adapter->client_unregister(client);
 		if (res) {
-			printk(KERN_ERR "i2c-core.o: client_unregister [%s] failed, "
-			       "client not detached",client->name);
-			return res;
+			printk(KERN_ERR
+			       "i2c-core.o: client_unregister [%s] failed, "
+			       "client not detached", client->name);
+			goto out;
 		}
 	}
 
 	down(&adapter->list);
 	for (i = 0; i < I2C_CLIENT_MAX; i++) {
-		if (client == adapter->clients[i])
-			break;
+		if (client == adapter->clients[i]) {
+			adapter->clients[i] = NULL;
+			goto out_unlock;
+		}
 	}
 
-	if (I2C_CLIENT_MAX == i) {
-		printk(KERN_WARNING " i2c-core.o: unregister_client "
-				    "[%s] not found\n",
+	printk(KERN_WARNING
+	       " i2c-core.o: unregister_client [%s] not found\n",
 	       client->name);
-		return -ENODEV;
-	} else
-		adapter->clients[i] = NULL;
-	up(&adapter->list);
+	res = -ENODEV;
 
-	return 0;
+ out_unlock:
+	up(&adapter->list);
+ out:
+	return res;
 }
 
 static int i2c_inc_use_client(struct i2c_client *client)
@@ -456,45 +462,7 @@ int i2c_release_client(struct i2c_client *client)
 	return 0;
 }
 
-/* ----------------------------------------------------
- * The /proc functions
- * ----------------------------------------------------
- */
-
 #ifdef CONFIG_PROC_FS
-/* This function generates the output for /proc/bus/i2c */
-static int read_bus_i2c(char *buf, char **start, off_t offset,
-			int len, int *eof, void *private)
-{
-	int i;
-	int nr = 0;
-
-	/* Note that it is safe to write a `little' beyond len. Yes, really. */
-	/* Fuck you.  Will convert this to seq_file later.  --hch */
-
-	down(&core_lists);
-	for (i = 0; (i < I2C_ADAP_MAX) && (nr < len); i++) {
-		if (adapters[i]) {
-			nr += sprintf(buf+nr, "i2c-%d\t", i);
-			if (adapters[i]->algo->smbus_xfer) {
-				if (adapters[i]->algo->master_xfer)
-					nr += sprintf(buf+nr,"smbus/i2c");
-				else
-					nr += sprintf(buf+nr,"smbus    ");
-			} else if (adapters[i]->algo->master_xfer)
-				nr += sprintf(buf+nr,"i2c       ");
-			else
-				nr += sprintf(buf+nr,"dummy     ");
-			nr += sprintf(buf+nr,"\t%-32s\t%-32s\n",
-			              adapters[i]->name,
-			              adapters[i]->algo->name);
-		}
-	}
-	up(&core_lists);
-
-	return nr;
-}
-
 /* This function generates the output for /proc/bus/i2c-? */
 static ssize_t i2cproc_bus_read(struct file *file, char *buf,
 				size_t count, loff_t *ppos)
@@ -564,6 +532,50 @@ static struct file_operations i2cproc_operations = {
 	.read		= i2cproc_bus_read,
 };
 
+/* This function generates the output for /proc/bus/i2c */
+static int bus_i2c_show(struct seq_file *s, void *p)
+{
+	int i;
+
+	down(&core_lists);
+	for (i = 0; i < I2C_ADAP_MAX; i++) {
+		struct i2c_adapter *adapter = adapters[i];
+
+		if (!adapter)
+			continue;
+
+		seq_printf(s, "i2c-%d\t", i);
+
+		if (adapter->algo->smbus_xfer) {
+			if (adapter->algo->master_xfer)
+				seq_printf(s, "smbus/i2c");
+			else
+				seq_printf(s, "smbus    ");
+		} else if (adapter->algo->master_xfer)
+			seq_printf(s ,"i2c       ");
+		else
+			seq_printf(s, "dummy     ");
+
+		seq_printf(s, "\t%-32s\t%-32s\n",
+			      adapter->name, adapter->algo->name);
+	}
+	up(&core_lists);
+
+	return 0;
+}
+
+static int bus_i2c_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, bus_i2c_show, NULL);
+}
+
+static struct file_operations bus_i2c_fops = {
+	.open		= bus_i2c_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+ };
+
 static int i2cproc_register(struct i2c_adapter *adap, int bus)
 {
 	struct proc_dir_entry *proc_entry;
@@ -576,7 +588,7 @@ static int i2cproc_register(struct i2c_adapter *adap, int bus)
 		goto fail;
 
 	proc_entry->proc_fops = &i2cproc_operations;
-	proc_entry->owner = THIS_MODULE;
+	proc_entry->owner = adap->owner;
 	adap->inode = proc_entry->low_ino;
 	return 0;
  fail:
@@ -596,15 +608,16 @@ static int __init i2cproc_init(void)
 {
 	struct proc_dir_entry *proc_bus_i2c;
 
-	proc_bus_i2c = create_proc_entry("i2c",0,proc_bus);
-	if (!proc_bus_i2c) {
-		printk(KERN_ERR "i2c-core.o: Could not create /proc/bus/i2c");
-		return -ENOENT;
- 	}
-
-	proc_bus_i2c->read_proc = &read_bus_i2c;
+	proc_bus_i2c = create_proc_entry("i2c", 0, proc_bus);
+	if (!proc_bus_i2c)
+		goto fail;
+	proc_bus_i2c->proc_fops = &bus_i2c_fops;
  	proc_bus_i2c->owner = THIS_MODULE;
  	return 0;
+
+ fail:
+	printk(KERN_ERR "i2c-core.o: Could not create /proc/bus/i2c");
+	return -ENOENT;
 }
 
 static void __exit i2cproc_cleanup(void)

drivers/i2c/i2c-proc.c

@@ -35,8 +35,6 @@
 #include <linux/i2c-proc.h>
 #include <asm/uaccess.h>
 
-static int i2c_create_name(char **name, const char *prefix,
-			       struct i2c_adapter *adapter, int addr);
 static int i2c_parse_reals(int *nrels, void *buffer, int bufsize,
 			       long *results, int magnitude);
 static int i2c_write_reals(int nrels, void *buffer, size_t *bufsize,
@@ -54,15 +52,6 @@ static struct ctl_table_header *i2c_entries[SENSORS_ENTRY_MAX];
 
 static struct i2c_client *i2c_clients[SENSORS_ENTRY_MAX];
 
-static ctl_table sysctl_table[] = {
-	{CTL_DEV, "dev", NULL, 0, 0555},
-	{0},
-	{DEV_SENSORS, "sensors", NULL, 0, 0555},
-	{0},
-	{0, NULL, NULL, 0, 0555},
-	{0}
-};
-
 static ctl_table i2c_proc_dev_sensors[] = {
 	{SENSORS_CHIPS, "chips", NULL, 0, 0644, NULL, &i2c_proc_chips,
 	 &i2c_sysctl_chips},
@@ -87,36 +76,40 @@ static struct ctl_table_header *i2c_proc_header;
    (for a LM78 chip on the ISA bus at port 0x310), or lm75-i2c-3-4e (for
    a LM75 chip on the third i2c bus at address 0x4e).  
    name is allocated first. */
-static int i2c_create_name(char **name, const char *prefix,
-			struct i2c_adapter *adapter, int addr)
+static char *generate_name(struct i2c_client *client, const char *prefix)
 {
-	char name_buffer[50];
-	int id, i, end;
-	if (i2c_is_isa_adapter(adapter))
+	struct i2c_adapter *adapter = client->adapter;
+	int addr = client->addr;
+	char name_buffer[50], *name;
+
+	if (i2c_is_isa_adapter(adapter)) {
 		sprintf(name_buffer, "%s-isa-%04x", prefix, addr);
-	else if (!adapter->algo->smbus_xfer && !adapter->algo->master_xfer) {
-		/* dummy adapter, generate prefix */
+	} else if (adapter->algo->smbus_xfer || adapter->algo->master_xfer) {
+		int id = i2c_adapter_id(adapter);
+		if (id < 0)
+			return ERR_PTR(-ENOENT);
+		sprintf(name_buffer, "%s-i2c-%d-%02x", prefix, id, addr);
+	} else {	/* dummy adapter, generate prefix */
+		int end, i;
+
 		sprintf(name_buffer, "%s-", prefix);
 		end = strlen(name_buffer);
-		for(i = 0; i < 32; i++) {
-			if(adapter->algo->name[i] == ' ')
+
+		for (i = 0; i < 32; i++) {
+			if (adapter->algo->name[i] == ' ')
 				break;
 			name_buffer[end++] = tolower(adapter->algo->name[i]);
 		}
+
 		name_buffer[end] = 0;
 		sprintf(name_buffer + end, "-%04x", addr);
-	} else {
-		if ((id = i2c_adapter_id(adapter)) < 0)
-			return -ENOENT;
-		sprintf(name_buffer, "%s-i2c-%d-%02x", prefix, id, addr);
 	}
-	*name = kmalloc(strlen(name_buffer) + 1, GFP_KERNEL);
-	if (!*name) {
-		printk (KERN_WARNING "i2c_create_name: not enough memory\n");
-		return -ENOMEM;
-	}
-	strcpy(*name, name_buffer);
-	return 0;
+
+	name = kmalloc(strlen(name_buffer) + 1, GFP_KERNEL);
+	if (unlikely(!name))
+		return ERR_PTR(-ENOMEM);
+	strcpy(name, name_buffer);
+	return name;
 }
 
 /* This rather complex function must be called when you want to add an entry
@@ -127,93 +120,80 @@ static int i2c_create_name(char **name, const char *prefix,
    If any driver wants subdirectories within the newly created directory,
    this function must be updated!  */
 int i2c_register_entry(struct i2c_client *client, const char *prefix,
-			   ctl_table * ctl_template)
+		       struct ctl_table *leaf)
 {
-	int i, res, len, id;
-	ctl_table *new_table, *client_tbl, *tbl;
-	char *name;
-	struct ctl_table_header *new_header;
+	struct { struct ctl_table root[2], dev[2], sensors[2]; } *tbl;
+	struct ctl_table_header *hdr;
+	struct ctl_table *tmp;
+	const char *name;
+	int id;
 
-	if ((res = i2c_create_name(&name, prefix, client->adapter,
-				       client->addr))) return res;
+	name = generate_name(client, prefix);
+	if (IS_ERR(name))
+		return PTR_ERR(name);
 
-	for (id = 0; id < SENSORS_ENTRY_MAX; id++)
-		if (!i2c_entries[id]) {
-			break;
-		}
-	if (id == SENSORS_ENTRY_MAX) {
-		kfree(name);
-		return -ENOMEM;
+	for (id = 0; id < SENSORS_ENTRY_MAX; id++) {
+		if (!i2c_entries[id])
+			goto free_slot;
 	}
 
-	id += 256;
+	goto out_free_name;
 
-	len = 0;
-	while (ctl_template[len].procname)
-		len++;
-	if (!(new_table = kmalloc(sizeof(sysctl_table) + sizeof(ctl_table) * (len + 1), 
-				  GFP_KERNEL))) {
-		kfree(name);
-		return -ENOMEM;
-	}
+ free_slot:
+	tbl = kmalloc(sizeof(*tbl), GFP_KERNEL);
+	if (unlikely(!tbl))
+		goto out_free_name;
+	memset(tbl, 0, sizeof(*tbl));
 
-	memcpy(new_table, sysctl_table, sizeof(sysctl_table));
-	tbl = new_table; /* sys/ */
-	tbl = tbl->child = tbl + 2; /* dev/ */
-	tbl = tbl->child = tbl + 2; /* sensors/ */	
-       	client_tbl = tbl->child = tbl + 2; /* XX-chip-YY-ZZ/ */
+	for (tmp = leaf; tmp->ctl_name; tmp++)
+		tmp->extra2 = client;
 
-	client_tbl->procname = name;
-	client_tbl->ctl_name = id;
-	client_tbl->child = client_tbl + 2;
+	tbl->sensors->ctl_name = id+256;
+	tbl->sensors->procname = name;
+	tbl->sensors->mode = 0555;
+	tbl->sensors->child = leaf;
 
-	/* Next the client sysctls. --km */
-	tbl = client_tbl->child;
-	memcpy(tbl, ctl_template, sizeof(ctl_table) * (len+1));
-	for (i = 0; i < len; i++)
-		tbl[i].extra2 = client;
+	tbl->dev->ctl_name = DEV_SENSORS;
+	tbl->dev->procname = "sensors";
+	tbl->dev->mode = 0555;
+	tbl->dev->child = tbl->sensors;
 
-	if (!(new_header = register_sysctl_table(new_table, 0))) {
-		printk(KERN_ERR "i2c-proc.o: error: sysctl interface not supported by kernel!\n");
-		kfree(new_table);
-		kfree(name);
-		return -EPERM;
-	}
+	tbl->root->ctl_name = CTL_DEV;
+	tbl->root->procname = "dev";
+	tbl->root->mode = 0555;
+	tbl->root->child = tbl->dev;
 
- 	i2c_entries[id - 256] = new_header;
+	hdr = register_sysctl_table(tbl->root, 0);
+	if (unlikely(!hdr))
+		goto out_free_tbl;
 
-	i2c_clients[id - 256] = client;
+	i2c_entries[id] = hdr;
+	i2c_clients[id] = client;
 
-#ifdef DEBUG
-	if (!new_header || !new_header->ctl_table ||
-	    !new_header->ctl_table->child ||
-	    !new_header->ctl_table->child->child ||
-	    !new_header->ctl_table->child->child->de ) {
-		printk
-		    (KERN_ERR "i2c-proc.o: NULL pointer when trying to install fill_inode fix!\n");
-		return id;
-	}
-#endif				/* DEBUG */
-	client_tbl->de->owner = client->driver->owner;
-	return id;
+	return (id + 256);	/* XXX(hch) why?? */
+
+ out_free_tbl:
+	kfree(tbl);
+ out_free_name:
+	kfree(name);
+	return -ENOMEM;
 }
 
 void i2c_deregister_entry(int id)
 {
-	ctl_table *table;
-	char *temp;
-
 	id -= 256;
+
 	if (i2c_entries[id]) {
-		table = i2c_entries[id]->ctl_table;
-		unregister_sysctl_table(i2c_entries[id]);
-		/* 2-step kfree needed to keep gcc happy about const points */
-		(const char *) temp = table[4].procname;
-		kfree(temp);
-		kfree(table);
+		struct ctl_table_header *hdr = i2c_entries[id];
+		struct ctl_table *tbl = hdr->ctl_table;
+
+		unregister_sysctl_table(hdr);
+		kfree(tbl->child->child->procname);
+		kfree(tbl); /* actually the whole anonymous struct */
+	}
+
 	i2c_entries[id] = NULL;
 	i2c_clients[id] = NULL;
-	}
 }
 
 static int i2c_proc_chips(ctl_table * ctl, int write, struct file *filp,

drivers/md/linear.c

@@ -203,36 +203,34 @@ static int linear_make_request (request_queue_t *q, struct bio *bio)
 		return 0;
 	}
 	bio->bi_bdev = tmp_dev->rdev->bdev;
-	bio->bi_sector = bio->bi_sector - (tmp_dev->offset << 1);
+	bio->bi_sector = bio->bi_sector - (tmp_dev->offset << 1) + tmp_dev->rdev->data_offset;
 
 	return 1;
 }
 
-static int linear_status (char *page, mddev_t *mddev)
+static void linear_status (struct seq_file *seq, mddev_t *mddev)
 {
-	int sz = 0;
 
 #undef MD_DEBUG
 #ifdef MD_DEBUG
 	int j;
 	linear_conf_t *conf = mddev_to_conf(mddev);
   
-	sz += sprintf(page+sz, "      ");
+	seq_printf(seq, "      ");
 	for (j = 0; j < conf->nr_zones; j++)
 	{
-		sz += sprintf(page+sz, "[%s",
+		seq_printf(seq, "[%s",
 			bdev_partition_name(conf->hash_table[j].dev0->rdev->bdev));
 
 		if (conf->hash_table[j].dev1)
-			sz += sprintf(page+sz, "/%s] ",
+			seq_printf(seq, "/%s] ",
 			  bdev_partition_name(conf->hash_table[j].dev1->rdev->bdev));
 		else
-			sz += sprintf(page+sz, "] ");
+			seq_printf(seq, "] ");
 	}
-	sz += sprintf(page+sz, "\n");
+	seq_printf(seq, "\n");
 #endif
-	sz += sprintf(page+sz, " %dk rounding", mddev->chunk_size/1024);
-	return sz;
+	seq_printf(seq, " %dk rounding", mddev->chunk_size/1024);
 }
 
 

drivers/md/md.c

@@ -124,9 +124,6 @@ static ctl_table raid_root_table[] = {
 	{ .ctl_name = 0 }
 };
 
-static void md_recover_arrays(void);
-static mdk_thread_t *md_recovery_thread;
-
 sector_t md_size[MAX_MD_DEVS];
 
 static struct block_device_operations md_fops;
@@ -222,6 +219,7 @@ static mddev_t * mddev_find(int unit)
 	init_MUTEX(&new->reconfig_sem);
 	INIT_LIST_HEAD(&new->disks);
 	INIT_LIST_HEAD(&new->all_mddevs);
+	init_timer(&new->safemode_timer);
 	atomic_set(&new->active, 1);
 	blk_queue_make_request(&new->queue, md_fail_request);
 
@@ -272,40 +270,35 @@ static mdk_rdev_t * find_rdev(mddev_t * mddev, dev_t dev)
 	return NULL;
 }
 
-static sector_t calc_dev_sboffset(struct block_device *bdev)
+inline static sector_t calc_dev_sboffset(struct block_device *bdev)
 {
 	sector_t size = bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
 	return MD_NEW_SIZE_BLOCKS(size);
 }
 
-static sector_t calc_dev_size(struct block_device *bdev, mddev_t *mddev)
+static sector_t calc_dev_size(mdk_rdev_t *rdev, unsigned chunk_size)
 {
 	sector_t size;
 
-	if (mddev->persistent)
-		size = calc_dev_sboffset(bdev);
-	else
-		size = bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
-	if (mddev->chunk_size)
-		size &= ~((sector_t)mddev->chunk_size/1024 - 1);
+	size = rdev->sb_offset;
+
+	if (chunk_size)
+		size &= ~((sector_t)chunk_size/1024 - 1);
 	return size;
 }
 
 static sector_t zoned_raid_size(mddev_t *mddev)
 {
-	sector_t mask;
 	mdk_rdev_t * rdev;
 	struct list_head *tmp;
 
 	/*
 	 * do size and offset calculations.
 	 */
-	mask = ~((sector_t)mddev->chunk_size/1024 - 1);
 
-	ITERATE_RDEV(mddev,rdev,tmp) {
-		rdev->size &= mask;
+	ITERATE_RDEV(mddev,rdev,tmp)
 		md_size[mdidx(mddev)] += rdev->size;
-	}
+
 	return 0;
 }
 
@@ -389,7 +382,6 @@ static int sync_page_io(struct block_device *bdev, sector_t sector, int size,
 
 static int read_disk_sb(mdk_rdev_t * rdev)
 {
-	sector_t sb_offset;
 
 	if (!rdev->sb_page) {
 		MD_BUG();
@@ -398,16 +390,8 @@ static int read_disk_sb(mdk_rdev_t * rdev)
 	if (rdev->sb_loaded)
 		return 0;
 
-	/*
-	 * Calculate the position of the superblock,
-	 * it's at the end of the disk.
-	 *
-	 * It also happens to be a multiple of 4Kb.
-	 */
-	sb_offset = calc_dev_sboffset(rdev->bdev);
-	rdev->sb_offset = sb_offset;
 
-	if (!sync_page_io(rdev->bdev, sb_offset<<1, MD_SB_BYTES, rdev->sb_page, READ))
+	if (!sync_page_io(rdev->bdev, rdev->sb_offset<<1, MD_SB_BYTES, rdev->sb_page, READ))
 		goto fail;
 	rdev->sb_loaded = 1;
 	return 0;
@@ -486,7 +470,7 @@ static unsigned int calc_sb_csum(mdp_super_t * sb)
  * We rely on user-space to write the initial superblock, and support
  * reading and updating of superblocks.
  * Interface methods are:
- *   int load_super(mdk_rdev_t *dev, mdk_rdev_t *refdev)
+ *   int load_super(mdk_rdev_t *dev, mdk_rdev_t *refdev, int minor_version)
  *      loads and validates a superblock on dev.
  *      if refdev != NULL, compare superblocks on both devices
  *    Return:
@@ -511,7 +495,7 @@ static unsigned int calc_sb_csum(mdp_super_t * sb)
 struct super_type  {
 	char 		*name;
 	struct module	*owner;
-	int		(*load_super)(mdk_rdev_t *rdev, mdk_rdev_t *refdev);
+	int		(*load_super)(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version);
 	int		(*validate_super)(mddev_t *mddev, mdk_rdev_t *rdev);
 	void		(*sync_super)(mddev_t *mddev, mdk_rdev_t *rdev);
 };
@@ -519,10 +503,20 @@ struct super_type  {
 /*
  * load_super for 0.90.0 
  */
-static int super_90_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev)
+static int super_90_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version)
 {
 	mdp_super_t *sb;
 	int ret;
+	sector_t sb_offset;
+
+	/*
+	 * Calculate the position of the superblock,
+	 * it's at the end of the disk.
+	 *
+	 * It also happens to be a multiple of 4Kb.
+	 */
+	sb_offset = calc_dev_sboffset(rdev->bdev);
+	rdev->sb_offset = sb_offset;
 
 	ret = read_disk_sb(rdev);
 	if (ret) return ret;
@@ -557,6 +551,12 @@ static int super_90_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev)
 	}
 
 	rdev->preferred_minor = sb->md_minor;
+	rdev->data_offset = 0;
+
+	if (sb->level == MULTIPATH)
+		rdev->desc_nr = -1;
+	else
+		rdev->desc_nr = sb->this_disk.number;
 
 	if (refdev == 0)
 		ret = 1;
@@ -582,7 +582,7 @@ static int super_90_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev)
 		else 
 			ret = 0;
 	}
-
+	rdev->size = calc_dev_size(rdev, sb->chunk_size);
 
  abort:
 	return ret;
@@ -597,7 +597,7 @@ static int super_90_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 	mdp_super_t *sb = (mdp_super_t *)page_address(rdev->sb_page);
 
 	if (mddev->raid_disks == 0) {
-		mddev->major_version = sb->major_version;
+		mddev->major_version = 0;
 		mddev->minor_version = sb->minor_version;
 		mddev->patch_version = sb->patch_version;
 		mddev->persistent = ! sb->not_persistent;
@@ -634,7 +634,6 @@ static int super_90_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 			return -EINVAL;
 	}
 	if (mddev->level != LEVEL_MULTIPATH) {
-		rdev->desc_nr = sb->this_disk.number;
 		rdev->raid_disk = -1;
 		rdev->in_sync = rdev->faulty = 0;
 		desc = sb->disks + rdev->desc_nr;
@@ -704,10 +703,8 @@ static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 		sb->recovery_cp = mddev->recovery_cp;
 		sb->cp_events_hi = (mddev->events>>32);
 		sb->cp_events_lo = (u32)mddev->events;
-		if (mddev->recovery_cp == MaxSector) {
-			printk(KERN_INFO "md: marking sb clean...\n");
+		if (mddev->recovery_cp == MaxSector)
 			sb->state = (1<< MD_SB_CLEAN);
-		}
 	} else
 		sb->recovery_cp = 0;
 
@@ -717,7 +714,7 @@ static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 	sb->disks[0].state = (1<<MD_DISK_REMOVED);
 	ITERATE_RDEV(mddev,rdev2,tmp) {
 		mdp_disk_t *d;
-		if (rdev2->raid_disk >= 0)
+		if (rdev2->raid_disk >= 0 && rdev2->in_sync && !rdev2->faulty)
 			rdev2->desc_nr = rdev2->raid_disk;
 		else
 			rdev2->desc_nr = next_spare++;
@@ -726,7 +723,7 @@ static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 		d->number = rdev2->desc_nr;
 		d->major = MAJOR(rdev2->bdev->bd_dev);
 		d->minor = MINOR(rdev2->bdev->bd_dev);
-		if (rdev2->raid_disk >= 0)
+		if (rdev2->raid_disk >= 0 && rdev->in_sync && !rdev2->faulty)
 			d->raid_disk = rdev2->raid_disk;
 		else
 			d->raid_disk = rdev2->desc_nr; /* compatibility */
@@ -766,6 +763,210 @@ static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 	sb->sb_csum = calc_sb_csum(sb);
 }
 
+/*
+ * version 1 superblock
+ */
+
+static unsigned int calc_sb_1_csum(struct mdp_superblock_1 * sb)
+{
+	unsigned int disk_csum, csum;
+	int size = 256 + sb->max_dev*2;
+
+	disk_csum = sb->sb_csum;
+	sb->sb_csum = 0;
+	csum = csum_partial((void *)sb, size, 0);
+	sb->sb_csum = disk_csum;
+	return csum;
+}
+
+static int super_1_load(mdk_rdev_t *rdev, mdk_rdev_t *refdev, int minor_version)
+{
+	struct mdp_superblock_1 *sb;
+	int ret;
+	sector_t sb_offset;
+
+	/*
+	 * Calculate the position of the superblock.
+	 * It is always aligned to a 4K boundary and
+	 * depeding on minor_version, it can be:
+	 * 0: At least 8K, but less than 12K, from end of device
+	 * 1: At start of device
+	 * 2: 4K from start of device.
+	 */
+	switch(minor_version) {
+	case 0:
+		sb_offset = rdev->bdev->bd_inode->i_size >> 9;
+		sb_offset -= 8*2;
+		sb_offset &= ~(4*2);
+		/* convert from sectors to K */
+		sb_offset /= 2;
+		break;
+	case 1:
+		sb_offset = 0;
+		break;
+	case 2:
+		sb_offset = 4;
+		break;
+	default:
+		return -EINVAL;
+	}
+	rdev->sb_offset = sb_offset;
+
+	ret = read_disk_sb(rdev);
+	if (ret) return ret;
+
+
+	sb = (struct mdp_superblock_1*)page_address(rdev->sb_page);
+
+	if (sb->magic != cpu_to_le32(MD_SB_MAGIC) ||
+	    sb->major_version != cpu_to_le32(1) ||
+	    le32_to_cpu(sb->max_dev) > (4096-256)/2 ||
+	    le64_to_cpu(sb->super_offset) != (rdev->sb_offset<<1) ||
+	    sb->feature_map != 0)
+		return -EINVAL;
+
+	if (calc_sb_1_csum(sb) != sb->sb_csum) {
+		printk(BAD_CSUM, bdev_partition_name(rdev->bdev));
+		return -EINVAL;
+	}
+	rdev->preferred_minor = 0xffff;
+	rdev->data_offset = le64_to_cpu(sb->data_offset);
+
+	if (refdev == 0)
+		return 1;
+	else {
+		__u64 ev1, ev2;
+		struct mdp_superblock_1 *refsb = 
+			(struct mdp_superblock_1*)page_address(refdev->sb_page);
+
+		if (memcmp(sb->set_uuid, refsb->set_uuid, 16) != 0 ||
+		    sb->level != refsb->level ||
+		    sb->layout != refsb->layout ||
+		    sb->chunksize != refsb->chunksize) {
+			printk(KERN_WARNING "md: %s has strangely different superblock to %s\n",
+			       bdev_partition_name(rdev->bdev),
+			       bdev_partition_name(refdev->bdev));
+			return -EINVAL;
+		}
+		ev1 = le64_to_cpu(sb->events);
+		ev2 = le64_to_cpu(refsb->events);
+
+		if (ev1 > ev2)
+			return 1;
+	}
+	if (minor_version) 
+		rdev->size = ((rdev->bdev->bd_inode->i_size>>9) - le64_to_cpu(sb->data_offset)) / 2;
+	else
+		rdev->size = rdev->sb_offset;
+	if (rdev->size < le64_to_cpu(sb->data_size)/2)
+		return -EINVAL;
+	rdev->size = le64_to_cpu(sb->data_size)/2;
+	if (le32_to_cpu(sb->chunksize))
+		rdev->size &= ~((sector_t)le32_to_cpu(sb->chunksize)/2 - 1);
+	return 0;
+}
+
+static int super_1_validate(mddev_t *mddev, mdk_rdev_t *rdev)
+{
+	struct mdp_superblock_1 *sb = (struct mdp_superblock_1*)page_address(rdev->sb_page);
+
+	if (mddev->raid_disks == 0) {
+		mddev->major_version = 1;
+		mddev->minor_version = 0;
+		mddev->patch_version = 0;
+		mddev->persistent = 1;
+		mddev->chunk_size = le32_to_cpu(sb->chunksize) << 9;
+		mddev->ctime = le64_to_cpu(sb->ctime) & ((1ULL << 32)-1);
+		mddev->utime = le64_to_cpu(sb->utime) & ((1ULL << 32)-1);
+		mddev->level = le32_to_cpu(sb->level);
+		mddev->layout = le32_to_cpu(sb->layout);
+		mddev->raid_disks = le32_to_cpu(sb->raid_disks);
+		mddev->size = (u32)le64_to_cpu(sb->size);
+		mddev->events = le64_to_cpu(sb->events);
+		
+		mddev->recovery_cp = le64_to_cpu(sb->resync_offset);
+		memcpy(mddev->uuid, sb->set_uuid, 16);
+
+		mddev->max_disks =  (4096-256)/2;
+	} else {
+		__u64 ev1;
+		ev1 = le64_to_cpu(sb->events);
+		++ev1;
+		if (ev1 < mddev->events)
+			return -EINVAL;
+	}
+
+	if (mddev->level != LEVEL_MULTIPATH) {
+		int role;
+		rdev->desc_nr = le32_to_cpu(sb->dev_number);
+		role = le16_to_cpu(sb->dev_roles[rdev->desc_nr]);
+		switch(role) {
+		case 0xffff: /* spare */
+			rdev->in_sync = 0;
+			rdev->faulty = 0;
+			rdev->raid_disk = -1;
+			break;
+		case 0xfffe: /* faulty */
+			rdev->in_sync = 0;
+			rdev->faulty = 1;
+			rdev->raid_disk = -1;
+			break;
+		default:
+			rdev->in_sync = 1;
+			rdev->faulty = 0;
+			rdev->raid_disk = role;
+			break;
+		}
+	}
+	return 0;
+}
+
+static void super_1_sync(mddev_t *mddev, mdk_rdev_t *rdev)
+{
+	struct mdp_superblock_1 *sb;
+	struct list_head *tmp;
+	mdk_rdev_t *rdev2;
+	int max_dev, i;
+	/* make rdev->sb match mddev and rdev data. */
+
+	sb = (struct mdp_superblock_1*)page_address(rdev->sb_page);
+
+	sb->feature_map = 0;
+	sb->pad0 = 0;
+	memset(sb->pad1, 0, sizeof(sb->pad1));
+	memset(sb->pad2, 0, sizeof(sb->pad2));
+	memset(sb->pad3, 0, sizeof(sb->pad3));
+
+	sb->utime = cpu_to_le64((__u64)mddev->utime);
+	sb->events = cpu_to_le64(mddev->events);
+	if (mddev->in_sync)
+		sb->resync_offset = cpu_to_le64(mddev->recovery_cp);
+	else
+		sb->resync_offset = cpu_to_le64(0);
+
+	max_dev = 0;
+	ITERATE_RDEV(mddev,rdev2,tmp)
+		if (rdev2->desc_nr > max_dev)
+			max_dev = rdev2->desc_nr;
+	
+	sb->max_dev = max_dev;
+	for (i=0; i<max_dev;i++)
+		sb->dev_roles[max_dev] = cpu_to_le16(0xfffe);
+	
+	ITERATE_RDEV(mddev,rdev2,tmp) {
+		i = rdev2->desc_nr;
+		if (rdev2->faulty)
+			sb->dev_roles[i] = cpu_to_le16(0xfffe);
+		else if (rdev2->in_sync)
+			sb->dev_roles[i] = cpu_to_le16(rdev2->raid_disk);
+		else
+			sb->dev_roles[i] = cpu_to_le16(0xffff);
+	}
+
+	sb->recovery_offset = cpu_to_le64(0); /* not supported yet */
+}
+
+
 struct super_type super_types[] = {
 	[0] = {
 		.name	= "0.90.0",
@@ -774,10 +975,15 @@ struct super_type super_types[] = {
 		.validate_super	= super_90_validate,
 		.sync_super	= super_90_sync,
 	},
+	[1] = {
+		.name	= "md-1",
+		.owner	= THIS_MODULE,
+		.load_super	= super_1_load,
+		.validate_super	= super_1_validate,
+		.sync_super	= super_1_sync,
+	},
 };
 	
-
-	
 static mdk_rdev_t * match_dev_unit(mddev_t *mddev, mdk_rdev_t *dev)
 {
 	struct list_head *tmp;
@@ -804,13 +1010,13 @@ static int match_mddev_units(mddev_t *mddev1, mddev_t *mddev2)
 
 static LIST_HEAD(pending_raid_disks);
 
-static void bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
+static int bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
 {
 	mdk_rdev_t *same_pdev;
 
 	if (rdev->mddev) {
 		MD_BUG();
-		return;
+		return -EINVAL;
 	}
 	same_pdev = match_dev_unit(mddev, rdev);
 	if (same_pdev)
@@ -820,9 +1026,25 @@ static void bind_rdev_to_array(mdk_rdev_t * rdev, mddev_t * mddev)
 			mdidx(mddev), bdev_partition_name(rdev->bdev),
 				bdev_partition_name(same_pdev->bdev));
 
+	/* Verify rdev->desc_nr is unique.
+	 * If it is -1, assign a free number, else
+	 * check number is not in use
+	 */
+	if (rdev->desc_nr < 0) {
+		int choice = 0;
+		if (mddev->pers) choice = mddev->raid_disks;
+		while (find_rdev_nr(mddev, choice))
+			choice++;
+		rdev->desc_nr = choice;
+	} else {
+		if (find_rdev_nr(mddev, rdev->desc_nr))
+			return -EBUSY;
+	}
+			
 	list_add(&rdev->same_set, &mddev->disks);
 	rdev->mddev = mddev;
 	printk(KERN_INFO "md: bind<%s>\n", bdev_partition_name(rdev->bdev));
+	return 0;
 }
 
 static void unbind_rdev_from_array(mdk_rdev_t * rdev)
@@ -910,6 +1132,7 @@ static void export_array(mddev_t *mddev)
 	if (!list_empty(&mddev->disks))
 		MD_BUG();
 	mddev->raid_disks = 0;
+	mddev->major_version = 0;
 }
 
 #undef BAD_CSUM
@@ -994,8 +1217,6 @@ void md_print_devices(void)
 
 static int write_disk_sb(mdk_rdev_t * rdev)
 {
-	sector_t sb_offset;
-	sector_t size;
 
 	if (!rdev->sb_loaded) {
 		MD_BUG();
@@ -1006,35 +1227,12 @@ static int write_disk_sb(mdk_rdev_t * rdev)
 		return 1;
 	}
 
-	sb_offset = calc_dev_sboffset(rdev->bdev);
-	if (rdev->sb_offset != sb_offset) {
-		printk(KERN_INFO "%s's sb offset has changed from %llu to %llu, skipping\n",
-		       bdev_partition_name(rdev->bdev), 
-		    (unsigned long long)rdev->sb_offset, 
-		    (unsigned long long)sb_offset);
-		goto skip;
-	}
-	/*
-	 * If the disk went offline meanwhile and it's just a spare, then
-	 * its size has changed to zero silently, and the MD code does
-	 * not yet know that it's faulty.
-	 */
-	size = calc_dev_size(rdev->bdev, rdev->mddev);
-	if (size != rdev->size) {
-		printk(KERN_INFO "%s's size has changed from %llu to %llu since import, skipping\n",
-		       bdev_partition_name(rdev->bdev),
-		       (unsigned long long)rdev->size, 
-		       (unsigned long long)size);
-		goto skip;
-	}
-
-	printk(KERN_INFO "(write) %s's sb offset: %llu\n", bdev_partition_name(rdev->bdev), (unsigned long long)sb_offset);
+	dprintk(KERN_INFO "(write) %s's sb offset: %llu\n", bdev_partition_name(rdev->bdev),
+	       (unsigned long long)rdev->sb_offset);
   
-	if (!sync_page_io(rdev->bdev, sb_offset<<1, MD_SB_BYTES, rdev->sb_page, WRITE))
-		goto fail;
-skip:
+	if (sync_page_io(rdev->bdev, rdev->sb_offset<<1, MD_SB_BYTES, rdev->sb_page, WRITE))
 		return 0;
-fail:
+
 	printk("md: write_disk_sb failed for device %s\n", bdev_partition_name(rdev->bdev));
 	return 1;
 }
@@ -1045,7 +1243,8 @@ static void sync_sbs(mddev_t * mddev)
 	struct list_head *tmp;
 
 	ITERATE_RDEV(mddev,rdev,tmp) {
-		super_90_sync(mddev, rdev);
+		super_types[mddev->major_version].
+			sync_super(mddev, rdev);
 		rdev->sb_loaded = 1;
 	}
 }
@@ -1079,20 +1278,20 @@ static void md_update_sb(mddev_t * mddev)
 	if (!mddev->persistent)
 		return;
 
-	printk(KERN_INFO "md: updating md%d RAID superblock on device (in sync %d)\n",
+	dprintk(KERN_INFO "md: updating md%d RAID superblock on device (in sync %d)\n",
 					mdidx(mddev),mddev->in_sync);
 
 	err = 0;
 	ITERATE_RDEV(mddev,rdev,tmp) {
-		printk(KERN_INFO "md: ");
+		dprintk(KERN_INFO "md: ");
 		if (rdev->faulty)
-			printk("(skipping faulty ");
+			dprintk("(skipping faulty ");
 
-		printk("%s ", bdev_partition_name(rdev->bdev));
+		dprintk("%s ", bdev_partition_name(rdev->bdev));
 		if (!rdev->faulty) {
 			err += write_disk_sb(rdev);
 		} else
-			printk(")\n");
+			dprintk(")\n");
 		if (!err && mddev->level == LEVEL_MULTIPATH)
 			/* only need to write one superblock... */
 			break;
@@ -1107,7 +1306,7 @@ static void md_update_sb(mddev_t * mddev)
 }
 
 /*
- * Import a device. If 'on_disk', then sanity check the superblock
+ * Import a device. If 'super_format' >= 0, then sanity check the superblock
  *
  * mark the device faulty if:
  *
@@ -1116,7 +1315,7 @@ static void md_update_sb(mddev_t * mddev)
  *
  * a faulty rdev _never_ has rdev->sb set.
  */
-static mdk_rdev_t *md_import_device(dev_t newdev, int on_disk)
+static mdk_rdev_t *md_import_device(dev_t newdev, int super_format, int super_minor)
 {
 	int err;
 	mdk_rdev_t *rdev;
@@ -1141,6 +1340,7 @@ static mdk_rdev_t *md_import_device(dev_t newdev, int on_disk)
 	rdev->desc_nr = -1;
 	rdev->faulty = 0;
 	rdev->in_sync = 0;
+	rdev->data_offset = 0;
 	atomic_set(&rdev->nr_pending, 0);
 
 	size = rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
@@ -1152,8 +1352,9 @@ static mdk_rdev_t *md_import_device(dev_t newdev, int on_disk)
 		goto abort_free;
 	}
 
-	if (on_disk) {
-		err = super_90_load(rdev, NULL);
+	if (super_format >= 0) {
+		err = super_types[super_format].
+			load_super(rdev, NULL, super_minor);
 		if (err == -EINVAL) {
 			printk(KERN_WARNING "md: %s has invalid sb, not importing!\n",
 			       bdev_partition_name(rdev->bdev));
@@ -1206,7 +1407,8 @@ static int analyze_sbs(mddev_t * mddev)
 
 	freshest = NULL;
 	ITERATE_RDEV(mddev,rdev,tmp)
-		switch (super_90_load(rdev, freshest)) {
+		switch (super_types[mddev->major_version].
+			load_super(rdev, freshest, mddev->minor_version)) {
 		case 1:
 			freshest = rdev;
 			break;
@@ -1218,12 +1420,14 @@ static int analyze_sbs(mddev_t * mddev)
 		}
 
 
-	super_90_validate(mddev, freshest);
+	super_types[mddev->major_version].
+		validate_super(mddev, freshest);
 
 	i = 0;
 	ITERATE_RDEV(mddev,rdev,tmp) {
 		if (rdev != freshest)
-			if (super_90_validate(mddev, rdev)) {
+			if (super_types[mddev->major_version].
+			    validate_super(mddev, rdev)) {
 				printk(KERN_WARNING "md: kicking non-fresh %s from array!\n",
 				       bdev_partition_name(rdev->bdev));
 				kick_rdev_from_array(rdev);
@@ -1278,11 +1482,6 @@ static int device_size_calculation(mddev_t * mddev)
 	ITERATE_RDEV(mddev,rdev,tmp) {
 		if (rdev->faulty)
 			continue;
-		if (rdev->size) {
-			MD_BUG();
-			continue;
-		}
-		rdev->size = calc_dev_size(rdev->bdev, mddev);
 		if (rdev->size < mddev->chunk_size / 1024) {
 			printk(KERN_WARNING
 				"md: Dev %s smaller than chunk_size: %lluk < %dk\n",
@@ -1380,6 +1579,16 @@ static struct gendisk *md_probe(dev_t dev, int *part, void *data)
 	return NULL;
 }
 
+void md_wakeup_thread(mdk_thread_t *thread);
+
+static void md_safemode_timeout(unsigned long data)
+{
+	mddev_t *mddev = (mddev_t *) data;
+
+	mddev->safemode = 1;
+	md_wakeup_thread(mddev->thread);
+}
+
 #define TOO_BIG_CHUNKSIZE KERN_ERR \
 "too big chunk_size: %d > %d\n"
 
@@ -1521,13 +1730,14 @@ static int do_md_run(mddev_t * mddev)
 	}
  	atomic_set(&mddev->writes_pending,0);
 	mddev->safemode = 0;
-	if (mddev->pers->sync_request)
-		mddev->in_sync = 0;
-	else
+	mddev->safemode_timer.function = md_safemode_timeout;
+	mddev->safemode_timer.data = (unsigned long) mddev;
+	mddev->safemode_delay = (20 * HZ)/1000 +1; /* 20 msec delay */
 	mddev->in_sync = 1;
 	
 	md_update_sb(mddev);
-	md_recover_arrays();
+	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+	md_wakeup_thread(mddev->thread);
 	set_capacity(disk, md_size[mdidx(mddev)]<<1);
 	return (0);
 }
@@ -1553,7 +1763,6 @@ static int restart_array(mddev_t *mddev)
 			goto out;
 
 		mddev->safemode = 0;
-		mddev->in_sync = 0;
 		md_update_sb(mddev);
 		mddev->ro = 0;
 		set_disk_ro(disk, 0);
@@ -1563,7 +1772,8 @@ static int restart_array(mddev_t *mddev)
 		/*
 		 * Kick recovery or resync if necessary
 		 */
-		md_recover_arrays();
+		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+		md_wakeup_thread(mddev->thread);
 		err = 0;
 	} else {
 		printk(KERN_ERR "md: md%d has no personality assigned.\n",
@@ -1593,12 +1803,13 @@ static int do_md_stop(mddev_t * mddev, int ro)
 
 	if (mddev->pers) {
 		if (mddev->sync_thread) {
-			if (mddev->recovery_running > 0)
-				mddev->recovery_running = -1;
+			set_bit(MD_RECOVERY_INTR, &mddev->recovery);
 			md_unregister_thread(mddev->sync_thread);
 			mddev->sync_thread = NULL;
 		}
 
+		del_timer_sync(&mddev->safemode_timer);
+
 		invalidate_device(mk_kdev(disk->major, disk->first_minor), 1);
 
 		if (ro) {
@@ -1699,7 +1910,7 @@ static void autorun_devices(void)
 		printk(KERN_INFO "md: considering %s ...\n", bdev_partition_name(rdev0->bdev));
 		INIT_LIST_HEAD(&candidates);
 		ITERATE_RDEV_PENDING(rdev,tmp)
-			if (super_90_load(rdev, rdev0) >= 0) {
+			if (super_90_load(rdev, rdev0, 0) >= 0) {
 				printk(KERN_INFO "md:  adding %s ...\n", bdev_partition_name(rdev->bdev));
 				list_move(&rdev->same_set, &candidates);
 			}
@@ -1717,7 +1928,8 @@ static void autorun_devices(void)
 		if (mddev_lock(mddev)) 
 			printk(KERN_WARNING "md: md%d locked, cannot run\n",
 			       mdidx(mddev));
-		else if (mddev->raid_disks || !list_empty(&mddev->disks)) {
+		else if (mddev->raid_disks || mddev->major_version
+			 || !list_empty(&mddev->disks)) {
 			printk(KERN_WARNING "md: md%d already running, cannot run %s\n",
 			       mdidx(mddev), bdev_partition_name(rdev0->bdev));
 			mddev_unlock(mddev);
@@ -1725,7 +1937,8 @@ static void autorun_devices(void)
 			printk(KERN_INFO "md: created md%d\n", mdidx(mddev));
 			ITERATE_RDEV_GENERIC(candidates,rdev,tmp) {
 				list_del_init(&rdev->same_set);
-				bind_rdev_to_array(rdev, mddev);
+				if (bind_rdev_to_array(rdev, mddev))
+					export_rdev(rdev);
 			}
 			autorun_array(mddev);
 			mddev_unlock(mddev);
@@ -1778,7 +1991,7 @@ static int autostart_array(dev_t startdev)
 	mdp_super_t *sb = NULL;
 	mdk_rdev_t *start_rdev = NULL, *rdev;
 
-	start_rdev = md_import_device(startdev, 1);
+	start_rdev = md_import_device(startdev, 0, 0);
 	if (IS_ERR(start_rdev)) {
 		printk(KERN_WARNING "md: could not import %s!\n", partition_name(startdev));
 		return err;
@@ -1812,7 +2025,7 @@ static int autostart_array(dev_t startdev)
 			continue;
 		if (dev == startdev)
 			continue;
-		rdev = md_import_device(dev, 1);
+		rdev = md_import_device(dev, 0, 0);
 		if (IS_ERR(rdev)) {
 			printk(KERN_WARNING "md: could not import %s, trying to run array nevertheless.\n",
 			       partition_name(dev));
@@ -1874,10 +2087,9 @@ static int get_array_info(mddev_t * mddev, void * arg)
 		}
 	}
 
-	info.major_version = mddev->major_version;
 	info.major_version = mddev->major_version;
 	info.minor_version = mddev->minor_version;
-	info.patch_version = mddev->patch_version;
+	info.patch_version = 1;
 	info.ctime         = mddev->ctime;
 	info.level         = mddev->level;
 	info.size          = mddev->size;
@@ -1888,7 +2100,7 @@ static int get_array_info(mddev_t * mddev, void * arg)
 
 	info.utime         = mddev->utime;
 	info.state         = 0;
-	if (mddev->recovery_cp == MaxSector)
+	if (mddev->in_sync)
 		info.state = (1<<MD_SB_CLEAN);
 	info.active_disks  = active;
 	info.working_disks = working;
@@ -1943,13 +2155,13 @@ static int get_disk_info(mddev_t * mddev, void * arg)
 
 static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
 {
-	sector_t size;
 	mdk_rdev_t *rdev;
 	dev_t dev;
 	dev = MKDEV(info->major,info->minor);
 	if (!mddev->raid_disks) {
+		int err;
 		/* expecting a device which has a superblock */
-		rdev = md_import_device(dev, 1);
+		rdev = md_import_device(dev, mddev->major_version, mddev->minor_version);
 		if (IS_ERR(rdev)) {
 			printk(KERN_WARNING "md: md_import_device returned %ld\n", PTR_ERR(rdev));
 			return PTR_ERR(rdev);
@@ -1957,7 +2169,8 @@ static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
 		if (!list_empty(&mddev->disks)) {
 			mdk_rdev_t *rdev0 = list_entry(mddev->disks.next,
 							mdk_rdev_t, same_set);
-			int err = super_90_load(rdev, rdev0);
+			int err = super_types[mddev->major_version]
+				.load_super(rdev, rdev0, mddev->minor_version);
 			if (err < 0) {
 				printk(KERN_WARNING "md: %s has different UUID to %s\n",
 				       bdev_partition_name(rdev->bdev), bdev_partition_name(rdev0->bdev));
@@ -1965,12 +2178,52 @@ static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
 				return -EINVAL;
 			}
 		}
-		bind_rdev_to_array(rdev, mddev);
-		return 0;
+		err = bind_rdev_to_array(rdev, mddev);
+		if (err)
+			export_rdev(rdev);
+		return err;
+	}
+
+	/*
+	 * add_new_disk can be used once the array is assembled
+	 * to add "hot spares".  They must already have a superblock
+	 * written
+	 */
+	if (mddev->pers) {
+		int err;
+		if (!mddev->pers->hot_add_disk) {
+			printk(KERN_WARNING "md%d: personality does not support diskops!\n",
+			       mdidx(mddev));
+			return -EINVAL;
+		}
+		rdev = md_import_device(dev, mddev->major_version,
+					mddev->minor_version);
+		if (IS_ERR(rdev)) {
+			printk(KERN_WARNING "md: md_import_device returned %ld\n", PTR_ERR(rdev));
+			return PTR_ERR(rdev);
+		}
+		rdev->in_sync = 0; /* just to be sure */
+		rdev->raid_disk = -1;
+		err = bind_rdev_to_array(rdev, mddev);
+		if (err)
+			export_rdev(rdev);
+		if (mddev->thread)
+			md_wakeup_thread(mddev->thread);
+		return err;
+	}
+
+	/* otherwise, add_new_disk is only allowed
+	 * for major_version==0 superblocks
+	 */
+	if (mddev->major_version != 0) {
+		printk(KERN_WARNING "md%d: ADD_NEW_DISK not supported\n",
+		       mdidx(mddev));
+		return -EINVAL;
 	}
 
 	if (!(info->state & (1<<MD_DISK_FAULTY))) {
-		rdev = md_import_device (dev, 0);
+		int err;
+		rdev = md_import_device (dev, -1, 0);
 		if (IS_ERR(rdev)) {
 			printk(KERN_WARNING "md: error, md_import_device() returned %ld\n", PTR_ERR(rdev));
 			return PTR_ERR(rdev);
@@ -1987,16 +2240,21 @@ static int add_new_disk(mddev_t * mddev, mdu_disk_info_t *info)
 		else
 			rdev->in_sync = 0;
 
-		bind_rdev_to_array(rdev, mddev);
+		err = bind_rdev_to_array(rdev, mddev);
+		if (err) {
+			export_rdev(rdev);
+			return err;
+		}
 
-		if (!mddev->persistent)
+		if (!mddev->persistent) {
 			printk(KERN_INFO "md: nonpersistent superblock ...\n");
-
-		size = calc_dev_size(rdev->bdev, mddev);
+			rdev->sb_offset = rdev->bdev->bd_inode->i_size >> BLOCK_SIZE_BITS;
+		} else 
 			rdev->sb_offset = calc_dev_sboffset(rdev->bdev);
+		rdev->size = calc_dev_size(rdev, mddev->chunk_size);
 
-		if (!mddev->size || (mddev->size > size))
-			mddev->size = size;
+		if (!mddev->size || (mddev->size > rdev->size))
+			mddev->size = rdev->size;
 	}
 
 	return 0;
@@ -2066,7 +2324,7 @@ static int hot_remove_disk(mddev_t * mddev, dev_t dev)
 
 static int hot_add_disk(mddev_t * mddev, dev_t dev)
 {
-	int i, err;
+	int err;
 	unsigned int size;
 	mdk_rdev_t *rdev;
 
@@ -2076,19 +2334,26 @@ static int hot_add_disk(mddev_t * mddev, dev_t dev)
 	printk(KERN_INFO "md: trying to hot-add %s to md%d ... \n",
 		partition_name(dev), mdidx(mddev));
 
+	if (mddev->major_version != 0) {
+		printk(KERN_WARNING "md%d: HOT_ADD may only be used with version-0 superblocks.\n",
+		       mdidx(mddev));
+		return -EINVAL;
+	}
 	if (!mddev->pers->hot_add_disk) {
 		printk(KERN_WARNING "md%d: personality does not support diskops!\n",
 		       mdidx(mddev));
 		return -EINVAL;
 	}
 
-	rdev = md_import_device (dev, 0);
+	rdev = md_import_device (dev, -1, 0);
 	if (IS_ERR(rdev)) {
 		printk(KERN_WARNING "md: error, md_import_device() returned %ld\n", PTR_ERR(rdev));
 		return -EINVAL;
 	}
 
-	size = calc_dev_size(rdev->bdev, mddev);
+	rdev->sb_offset = calc_dev_sboffset(rdev->bdev);
+	size = calc_dev_size(rdev, mddev->chunk_size);
+	rdev->size = size;
 
 	if (size < mddev->size) {
 		printk(KERN_WARNING "md%d: disk size %llu blocks < array size %llu\n",
@@ -2105,27 +2370,21 @@ static int hot_add_disk(mddev_t * mddev, dev_t dev)
 		goto abort_export;
 	}
 	rdev->in_sync = 0;
+	rdev->desc_nr = -1;
 	bind_rdev_to_array(rdev, mddev);
 
 	/*
 	 * The rest should better be atomic, we can have disk failures
 	 * noticed in interrupt contexts ...
 	 */
-	rdev->size = size;
-	rdev->sb_offset = calc_dev_sboffset(rdev->bdev);
-
-	for (i = mddev->raid_disks; i < mddev->max_disks; i++)
-		if (find_rdev_nr(mddev,i)==NULL)
-			break;
 
-	if (i == mddev->max_disks) {
+	if (rdev->desc_nr == mddev->max_disks) {
 		printk(KERN_WARNING "md%d: can not hot-add to full array!\n",
 		       mdidx(mddev));
 		err = -EBUSY;
 		goto abort_unbind_export;
 	}
 
-	rdev->desc_nr = i;
 	rdev->raid_disk = -1;
 
 	md_update_sb(mddev);
@@ -2134,7 +2393,8 @@ static int hot_add_disk(mddev_t * mddev, dev_t dev)
 	 * Kick recovery, maybe this spare has to be added to the
 	 * array immediately.
 	 */
-	md_recover_arrays();
+	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+	md_wakeup_thread(mddev->thread);
 
 	return 0;
 
@@ -2146,9 +2406,37 @@ static int hot_add_disk(mddev_t * mddev, dev_t dev)
 	return err;
 }
 
+/*
+ * set_array_info is used two different ways
+ * The original usage is when creating a new array.
+ * In this usage, raid_disks is > = and it together with
+ *  level, size, not_persistent,layout,chunksize determine the
+ *  shape of the array.
+ *  This will always create an array with a type-0.90.0 superblock.
+ * The newer usage is when assembling an array.
+ *  In this case raid_disks will be 0, and the major_version field is
+ *  use to determine which style super-blocks are to be found on the devices.
+ *  The minor and patch _version numbers are also kept incase the
+ *  super_block handler wishes to interpret them.
+ */
 static int set_array_info(mddev_t * mddev, mdu_array_info_t *info)
 {
 
+	if (info->raid_disks == 0) {
+		/* just setting version number for superblock loading */
+		if (info->major_version < 0 ||
+		    info->major_version >= sizeof(super_types)/sizeof(super_types[0]) ||
+		    super_types[info->major_version].name == NULL) {
+			/* maybe try to auto-load a module? */
+			printk(KERN_INFO "md: superblock version %d not known\n",
+			       info->major_version);
+			return -EINVAL;
+		}
+		mddev->major_version = info->major_version;
+		mddev->minor_version = info->minor_version;
+		mddev->patch_version = info->patch_version;
+		return 0;
+	}
 	mddev->major_version = MD_MAJOR_VERSION;
 	mddev->minor_version = MD_MINOR_VERSION;
 	mddev->patch_version = MD_PATCHLEVEL_VERSION;
@@ -2169,6 +2457,7 @@ static int set_array_info(mddev_t * mddev, mdu_array_info_t *info)
 	mddev->layout        = info->layout;
 	mddev->chunk_size    = info->chunk_size;
 
+	mddev->max_disks     = MD_SB_DISKS;
 
 
 	/*
@@ -2282,9 +2571,11 @@ static int md_ioctl(struct inode *inode, struct file *file,
 				err = -EBUSY;
 				goto abort_unlock;
 			}
-			if (arg) {
+			{
 				mdu_array_info_t info;
-				if (copy_from_user(&info, (void*)arg, sizeof(info))) {
+				if (!arg)
+					memset(&info, 0, sizeof(info));
+				else if (copy_from_user(&info, (void*)arg, sizeof(info))) {
 					err = -EFAULT;
 					goto abort_unlock;
 				}
@@ -2473,12 +2764,6 @@ static struct block_device_operations md_fops =
 	.ioctl		= md_ioctl,
 };
 
-
-static inline void flush_curr_signals(void)
-{
-	flush_signals(current);
-}
-
 int md_thread(void * arg)
 {
 	mdk_thread_t *thread = arg;
@@ -2489,7 +2774,7 @@ int md_thread(void * arg)
 	 * Detach thread
 	 */
 
-	daemonize(thread->name);
+	daemonize(thread->name, mdidx(thread->mddev));
 
 	current->exit_signal = SIGCHLD;
 	allow_signal(SIGKILL);
@@ -2510,7 +2795,7 @@ int md_thread(void * arg)
 
 	complete(thread->event);
 	while (thread->run) {
-		void (*run)(void *data);
+		void (*run)(mddev_t *);
 
 		wait_event_interruptible(thread->wqueue,
 					 test_bit(THREAD_WAKEUP, &thread->flags));
@@ -2521,11 +2806,11 @@ int md_thread(void * arg)
 
 		run = thread->run;
 		if (run) {
-			run(thread->data);
+			run(thread->mddev);
 			blk_run_queues();
 		}
 		if (signal_pending(current))
-			flush_curr_signals();
+			flush_signals(current);
 	}
 	complete(thread->event);
 	return 0;
@@ -2538,8 +2823,8 @@ void md_wakeup_thread(mdk_thread_t *thread)
 	wake_up(&thread->wqueue);
 }
 
-mdk_thread_t *md_register_thread(void (*run) (void *),
-						void *data, const char *name)
+mdk_thread_t *md_register_thread(void (*run) (mddev_t *), mddev_t *mddev,
+				 const char *name)
 {
 	mdk_thread_t *thread;
 	int ret;
@@ -2556,7 +2841,7 @@ mdk_thread_t *md_register_thread(void (*run) (void *),
 	init_completion(&event);
 	thread->event = &event;
 	thread->run = run;
-	thread->data = data;
+	thread->mddev = mddev;
 	thread->name = name;
 	ret = kernel_thread(md_thread, thread, 0);
 	if (ret < 0) {
@@ -2591,16 +2876,6 @@ void md_unregister_thread(mdk_thread_t *thread)
 	kfree(thread);
 }
 
-static void md_recover_arrays(void)
-{
-	if (!md_recovery_thread) {
-		MD_BUG();
-		return;
-	}
-	md_wakeup_thread(md_recovery_thread);
-}
-
-
 void md_error(mddev_t *mddev, mdk_rdev_t *rdev)
 {
 	dprintk("md_error dev:(%d:%d), rdev:(%d:%d), (caller: %p,%p,%p,%p).\n",
@@ -2618,33 +2893,34 @@ void md_error(mddev_t *mddev, mdk_rdev_t *rdev)
 	if (!mddev->pers->error_handler)
 		return;
 	mddev->pers->error_handler(mddev,rdev);
-	md_recover_arrays();
+	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+	md_wakeup_thread(mddev->thread);
 }
 
-static int status_unused(char * page)
+/* seq_file implementation /proc/mdstat */
+
+static void status_unused(struct seq_file *seq)
 {
-	int sz = 0, i = 0;
+	int i = 0;
 	mdk_rdev_t *rdev;
 	struct list_head *tmp;
 
-	sz += sprintf(page + sz, "unused devices: ");
+	seq_printf(seq, "unused devices: ");
 
 	ITERATE_RDEV_PENDING(rdev,tmp) {
 		i++;
-		sz += sprintf(page + sz, "%s ",
+		seq_printf(seq, "%s ",
 			      bdev_partition_name(rdev->bdev));
 	}
 	if (!i)
-		sz += sprintf(page + sz, "<none>");
+		seq_printf(seq, "<none>");
 
-	sz += sprintf(page + sz, "\n");
-	return sz;
+	seq_printf(seq, "\n");
 }
 
 
-static int status_resync(char * page, mddev_t * mddev)
+static void status_resync(struct seq_file *seq, mddev_t * mddev)
 {
-	int sz = 0;
 	unsigned long max_blocks, resync, res, dt, db, rt;
 
 	resync = (mddev->curr_resync - atomic_read(&mddev->recovery_active))/2;
@@ -2655,21 +2931,22 @@ static int status_resync(char * page, mddev_t * mddev)
 	 */
 	if (!max_blocks) {
 		MD_BUG();
-		return 0;
+		return;
 	}
 	res = (resync/1024)*1000/(max_blocks/1024 + 1);
 	{
 		int i, x = res/50, y = 20-x;
-		sz += sprintf(page + sz, "[");
+		seq_printf(seq, "[");
 		for (i = 0; i < x; i++)
-			sz += sprintf(page + sz, "=");
-		sz += sprintf(page + sz, ">");
+			seq_printf(seq, "=");
+		seq_printf(seq, ">");
 		for (i = 0; i < y; i++)
-			sz += sprintf(page + sz, ".");
-		sz += sprintf(page + sz, "] ");
+			seq_printf(seq, ".");
+		seq_printf(seq, "] ");
 	}
-	sz += sprintf(page + sz, " %s =%3lu.%lu%% (%lu/%lu)",
-		      (mddev->spares ? "recovery" : "resync"),
+	seq_printf(seq, " %s =%3lu.%lu%% (%lu/%lu)",
+		      (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) ?
+		       "resync" : "recovery"),
 		      res/10, res % 10, resync, max_blocks);
 
 	/*
@@ -2686,44 +2963,110 @@ static int status_resync(char * page, mddev_t * mddev)
 	db = resync - (mddev->resync_mark_cnt/2);
 	rt = (dt * ((max_blocks-resync) / (db/100+1)))/100;
 
-	sz += sprintf(page + sz, " finish=%lu.%lumin", rt / 60, (rt % 60)/6);
+	seq_printf(seq, " finish=%lu.%lumin", rt / 60, (rt % 60)/6);
 
-	sz += sprintf(page + sz, " speed=%ldK/sec", db/dt);
+	seq_printf(seq, " speed=%ldK/sec", db/dt);
+}
+
+static void *md_seq_start(struct seq_file *seq, loff_t *pos)
+{
+	struct list_head *tmp;
+	loff_t l = *pos;
+	mddev_t *mddev;
+
+	if (l > 0x10000)
+		return NULL;
+	if (!l--)
+		/* header */
+		return (void*)1;
+
+	spin_lock(&all_mddevs_lock);
+	list_for_each(tmp,&all_mddevs)
+		if (!l--) {
+			mddev = list_entry(tmp, mddev_t, all_mddevs);
+			mddev_get(mddev);
+			spin_unlock(&all_mddevs_lock);
+			return mddev;
+		}
+	spin_unlock(&all_mddevs_lock);
+	return (void*)2;/* tail */
+}
+
+static void *md_seq_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+	struct list_head *tmp;
+	mddev_t *next_mddev, *mddev = v;
+	
+	++*pos;
+	if (v == (void*)2)
+		return NULL;
+
+	spin_lock(&all_mddevs_lock);
+	if (v == (void*)1)
+		tmp = all_mddevs.next;
+	else
+		tmp = mddev->all_mddevs.next;
+	if (tmp != &all_mddevs)
+		next_mddev = mddev_get(list_entry(tmp,mddev_t,all_mddevs));
+	else {
+		next_mddev = (void*)2;
+		*pos = 0x10000;
+	}		
+	spin_unlock(&all_mddevs_lock);
+
+	if (v != (void*)1)
+		mddev_put(mddev);
+	return next_mddev;
 
-	return sz;
 }
 
-static int md_status_read_proc(char *page, char **start, off_t off,
-			int count, int *eof, void *data)
+static void md_seq_stop(struct seq_file *seq, void *v)
 {
-	int sz = 0, j;
+	mddev_t *mddev = v;
+
+	if (mddev && v != (void*)1 && v != (void*)2)
+		mddev_put(mddev);
+}
+
+static int md_seq_show(struct seq_file *seq, void *v)
+{
+	mddev_t *mddev = v;
 	sector_t size;
-	struct list_head *tmp, *tmp2;
+	struct list_head *tmp2;
 	mdk_rdev_t *rdev;
-	mddev_t *mddev;
+	int i;
 
-	sz += sprintf(page + sz, "Personalities : ");
-	for (j = 0; j < MAX_PERSONALITY; j++)
-	if (pers[j])
-		sz += sprintf(page+sz, "[%s] ", pers[j]->name);
+	if (v == (void*)1) {
+		seq_printf(seq, "Personalities : ");
+		for (i = 0; i < MAX_PERSONALITY; i++)
+			if (pers[i])
+				seq_printf(seq, "[%s] ", pers[i]->name);
 
-	sz += sprintf(page+sz, "\n");
+		seq_printf(seq, "\n");
+		return 0;
+	}
+	if (v == (void*)2) {
+		status_unused(seq);
+		return 0;
+	}
 
-	ITERATE_MDDEV(mddev,tmp) if (mddev_lock(mddev)==0) {
-		sz += sprintf(page + sz, "md%d : %sactive", mdidx(mddev),
+	if (mddev_lock(mddev)!=0) 
+		return -EINTR;
+	if (mddev->pers || mddev->raid_disks || !list_empty(&mddev->disks)) {
+		seq_printf(seq, "md%d : %sactive", mdidx(mddev),
 						mddev->pers ? "" : "in");
 		if (mddev->pers) {
 			if (mddev->ro)
-				sz += sprintf(page + sz, " (read-only)");
-			sz += sprintf(page + sz, " %s", mddev->pers->name);
+				seq_printf(seq, " (read-only)");
+			seq_printf(seq, " %s", mddev->pers->name);
 		}
 
 		size = 0;
 		ITERATE_RDEV(mddev,rdev,tmp2) {
-			sz += sprintf(page + sz, " %s[%d]",
+			seq_printf(seq, " %s[%d]",
 				bdev_partition_name(rdev->bdev), rdev->desc_nr);
 			if (rdev->faulty) {
-				sz += sprintf(page + sz, "(F)");
+				seq_printf(seq, "(F)");
 				continue;
 			}
 			size += rdev->size;
@@ -2731,34 +3074,50 @@ static int md_status_read_proc(char *page, char **start, off_t off,
 
 		if (!list_empty(&mddev->disks)) {
 			if (mddev->pers)
-				sz += sprintf(page + sz, "\n      %llu blocks",
+				seq_printf(seq, "\n      %llu blocks",
 						 (unsigned long long)md_size[mdidx(mddev)]);
 			else
-				sz += sprintf(page + sz, "\n      %llu blocks", (unsigned long long)size);
+				seq_printf(seq, "\n      %llu blocks", (unsigned long long)size);
 		}
 
-		if (!mddev->pers) {
-			sz += sprintf(page+sz, "\n");
-			mddev_unlock(mddev);
-			continue;
-		}
-
-		sz += mddev->pers->status (page+sz, mddev);
-
-		sz += sprintf(page+sz, "\n      ");
+		if (mddev->pers) {
+			mddev->pers->status (seq, mddev);
+	 		seq_printf(seq, "\n      ");
 			if (mddev->curr_resync > 2)
-			sz += status_resync (page+sz, mddev);
+				status_resync (seq, mddev);
 			else if (mddev->curr_resync == 1 || mddev->curr_resync == 2)
-				sz += sprintf(page + sz, "	resync=DELAYED");
+				seq_printf(seq, "	resync=DELAYED");
+		}
 
-		sz += sprintf(page + sz, "\n");
-		mddev_unlock(mddev);
+		seq_printf(seq, "\n");
 	}
-	sz += status_unused(page + sz);
+	mddev_unlock(mddev);
+	
+	return 0;
+}
+
+static struct seq_operations md_seq_ops = {
+	.start  = md_seq_start,
+	.next   = md_seq_next,
+	.stop   = md_seq_stop,
+	.show   = md_seq_show,
+};
+
+static int md_seq_open(struct inode *inode, struct file *file)
+{
+	int error;
 
-	return sz;
+	error = seq_open(file, &md_seq_ops);
+	return error;
 }
 
+static struct file_operations md_seq_fops = {
+	.open           = md_seq_open,
+	.read           = seq_read,
+	.llseek         = seq_lseek,
+	.release	= seq_release,
+};
+
 int register_md_personality(int pnum, mdk_personality_t *p)
 {
 	if (pnum >= MAX_PERSONALITY) {
@@ -2820,9 +3179,8 @@ void md_done_sync(mddev_t *mddev, int blocks, int ok)
 	atomic_sub(blocks, &mddev->recovery_active);
 	wake_up(&mddev->recovery_wait);
 	if (!ok) {
-		mddev->recovery_error = -EIO;
-		mddev->recovery_running = -1;
-		md_recover_arrays();
+		set_bit(MD_RECOVERY_ERR, &mddev->recovery);
+		md_wakeup_thread(mddev->thread);
 		// stop recovery, signal do_sync ....
 	}
 }
@@ -2830,40 +3188,49 @@ void md_done_sync(mddev_t *mddev, int blocks, int ok)
 
 void md_write_start(mddev_t *mddev)
 {
-	if (mddev->safemode && !atomic_read(&mddev->writes_pending)) {
+	if (!atomic_read(&mddev->writes_pending)) {
 		mddev_lock_uninterruptible(mddev);
-		atomic_inc(&mddev->writes_pending);
 		if (mddev->in_sync) {
 			mddev->in_sync = 0;
+ 			del_timer(&mddev->safemode_timer);
 			md_update_sb(mddev);
 		}
+		atomic_inc(&mddev->writes_pending);
 		mddev_unlock(mddev);
 	} else
 		atomic_inc(&mddev->writes_pending);
 }
 
-void md_write_end(mddev_t *mddev, mdk_thread_t *thread)
+void md_write_end(mddev_t *mddev)
 {
-	if (atomic_dec_and_test(&mddev->writes_pending) && mddev->safemode)
-		md_wakeup_thread(thread);
+	if (atomic_dec_and_test(&mddev->writes_pending)) {
+		if (mddev->safemode == 2)
+			md_wakeup_thread(mddev->thread);
+		else
+			mod_timer(&mddev->safemode_timer, jiffies + mddev->safemode_delay);
+	}
 }
+
 static inline void md_enter_safemode(mddev_t *mddev)
 {
-
 	mddev_lock_uninterruptible(mddev);
-	if (mddev->safemode && !atomic_read(&mddev->writes_pending) && !mddev->in_sync && !mddev->recovery_running) {
+	if (mddev->safemode && !atomic_read(&mddev->writes_pending) &&
+	    !mddev->in_sync && mddev->recovery_cp == MaxSector) {
 		mddev->in_sync = 1;
 		md_update_sb(mddev);
 	}
 	mddev_unlock(mddev);
+
+	if (mddev->safemode == 1)
+		mddev->safemode = 0;
 }
 
 void md_handle_safemode(mddev_t *mddev)
 {
 	if (signal_pending(current)) {
-		printk(KERN_INFO "md: md%d in safe mode\n",mdidx(mddev));
-		mddev->safemode= 1;
-		flush_curr_signals();
+		printk(KERN_INFO "md: md%d in immediate safe mode\n",mdidx(mddev));
+		mddev->safemode = 2;
+		flush_signals(current);
 	}
 	if (mddev->safemode)
 		md_enter_safemode(mddev);
@@ -2874,9 +3241,8 @@ DECLARE_WAIT_QUEUE_HEAD(resync_wait);
 
 #define SYNC_MARKS	10
 #define	SYNC_MARK_STEP	(3*HZ)
-static void md_do_sync(void *data)
+static void md_do_sync(mddev_t *mddev)
 {
-	mddev_t *mddev = data;
 	mddev_t *mddev2;
 	unsigned int max_sectors, currspeed = 0,
 		j, window, err;
@@ -2887,7 +3253,7 @@ static void md_do_sync(void *data)
 	unsigned long last_check;
 
 	/* just incase thread restarts... */
-	if (mddev->recovery_running <= 0)
+	if (test_bit(MD_RECOVERY_DONE, &mddev->recovery))
 		return;
 
 	/* we overload curr_resync somewhat here.
@@ -2914,15 +3280,17 @@ static void md_do_sync(void *data)
 				}
 				if (wait_event_interruptible(resync_wait,
 							     mddev2->curr_resync < mddev->curr_resync)) {
-					flush_curr_signals();
+					flush_signals(current);
 					err = -EINTR;
 					mddev_put(mddev2);
 					goto skip;
 				}
 			}
-			if (mddev->curr_resync == 1)
+			if (mddev->curr_resync == 1) {
+				mddev_put(mddev2);
 				break;
 			}
+		}
 	} while (mddev->curr_resync < 2);
 
 	max_sectors = mddev->size << 1;
@@ -2934,9 +3302,13 @@ static void md_do_sync(void *data)
 	       sysctl_speed_limit_max);
 
 	is_mddev_idle(mddev); /* this also initializes IO event counters */
+	if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery))
+		j = mddev->recovery_cp;
+	else
+		j = 0;
 	for (m = 0; m < SYNC_MARKS; m++) {
 		mark[m] = jiffies;
-		mark_cnt[m] = mddev->recovery_cp;
+		mark_cnt[m] = j;
 	}
 	last_mark = 0;
 	mddev->resync_mark = mark[last_mark];
@@ -2953,12 +3325,10 @@ static void md_do_sync(void *data)
 	init_waitqueue_head(&mddev->recovery_wait);
 	last_check = 0;
 
-	mddev->recovery_error = 0;
-
-	if (mddev->recovery_cp)
+	if (j)
 		printk(KERN_INFO "md: resuming recovery of md%d from checkpoint.\n", mdidx(mddev));
 
-	for (j = mddev->recovery_cp; j < max_sectors;) {
+	while (j < max_sectors) {
 		int sectors;
 
 		sectors = mddev->pers->sync_request(mddev, j, currspeed < sysctl_speed_limit_min);
@@ -2975,6 +3345,10 @@ static void md_do_sync(void *data)
 
 		last_check = j;
 
+		if (test_bit(MD_RECOVERY_INTR, &mddev->recovery) ||
+		    test_bit(MD_RECOVERY_ERR, &mddev->recovery))
+			break;
+
 		blk_run_queues();
 
 	repeat:
@@ -2995,7 +3369,7 @@ static void md_do_sync(void *data)
 			 * got a signal, exit.
 			 */
 			printk(KERN_INFO "md: md_do_sync() got signal ... exiting\n");
-			flush_curr_signals();
+			flush_signals(current);
 			err = -EINTR;
 			goto out;
 		}
@@ -3029,39 +3403,42 @@ static void md_do_sync(void *data)
  out:
 	wait_event(mddev->recovery_wait, !atomic_read(&mddev->recovery_active));
 
-	if (mddev->recovery_running < 0 && 
-		!mddev->recovery_error && mddev->curr_resync > 2)
-	{
-		/* interrupted but no write errors */
-		printk(KERN_INFO "md: checkpointing recovery of md%d.\n", mdidx(mddev));
-		mddev->recovery_cp = mddev->curr_resync;
-	}
-
 	/* tell personality that we are finished */
 	mddev->pers->sync_request(mddev, max_sectors, 1);
- skip:
-	mddev->curr_resync = 0;
+
 	if (err)
-		mddev->recovery_running = -1;
-	if (mddev->recovery_running > 0)
-		mddev->recovery_running = 0;
-	if (mddev->recovery_running == 0)
+		set_bit(MD_RECOVERY_ERR, &mddev->recovery);
+	if (!test_bit(MD_RECOVERY_ERR, &mddev->recovery) &&
+	    mddev->curr_resync > 2 &&
+	    mddev->curr_resync > mddev->recovery_cp) {
+		if (test_bit(MD_RECOVERY_INTR, &mddev->recovery)) {
+			printk(KERN_INFO "md: checkpointing recovery of md%d.\n", mdidx(mddev));
+			mddev->recovery_cp = mddev->curr_resync;
+		} else
 			mddev->recovery_cp = MaxSector;
+	}
+
 	if (mddev->safemode)
 		md_enter_safemode(mddev);
-	md_recover_arrays();
+ skip:
+	mddev->curr_resync = 0;
+	set_bit(MD_RECOVERY_DONE, &mddev->recovery);
+	md_wakeup_thread(mddev->thread);
 }
 
 
 /*
- * This is the kernel thread that watches all md arrays for re-sync and other
- * action that might be needed.
+ * This routine is regularly called by all per-raid-array threads to
+ * deal with generic issues like resync and super-block update.
+ * Raid personalities that don't have a thread (linear/raid0) do not
+ * need this as they never do any recovery or update the superblock.
+ *
  * It does not do any resync itself, but rather "forks" off other threads
  * to do that as needed.
- * When it is determined that resync is needed, we set "->recovery_running" and
- * create a thread at ->sync_thread.
- * When the thread finishes it clears recovery_running (or sets an error)
- * and wakeup up this thread which will reap the thread and finish up.
+ * When it is determined that resync is needed, we set MD_RECOVERY_RUNNING in
+ * "->recovery" and create a thread at ->sync_thread.
+ * When the thread finishes it sets MD_RECOVERY_DONE (and might set MD_RECOVERY_ERR)
+ * and wakeups up this thread which will reap the thread and finish up.
  * This thread also removes any faulty devices (with nr_pending == 0).
  *
  * The overall approach is:
@@ -3072,41 +3449,47 @@ static void md_do_sync(void *data)
  *  5/ If array is degraded, try to add spares devices
  *  6/ If array has spares or is not in-sync, start a resync thread.
  */
-void md_do_recovery(void *data)
+void md_check_recovery(mddev_t *mddev)
 {
-	mddev_t *mddev;
 	mdk_rdev_t *rdev;
-	struct list_head *tmp, *rtmp;
+	struct list_head *rtmp;
 
 
 	dprintk(KERN_INFO "md: recovery thread got woken up ...\n");
 
-	ITERATE_MDDEV(mddev,tmp) if (mddev_lock(mddev)==0) {
-		if (!mddev->raid_disks || !mddev->pers || mddev->ro)
-			goto unlock;
+	if (mddev->ro)
+		return;
+	if ( ! (
+		mddev->sb_dirty ||
+		test_bit(MD_RECOVERY_NEEDED, &mddev->recovery) ||
+		test_bit(MD_RECOVERY_DONE, &mddev->recovery)
+		))
+		return;
+	if (mddev_trylock(mddev)==0) {
+		int spares =0;
 		if (mddev->sb_dirty)
 			md_update_sb(mddev);
-		if (mddev->recovery_running > 0)
+		if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) &&
+		    !test_bit(MD_RECOVERY_DONE, &mddev->recovery))
 			/* resync/recovery still happening */
 			goto unlock;
 		if (mddev->sync_thread) {
 			/* resync has finished, collect result */
 			md_unregister_thread(mddev->sync_thread);
 			mddev->sync_thread = NULL;
-			if (mddev->recovery_running == 0) {
+			if (!test_bit(MD_RECOVERY_ERR, &mddev->recovery)) {
 				/* success...*/
 				/* activate any spares */
 				mddev->pers->spare_active(mddev);
-				mddev->spares = 0;
 			}
 			md_update_sb(mddev);
-			mddev->recovery_running = 0;
+			mddev->recovery = 0;
 			wake_up(&resync_wait);
 			goto unlock;
 		}
-		if (mddev->recovery_running) {
+		if (mddev->recovery) {
 			/* that's odd.. */
-			mddev->recovery_running = 0;
+			mddev->recovery = 0;
 			wake_up(&resync_wait);
 		}
 
@@ -3114,7 +3497,6 @@ void md_do_recovery(void *data)
 		 * remove any failed drives, then
 		 * add spares if possible
 		 */
-		mddev->spares = 0;
 		ITERATE_RDEV(mddev,rdev,rtmp) {
 			if (rdev->raid_disk >= 0 &&
 			    rdev->faulty &&
@@ -3123,43 +3505,41 @@ void md_do_recovery(void *data)
 				rdev->raid_disk = -1;
 			}
 			if (!rdev->faulty && rdev->raid_disk >= 0 && !rdev->in_sync)
-				mddev->spares++;
+				spares++;
 		}
 		if (mddev->degraded) {
 			ITERATE_RDEV(mddev,rdev,rtmp)
 				if (rdev->raid_disk < 0
 				    && !rdev->faulty) {
-					if (mddev->pers->hot_add_disk(mddev,rdev)) {
-						mddev->spares++;
-						mddev->recovery_cp = 0;
-					}
+					if (mddev->pers->hot_add_disk(mddev,rdev))
+						spares++;
 					else
 						break;
 				}
 		}
 
-		if (!mddev->spares && (mddev->recovery_cp == MaxSector )) {
+		if (!spares && (mddev->recovery_cp == MaxSector )) {
 			/* nothing we can do ... */
 			goto unlock;
 		}
 		if (mddev->pers->sync_request) {
+			set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
+			if (!spares)
+				set_bit(MD_RECOVERY_SYNC, &mddev->recovery);
 			mddev->sync_thread = md_register_thread(md_do_sync,
 								mddev,
-								"md_resync");
+								"md%d_resync");
 			if (!mddev->sync_thread) {
 				printk(KERN_ERR "md%d: could not start resync thread...\n", mdidx(mddev));
 				/* leave the spares where they are, it shouldn't hurt */
-				mddev->recovery_running = 0;
+				mddev->recovery = 0;
 			} else {
-				mddev->recovery_running = 1;
 				md_wakeup_thread(mddev->sync_thread);
 			}
 		}
 	unlock:
 		mddev_unlock(mddev);
 	}
-	dprintk(KERN_INFO "md: recovery thread finished ...\n");
-
 }
 
 int md_notify_reboot(struct notifier_block *this,
@@ -3194,6 +3574,7 @@ struct notifier_block md_notifier = {
 
 static void md_geninit(void)
 {
+	struct proc_dir_entry *p;
 	int i;
 
 	for(i = 0; i < MAX_MD_DEVS; i++) {
@@ -3203,13 +3584,14 @@ static void md_geninit(void)
 	dprintk("md: sizeof(mdp_super_t) = %d\n", (int)sizeof(mdp_super_t));
 
 #ifdef CONFIG_PROC_FS
-	create_proc_read_entry("mdstat", 0, NULL, md_status_read_proc, NULL);
+	p = create_proc_entry("mdstat", S_IRUGO, NULL);
+	if (p)
+		p->proc_fops = &md_seq_fops;
 #endif
 }
 
 int __init md_init(void)
 {
-	static char * name = "mdrecoveryd";
 	int minor;
 
 	printk(KERN_INFO "md: md driver %d.%d.%d MAX_MD_DEVS=%d, MD_SB_DISKS=%d\n",
@@ -3229,11 +3611,6 @@ int __init md_init(void)
 			       S_IFBLK | S_IRUSR | S_IWUSR, &md_fops, NULL);
 	}
 
-	md_recovery_thread = md_register_thread(md_do_recovery, NULL, name);
-	if (!md_recovery_thread)
-		printk(KERN_ALERT
-		       "md: bug: couldn't allocate md_recovery_thread\n");
-
 	register_reboot_notifier(&md_notifier);
 	raid_table_header = register_sysctl_table(raid_root_table, 1);
 
@@ -3268,7 +3645,7 @@ static void autostart_arrays(void)
 	for (i = 0; i < dev_cnt; i++) {
 		dev_t dev = detected_devices[i];
 
-		rdev = md_import_device(dev,1);
+		rdev = md_import_device(dev,0, 0);
 		if (IS_ERR(rdev)) {
 			printk(KERN_ALERT "md: could not import %s!\n",
 				partition_name(dev));
@@ -3291,7 +3668,6 @@ static __exit void md_exit(void)
 {
 	int i;
 	blk_unregister_region(MKDEV(MAJOR_NR,0), MAX_MD_DEVS);
-	md_unregister_thread(md_recovery_thread);
 	for (i=0; i < MAX_MD_DEVS; i++)
 		devfs_remove("md/%d", i);
 	devfs_remove("md");
@@ -3331,4 +3707,5 @@ EXPORT_SYMBOL(md_unregister_thread);
 EXPORT_SYMBOL(md_wakeup_thread);
 EXPORT_SYMBOL(md_print_devices);
 EXPORT_SYMBOL(md_interrupt_thread);
+EXPORT_SYMBOL(md_check_recovery);
 MODULE_LICENSE("GPL");

drivers/md/multipath.c

@@ -86,7 +86,6 @@ static void multipath_reschedule_retry (struct multipath_bh *mp_bh)
 {
 	unsigned long flags;
 	mddev_t *mddev = mp_bh->mddev;
-	multipath_conf_t *conf = mddev_to_conf(mddev);
 
 	spin_lock_irqsave(&retry_list_lock, flags);
 	if (multipath_retry_list == NULL)
@@ -95,7 +94,7 @@ static void multipath_reschedule_retry (struct multipath_bh *mp_bh)
 	multipath_retry_tail = &mp_bh->next_mp;
 	mp_bh->next_mp = NULL;
 	spin_unlock_irqrestore(&retry_list_lock, flags);
-	md_wakeup_thread(conf->thread);
+	md_wakeup_thread(mddev->thread);
 }
 
 
@@ -185,19 +184,18 @@ static int multipath_make_request (request_queue_t *q, struct bio * bio)
 	return 0;
 }
 
-static int multipath_status (char *page, mddev_t *mddev)
+static void multipath_status (struct seq_file *seq, mddev_t *mddev)
 {
 	multipath_conf_t *conf = mddev_to_conf(mddev);
-	int sz = 0, i;
+	int i;
 	
-	sz += sprintf (page+sz, " [%d/%d] [", conf->raid_disks,
+	seq_printf (seq, " [%d/%d] [", conf->raid_disks,
 						 conf->working_disks);
 	for (i = 0; i < conf->raid_disks; i++)
-		sz += sprintf (page+sz, "%s",
+		seq_printf (seq, "%s",
 			       conf->multipaths[i].rdev && 
 			       conf->multipaths[i].rdev->in_sync ? "U" : "_");
-	sz += sprintf (page+sz, "]");
-	return sz;
+	seq_printf (seq, "]");
 }
 
 #define LAST_DISK KERN_ALERT \
@@ -334,14 +332,14 @@ static int multipath_remove_disk(mddev_t *mddev, int number)
  *	3.	Performs writes following reads for array syncronising.
  */
 
-static void multipathd (void *data)
+static void multipathd (mddev_t *mddev)
 {
 	struct multipath_bh *mp_bh;
 	struct bio *bio;
 	unsigned long flags;
-	mddev_t *mddev;
 	mdk_rdev_t *rdev;
 
+	md_check_recovery(mddev);
 	for (;;) {
 		spin_lock_irqsave(&retry_list_lock, flags);
 		mp_bh = multipath_retry_list;
@@ -471,10 +469,10 @@ static int multipath_run (mddev_t *mddev)
 	}
 
 	{
-		const char * name = "multipathd";
+		const char * name = "md%d_multipath";
 
-		conf->thread = md_register_thread(multipathd, conf, name);
-		if (!conf->thread) {
+		mddev->thread = md_register_thread(multipathd, mddev, name);
+		if (!mddev->thread) {
 			printk(THREAD_ERROR, mdidx(mddev));
 			goto out_free_conf;
 		}
@@ -513,7 +511,7 @@ static int multipath_stop (mddev_t *mddev)
 {
 	multipath_conf_t *conf = mddev_to_conf(mddev);
 
-	md_unregister_thread(conf->thread);
+	md_unregister_thread(mddev->thread);
 	mempool_destroy(conf->pool);
 	kfree(conf);
 	mddev->private = NULL;

drivers/md/raid0.c

@@ -349,7 +349,7 @@ static int raid0_make_request (request_queue_t *q, struct bio *bio)
 	 * is the only IO operation happening on this bh.
 	 */
 	bio->bi_bdev = tmp_dev->bdev;
-	bio->bi_sector = rsect;
+	bio->bi_sector = rsect + tmp_dev->data_offset;
 
 	/*
 	 * Let the main block layer submit the IO and resolve recursion:
@@ -372,41 +372,40 @@ static int raid0_make_request (request_queue_t *q, struct bio *bio)
 	return 0;
 }
 			   
-static int raid0_status (char *page, mddev_t *mddev)
+static void raid0_status (struct seq_file *seq, mddev_t *mddev)
 {
-	int sz = 0;
 #undef MD_DEBUG
 #ifdef MD_DEBUG
 	int j, k;
 	raid0_conf_t *conf = mddev_to_conf(mddev);
   
-	sz += sprintf(page + sz, "      ");
+	seq_printf(seq, "      ");
 	for (j = 0; j < conf->nr_zones; j++) {
-		sz += sprintf(page + sz, "[z%d",
+		seq_printf(seq, "[z%d",
 				conf->hash_table[j].zone0 - conf->strip_zone);
 		if (conf->hash_table[j].zone1)
-			sz += sprintf(page+sz, "/z%d] ",
+			seq_printf(seq, "/z%d] ",
 				conf->hash_table[j].zone1 - conf->strip_zone);
 		else
-			sz += sprintf(page+sz, "] ");
+			seq_printf(seq, "] ");
 	}
   
-	sz += sprintf(page + sz, "\n");
+	seq_printf(seq, "\n");
   
 	for (j = 0; j < conf->nr_strip_zones; j++) {
-		sz += sprintf(page + sz, "      z%d=[", j);
+		seq_printf(seq, "      z%d=[", j);
 		for (k = 0; k < conf->strip_zone[j].nb_dev; k++)
-			sz += sprintf (page+sz, "%s/", bdev_partition_name(
+			seq_printf (seq, "%s/", bdev_partition_name(
 				conf->strip_zone[j].dev[k]->bdev));
-		sz--;
-		sz += sprintf (page+sz, "] zo=%d do=%d s=%d\n",
+
+		seq_printf (seq, "] zo=%d do=%d s=%d\n",
 				conf->strip_zone[j].zone_offset,
 				conf->strip_zone[j].dev_offset,
 				conf->strip_zone[j].size);
 	}
 #endif
-	sz += sprintf(page + sz, " %dk chunks", mddev->chunk_size/1024);
-	return sz;
+	seq_printf(seq, " %dk chunks", mddev->chunk_size/1024);
+	return;
 }
 
 static mdk_personality_t raid0_personality=

drivers/md/raid1.c

@@ -225,13 +225,12 @@ static void reschedule_retry(r1bio_t *r1_bio)
 {
 	unsigned long flags;
 	mddev_t *mddev = r1_bio->mddev;
-	conf_t *conf = mddev_to_conf(mddev);
 
 	spin_lock_irqsave(&retry_list_lock, flags);
 	list_add(&r1_bio->retry_list, &retry_list_head);
 	spin_unlock_irqrestore(&retry_list_lock, flags);
 
-	md_wakeup_thread(conf->thread);
+	md_wakeup_thread(mddev->thread);
 }
 
 /*
@@ -320,7 +319,7 @@ static int end_request(struct bio *bio, unsigned int bytes_done, int error)
 		 * already.
 		 */
 		if (atomic_dec_and_test(&r1_bio->remaining)) {
-			md_write_end(r1_bio->mddev,conf->thread);
+			md_write_end(r1_bio->mddev);
 			raid_end_bio_io(r1_bio, uptodate);
 		}	
 	}
@@ -494,7 +493,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 			BUG();
 		r1_bio->read_bio = read_bio;
 
-		read_bio->bi_sector = r1_bio->sector;
+		read_bio->bi_sector = r1_bio->sector + mirror->rdev->data_offset;
 		read_bio->bi_bdev = mirror->rdev->bdev;
 		read_bio->bi_end_io = end_request;
 		read_bio->bi_rw = r1_bio->cmd;
@@ -529,7 +528,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 		mbio = bio_clone(bio, GFP_NOIO);
 		r1_bio->write_bios[i] = mbio;
 
-		mbio->bi_sector	= r1_bio->sector;
+		mbio->bi_sector	= r1_bio->sector + conf->mirrors[i].rdev->data_offset;
 		mbio->bi_bdev = conf->mirrors[i].rdev->bdev;
 		mbio->bi_end_io	= end_request;
 		mbio->bi_rw = r1_bio->cmd;
@@ -542,7 +541,7 @@ static int make_request(request_queue_t *q, struct bio * bio)
 		 * If all mirrors are non-operational
 		 * then return an IO error:
 		 */
-		md_write_end(mddev,conf->thread);
+		md_write_end(mddev);
 		raid_end_bio_io(r1_bio, 0);
 		return 0;
 	}
@@ -571,19 +570,18 @@ static int make_request(request_queue_t *q, struct bio * bio)
 	return 0;
 }
 
-static int status(char *page, mddev_t *mddev)
+static void status(struct seq_file *seq, mddev_t *mddev)
 {
 	conf_t *conf = mddev_to_conf(mddev);
-	int sz = 0, i;
+	int i;
 
-	sz += sprintf(page+sz, " [%d/%d] [", conf->raid_disks,
+	seq_printf(seq, " [%d/%d] [", conf->raid_disks,
 						conf->working_disks);
 	for (i = 0; i < conf->raid_disks; i++)
-		sz += sprintf(page+sz, "%s",
+		seq_printf(seq, "%s",
 			      conf->mirrors[i].rdev &&
 			      conf->mirrors[i].rdev->in_sync ? "U" : "_");
-	sz += sprintf (page+sz, "]");
-	return sz;
+	seq_printf(seq, "]");
 }
 
 #define LAST_DISK KERN_ALERT \
@@ -624,10 +622,9 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev)
 		mddev->degraded++;
 		conf->working_disks--;
 		/*
-		 * if recovery was running, stop it now.
+		 * if recovery is running, make sure it aborts.
 		 */
-		if (mddev->recovery_running) 
-			mddev->recovery_running = -EIO;
+		set_bit(MD_RECOVERY_ERR, &mddev->recovery);
 	}
 	rdev->in_sync = 0;
 	rdev->faulty = 1;
@@ -859,7 +856,7 @@ static void sync_request_write(mddev_t *mddev, r1bio_t *r1_bio)
 		mbio = bio_clone(bio, GFP_NOIO);
 		r1_bio->write_bios[i] = mbio;
 		mbio->bi_bdev = conf->mirrors[i].rdev->bdev;
-		mbio->bi_sector = r1_bio->sector;
+		mbio->bi_sector = r1_bio->sector | conf->mirrors[i].rdev->data_offset;
 		mbio->bi_end_io	= end_sync_write;
 		mbio->bi_rw = WRITE;
 		mbio->bi_private = r1_bio;
@@ -900,17 +897,17 @@ static void sync_request_write(mddev_t *mddev, r1bio_t *r1_bio)
  *	3.	Performs writes following reads for array syncronising.
  */
 
-static void raid1d(void *data)
+static void raid1d(mddev_t *mddev)
 {
 	struct list_head *head = &retry_list_head;
 	r1bio_t *r1_bio;
 	struct bio *bio;
 	unsigned long flags;
-	mddev_t *mddev;
-	conf_t *conf = data;
+	conf_t *conf = mddev_to_conf(mddev);
 	mdk_rdev_t *rdev;
 
-	md_handle_safemode(conf->mddev);
+	md_check_recovery(mddev);
+	md_handle_safemode(mddev);
 	
 	for (;;) {
 		spin_lock_irqsave(&retry_list_lock, flags);
@@ -937,7 +934,7 @@ static void raid1d(void *data)
 			printk(REDIRECT_SECTOR,
 				bdev_partition_name(rdev->bdev), (unsigned long long)r1_bio->sector);
 			bio->bi_bdev = rdev->bdev;
-			bio->bi_sector = r1_bio->sector;
+			bio->bi_sector = r1_bio->sector + rdev->data_offset;
 			bio->bi_rw = r1_bio->cmd;
 
 			generic_make_request(bio);
@@ -1048,7 +1045,7 @@ static int sync_request(mddev_t *mddev, sector_t sector_nr, int go_faster)
 
 	read_bio = bio_clone(r1_bio->master_bio, GFP_NOIO);
 
-	read_bio->bi_sector = sector_nr;
+	read_bio->bi_sector = sector_nr + mirror->rdev->data_offset;
 	read_bio->bi_bdev = mirror->rdev->bdev;
 	read_bio->bi_end_io = end_sync_read;
 	read_bio->bi_rw = READ;
@@ -1190,10 +1187,8 @@ static int run(mddev_t *mddev)
 
 
 	{
-		snprintf(conf->thread_name,MD_THREAD_NAME_MAX,"raid1d_md%d",mdidx(mddev));
-
-		conf->thread = md_register_thread(raid1d, conf, conf->thread_name);
-		if (!conf->thread) {
+		mddev->thread = md_register_thread(raid1d, mddev, "md%d_raid1");
+		if (!mddev->thread) {
 			printk(THREAD_ERROR, mdidx(mddev));
 			goto out_free_conf;
 		}
@@ -1219,7 +1214,8 @@ static int stop(mddev_t *mddev)
 {
 	conf_t *conf = mddev_to_conf(mddev);
 
-	md_unregister_thread(conf->thread);
+	md_unregister_thread(mddev->thread);
+	mddev->thread = NULL;
 	if (conf->r1bio_pool)
 		mempool_destroy(conf->r1bio_pool);
 	kfree(conf);

drivers/md/raid5.c

@@ -71,12 +71,12 @@ static inline void __release_stripe(raid5_conf_t *conf, struct stripe_head *sh)
 				list_add_tail(&sh->lru, &conf->delayed_list);
 			else
 				list_add_tail(&sh->lru, &conf->handle_list);
-			md_wakeup_thread(conf->thread);
+			md_wakeup_thread(conf->mddev->thread);
 		} else {
 			if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
 				atomic_dec(&conf->preread_active_stripes);
 				if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD)
-					md_wakeup_thread(conf->thread);
+					md_wakeup_thread(conf->mddev->thread);
 			}
 			list_add_tail(&sh->lru, &conf->inactive_list);
 			atomic_dec(&conf->active_stripes);
@@ -463,10 +463,9 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev)
 			conf->failed_disks++;
 			rdev->in_sync = 0;
 			/*
-			 * if recovery was running, stop it now.
+			 * if recovery was running, make sure it aborts.
 			 */
-			if (mddev->recovery_running) 
-				mddev->recovery_running = -EIO;
+			set_bit(MD_RECOVERY_ERR, &mddev->recovery);
 		}
 		rdev->faulty = 1;
 		printk (KERN_ALERT
@@ -913,7 +912,7 @@ static void handle_stripe(struct stripe_head *sh)
 				struct bio *nextbi = bi->bi_next;
 				clear_bit(BIO_UPTODATE, &bi->bi_flags);
 				if (--bi->bi_phys_segments == 0) {
-					md_write_end(conf->mddev, conf->thread);
+					md_write_end(conf->mddev);
 					bi->bi_next = return_bi;
 					return_bi = bi;
 				}
@@ -970,7 +969,7 @@ static void handle_stripe(struct stripe_head *sh)
 			    while (wbi && wbi->bi_sector < dev->sector + STRIPE_SECTORS) {
 				    wbi2 = wbi->bi_next;
 				    if (--wbi->bi_phys_segments == 0) {
-					    md_write_end(conf->mddev, conf->thread);
+					    md_write_end(conf->mddev);
 					    wbi->bi_next = return_bi;
 					    return_bi = wbi;
 				    }
@@ -1113,7 +1112,7 @@ static void handle_stripe(struct stripe_head *sh)
 			if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
 				atomic_dec(&conf->preread_active_stripes);
 				if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD)
-					md_wakeup_thread(conf->thread);
+					md_wakeup_thread(conf->mddev->thread);
 			}
 		}
 	}
@@ -1207,7 +1206,7 @@ static void handle_stripe(struct stripe_head *sh)
 			bi->bi_bdev = rdev->bdev;
 			PRINTK("for %llu schedule op %ld on disc %d\n", (unsigned long long)sh->sector, bi->bi_rw, i);
 			atomic_inc(&sh->count);
-			bi->bi_sector = sh->sector;
+			bi->bi_sector = sh->sector + rdev->data_offset;
 			bi->bi_flags = 1 << BIO_UPTODATE;
 			bi->bi_vcnt = 1;	
 			bi->bi_idx = 0;
@@ -1251,7 +1250,7 @@ static void raid5_unplug_device(void *data)
 
 	if (blk_remove_plug(q))
 		raid5_activate_delayed(conf);
-	md_wakeup_thread(conf->thread);
+	md_wakeup_thread(mddev->thread);
 
 	spin_unlock_irqrestore(&conf->device_lock, flags);
 }
@@ -1304,7 +1303,7 @@ static int make_request (request_queue_t *q, struct bio * bi)
 		int bytes = bi->bi_size;
 
 		if ( bio_data_dir(bi) == WRITE )
-			md_write_end(mddev,conf->thread);
+			md_write_end(mddev);
 		bi->bi_size = 0;
 		bi->bi_end_io(bi, bytes, 0);
 	}
@@ -1356,16 +1355,17 @@ static int sync_request (mddev_t *mddev, sector_t sector_nr, int go_faster)
  * During the scan, completed stripes are saved for us by the interrupt
  * handler, so that they will not have to wait for our next wakeup.
  */
-static void raid5d (void *data)
+static void raid5d (mddev_t *mddev)
 {
 	struct stripe_head *sh;
-	raid5_conf_t *conf = data;
-	mddev_t *mddev = conf->mddev;
+	raid5_conf_t *conf = mddev_to_conf(mddev);
 	int handled;
 
 	PRINTK("+++ raid5d active\n");
 
+	md_check_recovery(mddev);
 	md_handle_safemode(mddev);
+
 	handled = 0;
 	spin_lock_irq(&conf->device_lock);
 	while (1) {
@@ -1486,10 +1486,8 @@ static int run (mddev_t *mddev)
 	}
 
 	{
-		snprintf(conf->thread_name,MD_THREAD_NAME_MAX,"raid5d_md%d",mdidx(mddev));
-
-		conf->thread = md_register_thread(raid5d, conf, conf->thread_name);
-		if (!conf->thread) {
+		mddev->thread = md_register_thread(raid5d, mddev, "md%d_raid5");
+		if (!mddev->thread) {
 			printk(KERN_ERR "raid5: couldn't allocate thread for md%d\n", mdidx(mddev));
 			goto abort;
 		}
@@ -1500,7 +1498,7 @@ static int run (mddev_t *mddev)
 	if (grow_stripes(conf, conf->max_nr_stripes)) {
 		printk(KERN_ERR "raid5: couldn't allocate %dkB for buffers\n", memory);
 		shrink_stripes(conf);
-		md_unregister_thread(conf->thread);
+		md_unregister_thread(mddev->thread);
 		goto abort;
 	} else
 		printk(KERN_INFO "raid5: allocated %dkB for md%d\n", memory, mdidx(mddev));
@@ -1536,7 +1534,8 @@ static int stop (mddev_t *mddev)
 {
 	raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
 
-	md_unregister_thread(conf->thread);
+	md_unregister_thread(mddev->thread);
+	mddev->thread = NULL;
 	shrink_stripes(conf);
 	free_pages((unsigned long) conf->stripe_hashtbl, HASH_PAGES_ORDER);
 	kfree(conf);
@@ -1574,29 +1573,26 @@ static void printall (raid5_conf_t *conf)
 		}
 	}
 	spin_unlock_irq(&conf->device_lock);
-
-	PRINTK("--- raid5d inactive\n");
 }
 #endif
 
-static int status (char *page, mddev_t *mddev)
+static void status (struct seq_file *seq, mddev_t *mddev)
 {
 	raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
-	int sz = 0, i;
+	int i;
 
-	sz += sprintf (page+sz, " level %d, %dk chunk, algorithm %d", mddev->level, mddev->chunk_size >> 10, mddev->layout);
-	sz += sprintf (page+sz, " [%d/%d] [", conf->raid_disks, conf->working_disks);
+	seq_printf (seq, " level %d, %dk chunk, algorithm %d", mddev->level, mddev->chunk_size >> 10, mddev->layout);
+	seq_printf (seq, " [%d/%d] [", conf->raid_disks, conf->working_disks);
 	for (i = 0; i < conf->raid_disks; i++)
-		sz += sprintf (page+sz, "%s",
+		seq_printf (seq, "%s",
 			       conf->disks[i].rdev &&
 			       conf->disks[i].rdev->in_sync ? "U" : "_");
-	sz += sprintf (page+sz, "]");
+	seq_printf (seq, "]");
 #if RAID5_DEBUG
 #define D(x) \
-	sz += sprintf (page+sz, "<"#x":%d>", atomic_read(&conf->x))
+	seq_printf (seq, "<"#x":%d>", atomic_read(&conf->x))
 	printall(conf);
 #endif
-	return sz;
 }
 
 static void print_raid5_conf (raid5_conf_t *conf)

fs/lockd/svclock.c

@@ -305,8 +305,6 @@ nlmsvc_lock(struct svc_rqst *rqstp, struct nlm_file *file,
 				(long long)lock->fl.fl_end,
 				wait);
 
-	/* Lock file against concurrent access */
-	down(&file->f_sema);
 
 	/* Get existing block (in case client is busy-waiting) */
 	block = nlmsvc_lookup_block(file, lock, 0);
@@ -314,6 +312,9 @@ nlmsvc_lock(struct svc_rqst *rqstp, struct nlm_file *file,
 	lock->fl.fl_flags |= FL_LOCKD;
 
 again:
+	/* Lock file against concurrent access */
+	down(&file->f_sema);
+
 	if (!(conflock = posix_test_lock(&file->f_file, &lock->fl))) {
 		error = posix_lock_file(&file->f_file, &lock->fl);
 
@@ -346,7 +347,10 @@ nlmsvc_lock(struct svc_rqst *rqstp, struct nlm_file *file,
 
 	/* If we don't have a block, create and initialize it. Then
 	 * retry because we may have slept in kmalloc. */
+	/* We have to release f_sema as nlmsvc_create_block may try to
+	 * to claim it while doing host garbage collection */
 	if (block == NULL) {
+		up(&file->f_sema);
 		dprintk("lockd: blocking on this lock (allocating).\n");
 		if (!(block = nlmsvc_create_block(rqstp, file, lock, cookie)))
 			return nlm_lck_denied_nolocks;

fs/nfsd/export.c

@@ -294,7 +294,9 @@ int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen)
 
 	/* client */
 	len = qword_get(&mesg, buf, PAGE_SIZE);
-	if (len <= 0) return -EINVAL;
+	err = -EINVAL;
+	if (len <= 0) goto out;
+
 	err = -ENOENT;
 	dom = auth_domain_find(buf);
 	if (!dom)
@@ -473,8 +475,14 @@ exp_get_by_name(svc_client *clp, struct vfsmount *mnt, struct dentry *dentry,
 
 	exp = svc_export_lookup(&key, 0);
 	if (exp != NULL) 
-		if (cache_check(&svc_export_cache, &exp->h, reqp))
+		switch (cache_check(&svc_export_cache, &exp->h, reqp)) {
+		case 0: break;
+		case -EAGAIN:
+			exp = ERR_PTR(-EAGAIN);
+			break;
+		default:
 			exp = NULL;
+		}
 
 	return exp;
 }
@@ -915,7 +923,8 @@ struct flags {
 	{ NFSEXP_UIDMAP, {"uidmap", ""}},
 	{ NFSEXP_KERBEROS, { "kerberos", ""}},
 	{ NFSEXP_SUNSECURE, { "sunsecure", ""}},
-	{ NFSEXP_CROSSMNT, {"nohide", ""}},
+	{ NFSEXP_NOHIDE, {"nohide", ""}},
+	{ NFSEXP_CROSSMNT, {"crossmnt", ""}},
 	{ NFSEXP_NOSUBTREECHECK, {"no_subtree_check", ""}},
 	{ NFSEXP_NOAUTHNLM, {"insecure_locks", ""}},
 #ifdef MSNFS

fs/nfsd/vfs.c

@@ -79,7 +79,7 @@ static struct raparms *		raparm_cache;
  * N.B. After this call _both_ fhp and resfh need an fh_put
  *
  * If the lookup would cross a mountpoint, and the mounted filesystem
- * is exported to the client with NFSEXP_CROSSMNT, then the lookup is
+ * is exported to the client with NFSEXP_NOHIDE, then the lookup is
  * accepted as it stands and the mounted directory is
  * returned. Otherwise the covered directory is returned.
  * NOTE: this mountpoint crossing is not supported properly by all
@@ -115,7 +115,7 @@ nfsd_lookup(struct svc_rqst *rqstp, struct svc_fh *fhp, const char *name,
 			read_lock(&dparent_lock);
 			dentry = dget(dparent->d_parent);
 			read_unlock(&dparent_lock);
-		} else  if (!EX_CROSSMNT(exp))
+		} else  if (!EX_NOHIDE(exp))
 			dentry = dget(dparent); /* .. == . just like at / */
 		else {
 			/* checking mountpoint crossing is very different when stepping up */
@@ -133,6 +133,12 @@ nfsd_lookup(struct svc_rqst *rqstp, struct svc_fh *fhp, const char *name,
 
 			exp2 = exp_parent(exp->ex_client, mnt, dentry,
 					  &rqstp->rq_chandle);
+			if (IS_ERR(exp2)) {
+				err = PTR_ERR(exp2);
+				dput(dentry);
+				mntput(mnt);
+				goto out;
+			}
 			if (!exp2) {
 				dput(dentry);
 				dentry = dget(dparent);
@@ -157,9 +163,19 @@ nfsd_lookup(struct svc_rqst *rqstp, struct svc_fh *fhp, const char *name,
 			struct dentry *mounts = dget(dentry);
 			while (follow_down(&mnt,&mounts)&&d_mountpoint(mounts))
 				;
+
 			exp2 = exp_get_by_name(exp->ex_client, mnt, 
 					       mounts, &rqstp->rq_chandle);
-			if (exp2 && EX_CROSSMNT(exp2)) {
+			if (IS_ERR(exp2)) {
+				err = PTR_ERR(exp2);
+				dput(mounts);
+				dput(dentry);
+				mntput(mnt);
+				goto out;
+			}
+			if (exp2 &&
+			    ((exp->ex_flags & NFSEXP_CROSSMNT)
+			     || EX_NOHIDE(exp2))) {
 				/* successfully crossed mount point */
 				exp_put(exp);
 				exp = exp2;

fs/reiserfs/journal.c

@@ -1310,6 +1310,10 @@ static void free_journal_ram(struct super_block *p_s_sb) {
   if (SB_JOURNAL(p_s_sb)->j_header_bh) {
     brelse(SB_JOURNAL(p_s_sb)->j_header_bh) ;
   }
+  /* j_header_bh is on the journal dev, make sure not to release the journal
+   * dev until we brelse j_header_bh
+   */
+  release_journal_dev(p_s_sb, SB_JOURNAL(p_s_sb));
   vfree(SB_JOURNAL(p_s_sb)) ;
 }
 
@@ -1341,7 +1345,6 @@ static int do_journal_release(struct reiserfs_transaction_handle *th, struct sup
     commit_wq = NULL;
   }
 
-  release_journal_dev( p_s_sb, SB_JOURNAL( p_s_sb ) );
   free_journal_ram(p_s_sb) ;
 
   return 0 ;
@@ -1868,23 +1871,17 @@ static int release_journal_dev( struct super_block *super,
     
     result = 0;
 
-
     if( journal -> j_dev_file != NULL ) {
-	/*
-	 * journal block device was taken via filp_open
-	 */
 	result = filp_close( journal -> j_dev_file, NULL );
 	journal -> j_dev_file = NULL;
 	journal -> j_dev_bd = NULL;
     } else if( journal -> j_dev_bd != NULL ) {
-	/*
-	 * journal block device was taken via bdget and blkdev_get
-	 */
 	result = blkdev_put( journal -> j_dev_bd, BDEV_FS );
 	journal -> j_dev_bd = NULL;
     }
+
     if( result != 0 ) {
-	reiserfs_warning("sh-457: release_journal_dev: Cannot release journal device: %i", result );
+	reiserfs_warning("sh-457: release_journal_dev: Cannot release journal device: %i\n", result );
     }
     return result;
 }
@@ -1895,6 +1892,7 @@ static int journal_init_dev( struct super_block *super,
 {
 	int result;
 	dev_t jdev;
+	int blkdev_mode = FMODE_READ | FMODE_WRITE;
 
 	result = 0;
 
@@ -1902,12 +1900,16 @@ static int journal_init_dev( struct super_block *super,
 	journal -> j_dev_file = NULL;
 	jdev = SB_ONDISK_JOURNAL_DEVICE( super ) ?
 		SB_ONDISK_JOURNAL_DEVICE( super ) : super->s_dev;	
+
+	if (bdev_read_only(super->s_bdev))
+	    blkdev_mode = FMODE_READ;
+
 	/* there is no "jdev" option and journal is on separate device */
 	if( ( !jdev_name || !jdev_name[ 0 ] ) ) {
 		journal -> j_dev_bd = bdget(jdev);
 		if( journal -> j_dev_bd )
 			result = blkdev_get( journal -> j_dev_bd, 
-					     FMODE_READ | FMODE_WRITE, 0, 
+					     blkdev_mode, 0, 
 					     BDEV_FS );
 		else
 			result = -ENOMEM;
@@ -1928,10 +1930,10 @@ static int journal_init_dev( struct super_block *super,
 		jdev_inode = journal -> j_dev_file -> f_dentry -> d_inode;
 		journal -> j_dev_bd = jdev_inode -> i_bdev;
 		if( !S_ISBLK( jdev_inode -> i_mode ) ) {
-			printk( "journal_init_dev: '%s' is not a block device", jdev_name );
+			printk( "journal_init_dev: '%s' is not a block device\n", jdev_name );
 			result = -ENOTBLK;
 		} else if( jdev_inode -> i_bdev == NULL ) {
-			printk( "journal_init_dev: bdev uninitialized for '%s'", jdev_name );
+			printk( "journal_init_dev: bdev uninitialized for '%s'\n", jdev_name );
 			result = -ENOMEM;
 		} else  {
 			/* ok */
@@ -1941,12 +1943,12 @@ static int journal_init_dev( struct super_block *super,
 	} else {
 		result = PTR_ERR( journal -> j_dev_file );
 		journal -> j_dev_file = NULL;
-		printk( "journal_init_dev: Cannot open '%s': %i", jdev_name, result );
+		printk( "journal_init_dev: Cannot open '%s': %i\n", jdev_name, result );
 	}
 	if( result != 0 ) {
 		release_journal_dev( super, journal );
 	}
-	printk( "journal_init_dev: journal device: %s", bdevname(journal->j_dev_bd));
+	printk( "journal_init_dev: journal device: %s\n", bdevname(journal->j_dev_bd));
 	return result;
 }
 
@@ -1961,8 +1963,7 @@ int journal_init(struct super_block *p_s_sb, const char * j_dev_name, int old_fo
     struct reiserfs_journal *journal;
 
     if (sizeof(struct reiserfs_journal_commit) != 4096 ||
-      sizeof(struct reiserfs_journal_desc) != 4096
-     ) {
+      sizeof(struct reiserfs_journal_desc) != 4096) {
 	printk("journal-1249: commit or desc struct not 4096 %Zd %Zd\n", sizeof(struct reiserfs_journal_commit), 
         sizeof(struct reiserfs_journal_desc)) ;
 	return 1 ;
@@ -1974,6 +1975,11 @@ int journal_init(struct super_block *p_s_sb, const char * j_dev_name, int old_fo
 	return 1 ;
     }
     memset(journal, 0, sizeof(struct reiserfs_journal)) ;
+    INIT_LIST_HEAD(&SB_JOURNAL(p_s_sb)->j_bitmap_nodes) ;
+    INIT_LIST_HEAD (&SB_JOURNAL(p_s_sb)->j_prealloc_list);
+    reiserfs_allocate_list_bitmaps(p_s_sb, SB_JOURNAL(p_s_sb)->j_list_bitmap, 
+ 				   SB_BMAP_NR(p_s_sb)) ;
+    allocate_bitmap_nodes(p_s_sb) ;
 
     /* reserved for journal area support */
     SB_JOURNAL_1st_RESERVED_BLOCK(p_s_sb) = (old_format ?
@@ -1983,7 +1989,7 @@ int journal_init(struct super_block *p_s_sb, const char * j_dev_name, int old_fo
     
     if( journal_init_dev( p_s_sb, journal, j_dev_name ) != 0 ) {
       printk( "sh-462: unable to initialize jornal device\n");
-      return 1;
+      goto free_and_return;
     }
 
      rs = SB_DISK_SUPER_BLOCK(p_s_sb);
@@ -1993,8 +1999,7 @@ int journal_init(struct super_block *p_s_sb, const char * j_dev_name, int old_fo
 		   SB_ONDISK_JOURNAL_1st_BLOCK(p_s_sb) + SB_ONDISK_JOURNAL_SIZE(p_s_sb));
      if (!bhjh) {
 	 printk("sh-459: unable to read  journal header\n") ;
-	 release_journal_dev(p_s_sb, journal);
-	 return 1 ;
+	 goto free_and_return;
      }
      jh = (struct reiserfs_journal_header *)(bhjh->b_data);
      
@@ -2005,8 +2010,7 @@ int journal_init(struct super_block *p_s_sb, const char * j_dev_name, int old_fo
 		jh->jh_journal.jp_journal_magic, bdevname( SB_JOURNAL(p_s_sb)->j_dev_bd ),
 		sb_jp_journal_magic(rs), reiserfs_bdevname (p_s_sb));
 	 brelse (bhjh);
-	 release_journal_dev(p_s_sb, journal);
-	 return 1 ;
+	 goto free_and_return;
   }
      
   SB_JOURNAL_TRANS_MAX(p_s_sb)      = le32_to_cpu (jh->jh_journal.jp_journal_trans_max);
@@ -2064,7 +2068,6 @@ int journal_init(struct super_block *p_s_sb, const char * j_dev_name, int old_fo
 
   brelse (bhjh);
      
-
   SB_JOURNAL(p_s_sb)->j_list_bitmap_index = 0 ;
   SB_JOURNAL_LIST_INDEX(p_s_sb) = -10000 ; /* make sure flush_old_commits does not try to flush a list while replay is on */
 
@@ -2075,12 +2078,8 @@ int journal_init(struct super_block *p_s_sb, const char * j_dev_name, int old_fo
   memset(SB_JOURNAL(p_s_sb)->j_list_hash_table, 0, JOURNAL_HASH_SIZE * sizeof(struct reiserfs_journal_cnode *)) ;
   memset(journal_writers, 0, sizeof(char *) * 512) ; /* debug code */
 
-  INIT_LIST_HEAD(&SB_JOURNAL(p_s_sb)->j_bitmap_nodes) ;
   INIT_LIST_HEAD(&SB_JOURNAL(p_s_sb)->j_dirty_buffers) ;
   spin_lock_init(&SB_JOURNAL(p_s_sb)->j_dirty_buffers_lock) ;
-  reiserfs_allocate_list_bitmaps(p_s_sb, SB_JOURNAL(p_s_sb)->j_list_bitmap, 
-                                 SB_BMAP_NR(p_s_sb)) ;
-  allocate_bitmap_nodes(p_s_sb) ;
 
   SB_JOURNAL(p_s_sb)->j_start = 0 ;
   SB_JOURNAL(p_s_sb)->j_len = 0 ;
@@ -2107,20 +2106,15 @@ int journal_init(struct super_block *p_s_sb, const char * j_dev_name, int old_fo
   SB_JOURNAL_LIST(p_s_sb)[0].j_list_bitmap = get_list_bitmap(p_s_sb, SB_JOURNAL_LIST(p_s_sb)) ;
   if (!(SB_JOURNAL_LIST(p_s_sb)[0].j_list_bitmap)) {
     reiserfs_warning("journal-2005, get_list_bitmap failed for journal list 0\n") ;
-    release_journal_dev(p_s_sb, journal);
-    return 1 ;
+    goto free_and_return;
   }
   if (journal_read(p_s_sb) < 0) {
     reiserfs_warning("Replay Failure, unable to mount\n") ;
-    free_journal_ram(p_s_sb) ;
-    release_journal_dev(p_s_sb, journal);
-    return 1 ;
+    goto free_and_return;
   }
   SB_JOURNAL_LIST_INDEX(p_s_sb) = 0 ; /* once the read is done, we can set this
                                          where it belongs */
 
-  INIT_LIST_HEAD (&SB_JOURNAL(p_s_sb)->j_prealloc_list);
-
   if (reiserfs_dont_log (p_s_sb))
     return 0;
 
@@ -2129,7 +2123,9 @@ int journal_init(struct super_block *p_s_sb, const char * j_dev_name, int old_fo
     commit_wq = create_workqueue("reiserfs");
 
   return 0 ;
-
+free_and_return:
+  free_journal_ram(p_s_sb);
+  return 1;
 }
 
 /*

include/asm-generic/xor.h

@@ -678,35 +678,35 @@ xor_32regs_p_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
 }
 
 static struct xor_block_template xor_block_8regs = {
-	name: "8regs",
-	do_2: xor_8regs_2,
-	do_3: xor_8regs_3,
-	do_4: xor_8regs_4,
-	do_5: xor_8regs_5,
+	.name = "8regs",
+	.do_2 = xor_8regs_2,
+	.do_3 = xor_8regs_3,
+	.do_4 = xor_8regs_4,
+	.do_5 = xor_8regs_5,
 };
 
 static struct xor_block_template xor_block_32regs = {
-	name: "32regs",
-	do_2: xor_32regs_2,
-	do_3: xor_32regs_3,
-	do_4: xor_32regs_4,
-	do_5: xor_32regs_5,
+	.name = "32regs",
+	.do_2 = xor_32regs_2,
+	.do_3 = xor_32regs_3,
+	.do_4 = xor_32regs_4,
+	.do_5 = xor_32regs_5,
 };
 
 static struct xor_block_template xor_block_8regs_p = {
-	name: "8regs_prefetch",
-	do_2: xor_8regs_p_2,
-	do_3: xor_8regs_p_3,
-	do_4: xor_8regs_p_4,
-	do_5: xor_8regs_p_5,
+	.name = "8regs_prefetch",
+	.do_2 = xor_8regs_p_2,
+	.do_3 = xor_8regs_p_3,
+	.do_4 = xor_8regs_p_4,
+	.do_5 = xor_8regs_p_5,
 };
 
 static struct xor_block_template xor_block_32regs_p = {
-	name: "32regs_prefetch",
-	do_2: xor_32regs_p_2,
-	do_3: xor_32regs_p_3,
-	do_4: xor_32regs_p_4,
-	do_5: xor_32regs_p_5,
+	.name = "32regs_prefetch",
+	.do_2 = xor_32regs_p_2,
+	.do_3 = xor_32regs_p_3,
+	.do_4 = xor_32regs_p_4,
+	.do_5 = xor_32regs_p_5,
 };
 
 #define XOR_TRY_TEMPLATES			\

include/asm-i386/cpufeature.h

@@ -9,9 +9,9 @@
 
 #include <linux/bitops.h>
 
-#define NCAPINTS	4	/* Currently we have 4 32-bit words worth of info */
+#define NCAPINTS	6	/* Currently we have 6 32-bit words worth of info */
 
-/* Intel-defined CPU features, CPUID level 0x00000001, word 0 */
+/* Intel-defined CPU features, CPUID level 0x00000001 (edx), word 0 */
 #define X86_FEATURE_FPU		(0*32+ 0) /* Onboard FPU */
 #define X86_FEATURE_VME		(0*32+ 1) /* Virtual Mode Extensions */
 #define X86_FEATURE_DE		(0*32+ 2) /* Debugging Extensions */
@@ -64,6 +64,11 @@
 #define X86_FEATURE_CYRIX_ARR	(3*32+ 2) /* Cyrix ARRs (= MTRRs) */
 #define X86_FEATURE_CENTAUR_MCR	(3*32+ 3) /* Centaur MCRs (= MTRRs) */
 
+/* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */
+
+/* VIA/Cyrix/Centaur-defined CPU features, CPUID level 0xC0000001, word 5 */
+#define X86_FEATURE_XSTORE	(5*32+ 2) /* on-CPU RNG present (xstore insn) */
+
 
 #define cpu_has(c, bit)		test_bit(bit, (c)->x86_capability)
 #define boot_cpu_has(bit)	test_bit(bit, boot_cpu_data.x86_capability)
@@ -87,6 +92,7 @@
 #define cpu_has_k6_mtrr		boot_cpu_has(X86_FEATURE_K6_MTRR)
 #define cpu_has_cyrix_arr	boot_cpu_has(X86_FEATURE_CYRIX_ARR)
 #define cpu_has_centaur_mcr	boot_cpu_has(X86_FEATURE_CENTAUR_MCR)
+#define cpu_has_xstore		boot_cpu_has(X86_FEATURE_XSTORE)
 
 #endif /* __ASM_I386_CPUFEATURE_H */
 

include/asm-i386/msr.h

@@ -218,6 +218,7 @@
 /* VIA Cyrix defined MSRs*/
 #define MSR_VIA_FCR			0x1107
 #define MSR_VIA_LONGHAUL		0x110a
+#define MSR_VIA_RNG			0x110b
 #define MSR_VIA_BCR2			0x1147
 
 /* Transmeta defined MSRs */

include/asm-i386/xor.h

@@ -25,6 +25,7 @@
 #define XO3(x,y)	"       pxor   8*("#x")(%4), %%mm"#y"   ;\n"
 #define XO4(x,y)	"       pxor   8*("#x")(%5), %%mm"#y"   ;\n"
 
+#include <asm/i387.h>
 
 static void
 xor_pII_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)

include/linux/bio.h

@@ -239,7 +239,6 @@ extern inline char *bio_kmap_irq(struct bio *bio, unsigned long *flags)
 	 * balancing is a lot nicer this way
 	 */
 	local_save_flags(*flags);
-	local_irq_disable();
 	addr = (unsigned long) kmap_atomic(bio_page(bio), KM_BIO_SRC_IRQ);
 
 	if (addr & ~PAGE_MASK)

include/linux/nfsd/export.h

@@ -35,12 +35,13 @@
 #define NFSEXP_UIDMAP		0x0040
 #define NFSEXP_KERBEROS		0x0080		/* not available */
 #define NFSEXP_SUNSECURE	0x0100
-#define NFSEXP_CROSSMNT		0x0200
+#define NFSEXP_NOHIDE		0x0200
 #define NFSEXP_NOSUBTREECHECK	0x0400
 #define	NFSEXP_NOAUTHNLM	0x0800		/* Don't authenticate NLM requests - just trust */
 #define NFSEXP_MSNFS		0x1000	/* do silly things that MS clients expect */
 #define NFSEXP_FSID		0x2000
-#define NFSEXP_ALLFLAGS		0x3FFF
+#define	NFSEXP_CROSSMNT		0x4000
+#define NFSEXP_ALLFLAGS		0x7FFF
 
 
 #ifdef __KERNEL__
@@ -73,7 +74,7 @@ struct svc_expkey {
 #define EX_SECURE(exp)		(!((exp)->ex_flags & NFSEXP_INSECURE_PORT))
 #define EX_ISSYNC(exp)		(!((exp)->ex_flags & NFSEXP_ASYNC))
 #define EX_RDONLY(exp)		((exp)->ex_flags & NFSEXP_READONLY)
-#define EX_CROSSMNT(exp)	((exp)->ex_flags & NFSEXP_CROSSMNT)
+#define EX_NOHIDE(exp)	((exp)->ex_flags & NFSEXP_NOHIDE)
 #define EX_SUNSECURE(exp)	((exp)->ex_flags & NFSEXP_SUNSECURE)
 #define EX_WGATHER(exp)		((exp)->ex_flags & NFSEXP_GATHERED_WRITES)
 

include/linux/raid/md.h

@@ -27,6 +27,7 @@
 #include <linux/module.h>
 #include <linux/hdreg.h>
 #include <linux/proc_fs.h>
+#include <linux/seq_file.h>
 #include <linux/smp_lock.h>
 #include <linux/delay.h>
 #include <net/checksum.h>
@@ -68,13 +69,14 @@ extern inline char * bdev_partition_name (struct block_device *bdev)
 }
 extern int register_md_personality (int p_num, mdk_personality_t *p);
 extern int unregister_md_personality (int p_num);
-extern mdk_thread_t * md_register_thread (void (*run) (void *data),
-				void *data, const char *name);
+extern mdk_thread_t * md_register_thread (void (*run) (mddev_t *mddev),
+				mddev_t *mddev, const char *name);
 extern void md_unregister_thread (mdk_thread_t *thread);
 extern void md_wakeup_thread(mdk_thread_t *thread);
+extern void md_check_recovery(mddev_t *mddev);
 extern void md_interrupt_thread (mdk_thread_t *thread);
 extern void md_write_start(mddev_t *mddev);
-extern void md_write_end(mddev_t *mddev, mdk_thread_t *thread);
+extern void md_write_end(mddev_t *mddev);
 extern void md_handle_safemode(mddev_t *mddev);
 extern void md_done_sync(mddev_t *mddev, int blocks, int ok);
 extern void md_sync_acct(mdk_rdev_t *rdev, unsigned long nr_sectors);

include/linux/raid/md_k.h

@@ -155,6 +155,7 @@ struct mdk_rdev_s
 
 	struct page	*sb_page;
 	int		sb_loaded;
+	sector_t	data_offset;	/* start of data in array */
 	sector_t	sb_offset;
 	int		preferred_minor;	/* autorun support */
 
@@ -206,22 +207,31 @@ struct mddev_s
 
 	char				uuid[16];
 
+	struct mdk_thread_s		*thread;	/* management thread */
 	struct mdk_thread_s		*sync_thread;	/* doing resync or reconstruct */
 	unsigned long			curr_resync;	/* blocks scheduled */
 	unsigned long			resync_mark;	/* a recent timestamp */
 	unsigned long			resync_mark_cnt;/* blocks written at resync_mark */
-	/* recovery_running is 0 for no recovery/resync,
-	 * 1 for active recovery
-	 * 2 for active resync
-	 * -error for an error (e.g. -EINTR)
-	 * it can only be set > 0 under reconfig_sem
+
+	/* recovery/resync flags 
+	 * NEEDED:   we might need to start a resync/recover
+	 * RUNNING:  a thread is running, or about to be started
+	 * SYNC:     actually doing a resync, not a recovery
+	 * ERR:      and IO error was detected - abort the resync/recovery
+	 * INTR:     someone requested a (clean) early abort.
+	 * DONE:     thread is done and is waiting to be reaped
 	 */
-	int				recovery_running;
-	int				recovery_error;	/* error from recovery write */
+#define	MD_RECOVERY_RUNNING	0
+#define	MD_RECOVERY_SYNC	1
+#define	MD_RECOVERY_ERR		2
+#define	MD_RECOVERY_INTR	3
+#define	MD_RECOVERY_DONE	4
+#define	MD_RECOVERY_NEEDED	5
+	unsigned long			recovery;
+
 	int				in_sync;	/* know to not need resync */
 	struct semaphore		reconfig_sem;
 	atomic_t			active;
-	int				spares;
 
 	int				degraded;	/* whether md should consider
 							 * adding a spare
@@ -230,9 +240,11 @@ struct mddev_s
 	atomic_t			recovery_active; /* blocks scheduled, but not written */
 	wait_queue_head_t		recovery_wait;
 	sector_t			recovery_cp;
-	int				safemode;	/* if set, update "clean" superblock
+	unsigned int			safemode;	/* if set, update "clean" superblock
 							 * when no writes pending.
 							 */ 
+	unsigned int			safemode_delay;
+	struct timer_list		safemode_timer;
 	atomic_t			writes_pending; 
 	request_queue_t			queue;	/* for plugging ... */
 
@@ -245,7 +257,7 @@ struct mdk_personality_s
 	int (*make_request)(request_queue_t *q, struct bio *bio);
 	int (*run)(mddev_t *mddev);
 	int (*stop)(mddev_t *mddev);
-	int (*status)(char *page, mddev_t *mddev);
+	void (*status)(struct seq_file *seq, mddev_t *mddev);
 	/* error_handler must set ->faulty and clear ->in_sync
 	 * if appropriate, and should abort recovery if needed 
 	 */
@@ -292,8 +304,8 @@ extern mdk_rdev_t * find_rdev_nr(mddev_t *mddev, int nr);
 	ITERATE_RDEV_GENERIC(pending_raid_disks,rdev,tmp)
 
 typedef struct mdk_thread_s {
-	void			(*run) (void *data);
-	void			*data;
+	void			(*run) (mddev_t *mddev);
+	mddev_t			*mddev;
 	wait_queue_head_t	wqueue;
 	unsigned long           flags;
 	struct completion	*event;

include/linux/raid/md_p.h

@@ -173,5 +173,58 @@ static inline __u64 md_event(mdp_super_t *sb) {
 	return (ev<<32)| sb->events_lo;
 }
 
+/*
+ * The version-1 superblock :
+ * All numeric fields are little-endian.
+ *
+ * total size: 256 bytes plus 2 per device.
+ *  1K allows 384 devices.
+ */
+struct mdp_superblock_1 {
+	/* constant array information - 128 bytes */
+	__u32	magic;		/* MD_SB_MAGIC: 0xa92b4efc - little endian */
+	__u32	major_version;	/* 1 */
+	__u32	feature_map;	/* 0 for now */
+	__u32	pad0;		/* always set to 0 when writing */
+
+	__u8	set_uuid[16];	/* user-space generated. */
+	char	set_name[32];	/* set and interpreted by user-space */
+
+	__u64	ctime;		/* lo 40 bits are seconds, top 24 are microseconds or 0*/
+	__u32	level;		/* -4 (multipath), -1 (linear), 0,1,4,5 */
+	__u32	layout;		/* only for raid5 currently */
+	__u64	size;		/* used size of component devices, in 512byte sectors */
+
+	__u32	chunksize;	/* in 512byte sectors */
+	__u32	raid_disks;
+	__u8	pad1[128-92];	/* set to 0 when written */
+
+	/* constant this-device information - 64 bytes */
+	__u64	data_offset;	/* sector start of data, often 0 */
+	__u64	data_size;	/* sectors in this device that can be used for data */
+	__u64	super_offset;	/* sector start of this superblock */
+	__u64	recovery_offset;/* sectors before this offset (from data_offset) have been recovered */
+	__u32	dev_number;	/* permanent identifier of this  device - not role in raid */
+	__u32	cnt_corrected_read; /* number of read errors that were corrected by re-writing */
+	__u8	device_uuid[16]; /* user-space setable, ignored by kernel */
+	__u8	pad2[64-56];	/* set to 0 when writing */
+
+	/* array state information - 64 bytes */
+	__u64	utime;		/* 40 bits second, 24 btes microseconds */
+	__u64	events;		/* incremented when superblock updated */
+	__u64	resync_offset;	/* data before this offset (from data_offset) known to be in sync */
+	__u32	sb_csum;	/* checksum upto devs[max_dev] */
+	__u32	max_dev;	/* size of devs[] array to consider */
+	__u8	pad3[64-40];	/* set to 0 when writing */
+
+	/* device state information. Indexed by dev_number.
+	 * 2 bytes per device
+	 * Note there are no per-device state flags. State information is rolled
+	 * into the 'roles' value.  If a device is spare or faulty, then it doesn't
+	 * have a meaningful role.
+	 */
+	__u16	dev_roles[0];	/* role in array, or 0xffff for a spare, or 0xfffe for faulty */
+};
+
 #endif 
 

include/linux/raid/multipath.h

@@ -13,7 +13,6 @@ struct multipath_private_data {
 	struct multipath_info	multipaths[MD_SB_DISKS];
 	int			raid_disks;
 	int			working_disks;
-	mdk_thread_t		*thread;
 	spinlock_t		device_lock;
 
 	mempool_t		*pool;

include/linux/raid/raid1.h

@@ -19,7 +19,6 @@ struct r1_private_data_s {
 	int			working_disks;
 	int			last_used;
 	sector_t		next_seq_sect;
-	mdk_thread_t		*thread;
 	spinlock_t		device_lock;
 
 	/* for use when syncing mirrors: */
@@ -34,7 +33,6 @@ struct r1_private_data_s {
 
 	mempool_t *r1bio_pool;
 	mempool_t *r1buf_pool;
-	char thread_name[MD_THREAD_NAME_MAX];
 };
 
 typedef struct r1_private_data_s conf_t;

include/linux/raid/raid5.h

@@ -203,7 +203,6 @@ struct disk_info {
 struct raid5_private_data {
 	struct stripe_head	**stripe_hashtbl;
 	mddev_t			*mddev;
-	mdk_thread_t		*thread;
 	struct disk_info	disks[MD_SB_DISKS];
 	struct disk_info	*spare;
 	int			chunk_size, level, algorithm;
@@ -226,7 +225,6 @@ struct raid5_private_data {
 							 * waiting for 25% to be free
 							 */        
 	spinlock_t		device_lock;
-	char			thread_name[MD_THREAD_NAME_MAX];
 };
 
 typedef struct raid5_private_data raid5_conf_t;

include/linux/sunrpc/cache.h

@@ -190,6 +190,7 @@ RTN *FNAME ARGS										\
 			else read_unlock(&(DETAIL)->hash_lock);				\
 			if (set)							\
 				cache_fresh(DETAIL, &tmp->MEMBER, item->MEMBER.expiry_time); \
+			if (set==1 && new) cache_fresh(DETAIL, &new->MEMBER, 0);	\
 			if (new) (DETAIL)->cache_put(&new->MEMBER, DETAIL);		\
 			return tmp;							\
 		}									\

net/sunrpc/svcauth_unix.c

@@ -441,9 +441,6 @@ svcauth_unix_accept(struct svc_rqst *rqstp, u32 *authp)
 		return SVC_DENIED;
 	}
 
-	/* Put NULL verifier */
-	svc_putu32(resv, RPC_AUTH_NULL);
-	svc_putu32(resv, 0);
 
 	key.m_class = rqstp->rq_server->sv_program->pg_class;
 	key.m_addr = rqstp->rq_addr.sin_addr;
@@ -470,8 +467,13 @@ svcauth_unix_accept(struct svc_rqst *rqstp, u32 *authp)
 		}
 	else rv = SVC_DROP;
 
-	if (rqstp->rq_client == NULL && rqstp->rq_proc != 0)
+	if (rv  == SVC_OK && rqstp->rq_client == NULL && rqstp->rq_proc != 0)
 		goto badcred;
+
+	/* Put NULL verifier */
+	svc_putu32(resv, RPC_AUTH_NULL);
+	svc_putu32(resv, 0);
+
 	return rv;
 
 badcred:

net/sunrpc/svcsock.c

@@ -577,12 +577,15 @@ svc_udp_recvfrom(struct svc_rqst *rqstp)
 
 	if (skb_is_nonlinear(skb)) {
 		/* we have to copy */
+		local_bh_disable();
 		if (csum_partial_copy_to_xdr(&rqstp->rq_arg, skb)) {
+			local_bh_enable();
 			/* checksum error */
 			skb_free_datagram(svsk->sk_sk, skb);
 			svc_sock_received(svsk);
 			return 0;
 		}
+		local_bh_enable();
 		skb_free_datagram(svsk->sk_sk, skb); 
 	} else {
 		/* we can use it in-place */
@@ -1435,7 +1438,7 @@ static struct cache_deferred_req *
 svc_defer(struct cache_req *req)
 {
 	struct svc_rqst *rqstp = container_of(req, struct svc_rqst, rq_chandle);
-	int size = sizeof(struct svc_deferred_req) + (rqstp->rq_arg.head[0].iov_len);
+	int size = sizeof(struct svc_deferred_req) + (rqstp->rq_arg.len);
 	struct svc_deferred_req *dr;
 
 	if (rqstp->rq_arg.page_len)
@@ -1444,6 +1447,7 @@ svc_defer(struct cache_req *req)
 		dr = rqstp->rq_deferred;
 		rqstp->rq_deferred = NULL;
 	} else {
+		int skip  = rqstp->rq_arg.len - rqstp->rq_arg.head[0].iov_len;
 		/* FIXME maybe discard if size too large */
 		dr = kmalloc(size, GFP_KERNEL);
 		if (dr == NULL)
@@ -1452,8 +1456,8 @@ svc_defer(struct cache_req *req)
 		dr->serv = rqstp->rq_server;
 		dr->prot = rqstp->rq_prot;
 		dr->addr = rqstp->rq_addr;
-		dr->argslen = rqstp->rq_arg.head[0].iov_len >> 2;
-		memcpy(dr->args, rqstp->rq_arg.head[0].iov_base, dr->argslen<<2);
+		dr->argslen = rqstp->rq_arg.len >> 2;
+		memcpy(dr->args, rqstp->rq_arg.head[0].iov_base-skip, dr->argslen<<2);
 	}
 	spin_lock(&rqstp->rq_server->sv_lock);
 	rqstp->rq_sock->sk_inuse++;