diff --git a/Documentation/connector/connector.txt b/Documentation/connector/connector.txt
index 54a0a14b..57a314b 100644
--- a/Documentation/connector/connector.txt
+++ b/Documentation/connector/connector.txt
@@ -131,3 +131,47 @@
 be lost due to memory pressure or process' receiving queue overflowed,
 so caller is warned must be prepared. That is why struct cn_msg [main
 connector's message header] contains u32 seq and u32 ack fields.
+
+/*****************************************/
+Userspace usage.
+/*****************************************/
+2.6.14 has a new netlink socket implementation, which by default does not
+allow to send data to netlink groups other than 1.
+So, if to use netlink socket (for example using connector) 
+with different group number userspace application must subscribe to 
+that group. It can be achieved by following pseudocode:
+
+s = socket(PF_NETLINK, SOCK_DGRAM, NETLINK_CONNECTOR);
+
+l_local.nl_family = AF_NETLINK;
+l_local.nl_groups = 12345;
+l_local.nl_pid = 0;
+
+if (bind(s, (struct sockaddr *)&l_local, sizeof(struct sockaddr_nl)) == -1) {
+	perror("bind");
+	close(s);
+	return -1;
+}
+
+{
+	int on = l_local.nl_groups;
+	setsockopt(s, 270, 1, &on, sizeof(on));
+}
+
+Where 270 above is SOL_NETLINK, and 1 is a NETLINK_ADD_MEMBERSHIP socket
+option. To drop multicast subscription one should call above socket option
+with NETLINK_DROP_MEMBERSHIP parameter which is defined as 0.
+
+2.6.14 netlink code only allows to select a group which is less or equal to
+the maximum group number, which is used at netlink_kernel_create() time.
+In case of connector it is CN_NETLINK_USERS + 0xf, so if you want to use
+group number 12345, you must increment CN_NETLINK_USERS to that number.
+Additional 0xf numbers are allocated to be used by non-in-kernel users.
+
+Due to this limitation, group 0xffffffff does not work now, so one can
+not use add/remove connector's group notifications, but as far as I know, 
+only cn_test.c test module used it.
+
+Some work in netlink area is still being done, so things can be changed in
+2.6.15 timeframe, if it will happen, documentation will be updated for that
+kernel.
diff --git a/Documentation/dell_rbu.txt b/Documentation/dell_rbu.txt
index 95d7f62..941343a 100644
--- a/Documentation/dell_rbu.txt
+++ b/Documentation/dell_rbu.txt
@@ -35,6 +35,7 @@
 /sys/class/firmware/dell_rbu/data
 /sys/devices/platform/dell_rbu/image_type
 /sys/devices/platform/dell_rbu/data
+/sys/devices/platform/dell_rbu/packet_size
 
 The driver supports two types of update mechanism; monolithic and packetized.
 These update mechanism depends upon the BIOS currently running on the system.
@@ -47,8 +48,26 @@
 changed to packets during the driver load time by specifying the load
 parameter image_type=packet.  This can also be changed later as below
 echo packet > /sys/devices/platform/dell_rbu/image_type
-Also echoing either mono ,packet or init in to image_type will free up the
-memory allocated by the driver.
+
+In packet update mode the packet size has to be given before any packets can
+be downloaded. It is done as below
+echo XXXX > /sys/devices/platform/dell_rbu/packet_size
+In the packet update mechanism, the user neesd to create a new file having
+packets of data arranged back to back. It can be done as follows
+The user creates packets header, gets the chunk of the BIOS image and
+placs it next to the packetheader; now, the packetheader + BIOS image chunk
+added to geather should match the specified packet_size. This makes one
+packet, the user needs to create more such packets out of the entire BIOS
+image file and then arrange all these packets back to back in to one single
+file.
+This file is then copied to /sys/class/firmware/dell_rbu/data.
+Once this file gets to the driver, the driver extracts packet_size data from
+the file and spreads it accross the physical memory in contiguous packet_sized
+space.
+This method makes sure that all the packets get to the driver in a single operation.
+
+In monolithic update the user simply get the BIOS image (.hdr file) and copies
+to the data file as is without any change to the BIOS image itself.
 
 Do the steps below to download the BIOS image.
 1) echo 1 > /sys/class/firmware/dell_rbu/loading
@@ -58,7 +77,10 @@
 The /sys/class/firmware/dell_rbu/ entries will remain till the following is
 done.
 echo -1 > /sys/class/firmware/dell_rbu/loading.
-Until this step is completed the drivr cannot be unloaded.
+Until this step is completed the driver cannot be unloaded.
+Also echoing either mono ,packet or init in to image_type will free up the
+memory allocated by the driver.
+
 If an user by accident executes steps 1 and 3 above without executing step 2;
 it will make the /sys/class/firmware/dell_rbu/ entries to disappear.
 The entries can be recreated by doing the following
@@ -66,15 +88,11 @@
 NOTE: echoing init in image_type does not change it original value.
 
 Also the driver provides /sys/devices/platform/dell_rbu/data readonly file to
-read back the image downloaded. This is useful in case of packet update
-mechanism where the above steps 1,2,3 will repeated for every packet.
-By reading the /sys/devices/platform/dell_rbu/data file all packet data
-downloaded can be verified in a single file.
-The packets are arranged in this file one after the other in a FIFO order.
+read back the image downloaded.
 
 NOTE:
-This driver requires a patch for firmware_class.c which has the addition
-of request_firmware_nowait_nohotplug function to wortk
+This driver requires a patch for firmware_class.c which has the modified
+request_firmware_nowait function.
 Also after updating the BIOS image an user mdoe application neeeds to execute
 code which message the BIOS update request to the BIOS. So on the next reboot
 the BIOS knows about the new image downloaded and it updates it self.
diff --git a/Makefile b/Makefile
index fdb96bc..504ba3c 100644
--- a/Makefile
+++ b/Makefile
@@ -1,7 +1,7 @@
 VERSION = 2
 PATCHLEVEL = 6
 SUBLEVEL = 14
-EXTRAVERSION =-rc3
+EXTRAVERSION =-rc4
 NAME=Affluent Albatross
 
 # *DOCUMENTATION*
diff --git a/arch/arm/Makefile b/arch/arm/Makefile
index 7779f2d..299bc04 100644
--- a/arch/arm/Makefile
+++ b/arch/arm/Makefile
@@ -53,7 +53,7 @@
 tune-$(CONFIG_CPU_SA110)	:=-mtune=strongarm110
 tune-$(CONFIG_CPU_SA1100)	:=-mtune=strongarm1100
 tune-$(CONFIG_CPU_XSCALE)	:=$(call cc-option,-mtune=xscale,-mtune=strongarm110) -Wa,-mcpu=xscale
-tune-$(CONFIG_CPU_V6)		:=-mtune=strongarm
+tune-$(CONFIG_CPU_V6)		:=$(call cc-option,-mtune=arm1136j-s,-mtune=strongarm)
 
 # Need -Uarm for gcc < 3.x
 CFLAGS_ABI	:=$(call cc-option,-mapcs-32,-mabi=apcs-gnu) $(call cc-option,-mno-thumb-interwork,)
diff --git a/arch/arm/kernel/armksyms.c b/arch/arm/kernel/armksyms.c
index 835d450..7b17a87 100644
--- a/arch/arm/kernel/armksyms.c
+++ b/arch/arm/kernel/armksyms.c
@@ -45,8 +45,8 @@
 
 #define EXPORT_SYMBOL_ALIAS(sym,orig)		\
  EXPORT_CRC_ALIAS(sym)				\
- const struct kernel_symbol __ksymtab_##sym	\
-  __attribute__((section("__ksymtab"))) =	\
+ static const struct kernel_symbol __ksymtab_##sym	\
+  __attribute_used__ __attribute__((section("__ksymtab"))) =	\
     { (unsigned long)&orig, #sym };
 
 /*
diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S
index 81d450a..066597f 100644
--- a/arch/arm/kernel/entry-common.S
+++ b/arch/arm/kernel/entry-common.S
@@ -106,15 +106,10 @@
 	.endm
 
 .Larm700bug:
-	ldr	r0, [sp, #S_PSR]		@ Get calling cpsr
-	sub	lr, lr, #4
-	str	lr, [r8]
-	msr	spsr_cxsf, r0
 	ldmia	sp, {r0 - lr}^			@ Get calling r0 - lr
 	mov	r0, r0
-	ldr	lr, [sp, #S_PC]			@ Get PC
 	add	sp, sp, #S_FRAME_SIZE
-	movs	pc, lr
+	subs	pc, lr, #4
 #else
 	.macro	arm710_bug_check, instr, temp
 	.endm
diff --git a/arch/arm/mach-s3c2410/mach-anubis.c b/arch/arm/mach-s3c2410/mach-anubis.c
index 7c05f27..5ae80f4 100644
--- a/arch/arm/mach-s3c2410/mach-anubis.c
+++ b/arch/arm/mach-s3c2410/mach-anubis.c
@@ -125,7 +125,7 @@
 static int chip0_map[]      = { 0 };
 static int chip1_map[]      = { 1 };
 
-struct mtd_partition anubis_default_nand_part[] = {
+static struct mtd_partition anubis_default_nand_part[] = {
 	[0] = {
 		.name	= "Boot Agent",
 		.size	= SZ_16K,
diff --git a/arch/arm/mach-s3c2410/mach-bast.c b/arch/arm/mach-s3c2410/mach-bast.c
index ed1f07d..8ca9559 100644
--- a/arch/arm/mach-s3c2410/mach-bast.c
+++ b/arch/arm/mach-s3c2410/mach-bast.c
@@ -230,7 +230,7 @@
 static int chip1_map[] = { 2 };
 static int chip2_map[] = { 3 };
 
-struct mtd_partition bast_default_nand_part[] = {
+static struct mtd_partition bast_default_nand_part[] = {
 	[0] = {
 		.name	= "Boot Agent",
 		.size	= SZ_16K,
@@ -340,7 +340,7 @@
  * better IO routines can be written and tested
 */
 
-struct dm9000_plat_data bast_dm9k_platdata = {
+static struct dm9000_plat_data bast_dm9k_platdata = {
 	.flags		= DM9000_PLATF_16BITONLY
 };
 
diff --git a/arch/arm/mach-s3c2410/mach-vr1000.c b/arch/arm/mach-s3c2410/mach-vr1000.c
index 663a7f9..46b2596 100644
--- a/arch/arm/mach-s3c2410/mach-vr1000.c
+++ b/arch/arm/mach-s3c2410/mach-vr1000.c
@@ -288,7 +288,7 @@
  * better IO routines can be written and tested
 */
 
-struct dm9000_plat_data vr1000_dm9k_platdata = {
+static struct dm9000_plat_data vr1000_dm9k_platdata = {
 	.flags		= DM9000_PLATF_16BITONLY,
 };
 
diff --git a/arch/arm/mach-s3c2410/s3c2410.c b/arch/arm/mach-s3c2410/s3c2410.c
index 0b88993..a8bf5ec 100644
--- a/arch/arm/mach-s3c2410/s3c2410.c
+++ b/arch/arm/mach-s3c2410/s3c2410.c
@@ -125,9 +125,6 @@
 	&s3c_uart2
 };
 
-/* store our uart devices for the serial driver console */
-struct platform_device *s3c2410_uart_devices[3];
-
 static int s3c2410_uart_count = 0;
 
 /* uart registration process */
diff --git a/arch/arm/mach-s3c2410/s3c2440.c b/arch/arm/mach-s3c2410/s3c2440.c
index d4c8281..833fa36 100644
--- a/arch/arm/mach-s3c2410/s3c2440.c
+++ b/arch/arm/mach-s3c2410/s3c2440.c
@@ -151,7 +151,7 @@
 
 #ifdef CONFIG_PM
 
-struct sleep_save s3c2440_sleep[] = {
+static struct sleep_save s3c2440_sleep[] = {
 	SAVE_ITEM(S3C2440_DSC0),
 	SAVE_ITEM(S3C2440_DSC1),
 	SAVE_ITEM(S3C2440_GPJDAT),
@@ -260,7 +260,7 @@
  * as a driver which may support both 2410 and 2440 may try and use it.
 */
 
-int __init s3c2440_core_init(void)
+static int __init s3c2440_core_init(void)
 {
 	return sysdev_class_register(&s3c2440_sysclass);
 }
diff --git a/arch/arm/mach-s3c2410/time.c b/arch/arm/mach-s3c2410/time.c
index c0acfb2..8a00e3c 100644
--- a/arch/arm/mach-s3c2410/time.c
+++ b/arch/arm/mach-s3c2410/time.c
@@ -38,6 +38,7 @@
 #include <asm/hardware/clock.h>
 
 #include "clock.h"
+#include "cpu.h"
 
 static unsigned long timer_startval;
 static unsigned long timer_usec_ticks;
diff --git a/arch/arm/mm/alignment.c b/arch/arm/mm/alignment.c
index f35e69e..705c989 100644
--- a/arch/arm/mm/alignment.c
+++ b/arch/arm/mm/alignment.c
@@ -111,7 +111,7 @@
 }
 
 static int proc_alignment_write(struct file *file, const char __user *buffer,
-			       unsigned long count, void *data)
+				unsigned long count, void *data)
 {
 	char mode;
 
@@ -119,7 +119,7 @@
 		if (get_user(mode, buffer))
 			return -EFAULT;
 		if (mode >= '0' && mode <= '5')
-			   ai_usermode = mode - '0';
+			ai_usermode = mode - '0';
 	}
 	return count;
 }
@@ -262,7 +262,7 @@
 			goto fault;				\
 	} while (0)
 
-#define put32_unaligned_check(val,addr)	 \
+#define put32_unaligned_check(val,addr) \
 	__put32_unaligned_check("strb", val, addr)
 
 #define put32t_unaligned_check(val,addr) \
@@ -306,19 +306,19 @@
 	return TYPE_LDST;
 
  user:
- 	if (LDST_L_BIT(instr)) {
- 		unsigned long val;
- 		get16t_unaligned_check(val, addr);
+	if (LDST_L_BIT(instr)) {
+		unsigned long val;
+		get16t_unaligned_check(val, addr);
 
- 		/* signed half-word? */
- 		if (instr & 0x40)
- 			val = (signed long)((signed short) val);
+		/* signed half-word? */
+		if (instr & 0x40)
+			val = (signed long)((signed short) val);
 
- 		regs->uregs[rd] = val;
- 	} else
- 		put16t_unaligned_check(regs->uregs[rd], addr);
+		regs->uregs[rd] = val;
+	} else
+		put16t_unaligned_check(regs->uregs[rd], addr);
 
- 	return TYPE_LDST;
+	return TYPE_LDST;
 
  fault:
 	return TYPE_FAULT;
@@ -342,11 +342,11 @@
 		unsigned long val;
 		get32_unaligned_check(val, addr);
 		regs->uregs[rd] = val;
-		get32_unaligned_check(val, addr+4);
-		regs->uregs[rd+1] = val;
+		get32_unaligned_check(val, addr + 4);
+		regs->uregs[rd + 1] = val;
 	} else {
 		put32_unaligned_check(regs->uregs[rd], addr);
-		put32_unaligned_check(regs->uregs[rd+1], addr+4);
+		put32_unaligned_check(regs->uregs[rd + 1], addr + 4);
 	}
 
 	return TYPE_LDST;
@@ -356,11 +356,11 @@
 		unsigned long val;
 		get32t_unaligned_check(val, addr);
 		regs->uregs[rd] = val;
-		get32t_unaligned_check(val, addr+4);
-		regs->uregs[rd+1] = val;
+		get32t_unaligned_check(val, addr + 4);
+		regs->uregs[rd + 1] = val;
 	} else {
 		put32t_unaligned_check(regs->uregs[rd], addr);
-		put32t_unaligned_check(regs->uregs[rd+1], addr+4);
+		put32t_unaligned_check(regs->uregs[rd + 1], addr + 4);
 	}
 
 	return TYPE_LDST;
@@ -443,7 +443,7 @@
 	if (LDST_P_EQ_U(instr))	/* U = P */
 		eaddr += 4;
 
-	/* 
+	/*
 	 * For alignment faults on the ARM922T/ARM920T the MMU  makes
 	 * the FSR (and hence addr) equal to the updated base address
 	 * of the multiple access rather than the restored value.
@@ -570,7 +570,7 @@
 	/* 6.5.1 Format 3: */
 	case 0x4800 >> 11:				/* 7.1.28 LDR(3) */
 		/* NOTE: This case is not technically possible. We're
-		 * 	 loading 32-bit memory data via PC relative
+		 *	 loading 32-bit memory data via PC relative
 		 *	 addressing mode. So we can and should eliminate
 		 *	 this case. But I'll leave it here for now.
 		 */
@@ -642,7 +642,7 @@
 
 	if (fault) {
 		type = TYPE_FAULT;
- 		goto bad_or_fault;
+		goto bad_or_fault;
 	}
 
 	if (user_mode(regs))
diff --git a/arch/arm/nwfpe/fpa11.c b/arch/arm/nwfpe/fpa11.c
index 7690f73..7b3d74d 100644
--- a/arch/arm/nwfpe/fpa11.c
+++ b/arch/arm/nwfpe/fpa11.c
@@ -31,11 +31,6 @@
 #include <linux/string.h>
 #include <asm/system.h>
 
-/* forward declarations */
-unsigned int EmulateCPDO(const unsigned int);
-unsigned int EmulateCPDT(const unsigned int);
-unsigned int EmulateCPRT(const unsigned int);
-
 /* Reset the FPA11 chip.  Called to initialize and reset the emulator. */
 static void resetFPA11(void)
 {
diff --git a/arch/arm/nwfpe/fpa11.h b/arch/arm/nwfpe/fpa11.h
index 93523ae..9677ae8 100644
--- a/arch/arm/nwfpe/fpa11.h
+++ b/arch/arm/nwfpe/fpa11.h
@@ -95,4 +95,24 @@
 extern int8 SetRoundingPrecision(const unsigned int);
 extern void nwfpe_init_fpa(union fp_state *fp);
 
+extern unsigned int EmulateAll(unsigned int opcode);
+
+extern unsigned int EmulateCPDT(const unsigned int opcode);
+extern unsigned int EmulateCPDO(const unsigned int opcode);
+extern unsigned int EmulateCPRT(const unsigned int opcode);
+
+/* fpa11_cpdt.c */
+extern unsigned int PerformLDF(const unsigned int opcode);
+extern unsigned int PerformSTF(const unsigned int opcode);
+extern unsigned int PerformLFM(const unsigned int opcode);
+extern unsigned int PerformSFM(const unsigned int opcode);
+
+/* single_cpdo.c */
+
+extern unsigned int SingleCPDO(struct roundingData *roundData,
+			       const unsigned int opcode, FPREG * rFd);
+/* double_cpdo.c */
+extern unsigned int DoubleCPDO(struct roundingData *roundData,
+			       const unsigned int opcode, FPREG * rFd);
+
 #endif
diff --git a/arch/arm/nwfpe/fpa11_cprt.c b/arch/arm/nwfpe/fpa11_cprt.c
index adf8d30..7c67023 100644
--- a/arch/arm/nwfpe/fpa11_cprt.c
+++ b/arch/arm/nwfpe/fpa11_cprt.c
@@ -26,12 +26,11 @@
 #include "fpa11.inl"
 #include "fpmodule.h"
 #include "fpmodule.inl"
+#include "softfloat.h"
 
 #ifdef CONFIG_FPE_NWFPE_XP
 extern flag floatx80_is_nan(floatx80);
 #endif
-extern flag float64_is_nan(float64);
-extern flag float32_is_nan(float32);
 
 unsigned int PerformFLT(const unsigned int opcode);
 unsigned int PerformFIX(const unsigned int opcode);
diff --git a/arch/arm/nwfpe/fpopcode.h b/arch/arm/nwfpe/fpopcode.h
index 1777e92..6528e08 100644
--- a/arch/arm/nwfpe/fpopcode.h
+++ b/arch/arm/nwfpe/fpopcode.h
@@ -476,4 +476,10 @@
 	return (nRc);
 }
 
+extern unsigned int checkCondition(const unsigned int opcode,
+				   const unsigned int ccodes);
+
+extern const float64 float64Constant[];
+extern const float32 float32Constant[];
+
 #endif
diff --git a/arch/arm/nwfpe/softfloat.h b/arch/arm/nwfpe/softfloat.h
index 1c8799b..1415170 100644
--- a/arch/arm/nwfpe/softfloat.h
+++ b/arch/arm/nwfpe/softfloat.h
@@ -265,4 +265,7 @@
 	return (a != b) && (aSign ^ (a < b));
 }
 
+extern flag float32_is_nan( float32 a );
+extern flag float64_is_nan( float64 a );
+
 #endif
diff --git a/arch/cris/arch-v32/kernel/smp.c b/arch/cris/arch-v32/kernel/smp.c
index 2c5cae0..957f551 100644
--- a/arch/cris/arch-v32/kernel/smp.c
+++ b/arch/cris/arch-v32/kernel/smp.c
@@ -15,6 +15,7 @@
 #include <linux/kernel.h>
 #include <linux/cpumask.h>
 #include <linux/interrupt.h>
+#include <linux/module.h>
 
 #define IPI_SCHEDULE 1
 #define IPI_CALL 2
@@ -28,6 +29,7 @@
 /* CPU masks */
 cpumask_t cpu_online_map = CPU_MASK_NONE;
 cpumask_t phys_cpu_present_map = CPU_MASK_NONE;
+EXPORT_SYMBOL(phys_cpu_present_map);
 
 /* Variables used during SMP boot */
 volatile int cpu_now_booting = 0;
diff --git a/arch/m32r/kernel/entry.S b/arch/m32r/kernel/entry.S
index dddbf6b..85920fb 100644
--- a/arch/m32r/kernel/entry.S
+++ b/arch/m32r/kernel/entry.S
@@ -681,6 +681,15 @@
 	bl	do_debug_trap
 	bra	error_code
 
+ENTRY(ill_trap)
+	/* void ill_trap(void) */
+	SWITCH_TO_KERNEL_STACK
+	SAVE_ALL
+	ldi	r1, #0				; error_code ; FIXME
+	mv	r0, sp				; pt_regs
+	bl	do_ill_trap
+	bra	error_code
+
 
 /* Cache flushing handler */
 ENTRY(cache_flushing_handler)
diff --git a/arch/m32r/kernel/traps.c b/arch/m32r/kernel/traps.c
index 0192227..5fe8ed6 100644
--- a/arch/m32r/kernel/traps.c
+++ b/arch/m32r/kernel/traps.c
@@ -5,8 +5,6 @@
  *                            Hitoshi Yamamoto
  */
 
-/* $Id$ */
-
 /*
  * 'traps.c' handles hardware traps and faults after we have saved some
  * state in 'entry.S'.
@@ -35,6 +33,7 @@
 asmlinkage void rie_handler(void);
 asmlinkage void debug_trap(void);
 asmlinkage void cache_flushing_handler(void);
+asmlinkage void ill_trap(void);
 
 #ifdef CONFIG_SMP
 extern void smp_reschedule_interrupt(void);
@@ -77,22 +76,22 @@
 	eit_vector[5] = BRA_INSN(default_eit_handler, 5);
 	eit_vector[8] = BRA_INSN(rie_handler, 8);
 	eit_vector[12] = BRA_INSN(alignment_check, 12);
-	eit_vector[16] = 0xff000000UL;
+	eit_vector[16] = BRA_INSN(ill_trap, 16);
 	eit_vector[17] = BRA_INSN(debug_trap, 17);
 	eit_vector[18] = BRA_INSN(system_call, 18);
-	eit_vector[19] = 0xff000000UL;
-	eit_vector[20] = 0xff000000UL;
-	eit_vector[21] = 0xff000000UL;
-	eit_vector[22] = 0xff000000UL;
-	eit_vector[23] = 0xff000000UL;
-	eit_vector[24] = 0xff000000UL;
-	eit_vector[25] = 0xff000000UL;
-	eit_vector[26] = 0xff000000UL;
-	eit_vector[27] = 0xff000000UL;
+	eit_vector[19] = BRA_INSN(ill_trap, 19);
+	eit_vector[20] = BRA_INSN(ill_trap, 20);
+	eit_vector[21] = BRA_INSN(ill_trap, 21);
+	eit_vector[22] = BRA_INSN(ill_trap, 22);
+	eit_vector[23] = BRA_INSN(ill_trap, 23);
+	eit_vector[24] = BRA_INSN(ill_trap, 24);
+	eit_vector[25] = BRA_INSN(ill_trap, 25);
+	eit_vector[26] = BRA_INSN(ill_trap, 26);
+	eit_vector[27] = BRA_INSN(ill_trap, 27);
 	eit_vector[28] = BRA_INSN(cache_flushing_handler, 28);
-	eit_vector[29] = 0xff000000UL;
-	eit_vector[30] = 0xff000000UL;
-	eit_vector[31] = 0xff000000UL;
+	eit_vector[29] = BRA_INSN(ill_trap, 29);
+	eit_vector[30] = BRA_INSN(ill_trap, 30);
+	eit_vector[31] = BRA_INSN(ill_trap, 31);
 	eit_vector[32] = BRA_INSN(ei_handler, 32);
 	eit_vector[64] = BRA_INSN(pie_handler, 64);
 #ifdef CONFIG_MMU
@@ -286,7 +285,8 @@
 
 DO_ERROR( 1, SIGTRAP, "debug trap", debug_trap)
 DO_ERROR_INFO(0x20, SIGILL,  "reserved instruction ", rie_handler, ILL_ILLOPC, regs->bpc)
-DO_ERROR_INFO(0x100, SIGILL,  "privilege instruction", pie_handler, ILL_PRVOPC, regs->bpc)
+DO_ERROR_INFO(0x100, SIGILL,  "privileged instruction", pie_handler, ILL_PRVOPC, regs->bpc)
+DO_ERROR_INFO(-1, SIGILL,  "illegal trap", ill_trap, ILL_ILLTRP, regs->bpc)
 
 extern int handle_unaligned_access(unsigned long, struct pt_regs *);
 
@@ -329,4 +329,3 @@
 		set_fs(oldfs);
 	}
 }
-
diff --git a/arch/ppc/kernel/cputable.c b/arch/ppc/kernel/cputable.c
index 546e1ea..6b76cf5 100644
--- a/arch/ppc/kernel/cputable.c
+++ b/arch/ppc/kernel/cputable.c
@@ -91,7 +91,7 @@
 		.cpu_features		= CPU_FTR_COMMON | CPU_FTR_601 |
 			CPU_FTR_HPTE_TABLE,
 		.cpu_user_features 	= COMMON_PPC | PPC_FEATURE_601_INSTR |
-			PPC_FEATURE_UNIFIED_CACHE,
+			PPC_FEATURE_UNIFIED_CACHE | PPC_FEATURE_NO_TB,
 		.icache_bsize		= 32,
 		.dcache_bsize		= 32,
 		.cpu_setup		= __setup_cpu_601
@@ -745,7 +745,8 @@
 		.cpu_name		= "403GCX",
 		.cpu_features		= CPU_FTR_SPLIT_ID_CACHE |
 			CPU_FTR_USE_TB,
-		.cpu_user_features	= PPC_FEATURE_32 | PPC_FEATURE_HAS_MMU,
+		.cpu_user_features	= PPC_FEATURE_32 |
+		 	PPC_FEATURE_HAS_MMU | PPC_FEATURE_NO_TB,
 		.icache_bsize		= 16,
 		.dcache_bsize		= 16,
 	},
diff --git a/arch/ppc/kernel/dma-mapping.c b/arch/ppc/kernel/dma-mapping.c
index b566d98..8edee80 100644
--- a/arch/ppc/kernel/dma-mapping.c
+++ b/arch/ppc/kernel/dma-mapping.c
@@ -401,10 +401,10 @@
 static inline void __dma_sync_page_highmem(struct page *page,
 		unsigned long offset, size_t size, int direction)
 {
-	size_t seg_size = min((size_t)PAGE_SIZE, size) - offset;
+	size_t seg_size = min((size_t)(PAGE_SIZE - offset), size);
 	size_t cur_size = seg_size;
 	unsigned long flags, start, seg_offset = offset;
-	int nr_segs = PAGE_ALIGN(size + (PAGE_SIZE - offset))/PAGE_SIZE;
+	int nr_segs = 1 + ((size - seg_size) + PAGE_SIZE - 1)/PAGE_SIZE;
 	int seg_nr = 0;
 
 	local_irq_save(flags);
diff --git a/arch/ppc64/kernel/module.c b/arch/ppc64/kernel/module.c
index c683bf8..928b858 100644
--- a/arch/ppc64/kernel/module.c
+++ b/arch/ppc64/kernel/module.c
@@ -341,6 +341,19 @@
 			*(unsigned long *)location = my_r2(sechdrs, me);
 			break;
 
+		case R_PPC64_TOC16:
+			/* Subtact TOC pointer */
+			value -= my_r2(sechdrs, me);
+			if (value + 0x8000 > 0xffff) {
+				printk("%s: bad TOC16 relocation (%lu)\n",
+				       me->name, value);
+				return -ENOEXEC;
+			}
+			*((uint16_t *) location)
+				= (*((uint16_t *) location) & ~0xffff)
+				| (value & 0xffff);
+			break;
+
 		case R_PPC64_TOC16_DS:
 			/* Subtact TOC pointer */
 			value -= my_r2(sechdrs, me);
diff --git a/arch/ppc64/kernel/pSeries_pci.c b/arch/ppc64/kernel/pSeries_pci.c
index 1f5f141..928f8fe 100644
--- a/arch/ppc64/kernel/pSeries_pci.c
+++ b/arch/ppc64/kernel/pSeries_pci.c
@@ -32,7 +32,7 @@
 
 #include "pci.h"
 
-static int __initdata s7a_workaround = -1;
+static int __devinitdata s7a_workaround = -1;
 
 #if 0
 void pcibios_name_device(struct pci_dev *dev)
@@ -60,7 +60,7 @@
 DECLARE_PCI_FIXUP_HEADER(PCI_ANY_ID, PCI_ANY_ID, pcibios_name_device);
 #endif
 
-static void __init check_s7a(void)
+static void __devinit check_s7a(void)
 {
 	struct device_node *root;
 	char *model;
diff --git a/arch/sh/kernel/smp.c b/arch/sh/kernel/smp.c
index 56a39d6..5ecefc0 100644
--- a/arch/sh/kernel/smp.c
+++ b/arch/sh/kernel/smp.c
@@ -22,6 +22,7 @@
 #include <linux/time.h>
 #include <linux/timex.h>
 #include <linux/sched.h>
+#include <linux/module.h>
 
 #include <asm/atomic.h>
 #include <asm/processor.h>
@@ -39,6 +40,8 @@
 extern void per_cpu_trap_init(void);
 
 cpumask_t cpu_possible_map;
+EXPORT_SYMBOL(cpu_possible_map);
+
 cpumask_t cpu_online_map;
 static atomic_t cpus_booted = ATOMIC_INIT(0);
 
diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig
index f7c51b8..6537445 100644
--- a/arch/sparc/Kconfig
+++ b/arch/sparc/Kconfig
@@ -21,10 +21,6 @@
 	bool
 	default y
 
-config GENERIC_IOMAP
-	bool
-	default y
-
 source "init/Kconfig"
 
 menu "General machine setup"
diff --git a/arch/sparc/defconfig b/arch/sparc/defconfig
index 8a3aef1e..a698562 100644
--- a/arch/sparc/defconfig
+++ b/arch/sparc/defconfig
@@ -5,7 +5,6 @@
 CONFIG_UID16=y
 CONFIG_HIGHMEM=y
 CONFIG_GENERIC_ISA_DMA=y
-CONFIG_GENERIC_IOMAP=y
 
 #
 # Code maturity level options
diff --git a/arch/sparc64/kernel/dtlb_base.S b/arch/sparc64/kernel/dtlb_base.S
index 702d349..6528786 100644
--- a/arch/sparc64/kernel/dtlb_base.S
+++ b/arch/sparc64/kernel/dtlb_base.S
@@ -53,19 +53,18 @@
  * be guaranteed to be 0 ... mmu_context.h does guarantee this
  * by only using 10 bits in the hwcontext value.
  */
-#define CREATE_VPTE_OFFSET1(r1, r2)
+#define CREATE_VPTE_OFFSET1(r1, r2) nop
 #define CREATE_VPTE_OFFSET2(r1, r2) \
 				srax	r1, 10, r2
-#define CREATE_VPTE_NOP		nop
 #else
 #define CREATE_VPTE_OFFSET1(r1, r2) \
 				srax	r1, PAGE_SHIFT, r2
 #define CREATE_VPTE_OFFSET2(r1, r2) \
 				sllx	r2, 3, r2
-#define CREATE_VPTE_NOP
 #endif
 
 /* DTLB ** ICACHE line 1: Quick user TLB misses		*/
+	mov		TLB_SFSR, %g1
 	ldxa		[%g1 + %g1] ASI_DMMU, %g4	! Get TAG_ACCESS
 	andcc		%g4, TAG_CONTEXT_BITS, %g0	! From Nucleus?
 from_tl1_trap:
@@ -74,18 +73,16 @@
 	be,pn		%xcc, kvmap			! Yep, special processing
 	 CREATE_VPTE_OFFSET2(%g4, %g6)			! Create VPTE offset
 	cmp		%g5, 4				! Last trap level?
-	be,pn		%xcc, longpath			! Yep, cannot risk VPTE miss
-	 nop						! delay slot
 
 /* DTLB ** ICACHE line 2: User finish + quick kernel TLB misses	*/
+	be,pn		%xcc, longpath			! Yep, cannot risk VPTE miss
+	 nop						! delay slot
 	ldxa		[%g3 + %g6] ASI_S, %g5		! Load VPTE
 1:	brgez,pn	%g5, longpath			! Invalid, branch out
 	 nop						! Delay-slot
 9:	stxa		%g5, [%g0] ASI_DTLB_DATA_IN	! Reload TLB
 	retry						! Trap return
 	nop
-	nop
-	nop
 
 /* DTLB ** ICACHE line 3: winfixups+real_faults		*/
 longpath:
@@ -106,8 +103,7 @@
 	nop
 	nop
 	nop
-	CREATE_VPTE_NOP
+	nop
 
 #undef CREATE_VPTE_OFFSET1
 #undef CREATE_VPTE_OFFSET2
-#undef CREATE_VPTE_NOP
diff --git a/arch/sparc64/kernel/dtlb_prot.S b/arch/sparc64/kernel/dtlb_prot.S
index d848bb7..e0a9201 100644
--- a/arch/sparc64/kernel/dtlb_prot.S
+++ b/arch/sparc64/kernel/dtlb_prot.S
@@ -14,14 +14,14 @@
  */
 
 /* PROT ** ICACHE line 1: User DTLB protection trap	*/
-	stxa		%g0, [%g1] ASI_DMMU		! Clear SFSR FaultValid bit
-	membar		#Sync				! Synchronize ASI stores
-	rdpr		%pstate, %g5			! Move into alternate globals
+	mov		TLB_SFSR, %g1
+	stxa		%g0, [%g1] ASI_DMMU		! Clear FaultValid bit
+	membar		#Sync				! Synchronize stores
+	rdpr		%pstate, %g5			! Move into alt-globals
 	wrpr		%g5, PSTATE_AG|PSTATE_MG, %pstate
-	rdpr		%tl, %g1			! Need to do a winfixup?
+	rdpr		%tl, %g1			! Need a winfixup?
 	cmp		%g1, 1				! Trap level >1?
-	mov		TLB_TAG_ACCESS, %g4		! Prepare reload of vaddr
-	nop
+	mov		TLB_TAG_ACCESS, %g4		! For reload of vaddr
 
 /* PROT ** ICACHE line 2: More real fault processing */
 	bgu,pn		%xcc, winfix_trampoline		! Yes, perform winfixup
diff --git a/arch/sparc64/kernel/head.S b/arch/sparc64/kernel/head.S
index f1dcdf8..b49dcd4 100644
--- a/arch/sparc64/kernel/head.S
+++ b/arch/sparc64/kernel/head.S
@@ -28,19 +28,14 @@
 #include <asm/mmu.h>
 	
 /* This section from from _start to sparc64_boot_end should fit into
- * 0x0000.0000.0040.4000 to 0x0000.0000.0040.8000 and will be sharing space
- * with bootup_user_stack, which is from 0x0000.0000.0040.4000 to
- * 0x0000.0000.0040.6000 and empty_bad_page, which is from
- * 0x0000.0000.0040.6000 to 0x0000.0000.0040.8000. 
+ * 0x0000000000404000 to 0x0000000000408000.
  */
-
 	.text
 	.globl	start, _start, stext, _stext
 _start:
 start:
 _stext:
 stext:
-bootup_user_stack:
 ! 0x0000000000404000
 	b	sparc64_boot
 	 flushw					/* Flush register file.      */
@@ -191,8 +186,9 @@
 	stx	%l3, [%sp + 2047 + 128 + 0x10]	! num_rets, 5
 	stx	%l2, [%sp + 2047 + 128 + 0x18]	! arg1: "translate"
 	stx	%l5, [%sp + 2047 + 128 + 0x20]	! arg2: prom_mmu_ihandle_cache
-	srlx	%l0, 22, %l3
-	sllx	%l3, 22, %l3
+	/* PAGE align */
+	srlx	%l0, 13, %l3
+	sllx	%l3, 13, %l3
 	stx	%l3, [%sp + 2047 + 128 + 0x28]	! arg3: vaddr, our PC
 	stx	%g0, [%sp + 2047 + 128 + 0x30]	! res1
 	stx	%g0, [%sp + 2047 + 128 + 0x38]	! res2
@@ -211,6 +207,9 @@
 	ldx	[%sp + 2047 + 128 + 0x48], %l2	! physaddr high
 	stx	%l2, [%l4 + 0x0]
 	ldx	[%sp + 2047 + 128 + 0x50], %l3	! physaddr low
+	/* 4MB align */
+	srlx	%l3, 22, %l3
+	sllx	%l3, 22, %l3
 	stx	%l3, [%l4 + 0x8]
 
 	/* Leave service as-is, "call-method" */
@@ -388,31 +387,30 @@
 	 * former does use this code, the latter does not yet due
 	 * to some complexities.  That should be fixed up at some
 	 * point.
+	 *
+	 * There used to be enormous complexity wrt. transferring
+	 * over from the firwmare's trap table to the Linux kernel's.
+	 * For example, there was a chicken & egg problem wrt. building
+	 * the OBP page tables, yet needing to be on the Linux kernel
+	 * trap table (to translate PAGE_OFFSET addresses) in order to
+	 * do that.
+	 *
+	 * We now handle OBP tlb misses differently, via linear lookups
+	 * into the prom_trans[] array.  So that specific problem no
+	 * longer exists.  Yet, unfortunately there are still some issues
+	 * preventing trampoline.S from using this code... ho hum.
 	 */
 	.globl	setup_trap_table
 setup_trap_table:
 	save	%sp, -192, %sp
 
-	/* Force interrupts to be disabled.  Transferring over to
-	 * the Linux trap table is a very delicate operation.
-	 * Until we are actually on the Linux trap table, we cannot
-	 * get the PAGE_OFFSET linear mappings translated.  We need
-	 * that mapping to be setup in order to initialize the firmware
-	 * page tables.
-	 *
-	 * So there is this window of time, from the return from
-	 * prom_set_trap_table() until inherit_prom_mappings_post()
-	 * (in arch/sparc64/mm/init.c) completes, during which no
-	 * firmware address space accesses can be made.
-	 */
+	/* Force interrupts to be disabled. */
 	rdpr	%pstate, %o1
 	andn	%o1, PSTATE_IE, %o1
 	wrpr	%o1, 0x0, %pstate
 	wrpr	%g0, 15, %pil
 
-	/* Ok, now make the final valid firmware call to jump over
-	 * to the Linux trap table.
-	 */
+	/* Make the firmware call to jump over to the Linux trap table.  */
 	call	prom_set_trap_table
 	 sethi	%hi(sparc64_ttable_tl0), %o0
 
@@ -536,15 +534,21 @@
 
 	ret
 	 restore
+sparc64_boot_end:
+
+#include "systbls.S"
+#include "ktlb.S"
+#include "etrap.S"
+#include "rtrap.S"
+#include "winfixup.S"
+#include "entry.S"
 
 /*
- * The following skips make sure the trap table in ttable.S is aligned
+ * The following skip makes sure the trap table in ttable.S is aligned
  * on a 32K boundary as required by the v9 specs for TBA register.
  */
-sparc64_boot_end:
-	.skip	0x2000 + _start - sparc64_boot_end
-bootup_user_stack_end:
-	.skip	0x2000
+1:
+	.skip	0x4000 + _start - 1b
 
 #ifdef CONFIG_SBUS
 /* This is just a hack to fool make depend config.h discovering
@@ -556,15 +560,6 @@
 ! 0x0000000000408000
 
 #include "ttable.S"
-#include "systbls.S"
-#include "ktlb.S"
-#include "etrap.S"
-#include "rtrap.S"
-#include "winfixup.S"
-#include "entry.S"
-
-	/* This is just anal retentiveness on my part... */
-	.align	16384
 
 	.data
 	.align	8
diff --git a/arch/sparc64/kernel/itlb_base.S b/arch/sparc64/kernel/itlb_base.S
index b5e32df..4951ff8 100644
--- a/arch/sparc64/kernel/itlb_base.S
+++ b/arch/sparc64/kernel/itlb_base.S
@@ -15,14 +15,12 @@
  */
 #define CREATE_VPTE_OFFSET1(r1, r2) \
 				srax	r1, 10, r2
-#define CREATE_VPTE_OFFSET2(r1, r2)
-#define CREATE_VPTE_NOP		nop
+#define CREATE_VPTE_OFFSET2(r1, r2) nop
 #else /* PAGE_SHIFT */
 #define CREATE_VPTE_OFFSET1(r1, r2) \
 				srax	r1, PAGE_SHIFT, r2
 #define CREATE_VPTE_OFFSET2(r1, r2) \
 				sllx	r2, 3, r2
-#define CREATE_VPTE_NOP
 #endif /* PAGE_SHIFT */
 
 
@@ -36,6 +34,7 @@
  */
 
 /* ITLB ** ICACHE line 1: Quick user TLB misses		*/
+	mov		TLB_SFSR, %g1
 	ldxa		[%g1 + %g1] ASI_IMMU, %g4	! Get TAG_ACCESS
 	CREATE_VPTE_OFFSET1(%g4, %g6)			! Create VPTE offset
 	CREATE_VPTE_OFFSET2(%g4, %g6)			! Create VPTE offset
@@ -43,41 +42,38 @@
 1:	brgez,pn	%g5, 3f				! Not valid, branch out
 	 sethi		%hi(_PAGE_EXEC), %g4		! Delay-slot
 	andcc		%g5, %g4, %g0			! Executable?
+
+/* ITLB ** ICACHE line 2: Real faults			*/
 	be,pn		%xcc, 3f			! Nope, branch.
 	 nop						! Delay-slot
 2:	stxa		%g5, [%g0] ASI_ITLB_DATA_IN	! Load PTE into TLB
 	retry						! Trap return
-3:	rdpr		%pstate, %g4			! Move into alternate globals
-
-/* ITLB ** ICACHE line 2: Real faults			*/
+3:	rdpr		%pstate, %g4			! Move into alt-globals
 	wrpr		%g4, PSTATE_AG|PSTATE_MG, %pstate
 	rdpr		%tpc, %g5			! And load faulting VA
 	mov		FAULT_CODE_ITLB, %g4		! It was read from ITLB
-sparc64_realfault_common:				! Called by TL0 dtlb_miss too
+
+/* ITLB ** ICACHE line 3: Finish faults	*/
+sparc64_realfault_common:				! Called by dtlb_miss
 	stb		%g4, [%g6 + TI_FAULT_CODE]
 	stx		%g5, [%g6 + TI_FAULT_ADDR]
 	ba,pt		%xcc, etrap			! Save state
 1:	 rd		%pc, %g7			! ...
-	nop
-
-/* ITLB ** ICACHE line 3: Finish faults + window fixups	*/
 	call		do_sparc64_fault		! Call fault handler
 	 add		%sp, PTREGS_OFF, %o0! Compute pt_regs arg
 	ba,pt		%xcc, rtrap_clr_l6		! Restore cpu state
 	 nop
+
+/* ITLB ** ICACHE line 4: Window fixups */
 winfix_trampoline:
 	rdpr		%tpc, %g3			! Prepare winfixup TNPC
-	or		%g3, 0x7c, %g3			! Compute offset to branch
+	or		%g3, 0x7c, %g3			! Compute branch offset
 	wrpr		%g3, %tnpc			! Write it into TNPC
 	done						! Do it to it
-
-/* ITLB ** ICACHE line 4: Unused...	*/
 	nop
 	nop
 	nop
 	nop
-	CREATE_VPTE_NOP
 
 #undef CREATE_VPTE_OFFSET1
 #undef CREATE_VPTE_OFFSET2
-#undef CREATE_VPTE_NOP
diff --git a/arch/sparc64/kernel/ktlb.S b/arch/sparc64/kernel/ktlb.S
index 7796b37..d9244d3 100644
--- a/arch/sparc64/kernel/ktlb.S
+++ b/arch/sparc64/kernel/ktlb.S
@@ -58,9 +58,6 @@
 	done
 
 vpte_insn_obp:
-	sethi		%hi(prom_pmd_phys), %g5
-	ldx		[%g5 + %lo(prom_pmd_phys)], %g5
-
 	/* Behave as if we are at TL0.  */
 	wrpr		%g0, 1, %tl
 	rdpr		%tpc, %g4	/* Find original faulting iaddr */
@@ -71,58 +68,57 @@
 	mov		TLB_SFSR, %g1
 	stxa		%g4, [%g1 + %g1] ASI_IMMU
 
-	/* Get PMD offset.  */
-	srlx		%g4, 23, %g6
-	and		%g6, 0x7ff, %g6
-	sllx		%g6, 2, %g6
+	sethi		%hi(prom_trans), %g5
+	or		%g5, %lo(prom_trans), %g5
 
-	/* Load PMD, is it valid?  */
-	lduwa		[%g5 + %g6] ASI_PHYS_USE_EC, %g5
-	brz,pn		%g5, longpath
-	 sllx		%g5, 11, %g5
+1:	ldx		[%g5 + 0x00], %g6	! base
+	brz,a,pn	%g6, longpath		! no more entries, fail
+	 mov		TLB_SFSR, %g1		! and restore %g1
+	ldx		[%g5 + 0x08], %g1	! len
+	add		%g6, %g1, %g1		! end
+	cmp		%g6, %g4
+	bgu,pt		%xcc, 2f
+	 cmp		%g4, %g1
+	bgeu,pt		%xcc, 2f
+	 ldx		[%g5 + 0x10], %g1	! PTE
 
-	/* Get PTE offset.  */
-	srlx		%g4, 13, %g6
-	and		%g6, 0x3ff, %g6
-	sllx		%g6, 3, %g6
-
-	/* Load PTE.  */
-	ldxa		[%g5 + %g6] ASI_PHYS_USE_EC, %g5
-	brgez,pn	%g5, longpath
-	 nop
-
-	/* TLB load and return from trap.  */
+	/* TLB load, restore %g1, and return from trap.  */
+	sub		%g4, %g6, %g6
+	add		%g1, %g6, %g5
+	mov		TLB_SFSR, %g1
 	stxa		%g5, [%g0] ASI_ITLB_DATA_IN
 	retry
 
+2:	ba,pt		%xcc, 1b
+	 add		%g5, (3 * 8), %g5	! next entry
+
 kvmap_do_obp:
-	sethi		%hi(prom_pmd_phys), %g5
-	ldx		[%g5 + %lo(prom_pmd_phys)], %g5
+	sethi		%hi(prom_trans), %g5
+	or		%g5, %lo(prom_trans), %g5
+	srlx		%g4, 13, %g4
+	sllx		%g4, 13, %g4
 
-	/* Get PMD offset.  */
-	srlx		%g4, 23, %g6
-	and		%g6, 0x7ff, %g6
-	sllx		%g6, 2, %g6
+1:	ldx		[%g5 + 0x00], %g6	! base
+	brz,a,pn	%g6, longpath		! no more entries, fail
+	 mov		TLB_SFSR, %g1		! and restore %g1
+	ldx		[%g5 + 0x08], %g1	! len
+	add		%g6, %g1, %g1		! end
+	cmp		%g6, %g4
+	bgu,pt		%xcc, 2f
+	 cmp		%g4, %g1
+	bgeu,pt		%xcc, 2f
+	 ldx		[%g5 + 0x10], %g1	! PTE
 
-	/* Load PMD, is it valid?  */
-	lduwa		[%g5 + %g6] ASI_PHYS_USE_EC, %g5
-	brz,pn		%g5, longpath
-	 sllx		%g5, 11, %g5
-
-	/* Get PTE offset.  */
-	srlx		%g4, 13, %g6
-	and		%g6, 0x3ff, %g6
-	sllx		%g6, 3, %g6
-
-	/* Load PTE.  */
-	ldxa		[%g5 + %g6] ASI_PHYS_USE_EC, %g5
-	brgez,pn	%g5, longpath
-	 nop
-
-	/* TLB load and return from trap.  */
+	/* TLB load, restore %g1, and return from trap.  */
+	sub		%g4, %g6, %g6
+	add		%g1, %g6, %g5
+	mov		TLB_SFSR, %g1
 	stxa		%g5, [%g0] ASI_DTLB_DATA_IN
 	retry
 
+2:	ba,pt		%xcc, 1b
+	 add		%g5, (3 * 8), %g5	! next entry
+
 /*
  * On a first level data miss, check whether this is to the OBP range (note
  * that such accesses can be made by prom, as well as by kernel using
diff --git a/arch/sparc64/mm/init.c b/arch/sparc64/mm/init.c
index 0d2e967..1e44ee2 100644
--- a/arch/sparc64/mm/init.c
+++ b/arch/sparc64/mm/init.c
@@ -105,7 +105,7 @@
 		regs[i].phys_addr = base;
 		regs[i].reg_size = size;
 	}
-	sort(regs, ents, sizeof(struct  linux_prom64_registers),
+	sort(regs, ents, sizeof(struct linux_prom64_registers),
 	     cmp_p64, NULL);
 }
 
@@ -367,8 +367,11 @@
 	unsigned long size;
 	unsigned long data;
 };
-static struct linux_prom_translation prom_trans[512] __initdata;
-static unsigned int prom_trans_ents __initdata;
+
+/* Exported for kernel TLB miss handling in ktlb.S */
+struct linux_prom_translation prom_trans[512] __read_mostly;
+unsigned int prom_trans_ents __read_mostly;
+unsigned int swapper_pgd_zero __read_mostly;
 
 extern unsigned long prom_boot_page;
 extern void prom_remap(unsigned long physpage, unsigned long virtpage, int mmu_ihandle);
@@ -378,122 +381,57 @@
 /* Exported for SMP bootup purposes. */
 unsigned long kern_locked_tte_data;
 
-/* Exported for kernel TLB miss handling in ktlb.S */
-unsigned long prom_pmd_phys __read_mostly;
-unsigned int swapper_pgd_zero __read_mostly;
-
-static pmd_t *prompmd __read_mostly;
-
-#define BASE_PAGE_SIZE 8192
-
 /*
  * Translate PROM's mapping we capture at boot time into physical address.
  * The second parameter is only set from prom_callback() invocations.
  */
 unsigned long prom_virt_to_phys(unsigned long promva, int *error)
 {
-	pmd_t *pmdp = prompmd + ((promva >> 23) & 0x7ff);
-	pte_t *ptep;
-	unsigned long base;
+	int i;
 
-	if (pmd_none(*pmdp)) {
-		if (error)
-			*error = 1;
-		return 0;
-	}
-	ptep = (pte_t *)__pmd_page(*pmdp) + ((promva >> 13) & 0x3ff);
-	if (!pte_present(*ptep)) {
-		if (error)
-			*error = 1;
-		return 0;
-	}
-	if (error) {
-		*error = 0;
-		return pte_val(*ptep);
-	}
-	base = pte_val(*ptep) & _PAGE_PADDR;
+	for (i = 0; i < prom_trans_ents; i++) {
+		struct linux_prom_translation *p = &prom_trans[i];
 
-	return base + (promva & (BASE_PAGE_SIZE - 1));
+		if (promva >= p->virt &&
+		    promva < (p->virt + p->size)) {
+			unsigned long base = p->data & _PAGE_PADDR;
+
+			if (error)
+				*error = 0;
+			return base + (promva & (8192 - 1));
+		}
+	}
+	if (error)
+		*error = 1;
+	return 0UL;
 }
 
 /* The obp translations are saved based on 8k pagesize, since obp can
  * use a mixture of pagesizes. Misses to the LOW_OBP_ADDRESS ->
- * HI_OBP_ADDRESS range are handled in entry.S and do not use the vpte
+ * HI_OBP_ADDRESS range are handled in ktlb.S and do not use the vpte
  * scheme (also, see rant in inherit_locked_prom_mappings()).
  */
-static void __init build_obp_range(unsigned long start, unsigned long end, unsigned long data)
-{
-	unsigned long vaddr;
-
-	for (vaddr = start; vaddr < end; vaddr += BASE_PAGE_SIZE) {
-		unsigned long val;
-		pmd_t *pmd;
-		pte_t *pte;
-
-		pmd = prompmd + ((vaddr >> 23) & 0x7ff);
-		if (pmd_none(*pmd)) {
-			pte = __alloc_bootmem(BASE_PAGE_SIZE, BASE_PAGE_SIZE,
-					      PAGE_SIZE);
-			if (!pte)
-				prom_halt();
-			memset(pte, 0, BASE_PAGE_SIZE);
-			pmd_set(pmd, pte);
-		}
-		pte = (pte_t *) __pmd_page(*pmd) + ((vaddr >> 13) & 0x3ff);
-
-		val = data;
-
-		/* Clear diag TTE bits. */
-		if (tlb_type == spitfire)
-			val &= ~0x0003fe0000000000UL;
-
-		set_pte_at(&init_mm, vaddr, pte,
-			   __pte(val | _PAGE_MODIFIED));
-
-		data += BASE_PAGE_SIZE;
-	}
-}
-
 static inline int in_obp_range(unsigned long vaddr)
 {
 	return (vaddr >= LOW_OBP_ADDRESS &&
 		vaddr < HI_OBP_ADDRESS);
 }
 
-#define OBP_PMD_SIZE 2048
-static void __init build_obp_pgtable(void)
+static int cmp_ptrans(const void *a, const void *b)
 {
-	unsigned long i;
+	const struct linux_prom_translation *x = a, *y = b;
 
-	prompmd = __alloc_bootmem(OBP_PMD_SIZE, OBP_PMD_SIZE, PAGE_SIZE);
-	if (!prompmd)
-		prom_halt();
-
-	memset(prompmd, 0, OBP_PMD_SIZE);
-
-	prom_pmd_phys = __pa(prompmd);
-
-	for (i = 0; i < prom_trans_ents; i++) {
-		unsigned long start, end;
-
-		if (!in_obp_range(prom_trans[i].virt))
-			continue;
-
-		start = prom_trans[i].virt;
-		end = start + prom_trans[i].size;
-		if (end > HI_OBP_ADDRESS)
-			end = HI_OBP_ADDRESS;
-
-		build_obp_range(start, end, prom_trans[i].data);
-	}
+	if (x->virt > y->virt)
+		return 1;
+	if (x->virt < y->virt)
+		return -1;
+	return 0;
 }
 
-/* Read OBP translations property into 'prom_trans[]'.
- * Return the number of entries.
- */
+/* Read OBP translations property into 'prom_trans[]'.  */
 static void __init read_obp_translations(void)
 {
-	int n, node;
+	int n, node, ents, first, last, i;
 
 	node = prom_finddevice("/virtual-memory");
 	n = prom_getproplen(node, "translations");
@@ -515,7 +453,41 @@
 
 	n = n / sizeof(struct linux_prom_translation);
 
-	prom_trans_ents = n;
+	ents = n;
+
+	sort(prom_trans, ents, sizeof(struct linux_prom_translation),
+	     cmp_ptrans, NULL);
+
+	/* Now kick out all the non-OBP entries.  */
+	for (i = 0; i < ents; i++) {
+		if (in_obp_range(prom_trans[i].virt))
+			break;
+	}
+	first = i;
+	for (; i < ents; i++) {
+		if (!in_obp_range(prom_trans[i].virt))
+			break;
+	}
+	last = i;
+
+	for (i = 0; i < (last - first); i++) {
+		struct linux_prom_translation *src = &prom_trans[i + first];
+		struct linux_prom_translation *dest = &prom_trans[i];
+
+		*dest = *src;
+	}
+	for (; i < ents; i++) {
+		struct linux_prom_translation *dest = &prom_trans[i];
+		dest->virt = dest->size = dest->data = 0x0UL;
+	}
+
+	prom_trans_ents = last - first;
+
+	if (tlb_type == spitfire) {
+		/* Clear diag TTE bits. */
+		for (i = 0; i < prom_trans_ents; i++)
+			prom_trans[i].data &= ~0x0003fe0000000000UL;
+	}
 }
 
 static void __init remap_kernel(void)
@@ -553,21 +525,18 @@
 }
 
 
-static void __init inherit_prom_mappings_pre(void)
+static void __init inherit_prom_mappings(void)
 {
 	read_obp_translations();
 
 	/* Now fixup OBP's idea about where we really are mapped. */
 	prom_printf("Remapping the kernel... ");
 	remap_kernel();
-
 	prom_printf("done.\n");
-}
 
-static void __init inherit_prom_mappings_post(void)
-{
-	build_obp_pgtable();
+	prom_printf("Registering callbacks... ");
 	register_prom_callbacks();
+	prom_printf("done.\n");
 }
 
 /* The OBP specifications for sun4u mark 0xfffffffc00000000 and
@@ -1519,7 +1488,7 @@
 	
 	swapper_pgd_zero = pgd_val(swapper_pg_dir[0]);
 	
-	inherit_prom_mappings_pre();
+	inherit_prom_mappings();
 	
 	/* Ok, we can use our TLB miss and window trap handlers safely.
 	 * We need to do a quick peek here to see if we are on StarFire
@@ -1530,23 +1499,15 @@
 		extern void setup_tba(int);
 		setup_tba(this_is_starfire);
 	}
-	__flush_tlb_all();
 
-	/* Everything from this point forward, until we are done with
-	 * inherit_prom_mappings_post(), must complete successfully
-	 * without calling into the firmware.  The firwmare page tables
-	 * have not been built, but we are running on the Linux kernel's
-	 * trap table.
-	 */
+	inherit_locked_prom_mappings(1);
+
+	__flush_tlb_all();
 
 	/* Setup bootmem... */
 	pages_avail = 0;
 	last_valid_pfn = end_pfn = bootmem_init(&pages_avail);
 
-	inherit_prom_mappings_post();
-
-	inherit_locked_prom_mappings(1);
-
 #ifdef CONFIG_DEBUG_PAGEALLOC
 	kernel_physical_mapping_init();
 #endif
diff --git a/arch/um/drivers/Makefile b/arch/um/drivers/Makefile
index 783e18c..de17d4c 100644
--- a/arch/um/drivers/Makefile
+++ b/arch/um/drivers/Makefile
@@ -13,7 +13,7 @@
 net-objs := net_kern.o net_user.o
 mconsole-objs := mconsole_kern.o mconsole_user.o
 hostaudio-objs := hostaudio_kern.o
-ubd-objs := ubd_kern.o
+ubd-objs := ubd_kern.o ubd_user.o
 port-objs := port_kern.o port_user.o
 harddog-objs := harddog_kern.o harddog_user.o
 
diff --git a/arch/um/drivers/ubd_kern.c b/arch/um/drivers/ubd_kern.c
index e77a38d..f731343 100644
--- a/arch/um/drivers/ubd_kern.c
+++ b/arch/um/drivers/ubd_kern.c
@@ -35,7 +35,6 @@
 #include "linux/blkpg.h"
 #include "linux/genhd.h"
 #include "linux/spinlock.h"
-#include "asm/atomic.h"
 #include "asm/segment.h"
 #include "asm/uaccess.h"
 #include "asm/irq.h"
@@ -54,21 +53,20 @@
 #include "mem.h"
 #include "mem_kern.h"
 #include "cow.h"
-#include "aio.h"
 
 enum ubd_req { UBD_READ, UBD_WRITE };
 
 struct io_thread_req {
-	enum aio_type op;
+	enum ubd_req op;
 	int fds[2];
 	unsigned long offsets[2];
 	unsigned long long offset;
 	unsigned long length;
 	char *buffer;
 	int sectorsize;
-	int bitmap_offset;
-	long bitmap_start;
-	long bitmap_end;
+	unsigned long sector_mask;
+	unsigned long long cow_offset;
+	unsigned long bitmap_words[2];
 	int error;
 };
 
@@ -82,31 +80,28 @@
 			   unsigned long *bitmap_len_out,
 			   int *data_offset_out);
 extern int read_cow_bitmap(int fd, void *buf, int offset, int len);
-extern void do_io(struct io_thread_req *req, struct request *r,
-		  unsigned long *bitmap);
+extern void do_io(struct io_thread_req *req);
 
-static inline int ubd_test_bit(__u64 bit, void *data)
+static inline int ubd_test_bit(__u64 bit, unsigned char *data)
 {
-	unsigned char *buffer = data;
 	__u64 n;
 	int bits, off;
 
-	bits = sizeof(buffer[0]) * 8;
+	bits = sizeof(data[0]) * 8;
 	n = bit / bits;
 	off = bit % bits;
-	return((buffer[n] & (1 << off)) != 0);
+	return((data[n] & (1 << off)) != 0);
 }
 
-static inline void ubd_set_bit(__u64 bit, void *data)
+static inline void ubd_set_bit(__u64 bit, unsigned char *data)
 {
-	unsigned char *buffer = data;
 	__u64 n;
 	int bits, off;
 
-	bits = sizeof(buffer[0]) * 8;
+	bits = sizeof(data[0]) * 8;
 	n = bit / bits;
 	off = bit % bits;
-	buffer[n] |= (1 << off);
+	data[n] |= (1 << off);
 }
 /*End stuff from ubd_user.h*/
 
@@ -115,6 +110,8 @@
 static DEFINE_SPINLOCK(ubd_io_lock);
 static DEFINE_SPINLOCK(ubd_lock);
 
+static void (*do_ubd)(void);
+
 static int ubd_open(struct inode * inode, struct file * filp);
 static int ubd_release(struct inode * inode, struct file * file);
 static int ubd_ioctl(struct inode * inode, struct file * file,
@@ -161,8 +158,6 @@
         int data_offset;
 };
 
-#define MAX_SG 64
-
 struct ubd {
 	char *file;
 	int count;
@@ -173,7 +168,6 @@
 	int no_cow;
 	struct cow cow;
 	struct platform_device pdev;
-        struct scatterlist sg[MAX_SG];
 };
 
 #define DEFAULT_COW { \
@@ -466,114 +460,81 @@
 );
 
 static void do_ubd_request(request_queue_t * q);
-static int in_ubd;
+
+/* Only changed by ubd_init, which is an initcall. */
+int thread_fd = -1;
 
 /* Changed by ubd_handler, which is serialized because interrupts only
  * happen on CPU 0.
  */
 int intr_count = 0;
 
-static void ubd_end_request(struct request *req, int bytes, int uptodate)
-{
-	if (!end_that_request_first(req, uptodate, bytes >> 9)) {
-		add_disk_randomness(req->rq_disk);
-		end_that_request_last(req);
-	}
-}
-
 /* call ubd_finish if you need to serialize */
-static void __ubd_finish(struct request *req, int bytes)
+static void __ubd_finish(struct request *req, int error)
 {
-	if(bytes < 0){
-		ubd_end_request(req, 0, 0);
-  		return;
-  	}
+	int nsect;
 
-	ubd_end_request(req, bytes, 1);
+	if(error){
+		end_request(req, 0);
+		return;
+	}
+	nsect = req->current_nr_sectors;
+	req->sector += nsect;
+	req->buffer += nsect << 9;
+	req->errors = 0;
+	req->nr_sectors -= nsect;
+	req->current_nr_sectors = 0;
+	end_request(req, 1);
 }
 
-static inline void ubd_finish(struct request *req, int bytes)
+static inline void ubd_finish(struct request *req, int error)
 {
-   	spin_lock(&ubd_io_lock);
-	__ubd_finish(req, bytes);
-  	spin_unlock(&ubd_io_lock);
+ 	spin_lock(&ubd_io_lock);
+	__ubd_finish(req, error);
+	spin_unlock(&ubd_io_lock);
 }
 
-struct bitmap_io {
-        atomic_t count;
-        struct aio_context aio;
-};
+/* Called without ubd_io_lock held */
+static void ubd_handler(void)
+{
+	struct io_thread_req req;
+	struct request *rq = elv_next_request(ubd_queue);
+	int n;
 
-struct ubd_aio {
-        struct aio_context aio;
-        struct request *req;
-        int len;
-        struct bitmap_io *bitmap;
-        void *bitmap_buf;
-};
-
-static int ubd_reply_fd = -1;
+	do_ubd = NULL;
+	intr_count++;
+	n = os_read_file(thread_fd, &req, sizeof(req));
+	if(n != sizeof(req)){
+		printk(KERN_ERR "Pid %d - spurious interrupt in ubd_handler, "
+		       "err = %d\n", os_getpid(), -n);
+		spin_lock(&ubd_io_lock);
+		end_request(rq, 0);
+		spin_unlock(&ubd_io_lock);
+		return;
+	}
+        
+	ubd_finish(rq, req.error);
+	reactivate_fd(thread_fd, UBD_IRQ);	
+	do_ubd_request(ubd_queue);
+}
 
 static irqreturn_t ubd_intr(int irq, void *dev, struct pt_regs *unused)
 {
-	struct aio_thread_reply reply;
-	struct ubd_aio *aio;
-	struct request *req;
-	int err, n, fd = (int) (long) dev;
-
-	while(1){
-		err = os_read_file(fd, &reply, sizeof(reply));
-		if(err == -EAGAIN)
-			break;
-		if(err < 0){
-			printk("ubd_aio_handler - read returned err %d\n",
-			       -err);
-			break;
-		}
-
-                aio = container_of(reply.data, struct ubd_aio, aio);
-                n = reply.err;
-
-		if(n == 0){
-			req = aio->req;
-			req->nr_sectors -= aio->len >> 9;
-
-			if((aio->bitmap != NULL) &&
-			   (atomic_dec_and_test(&aio->bitmap->count))){
-                                aio->aio = aio->bitmap->aio;
-                                aio->len = 0;
-                                kfree(aio->bitmap);
-                                aio->bitmap = NULL;
-                                submit_aio(&aio->aio);
-			}
-			else {
-				if((req->nr_sectors == 0) &&
-                                   (aio->bitmap == NULL)){
-					int len = req->hard_nr_sectors << 9;
-					ubd_finish(req, len);
-				}
-
-                                if(aio->bitmap_buf != NULL)
-                                        kfree(aio->bitmap_buf);
-				kfree(aio);
-			}
-		}
-                else if(n < 0){
-                        ubd_finish(aio->req, n);
-                        if(aio->bitmap != NULL)
-                                kfree(aio->bitmap);
-                        if(aio->bitmap_buf != NULL)
-                                kfree(aio->bitmap_buf);
-                        kfree(aio);
-                }
-	}
-	reactivate_fd(fd, UBD_IRQ);
-
-        do_ubd_request(ubd_queue);
-
+	ubd_handler();
 	return(IRQ_HANDLED);
 }
 
+/* Only changed by ubd_init, which is an initcall. */
+static int io_pid = -1;
+
+void kill_io_thread(void)
+{
+	if(io_pid != -1) 
+		os_kill_process(io_pid, 1);
+}
+
+__uml_exitcall(kill_io_thread);
+
 static int ubd_file_size(struct ubd *dev, __u64 *size_out)
 {
 	char *file;
@@ -608,7 +569,7 @@
 				&dev->cow.data_offset, create_ptr);
 
 	if((dev->fd == -ENOENT) && create_cow){
-		dev->fd = create_cow_file(dev->file, dev->cow.file,
+		dev->fd = create_cow_file(dev->file, dev->cow.file, 
 					  dev->openflags, 1 << 9, PAGE_SIZE,
 					  &dev->cow.bitmap_offset, 
 					  &dev->cow.bitmap_len,
@@ -870,10 +831,6 @@
 {
         int i;
 
-	ubd_reply_fd = init_aio_irq(UBD_IRQ, "ubd", ubd_intr);
-	if(ubd_reply_fd < 0)
-		printk("Setting up ubd AIO failed, err = %d\n", ubd_reply_fd);
-
 	devfs_mk_dir("ubd");
 	if (register_blkdev(MAJOR_NR, "ubd"))
 		return -1;
@@ -884,7 +841,6 @@
 		return -1;
 	}
 		
-	blk_queue_max_hw_segments(ubd_queue, MAX_SG);
 	if (fake_major != MAJOR_NR) {
 		char name[sizeof("ubd_nnn\0")];
 
@@ -896,12 +852,40 @@
 	driver_register(&ubd_driver);
 	for (i = 0; i < MAX_DEV; i++) 
 		ubd_add(i);
-
 	return 0;
 }
 
 late_initcall(ubd_init);
 
+int ubd_driver_init(void){
+	unsigned long stack;
+	int err;
+
+	/* Set by CONFIG_BLK_DEV_UBD_SYNC or ubd=sync.*/
+	if(global_openflags.s){
+		printk(KERN_INFO "ubd: Synchronous mode\n");
+		/* Letting ubd=sync be like using ubd#s= instead of ubd#= is
+		 * enough. So use anyway the io thread. */
+	}
+	stack = alloc_stack(0, 0);
+	io_pid = start_io_thread(stack + PAGE_SIZE - sizeof(void *), 
+				 &thread_fd);
+	if(io_pid < 0){
+		printk(KERN_ERR 
+		       "ubd : Failed to start I/O thread (errno = %d) - "
+		       "falling back to synchronous I/O\n", -io_pid);
+		io_pid = -1;
+		return(0);
+	}
+	err = um_request_irq(UBD_IRQ, thread_fd, IRQ_READ, ubd_intr, 
+			     SA_INTERRUPT, "ubd", ubd_dev);
+	if(err != 0)
+		printk(KERN_ERR "um_request_irq failed - errno = %d\n", -err);
+	return(err);
+}
+
+device_initcall(ubd_driver_init);
+
 static int ubd_open(struct inode *inode, struct file *filp)
 {
 	struct gendisk *disk = inode->i_bdev->bd_disk;
@@ -939,55 +923,105 @@
 	return(0);
 }
 
-static void cowify_bitmap(struct io_thread_req *req, unsigned long *bitmap)
+static void cowify_bitmap(__u64 io_offset, int length, unsigned long *cow_mask,
+			  __u64 *cow_offset, unsigned long *bitmap,
+			  __u64 bitmap_offset, unsigned long *bitmap_words,
+			  __u64 bitmap_len)
 {
-        __u64 sector = req->offset / req->sectorsize;
-        int i;
+	__u64 sector = io_offset >> 9;
+	int i, update_bitmap = 0;
 
-        for(i = 0; i < req->length / req->sectorsize; i++){
-                if(ubd_test_bit(sector + i, bitmap))
-                        continue;
+	for(i = 0; i < length >> 9; i++){
+		if(cow_mask != NULL)
+			ubd_set_bit(i, (unsigned char *) cow_mask);
+		if(ubd_test_bit(sector + i, (unsigned char *) bitmap))
+			continue;
 
-                if(req->bitmap_start == -1)
-                        req->bitmap_start = sector + i;
-                req->bitmap_end = sector + i + 1;
+		update_bitmap = 1;
+		ubd_set_bit(sector + i, (unsigned char *) bitmap);
+	}
 
-                ubd_set_bit(sector + i, bitmap);
-        }
+	if(!update_bitmap)
+		return;
+
+	*cow_offset = sector / (sizeof(unsigned long) * 8);
+
+	/* This takes care of the case where we're exactly at the end of the
+	 * device, and *cow_offset + 1 is off the end.  So, just back it up
+	 * by one word.  Thanks to Lynn Kerby for the fix and James McMechan
+	 * for the original diagnosis.
+	 */
+	if(*cow_offset == ((bitmap_len + sizeof(unsigned long) - 1) /
+			   sizeof(unsigned long) - 1))
+		(*cow_offset)--;
+
+	bitmap_words[0] = bitmap[*cow_offset];
+	bitmap_words[1] = bitmap[*cow_offset + 1];
+
+	*cow_offset *= sizeof(unsigned long);
+	*cow_offset += bitmap_offset;
+}
+
+static void cowify_req(struct io_thread_req *req, unsigned long *bitmap,
+		       __u64 bitmap_offset, __u64 bitmap_len)
+{
+	__u64 sector = req->offset >> 9;
+	int i;
+
+	if(req->length > (sizeof(req->sector_mask) * 8) << 9)
+		panic("Operation too long");
+
+	if(req->op == UBD_READ) {
+		for(i = 0; i < req->length >> 9; i++){
+			if(ubd_test_bit(sector + i, (unsigned char *) bitmap))
+				ubd_set_bit(i, (unsigned char *) 
+					    &req->sector_mask);
+                }
+	}
+	else cowify_bitmap(req->offset, req->length, &req->sector_mask,
+			   &req->cow_offset, bitmap, bitmap_offset,
+			   req->bitmap_words, bitmap_len);
 }
 
 /* Called with ubd_io_lock held */
-static int prepare_request(struct request *req, struct io_thread_req *io_req,
-                           unsigned long long offset, int page_offset,
-                           int len, struct page *page)
+static int prepare_request(struct request *req, struct io_thread_req *io_req)
 {
 	struct gendisk *disk = req->rq_disk;
 	struct ubd *dev = disk->private_data;
+	__u64 offset;
+	int len;
+
+	if(req->rq_status == RQ_INACTIVE) return(1);
 
 	/* This should be impossible now */
 	if((rq_data_dir(req) == WRITE) && !dev->openflags.w){
 		printk("Write attempted on readonly ubd device %s\n", 
 		       disk->disk_name);
-                ubd_end_request(req, 0, 0);
+		end_request(req, 0);
 		return(1);
 	}
 
+	offset = ((__u64) req->sector) << 9;
+	len = req->current_nr_sectors << 9;
+
 	io_req->fds[0] = (dev->cow.file != NULL) ? dev->cow.fd : dev->fd;
 	io_req->fds[1] = dev->fd;
+	io_req->cow_offset = -1;
 	io_req->offset = offset;
 	io_req->length = len;
 	io_req->error = 0;
-	io_req->op = (rq_data_dir(req) == READ) ? AIO_READ : AIO_WRITE;
+	io_req->sector_mask = 0;
+
+	io_req->op = (rq_data_dir(req) == READ) ? UBD_READ : UBD_WRITE;
 	io_req->offsets[0] = 0;
 	io_req->offsets[1] = dev->cow.data_offset;
-        io_req->buffer = page_address(page) + page_offset;
+	io_req->buffer = req->buffer;
 	io_req->sectorsize = 1 << 9;
-        io_req->bitmap_offset = dev->cow.bitmap_offset;
-        io_req->bitmap_start = -1;
-        io_req->bitmap_end = -1;
 
-        if((dev->cow.file != NULL) && (io_req->op == UBD_WRITE))
-                cowify_bitmap(io_req, dev->cow.bitmap);
+	if(dev->cow.file != NULL)
+		cowify_req(io_req, dev->cow.bitmap, dev->cow.bitmap_offset,
+			   dev->cow.bitmap_len);
+
 	return(0);
 }
 
@@ -996,36 +1030,30 @@
 {
 	struct io_thread_req io_req;
 	struct request *req;
-	__u64 sector;
-	int err;
+	int err, n;
 
-	if(in_ubd)
-		return;
-	in_ubd = 1;
-	while((req = elv_next_request(q)) != NULL){
-		struct gendisk *disk = req->rq_disk;
-		struct ubd *dev = disk->private_data;
-		int n, i;
-
-		blkdev_dequeue_request(req);
-
-		sector = req->sector;
-		n = blk_rq_map_sg(q, req, dev->sg);
-
-		for(i = 0; i < n; i++){
-			struct scatterlist *sg = &dev->sg[i];
-
-			err = prepare_request(req, &io_req, sector << 9,
-					      sg->offset, sg->length,
-					      sg->page);
-			if(err)
-				continue;
-
-			sector += sg->length >> 9;
-			do_io(&io_req, req, dev->cow.bitmap);
+	if(thread_fd == -1){
+		while((req = elv_next_request(q)) != NULL){
+			err = prepare_request(req, &io_req);
+			if(!err){
+				do_io(&io_req);
+				__ubd_finish(req, io_req.error);
+			}
 		}
 	}
-	in_ubd = 0;
+	else {
+		if(do_ubd || (req = elv_next_request(q)) == NULL)
+			return;
+		err = prepare_request(req, &io_req);
+		if(!err){
+			do_ubd = ubd_handler;
+			n = os_write_file(thread_fd, (char *) &io_req,
+					 sizeof(io_req));
+			if(n != sizeof(io_req))
+				printk("write to io thread failed, "
+				       "errno = %d\n", -n);
+		}
+	}
 }
 
 static int ubd_ioctl(struct inode * inode, struct file * file,
@@ -1241,95 +1269,131 @@
 	return(err);
 }
 
-void do_io(struct io_thread_req *req, struct request *r, unsigned long *bitmap)
+static int update_bitmap(struct io_thread_req *req)
 {
-        struct ubd_aio *aio;
-        struct bitmap_io *bitmap_io = NULL;
-        char *buf;
-        void *bitmap_buf = NULL;
-        unsigned long len, sector;
-        int nsectors, start, end, bit, err;
-        __u64 off;
+	int n;
 
-        if(req->bitmap_start != -1){
-                /* Round up to the nearest word */
-                int round = sizeof(unsigned long);
-                len = (req->bitmap_end - req->bitmap_start +
-                       round * 8 - 1) / (round * 8);
-                len *= round;
+	if(req->cow_offset == -1)
+		return(0);
 
-                off = req->bitmap_start / (8 * round);
-                off *= round;
+	n = os_seek_file(req->fds[1], req->cow_offset);
+	if(n < 0){
+		printk("do_io - bitmap lseek failed : err = %d\n", -n);
+		return(1);
+	}
 
-                bitmap_io = kmalloc(sizeof(*bitmap_io), GFP_KERNEL);
-                if(bitmap_io == NULL){
-                        printk("Failed to kmalloc bitmap IO\n");
-                        req->error = 1;
-                        return;
-                }
+	n = os_write_file(req->fds[1], &req->bitmap_words,
+		          sizeof(req->bitmap_words));
+	if(n != sizeof(req->bitmap_words)){
+		printk("do_io - bitmap update failed, err = %d fd = %d\n", -n,
+		       req->fds[1]);
+		return(1);
+	}
 
-                bitmap_buf = kmalloc(len, GFP_KERNEL);
-                if(bitmap_buf == NULL){
-                        printk("do_io : kmalloc of bitmap chunk "
-                               "failed\n");
-                        kfree(bitmap_io);
-                        req->error = 1;
-                        return;
-                }
-                memcpy(bitmap_buf, &bitmap[off / sizeof(bitmap[0])], len);
-
-                *bitmap_io = ((struct bitmap_io)
-                        { .count	= ATOMIC_INIT(0),
-                          .aio		= INIT_AIO(AIO_WRITE, req->fds[1],
-                                                   bitmap_buf, len,
-                                                   req->bitmap_offset + off,
-                                                   ubd_reply_fd) } );
-        }
-
-        nsectors = req->length / req->sectorsize;
-        start = 0;
-        end = nsectors;
-        bit = 0;
-        do {
-                if(bitmap != NULL){
-                        sector = req->offset / req->sectorsize;
-                        bit = ubd_test_bit(sector + start, bitmap);
-                        end = start;
-                        while((end < nsectors) &&
-                              (ubd_test_bit(sector + end, bitmap) == bit))
-                                end++;
-                }
-
-                off = req->offsets[bit] + req->offset +
-                        start * req->sectorsize;
-                len = (end - start) * req->sectorsize;
-                buf = &req->buffer[start * req->sectorsize];
-
-                aio = kmalloc(sizeof(*aio), GFP_KERNEL);
-                if(aio == NULL){
-                        req->error = 1;
-                        return;
-                }
-
-                *aio = ((struct ubd_aio)
-                        { .aio		= INIT_AIO(req->op, req->fds[bit], buf,
-                                                   len, off, ubd_reply_fd),
-                          .len		= len,
-                          .req		= r,
-                          .bitmap	= bitmap_io,
-                          .bitmap_buf 	= bitmap_buf });
-
-                if(aio->bitmap != NULL)
-                        atomic_inc(&aio->bitmap->count);
-
-                err = submit_aio(&aio->aio);
-                if(err){
-                        printk("do_io - submit_aio failed, "
-                               "err = %d\n", err);
-                        req->error = 1;
-                        return;
-                }
-
-                start = end;
-        } while(start < nsectors);
+	return(0);
 }
+
+void do_io(struct io_thread_req *req)
+{
+	char *buf;
+	unsigned long len;
+	int n, nsectors, start, end, bit;
+	int err;
+	__u64 off;
+
+	nsectors = req->length / req->sectorsize;
+	start = 0;
+	do {
+		bit = ubd_test_bit(start, (unsigned char *) &req->sector_mask);
+		end = start;
+		while((end < nsectors) &&
+		      (ubd_test_bit(end, (unsigned char *)
+				    &req->sector_mask) == bit))
+			end++;
+
+		off = req->offset + req->offsets[bit] +
+			start * req->sectorsize;
+		len = (end - start) * req->sectorsize;
+		buf = &req->buffer[start * req->sectorsize];
+
+		err = os_seek_file(req->fds[bit], off);
+		if(err < 0){
+			printk("do_io - lseek failed : err = %d\n", -err);
+			req->error = 1;
+			return;
+		}
+		if(req->op == UBD_READ){
+			n = 0;
+			do {
+				buf = &buf[n];
+				len -= n;
+				n = os_read_file(req->fds[bit], buf, len);
+				if (n < 0) {
+					printk("do_io - read failed, err = %d "
+					       "fd = %d\n", -n, req->fds[bit]);
+					req->error = 1;
+					return;
+				}
+			} while((n < len) && (n != 0));
+			if (n < len) memset(&buf[n], 0, len - n);
+		} else {
+			n = os_write_file(req->fds[bit], buf, len);
+			if(n != len){
+				printk("do_io - write failed err = %d "
+				       "fd = %d\n", -n, req->fds[bit]);
+				req->error = 1;
+				return;
+			}
+		}
+
+		start = end;
+	} while(start < nsectors);
+
+	req->error = update_bitmap(req);
+}
+
+/* Changed in start_io_thread, which is serialized by being called only
+ * from ubd_init, which is an initcall.
+ */
+int kernel_fd = -1;
+
+/* Only changed by the io thread */
+int io_count = 0;
+
+int io_thread(void *arg)
+{
+	struct io_thread_req req;
+	int n;
+
+	ignore_sigwinch_sig();
+	while(1){
+		n = os_read_file(kernel_fd, &req, sizeof(req));
+		if(n != sizeof(req)){
+			if(n < 0)
+				printk("io_thread - read failed, fd = %d, "
+				       "err = %d\n", kernel_fd, -n);
+			else {
+				printk("io_thread - short read, fd = %d, "
+				       "length = %d\n", kernel_fd, n);
+			}
+			continue;
+		}
+		io_count++;
+		do_io(&req);
+		n = os_write_file(kernel_fd, &req, sizeof(req));
+		if(n != sizeof(req))
+			printk("io_thread - write failed, fd = %d, err = %d\n",
+			       kernel_fd, -n);
+	}
+}
+
+/*
+ * Overrides for Emacs so that we follow Linus's tabbing style.
+ * Emacs will notice this stuff at the end of the file and automatically
+ * adjust the settings for this buffer only.  This must remain at the end
+ * of the file.
+ * ---------------------------------------------------------------------------
+ * Local variables:
+ * c-file-style: "linux"
+ * End:
+ */
diff --git a/arch/um/drivers/ubd_user.c b/arch/um/drivers/ubd_user.c
new file mode 100644
index 0000000..b94d2bc4f
--- /dev/null
+++ b/arch/um/drivers/ubd_user.c
@@ -0,0 +1,75 @@
+/* 
+ * Copyright (C) 2000, 2001, 2002 Jeff Dike (jdike@karaya.com)
+ * Copyright (C) 2001 Ridgerun,Inc (glonnon@ridgerun.com)
+ * Licensed under the GPL
+ */
+
+#include <stddef.h>
+#include <unistd.h>
+#include <errno.h>
+#include <sched.h>
+#include <signal.h>
+#include <string.h>
+#include <netinet/in.h>
+#include <sys/time.h>
+#include <sys/socket.h>
+#include <sys/mman.h>
+#include <sys/param.h>
+#include "asm/types.h"
+#include "user_util.h"
+#include "kern_util.h"
+#include "user.h"
+#include "ubd_user.h"
+#include "os.h"
+#include "cow.h"
+
+#include <endian.h>
+#include <byteswap.h>
+
+void ignore_sigwinch_sig(void)
+{
+	signal(SIGWINCH, SIG_IGN);
+}
+
+int start_io_thread(unsigned long sp, int *fd_out)
+{
+	int pid, fds[2], err;
+
+	err = os_pipe(fds, 1, 1);
+	if(err < 0){
+		printk("start_io_thread - os_pipe failed, err = %d\n", -err);
+		goto out;
+	}
+
+	kernel_fd = fds[0];
+	*fd_out = fds[1];
+
+	pid = clone(io_thread, (void *) sp, CLONE_FILES | CLONE_VM | SIGCHLD,
+		    NULL);
+	if(pid < 0){
+		printk("start_io_thread - clone failed : errno = %d\n", errno);
+		err = -errno;
+		goto out_close;
+	}
+
+	return(pid);
+
+ out_close:
+	os_close_file(fds[0]);
+	os_close_file(fds[1]);
+	kernel_fd = -1;
+	*fd_out = -1;
+ out:
+	return(err);
+}
+
+/*
+ * Overrides for Emacs so that we follow Linus's tabbing style.
+ * Emacs will notice this stuff at the end of the file and automatically
+ * adjust the settings for this buffer only.  This must remain at the end
+ * of the file.
+ * ---------------------------------------------------------------------------
+ * Local variables:
+ * c-file-style: "linux"
+ * End:
+ */
diff --git a/arch/um/include/aio.h b/arch/um/include/aio.h
index 83f1687..423bae91 100644
--- a/arch/um/include/aio.h
+++ b/arch/um/include/aio.h
@@ -14,27 +14,15 @@
 };
 
 struct aio_context {
-	enum aio_type type;
-	int fd;
-	void *data;
-	int len;
-	unsigned long long offset;
 	int reply_fd;
 	struct aio_context *next;
 };
 
-#define INIT_AIO(aio_type, aio_fd, aio_data, aio_len, aio_offset, \
-		 aio_reply_fd) \
-	{ .type 	= aio_type, \
-	  .fd		= aio_fd, \
-	  .data		= aio_data, \
-	  .len		= aio_len, \
-	  .offset	= aio_offset, \
-	  .reply_fd	= aio_reply_fd }
-
 #define INIT_AIO_CONTEXT { .reply_fd	= -1, \
 			   .next	= NULL }
 
-extern int submit_aio(struct aio_context *aio);
+extern int submit_aio(enum aio_type type, int fd, char *buf, int len,
+		      unsigned long long offset, int reply_fd,
+                      struct aio_context *aio);
 
 #endif
diff --git a/arch/um/include/os.h b/arch/um/include/os.h
index 6f766e1..2e58e30 100644
--- a/arch/um/include/os.h
+++ b/arch/um/include/os.h
@@ -6,6 +6,7 @@
 #ifndef __OS_H__
 #define __OS_H__
 
+#include "uml-config.h"
 #include "asm/types.h"
 #include "../os/include/file.h"
 
@@ -159,7 +160,11 @@
 
 /* Make sure they are clear when running in TT mode. Required by
  * SEGV_MAYBE_FIXABLE */
+#ifdef UML_CONFIG_MODE_SKAS
 #define clear_can_do_skas() do { ptrace_faultinfo = proc_mm = 0; } while (0)
+#else
+#define clear_can_do_skas() do {} while (0)
+#endif
 
 /* mem.c */
 extern int create_mem_file(unsigned long len);
diff --git a/arch/um/os-Linux/aio.c b/arch/um/os-Linux/aio.c
index f6e6402..41cfb09 100644
--- a/arch/um/os-Linux/aio.c
+++ b/arch/um/os-Linux/aio.c
@@ -6,7 +6,6 @@
 #include <stdlib.h>
 #include <unistd.h>
 #include <signal.h>
-#include <string.h>
 #include <errno.h>
 #include <sched.h>
 #include <sys/syscall.h>
@@ -17,31 +16,18 @@
 #include "user.h"
 #include "mode.h"
 
+struct aio_thread_req {
+        enum aio_type type;
+        int io_fd;
+        unsigned long long offset;
+        char *buf;
+        int len;
+        struct aio_context *aio;
+};
+
 static int aio_req_fd_r = -1;
 static int aio_req_fd_w = -1;
 
-static int update_aio(struct aio_context *aio, int res)
-{
-        if(res < 0)
-                aio->len = res;
-        else if((res == 0) && (aio->type == AIO_READ)){
-                /* This is the EOF case - we have hit the end of the file
-                 * and it ends in a partial block, so we fill the end of
-                 * the block with zeros and claim success.
-                 */
-                memset(aio->data, 0, aio->len);
-                aio->len = 0;
-        }
-        else if(res > 0){
-                aio->len -= res;
-                aio->data += res;
-                aio->offset += res;
-                return aio->len;
-        }
-
-        return 0;
-}
-
 #if defined(HAVE_AIO_ABI)
 #include <linux/aio_abi.h>
 
@@ -80,7 +66,8 @@
  * that it now backs the mmapped area.
  */
 
-static int do_aio(aio_context_t ctx, struct aio_context *aio)
+static int do_aio(aio_context_t ctx, enum aio_type type, int fd, char *buf,
+                  int len, unsigned long long offset, struct aio_context *aio)
 {
         struct iocb iocb, *iocbp = &iocb;
         char c;
@@ -88,39 +75,40 @@
 
         iocb = ((struct iocb) { .aio_data 	= (unsigned long) aio,
                                 .aio_reqprio	= 0,
-                                .aio_fildes	= aio->fd,
-                                .aio_buf	= (unsigned long) aio->data,
-                                .aio_nbytes	= aio->len,
-                                .aio_offset	= aio->offset,
+                                .aio_fildes	= fd,
+                                .aio_buf	= (unsigned long) buf,
+                                .aio_nbytes	= len,
+                                .aio_offset	= offset,
                                 .aio_reserved1	= 0,
                                 .aio_reserved2	= 0,
                                 .aio_reserved3	= 0 });
 
-        switch(aio->type){
+        switch(type){
         case AIO_READ:
                 iocb.aio_lio_opcode = IOCB_CMD_PREAD;
+                err = io_submit(ctx, 1, &iocbp);
                 break;
         case AIO_WRITE:
                 iocb.aio_lio_opcode = IOCB_CMD_PWRITE;
+                err = io_submit(ctx, 1, &iocbp);
                 break;
         case AIO_MMAP:
                 iocb.aio_lio_opcode = IOCB_CMD_PREAD;
                 iocb.aio_buf = (unsigned long) &c;
                 iocb.aio_nbytes = sizeof(c);
+                err = io_submit(ctx, 1, &iocbp);
                 break;
         default:
-                printk("Bogus op in do_aio - %d\n", aio->type);
+                printk("Bogus op in do_aio - %d\n", type);
                 err = -EINVAL;
-                goto out;
+                break;
         }
 
-        err = io_submit(ctx, 1, &iocbp);
         if(err > 0)
                 err = 0;
 	else
 		err = -errno;
 
- out:
         return err;
 }
 
@@ -129,9 +117,8 @@
 static int aio_thread(void *arg)
 {
         struct aio_thread_reply reply;
-        struct aio_context *aio;
         struct io_event event;
-        int err, n;
+        int err, n, reply_fd;
 
         signal(SIGWINCH, SIG_IGN);
 
@@ -144,22 +131,14 @@
                                "errno = %d\n", errno);
                 }
                 else {
-			/* This is safe as we've just a pointer here. */
-			aio = (struct aio_context *) (long) event.data;
-			if(update_aio(aio, event.res)){
-				do_aio(ctx, aio);
-				continue;
-			}
-
                         reply = ((struct aio_thread_reply)
-				{ .data = aio,
-				  .err	= aio->len });
-			err = os_write_file(aio->reply_fd, &reply,
-					    sizeof(reply));
+				 { .data = (void *) (long) event.data,
+				   .err	= event.res });
+			reply_fd = ((struct aio_context *) reply.data)->reply_fd;
+			err = os_write_file(reply_fd, &reply, sizeof(reply));
                         if(err != sizeof(reply))
-				printk("aio_thread - write failed, "
-				       "fd = %d, err = %d\n", aio->reply_fd,
-				       -err);
+				printk("aio_thread - write failed, fd = %d, "
+                                       "err = %d\n", aio_req_fd_r, -err);
                 }
         }
         return 0;
@@ -167,35 +146,35 @@
 
 #endif
 
-static int do_not_aio(struct aio_context *aio)
+static int do_not_aio(struct aio_thread_req *req)
 {
         char c;
         int err;
 
-        switch(aio->type){
+        switch(req->type){
         case AIO_READ:
-                err = os_seek_file(aio->fd, aio->offset);
+                err = os_seek_file(req->io_fd, req->offset);
                 if(err)
                         goto out;
 
-                err = os_read_file(aio->fd, aio->data, aio->len);
+                err = os_read_file(req->io_fd, req->buf, req->len);
                 break;
         case AIO_WRITE:
-                err = os_seek_file(aio->fd, aio->offset);
+                err = os_seek_file(req->io_fd, req->offset);
                 if(err)
                         goto out;
 
-                err = os_write_file(aio->fd, aio->data, aio->len);
+                err = os_write_file(req->io_fd, req->buf, req->len);
                 break;
         case AIO_MMAP:
-                err = os_seek_file(aio->fd, aio->offset);
+                err = os_seek_file(req->io_fd, req->offset);
                 if(err)
                         goto out;
 
-                err = os_read_file(aio->fd, &c, sizeof(c));
+                err = os_read_file(req->io_fd, &c, sizeof(c));
                 break;
         default:
-                printk("do_not_aio - bad request type : %d\n", aio->type);
+                printk("do_not_aio - bad request type : %d\n", req->type);
                 err = -EINVAL;
                 break;
         }
@@ -206,14 +185,14 @@
 
 static int not_aio_thread(void *arg)
 {
-        struct aio_context *aio;
+        struct aio_thread_req req;
         struct aio_thread_reply reply;
         int err;
 
         signal(SIGWINCH, SIG_IGN);
         while(1){
-                err = os_read_file(aio_req_fd_r, &aio, sizeof(aio));
-                if(err != sizeof(aio)){
+                err = os_read_file(aio_req_fd_r, &req, sizeof(req));
+                if(err != sizeof(req)){
                         if(err < 0)
                                 printk("not_aio_thread - read failed, "
                                        "fd = %d, err = %d\n", aio_req_fd_r,
@@ -224,34 +203,17 @@
                         }
                         continue;
                 }
- again:
-                err = do_not_aio(aio);
-
-                if(update_aio(aio, err))
-                        goto again;
-
-                reply = ((struct aio_thread_reply) { .data 	= aio,
-                                                     .err	= aio->len });
-                err = os_write_file(aio->reply_fd, &reply, sizeof(reply));
+                err = do_not_aio(&req);
+                reply = ((struct aio_thread_reply) { .data 	= req.aio,
+                                                     .err	= err });
+                err = os_write_file(req.aio->reply_fd, &reply, sizeof(reply));
                 if(err != sizeof(reply))
                         printk("not_aio_thread - write failed, fd = %d, "
                                "err = %d\n", aio_req_fd_r, -err);
         }
 }
 
-static int submit_aio_24(struct aio_context *aio)
-{
-        int err;
-
-        err = os_write_file(aio_req_fd_w, &aio, sizeof(aio));
-        if(err == sizeof(aio))
-                err = 0;
-
-        return err;
-}
-
 static int aio_pid = -1;
-static int (*submit_proc)(struct aio_context *aio);
 
 static int init_aio_24(void)
 {
@@ -283,33 +245,11 @@
 #endif
 	printk("2.6 host AIO support not used - falling back to I/O "
 	       "thread\n");
-
-	submit_proc = submit_aio_24;
-
         return 0;
 }
 
 #ifdef HAVE_AIO_ABI
 #define DEFAULT_24_AIO 0
-static int submit_aio_26(struct aio_context *aio)
-{
-	struct aio_thread_reply reply;
-	int err;
-
-	err = do_aio(ctx, aio);
-	if(err){
-		reply = ((struct aio_thread_reply) { .data = aio,
-					             .err  = err });
-		err = os_write_file(aio->reply_fd, &reply, sizeof(reply));
-		if(err != sizeof(reply))
-			printk("submit_aio_26 - write failed, "
-			       "fd = %d, err = %d\n", aio->reply_fd, -err);
-		else err = 0;
-	}
-
-	return err;
-}
-
 static int init_aio_26(void)
 {
         unsigned long stack;
@@ -330,22 +270,39 @@
         aio_pid = err;
 
 	printk("Using 2.6 host AIO\n");
-
-	submit_proc = submit_aio_26;
-
         return 0;
 }
 
+static int submit_aio_26(enum aio_type type, int io_fd, char *buf, int len,
+			 unsigned long long offset, struct aio_context *aio)
+{
+        struct aio_thread_reply reply;
+        int err;
+
+        err = do_aio(ctx, type, io_fd, buf, len, offset, aio);
+        if(err){
+                reply = ((struct aio_thread_reply) { .data = aio,
+                                                     .err  = err });
+                err = os_write_file(aio->reply_fd, &reply, sizeof(reply));
+                if(err != sizeof(reply))
+                        printk("submit_aio_26 - write failed, "
+                               "fd = %d, err = %d\n", aio->reply_fd, -err);
+                else err = 0;
+        }
+
+        return err;
+}
+
 #else
 #define DEFAULT_24_AIO 1
-static int submit_aio_26(struct aio_context *aio)
+static int init_aio_26(void)
 {
         return -ENOSYS;
 }
 
-static int init_aio_26(void)
+static int submit_aio_26(enum aio_type type, int io_fd, char *buf, int len,
+			 unsigned long long offset, struct aio_context *aio)
 {
-	submit_proc = submit_aio_26;
         return -ENOSYS;
 }
 #endif
@@ -412,7 +369,33 @@
 
 __uml_exitcall(exit_aio);
 
-int submit_aio(struct aio_context *aio)
+static int submit_aio_24(enum aio_type type, int io_fd, char *buf, int len,
+			 unsigned long long offset, struct aio_context *aio)
 {
-	return (*submit_proc)(aio);
+        struct aio_thread_req req = { .type 		= type,
+                                      .io_fd		= io_fd,
+                                      .offset		= offset,
+                                      .buf		= buf,
+                                      .len		= len,
+                                      .aio		= aio,
+        };
+        int err;
+
+        err = os_write_file(aio_req_fd_w, &req, sizeof(req));
+        if(err == sizeof(req))
+                err = 0;
+
+        return err;
+}
+
+int submit_aio(enum aio_type type, int io_fd, char *buf, int len,
+               unsigned long long offset, int reply_fd,
+               struct aio_context *aio)
+{
+        aio->reply_fd = reply_fd;
+        if(aio_24)
+                return submit_aio_24(type, io_fd, buf, len, offset, aio);
+        else {
+                return submit_aio_26(type, io_fd, buf, len, offset, aio);
+        }
 }
diff --git a/drivers/firmware/dell_rbu.c b/drivers/firmware/dell_rbu.c
index b667823..4f4ba9b 100644
--- a/drivers/firmware/dell_rbu.c
+++ b/drivers/firmware/dell_rbu.c
@@ -50,7 +50,7 @@
 MODULE_AUTHOR("Abhay Salunke <abhay_salunke@dell.com>");
 MODULE_DESCRIPTION("Driver for updating BIOS image on DELL systems");
 MODULE_LICENSE("GPL");
-MODULE_VERSION("2.0");
+MODULE_VERSION("3.0");
 
 #define BIOS_SCAN_LIMIT 0xffffffff
 #define MAX_IMAGE_LENGTH 16
@@ -62,15 +62,16 @@
 	int dma_alloc;
 	spinlock_t lock;
 	unsigned long packet_read_count;
-	unsigned long packet_write_count;
 	unsigned long num_packets;
 	unsigned long packetsize;
+	unsigned long imagesize;
 	int entry_created;
 } rbu_data;
 
 static char image_type[MAX_IMAGE_LENGTH + 1] = "mono";
 module_param_string(image_type, image_type, sizeof (image_type), 0);
-MODULE_PARM_DESC(image_type, "BIOS image type. choose- mono or packet");
+MODULE_PARM_DESC(image_type,
+	"BIOS image type. choose- mono or packet or init");
 
 struct packet_data {
 	struct list_head list;
@@ -88,55 +89,13 @@
 static void init_packet_head(void)
 {
 	INIT_LIST_HEAD(&packet_data_head.list);
-	rbu_data.packet_write_count = 0;
 	rbu_data.packet_read_count = 0;
 	rbu_data.num_packets = 0;
 	rbu_data.packetsize = 0;
+	rbu_data.imagesize = 0;
 }
 
-static int fill_last_packet(void *data, size_t length)
-{
-	struct list_head *ptemp_list;
-	struct packet_data *packet = NULL;
-	int packet_count = 0;
-
-	pr_debug("fill_last_packet: entry \n");
-
-	if (!rbu_data.num_packets) {
-		pr_debug("fill_last_packet: num_packets=0\n");
-		return -ENOMEM;
-	}
-
-	packet_count = rbu_data.num_packets;
-
-	ptemp_list = (&packet_data_head.list)->prev;
-
-	packet = list_entry(ptemp_list, struct packet_data, list);
-
-	if ((rbu_data.packet_write_count + length) > rbu_data.packetsize) {
-		pr_debug("dell_rbu:%s: packet size data "
-			"overrun\n", __FUNCTION__);
-		return -EINVAL;
-	}
-
-	pr_debug("fill_last_packet : buffer = %p\n", packet->data);
-
-	memcpy((packet->data + rbu_data.packet_write_count), data, length);
-
-	if ((rbu_data.packet_write_count + length) == rbu_data.packetsize) {
-		/*
-		 * this was the last data chunk in the packet
-		 * so reinitialize the packet data counter to zero
-		 */
-		rbu_data.packet_write_count = 0;
-	} else
-		rbu_data.packet_write_count += length;
-
-	pr_debug("fill_last_packet: exit \n");
-	return 0;
-}
-
-static int create_packet(size_t length)
+static int create_packet(void *data, size_t length)
 {
 	struct packet_data *newpacket;
 	int ordernum = 0;
@@ -186,9 +145,11 @@
 	INIT_LIST_HEAD(&newpacket->list);
 	list_add_tail(&newpacket->list, &packet_data_head.list);
 	/*
-	 * packets have fixed size
+	 * packets may not have fixed size
 	 */
-	newpacket->length = rbu_data.packetsize;
+	newpacket->length = length;
+
+	memcpy(newpacket->data, data, length);
 
 	pr_debug("create_packet: exit \n");
 
@@ -198,13 +159,37 @@
 static int packetize_data(void *data, size_t length)
 {
 	int rc = 0;
-
-	if (!rbu_data.packet_write_count) {
-		if ((rc = create_packet(length)))
-			return rc;
+	int done = 0;
+	int packet_length;
+	u8 *temp;
+	u8 *end = (u8 *) data + length;
+	pr_debug("packetize_data: data length %d\n", length);
+	if (!rbu_data.packetsize) {
+		printk(KERN_WARNING
+			"dell_rbu: packetsize not specified\n");
+		return -EIO;
 	}
-	if ((rc = fill_last_packet(data, length)))
-		return rc;
+
+	temp = (u8 *) data;
+
+	/* packetize the hunk */
+	while (!done) {
+		if ((temp + rbu_data.packetsize) < end)
+			packet_length = rbu_data.packetsize;
+		else {
+			/* this is the last packet */
+			packet_length = end - temp;
+			done = 1;
+		}
+
+		if ((rc = create_packet(temp, packet_length)))
+			return rc;
+
+		pr_debug("%lu:%lu\n", temp, (end - temp));
+		temp += packet_length;
+	}
+
+	rbu_data.imagesize = length;
 
 	return rc;
 }
@@ -243,7 +228,7 @@
 	return bytes_copied;
 }
 
-static int packet_read_list(char *data, size_t *pread_length)
+static int packet_read_list(char *data, size_t * pread_length)
 {
 	struct list_head *ptemp_list;
 	int temp_count = 0;
@@ -303,10 +288,9 @@
 			newpacket->ordernum);
 		kfree(newpacket);
 	}
-	rbu_data.packet_write_count = 0;
 	rbu_data.packet_read_count = 0;
 	rbu_data.num_packets = 0;
-	rbu_data.packetsize = 0;
+	rbu_data.imagesize = 0;
 }
 
 /*
@@ -425,7 +409,6 @@
 	size_t bytes_left;
 	size_t data_length;
 	char *ptempBuf = buffer;
-	unsigned long imagesize;
 
 	/* check to see if we have something to return */
 	if (rbu_data.num_packets == 0) {
@@ -434,22 +417,20 @@
 		goto read_rbu_data_exit;
 	}
 
-	imagesize = rbu_data.num_packets * rbu_data.packetsize;
-
-	if (pos > imagesize) {
+	if (pos > rbu_data.imagesize) {
 		retval = 0;
 		printk(KERN_WARNING "dell_rbu:read_packet_data: "
 			"data underrun\n");
 		goto read_rbu_data_exit;
 	}
 
-	bytes_left = imagesize - pos;
+	bytes_left = rbu_data.imagesize - pos;
 	data_length = min(bytes_left, count);
 
 	if ((retval = packet_read_list(ptempBuf, &data_length)) < 0)
 		goto read_rbu_data_exit;
 
-	if ((pos + count) > imagesize) {
+	if ((pos + count) > rbu_data.imagesize) {
 		rbu_data.packet_read_count = 0;
 		/* this was the last copy */
 		retval = bytes_left;
@@ -499,7 +480,7 @@
 }
 
 static ssize_t read_rbu_data(struct kobject *kobj, char *buffer,
-			loff_t pos, size_t count)
+	loff_t pos, size_t count)
 {
 	ssize_t ret_count = 0;
 
@@ -531,13 +512,18 @@
 			memcpy(rbu_data.image_update_buffer,
 				fw->data, fw->size);
 	} else if (!strcmp(image_type, "packet")) {
-		if (!rbu_data.packetsize)
-			rbu_data.packetsize = fw->size;
-		else if (rbu_data.packetsize != fw->size) {
+		/*
+		 * we need to free previous packets if a
+		 * new hunk of packets needs to be downloaded
+		 */
+		packet_empty_list();
+		if (packetize_data(fw->data, fw->size))
+			/* Incase something goes wrong when we are
+			 * in middle of packetizing the data, we
+			 * need to free up whatever packets might
+			 * have been created before we quit.
+			 */
 			packet_empty_list();
-			rbu_data.packetsize = fw->size;
-		}
-		packetize_data(fw->data, fw->size);
 	} else
 		pr_debug("invalid image type specified.\n");
 	spin_unlock(&rbu_data.lock);
@@ -553,7 +539,7 @@
 }
 
 static ssize_t read_rbu_image_type(struct kobject *kobj, char *buffer,
-			loff_t pos, size_t count)
+	loff_t pos, size_t count)
 {
 	int size = 0;
 	if (!pos)
@@ -562,7 +548,7 @@
 }
 
 static ssize_t write_rbu_image_type(struct kobject *kobj, char *buffer,
-			loff_t pos, size_t count)
+	loff_t pos, size_t count)
 {
 	int rc = count;
 	int req_firm_rc = 0;
@@ -621,25 +607,49 @@
 	return rc;
 }
 
+static ssize_t read_rbu_packet_size(struct kobject *kobj, char *buffer,
+	loff_t pos, size_t count)
+{
+	int size = 0;
+	if (!pos) {
+		spin_lock(&rbu_data.lock);
+		size = sprintf(buffer, "%lu\n", rbu_data.packetsize);
+		spin_unlock(&rbu_data.lock);
+	}
+	return size;
+}
+
+static ssize_t write_rbu_packet_size(struct kobject *kobj, char *buffer,
+	loff_t pos, size_t count)
+{
+	unsigned long temp;
+	spin_lock(&rbu_data.lock);
+	packet_empty_list();
+	sscanf(buffer, "%lu", &temp);
+	if (temp < 0xffffffff)
+		rbu_data.packetsize = temp;
+
+	spin_unlock(&rbu_data.lock);
+	return count;
+}
+
 static struct bin_attribute rbu_data_attr = {
-	.attr = {
-		.name = "data",
-		.owner = THIS_MODULE,
-		.mode = 0444,
-	},
+	.attr = {.name = "data",.owner = THIS_MODULE,.mode = 0444},
 	.read = read_rbu_data,
 };
 
 static struct bin_attribute rbu_image_type_attr = {
-	.attr = {
-		.name = "image_type",
-		.owner = THIS_MODULE,
-		.mode = 0644,
-	},
+	.attr = {.name = "image_type",.owner = THIS_MODULE,.mode = 0644},
 	.read = read_rbu_image_type,
 	.write = write_rbu_image_type,
 };
 
+static struct bin_attribute rbu_packet_size_attr = {
+	.attr = {.name = "packet_size",.owner = THIS_MODULE,.mode = 0644},
+	.read = read_rbu_packet_size,
+	.write = write_rbu_packet_size,
+};
+
 static int __init dcdrbu_init(void)
 {
 	int rc = 0;
@@ -657,6 +667,8 @@
 
 	sysfs_create_bin_file(&rbu_device->dev.kobj, &rbu_data_attr);
 	sysfs_create_bin_file(&rbu_device->dev.kobj, &rbu_image_type_attr);
+	sysfs_create_bin_file(&rbu_device->dev.kobj,
+		&rbu_packet_size_attr);
 
 	rc = request_firmware_nowait(THIS_MODULE, FW_ACTION_NOHOTPLUG,
 		"dell_rbu", &rbu_device->dev, &context, callbackfn_rbu);
diff --git a/drivers/media/video/bttv-cards.c b/drivers/media/video/bttv-cards.c
index 6c33280..0881a17 100644
--- a/drivers/media/video/bttv-cards.c
+++ b/drivers/media/video/bttv-cards.c
@@ -2393,10 +2393,10 @@
 	.tuner          = 0,
 	.tuner_type     = TUNER_LG_TDVS_H062F,
 	.tuner_addr	= ADDR_UNSET,
-	.video_inputs   = 2,
+	.video_inputs   = 3,
 	.audio_inputs   = 1,
 	.svhs           = 2,
-	.muxsel		= { 2, 3 },
+	.muxsel		= { 2, 3, 1 },
 	.gpiomask       = 0x00e00007,
 	.audiomux       = { 0x00400005, 0, 0x00000001, 0, 0x00c00007, 0 },
 	.no_msp34xx     = 1,
diff --git a/drivers/net/e100.c b/drivers/net/e100.c
index fbf1c06..40887f0 100644
--- a/drivers/net/e100.c
+++ b/drivers/net/e100.c
@@ -903,8 +903,8 @@
 
 static void e100_get_defaults(struct nic *nic)
 {
-	struct param_range rfds = { .min = 16, .max = 256, .count = 256 };
-	struct param_range cbs  = { .min = 64, .max = 256, .count = 128 };
+	struct param_range rfds = { .min = 16, .max = 256, .count = 64 };
+	struct param_range cbs  = { .min = 64, .max = 256, .count = 64 };
 
 	pci_read_config_byte(nic->pdev, PCI_REVISION_ID, &nic->rev_id);
 	/* MAC type is encoded as rev ID; exception: ICH is treated as 82559 */
@@ -1007,213 +1007,25 @@
 		c[16], c[17], c[18], c[19], c[20], c[21], c[22], c[23]);
 }
 
-/********************************************************/
-/*  Micro code for 8086:1229 Rev 8                      */
-/********************************************************/
-
-/*  Parameter values for the D101M B-step  */
-#define D101M_CPUSAVER_TIMER_DWORD		78
-#define D101M_CPUSAVER_BUNDLE_DWORD		65
-#define D101M_CPUSAVER_MIN_SIZE_DWORD		126
-
-#define D101M_B_RCVBUNDLE_UCODE \
-{\
-0x00550215, 0xFFFF0437, 0xFFFFFFFF, 0x06A70789, 0xFFFFFFFF, 0x0558FFFF, \
-0x000C0001, 0x00101312, 0x000C0008, 0x00380216, \
-0x0010009C, 0x00204056, 0x002380CC, 0x00380056, \
-0x0010009C, 0x00244C0B, 0x00000800, 0x00124818, \
-0x00380438, 0x00000000, 0x00140000, 0x00380555, \
-0x00308000, 0x00100662, 0x00100561, 0x000E0408, \
-0x00134861, 0x000C0002, 0x00103093, 0x00308000, \
-0x00100624, 0x00100561, 0x000E0408, 0x00100861, \
-0x000C007E, 0x00222C21, 0x000C0002, 0x00103093, \
-0x00380C7A, 0x00080000, 0x00103090, 0x00380C7A, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x0010009C, 0x00244C2D, 0x00010004, 0x00041000, \
-0x003A0437, 0x00044010, 0x0038078A, 0x00000000, \
-0x00100099, 0x00206C7A, 0x0010009C, 0x00244C48, \
-0x00130824, 0x000C0001, 0x00101213, 0x00260C75, \
-0x00041000, 0x00010004, 0x00130826, 0x000C0006, \
-0x002206A8, 0x0013C926, 0x00101313, 0x003806A8, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00080600, 0x00101B10, 0x00050004, 0x00100826, \
-0x00101210, 0x00380C34, 0x00000000, 0x00000000, \
-0x0021155B, 0x00100099, 0x00206559, 0x0010009C, \
-0x00244559, 0x00130836, 0x000C0000, 0x00220C62, \
-0x000C0001, 0x00101B13, 0x00229C0E, 0x00210C0E, \
-0x00226C0E, 0x00216C0E, 0x0022FC0E, 0x00215C0E, \
-0x00214C0E, 0x00380555, 0x00010004, 0x00041000, \
-0x00278C67, 0x00040800, 0x00018100, 0x003A0437, \
-0x00130826, 0x000C0001, 0x00220559, 0x00101313, \
-0x00380559, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00130831, 0x0010090B, 0x00124813, \
-0x000CFF80, 0x002606AB, 0x00041000, 0x00010004, \
-0x003806A8, 0x00000000, 0x00000000, 0x00000000, \
-}
-
-/********************************************************/
-/*  Micro code for 8086:1229 Rev 9                      */
-/********************************************************/
-
-/*  Parameter values for the D101S  */
-#define D101S_CPUSAVER_TIMER_DWORD		78
-#define D101S_CPUSAVER_BUNDLE_DWORD		67
-#define D101S_CPUSAVER_MIN_SIZE_DWORD		128
-
-#define D101S_RCVBUNDLE_UCODE \
-{\
-0x00550242, 0xFFFF047E, 0xFFFFFFFF, 0x06FF0818, 0xFFFFFFFF, 0x05A6FFFF, \
-0x000C0001, 0x00101312, 0x000C0008, 0x00380243, \
-0x0010009C, 0x00204056, 0x002380D0, 0x00380056, \
-0x0010009C, 0x00244F8B, 0x00000800, 0x00124818, \
-0x0038047F, 0x00000000, 0x00140000, 0x003805A3, \
-0x00308000, 0x00100610, 0x00100561, 0x000E0408, \
-0x00134861, 0x000C0002, 0x00103093, 0x00308000, \
-0x00100624, 0x00100561, 0x000E0408, 0x00100861, \
-0x000C007E, 0x00222FA1, 0x000C0002, 0x00103093, \
-0x00380F90, 0x00080000, 0x00103090, 0x00380F90, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x0010009C, 0x00244FAD, 0x00010004, 0x00041000, \
-0x003A047E, 0x00044010, 0x00380819, 0x00000000, \
-0x00100099, 0x00206FFD, 0x0010009A, 0x0020AFFD, \
-0x0010009C, 0x00244FC8, 0x00130824, 0x000C0001, \
-0x00101213, 0x00260FF7, 0x00041000, 0x00010004, \
-0x00130826, 0x000C0006, 0x00220700, 0x0013C926, \
-0x00101313, 0x00380700, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00080600, 0x00101B10, 0x00050004, 0x00100826, \
-0x00101210, 0x00380FB6, 0x00000000, 0x00000000, \
-0x002115A9, 0x00100099, 0x002065A7, 0x0010009A, \
-0x0020A5A7, 0x0010009C, 0x002445A7, 0x00130836, \
-0x000C0000, 0x00220FE4, 0x000C0001, 0x00101B13, \
-0x00229F8E, 0x00210F8E, 0x00226F8E, 0x00216F8E, \
-0x0022FF8E, 0x00215F8E, 0x00214F8E, 0x003805A3, \
-0x00010004, 0x00041000, 0x00278FE9, 0x00040800, \
-0x00018100, 0x003A047E, 0x00130826, 0x000C0001, \
-0x002205A7, 0x00101313, 0x003805A7, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00130831, \
-0x0010090B, 0x00124813, 0x000CFF80, 0x00260703, \
-0x00041000, 0x00010004, 0x00380700  \
-}
-
-/********************************************************/
-/*  Micro code for the 8086:1229 Rev F/10               */
-/********************************************************/
-
-/*  Parameter values for the D102 E-step  */
-#define D102_E_CPUSAVER_TIMER_DWORD		42
-#define D102_E_CPUSAVER_BUNDLE_DWORD		54
-#define D102_E_CPUSAVER_MIN_SIZE_DWORD		46
-
-#define     D102_E_RCVBUNDLE_UCODE \
-{\
-0x007D028F, 0x0E4204F9, 0x14ED0C85, 0x14FA14E9, 0x0EF70E36, 0x1FFF1FFF, \
-0x00E014B9, 0x00000000, 0x00000000, 0x00000000, \
-0x00E014BD, 0x00000000, 0x00000000, 0x00000000, \
-0x00E014D5, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00E014C1, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00E014C8, 0x00000000, 0x00000000, 0x00000000, \
-0x00200600, 0x00E014EE, 0x00000000, 0x00000000, \
-0x0030FF80, 0x00940E46, 0x00038200, 0x00102000, \
-0x00E00E43, 0x00000000, 0x00000000, 0x00000000, \
-0x00300006, 0x00E014FB, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00906E41, 0x00800E3C, 0x00E00E39, 0x00000000, \
-0x00906EFD, 0x00900EFD, 0x00E00EF8, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-0x00000000, 0x00000000, 0x00000000, 0x00000000, \
-}
-
 static void e100_load_ucode(struct nic *nic, struct cb *cb, struct sk_buff *skb)
 {
-/* *INDENT-OFF* */
-	static struct {
-		u32 ucode[UCODE_SIZE + 1];
-		u8 mac;
-		u8 timer_dword;
-		u8 bundle_dword;
-		u8 min_size_dword;
-	} ucode_opts[] = {
-		{ D101M_B_RCVBUNDLE_UCODE,
-		  mac_82559_D101M,
-		  D101M_CPUSAVER_TIMER_DWORD,
-		  D101M_CPUSAVER_BUNDLE_DWORD,
-		  D101M_CPUSAVER_MIN_SIZE_DWORD },
-		{ D101S_RCVBUNDLE_UCODE,
-		  mac_82559_D101S,
-		  D101S_CPUSAVER_TIMER_DWORD,
-		  D101S_CPUSAVER_BUNDLE_DWORD,
-		  D101S_CPUSAVER_MIN_SIZE_DWORD },
-		{ D102_E_RCVBUNDLE_UCODE,
-		  mac_82551_F,
-		  D102_E_CPUSAVER_TIMER_DWORD,
-		  D102_E_CPUSAVER_BUNDLE_DWORD,
-		  D102_E_CPUSAVER_MIN_SIZE_DWORD },
-		{ D102_E_RCVBUNDLE_UCODE,
-		  mac_82551_10,
-		  D102_E_CPUSAVER_TIMER_DWORD,
-		  D102_E_CPUSAVER_BUNDLE_DWORD,
-		  D102_E_CPUSAVER_MIN_SIZE_DWORD },
-		{ {0}, 0, 0, 0, 0}
-	}, *opts;
-/* *INDENT-ON* */
+	int i;
+	static const u32 ucode[UCODE_SIZE] = {
+		/* NFS packets are misinterpreted as TCO packets and
+		 * incorrectly routed to the BMC over SMBus.  This
+		 * microcode patch checks the fragmented IP bit in the
+		 * NFS/UDP header to distinguish between NFS and TCO. */
+		0x0EF70E36, 0x1FFF1FFF, 0x1FFF1FFF, 0x1FFF1FFF, 0x1FFF1FFF,
+		0x1FFF1FFF, 0x00906E41, 0x00800E3C, 0x00E00E39, 0x00000000,
+		0x00906EFD, 0x00900EFD,	0x00E00EF8,
+	};
 
-#define BUNDLESMALL 1
-#define BUNDLEMAX 50
-#define INTDELAY 15000
-
-	opts = ucode_opts;
-
-	/* do not load u-code for ICH devices */
-	if (nic->flags & ich)
-		return;
-
-	/* Search for ucode match against h/w rev_id */
-	while (opts->mac) {
-		if (nic->mac == opts->mac) {
-			int i;
-			u32 *ucode = opts->ucode;
-
-			/* Insert user-tunable settings */
-			ucode[opts->timer_dword] &= 0xFFFF0000;
-			ucode[opts->timer_dword] |=
-				(u16) INTDELAY;
-			ucode[opts->bundle_dword] &= 0xFFFF0000;
-			ucode[opts->bundle_dword] |= (u16) BUNDLEMAX;
-			ucode[opts->min_size_dword] &= 0xFFFF0000;
-			ucode[opts->min_size_dword] |=
-				(BUNDLESMALL) ?  0xFFFF : 0xFF80;
-
-			for(i = 0; i < UCODE_SIZE; i++)
-				cb->u.ucode[i] = cpu_to_le32(ucode[i]);
-			cb->command = cpu_to_le16(cb_ucode);
-			return;
-		}
-		opts++;
-	}
-
-	cb->command = cpu_to_le16(cb_nop);
+	if(nic->mac == mac_82551_F || nic->mac == mac_82551_10) {
+		for(i = 0; i < UCODE_SIZE; i++)
+			cb->u.ucode[i] = cpu_to_le32(ucode[i]);
+		cb->command = cpu_to_le16(cb_ucode);
+	} else
+		cb->command = cpu_to_le16(cb_nop);
 }
 
 static void e100_setup_iaaddr(struct nic *nic, struct cb *cb,
diff --git a/drivers/s390/cio/device.c b/drivers/s390/cio/device.c
index 14c76f5..9adc11e 100644
--- a/drivers/s390/cio/device.c
+++ b/drivers/s390/cio/device.c
@@ -544,7 +544,7 @@
 		.sibling = sibling,
 	};
 
-	dev = bus_find_device(&css_bus_type, NULL, &data, match_devno);
+	dev = bus_find_device(&ccw_bus_type, NULL, &data, match_devno);
 
 	return dev ? to_ccwdev(dev) : NULL;
 }
diff --git a/drivers/serial/imx.c b/drivers/serial/imx.c
index 53e0323..bdb4e45 100644
--- a/drivers/serial/imx.c
+++ b/drivers/serial/imx.c
@@ -73,7 +73,7 @@
 	struct uart_port	port;
 	struct timer_list	timer;
 	unsigned int		old_status;
-	int txirq,rxirq;
+	int txirq,rxirq,rtsirq;
 };
 
 /*
@@ -181,6 +181,22 @@
 		imx_transmit_buffer(sport);
 }
 
+static irqreturn_t imx_rtsint(int irq, void *dev_id, struct pt_regs *regs)
+{
+	struct imx_port *sport = (struct imx_port *)dev_id;
+	unsigned int val = USR1((u32)sport->port.membase)&USR1_RTSS;
+	unsigned long flags;
+
+	spin_lock_irqsave(&sport->port.lock, flags);
+
+	USR1((u32)sport->port.membase) = USR1_RTSD;
+	uart_handle_cts_change(&sport->port, !!val);
+	wake_up_interruptible(&sport->port.info->delta_msr_wait);
+
+	spin_unlock_irqrestore(&sport->port.lock, flags);
+	return IRQ_HANDLED;
+}
+
 static irqreturn_t imx_txint(int irq, void *dev_id, struct pt_regs *regs)
 {
 	struct imx_port *sport = (struct imx_port *)dev_id;
@@ -386,15 +402,21 @@
 	if (retval) goto error_out1;
 
 	retval = request_irq(sport->txirq, imx_txint, 0,
-			     "imx-uart", sport);
+			     DRIVER_NAME, sport);
 	if (retval) goto error_out2;
 
+	retval = request_irq(sport->rtsirq, imx_rtsint, 0,
+			     DRIVER_NAME, sport);
+	if (retval) goto error_out3;
+	set_irq_type(sport->rtsirq, IRQT_BOTHEDGE);
+
 	/*
 	 * Finally, clear and enable interrupts
 	 */
 
+	USR1((u32)sport->port.membase) = USR1_RTSD;
 	UCR1((u32)sport->port.membase) |=
-	                 (UCR1_TXMPTYEN | UCR1_RRDYEN | UCR1_UARTEN);
+	                 (UCR1_TXMPTYEN | UCR1_RRDYEN | UCR1_RTSDEN | UCR1_UARTEN);
 
 	UCR2((u32)sport->port.membase) |= (UCR2_RXEN | UCR2_TXEN);
 	/*
@@ -406,6 +428,8 @@
 
 	return 0;
 
+error_out3:
+	free_irq(sport->txirq, sport);
 error_out2:
 	free_irq(sport->rxirq, sport);
 error_out1:
@@ -424,6 +448,7 @@
 	/*
 	 * Free the interrupts
 	 */
+	free_irq(sport->rtsirq, sport);
 	free_irq(sport->txirq, sport);
 	free_irq(sport->rxirq, sport);
 
@@ -432,7 +457,7 @@
 	 */
 
 	UCR1((u32)sport->port.membase) &=
-	                 ~(UCR1_TXMPTYEN | UCR1_RRDYEN | UCR1_UARTEN);
+	                 ~(UCR1_TXMPTYEN | UCR1_RRDYEN | UCR1_RTSDEN | UCR1_UARTEN);
 }
 
 static void
@@ -522,7 +547,7 @@
 	 * disable interrupts and drain transmitter
 	 */
 	old_ucr1 = UCR1((u32)sport->port.membase);
-	UCR1((u32)sport->port.membase) &= ~(UCR1_TXMPTYEN | UCR1_RRDYEN);
+	UCR1((u32)sport->port.membase) &= ~(UCR1_TXMPTYEN | UCR1_RRDYEN | UCR1_RTSDEN);
 
 	while ( !(USR2((u32)sport->port.membase) & USR2_TXDC))
 		barrier();
@@ -643,6 +668,7 @@
 	{
 	.txirq  = UART1_MINT_TX,
 	.rxirq  = UART1_MINT_RX,
+	.rtsirq = UART1_MINT_RTS,
 	.port	= {
 		.type		= PORT_IMX,
 		.iotype		= SERIAL_IO_MEM,
@@ -658,6 +684,7 @@
 	}, {
 	.txirq  = UART2_MINT_TX,
 	.rxirq  = UART2_MINT_RX,
+	.rtsirq = UART2_MINT_RTS,
 	.port	= {
 		.type		= PORT_IMX,
 		.iotype		= SERIAL_IO_MEM,
@@ -737,7 +764,7 @@
 
 	UCR1((u32)sport->port.membase) =
 	                   (old_ucr1 | UCR1_UARTCLKEN | UCR1_UARTEN)
-	                   & ~(UCR1_TXMPTYEN | UCR1_RRDYEN);
+	                   & ~(UCR1_TXMPTYEN | UCR1_RRDYEN | UCR1_RTSDEN);
 	UCR2((u32)sport->port.membase) = old_ucr2 | UCR2_TXEN;
 
 	/*
diff --git a/drivers/serial/pxa.c b/drivers/serial/pxa.c
index 672b359..90c2a86 100644
--- a/drivers/serial/pxa.c
+++ b/drivers/serial/pxa.c
@@ -499,7 +499,7 @@
 	/*
 	 * Update the per-port timeout.
 	 */
-	uart_update_timeout(port, termios->c_cflag, quot);
+	uart_update_timeout(port, termios->c_cflag, baud);
 
 	up->port.read_status_mask = UART_LSR_OE | UART_LSR_THRE | UART_LSR_DR;
 	if (termios->c_iflag & INPCK)
diff --git a/drivers/serial/s3c2410.c b/drivers/serial/s3c2410.c
index eff2158..52692aa 100644
--- a/drivers/serial/s3c2410.c
+++ b/drivers/serial/s3c2410.c
@@ -1092,8 +1092,8 @@
 
 static int probe_index = 0;
 
-int s3c24xx_serial_probe(struct device *_dev,
-			 struct s3c24xx_uart_info *info)
+static int s3c24xx_serial_probe(struct device *_dev,
+				struct s3c24xx_uart_info *info)
 {
 	struct s3c24xx_uart_port *ourport;
 	struct platform_device *dev = to_platform_device(_dev);
@@ -1120,7 +1120,7 @@
 	return ret;
 }
 
-int s3c24xx_serial_remove(struct device *_dev)
+static int s3c24xx_serial_remove(struct device *_dev)
 {
 	struct uart_port *port = s3c24xx_dev_to_port(_dev);
 
@@ -1134,7 +1134,8 @@
 
 #ifdef CONFIG_PM
 
-int s3c24xx_serial_suspend(struct device *dev, pm_message_t state, u32 level)
+static int s3c24xx_serial_suspend(struct device *dev, pm_message_t state,
+				  u32 level)
 {
 	struct uart_port *port = s3c24xx_dev_to_port(dev);
 
@@ -1144,7 +1145,7 @@
 	return 0;
 }
 
-int s3c24xx_serial_resume(struct device *dev, u32 level)
+static int s3c24xx_serial_resume(struct device *dev, u32 level)
 {
 	struct uart_port *port = s3c24xx_dev_to_port(dev);
 	struct s3c24xx_uart_port *ourport = to_ourport(port);
@@ -1165,8 +1166,8 @@
 #define s3c24xx_serial_resume  NULL
 #endif
 
-int s3c24xx_serial_init(struct device_driver *drv,
-			struct s3c24xx_uart_info *info)
+static int s3c24xx_serial_init(struct device_driver *drv,
+			       struct s3c24xx_uart_info *info)
 {
 	dbg("s3c24xx_serial_init(%p,%p)\n", drv, info);
 	return driver_register(drv);
diff --git a/drivers/serial/sunsab.c b/drivers/serial/sunsab.c
index e971156..ba9381f 100644
--- a/drivers/serial/sunsab.c
+++ b/drivers/serial/sunsab.c
@@ -274,7 +274,6 @@
 	if (uart_circ_empty(xmit) || uart_tx_stopped(&up->port)) {
 		up->interrupt_mask1 |= SAB82532_IMR1_XPR;
 		writeb(up->interrupt_mask1, &up->regs->w.imr1);
-		uart_write_wakeup(&up->port);
 		return;
 	}
 
diff --git a/drivers/serial/sunzilog.c b/drivers/serial/sunzilog.c
index d754457..7653d6c 100644
--- a/drivers/serial/sunzilog.c
+++ b/drivers/serial/sunzilog.c
@@ -517,10 +517,9 @@
 	if (up->port.info == NULL)
 		goto ack_tx_int;
 	xmit = &up->port.info->xmit;
-	if (uart_circ_empty(xmit)) {
-		uart_write_wakeup(&up->port);
+	if (uart_circ_empty(xmit))
 		goto ack_tx_int;
-	}
+
 	if (uart_tx_stopped(&up->port))
 		goto ack_tx_int;
 
diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c
index a4799e9..bbc3cc6 100644
--- a/fs/9p/vfs_file.c
+++ b/fs/9p/vfs_file.c
@@ -175,16 +175,16 @@
 }
 
 /**
- * v9fs_read - read from a file (internal)
+ * v9fs_file_read - read from a file
  * @filep: file pointer to read
  * @data: data buffer to read data into
  * @count: size of buffer
  * @offset: offset at which to read data
  *
  */
-
 static ssize_t
-v9fs_read(struct file *filp, char *buffer, size_t count, loff_t * offset)
+v9fs_file_read(struct file *filp, char __user * data, size_t count,
+	       loff_t * offset)
 {
 	struct inode *inode = filp->f_dentry->d_inode;
 	struct v9fs_session_info *v9ses = v9fs_inode2v9ses(inode);
@@ -194,6 +194,7 @@
 	int rsize = 0;
 	int result = 0;
 	int total = 0;
+	int n;
 
 	dprintk(DEBUG_VFS, "\n");
 
@@ -216,10 +217,15 @@
 		} else
 			*offset += result;
 
-		/* XXX - extra copy */
-		memcpy(buffer, fcall->params.rread.data, result);
+		n = copy_to_user(data, fcall->params.rread.data, result);
+		if (n) {
+			dprintk(DEBUG_ERROR, "Problem copying to user %d\n", n);
+			kfree(fcall);
+			return -EFAULT;
+		}
+
 		count -= result;
-		buffer += result;
+		data += result;
 		total += result;
 
 		kfree(fcall);
@@ -232,101 +238,6 @@
 }
 
 /**
- * v9fs_file_read - read from a file
- * @filep: file pointer to read
- * @data: data buffer to read data into
- * @count: size of buffer
- * @offset: offset at which to read data
- *
- */
-
-static ssize_t
-v9fs_file_read(struct file *filp, char __user * data, size_t count,
-	       loff_t * offset)
-{
-	int retval = -1;
-	int ret = 0;
-	char *buffer;
-
-	buffer = kmalloc(count, GFP_KERNEL);
-	if (!buffer)
-		return -ENOMEM;
-
-	retval = v9fs_read(filp, buffer, count, offset);
-	if (retval > 0) {
-		if ((ret = copy_to_user(data, buffer, retval)) != 0) {
-			dprintk(DEBUG_ERROR, "Problem copying to user %d\n",
-				ret);
-			retval = ret;
-		}
-	}
-
-	kfree(buffer);
-
-	return retval;
-}
-
-/**
- * v9fs_write - write to a file
- * @filep: file pointer to write
- * @data: data buffer to write data from
- * @count: size of buffer
- * @offset: offset at which to write data
- *
- */
-
-static ssize_t
-v9fs_write(struct file *filp, char *buffer, size_t count, loff_t * offset)
-{
-	struct inode *inode = filp->f_dentry->d_inode;
-	struct v9fs_session_info *v9ses = v9fs_inode2v9ses(inode);
-	struct v9fs_fid *v9fid = filp->private_data;
-	struct v9fs_fcall *fcall;
-	int fid = v9fid->fid;
-	int result = -EIO;
-	int rsize = 0;
-	int total = 0;
-
-	dprintk(DEBUG_VFS, "data %p count %d offset %x\n", buffer, (int)count,
-		(int)*offset);
-	rsize = v9ses->maxdata - V9FS_IOHDRSZ;
-	if (v9fid->iounit != 0 && rsize > v9fid->iounit)
-		rsize = v9fid->iounit;
-
-	dump_data(buffer, count);
-
-	do {
-		if (count < rsize)
-			rsize = count;
-
-		result =
-		    v9fs_t_write(v9ses, fid, *offset, rsize, buffer, &fcall);
-		if (result < 0) {
-			eprintk(KERN_ERR, "error while writing: %s(%d)\n",
-				FCALL_ERROR(fcall), result);
-			kfree(fcall);
-			return result;
-		} else
-			*offset += result;
-
-		kfree(fcall);
-
-		if (result != rsize) {
-			eprintk(KERN_ERR,
-				"short write: v9fs_t_write returned %d\n",
-				result);
-			break;
-		}
-
-		count -= result;
-		buffer += result;
-		total += result;
-	} while (count);
-
-	return total;
-}
-
-/**
  * v9fs_file_write - write to a file
  * @filep: file pointer to write
  * @data: data buffer to write data from
@@ -339,24 +250,65 @@
 v9fs_file_write(struct file *filp, const char __user * data,
 		size_t count, loff_t * offset)
 {
-	int ret = -1;
-	char *buffer;
+	struct inode *inode = filp->f_dentry->d_inode;
+	struct v9fs_session_info *v9ses = v9fs_inode2v9ses(inode);
+	struct v9fs_fid *v9fid = filp->private_data;
+	struct v9fs_fcall *fcall;
+	int fid = v9fid->fid;
+	int result = -EIO;
+	int rsize = 0;
+	int total = 0;
+	char *buf;
 
-	buffer = kmalloc(count, GFP_KERNEL);
-	if (buffer == NULL)
+	dprintk(DEBUG_VFS, "data %p count %d offset %x\n", data, (int)count,
+		(int)*offset);
+	rsize = v9ses->maxdata - V9FS_IOHDRSZ;
+	if (v9fid->iounit != 0 && rsize > v9fid->iounit)
+		rsize = v9fid->iounit;
+
+	buf = kmalloc(v9ses->maxdata - V9FS_IOHDRSZ, GFP_KERNEL);
+	if (!buf)
 		return -ENOMEM;
 
-	ret = copy_from_user(buffer, data, count);
-	if (ret) {
-		dprintk(DEBUG_ERROR, "Problem copying from user\n");
-		ret = -EFAULT;
-	} else {
-		ret = v9fs_write(filp, buffer, count, offset);
-	}
+	do {
+		if (count < rsize)
+			rsize = count;
 
-	kfree(buffer);
+		result = copy_from_user(buf, data, rsize);
+		if (result) {
+			dprintk(DEBUG_ERROR, "Problem copying from user\n");
+			kfree(buf);
+			return -EFAULT;
+		}
 
-	return ret;
+		dump_data(buf, rsize);
+		result = v9fs_t_write(v9ses, fid, *offset, rsize, buf, &fcall);
+		if (result < 0) {
+			eprintk(KERN_ERR, "error while writing: %s(%d)\n",
+				FCALL_ERROR(fcall), result);
+			kfree(fcall);
+			kfree(buf);
+			return result;
+		} else
+			*offset += result;
+
+		kfree(fcall);
+		fcall = NULL;
+
+		if (result != rsize) {
+			eprintk(KERN_ERR,
+				"short write: v9fs_t_write returned %d\n",
+				result);
+			break;
+		}
+
+		count -= result;
+		data += result;
+		total += result;
+	} while (count);
+
+	kfree(buf);
+	return total;
 }
 
 struct file_operations v9fs_file_operations = {
diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index 7976a23..d4b1557 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -905,7 +905,7 @@
 		send_sig(SIGKILL, current, 0);
 		goto out_free_dentry;
 	}
-	if (padzero(elf_bss)) {
+	if (likely(elf_bss != elf_brk) && unlikely(padzero(elf_bss))) {
 		send_sig(SIGSEGV, current, 0);
 		retval = -EFAULT; /* Nobody gets to see this, but.. */
 		goto out_free_dentry;
diff --git a/fs/nfs_common/nfsacl.c b/fs/nfs_common/nfsacl.c
index 251e5a1..0c2be8c 100644
--- a/fs/nfs_common/nfsacl.c
+++ b/fs/nfs_common/nfsacl.c
@@ -48,43 +48,26 @@
 		(struct nfsacl_encode_desc *) desc;
 	u32 *p = (u32 *) elem;
 
-	if (nfsacl_desc->count < nfsacl_desc->acl->a_count) {
-		struct posix_acl_entry *entry =
-			&nfsacl_desc->acl->a_entries[nfsacl_desc->count++];
+	struct posix_acl_entry *entry =
+		&nfsacl_desc->acl->a_entries[nfsacl_desc->count++];
 
-		*p++ = htonl(entry->e_tag | nfsacl_desc->typeflag);
-		switch(entry->e_tag) {
-			case ACL_USER_OBJ:
-				*p++ = htonl(nfsacl_desc->uid);
-				break;
-			case ACL_GROUP_OBJ:
-				*p++ = htonl(nfsacl_desc->gid);
-				break;
-			case ACL_USER:
-			case ACL_GROUP:
-				*p++ = htonl(entry->e_id);
-				break;
-			default:  /* Solaris depends on that! */
-				*p++ = 0;
-				break;
-		}
-		*p++ = htonl(entry->e_perm & S_IRWXO);
-	} else {
-		const struct posix_acl_entry *pa, *pe;
-		int group_obj_perm = ACL_READ|ACL_WRITE|ACL_EXECUTE;
-
-		FOREACH_ACL_ENTRY(pa, nfsacl_desc->acl, pe) {
-			if (pa->e_tag == ACL_GROUP_OBJ) {
-				group_obj_perm = pa->e_perm & S_IRWXO;
-				break;
-			}
-		}
-		/* fake up ACL_MASK entry */
-		*p++ = htonl(ACL_MASK | nfsacl_desc->typeflag);
-		*p++ = htonl(0);
-		*p++ = htonl(group_obj_perm);
+	*p++ = htonl(entry->e_tag | nfsacl_desc->typeflag);
+	switch(entry->e_tag) {
+		case ACL_USER_OBJ:
+			*p++ = htonl(nfsacl_desc->uid);
+			break;
+		case ACL_GROUP_OBJ:
+			*p++ = htonl(nfsacl_desc->gid);
+			break;
+		case ACL_USER:
+		case ACL_GROUP:
+			*p++ = htonl(entry->e_id);
+			break;
+		default:  /* Solaris depends on that! */
+			*p++ = 0;
+			break;
 	}
-
+	*p++ = htonl(entry->e_perm & S_IRWXO);
 	return 0;
 }
 
@@ -105,11 +88,28 @@
 		.gid = inode->i_gid,
 	};
 	int err;
+	struct posix_acl *acl2 = NULL;
 
 	if (entries > NFS_ACL_MAX_ENTRIES ||
 	    xdr_encode_word(buf, base, entries))
 		return -EINVAL;
+	if (encode_entries && acl && acl->a_count == 3) {
+		/* Fake up an ACL_MASK entry. */
+		acl2 = posix_acl_alloc(4, GFP_KERNEL);
+		if (!acl2)
+			return -ENOMEM;
+		/* Insert entries in canonical order: other orders seem
+		 to confuse Solaris VxFS. */
+		acl2->a_entries[0] = acl->a_entries[0];  /* ACL_USER_OBJ */
+		acl2->a_entries[1] = acl->a_entries[1];  /* ACL_GROUP_OBJ */
+		acl2->a_entries[2] = acl->a_entries[1];  /* ACL_MASK */
+		acl2->a_entries[2].e_tag = ACL_MASK;
+		acl2->a_entries[3] = acl->a_entries[2];  /* ACL_OTHER */
+		nfsacl_desc.acl = acl2;
+	}
 	err = xdr_encode_array2(buf, base + 4, &nfsacl_desc.desc);
+	if (acl2)
+		posix_acl_release(acl2);
 	if (!err)
 		err = 8 + nfsacl_desc.desc.elem_size *
 			  nfsacl_desc.desc.array_len;
diff --git a/include/asm-arm/arch-pxa/pxa-regs.h b/include/asm-arm/arch-pxa/pxa-regs.h
index 939d9e5..13fa2de 100644
--- a/include/asm-arm/arch-pxa/pxa-regs.h
+++ b/include/asm-arm/arch-pxa/pxa-regs.h
@@ -126,8 +126,8 @@
 #define DRCMR12		__REG(0x40000130)  /* Request to Channel Map Register for AC97 audio transmit Request */
 #define DRCMR13		__REG(0x40000134)  /* Request to Channel Map Register for SSP receive Request */
 #define DRCMR14		__REG(0x40000138)  /* Request to Channel Map Register for SSP transmit Request */
-#define DRCMR15		__REG(0x4000013c)  /* Reserved */
-#define DRCMR16		__REG(0x40000140)  /* Reserved */
+#define DRCMR15		__REG(0x4000013c)  /* Request to Channel Map Register for SSP2 receive Request */
+#define DRCMR16		__REG(0x40000140)  /* Request to Channel Map Register for SSP2 transmit Request */
 #define DRCMR17		__REG(0x40000144)  /* Request to Channel Map Register for ICP receive Request */
 #define DRCMR18		__REG(0x40000148)  /* Request to Channel Map Register for ICP transmit Request */
 #define DRCMR19		__REG(0x4000014c)  /* Request to Channel Map Register for STUART receive Request */
@@ -151,7 +151,8 @@
 #define DRCMR37		__REG(0x40000194)  /* Request to Channel Map Register for USB endpoint 13 Request */
 #define DRCMR38		__REG(0x40000198)  /* Request to Channel Map Register for USB endpoint 14 Request */
 #define DRCMR39		__REG(0x4000019C)  /* Reserved */
-
+#define DRCMR66		__REG(0x40001108)  /* Request to Channel Map Register for SSP3 receive Request */
+#define DRCMR67		__REG(0x4000110C)  /* Request to Channel Map Register for SSP3 transmit Request */
 #define DRCMR68		__REG(0x40001110)  /* Request to Channel Map Register for Camera FIFO 0 Request */
 #define DRCMR69		__REG(0x40001114)  /* Request to Channel Map Register for Camera FIFO 1 Request */
 #define DRCMR70		__REG(0x40001118)  /* Request to Channel Map Register for Camera FIFO 2 Request */
diff --git a/include/asm-arm/arch-s3c2410/hardware.h b/include/asm-arm/arch-s3c2410/hardware.h
index 48a3991..1c9de29 100644
--- a/include/asm-arm/arch-s3c2410/hardware.h
+++ b/include/asm-arm/arch-s3c2410/hardware.h
@@ -92,6 +92,13 @@
 
 extern unsigned int s3c2410_modify_misccr(unsigned int clr, unsigned int chg);
 
+#ifdef CONFIG_CPU_S3C2440
+
+extern int s3c2440_set_dsc(unsigned int pin, unsigned int value);
+
+#endif /* CONFIG_CPU_S3C2440 */
+
+
 #endif /* __ASSEMBLY__ */
 
 #include <asm/sizes.h>
diff --git a/include/asm-powerpc/timex.h b/include/asm-powerpc/timex.h
index 51c5b31..c02d15ac 100644
--- a/include/asm-powerpc/timex.h
+++ b/include/asm-powerpc/timex.h
@@ -10,7 +10,7 @@
 #include <linux/config.h>
 #include <asm/cputable.h>
 
-#define CLOCK_TICK_RATE	1193180 /* Underlying HZ */
+#define CLOCK_TICK_RATE	1024000 /* Underlying HZ */
 
 typedef unsigned long cycles_t;
 
diff --git a/include/asm-ppc/cputable.h b/include/asm-ppc/cputable.h
index 41d8f84..e17c492 100644
--- a/include/asm-ppc/cputable.h
+++ b/include/asm-ppc/cputable.h
@@ -24,6 +24,7 @@
 #define PPC_FEATURE_HAS_SPE		0x00800000
 #define PPC_FEATURE_HAS_EFP_SINGLE	0x00400000
 #define PPC_FEATURE_HAS_EFP_DOUBLE	0x00200000
+#define PPC_FEATURE_NO_TB		0x00100000
 
 #ifdef __KERNEL__
 
diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h
index b15826f..fe97783 100644
--- a/include/linux/cpumask.h
+++ b/include/linux/cpumask.h
@@ -392,4 +392,16 @@
 #define for_each_online_cpu(cpu)  for_each_cpu_mask((cpu), cpu_online_map)
 #define for_each_present_cpu(cpu) for_each_cpu_mask((cpu), cpu_present_map)
 
+/* Find the highest possible smp_processor_id() */
+static inline unsigned int highest_possible_processor_id(void)
+{
+	unsigned int cpu, highest = 0;
+
+	for_each_cpu_mask(cpu, cpu_possible_map)
+		highest = cpu;
+
+	return highest;
+}
+
+
 #endif /* __LINUX_CPUMASK_H */
diff --git a/include/linux/netfilter/nfnetlink.h b/include/linux/netfilter/nfnetlink.h
index 1d5b10a..f08e870 100644
--- a/include/linux/netfilter/nfnetlink.h
+++ b/include/linux/netfilter/nfnetlink.h
@@ -41,11 +41,15 @@
 struct nfattr
 {
 	u_int16_t nfa_len;
-	u_int16_t nfa_type;
+	u_int16_t nfa_type;	/* we use 15 bits for the type, and the highest
+				 * bit to indicate whether the payload is nested */
 } __attribute__ ((packed));
 
-/* FIXME: Shamelessly copy and pasted from rtnetlink.h, it's time
- * 	  to put this in a generic file */
+/* FIXME: Apart from NFNL_NFA_NESTED shamelessly copy and pasted from
+ * rtnetlink.h, it's time to put this in a generic file */
+
+#define NFNL_NFA_NEST	0x8000
+#define NFA_TYPE(attr) 	((attr)->nfa_type & 0x7fff)
 
 #define NFA_ALIGNTO     4
 #define NFA_ALIGN(len)	(((len) + NFA_ALIGNTO - 1) & ~(NFA_ALIGNTO - 1))
@@ -59,7 +63,7 @@
 #define NFA_PAYLOAD(nfa) ((int)((nfa)->nfa_len) - NFA_LENGTH(0))
 #define NFA_NEST(skb, type) \
 ({	struct nfattr *__start = (struct nfattr *) (skb)->tail; \
-	NFA_PUT(skb, type, 0, NULL); \
+	NFA_PUT(skb, (NFNL_NFA_NEST | type), 0, NULL); \
 	__start;  })
 #define NFA_NEST_END(skb, start) \
 ({      (start)->nfa_len = ((skb)->tail - (unsigned char *) (start)); \
diff --git a/include/linux/netfilter/nfnetlink_conntrack.h b/include/linux/netfilter/nfnetlink_conntrack.h
index 5c55751..116fcac 100644
--- a/include/linux/netfilter/nfnetlink_conntrack.h
+++ b/include/linux/netfilter/nfnetlink_conntrack.h
@@ -70,15 +70,24 @@
 
 enum ctattr_protoinfo {
 	CTA_PROTOINFO_UNSPEC,
-	CTA_PROTOINFO_TCP_STATE,
+	CTA_PROTOINFO_TCP,
 	__CTA_PROTOINFO_MAX
 };
 #define CTA_PROTOINFO_MAX (__CTA_PROTOINFO_MAX - 1)
 
+enum ctattr_protoinfo_tcp {
+	CTA_PROTOINFO_TCP_UNSPEC,
+	CTA_PROTOINFO_TCP_STATE,
+	__CTA_PROTOINFO_TCP_MAX
+};
+#define CTA_PROTOINFO_TCP_MAX (__CTA_PROTOINFO_TCP_MAX - 1)
+
 enum ctattr_counters {
 	CTA_COUNTERS_UNSPEC,
-	CTA_COUNTERS_PACKETS,
-	CTA_COUNTERS_BYTES,
+	CTA_COUNTERS_PACKETS,		/* old 64bit counters */
+	CTA_COUNTERS_BYTES,		/* old 64bit counters */
+	CTA_COUNTERS32_PACKETS,
+	CTA_COUNTERS32_BYTES,
 	__CTA_COUNTERS_MAX
 };
 #define CTA_COUNTERS_MAX (__CTA_COUNTERS_MAX - 1)
diff --git a/include/linux/netfilter_ipv4/ip_conntrack.h b/include/linux/netfilter_ipv4/ip_conntrack.h
index 4ced387..d078bb9 100644
--- a/include/linux/netfilter_ipv4/ip_conntrack.h
+++ b/include/linux/netfilter_ipv4/ip_conntrack.h
@@ -117,6 +117,10 @@
 	/* NAT info */
 	IPCT_NATINFO_BIT = 10,
 	IPCT_NATINFO = (1 << IPCT_NATINFO_BIT),
+
+	/* Counter highest bit has been set */
+	IPCT_COUNTER_FILLING_BIT = 11,
+	IPCT_COUNTER_FILLING = (1 << IPCT_COUNTER_FILLING_BIT),
 };
 
 enum ip_conntrack_expect_events {
@@ -192,8 +196,8 @@
 
 struct ip_conntrack_counter
 {
-	u_int64_t packets;
-	u_int64_t bytes;
+	u_int32_t packets;
+	u_int32_t bytes;
 };
 
 struct ip_conntrack_helper;
diff --git a/include/linux/netfilter_ipv4/ip_conntrack_protocol.h b/include/linux/netfilter_ipv4/ip_conntrack_protocol.h
index b6b99be..2c76b87 100644
--- a/include/linux/netfilter_ipv4/ip_conntrack_protocol.h
+++ b/include/linux/netfilter_ipv4/ip_conntrack_protocol.h
@@ -52,6 +52,9 @@
 	int (*to_nfattr)(struct sk_buff *skb, struct nfattr *nfa,
 			 const struct ip_conntrack *ct);
 
+	/* convert nfnetlink attributes to protoinfo */
+	int (*from_nfattr)(struct nfattr *tb[], struct ip_conntrack *ct);
+
 	int (*tuple_to_nfattr)(struct sk_buff *skb,
 			       const struct ip_conntrack_tuple *t);
 	int (*nfattr_to_tuple)(struct nfattr *tb[],
diff --git a/include/linux/netfilter_ipv4/ip_conntrack_tuple.h b/include/linux/netfilter_ipv4/ip_conntrack_tuple.h
index 20e43f0..3232db1 100644
--- a/include/linux/netfilter_ipv4/ip_conntrack_tuple.h
+++ b/include/linux/netfilter_ipv4/ip_conntrack_tuple.h
@@ -1,6 +1,8 @@
 #ifndef _IP_CONNTRACK_TUPLE_H
 #define _IP_CONNTRACK_TUPLE_H
 
+#include <linux/types.h>
+
 /* A `tuple' is a structure containing the information to uniquely
   identify a connection.  ie. if two packets have the same tuple, they
   are in the same connection; if not, they are not.
diff --git a/include/linux/netfilter_ipv4/ip_nat.h b/include/linux/netfilter_ipv4/ip_nat.h
index e201ec6..41a107d 100644
--- a/include/linux/netfilter_ipv4/ip_nat.h
+++ b/include/linux/netfilter_ipv4/ip_nat.h
@@ -58,10 +58,6 @@
 struct ip_nat_info
 {
 	struct list_head bysource;
-
-	/* Helper (NULL if none). */
-	struct ip_nat_helper *helper;
-
 	struct ip_nat_seq seq[IP_CT_DIR_MAX];
 };
 
diff --git a/include/linux/netpoll.h b/include/linux/netpoll.h
index 5ade54a..ca5a873 100644
--- a/include/linux/netpoll.h
+++ b/include/linux/netpoll.h
@@ -86,7 +86,7 @@
 
 #else
 #define netpoll_rx(a) 0
-#define netpoll_poll_lock(a) 0
+#define netpoll_poll_lock(a) NULL
 #define netpoll_poll_unlock(a)
 #endif
 
diff --git a/include/net/inet_timewait_sock.h b/include/net/inet_timewait_sock.h
index 4ade56e..28f7b21 100644
--- a/include/net/inet_timewait_sock.h
+++ b/include/net/inet_timewait_sock.h
@@ -19,6 +19,7 @@
 
 #include <linux/ip.h>
 #include <linux/list.h>
+#include <linux/module.h>
 #include <linux/timer.h>
 #include <linux/types.h>
 #include <linux/workqueue.h>
@@ -193,11 +194,13 @@
 static inline void inet_twsk_put(struct inet_timewait_sock *tw)
 {
 	if (atomic_dec_and_test(&tw->tw_refcnt)) {
+		struct module *owner = tw->tw_prot->owner;
 #ifdef SOCK_REFCNT_DEBUG
 		printk(KERN_DEBUG "%s timewait_sock %p released\n",
 		       tw->tw_prot->name, tw);
 #endif
 		kmem_cache_free(tw->tw_prot->twsk_slab, tw);
+		module_put(owner);
 	}
 }
 
diff --git a/mm/fremap.c b/mm/fremap.c
index 3235fb7..ab23a06 100644
--- a/mm/fremap.c
+++ b/mm/fremap.c
@@ -89,6 +89,9 @@
 	size = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
 	if (!page->mapping || page->index >= size)
 		goto err_unlock;
+	err = -ENOMEM;
+	if (page_mapcount(page) > INT_MAX/2)
+		goto err_unlock;
 
 	zap_pte(mm, vma, addr, pte);
 
diff --git a/mm/madvise.c b/mm/madvise.c
index 4454936..20e075d 100644
--- a/mm/madvise.c
+++ b/mm/madvise.c
@@ -83,6 +83,9 @@
 {
 	struct file *file = vma->vm_file;
 
+	if (!file)
+		return -EBADF;
+
 	if (file->f_mapping->a_ops->get_xip_page) {
 		/* no bad return value, but ignore advice */
 		return 0;
@@ -141,11 +144,7 @@
 madvise_vma(struct vm_area_struct *vma, struct vm_area_struct **prev,
 		unsigned long start, unsigned long end, int behavior)
 {
-	struct file *filp = vma->vm_file;
-	long error = -EBADF;
-
-	if (!filp)
-		goto  out;
+	long error;
 
 	switch (behavior) {
 	case MADV_NORMAL:
@@ -166,8 +165,6 @@
 		error = -EINVAL;
 		break;
 	}
-		
-out:
 	return error;
 }
 
diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c
index 91bb895..defcf6a 100644
--- a/net/bridge/br_if.c
+++ b/net/bridge/br_if.c
@@ -79,7 +79,6 @@
 {
 	struct net_device *dev = p->dev;
 
-	dev->br_port = NULL;
 	p->br = NULL;
 	p->dev = NULL;
 	dev_put(dev);
@@ -100,6 +99,7 @@
 	struct net_bridge *br = p->br;
 	struct net_device *dev = p->dev;
 
+	dev->br_port = NULL;
 	dev_set_promiscuity(dev, -1);
 
 	spin_lock_bh(&br->lock);
diff --git a/net/bridge/netfilter/ebtables.c b/net/bridge/netfilter/ebtables.c
index c454014..f8ffbf6 100644
--- a/net/bridge/netfilter/ebtables.c
+++ b/net/bridge/netfilter/ebtables.c
@@ -26,6 +26,7 @@
 #include <linux/spinlock.h>
 #include <asm/uaccess.h>
 #include <linux/smp.h>
+#include <linux/cpumask.h>
 #include <net/sock.h>
 /* needed for logical [in,out]-dev filtering */
 #include "../br_private.h"
@@ -823,10 +824,11 @@
 		/* this will get free'd in do_replace()/ebt_register_table()
 		   if an error occurs */
 		newinfo->chainstack = (struct ebt_chainstack **)
-		   vmalloc(num_possible_cpus() * sizeof(struct ebt_chainstack));
+		   vmalloc((highest_possible_processor_id()+1) 
+				   		* sizeof(struct ebt_chainstack));
 		if (!newinfo->chainstack)
 			return -ENOMEM;
-		for (i = 0; i < num_possible_cpus(); i++) {
+		for_each_cpu(i) {
 			newinfo->chainstack[i] =
 			   vmalloc(udc_cnt * sizeof(struct ebt_chainstack));
 			if (!newinfo->chainstack[i]) {
@@ -895,9 +897,12 @@
 
 	/* counters of cpu 0 */
 	memcpy(counters, oldcounters,
-	   sizeof(struct ebt_counter) * nentries);
+	       sizeof(struct ebt_counter) * nentries);
+
 	/* add other counters to those of cpu 0 */
-	for (cpu = 1; cpu < num_possible_cpus(); cpu++) {
+	for_each_cpu(cpu) {
+		if (cpu == 0)
+			continue;
 		counter_base = COUNTER_BASE(oldcounters, nentries, cpu);
 		for (i = 0; i < nentries; i++) {
 			counters[i].pcnt += counter_base[i].pcnt;
@@ -929,7 +934,8 @@
 		BUGPRINT("Entries_size never zero\n");
 		return -EINVAL;
 	}
-	countersize = COUNTER_OFFSET(tmp.nentries) * num_possible_cpus();
+	countersize = COUNTER_OFFSET(tmp.nentries) * 
+					(highest_possible_processor_id()+1);
 	newinfo = (struct ebt_table_info *)
 	   vmalloc(sizeof(struct ebt_table_info) + countersize);
 	if (!newinfo)
@@ -1022,7 +1028,7 @@
 
 	vfree(table->entries);
 	if (table->chainstack) {
-		for (i = 0; i < num_possible_cpus(); i++)
+		for_each_cpu(i)
 			vfree(table->chainstack[i]);
 		vfree(table->chainstack);
 	}
@@ -1040,7 +1046,7 @@
 	vfree(counterstmp);
 	/* can be initialized in translate_table() */
 	if (newinfo->chainstack) {
-		for (i = 0; i < num_possible_cpus(); i++)
+		for_each_cpu(i)
 			vfree(newinfo->chainstack[i]);
 		vfree(newinfo->chainstack);
 	}
@@ -1132,7 +1138,8 @@
 		return -EINVAL;
 	}
 
-	countersize = COUNTER_OFFSET(table->table->nentries) * num_possible_cpus();
+	countersize = COUNTER_OFFSET(table->table->nentries) *
+					(highest_possible_processor_id()+1);
 	newinfo = (struct ebt_table_info *)
 	   vmalloc(sizeof(struct ebt_table_info) + countersize);
 	ret = -ENOMEM;
@@ -1186,7 +1193,7 @@
 	up(&ebt_mutex);
 free_chainstack:
 	if (newinfo->chainstack) {
-		for (i = 0; i < num_possible_cpus(); i++)
+		for_each_cpu(i)
 			vfree(newinfo->chainstack[i]);
 		vfree(newinfo->chainstack);
 	}
@@ -1209,7 +1216,7 @@
 	up(&ebt_mutex);
 	vfree(table->private->entries);
 	if (table->private->chainstack) {
-		for (i = 0; i < num_possible_cpus(); i++)
+		for_each_cpu(i)
 			vfree(table->private->chainstack[i]);
 		vfree(table->private->chainstack);
 	}
diff --git a/net/dccp/ccid.h b/net/dccp/ccid.h
index 21e5514..c37eeea 100644
--- a/net/dccp/ccid.h
+++ b/net/dccp/ccid.h
@@ -110,14 +110,14 @@
 
 static inline void ccid_hc_rx_exit(struct ccid *ccid, struct sock *sk)
 {
-	if (ccid->ccid_hc_rx_exit != NULL &&
+	if (ccid != NULL && ccid->ccid_hc_rx_exit != NULL &&
 	    dccp_sk(sk)->dccps_hc_rx_ccid_private != NULL)
 		ccid->ccid_hc_rx_exit(sk);
 }
 
 static inline void ccid_hc_tx_exit(struct ccid *ccid, struct sock *sk)
 {
-	if (ccid->ccid_hc_tx_exit != NULL &&
+	if (ccid != NULL && ccid->ccid_hc_tx_exit != NULL &&
 	    dccp_sk(sk)->dccps_hc_tx_ccid_private != NULL)
 		ccid->ccid_hc_tx_exit(sk);
 }
diff --git a/net/dccp/input.c b/net/dccp/input.c
index 1b6b2cb..3454d59 100644
--- a/net/dccp/input.c
+++ b/net/dccp/input.c
@@ -375,6 +375,9 @@
 	case DCCP_PKT_RESET:
 		inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK);
 		break;
+	case DCCP_PKT_DATA:
+		if (sk->sk_state == DCCP_RESPOND)
+			break;
 	case DCCP_PKT_DATAACK:
 	case DCCP_PKT_ACK:
 		/*
@@ -393,7 +396,8 @@
 		dccp_sk(sk)->dccps_osr = DCCP_SKB_CB(skb)->dccpd_seq;
 		dccp_set_state(sk, DCCP_OPEN);
 
-		if (dh->dccph_type == DCCP_PKT_DATAACK) {
+		if (dh->dccph_type == DCCP_PKT_DATAACK ||
+		    dh->dccph_type == DCCP_PKT_DATA) {
 			dccp_rcv_established(sk, skb, dh, len);
 			queued = 1; /* packet was queued
 				       (by dccp_rcv_established) */
diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c
index 1b5a09d..1b18ce6 100644
--- a/net/ipv4/esp4.c
+++ b/net/ipv4/esp4.c
@@ -5,6 +5,7 @@
 #include <net/esp.h>
 #include <asm/scatterlist.h>
 #include <linux/crypto.h>
+#include <linux/kernel.h>
 #include <linux/pfkeyv2.h>
 #include <linux/random.h>
 #include <net/icmp.h>
@@ -42,10 +43,10 @@
 	esp = x->data;
 	alen = esp->auth.icv_trunc_len;
 	tfm = esp->conf.tfm;
-	blksize = (crypto_tfm_alg_blocksize(tfm) + 3) & ~3;
-	clen = (clen + 2 + blksize-1)&~(blksize-1);
+	blksize = ALIGN(crypto_tfm_alg_blocksize(tfm), 4);
+	clen = ALIGN(clen + 2, blksize);
 	if (esp->conf.padlen)
-		clen = (clen + esp->conf.padlen-1)&~(esp->conf.padlen-1);
+		clen = ALIGN(clen, esp->conf.padlen);
 
 	if ((nfrags = skb_cow_data(skb, clen-skb->len+alen, &trailer)) < 0)
 		goto error;
@@ -143,7 +144,7 @@
 	struct ip_esp_hdr *esph;
 	struct esp_data *esp = x->data;
 	struct sk_buff *trailer;
-	int blksize = crypto_tfm_alg_blocksize(esp->conf.tfm);
+	int blksize = ALIGN(crypto_tfm_alg_blocksize(esp->conf.tfm), 4);
 	int alen = esp->auth.icv_trunc_len;
 	int elen = skb->len - sizeof(struct ip_esp_hdr) - esp->conf.ivlen - alen;
 	int nfrags;
@@ -304,16 +305,16 @@
 static u32 esp4_get_max_size(struct xfrm_state *x, int mtu)
 {
 	struct esp_data *esp = x->data;
-	u32 blksize = crypto_tfm_alg_blocksize(esp->conf.tfm);
+	u32 blksize = ALIGN(crypto_tfm_alg_blocksize(esp->conf.tfm), 4);
 
 	if (x->props.mode) {
-		mtu = (mtu + 2 + blksize-1)&~(blksize-1);
+		mtu = ALIGN(mtu + 2, blksize);
 	} else {
 		/* The worst case. */
-		mtu += 2 + blksize;
+		mtu = ALIGN(mtu + 2, 4) + blksize - 4;
 	}
 	if (esp->conf.padlen)
-		mtu = (mtu + esp->conf.padlen-1)&~(esp->conf.padlen-1);
+		mtu = ALIGN(mtu, esp->conf.padlen);
 
 	return mtu + x->props.header_len + esp->auth.icv_trunc_len;
 }
diff --git a/net/ipv4/inet_timewait_sock.c b/net/ipv4/inet_timewait_sock.c
index f9076ef..a010e9a 100644
--- a/net/ipv4/inet_timewait_sock.c
+++ b/net/ipv4/inet_timewait_sock.c
@@ -111,6 +111,7 @@
 		tw->tw_prot	    = sk->sk_prot_creator;
 		atomic_set(&tw->tw_refcnt, 1);
 		inet_twsk_dead_node_init(tw);
+		__module_get(tw->tw_prot->owner);
 	}
 
 	return tw;
diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig
index a765972..7d917e4 100644
--- a/net/ipv4/netfilter/Kconfig
+++ b/net/ipv4/netfilter/Kconfig
@@ -139,6 +139,7 @@
 
 config IP_NF_PPTP
 	tristate  'PPTP protocol support'
+	depends on IP_NF_CONNTRACK
 	help
 	  This module adds support for PPTP (Point to Point Tunnelling
 	  Protocol, RFC2637) connection tracking and NAT. 
@@ -498,9 +499,14 @@
 	  To compile it as a module, choose M here.  If unsure, say N.
 
 config IP_NF_TARGET_ULOG
-	tristate "ULOG target support"
+	tristate "ULOG target support (OBSOLETE)"
 	depends on IP_NF_IPTABLES
 	---help---
+
+	  This option enables the old IPv4-only "ipt_ULOG" implementation
+	  which has been obsoleted by the new "nfnetlink_log" code (see
+	  CONFIG_NETFILTER_NETLINK_LOG).
+
 	  This option adds a `ULOG' target, which allows you to create rules in
 	  any iptables table. The packet is passed to a userspace logging
 	  daemon using netlink multicast sockets; unlike the LOG target
diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c
index fa16342..a796928 100644
--- a/net/ipv4/netfilter/arp_tables.c
+++ b/net/ipv4/netfilter/arp_tables.c
@@ -716,8 +716,10 @@
 	}
 
 	/* And one copy for every other CPU */
-	for (i = 1; i < num_possible_cpus(); i++) {
-		memcpy(newinfo->entries + SMP_ALIGN(newinfo->size)*i,
+	for_each_cpu(i) {
+		if (i == 0)
+			continue;
+		memcpy(newinfo->entries + SMP_ALIGN(newinfo->size) * i,
 		       newinfo->entries,
 		       SMP_ALIGN(newinfo->size));
 	}
@@ -767,7 +769,7 @@
 	unsigned int cpu;
 	unsigned int i;
 
-	for (cpu = 0; cpu < num_possible_cpus(); cpu++) {
+	for_each_cpu(cpu) {
 		i = 0;
 		ARPT_ENTRY_ITERATE(t->entries + TABLE_OFFSET(t, cpu),
 				   t->size,
@@ -885,7 +887,8 @@
 		return -ENOMEM;
 
 	newinfo = vmalloc(sizeof(struct arpt_table_info)
-			  + SMP_ALIGN(tmp.size) * num_possible_cpus());
+			  + SMP_ALIGN(tmp.size) *
+			  		(highest_possible_processor_id()+1));
 	if (!newinfo)
 		return -ENOMEM;
 
@@ -1158,7 +1161,8 @@
 		= { 0, 0, 0, { 0 }, { 0 }, { } };
 
 	newinfo = vmalloc(sizeof(struct arpt_table_info)
-			  + SMP_ALIGN(repl->size) * num_possible_cpus());
+			  + SMP_ALIGN(repl->size) *
+			  		(highest_possible_processor_id()+1));
 	if (!newinfo) {
 		ret = -ENOMEM;
 		return ret;
diff --git a/net/ipv4/netfilter/ip_conntrack_core.c b/net/ipv4/netfilter/ip_conntrack_core.c
index ea65dd3..07a80b5 100644
--- a/net/ipv4/netfilter/ip_conntrack_core.c
+++ b/net/ipv4/netfilter/ip_conntrack_core.c
@@ -1119,7 +1119,7 @@
 			unsigned long extra_jiffies,
 			int do_acct)
 {
-	int do_event = 0;
+	int event = 0;
 
 	IP_NF_ASSERT(ct->timeout.data == (unsigned long)ct);
 	IP_NF_ASSERT(skb);
@@ -1129,13 +1129,13 @@
 	/* If not in hash table, timer will not be active yet */
 	if (!is_confirmed(ct)) {
 		ct->timeout.expires = extra_jiffies;
-		do_event = 1;
+		event = IPCT_REFRESH;
 	} else {
 		/* Need del_timer for race avoidance (may already be dying). */
 		if (del_timer(&ct->timeout)) {
 			ct->timeout.expires = jiffies + extra_jiffies;
 			add_timer(&ct->timeout);
-			do_event = 1;
+			event = IPCT_REFRESH;
 		}
 	}
 
@@ -1144,14 +1144,17 @@
 		ct->counters[CTINFO2DIR(ctinfo)].packets++;
 		ct->counters[CTINFO2DIR(ctinfo)].bytes += 
 						ntohs(skb->nh.iph->tot_len);
+		if ((ct->counters[CTINFO2DIR(ctinfo)].packets & 0x80000000)
+		    || (ct->counters[CTINFO2DIR(ctinfo)].bytes & 0x80000000))
+			event |= IPCT_COUNTER_FILLING;
 	}
 #endif
 
 	write_unlock_bh(&ip_conntrack_lock);
 
 	/* must be unlocked when calling event cache */
-	if (do_event)
-		ip_conntrack_event_cache(IPCT_REFRESH, skb);
+	if (event)
+		ip_conntrack_event_cache(event, skb);
 }
 
 #if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \
diff --git a/net/ipv4/netfilter/ip_conntrack_netlink.c b/net/ipv4/netfilter/ip_conntrack_netlink.c
index b08a432..166e606 100644
--- a/net/ipv4/netfilter/ip_conntrack_netlink.c
+++ b/net/ipv4/netfilter/ip_conntrack_netlink.c
@@ -177,11 +177,11 @@
 	struct nfattr *nest_count = NFA_NEST(skb, type);
 	u_int64_t tmp;
 
-	tmp = cpu_to_be64(ct->counters[dir].packets);
-	NFA_PUT(skb, CTA_COUNTERS_PACKETS, sizeof(u_int64_t), &tmp);
+	tmp = htonl(ct->counters[dir].packets);
+	NFA_PUT(skb, CTA_COUNTERS32_PACKETS, sizeof(u_int32_t), &tmp);
 
-	tmp = cpu_to_be64(ct->counters[dir].bytes);
-	NFA_PUT(skb, CTA_COUNTERS_BYTES, sizeof(u_int64_t), &tmp);
+	tmp = htonl(ct->counters[dir].bytes);
+	NFA_PUT(skb, CTA_COUNTERS32_BYTES, sizeof(u_int32_t), &tmp);
 
 	NFA_NEST_END(skb, nest_count);
 
@@ -833,7 +833,8 @@
 static inline int
 ctnetlink_change_status(struct ip_conntrack *ct, struct nfattr *cda[])
 {
-	unsigned long d, status = *(u_int32_t *)NFA_DATA(cda[CTA_STATUS-1]);
+	unsigned long d;
+	unsigned status = ntohl(*(u_int32_t *)NFA_DATA(cda[CTA_STATUS-1]));
 	d = ct->status ^ status;
 
 	if (d & (IPS_EXPECTED|IPS_CONFIRMED|IPS_DYING))
@@ -948,6 +949,31 @@
 	return 0;
 }
 
+static inline int
+ctnetlink_change_protoinfo(struct ip_conntrack *ct, struct nfattr *cda[])
+{
+	struct nfattr *tb[CTA_PROTOINFO_MAX], *attr = cda[CTA_PROTOINFO-1];
+	struct ip_conntrack_protocol *proto;
+	u_int16_t npt = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum;
+	int err = 0;
+
+	if (nfattr_parse_nested(tb, CTA_PROTOINFO_MAX, attr) < 0)
+		goto nfattr_failure;
+
+	proto = ip_conntrack_proto_find_get(npt);
+	if (!proto)
+		return -EINVAL;
+
+	if (proto->from_nfattr)
+		err = proto->from_nfattr(tb, ct);
+	ip_conntrack_proto_put(proto); 
+
+	return err;
+
+nfattr_failure:
+	return -ENOMEM;
+}
+
 static int
 ctnetlink_change_conntrack(struct ip_conntrack *ct, struct nfattr *cda[])
 {
@@ -973,6 +999,12 @@
 			return err;
 	}
 
+	if (cda[CTA_PROTOINFO-1]) {
+		err = ctnetlink_change_protoinfo(ct, cda);
+		if (err < 0)
+			return err;
+	}
+
 	DEBUGP("all done\n");
 	return 0;
 }
@@ -1002,6 +1034,12 @@
 	if (err < 0)
 		goto err;
 
+	if (cda[CTA_PROTOINFO-1]) {
+		err = ctnetlink_change_protoinfo(ct, cda);
+		if (err < 0)
+			return err;
+	}
+
 	ct->helper = ip_conntrack_helper_find_get(rtuple);
 
 	add_timer(&ct->timeout);
diff --git a/net/ipv4/netfilter/ip_conntrack_proto_icmp.c b/net/ipv4/netfilter/ip_conntrack_proto_icmp.c
index 838d1d6..98f0015 100644
--- a/net/ipv4/netfilter/ip_conntrack_proto_icmp.c
+++ b/net/ipv4/netfilter/ip_conntrack_proto_icmp.c
@@ -296,8 +296,7 @@
 				struct ip_conntrack_tuple *tuple)
 {
 	if (!tb[CTA_PROTO_ICMP_TYPE-1]
-	    || !tb[CTA_PROTO_ICMP_CODE-1]
-	    || !tb[CTA_PROTO_ICMP_ID-1])
+	    || !tb[CTA_PROTO_ICMP_CODE-1])
 		return -1;
 
 	tuple->dst.u.icmp.type = 
diff --git a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
index 121760d..d6701ca 100644
--- a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
+++ b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
@@ -341,17 +341,43 @@
 static int tcp_to_nfattr(struct sk_buff *skb, struct nfattr *nfa,
 			 const struct ip_conntrack *ct)
 {
+	struct nfattr *nest_parms = NFA_NEST(skb, CTA_PROTOINFO_TCP);
+	
 	read_lock_bh(&tcp_lock);
 	NFA_PUT(skb, CTA_PROTOINFO_TCP_STATE, sizeof(u_int8_t),
 		&ct->proto.tcp.state);
 	read_unlock_bh(&tcp_lock);
 
+	NFA_NEST_END(skb, nest_parms);
+
 	return 0;
 
 nfattr_failure:
 	read_unlock_bh(&tcp_lock);
 	return -1;
 }
+
+static int nfattr_to_tcp(struct nfattr *cda[], struct ip_conntrack *ct)
+{
+	struct nfattr *attr = cda[CTA_PROTOINFO_TCP-1];
+	struct nfattr *tb[CTA_PROTOINFO_TCP_MAX];
+
+        if (nfattr_parse_nested(tb, CTA_PROTOINFO_TCP_MAX, attr) < 0)
+                goto nfattr_failure;
+
+	if (!tb[CTA_PROTOINFO_TCP_STATE-1])
+		return -EINVAL;
+
+	write_lock_bh(&tcp_lock);
+	ct->proto.tcp.state = 
+		*(u_int8_t *)NFA_DATA(tb[CTA_PROTOINFO_TCP_STATE-1]);
+	write_unlock_bh(&tcp_lock);
+
+	return 0;
+
+nfattr_failure:
+	return -1;
+}
 #endif
 
 static unsigned int get_conntrack_index(const struct tcphdr *tcph)
@@ -1123,6 +1149,7 @@
 #if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \
     defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE)
 	.to_nfattr		= tcp_to_nfattr,
+	.from_nfattr		= nfattr_to_tcp,
 	.tuple_to_nfattr	= ip_ct_port_tuple_to_nfattr,
 	.nfattr_to_tuple	= ip_ct_port_nfattr_to_tuple,
 #endif
diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c
index eef99a1..75c27e9 100644
--- a/net/ipv4/netfilter/ip_tables.c
+++ b/net/ipv4/netfilter/ip_tables.c
@@ -27,6 +27,7 @@
 #include <asm/semaphore.h>
 #include <linux/proc_fs.h>
 #include <linux/err.h>
+#include <linux/cpumask.h>
 
 #include <linux/netfilter_ipv4/ip_tables.h>
 
@@ -921,8 +922,10 @@
 	}
 
 	/* And one copy for every other CPU */
-	for (i = 1; i < num_possible_cpus(); i++) {
-		memcpy(newinfo->entries + SMP_ALIGN(newinfo->size)*i,
+	for_each_cpu(i) {
+		if (i == 0)
+			continue;
+		memcpy(newinfo->entries + SMP_ALIGN(newinfo->size) * i,
 		       newinfo->entries,
 		       SMP_ALIGN(newinfo->size));
 	}
@@ -943,7 +946,7 @@
 		struct ipt_entry *table_base;
 		unsigned int i;
 
-		for (i = 0; i < num_possible_cpus(); i++) {
+		for_each_cpu(i) {
 			table_base =
 				(void *)newinfo->entries
 				+ TABLE_OFFSET(newinfo, i);
@@ -990,7 +993,7 @@
 	unsigned int cpu;
 	unsigned int i;
 
-	for (cpu = 0; cpu < num_possible_cpus(); cpu++) {
+	for_each_cpu(cpu) {
 		i = 0;
 		IPT_ENTRY_ITERATE(t->entries + TABLE_OFFSET(t, cpu),
 				  t->size,
@@ -1128,7 +1131,8 @@
 		return -ENOMEM;
 
 	newinfo = vmalloc(sizeof(struct ipt_table_info)
-			  + SMP_ALIGN(tmp.size) * num_possible_cpus());
+			  + SMP_ALIGN(tmp.size) * 
+			  	(highest_possible_processor_id()+1));
 	if (!newinfo)
 		return -ENOMEM;
 
@@ -1458,7 +1462,8 @@
 		= { 0, 0, 0, { 0 }, { 0 }, { } };
 
 	newinfo = vmalloc(sizeof(struct ipt_table_info)
-			  + SMP_ALIGN(repl->size) * num_possible_cpus());
+			  + SMP_ALIGN(repl->size) * 
+			  		(highest_possible_processor_id()+1));
 	if (!newinfo)
 		return -ENOMEM;
 
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 8225e42..7114031 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -435,7 +435,16 @@
 	int nsize, old_factor;
 	u16 flags;
 
-	BUG_ON(len >= skb->len);
+	if (unlikely(len >= skb->len)) {
+		if (net_ratelimit()) {
+			printk(KERN_DEBUG "TCP: seg_size=%u, mss=%u, seq=%u, "
+			       "end_seq=%u, skb->len=%u.\n", len, mss_now,
+			       TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq,
+			       skb->len);
+			WARN_ON(1);
+		}
+		return 0;
+	}
 
 	nsize = skb_headlen(skb) - len;
 	if (nsize < 0)
diff --git a/net/ipv6/esp6.c b/net/ipv6/esp6.c
index 9b27460..40d9a19 100644
--- a/net/ipv6/esp6.c
+++ b/net/ipv6/esp6.c
@@ -31,6 +31,7 @@
 #include <net/esp.h>
 #include <asm/scatterlist.h>
 #include <linux/crypto.h>
+#include <linux/kernel.h>
 #include <linux/pfkeyv2.h>
 #include <linux/random.h>
 #include <net/icmp.h>
@@ -66,10 +67,10 @@
 
 	alen = esp->auth.icv_trunc_len;
 	tfm = esp->conf.tfm;
-	blksize = (crypto_tfm_alg_blocksize(tfm) + 3) & ~3;
-	clen = (clen + 2 + blksize-1)&~(blksize-1);
+	blksize = ALIGN(crypto_tfm_alg_blocksize(tfm), 4);
+	clen = ALIGN(clen + 2, blksize);
 	if (esp->conf.padlen)
-		clen = (clen + esp->conf.padlen-1)&~(esp->conf.padlen-1);
+		clen = ALIGN(clen, esp->conf.padlen);
 
 	if ((nfrags = skb_cow_data(skb, clen-skb->len+alen, &trailer)) < 0) {
 		goto error;
@@ -133,7 +134,7 @@
 	struct ipv6_esp_hdr *esph;
 	struct esp_data *esp = x->data;
 	struct sk_buff *trailer;
-	int blksize = crypto_tfm_alg_blocksize(esp->conf.tfm);
+	int blksize = ALIGN(crypto_tfm_alg_blocksize(esp->conf.tfm), 4);
 	int alen = esp->auth.icv_trunc_len;
 	int elen = skb->len - sizeof(struct ipv6_esp_hdr) - esp->conf.ivlen - alen;
 
@@ -235,16 +236,17 @@
 static u32 esp6_get_max_size(struct xfrm_state *x, int mtu)
 {
 	struct esp_data *esp = x->data;
-	u32 blksize = crypto_tfm_alg_blocksize(esp->conf.tfm);
+	u32 blksize = ALIGN(crypto_tfm_alg_blocksize(esp->conf.tfm), 4);
 
 	if (x->props.mode) {
-		mtu = (mtu + 2 + blksize-1)&~(blksize-1);
+		mtu = ALIGN(mtu + 2, blksize);
 	} else {
 		/* The worst case. */
-		mtu += 2 + blksize;
+		u32 padsize = ((blksize - 1) & 7) + 1;
+		mtu = ALIGN(mtu + 2, padsize) + blksize - padsize;
 	}
 	if (esp->conf.padlen)
-		mtu = (mtu + esp->conf.padlen-1)&~(esp->conf.padlen-1);
+		mtu = ALIGN(mtu, esp->conf.padlen);
 
 	return mtu + x->props.header_len + esp->auth.icv_full_len;
 }
diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c
index 2da514b..b03e906 100644
--- a/net/ipv6/netfilter/ip6_tables.c
+++ b/net/ipv6/netfilter/ip6_tables.c
@@ -28,6 +28,7 @@
 #include <asm/uaccess.h>
 #include <asm/semaphore.h>
 #include <linux/proc_fs.h>
+#include <linux/cpumask.h>
 
 #include <linux/netfilter_ipv6/ip6_tables.h>
 
@@ -950,8 +951,10 @@
 	}
 
 	/* And one copy for every other CPU */
-	for (i = 1; i < num_possible_cpus(); i++) {
-		memcpy(newinfo->entries + SMP_ALIGN(newinfo->size)*i,
+	for_each_cpu(i) {
+		if (i == 0)
+			continue;
+		memcpy(newinfo->entries + SMP_ALIGN(newinfo->size) * i,
 		       newinfo->entries,
 		       SMP_ALIGN(newinfo->size));
 	}
@@ -973,6 +976,7 @@
 		unsigned int i;
 
 		for (i = 0; i < num_possible_cpus(); i++) {
+		for_each_cpu(i) {
 			table_base =
 				(void *)newinfo->entries
 				+ TABLE_OFFSET(newinfo, i);
@@ -1019,7 +1023,7 @@
 	unsigned int cpu;
 	unsigned int i;
 
-	for (cpu = 0; cpu < num_possible_cpus(); cpu++) {
+	for_each_cpu(cpu) {
 		i = 0;
 		IP6T_ENTRY_ITERATE(t->entries + TABLE_OFFSET(t, cpu),
 				  t->size,
@@ -1153,7 +1157,8 @@
 		return -ENOMEM;
 
 	newinfo = vmalloc(sizeof(struct ip6t_table_info)
-			  + SMP_ALIGN(tmp.size) * num_possible_cpus());
+			  + SMP_ALIGN(tmp.size) *
+			  		(highest_possible_processor_id()+1));
 	if (!newinfo)
 		return -ENOMEM;
 
@@ -1467,7 +1472,8 @@
 		= { 0, 0, 0, { 0 }, { 0 }, { } };
 
 	newinfo = vmalloc(sizeof(struct ip6t_table_info)
-			  + SMP_ALIGN(repl->size) * num_possible_cpus());
+			  + SMP_ALIGN(repl->size) *
+			  		(highest_possible_processor_id()+1));
 	if (!newinfo)
 		return -ENOMEM;
 
diff --git a/net/netfilter/nfnetlink.c b/net/netfilter/nfnetlink.c
index 1caaca0..4bc27a6 100644
--- a/net/netfilter/nfnetlink.c
+++ b/net/netfilter/nfnetlink.c
@@ -133,7 +133,7 @@
 	memset(tb, 0, sizeof(struct nfattr *) * maxattr);
 
 	while (NFA_OK(nfa, len)) {
-		unsigned flavor = nfa->nfa_type;
+		unsigned flavor = NFA_TYPE(nfa);
 		if (flavor && flavor <= maxattr)
 			tb[flavor-1] = nfa;
 		nfa = NFA_NEXT(nfa, len);
@@ -177,7 +177,7 @@
 		int attrlen = nlh->nlmsg_len - NLMSG_ALIGN(min_len);
 
 		while (NFA_OK(attr, attrlen)) {
-			unsigned flavor = attr->nfa_type;
+			unsigned flavor = NFA_TYPE(attr);
 			if (flavor) {
 				if (flavor > attr_count)
 					return -EINVAL;
diff --git a/net/sched/Kconfig b/net/sched/Kconfig
index 45d3bc0..81510da 100644
--- a/net/sched/Kconfig
+++ b/net/sched/Kconfig
@@ -72,9 +72,11 @@
 	  Choose this if you need a high resolution clock source but can't use
 	  the CPU's cycle counter.
 
+# don't allow on SMP x86 because they can have unsynchronized TSCs.
+# gettimeofday is a good alternative
 config NET_SCH_CLK_CPU
 	bool "CPU cycle counter"
-	depends on X86_TSC || X86_64 || ALPHA || SPARC64 || PPC64 || IA64
+	depends on ((X86_TSC || X86_64) && !SMP) || ALPHA || SPARC64 || PPC64 || IA64
 	help
 	  Say Y here if you want to use the CPU's cycle counter as clock source.
 	  This is a cheap and high resolution clock source, but on some