20 files changed, 798 insertions, 628 deletions

diff --git a/arch/arm/boot/compressed/.gitignore b/arch/arm/boot/compressed/.gitignore
index c6028967d33..0714e0334e3 100644
--- a/arch/arm/boot/compressed/.gitignore
+++ b/arch/arm/boot/compressed/.gitignore
@@ -1,7 +1,21 @@
+ashldi3.S
+bswapsdi2.S
 font.c
 lib1funcs.S
+hyp-stub.S
 piggy.gzip
 piggy.lzo
 piggy.lzma
+piggy.xzkern
+piggy.lz4
 vmlinux
 vmlinux.lds
+
+# borrowed libfdt files
+fdt.c
+fdt.h
+fdt_ro.c
+fdt_rw.c
+fdt_wip.c
+libfdt.h
+libfdt_internal.h
diff --git a/arch/arm/boot/compressed/Makefile b/arch/arm/boot/compressed/Makefile
index 0c74a6fab95..68c918362b7 100644
--- a/arch/arm/boot/compressed/Makefile
+++ b/arch/arm/boot/compressed/Makefile
@@ -24,7 +24,18 @@ endif
 AFLAGS_head.o += -DTEXT_OFFSET=$(TEXT_OFFSET)
 HEAD	= head.o
 OBJS	+= misc.o decompress.o
-FONTC	= $(srctree)/drivers/video/console/font_acorn_8x8.c
+ifeq ($(CONFIG_DEBUG_UNCOMPRESS),y)
+OBJS	+= debug.o
+endif
+FONTC	= $(srctree)/lib/fonts/font_acorn_8x8.c
+
+# string library code (-Os is enforced to keep it much smaller)
+OBJS		+= string.o
+CFLAGS_string.o	:= -Os
+
+ifeq ($(CONFIG_ARM_VIRT_EXT),y)
+OBJS		+= hyp-stub.o
+endif
 
 #
 # Architecture dependencies
@@ -33,23 +44,10 @@ ifeq ($(CONFIG_ARCH_ACORN),y)
 OBJS		+= ll_char_wr.o font.o
 endif
 
-ifeq ($(CONFIG_ARCH_SHARK),y)
-OBJS		+= head-shark.o ofw-shark.o
-endif
-
-ifeq ($(CONFIG_ARCH_P720T),y)
-# Borrow this code from SA1100
-OBJS		+= head-sa1100.o
-endif
-
 ifeq ($(CONFIG_ARCH_SA1100),y)
 OBJS		+= head-sa1100.o
 endif
 
-ifeq ($(CONFIG_ARCH_VT8500),y)
-OBJS		+= head-vt8500.o
-endif
-
 ifeq ($(CONFIG_CPU_XSCALE),y)
 OBJS		+= head-xscale.o
 endif
@@ -66,7 +64,7 @@ else
 endif
 endif
 
-ifeq ($(CONFIG_ARCH_SHMOBILE),y)
+ifeq ($(CONFIG_ARCH_SHMOBILE_LEGACY),y)
 OBJS		+= head-shmobile.o
 endif
 
@@ -88,22 +86,48 @@ SEDFLAGS	= s/TEXT_START/$(ZTEXTADDR)/;s/BSS_START/$(ZBSSADDR)/
 suffix_$(CONFIG_KERNEL_GZIP) = gzip
 suffix_$(CONFIG_KERNEL_LZO)  = lzo
 suffix_$(CONFIG_KERNEL_LZMA) = lzma
+suffix_$(CONFIG_KERNEL_XZ)   = xzkern
+suffix_$(CONFIG_KERNEL_LZ4)  = lz4
+
+# Borrowed libfdt files for the ATAG compatibility mode
+
+libfdt		:= fdt_rw.c fdt_ro.c fdt_wip.c fdt.c
+libfdt_hdrs	:= fdt.h libfdt.h libfdt_internal.h
+
+libfdt_objs	:= $(addsuffix .o, $(basename $(libfdt)))
+
+$(addprefix $(obj)/,$(libfdt) $(libfdt_hdrs)): $(obj)/%: $(srctree)/scripts/dtc/libfdt/%
+	$(call cmd,shipped)
+
+$(addprefix $(obj)/,$(libfdt_objs) atags_to_fdt.o): \
+	$(addprefix $(obj)/,$(libfdt_hdrs))
+
+ifeq ($(CONFIG_ARM_ATAG_DTB_COMPAT),y)
+OBJS	+= $(libfdt_objs) atags_to_fdt.o
+endif
 
 targets       := vmlinux vmlinux.lds \
 		 piggy.$(suffix_y) piggy.$(suffix_y).o \
-		 font.o font.c head.o misc.o $(OBJS)
+		 lib1funcs.o lib1funcs.S ashldi3.o ashldi3.S bswapsdi2.o \
+		 bswapsdi2.S font.o font.c head.o misc.o $(OBJS)
 
 # Make sure files are removed during clean
-extra-y       += piggy.gzip piggy.lzo piggy.lzma lib1funcs.S
+extra-y       += piggy.gzip piggy.lzo piggy.lzma piggy.xzkern piggy.lz4 \
+		 lib1funcs.S ashldi3.S bswapsdi2.S $(libfdt) $(libfdt_hdrs) \
+		 hyp-stub.S
 
 ifeq ($(CONFIG_FUNCTION_TRACER),y)
 ORIG_CFLAGS := $(KBUILD_CFLAGS)
 KBUILD_CFLAGS = $(subst -pg, , $(ORIG_CFLAGS))
 endif
 
-ccflags-y := -fpic -fno-builtin
-asflags-y := -Wa,-march=all
+ccflags-y := -fpic -mno-single-pic-base -fno-builtin -I$(obj)
+asflags-y := -DZIMAGE
 
+# Supply kernel BSS size to the decompressor via a linker symbol.
+KBSS_SZ = $(shell $(CROSS_COMPILE)size $(obj)/../../../../vmlinux | \
+		awk 'END{print $$3}')
+LDFLAGS_vmlinux = --defsym _kernel_bss_size=$(KBSS_SZ)
 # Supply ZRELADDR to the decompressor via a linker symbol.
 ifneq ($(CONFIG_AUTO_ZRELADDR),y)
 LDFLAGS_vmlinux += --defsym zreladdr=$(ZRELADDR)
@@ -123,7 +147,19 @@ LDFLAGS_vmlinux += -T
 # For __aeabi_uidivmod
 lib1funcs = $(obj)/lib1funcs.o
 
-$(obj)/lib1funcs.S: $(srctree)/arch/$(SRCARCH)/lib/lib1funcs.S FORCE
+$(obj)/lib1funcs.S: $(srctree)/arch/$(SRCARCH)/lib/lib1funcs.S
+	$(call cmd,shipped)
+
+# For __aeabi_llsl
+ashldi3 = $(obj)/ashldi3.o
+
+$(obj)/ashldi3.S: $(srctree)/arch/$(SRCARCH)/lib/ashldi3.S
+	$(call cmd,shipped)
+
+# For __bswapsi2, __bswapdi2
+bswapsdi2 = $(obj)/bswapsdi2.o
+
+$(obj)/bswapsdi2.S: $(srctree)/arch/$(SRCARCH)/lib/bswapsdi2.S
 	$(call cmd,shipped)
 
 # We need to prevent any GOTOFF relocs being used with references
@@ -139,8 +175,17 @@ bad_syms=$$($(CROSS_COMPILE)nm $@ | sed -n 's/^.\{8\} [bc] \(.*\)/\1/p') && \
   ( echo "following symbols must have non local/private scope:" >&2; \
     echo "$$bad_syms" >&2; rm -f $@; false )
 
+check_for_multiple_zreladdr = \
+if [ $(words $(ZRELADDR)) -gt 1 -a "$(CONFIG_AUTO_ZRELADDR)" = "" ]; then \
+	echo 'multiple zreladdrs: $(ZRELADDR)'; \
+	echo 'This needs CONFIG_AUTO_ZRELADDR to be set'; \
+	false; \
+fi
+
 $(obj)/vmlinux: $(obj)/vmlinux.lds $(obj)/$(HEAD) $(obj)/piggy.$(suffix_y).o \
-	 	$(addprefix $(obj)/, $(OBJS)) $(lib1funcs) FORCE
+		$(addprefix $(obj)/, $(OBJS)) $(lib1funcs) $(ashldi3) \
+		$(bswapsdi2) FORCE
+	@$(check_for_multiple_zreladdr)
 	$(call if_changed,ld)
 	@$(check_for_bad_syms)
 
@@ -156,3 +201,6 @@ $(obj)/font.c: $(FONTC)
 
 $(obj)/vmlinux.lds: $(obj)/vmlinux.lds.in arch/arm/boot/Makefile $(KCONFIG_CONFIG)
 	@sed "$(SEDFLAGS)" < $< > $@
+
+$(obj)/hyp-stub.S: $(srctree)/arch/$(SRCARCH)/kernel/hyp-stub.S
+	$(call cmd,shipped)
diff --git a/arch/arm/boot/compressed/atags_to_fdt.c b/arch/arm/boot/compressed/atags_to_fdt.c
new file mode 100644
index 00000000000..9448aa0c668
--- /dev/null
+++ b/arch/arm/boot/compressed/atags_to_fdt.c
@@ -0,0 +1,191 @@
+#include <asm/setup.h>
+#include <libfdt.h>
+
+#if defined(CONFIG_ARM_ATAG_DTB_COMPAT_CMDLINE_EXTEND)
+#define do_extend_cmdline 1
+#else
+#define do_extend_cmdline 0
+#endif
+
+#define NR_BANKS 16
+
+static int node_offset(void *fdt, const char *node_path)
+{
+	int offset = fdt_path_offset(fdt, node_path);
+	if (offset == -FDT_ERR_NOTFOUND)
+		offset = fdt_add_subnode(fdt, 0, node_path);
+	return offset;
+}
+
+static int setprop(void *fdt, const char *node_path, const char *property,
+		   uint32_t *val_array, int size)
+{
+	int offset = node_offset(fdt, node_path);
+	if (offset < 0)
+		return offset;
+	return fdt_setprop(fdt, offset, property, val_array, size);
+}
+
+static int setprop_string(void *fdt, const char *node_path,
+			  const char *property, const char *string)
+{
+	int offset = node_offset(fdt, node_path);
+	if (offset < 0)
+		return offset;
+	return fdt_setprop_string(fdt, offset, property, string);
+}
+
+static int setprop_cell(void *fdt, const char *node_path,
+			const char *property, uint32_t val)
+{
+	int offset = node_offset(fdt, node_path);
+	if (offset < 0)
+		return offset;
+	return fdt_setprop_cell(fdt, offset, property, val);
+}
+
+static const void *getprop(const void *fdt, const char *node_path,
+			   const char *property, int *len)
+{
+	int offset = fdt_path_offset(fdt, node_path);
+
+	if (offset == -FDT_ERR_NOTFOUND)
+		return NULL;
+
+	return fdt_getprop(fdt, offset, property, len);
+}
+
+static uint32_t get_cell_size(const void *fdt)
+{
+	int len;
+	uint32_t cell_size = 1;
+	const uint32_t *size_len =  getprop(fdt, "/", "#size-cells", &len);
+
+	if (size_len)
+		cell_size = fdt32_to_cpu(*size_len);
+	return cell_size;
+}
+
+static void merge_fdt_bootargs(void *fdt, const char *fdt_cmdline)
+{
+	char cmdline[COMMAND_LINE_SIZE];
+	const char *fdt_bootargs;
+	char *ptr = cmdline;
+	int len = 0;
+
+	/* copy the fdt command line into the buffer */
+	fdt_bootargs = getprop(fdt, "/chosen", "bootargs", &len);
+	if (fdt_bootargs)
+		if (len < COMMAND_LINE_SIZE) {
+			memcpy(ptr, fdt_bootargs, len);
+			/* len is the length of the string
+			 * including the NULL terminator */
+			ptr += len - 1;
+		}
+
+	/* and append the ATAG_CMDLINE */
+	if (fdt_cmdline) {
+		len = strlen(fdt_cmdline);
+		if (ptr - cmdline + len + 2 < COMMAND_LINE_SIZE) {
+			*ptr++ = ' ';
+			memcpy(ptr, fdt_cmdline, len);
+			ptr += len;
+		}
+	}
+	*ptr = '\0';
+
+	setprop_string(fdt, "/chosen", "bootargs", cmdline);
+}
+
+/*
+ * Convert and fold provided ATAGs into the provided FDT.
+ *
+ * REturn values:
+ *    = 0 -> pretend success
+ *    = 1 -> bad ATAG (may retry with another possible ATAG pointer)
+ *    < 0 -> error from libfdt
+ */
+int atags_to_fdt(void *atag_list, void *fdt, int total_space)
+{
+	struct tag *atag = atag_list;
+	/* In the case of 64 bits memory size, need to reserve 2 cells for
+	 * address and size for each bank */
+	uint32_t mem_reg_property[2 * 2 * NR_BANKS];
+	int memcount = 0;
+	int ret, memsize;
+
+	/* make sure we've got an aligned pointer */
+	if ((u32)atag_list & 0x3)
+		return 1;
+
+	/* if we get a DTB here we're done already */
+	if (*(u32 *)atag_list == fdt32_to_cpu(FDT_MAGIC))
+	       return 0;
+
+	/* validate the ATAG */
+	if (atag->hdr.tag != ATAG_CORE ||
+	    (atag->hdr.size != tag_size(tag_core) &&
+	     atag->hdr.size != 2))
+		return 1;
+
+	/* let's give it all the room it could need */
+	ret = fdt_open_into(fdt, fdt, total_space);
+	if (ret < 0)
+		return ret;
+
+	for_each_tag(atag, atag_list) {
+		if (atag->hdr.tag == ATAG_CMDLINE) {
+			/* Append the ATAGS command line to the device tree
+			 * command line.
+			 * NB: This means that if the same parameter is set in
+			 * the device tree and in the tags, the one from the
+			 * tags will be chosen.
+			 */
+			if (do_extend_cmdline)
+				merge_fdt_bootargs(fdt,
+						   atag->u.cmdline.cmdline);
+			else
+				setprop_string(fdt, "/chosen", "bootargs",
+					       atag->u.cmdline.cmdline);
+		} else if (atag->hdr.tag == ATAG_MEM) {
+			if (memcount >= sizeof(mem_reg_property)/4)
+				continue;
+			if (!atag->u.mem.size)
+				continue;
+			memsize = get_cell_size(fdt);
+
+			if (memsize == 2) {
+				/* if memsize is 2, that means that
+				 * each data needs 2 cells of 32 bits,
+				 * so the data are 64 bits */
+				uint64_t *mem_reg_prop64 =
+					(uint64_t *)mem_reg_property;
+				mem_reg_prop64[memcount++] =
+					cpu_to_fdt64(atag->u.mem.start);
+				mem_reg_prop64[memcount++] =
+					cpu_to_fdt64(atag->u.mem.size);
+			} else {
+				mem_reg_property[memcount++] =
+					cpu_to_fdt32(atag->u.mem.start);
+				mem_reg_property[memcount++] =
+					cpu_to_fdt32(atag->u.mem.size);
+			}
+
+		} else if (atag->hdr.tag == ATAG_INITRD2) {
+			uint32_t initrd_start, initrd_size;
+			initrd_start = atag->u.initrd.start;
+			initrd_size = atag->u.initrd.size;
+			setprop_cell(fdt, "/chosen", "linux,initrd-start",
+					initrd_start);
+			setprop_cell(fdt, "/chosen", "linux,initrd-end",
+					initrd_start + initrd_size);
+		}
+	}
+
+	if (memcount) {
+		setprop(fdt, "/memory", "reg", mem_reg_property,
+			4 * memcount * memsize);
+	}
+
+	return fdt_pack(fdt);
+}
diff --git a/arch/arm/boot/compressed/debug.S b/arch/arm/boot/compressed/debug.S
new file mode 100644
index 00000000000..5392ee63338
--- /dev/null
+++ b/arch/arm/boot/compressed/debug.S
@@ -0,0 +1,40 @@
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+
+#ifndef CONFIG_DEBUG_SEMIHOSTING
+
+#include CONFIG_DEBUG_LL_INCLUDE
+
+ENTRY(putc)
+	addruart r1, r2, r3
+	waituart r3, r1
+	senduart r0, r1
+	busyuart r3, r1
+	mov	 pc, lr
+ENDPROC(putc)
+
+#else
+
+ENTRY(putc)
+	adr	r1, 1f
+	ldmia	r1, {r2, r3}
+	add	r2, r2, r1
+	ldr	r1, [r2, r3]
+	strb	r0, [r1]
+	mov	r0, #0x03		@ SYS_WRITEC
+   ARM(	svc	#0x123456	)
+ THUMB(	svc	#0xab		)
+	mov	pc, lr
+	.align	2
+1:	.word	_GLOBAL_OFFSET_TABLE_ - .
+	.word	semi_writec_buf(GOT)
+ENDPROC(putc)
+
+	.bss
+	.global	semi_writec_buf
+	.type   semi_writec_buf, %object
+semi_writec_buf:
+	.space	4
+	.size	semi_writec_buf, 4
+
+#endif
diff --git a/arch/arm/boot/compressed/decompress.c b/arch/arm/boot/compressed/decompress.c
index 07be5a2f830..bd245d34952 100644
--- a/arch/arm/boot/compressed/decompress.c
+++ b/arch/arm/boot/compressed/decompress.c
@@ -13,8 +13,6 @@ extern void error(char *);
 #define STATIC static
 #define STATIC_RW_DATA	/* non-static please */
 
-#define ARCH_HAS_DECOMP_WDOG
-
 /* Diagnostic functions */
 #ifdef DEBUG
 #  define Assert(cond,msg) {if(!(cond)) error(msg);}
@@ -32,6 +30,9 @@ extern void error(char *);
 #  define Tracecv(c,x)
 #endif
 
+/* Not needed, but used in some headers pulled in by decompressors */
+extern char * strstr(const char * s1, const char *s2);
+
 #ifdef CONFIG_KERNEL_GZIP
 #include "../../../../lib/decompress_inflate.c"
 #endif
@@ -44,6 +45,16 @@ extern void error(char *);
 #include "../../../../lib/decompress_unlzma.c"
 #endif
 
+#ifdef CONFIG_KERNEL_XZ
+#define memmove memmove
+#define memcpy memcpy
+#include "../../../../lib/decompress_unxz.c"
+#endif
+
+#ifdef CONFIG_KERNEL_LZ4
+#include "../../../../lib/decompress_unlz4.c"
+#endif
+
 int do_decompress(u8 *input, int len, u8 *output, void (*error)(char *x))
 {
 	return decompress(input, len, NULL, NULL, output, NULL, error);
diff --git a/arch/arm/boot/compressed/head-sa1100.S b/arch/arm/boot/compressed/head-sa1100.S
index 6179d94dd5c..3115e313d9f 100644
--- a/arch/arm/boot/compressed/head-sa1100.S
+++ b/arch/arm/boot/compressed/head-sa1100.S
@@ -11,6 +11,7 @@
 #include <asm/mach-types.h>
 
 		.section        ".start", "ax"
+		.arch	armv4
 
 __SA1100_start:
 
diff --git a/arch/arm/boot/compressed/head-shark.S b/arch/arm/boot/compressed/head-shark.S
deleted file mode 100644
index 089c560e07f..00000000000
--- a/arch/arm/boot/compressed/head-shark.S
+++ /dev/null
@@ -1,139 +0,0 @@
-/* The head-file for the Shark
- * by Alexander Schulz
- *
- * Does the following:
- * - get the memory layout from firmware. This can only be done as long as the mmu
- *   is still on.
- * - switch the mmu off, so we have physical addresses
- * - copy the kernel to 0x08508000. This is done to have a fixed address where the
- *   C-parts (misc.c) are executed. This address must be known at compile-time,
- *   but the load-address of the kernel depends on how much memory is installed.
- * - Jump to this location.
- * - Set r8 with 0, r7 with the architecture ID for head.S
- */
-
-#include <linux/linkage.h>
-
-#include <asm/assembler.h>
-	
-		.section	".start", "ax"
-
-		b	__beginning
-	
-__ofw_data:	.long	0				@ the number of memory blocks
-		.space  128				@ (startaddr,size) ...
-		.space  128				@ bootargs
-		.align
-
-__beginning:	mov	r4, r0				@ save the entry to the firmware
-
-		mov	r0, #0xC0			@ disable irq and fiq
-		mov	r1, r0
-		mrs	r3, cpsr
-		bic	r2, r3, r0
-		eor	r2, r2, r1
-		msr	cpsr_c, r2
-
-		mov	r0, r4				@ get the Memory layout from firmware
-		adr	r1, __ofw_data
-		add	r2, r1, #4
-		mov	lr, pc
-		b	ofw_init
-		mov	r1, #0
-
-		adr	r2, __mmu_off			@ calculate physical address
-		sub	r2, r2, #0xf0000000		@ openprom maps us at f000 virt, 0e50 phys
-		adr	r0, __ofw_data
-		ldr	r0, [r0, #4]
-		add	r2, r2, r0
-		add	r2, r2, #0x00500000
-
-		mrc	p15, 0, r3, c1, c0
-		bic	r3, r3, #0xC			@ Write Buffer and DCache
-		bic	r3, r3, #0x1000			@ ICache
-		mcr	p15, 0, r3, c1, c0		@ disabled
-
-		mov	r0, #0
-		mcr	p15, 0, r0, c7, c7		@ flush I,D caches on v4
-		mcr	p15, 0, r0, c7, c10, 4		@ drain write buffer on v4
-		mcr	p15, 0, r0, c8, c7		@ flush I,D TLBs on v4
-
-		bic	r3, r3, #0x1			@ MMU
-		mcr	p15, 0, r3, c1, c0		@ disabled
-
-		mov	pc, r2
-
-__copy_target:	.long	0x08507FFC
-__copy_end:	.long	0x08607FFC
-		
-		.word	_start
-		.word	__bss_start
-
-		.align
-__temp_stack:	.space 128
-
-__mmu_off:
-		adr	r0, __ofw_data			@ read the 1. entry of the memory map
-		ldr	r0, [r0, #4]
-		orr	r0, r0, #0x00600000
-		sub	r0, r0, #4
-	
-		ldr	r1, __copy_end
-		ldr	r3, __copy_target
-
-/* r0 = 0x0e600000 (current end of kernelcode)
- * r3 = 0x08508000 (where it should begin)
- * r1 = 0x08608000 (end of copying area, 1MB)
- * The kernel is compressed, so 1 MB should be enough.
- * copy the kernel to the beginning of physical memory
- * We start from the highest address, so we can copy
- * from 0x08500000 to 0x08508000 if we have only 8MB
- */
-
-/* As we get more 2.6-kernels it gets more and more
- * uncomfortable to be bound to kernel images of 1MB only.
- * So we add a loop here, to be able to copy some more.
- * Alexander Schulz 2005-07-17
- */
-
-		mov	r4, #3				@ How many megabytes to copy
-
-
-__MoveCode:	sub	r4, r4, #1
-	
-__Copy:		ldr	r2, [r0], #-4
-		str	r2, [r1], #-4
-		teq	r1, r3
-		bne	__Copy
-
-		/* The firmware maps us in blocks of 1 MB, the next block is
-		   _below_ the last one. So our decrementing source pointer
-		   ist right here, but the destination pointer must be increased
-		   by 2 MB */
-		add	r1, r1, #0x00200000
-		add	r3, r3, #0x00100000
-
-		teq	r4, #0
-		bne	__MoveCode
-
-
-		/* and jump to it */
-		adr	r2, __go_on			@ where we want to jump
-		adr	r0, __ofw_data			@ read the 1. entry of the memory map
-		ldr	r0, [r0, #4]
-		sub	r2, r2, r0			@ we are mapped add 0e50 now, sub that (-0e00)
-		sub	r2, r2, #0x00500000		@ -0050
-		ldr	r0, __copy_target		@ and add 0850 8000 instead
-		add	r0, r0, #4
-		add	r2, r2, r0
-		mov	pc, r2				@ and jump there
-
-__go_on:
-		adr	sp, __temp_stack
-		add	sp, sp, #128
-		adr	r0, __ofw_data
-		mov	lr, pc
-		b	create_params
-	
-		mov	r8, #0
-		mov	r7, #15
diff --git a/arch/arm/boot/compressed/head-shmobile.S b/arch/arm/boot/compressed/head-shmobile.S
index fe3719b516f..e7f80928949 100644
--- a/arch/arm/boot/compressed/head-shmobile.S
+++ b/arch/arm/boot/compressed/head-shmobile.S
@@ -46,7 +46,7 @@ __image_start:
 __image_end:
 	.long	_got_end
 __load_base:
-	.long	CONFIG_MEMORY_START + 0x02000000 @ Load at 32Mb into SDRAM
+	.long	MEMORY_START + 0x02000000 @ Load at 32Mb into SDRAM
 __loaded:
 	.long	__continue
 	.align
@@ -55,29 +55,47 @@ __tmp_stack:
 __continue:
 #endif /* CONFIG_ZBOOT_ROM_MMC || CONFIG_ZBOOT_ROM_SH_MOBILE_SDHI */
 
-	b	1f
-__atags:@ tag #1
-	.long	12			@ tag->hdr.size = tag_size(tag_core);
-	.long	0x54410001		@ tag->hdr.tag = ATAG_CORE;
-	.long   0			@ tag->u.core.flags = 0;
-	.long	0			@ tag->u.core.pagesize = 0;
-	.long	0			@ tag->u.core.rootdev = 0;
-	@ tag #2
-	.long	8			@ tag->hdr.size = tag_size(tag_mem32);
-	.long	0x54410002		@ tag->hdr.tag = ATAG_MEM;
-	.long	CONFIG_MEMORY_SIZE	@ tag->u.mem.size = CONFIG_MEMORY_SIZE;
-	.long	CONFIG_MEMORY_START	@ @ tag->u.mem.start = CONFIG_MEMORY_START;
-	@ tag #3
-	.long	0			@ tag->hdr.size = 0
-	.long	0			@ tag->hdr.tag = ATAG_NONE;
-1:
+	adr	r0, dtb_info
+	ldmia	r0, {r1, r3, r4, r5, r7}
 
-	/* Set board ID necessary for boot */
-	ldr	r7, 1f				@ Set machine type register
-	adr	r8, __atags			@ Set atag register
+	sub	r0, r0, r1		@ calculate the delta offset
+	add	r5, r5, r0		@ _edata
+
+	ldr	lr, [r5, #0]		@ check if valid DTB is present
+	cmp	lr, r3
+	bne	0f
+
+	add	r9, r7, #31		@ rounded up to a multiple
+	bic	r9, r9, #31		@ ... of 32 bytes
+
+	add	r6, r9, r5		@ copy from _edata
+	add	r9, r9, r4		@ to MEMORY_START
+
+1:	ldmdb	r6!, {r0 - r3, r10 - r12, lr}
+	cmp	r6, r5
+	stmdb	r9!, {r0 - r3, r10 - r12, lr}
+	bhi	1b
+
+	/* Success: Zero board ID, pointer to start of memory for atag/dtb */
+	mov	r7, #0
+	mov	r8, r4
 	b	2f
 
-1 :	.long MACH_TYPE
+	.align	2
+dtb_info:
+	.word	dtb_info
+#ifndef __ARMEB__
+	.word	0xedfe0dd0		@ sig is 0xd00dfeed big endian
+#else
+	.word	0xd00dfeed
+#endif
+	.word	MEMORY_START
+	.word	_edata
+	.word	0x4000			@ maximum DTB size
+0:
+	/* Failure: Zero board ID, NULL atag/dtb */
+	mov 	r7, #0
+	mov	r8, #0			@ pass null pointer as atag
 2 :
 
 #endif /* CONFIG_ZBOOT_ROM */
diff --git a/arch/arm/boot/compressed/head-vt8500.S b/arch/arm/boot/compressed/head-vt8500.S
deleted file mode 100644
index 1dc1e21a3be..00000000000
--- a/arch/arm/boot/compressed/head-vt8500.S
+++ /dev/null
@@ -1,46 +0,0 @@
-/*
- * linux/arch/arm/boot/compressed/head-vt8500.S
- *
- * Copyright (C) 2010 Alexey Charkov <alchark@gmail.com>
- *
- * VIA VT8500 specific tweaks. This is merged into head.S by the linker.
- *
- */
-
-#include <linux/linkage.h>
-#include <asm/mach-types.h>
-
-		.section        ".start", "ax"
-
-__VT8500_start:
-	@ Compare the SCC ID register against a list of known values
-	ldr	r1, .SCCID
-	ldr	r3, [r1]
-
-	@ VT8500 override
-	ldr	r4, .VT8500SCC
-	cmp	r3, r4
-	ldreq	r7, .ID_BV07
-	beq	.Lendvt8500
-
-	@ WM8505 override
-	ldr	r4, .WM8505SCC
-	cmp	r3, r4
-	ldreq	r7, .ID_8505
-	beq	.Lendvt8500
-
-	@ Otherwise, leave the bootloader's machine id untouched
-
-.SCCID:
-	.word	0xd8120000
-.VT8500SCC:
-	.word	0x34000102
-.WM8505SCC:
-	.word	0x34260103
-
-.ID_BV07:
-	.word	MACH_TYPE_BV07
-.ID_8505:
-	.word	MACH_TYPE_WM8505_7IN_NETBOOK
-
-.Lendvt8500:
diff --git a/arch/arm/boot/compressed/head-xscale.S b/arch/arm/boot/compressed/head-xscale.S
index aa5ee49c5c5..6ab0599c02d 100644
--- a/arch/arm/boot/compressed/head-xscale.S
+++ b/arch/arm/boot/compressed/head-xscale.S
@@ -32,10 +32,3 @@ __XScale_start:
 		bic	r0, r0, #0x1000		@ clear Icache
 		mcr	p15, 0, r0, c1, c0, 0
 
-#ifdef CONFIG_ARCH_IXP2000
-		mov	r1, #-1
-		mov	r0, #0xd6000000
-		str	r1, [r0, #0x14]
-		str	r1, [r0, #0x18]
-#endif
-
diff --git a/arch/arm/boot/compressed/head.S b/arch/arm/boot/compressed/head.S
index e95a5989602..3a8b32df6b3 100644
--- a/arch/arm/boot/compressed/head.S
+++ b/arch/arm/boot/compressed/head.S
@@ -9,7 +9,9 @@
  * published by the Free Software Foundation.
  */
 #include <linux/linkage.h>
+#include <asm/assembler.h>
 
+	.arch	armv7-a
 /*
  * Debugging stuff
  *
@@ -43,7 +45,7 @@
 
 #else
 
-#include <mach/debug-macro.S>
+#include CONFIG_DEBUG_LL_INCLUDE
 
 		.macro	writeb,	ch, rb
 		senduart \ch, \rb
@@ -58,11 +60,6 @@
 		add	\rb, \rb, #0x00010000	@ Ser1
 #endif
 		.endm
-#elif defined(CONFIG_ARCH_S3C2410)
-		.macro loadsp, rb, tmp
-		mov	\rb, #0x50000000
-		add	\rb, \rb, #0x4000 * CONFIG_S3C_LOWLEVEL_UART_PORT
-		.endm
 #else
 		.macro	loadsp,	rb, tmp
 		addruart \rb, \tmp
@@ -132,10 +129,15 @@ start:
 		.word	start			@ absolute load/run zImage address
 		.word	_edata			@ zImage end address
  THUMB(		.thumb			)
-1:		mov	r7, r1			@ save architecture ID
+1:
+ ARM_BE8(	setend	be )			@ go BE8 if compiled for BE8
+		mrs	r9, cpsr
+#ifdef CONFIG_ARM_VIRT_EXT
+		bl	__hyp_stub_install	@ get into SVC mode, reversibly
+#endif
+		mov	r7, r1			@ save architecture ID
 		mov	r8, r2			@ save atags pointer
 
-#ifndef __ARM_ARCH_2__
 		/*
 		 * Booting from Angel - need to enter SVC mode and disable
 		 * FIQs/IRQs (numeric definitions from angel arm.h source).
@@ -148,13 +150,9 @@ start:
  ARM(		swi	0x123456	)	@ angel_SWI_ARM
  THUMB(		svc	0xab		)	@ angel_SWI_THUMB
 not_angel:
-		mrs	r2, cpsr		@ turn off interrupts to
-		orr	r2, r2, #0xc0		@ prevent angel from running
-		msr	cpsr_c, r2
-#else
-		teqp	pc, #0x0c000003		@ turn off interrupts
-#endif
-
+		safe_svcmode_maskall r0
+		msr	spsr_cxsf, r9		@ Save the CPU boot mode in
+						@ SPSR
 		/*
 		 * Note that some cache flushing and other stuff may
 		 * be needed here - is there an Angel SWI call for this?
@@ -176,7 +174,19 @@ not_angel:
 		ldr	r4, =zreladdr
 #endif
 
-		bl	cache_on
+		/*
+		 * Set up a page table only if it won't overwrite ourself.
+		 * That means r4 < pc && r4 - 16k page directory > &_end.
+		 * Given that r4 > &_end is most unfrequent, we add a rough
+		 * additional 1MB of room for a possible appended DTB.
+		 */
+		mov	r0, pc
+		cmp	r0, r4
+		ldrcc	r0, LC0+32
+		addcc	r0, r0, pc
+		cmpcc	r4, r0
+		orrcc	r4, r4, #1		@ remember we skipped cache_on
+		blcs	cache_on
 
 restart:	adr	r0, LC0
 		ldmia	r0, {r1, r2, r3, r6, r10, r11, r12}
@@ -216,22 +226,121 @@ restart:	adr	r0, LC0
 		mov	r10, r6
 #endif
 
+		mov	r5, #0			@ init dtb size to 0
+#ifdef CONFIG_ARM_APPENDED_DTB
+/*
+ *   r0  = delta
+ *   r2  = BSS start
+ *   r3  = BSS end
+ *   r4  = final kernel address (possibly with LSB set)
+ *   r5  = appended dtb size (still unknown)
+ *   r6  = _edata
+ *   r7  = architecture ID
+ *   r8  = atags/device tree pointer
+ *   r9  = size of decompressed image
+ *   r10 = end of this image, including  bss/stack/malloc space if non XIP
+ *   r11 = GOT start
+ *   r12 = GOT end
+ *   sp  = stack pointer
+ *
+ * if there are device trees (dtb) appended to zImage, advance r10 so that the
+ * dtb data will get relocated along with the kernel if necessary.
+ */
+
+		ldr	lr, [r6, #0]
+#ifndef __ARMEB__
+		ldr	r1, =0xedfe0dd0		@ sig is 0xd00dfeed big endian
+#else
+		ldr	r1, =0xd00dfeed
+#endif
+		cmp	lr, r1
+		bne	dtb_check_done		@ not found
+
+#ifdef CONFIG_ARM_ATAG_DTB_COMPAT
+		/*
+		 * OK... Let's do some funky business here.
+		 * If we do have a DTB appended to zImage, and we do have
+		 * an ATAG list around, we want the later to be translated
+		 * and folded into the former here.  To be on the safe side,
+		 * let's temporarily move  the stack away into the malloc
+		 * area.  No GOT fixup has occurred yet, but none of the
+		 * code we're about to call uses any global variable.
+		*/
+		add	sp, sp, #0x10000
+		stmfd	sp!, {r0-r3, ip, lr}
+		mov	r0, r8
+		mov	r1, r6
+		sub	r2, sp, r6
+		bl	atags_to_fdt
+
+		/*
+		 * If returned value is 1, there is no ATAG at the location
+		 * pointed by r8.  Try the typical 0x100 offset from start
+		 * of RAM and hope for the best.
+		 */
+		cmp	r0, #1
+		sub	r0, r4, #TEXT_OFFSET
+		bic	r0, r0, #1
+		add	r0, r0, #0x100
+		mov	r1, r6
+		sub	r2, sp, r6
+		bleq	atags_to_fdt
+
+		ldmfd	sp!, {r0-r3, ip, lr}
+		sub	sp, sp, #0x10000
+#endif
+
+		mov	r8, r6			@ use the appended device tree
+
+		/*
+		 * Make sure that the DTB doesn't end up in the final
+		 * kernel's .bss area. To do so, we adjust the decompressed
+		 * kernel size to compensate if that .bss size is larger
+		 * than the relocated code.
+		 */
+		ldr	r5, =_kernel_bss_size
+		adr	r1, wont_overwrite
+		sub	r1, r6, r1
+		subs	r1, r5, r1
+		addhi	r9, r9, r1
+
+		/* Get the dtb's size */
+		ldr	r5, [r6, #4]
+#ifndef __ARMEB__
+		/* convert r5 (dtb size) to little endian */
+		eor	r1, r5, r5, ror #16
+		bic	r1, r1, #0x00ff0000
+		mov	r5, r5, ror #8
+		eor	r5, r5, r1, lsr #8
+#endif
+
+		/* preserve 64-bit alignment */
+		add	r5, r5, #7
+		bic	r5, r5, #7
+
+		/* relocate some pointers past the appended dtb */
+		add	r6, r6, r5
+		add	r10, r10, r5
+		add	sp, sp, r5
+dtb_check_done:
+#endif
+
 /*
  * Check to see if we will overwrite ourselves.
- *   r4  = final kernel address
+ *   r4  = final kernel address (possibly with LSB set)
  *   r9  = size of decompressed image
  *   r10 = end of this image, including  bss/stack/malloc space if non XIP
  * We basically want:
  *   r4 - 16k page directory >= r10 -> OK
- *   r4 + image length <= current position (pc) -> OK
+ *   r4 + image length <= address of wont_overwrite -> OK
+ * Note: the possible LSB in r4 is harmless here.
  */
 		add	r10, r10, #16384
 		cmp	r4, r10
 		bhs	wont_overwrite
 		add	r10, r4, r9
-   ARM(		cmp	r10, pc		)
- THUMB(		mov	lr, pc		)
- THUMB(		cmp	r10, lr		)
+		adr	r9, wont_overwrite
+		cmp	r10, r9
 		bls	wont_overwrite
 
 /*
@@ -253,6 +362,20 @@ restart:	adr	r0, LC0
 		adr	r5, restart
 		bic	r5, r5, #31
 
+/* Relocate the hyp vector base if necessary */
+#ifdef CONFIG_ARM_VIRT_EXT
+		mrs	r0, spsr
+		and	r0, r0, #MODE_MASK
+		cmp	r0, #HYP_MODE
+		bne	1f
+
+		bl	__hyp_get_vectors
+		sub	r0, r0, r5
+		add	r0, r0, r10
+		bl	__hyp_set_vectors
+1:
+#endif
+
 		sub	r9, r6, r5		@ size to copy
 		add	r9, r9, #31		@ rounded up to a multiple
 		bic	r9, r9, #31		@ ... of 32 bytes
@@ -272,7 +395,8 @@ restart:	adr	r0, LC0
 		add	sp, sp, r6
 #endif
 
-		bl	cache_clean_flush
+		tst	r4, #1
+		bleq	cache_clean_flush
 
 		adr	r0, BSYM(restart)
 		add	r0, r0, r6
@@ -284,15 +408,17 @@ wont_overwrite:
  *   r0  = delta
  *   r2  = BSS start
  *   r3  = BSS end
- *   r4  = kernel execution address
+ *   r4  = kernel execution address (possibly with LSB set)
+ *   r5  = appended dtb size (0 if not present)
  *   r7  = architecture ID
  *   r8  = atags pointer
  *   r11 = GOT start
  *   r12 = GOT end
  *   sp  = stack pointer
  */
-		teq	r0, #0
+		orrs	r1, r0, r5
 		beq	not_relocated
+
 		add	r11, r11, r0
 		add	r12, r12, r0
 
@@ -307,12 +433,21 @@ wont_overwrite:
 
 		/*
 		 * Relocate all entries in the GOT table.
+		 * Bump bss entries to _edata + dtb size
 		 */
 1:		ldr	r1, [r11, #0]		@ relocate entries in the GOT
-		add	r1, r1, r0		@ table.  This fixes up the
-		str	r1, [r11], #4		@ C references.
+		add	r1, r1, r0		@ This fixes up C references
+		cmp	r1, r2			@ if entry >= bss_start &&
+		cmphs	r3, r1			@       bss_end > entry
+		addhi	r1, r1, r5		@    entry += dtb size
+		str	r1, [r11], #4		@ next entry
 		cmp	r11, r12
 		blo	1b
+
+		/* bump our bss pointers too */
+		add	r2, r2, r5
+		add	r3, r3, r5
+
 #else
 
 		/*
@@ -336,6 +471,15 @@ not_relocated:	mov	r0, #0
 		cmp	r2, r3
 		blo	1b
 
+		/*
+		 * Did we skip the cache setup earlier?
+		 * That is indicated by the LSB in r4.
+		 * Do it now if so.
+		 */
+		tst	r4, #1
+		bic	r4, r4, #1
+		blne	cache_on
+
 /*
  * The C runtime environment should now be setup sufficiently.
  * Set up some pointers, and start decompressing.
@@ -350,11 +494,29 @@ not_relocated:	mov	r0, #0
 		bl	decompress_kernel
 		bl	cache_clean_flush
 		bl	cache_off
-		mov	r0, #0			@ must be zero
 		mov	r1, r7			@ restore architecture number
 		mov	r2, r8			@ restore atags pointer
- ARM(		mov	pc, r4	)		@ call kernel
- THUMB(		bx	r4	)		@ entry point is always ARM
+
+#ifdef CONFIG_ARM_VIRT_EXT
+		mrs	r0, spsr		@ Get saved CPU boot mode
+		and	r0, r0, #MODE_MASK
+		cmp	r0, #HYP_MODE		@ if not booted in HYP mode...
+		bne	__enter_kernel		@ boot kernel directly
+
+		adr	r12, .L__hyp_reentry_vectors_offset
+		ldr	r0, [r12]
+		add	r0, r0, r12
+
+		bl	__hyp_set_vectors
+		__HVC(0)			@ otherwise bounce to hyp mode
+
+		b	.			@ should never be reached
+
+		.align	2
+.L__hyp_reentry_vectors_offset:	.long	__hyp_reentry_vectors - .
+#else
+		b	__enter_kernel
+#endif
 
 		.align	2
 		.type	LC0, #object
@@ -366,6 +528,7 @@ LC0:		.word	LC0			@ r1
 		.word	_got_start		@ r11
 		.word	_got_end		@ ip
 		.word	.L_user_stack_end	@ sp
+		.word	_end - restart + 16384 + 1024*1024
 		.size	LC0, . - LC0
 
 #ifdef CONFIG_ARCH_RPC
@@ -459,6 +622,12 @@ __armv3_mpu_cache_on:
 		mcr	p15, 0, r0, c7, c0, 0	@ invalidate whole cache v3
 		mov	pc, lr
 
+#ifdef CONFIG_CPU_DCACHE_WRITETHROUGH
+#define CB_BITS 0x08
+#else
+#define CB_BITS 0x0c
+#endif
+
 __setup_mmu:	sub	r3, r4, #16384		@ Page directory size
 		bic	r3, r3, #0xff		@ Align the pointer
 		bic	r3, r3, #0x3f00
@@ -470,17 +639,14 @@ __setup_mmu:	sub	r3, r4, #16384		@ Page directory size
 		mov	r9, r0, lsr #18
 		mov	r9, r9, lsl #18		@ start of RAM
 		add	r10, r9, #0x10000000	@ a reasonable RAM size
-		mov	r1, #0x12
-		orr	r1, r1, #3 << 10
+		mov	r1, #0x12		@ XN|U + section mapping
+		orr	r1, r1, #3 << 10	@ AP=11
 		add	r2, r3, #16384
 1:		cmp	r1, r9			@ if virt > start of RAM
-#ifdef CONFIG_CPU_DCACHE_WRITETHROUGH
-		orrhs	r1, r1, #0x08		@ set cacheable
-#else
-		orrhs	r1, r1, #0x0c		@ set cacheable, bufferable
-#endif
-		cmp	r1, r10			@ if virt > end of RAM
-		bichs	r1, r1, #0x0c		@ clear cacheable, bufferable
+		cmphs	r10, r1			@   && end of RAM > virt
+		bic	r1, r1, #0x1c		@ clear XN|U + C + B
+		orrlo	r1, r1, #0x10		@ Set XN|U for non-RAM
+		orrhs	r1, r1, r6		@ set RAM section settings
 		str	r1, [r0], #4		@ 1:1 mapping
 		add	r1, r1, #1048576
 		teq	r0, r2
@@ -491,7 +657,7 @@ __setup_mmu:	sub	r3, r4, #16384		@ Page directory size
  * so there is no map overlap problem for up to 1 MB compressed kernel.
  * If the execution is in RAM then we would only be duplicating the above.
  */
-		mov	r1, #0x1e
+		orr	r1, r6, #0x04		@ ensure B is set for this
 		orr	r1, r1, #3 << 10
 		mov	r2, pc
 		mov	r2, r2, lsr #20
@@ -503,6 +669,15 @@ __setup_mmu:	sub	r3, r4, #16384		@ Page directory size
 		mov	pc, lr
 ENDPROC(__setup_mmu)
 
+@ Enable unaligned access on v6, to allow better code generation
+@ for the decompressor C code:
+__armv6_mmu_cache_on:
+		mrc	p15, 0, r0, c1, c0, 0	@ read SCTLR
+		bic	r0, r0, #2		@ A (no unaligned access fault)
+		orr	r0, r0, #1 << 22	@ U (v6 unaligned access model)
+		mcr	p15, 0, r0, c1, c0, 0	@ write SCTLR
+		b	__armv4_mmu_cache_on
+
 __arm926ejs_mmu_cache_on:
 #ifdef CONFIG_CPU_DCACHE_WRITETHROUGH
 		mov	r0, #4			@ put dcache in WT mode
@@ -512,6 +687,7 @@ __arm926ejs_mmu_cache_on:
 __armv4_mmu_cache_on:
 		mov	r12, lr
 #ifdef CONFIG_MMU
+		mov	r6, #CB_BITS | 0x12	@ U
 		bl	__setup_mmu
 		mov	r0, #0
 		mcr	p15, 0, r0, c7, c10, 4	@ drain write buffer
@@ -519,9 +695,7 @@ __armv4_mmu_cache_on:
 		mrc	p15, 0, r0, c1, c0, 0	@ read control reg
 		orr	r0, r0, #0x5000		@ I-cache enable, RR cache replacement
 		orr	r0, r0, #0x0030
-#ifdef CONFIG_CPU_ENDIAN_BE8
-		orr	r0, r0, #1 << 25	@ big-endian page tables
-#endif
+ ARM_BE8(	orr	r0, r0, #1 << 25 )	@ big-endian page tables
 		bl	__common_mmu_cache_on
 		mov	r0, #0
 		mcr	p15, 0, r0, c8, c7, 0	@ flush I,D TLBs
@@ -533,6 +707,7 @@ __armv7_mmu_cache_on:
 #ifdef CONFIG_MMU
 		mrc	p15, 0, r11, c0, c1, 4	@ read ID_MMFR0
 		tst	r11, #0xf		@ VMSA
+		movne	r6, #CB_BITS | 0x02	@ !XN
 		blne	__setup_mmu
 		mov	r0, #0
 		mcr	p15, 0, r0, c7, c10, 4	@ drain write buffer
@@ -540,17 +715,24 @@ __armv7_mmu_cache_on:
 		mcrne	p15, 0, r0, c8, c7, 0	@ flush I,D TLBs
 #endif
 		mrc	p15, 0, r0, c1, c0, 0	@ read control reg
+		bic	r0, r0, #1 << 28	@ clear SCTLR.TRE
 		orr	r0, r0, #0x5000		@ I-cache enable, RR cache replacement
 		orr	r0, r0, #0x003c		@ write buffer
+		bic	r0, r0, #2		@ A (no unaligned access fault)
+		orr	r0, r0, #1 << 22	@ U (v6 unaligned access model)
+						@ (needed for ARM1176)
 #ifdef CONFIG_MMU
-#ifdef CONFIG_CPU_ENDIAN_BE8
-		orr	r0, r0, #1 << 25	@ big-endian page tables
-#endif
+ ARM_BE8(	orr	r0, r0, #1 << 25 )	@ big-endian page tables
+		mrcne   p15, 0, r6, c2, c0, 2   @ read ttb control reg
 		orrne	r0, r0, #1		@ MMU enabled
-		movne	r1, #-1
+		movne	r1, #0xfffffffd		@ domain 0 = client
+		bic     r6, r6, #1 << 31        @ 32-bit translation system
+		bic     r6, r6, #3 << 0         @ use only ttbr0
 		mcrne	p15, 0, r3, c2, c0, 0	@ load page table pointer
 		mcrne	p15, 0, r1, c3, c0, 0	@ load domain access control
+		mcrne   p15, 0, r6, c2, c0, 2   @ load ttb control
 #endif
+		mcr	p15, 0, r0, c7, c5, 4	@ ISB
 		mcr	p15, 0, r0, c1, c0, 0	@ load control register
 		mrc	p15, 0, r0, c1, c0, 0	@ and read it back
 		mov	r0, #0
@@ -559,6 +741,7 @@ __armv7_mmu_cache_on:
 
 __fa526_cache_on:
 		mov	r12, lr
+		mov	r6, #CB_BITS | 0x12	@ U
 		bl	__setup_mmu
 		mov	r0, #0
 		mcr	p15, 0, r0, c7, c7, 0	@ Invalidate whole cache
@@ -571,18 +754,6 @@ __fa526_cache_on:
 		mcr	p15, 0, r0, c8, c7, 0	@ flush UTLB
 		mov	pc, r12
 
-__arm6_mmu_cache_on:
-		mov	r12, lr
-		bl	__setup_mmu
-		mov	r0, #0
-		mcr	p15, 0, r0, c7, c0, 0	@ invalidate whole cache v3
-		mcr	p15, 0, r0, c5, c0, 0	@ invalidate whole TLB v3
-		mov	r0, #0x30
-		bl	__common_mmu_cache_on
-		mov	r0, #0
-		mcr	p15, 0, r0, c5, c0, 0	@ invalidate whole TLB v3
-		mov	pc, r12
-
 __common_mmu_cache_on:
 #ifndef CONFIG_THUMB2_KERNEL
 #ifndef DEBUG
@@ -647,18 +818,8 @@ call_cache_fn:	adr	r12, proc_types
 		.align	2
 		.type	proc_types,#object
 proc_types:
-		.word	0x41560600		@ ARM6/610
-		.word	0xffffffe0
-		W(b)	__arm6_mmu_cache_off	@ works, but slow
-		W(b)	__arm6_mmu_cache_off
-		mov	pc, lr
- THUMB(		nop				)
-@		b	__arm6_mmu_cache_on		@ untested
-@		b	__arm6_mmu_cache_off
-@		b	__armv3_mmu_cache_flush
-
-		.word	0x00000000		@ old ARM ID
-		.word	0x0000f000
+		.word	0x41000000		@ old ARM ID
+		.word	0xff00f000
 		mov	pc, lr
  THUMB(		nop				)
 		mov	pc, lr
@@ -668,8 +829,10 @@ proc_types:
 
 		.word	0x41007000		@ ARM7/710
 		.word	0xfff8fe00
-		W(b)	__arm7_mmu_cache_off
-		W(b)	__arm7_mmu_cache_off
+		mov	pc, lr
+ THUMB(		nop				)
+		mov	pc, lr
+ THUMB(		nop				)
 		mov	pc, lr
  THUMB(		nop				)
 
@@ -776,7 +939,7 @@ proc_types:
 
 		.word	0x0007b000		@ ARMv6
 		.word	0x000ff000
-		W(b)	__armv4_mmu_cache_on
+		W(b)	__armv6_mmu_cache_on
 		W(b)	__armv4_mmu_cache_off
 		W(b)	__armv6_mmu_cache_flush
 
@@ -868,21 +1031,6 @@ __armv7_mmu_cache_off:
 		mcr	p15, 0, r0, c7, c5, 4	@ ISB
 		mov	pc, r12
 
-__arm6_mmu_cache_off:
-		mov	r0, #0x00000030		@ ARM6 control reg.
-		b	__armv3_mmu_cache_off
-
-__arm7_mmu_cache_off:
-		mov	r0, #0x00000070		@ ARM7 control reg.
-		b	__armv3_mmu_cache_off
-
-__armv3_mmu_cache_off:
-		mcr	p15, 0, r0, c1, c0, 0	@ turn MMU and cache off
-		mov	r0, #0
-		mcr	p15, 0, r0, c7, c0, 0	@ invalidate whole cache v3
-		mcr	p15, 0, r0, c5, c0, 0	@ invalidate whole TLB v3
-		mov	pc, lr
-
 /*
  * Clean and flush the cache to maintain consistency.
  *
@@ -1111,6 +1259,25 @@ memdump:	mov	r12, r0
 #endif
 
 		.ltorg
+
+#ifdef CONFIG_ARM_VIRT_EXT
+.align 5
+__hyp_reentry_vectors:
+		W(b)	.			@ reset
+		W(b)	.			@ undef
+		W(b)	.			@ svc
+		W(b)	.			@ pabort
+		W(b)	.			@ dabort
+		W(b)	__enter_kernel		@ hyp
+		W(b)	.			@ irq
+		W(b)	.			@ fiq
+#endif /* CONFIG_ARM_VIRT_EXT */
+
+__enter_kernel:
+		mov	r0, #0			@ must be 0
+ ARM(		mov	pc, r4	)		@ call kernel
+ THUMB(		bx	r4	)		@ entry point is always ARM
+
 reloc_code_end:
 
 		.align
diff --git a/arch/arm/boot/compressed/libfdt_env.h b/arch/arm/boot/compressed/libfdt_env.h
new file mode 100644
index 00000000000..1f4e71876b0
--- /dev/null
+++ b/arch/arm/boot/compressed/libfdt_env.h
@@ -0,0 +1,15 @@
+#ifndef _ARM_LIBFDT_ENV_H
+#define _ARM_LIBFDT_ENV_H
+
+#include <linux/types.h>
+#include <linux/string.h>
+#include <asm/byteorder.h>
+
+#define fdt16_to_cpu(x)		be16_to_cpu(x)
+#define cpu_to_fdt16(x)		cpu_to_be16(x)
+#define fdt32_to_cpu(x)		be32_to_cpu(x)
+#define cpu_to_fdt32(x)		cpu_to_be32(x)
+#define fdt64_to_cpu(x)		be64_to_cpu(x)
+#define cpu_to_fdt64(x)		cpu_to_be64(x)
+
+#endif
diff --git a/arch/arm/boot/compressed/misc.c b/arch/arm/boot/compressed/misc.c
index 832d37236c5..d4f891f5699 100644
--- a/arch/arm/boot/compressed/misc.c
+++ b/arch/arm/boot/compressed/misc.c
@@ -18,19 +18,14 @@
 
 unsigned int __machine_arch_type;
 
-#define _LINUX_STRING_H_
-
 #include <linux/compiler.h>	/* for inline */
-#include <linux/types.h>	/* for size_t */
-#include <linux/stddef.h>	/* for NULL */
+#include <linux/types.h>
 #include <linux/linkage.h>
-#include <asm/string.h>
-
 
 static void putstr(const char *ptr);
 extern void error(char *x);
 
-#include <mach/uncompress.h>
+#include CONFIG_UNCOMPRESS_INCLUDE
 
 #ifdef CONFIG_DEBUG_ICEDCC
 
@@ -101,41 +96,6 @@ static void putstr(const char *ptr)
 	flush();
 }
 
-
-void *memcpy(void *__dest, __const void *__src, size_t __n)
-{
-	int i = 0;
-	unsigned char *d = (unsigned char *)__dest, *s = (unsigned char *)__src;
-
-	for (i = __n >> 3; i > 0; i--) {
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-	}
-
-	if (__n & 1 << 2) {
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-		*d++ = *s++;
-	}
-
-	if (__n & 1 << 1) {
-		*d++ = *s++;
-		*d++ = *s++;
-	}
-
-	if (__n & 1)
-		*d++ = *s++;
-
-	return __dest;
-}
-
 /*
  * gzip declarations
  */
@@ -167,6 +127,18 @@ asmlinkage void __div0(void)
 	error("Attempting division by 0!");
 }
 
+unsigned long __stack_chk_guard;
+
+void __stack_chk_guard_setup(void)
+{
+	__stack_chk_guard = 0x000a0dff;
+}
+
+void __stack_chk_fail(void)
+{
+	error("stack-protector: Kernel stack is corrupted\n");
+}
+
 extern int do_decompress(u8 *input, int len, u8 *output, void (*error)(char *x));
 
 
@@ -177,6 +149,8 @@ decompress_kernel(unsigned long output_start, unsigned long free_mem_ptr_p,
 {
 	int ret;
 
+	__stack_chk_guard_setup();
+
 	output_data		= (unsigned char *)output_start;
 	free_mem_ptr		= free_mem_ptr_p;
 	free_mem_end_ptr	= free_mem_ptr_end_p;
diff --git a/arch/arm/boot/compressed/mmcif-sh7372.c b/arch/arm/boot/compressed/mmcif-sh7372.c
index b6f61d9a5a1..672ae95db5c 100644
--- a/arch/arm/boot/compressed/mmcif-sh7372.c
+++ b/arch/arm/boot/compressed/mmcif-sh7372.c
@@ -82,7 +82,7 @@ asmlinkage void mmc_loader(unsigned char *buf, unsigned long len)
 
 
 	/* Disable clock to MMC hardware block */
-	__raw_writel(__raw_readl(SMSTPCR3) & (1 << 12), SMSTPCR3);
+	__raw_writel(__raw_readl(SMSTPCR3) | (1 << 12), SMSTPCR3);
 
 	mmc_update_progress(MMC_PROGRESS_DONE);
 }
diff --git a/arch/arm/boot/compressed/ofw-shark.c b/arch/arm/boot/compressed/ofw-shark.c
deleted file mode 100644
index 465c54b6b12..00000000000
--- a/arch/arm/boot/compressed/ofw-shark.c
+++ /dev/null
@@ -1,260 +0,0 @@
-/*
- * linux/arch/arm/boot/compressed/ofw-shark.c
- *
- * by Alexander Schulz
- *
- * This file is used to get some basic information
- * about the memory layout of the shark we are running
- * on. Memory is usually divided in blocks a 8 MB.
- * And bootargs are copied from OpenFirmware.
- */
-
-
-#include <linux/kernel.h>
-#include <linux/types.h>
-#include <asm/setup.h>
-#include <asm/page.h>
-
-
-asmlinkage void
-create_params (unsigned long *buffer)
-{
-	/* Is there a better address? Also change in mach-shark/core.c */
-	struct tag *tag = (struct tag *) 0x08003000;
-	int j,i,m,k,nr_banks,size;
-	unsigned char *c;
-
-	k = 0;
-
-	/* Head of the taglist */
-	tag->hdr.tag  = ATAG_CORE;
-	tag->hdr.size = tag_size(tag_core);
-	tag->u.core.flags = 1;
-	tag->u.core.pagesize = PAGE_SIZE;
-	tag->u.core.rootdev = 0;
-
-	/* Build up one tagged block for each memory region */
-	size=0;
-	nr_banks=(unsigned int) buffer[0];
-	for (j=0;j<nr_banks;j++){
-		/* search the lowest address and put it into the next entry   */
-		/* not a fast sort algorithm, but there are at most 8 entries */
-		/* and this is used only once anyway                          */
-		m=0xffffffff;
-		for (i=0;i<(unsigned int) buffer[0];i++){
-			if (buffer[2*i+1]<m) {
-				m=buffer[2*i+1];
-				k=i;
-			}
-		}
-	  
-		tag = tag_next(tag);
-		tag->hdr.tag = ATAG_MEM;
-		tag->hdr.size = tag_size(tag_mem32);
-		tag->u.mem.size = buffer[2*k+2];
-		tag->u.mem.start = buffer[2*k+1];
-
-		size += buffer[2*k+2];
-
-		buffer[2*k+1]=0xffffffff;                    /* mark as copied */
-	}
-	
-	/* The command line */
-	tag = tag_next(tag);
-	tag->hdr.tag = ATAG_CMDLINE;
-	
-	c=(unsigned char *)(&buffer[34]);
-	j=0;
-	while (*c) tag->u.cmdline.cmdline[j++]=*c++;
-
-	tag->u.cmdline.cmdline[j]=0;
-	tag->hdr.size = (j + 7 + sizeof(struct tag_header)) >> 2;
-
-	/* Hardware revision */
-	tag = tag_next(tag);
-	tag->hdr.tag = ATAG_REVISION;
-	tag->hdr.size = tag_size(tag_revision);
-	tag->u.revision.rev = ((unsigned char) buffer[33])-'0';
-
-	/* End of the taglist */
-	tag = tag_next(tag);
-	tag->hdr.tag = 0;
-	tag->hdr.size = 0;
-}
-
-
-typedef int (*ofw_handle_t)(void *);
-
-/* Everything below is called with a wrong MMU setting.
- * This means: no string constants, no initialization of
- * arrays, no global variables! This is ugly but I didn't
- * want to write this in assembler :-)
- */
-
-int
-of_decode_int(const unsigned char *p)
-{
-	unsigned int i = *p++ << 8;
-	i = (i + *p++) << 8;
-	i = (i + *p++) << 8;
-	return (i + *p);
-}
-  
-int
-OF_finddevice(ofw_handle_t openfirmware, char *name)
-{
-	unsigned int args[8];
-	char service[12];
-
-	service[0]='f';
-	service[1]='i';
-	service[2]='n';
-	service[3]='d';
-	service[4]='d';
-	service[5]='e';
-	service[6]='v';
-	service[7]='i';
-	service[8]='c';
-	service[9]='e';
-	service[10]='\0';
-
-	args[0]=(unsigned int)service;
-	args[1]=1;
-	args[2]=1;
-	args[3]=(unsigned int)name;
-
-	if (openfirmware(args) == -1)
-		return -1;
-	return args[4];
-}
-
-int
-OF_getproplen(ofw_handle_t openfirmware, int handle, char *prop)
-{
-	unsigned int args[8];
-	char service[12];
-
-	service[0]='g';
-	service[1]='e';
-	service[2]='t';
-	service[3]='p';
-	service[4]='r';
-	service[5]='o';
-	service[6]='p';
-	service[7]='l';
-	service[8]='e';
-	service[9]='n';
-	service[10]='\0';
-
-	args[0] = (unsigned int)service;
-	args[1] = 2;
-	args[2] = 1;
-	args[3] = (unsigned int)handle;
-	args[4] = (unsigned int)prop;
-
-	if (openfirmware(args) == -1)
-		return -1;
-	return args[5];
-}
-  
-int
-OF_getprop(ofw_handle_t openfirmware, int handle, char *prop, void *buf, unsigned int buflen)
-{
-	unsigned int args[8];
-	char service[8];
-
-	service[0]='g';
-	service[1]='e';
-	service[2]='t';
-	service[3]='p';
-	service[4]='r';
-	service[5]='o';
-	service[6]='p';
-	service[7]='\0';
-
-	args[0] = (unsigned int)service;
-	args[1] = 4;
-	args[2] = 1;
-	args[3] = (unsigned int)handle;
-	args[4] = (unsigned int)prop;
-	args[5] = (unsigned int)buf;
-	args[6] = buflen;
-
-	if (openfirmware(args) == -1)
-		return -1;
-	return args[7];
-}
-  
-asmlinkage void ofw_init(ofw_handle_t o, int *nomr, int *pointer)
-{
-	int phandle,i,mem_len,buffer[32];
-	char temp[15];
-  
-	temp[0]='/';
-	temp[1]='m';
-	temp[2]='e';
-	temp[3]='m';
-	temp[4]='o';
-	temp[5]='r';
-	temp[6]='y';
-	temp[7]='\0';
-
-	phandle=OF_finddevice(o,temp);
-
-	temp[0]='r';
-	temp[1]='e';
-	temp[2]='g';
-	temp[3]='\0';
-
-	mem_len = OF_getproplen(o,phandle, temp);
-	OF_getprop(o,phandle, temp, buffer, mem_len);
-	*nomr=mem_len >> 3;
-
-	for (i=0; i<=mem_len/4; i++) pointer[i]=of_decode_int((const unsigned char *)&buffer[i]);
-
-	temp[0]='/';
-	temp[1]='c';
-	temp[2]='h';
-	temp[3]='o';
-	temp[4]='s';
-	temp[5]='e';
-	temp[6]='n';
-	temp[7]='\0';
-
-	phandle=OF_finddevice(o,temp);
-
-	temp[0]='b';
-	temp[1]='o';
-	temp[2]='o';
-	temp[3]='t';
-	temp[4]='a';
-	temp[5]='r';
-	temp[6]='g';
-	temp[7]='s';
-	temp[8]='\0';
-
-	mem_len = OF_getproplen(o,phandle, temp);
-	OF_getprop(o,phandle, temp, buffer, mem_len);
-	if (mem_len > 128) mem_len=128;
-	for (i=0; i<=mem_len/4; i++) pointer[i+33]=buffer[i];
-	pointer[i+33]=0;
-
-	temp[0]='/';
-	temp[1]='\0';
-	phandle=OF_finddevice(o,temp);
-	temp[0]='b';
-	temp[1]='a';
-	temp[2]='n';
-	temp[3]='n';
-	temp[4]='e';
-	temp[5]='r';
-	temp[6]='-';
-	temp[7]='n';
-	temp[8]='a';
-	temp[9]='m';
-	temp[10]='e';
-	temp[11]='\0';
-	mem_len = OF_getproplen(o,phandle, temp);
-	OF_getprop(o,phandle, temp, buffer, mem_len);
-	* ((unsigned char *) &pointer[32]) = ((unsigned char *) buffer)[mem_len-2];
-}
diff --git a/arch/arm/boot/compressed/piggy.lz4.S b/arch/arm/boot/compressed/piggy.lz4.S
new file mode 100644
index 00000000000..3d9a575618a
--- /dev/null
+++ b/arch/arm/boot/compressed/piggy.lz4.S
@@ -0,0 +1,6 @@
+	.section .piggydata,#alloc
+	.globl	input_data
+input_data:
+	.incbin	"arch/arm/boot/compressed/piggy.lz4"
+	.globl	input_data_end
+input_data_end:
diff --git a/arch/arm/boot/compressed/piggy.xzkern.S b/arch/arm/boot/compressed/piggy.xzkern.S
new file mode 100644
index 00000000000..5703f300d02
--- /dev/null
+++ b/arch/arm/boot/compressed/piggy.xzkern.S
@@ -0,0 +1,6 @@
+	.section .piggydata,#alloc
+	.globl	input_data
+input_data:
+	.incbin	"arch/arm/boot/compressed/piggy.xzkern"
+	.globl	input_data_end
+input_data_end:
diff --git a/arch/arm/boot/compressed/sdhi-sh7372.c b/arch/arm/boot/compressed/sdhi-sh7372.c
index d403a8b24d7..d279294f238 100644
--- a/arch/arm/boot/compressed/sdhi-sh7372.c
+++ b/arch/arm/boot/compressed/sdhi-sh7372.c
@@ -85,7 +85,7 @@ asmlinkage void mmc_loader(unsigned short *buf, unsigned long len)
 		goto err;
 
         /* Disable clock to SDHI1 hardware block */
-        __raw_writel(__raw_readl(SMSTPCR3) & (1 << 13), SMSTPCR3);
+        __raw_writel(__raw_readl(SMSTPCR3) | (1 << 13), SMSTPCR3);
 
 	mmc_update_progress(MMC_PROGRESS_DONE);
 
diff --git a/arch/arm/boot/compressed/string.c b/arch/arm/boot/compressed/string.c
new file mode 100644
index 00000000000..36e53ef9200
--- /dev/null
+++ b/arch/arm/boot/compressed/string.c
@@ -0,0 +1,127 @@
+/*
+ * arch/arm/boot/compressed/string.c
+ *
+ * Small subset of simple string routines
+ */
+
+#include <linux/string.h>
+
+void *memcpy(void *__dest, __const void *__src, size_t __n)
+{
+	int i = 0;
+	unsigned char *d = (unsigned char *)__dest, *s = (unsigned char *)__src;
+
+	for (i = __n >> 3; i > 0; i--) {
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+	}
+
+	if (__n & 1 << 2) {
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+		*d++ = *s++;
+	}
+
+	if (__n & 1 << 1) {
+		*d++ = *s++;
+		*d++ = *s++;
+	}
+
+	if (__n & 1)
+		*d++ = *s++;
+
+	return __dest;
+}
+
+void *memmove(void *__dest, __const void *__src, size_t count)
+{
+	unsigned char *d = __dest;
+	const unsigned char *s = __src;
+
+	if (__dest == __src)
+		return __dest;
+
+	if (__dest < __src)
+		return memcpy(__dest, __src, count);
+
+	while (count--)
+		d[count] = s[count];
+	return __dest;
+}
+
+size_t strlen(const char *s)
+{
+	const char *sc = s;
+
+	while (*sc != '\0')
+		sc++;
+	return sc - s;
+}
+
+int memcmp(const void *cs, const void *ct, size_t count)
+{
+	const unsigned char *su1 = cs, *su2 = ct, *end = su1 + count;
+	int res = 0;
+
+	while (su1 < end) {
+		res = *su1++ - *su2++;
+		if (res)
+			break;
+	}
+	return res;
+}
+
+int strcmp(const char *cs, const char *ct)
+{
+	unsigned char c1, c2;
+	int res = 0;
+
+	do {
+		c1 = *cs++;
+		c2 = *ct++;
+		res = c1 - c2;
+		if (res)
+			break;
+	} while (c1);
+	return res;
+}
+
+void *memchr(const void *s, int c, size_t count)
+{
+	const unsigned char *p = s;
+
+	while (count--)
+		if ((unsigned char)c == *p++)
+			return (void *)(p - 1);
+	return NULL;
+}
+
+char *strchr(const char *s, int c)
+{
+	while (*s != (char)c)
+		if (*s++ == '\0')
+			return NULL;
+	return (char *)s;
+}
+
+#undef memset
+
+void *memset(void *s, int c, size_t count)
+{
+	char *xs = s;
+	while (count--)
+		*xs++ = c;
+	return s;
+}
+
+void __memzero(void *s, size_t count)
+{
+	memset(s, 0, count);
+}
diff --git a/arch/arm/boot/compressed/vmlinux.lds.in b/arch/arm/boot/compressed/vmlinux.lds.in
index 4e728834a1b..4919f2ac8b8 100644
--- a/arch/arm/boot/compressed/vmlinux.lds.in
+++ b/arch/arm/boot/compressed/vmlinux.lds.in
@@ -51,6 +51,10 @@ SECTIONS
   _got_start = .;
   .got			: { *(.got) }
   _got_end = .;
+
+  /* ensure the zImage file size is always a multiple of 64 bits */
+  /* (without a dummy byte, ld just ignores the empty section) */
+  .pad			: { BYTE(0); . = ALIGN(8); }
   _edata = .;
 
   . = BSS_START;