Merge commit 'origin/master' into next

Manual merge of:
	arch/powerpc/include/asm/elf.h
	drivers/i2c/busses/i2c-mpc.c

15 files changed, 1464 insertions, 56 deletions

diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig
index d490f3773c0..20979564e7e 100644
--- a/arch/arm/mm/Kconfig
+++ b/arch/arm/mm/Kconfig
@@ -186,6 +186,24 @@ config CPU_ARM926T
 	  Say Y if you want support for the ARM926T processor.
 	  Otherwise, say N.
 
+# FA526
+config CPU_FA526
+	bool
+	select CPU_32v4
+	select CPU_ABRT_EV4
+	select CPU_PABRT_NOIFAR
+	select CPU_CACHE_VIVT
+	select CPU_CP15_MMU
+	select CPU_CACHE_FA
+	select CPU_COPY_FA if MMU
+	select CPU_TLB_FA if MMU
+	help
+	  The FA526 is a version of the ARMv4 compatible processor with
+	  Branch Target Buffer, Unified TLB and cache line size 16.
+
+	  Say Y if you want support for the FA526 processor.
+	  Otherwise, say N.
+
 # ARM940T
 config CPU_ARM940T
 	bool "Support ARM940T processor" if ARCH_INTEGRATOR
@@ -340,6 +358,17 @@ config CPU_XSC3
 	select CPU_TLB_V4WBI if MMU
 	select IO_36
 
+# Marvell PJ1 (Mohawk)
+config CPU_MOHAWK
+	bool
+	select CPU_32v5
+	select CPU_ABRT_EV5T
+	select CPU_PABRT_NOIFAR
+	select CPU_CACHE_VIVT
+	select CPU_CP15_MMU
+	select CPU_TLB_V4WBI if MMU
+	select CPU_COPY_V4WB if MMU
+
 # Feroceon
 config CPU_FEROCEON
 	bool
@@ -484,6 +513,9 @@ config CPU_CACHE_VIVT
 config CPU_CACHE_VIPT
 	bool
 
+config CPU_CACHE_FA
+	bool
+
 if MMU
 # The copy-page model
 config CPU_COPY_V3
@@ -498,6 +530,9 @@ config CPU_COPY_V4WB
 config CPU_COPY_FEROCEON
 	bool
 
+config CPU_COPY_FA
+	bool
+
 config CPU_COPY_V6
 	bool
 
@@ -528,6 +563,13 @@ config CPU_TLB_FEROCEON
 	help
 	  Feroceon TLB (v4wbi with non-outer-cachable page table walks).
 
+config CPU_TLB_FA
+	bool
+	help
+	  Faraday ARM FA526 architecture, unified TLB with writeback cache
+	  and invalidate instruction cache entry. Branch target buffer is
+	  also supported.
+
 config CPU_TLB_V6
 	bool
 
@@ -569,7 +611,7 @@ comment "Processor Features"
 
 config ARM_THUMB
 	bool "Support Thumb user binaries"
-	depends on CPU_ARM720T || CPU_ARM740T || CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM940T || CPU_ARM946E || CPU_ARM1020 || CPU_ARM1020E || CPU_ARM1022 || CPU_ARM1026 || CPU_XSCALE || CPU_XSC3 || CPU_V6 || CPU_V7 || CPU_FEROCEON
+	depends on CPU_ARM720T || CPU_ARM740T || CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM940T || CPU_ARM946E || CPU_ARM1020 || CPU_ARM1020E || CPU_ARM1022 || CPU_ARM1026 || CPU_XSCALE || CPU_XSC3 || CPU_MOHAWK || CPU_V6 || CPU_V7 || CPU_FEROCEON
 	default y
 	help
 	  Say Y if you want to include kernel support for running user space
@@ -638,7 +680,7 @@ config CPU_DCACHE_SIZE
 
 config CPU_DCACHE_WRITETHROUGH
 	bool "Force write through D-cache"
-	depends on (CPU_ARM740T || CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM940T || CPU_ARM946E || CPU_ARM1020) && !CPU_DCACHE_DISABLE
+	depends on (CPU_ARM740T || CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM940T || CPU_ARM946E || CPU_ARM1020 || CPU_FA526) && !CPU_DCACHE_DISABLE
 	default y if CPU_ARM925T
 	help
 	  Say Y here to use the data cache in writethrough mode. Unless you
@@ -653,7 +695,7 @@ config CPU_CACHE_ROUND_ROBIN
 
 config CPU_BPREDICT_DISABLE
 	bool "Disable branch prediction"
-	depends on CPU_ARM1020 || CPU_V6 || CPU_XSC3 || CPU_V7
+	depends on CPU_ARM1020 || CPU_V6 || CPU_MOHAWK || CPU_XSC3 || CPU_V7 || CPU_FA526
 	help
 	  Say Y here to disable branch prediction.  If unsure, say N.
 
@@ -704,7 +746,8 @@ config CACHE_FEROCEON_L2_WRITETHROUGH
 
 config CACHE_L2X0
 	bool "Enable the L2x0 outer cache controller"
-	depends on REALVIEW_EB_ARM11MP || MACH_REALVIEW_PB11MP || MACH_REALVIEW_PB1176 || REALVIEW_EB_A9MP
+	depends on REALVIEW_EB_ARM11MP || MACH_REALVIEW_PB11MP || MACH_REALVIEW_PB1176 || \
+		   REALVIEW_EB_A9MP || ARCH_MX35 || ARCH_MX31
 	default y
 	select OUTER_CACHE
 	help
diff --git a/arch/arm/mm/Makefile b/arch/arm/mm/Makefile
index 480f78a3611..63e3f6dd0e2 100644
--- a/arch/arm/mm/Makefile
+++ b/arch/arm/mm/Makefile
@@ -16,6 +16,7 @@ obj-$(CONFIG_MODULES)		+= proc-syms.o
 
 obj-$(CONFIG_ALIGNMENT_TRAP)	+= alignment.o
 obj-$(CONFIG_DISCONTIGMEM)	+= discontig.o
+obj-$(CONFIG_HIGHMEM)		+= highmem.o
 
 obj-$(CONFIG_CPU_ABRT_NOMMU)	+= abort-nommu.o
 obj-$(CONFIG_CPU_ABRT_EV4)	+= abort-ev4.o
@@ -32,6 +33,7 @@ obj-$(CONFIG_CPU_CACHE_V4WT)	+= cache-v4wt.o
 obj-$(CONFIG_CPU_CACHE_V4WB)	+= cache-v4wb.o
 obj-$(CONFIG_CPU_CACHE_V6)	+= cache-v6.o
 obj-$(CONFIG_CPU_CACHE_V7)	+= cache-v7.o
+obj-$(CONFIG_CPU_CACHE_FA)	+= cache-fa.o
 
 obj-$(CONFIG_CPU_COPY_V3)	+= copypage-v3.o
 obj-$(CONFIG_CPU_COPY_V4WT)	+= copypage-v4wt.o
@@ -41,6 +43,7 @@ obj-$(CONFIG_CPU_COPY_V6)	+= copypage-v6.o context.o
 obj-$(CONFIG_CPU_SA1100)	+= copypage-v4mc.o
 obj-$(CONFIG_CPU_XSCALE)	+= copypage-xscale.o
 obj-$(CONFIG_CPU_XSC3)		+= copypage-xsc3.o
+obj-$(CONFIG_CPU_COPY_FA)	+= copypage-fa.o
 
 obj-$(CONFIG_CPU_TLB_V3)	+= tlb-v3.o
 obj-$(CONFIG_CPU_TLB_V4WT)	+= tlb-v4.o
@@ -49,6 +52,7 @@ obj-$(CONFIG_CPU_TLB_V4WBI)	+= tlb-v4wbi.o
 obj-$(CONFIG_CPU_TLB_FEROCEON)	+= tlb-v4wbi.o	# reuse v4wbi TLB functions
 obj-$(CONFIG_CPU_TLB_V6)	+= tlb-v6.o
 obj-$(CONFIG_CPU_TLB_V7)	+= tlb-v7.o
+obj-$(CONFIG_CPU_TLB_FA)	+= tlb-fa.o
 
 obj-$(CONFIG_CPU_ARM610)	+= proc-arm6_7.o
 obj-$(CONFIG_CPU_ARM710)	+= proc-arm6_7.o
@@ -62,6 +66,7 @@ obj-$(CONFIG_CPU_ARM925T)	+= proc-arm925.o
 obj-$(CONFIG_CPU_ARM926T)	+= proc-arm926.o
 obj-$(CONFIG_CPU_ARM940T)	+= proc-arm940.o
 obj-$(CONFIG_CPU_ARM946E)	+= proc-arm946.o
+obj-$(CONFIG_CPU_FA526)		+= proc-fa526.o
 obj-$(CONFIG_CPU_ARM1020)	+= proc-arm1020.o
 obj-$(CONFIG_CPU_ARM1020E)	+= proc-arm1020e.o
 obj-$(CONFIG_CPU_ARM1022)	+= proc-arm1022.o
@@ -70,6 +75,7 @@ obj-$(CONFIG_CPU_SA110)		+= proc-sa110.o
 obj-$(CONFIG_CPU_SA1100)	+= proc-sa1100.o
 obj-$(CONFIG_CPU_XSCALE)	+= proc-xscale.o
 obj-$(CONFIG_CPU_XSC3)		+= proc-xsc3.o
+obj-$(CONFIG_CPU_MOHAWK)	+= proc-mohawk.o
 obj-$(CONFIG_CPU_FEROCEON)	+= proc-feroceon.o
 obj-$(CONFIG_CPU_V6)		+= proc-v6.o
 obj-$(CONFIG_CPU_V7)		+= proc-v7.o
diff --git a/arch/arm/mm/cache-fa.S b/arch/arm/mm/cache-fa.S
new file mode 100644
index 00000000000..b63a8f7b95c
--- /dev/null
+++ b/arch/arm/mm/cache-fa.S
@@ -0,0 +1,220 @@
+/*
+ *  linux/arch/arm/mm/cache-fa.S
+ *
+ *  Copyright (C) 2005 Faraday Corp.
+ *  Copyright (C) 2008-2009 Paulius Zaleckas <paulius.zaleckas@teltonika.lt>
+ *
+ * Based on cache-v4wb.S:
+ *  Copyright (C) 1997-2002 Russell king
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ *  Processors: FA520 FA526 FA626	
+ */
+#include <linux/linkage.h>
+#include <linux/init.h>
+#include <asm/memory.h>
+#include <asm/page.h>
+
+#include "proc-macros.S"
+
+/*
+ * The size of one data cache line.
+ */
+#define CACHE_DLINESIZE	16
+
+/*
+ * The total size of the data cache.
+ */
+#ifdef CONFIG_ARCH_GEMINI
+#define CACHE_DSIZE	8192
+#else
+#define CACHE_DSIZE	16384 
+#endif 
+
+/* FIXME: put optimal value here. Current one is just estimation */
+#define CACHE_DLIMIT	(CACHE_DSIZE * 2)
+
+/*
+ *	flush_user_cache_all()
+ *
+ *	Clean and invalidate all cache entries in a particular address
+ *	space.
+ */
+ENTRY(fa_flush_user_cache_all)
+	/* FALLTHROUGH */
+/*
+ *	flush_kern_cache_all()
+ *
+ *	Clean and invalidate the entire cache.
+ */
+ENTRY(fa_flush_kern_cache_all)
+	mov	ip, #0
+	mov	r2, #VM_EXEC
+__flush_whole_cache:
+	mcr	p15, 0, ip, c7, c14, 0		@ clean/invalidate D cache
+	tst	r2, #VM_EXEC
+	mcrne	p15, 0, ip, c7, c5, 0		@ invalidate I cache
+	mcrne	p15, 0, ip, c7, c5, 6		@ invalidate BTB
+	mcrne	p15, 0, ip, c7, c10, 4		@ drain write buffer
+	mcrne	p15, 0, ip, c7, c5, 4		@ prefetch flush
+	mov	pc, lr
+
+/*
+ *	flush_user_cache_range(start, end, flags)
+ *
+ *	Invalidate a range of cache entries in the specified
+ *	address space.
+ *
+ *	- start - start address (inclusive, page aligned)
+ *	- end	- end address (exclusive, page aligned)
+ *	- flags	- vma_area_struct flags describing address space
+ */
+ENTRY(fa_flush_user_cache_range)
+	mov	ip, #0
+	sub	r3, r1, r0			@ calculate total size
+	cmp	r3, #CACHE_DLIMIT		@ total size >= limit?
+	bhs	__flush_whole_cache		@ flush whole D cache
+
+1:	tst	r2, #VM_EXEC
+	mcrne	p15, 0, r0, c7, c5, 1		@ invalidate I line
+	mcr	p15, 0, r0, c7, c14, 1		@ clean and invalidate D entry
+	add	r0, r0, #CACHE_DLINESIZE
+	cmp	r0, r1
+	blo	1b
+	tst	r2, #VM_EXEC
+	mcrne	p15, 0, ip, c7, c5, 6		@ invalidate BTB
+	mcrne	p15, 0, ip, c7, c10, 4		@ data write barrier
+	mcrne	p15, 0, ip, c7, c5, 4		@ prefetch flush
+	mov	pc, lr
+
+/*
+ *	coherent_kern_range(start, end)
+ *
+ *	Ensure coherency between the Icache and the Dcache in the
+ *	region described by start.  If you have non-snooping
+ *	Harvard caches, you need to implement this function.
+ *
+ *	- start  - virtual start address
+ *	- end	 - virtual end address
+ */
+ENTRY(fa_coherent_kern_range)
+	/* fall through */
+
+/*
+ *	coherent_user_range(start, end)
+ *
+ *	Ensure coherency between the Icache and the Dcache in the
+ *	region described by start.  If you have non-snooping
+ *	Harvard caches, you need to implement this function.
+ *
+ *	- start  - virtual start address
+ *	- end	 - virtual end address
+ */
+ENTRY(fa_coherent_user_range)
+	bic	r0, r0, #CACHE_DLINESIZE - 1
+1:	mcr	p15, 0, r0, c7, c14, 1		@ clean and invalidate D entry
+	mcr	p15, 0, r0, c7, c5, 1		@ invalidate I entry
+	add	r0, r0, #CACHE_DLINESIZE
+	cmp	r0, r1
+	blo	1b
+	mov	r0, #0
+	mcr	p15, 0, r0, c7, c5, 6		@ invalidate BTB
+	mcr	p15, 0, r0, c7, c10, 4		@ drain write buffer
+	mcr	p15, 0, r0, c7, c5, 4		@ prefetch flush
+	mov	pc, lr
+
+/*
+ *	flush_kern_dcache_page(kaddr)
+ *
+ *	Ensure that the data held in the page kaddr is written back
+ *	to the page in question.
+ *
+ *	- kaddr   - kernel address (guaranteed to be page aligned)
+ */
+ENTRY(fa_flush_kern_dcache_page)
+	add	r1, r0, #PAGE_SZ
+1:	mcr	p15, 0, r0, c7, c14, 1		@ clean & invalidate D line
+	add	r0, r0, #CACHE_DLINESIZE
+	cmp	r0, r1
+	blo	1b
+	mov	r0, #0
+	mcr	p15, 0, r0, c7, c5, 0		@ invalidate I cache
+	mcr	p15, 0, r0, c7, c10, 4		@ drain write buffer
+	mov	pc, lr
+
+/*
+ *	dma_inv_range(start, end)
+ *
+ *	Invalidate (discard) the specified virtual address range.
+ *	May not write back any entries.  If 'start' or 'end'
+ *	are not cache line aligned, those lines must be written
+ *	back.
+ *
+ *	- start  - virtual start address
+ *	- end	 - virtual end address
+ */
+ENTRY(fa_dma_inv_range)
+	tst	r0, #CACHE_DLINESIZE - 1
+	bic	r0, r0, #CACHE_DLINESIZE - 1
+	mcrne	p15, 0, r0, c7, c14, 1		@ clean & invalidate D entry
+	tst	r1, #CACHE_DLINESIZE - 1
+	bic	r1, r1, #CACHE_DLINESIZE - 1
+	mcrne	p15, 0, r1, c7, c14, 1		@ clean & invalidate D entry
+1:	mcr	p15, 0, r0, c7, c6, 1		@ invalidate D entry
+	add	r0, r0, #CACHE_DLINESIZE
+	cmp	r0, r1
+	blo	1b
+	mov	r0, #0
+	mcr	p15, 0, r0, c7, c10, 4		@ drain write buffer
+	mov	pc, lr
+
+/*
+ *	dma_clean_range(start, end)
+ *
+ *	Clean (write back) the specified virtual address range.
+ *
+ *	- start  - virtual start address
+ *	- end	 - virtual end address
+ */
+ENTRY(fa_dma_clean_range)
+	bic	r0, r0, #CACHE_DLINESIZE - 1
+1:	mcr	p15, 0, r0, c7, c10, 1		@ clean D entry
+	add	r0, r0, #CACHE_DLINESIZE
+	cmp	r0, r1
+	blo	1b
+	mov	r0, #0	
+	mcr	p15, 0, r0, c7, c10, 4		@ drain write buffer
+	mov	pc, lr
+
+/*
+ *	dma_flush_range(start,end)
+ *	- start   - virtual start address of region
+ *	- end     - virtual end address of region
+ */
+ENTRY(fa_dma_flush_range)
+	bic	r0, r0, #CACHE_DLINESIZE - 1
+1:	mcr	p15, 0, r0, c7, c14, 1		@ clean & invalidate D entry
+	add	r0, r0, #CACHE_DLINESIZE
+	cmp	r0, r1
+	blo	1b
+	mov	r0, #0	
+	mcr	p15, 0, r0, c7, c10, 4		@ drain write buffer
+	mov	pc, lr
+
+	__INITDATA
+
+	.type	fa_cache_fns, #object
+ENTRY(fa_cache_fns)
+	.long	fa_flush_kern_cache_all
+	.long	fa_flush_user_cache_all
+	.long	fa_flush_user_cache_range
+	.long	fa_coherent_kern_range
+	.long	fa_coherent_user_range
+	.long	fa_flush_kern_dcache_page
+	.long	fa_dma_inv_range
+	.long	fa_dma_clean_range
+	.long	fa_dma_flush_range
+	.size	fa_cache_fns, . - fa_cache_fns
diff --git a/arch/arm/mm/cache-feroceon-l2.c b/arch/arm/mm/cache-feroceon-l2.c
index 80cd207cbae..d6dd83826f8 100644
--- a/arch/arm/mm/cache-feroceon-l2.c
+++ b/arch/arm/mm/cache-feroceon-l2.c
@@ -14,8 +14,12 @@
 
 #include <linux/init.h>
 #include <asm/cacheflush.h>
+#include <asm/kmap_types.h>
+#include <asm/fixmap.h>
+#include <asm/pgtable.h>
+#include <asm/tlbflush.h>
 #include <plat/cache-feroceon-l2.h>
-
+#include "mm.h"
 
 /*
  * Low-level cache maintenance operations.
@@ -34,14 +38,36 @@
  * The range operations require two successive cp15 writes, in
  * between which we don't want to be preempted.
  */
+
+static inline unsigned long l2_start_va(unsigned long paddr)
+{
+#ifdef CONFIG_HIGHMEM
+	/*
+	 * Let's do our own fixmap stuff in a minimal way here.
+	 * Because range ops can't be done on physical addresses,
+	 * we simply install a virtual mapping for it only for the
+	 * TLB lookup to occur, hence no need to flush the untouched
+	 * memory mapping.  This is protected with the disabling of
+	 * interrupts by the caller.
+	 */
+	unsigned long idx = KM_L2_CACHE + KM_TYPE_NR * smp_processor_id();
+	unsigned long vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
+	set_pte_ext(TOP_PTE(vaddr), pfn_pte(paddr >> PAGE_SHIFT, PAGE_KERNEL), 0);
+	local_flush_tlb_kernel_page(vaddr);
+	return vaddr + (paddr & ~PAGE_MASK);
+#else
+	return __phys_to_virt(paddr);
+#endif
+}
+
 static inline void l2_clean_pa(unsigned long addr)
 {
 	__asm__("mcr p15, 1, %0, c15, c9, 3" : : "r" (addr));
 }
 
-static inline void l2_clean_mva_range(unsigned long start, unsigned long end)
+static inline void l2_clean_pa_range(unsigned long start, unsigned long end)
 {
-	unsigned long flags;
+	unsigned long va_start, va_end, flags;
 
 	/*
 	 * Make sure 'start' and 'end' reference the same page, as
@@ -51,17 +77,14 @@ static inline void l2_clean_mva_range(unsigned long start, unsigned long end)
 	BUG_ON((start ^ end) >> PAGE_SHIFT);
 
 	raw_local_irq_save(flags);
+	va_start = l2_start_va(start);
+	va_end = va_start + (end - start);
 	__asm__("mcr p15, 1, %0, c15, c9, 4\n\t"
 		"mcr p15, 1, %1, c15, c9, 5"
-		: : "r" (start), "r" (end));
+		: : "r" (va_start), "r" (va_end));
 	raw_local_irq_restore(flags);
 }
 
-static inline void l2_clean_pa_range(unsigned long start, unsigned long end)
-{
-	l2_clean_mva_range(__phys_to_virt(start), __phys_to_virt(end));
-}
-
 static inline void l2_clean_inv_pa(unsigned long addr)
 {
 	__asm__("mcr p15, 1, %0, c15, c10, 3" : : "r" (addr));
@@ -72,9 +95,9 @@ static inline void l2_inv_pa(unsigned long addr)
 	__asm__("mcr p15, 1, %0, c15, c11, 3" : : "r" (addr));
 }
 
-static inline void l2_inv_mva_range(unsigned long start, unsigned long end)
+static inline void l2_inv_pa_range(unsigned long start, unsigned long end)
 {
-	unsigned long flags;
+	unsigned long va_start, va_end, flags;
 
 	/*
 	 * Make sure 'start' and 'end' reference the same page, as
@@ -84,17 +107,14 @@ static inline void l2_inv_mva_range(unsigned long start, unsigned long end)
 	BUG_ON((start ^ end) >> PAGE_SHIFT);
 
 	raw_local_irq_save(flags);
+	va_start = l2_start_va(start);
+	va_end = va_start + (end - start);
 	__asm__("mcr p15, 1, %0, c15, c11, 4\n\t"
 		"mcr p15, 1, %1, c15, c11, 5"
-		: : "r" (start), "r" (end));
+		: : "r" (va_start), "r" (va_end));
 	raw_local_irq_restore(flags);
 }
 
-static inline void l2_inv_pa_range(unsigned long start, unsigned long end)
-{
-	l2_inv_mva_range(__phys_to_virt(start), __phys_to_virt(end));
-}
-
 
 /*
  * Linux primitives.
diff --git a/arch/arm/mm/cache-xsc3l2.c b/arch/arm/mm/cache-xsc3l2.c
index 464de893a98..5d180cb0bd9 100644
--- a/arch/arm/mm/cache-xsc3l2.c
+++ b/arch/arm/mm/cache-xsc3l2.c
@@ -17,12 +17,14 @@
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
  */
 #include <linux/init.h>
-#include <linux/spinlock.h>
-#include <linux/io.h>
-
 #include <asm/system.h>
 #include <asm/cputype.h>
 #include <asm/cacheflush.h>
+#include <asm/kmap_types.h>
+#include <asm/fixmap.h>
+#include <asm/pgtable.h>
+#include <asm/tlbflush.h>
+#include "mm.h"
 
 #define CR_L2	(1 << 26)
 
@@ -47,21 +49,11 @@ static inline void xsc3_l2_clean_mva(unsigned long addr)
 	__asm__("mcr p15, 1, %0, c7, c11, 1" : : "r" (addr));
 }
 
-static inline void xsc3_l2_clean_pa(unsigned long addr)
-{
-	xsc3_l2_clean_mva(__phys_to_virt(addr));
-}
-
 static inline void xsc3_l2_inv_mva(unsigned long addr)
 {
 	__asm__("mcr p15, 1, %0, c7, c7, 1" : : "r" (addr));
 }
 
-static inline void xsc3_l2_inv_pa(unsigned long addr)
-{
-	xsc3_l2_inv_mva(__phys_to_virt(addr));
-}
-
 static inline void xsc3_l2_inv_all(void)
 {
 	unsigned long l2ctype, set_way;
@@ -79,50 +71,103 @@ static inline void xsc3_l2_inv_all(void)
 	dsb();
 }
 
+#ifdef CONFIG_HIGHMEM
+#define l2_map_save_flags(x)		raw_local_save_flags(x)
+#define l2_map_restore_flags(x)		raw_local_irq_restore(x)
+#else
+#define l2_map_save_flags(x)		((x) = 0)
+#define l2_map_restore_flags(x)		((void)(x))
+#endif
+
+static inline unsigned long l2_map_va(unsigned long pa, unsigned long prev_va,
+				      unsigned long flags)
+{
+#ifdef CONFIG_HIGHMEM
+	unsigned long va = prev_va & PAGE_MASK;
+	unsigned long pa_offset = pa << (32 - PAGE_SHIFT);
+	if (unlikely(pa_offset < (prev_va << (32 - PAGE_SHIFT)))) {
+		/*
+		 * Switching to a new page.  Because cache ops are
+		 * using virtual addresses only, we must put a mapping
+		 * in place for it.  We also enable interrupts for a
+		 * short while and disable them again to protect this
+		 * mapping.
+		 */
+		unsigned long idx;
+		raw_local_irq_restore(flags);
+		idx = KM_L2_CACHE + KM_TYPE_NR * smp_processor_id();
+		va = __fix_to_virt(FIX_KMAP_BEGIN + idx);
+		raw_local_irq_restore(flags | PSR_I_BIT);
+		set_pte_ext(TOP_PTE(va), pfn_pte(pa >> PAGE_SHIFT, PAGE_KERNEL), 0);
+		local_flush_tlb_kernel_page(va);
+	}
+	return va + (pa_offset >> (32 - PAGE_SHIFT));
+#else
+	return __phys_to_virt(pa);
+#endif
+}
+
 static void xsc3_l2_inv_range(unsigned long start, unsigned long end)
 {
+	unsigned long vaddr, flags;
+
 	if (start == 0 && end == -1ul) {
 		xsc3_l2_inv_all();
 		return;
 	}
 
+	vaddr = -1;  /* to force the first mapping */
+	l2_map_save_flags(flags);
+
 	/*
 	 * Clean and invalidate partial first cache line.
 	 */
 	if (start & (CACHE_LINE_SIZE - 1)) {
-		xsc3_l2_clean_pa(start & ~(CACHE_LINE_SIZE - 1));
-		xsc3_l2_inv_pa(start & ~(CACHE_LINE_SIZE - 1));
+		vaddr = l2_map_va(start & ~(CACHE_LINE_SIZE - 1), vaddr, flags);
+		xsc3_l2_clean_mva(vaddr);
+		xsc3_l2_inv_mva(vaddr);
 		start = (start | (CACHE_LINE_SIZE - 1)) + 1;
 	}
 
 	/*
-	 * Clean and invalidate partial last cache line.
+	 * Invalidate all full cache lines between 'start' and 'end'.
 	 */
-	if (start < end && (end & (CACHE_LINE_SIZE - 1))) {
-		xsc3_l2_clean_pa(end & ~(CACHE_LINE_SIZE - 1));
-		xsc3_l2_inv_pa(end & ~(CACHE_LINE_SIZE - 1));
-		end &= ~(CACHE_LINE_SIZE - 1);
+	while (start < (end & ~(CACHE_LINE_SIZE - 1))) {
+		vaddr = l2_map_va(start, vaddr, flags);
+		xsc3_l2_inv_mva(vaddr);
+		start += CACHE_LINE_SIZE;
 	}
 
 	/*
-	 * Invalidate all full cache lines between 'start' and 'end'.
+	 * Clean and invalidate partial last cache line.
 	 */
-	while (start < end) {
-		xsc3_l2_inv_pa(start);
-		start += CACHE_LINE_SIZE;
+	if (start < end) {
+		vaddr = l2_map_va(start, vaddr, flags);
+		xsc3_l2_clean_mva(vaddr);
+		xsc3_l2_inv_mva(vaddr);
 	}
 
+	l2_map_restore_flags(flags);
+
 	dsb();
 }
 
 static void xsc3_l2_clean_range(unsigned long start, unsigned long end)
 {
+	unsigned long vaddr, flags;
+
+	vaddr = -1;  /* to force the first mapping */
+	l2_map_save_flags(flags);
+
 	start &= ~(CACHE_LINE_SIZE - 1);
 	while (start < end) {
-		xsc3_l2_clean_pa(start);
+		vaddr = l2_map_va(start, vaddr, flags);
+		xsc3_l2_clean_mva(vaddr);
 		start += CACHE_LINE_SIZE;
 	}
 
+	l2_map_restore_flags(flags);
+
 	dsb();
 }
 
@@ -148,18 +193,26 @@ static inline void xsc3_l2_flush_all(void)
 
 static void xsc3_l2_flush_range(unsigned long start, unsigned long end)
 {
+	unsigned long vaddr, flags;
+
 	if (start == 0 && end == -1ul) {
 		xsc3_l2_flush_all();
 		return;
 	}
 
+	vaddr = -1;  /* to force the first mapping */
+	l2_map_save_flags(flags);
+
 	start &= ~(CACHE_LINE_SIZE - 1);
 	while (start < end) {
-		xsc3_l2_clean_pa(start);
-		xsc3_l2_inv_pa(start);
+		vaddr = l2_map_va(start, vaddr, flags);
+		xsc3_l2_clean_mva(vaddr);
+		xsc3_l2_inv_mva(vaddr);
 		start += CACHE_LINE_SIZE;
 	}
 
+	l2_map_restore_flags(flags);
+
 	dsb();
 }
 
diff --git a/arch/arm/mm/copypage-fa.c b/arch/arm/mm/copypage-fa.c
new file mode 100644
index 00000000000..b2a6008b011
--- /dev/null
+++ b/arch/arm/mm/copypage-fa.c
@@ -0,0 +1,86 @@
+/*
+ *  linux/arch/arm/lib/copypage-fa.S
+ *
+ *  Copyright (C) 2005 Faraday Corp.
+ *  Copyright (C) 2008-2009 Paulius Zaleckas <paulius.zaleckas@teltonika.lt>
+ *
+ * Based on copypage-v4wb.S:
+ *  Copyright (C) 1995-1999 Russell King
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+#include <linux/init.h>
+#include <linux/highmem.h>
+
+/*
+ * Faraday optimised copy_user_page
+ */
+static void __naked
+fa_copy_user_page(void *kto, const void *kfrom)
+{
+	asm("\
+	stmfd	sp!, {r4, lr}			@ 2\n\
+	mov	r2, %0				@ 1\n\
+1:	ldmia	r1!, {r3, r4, ip, lr}		@ 4\n\
+	stmia	r0, {r3, r4, ip, lr}		@ 4\n\
+	mcr	p15, 0, r0, c7, c14, 1		@ 1   clean and invalidate D line\n\
+	add	r0, r0, #16			@ 1\n\
+	ldmia	r1!, {r3, r4, ip, lr}		@ 4\n\
+	stmia	r0, {r3, r4, ip, lr}		@ 4\n\
+	mcr	p15, 0, r0, c7, c14, 1		@ 1   clean and invalidate D line\n\
+	add	r0, r0, #16			@ 1\n\
+	subs	r2, r2, #1			@ 1\n\
+	bne	1b				@ 1\n\
+	mcr	p15, 0, r2, c7, c10, 4		@ 1   drain WB\n\
+	ldmfd	sp!, {r4, pc}			@ 3"
+	:
+	: "I" (PAGE_SIZE / 32));
+}
+
+void fa_copy_user_highpage(struct page *to, struct page *from,
+	unsigned long vaddr)
+{
+	void *kto, *kfrom;
+
+	kto = kmap_atomic(to, KM_USER0);
+	kfrom = kmap_atomic(from, KM_USER1);
+	fa_copy_user_page(kto, kfrom);
+	kunmap_atomic(kfrom, KM_USER1);
+	kunmap_atomic(kto, KM_USER0);
+}
+
+/*
+ * Faraday optimised clear_user_page
+ *
+ * Same story as above.
+ */
+void fa_clear_user_highpage(struct page *page, unsigned long vaddr)
+{
+	void *ptr, *kaddr = kmap_atomic(page, KM_USER0);
+	asm volatile("\
+	mov	r1, %2				@ 1\n\
+	mov	r2, #0				@ 1\n\
+	mov	r3, #0				@ 1\n\
+	mov	ip, #0				@ 1\n\
+	mov	lr, #0				@ 1\n\
+1:	stmia	%0, {r2, r3, ip, lr}		@ 4\n\
+	mcr	p15, 0, %0, c7, c14, 1		@ 1   clean and invalidate D line\n\
+	add	%0, %0, #16			@ 1\n\
+	stmia	%0, {r2, r3, ip, lr}		@ 4\n\
+	mcr	p15, 0, %0, c7, c14, 1		@ 1   clean and invalidate D line\n\
+	add	%0, %0, #16			@ 1\n\
+	subs	r1, r1, #1			@ 1\n\
+	bne	1b				@ 1\n\
+	mcr	p15, 0, r1, c7, c10, 4		@ 1   drain WB"
+	: "=r" (ptr)
+	: "0" (kaddr), "I" (PAGE_SIZE / 32)
+	: "r1", "r2", "r3", "ip", "lr");
+	kunmap_atomic(kaddr, KM_USER0);
+}
+
+struct cpu_user_fns fa_user_fns __initdata = {
+	.cpu_clear_user_highpage = fa_clear_user_highpage,
+	.cpu_copy_user_highpage	= fa_copy_user_highpage,
+};
diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c
index f1ef5613ccd..510c179b0ac 100644
--- a/arch/arm/mm/dma-mapping.c
+++ b/arch/arm/mm/dma-mapping.c
@@ -19,6 +19,7 @@
 #include <linux/dma-mapping.h>
 
 #include <asm/memory.h>
+#include <asm/highmem.h>
 #include <asm/cacheflush.h>
 #include <asm/tlbflush.h>
 #include <asm/sizes.h>
@@ -517,6 +518,74 @@ void dma_cache_maint(const void *start, size_t size, int direction)
 }
 EXPORT_SYMBOL(dma_cache_maint);
 
+static void dma_cache_maint_contiguous(struct page *page, unsigned long offset,
+				       size_t size, int direction)
+{
+	void *vaddr;
+	unsigned long paddr;
+	void (*inner_op)(const void *, const void *);
+	void (*outer_op)(unsigned long, unsigned long);
+
+	switch (direction) {
+	case DMA_FROM_DEVICE:		/* invalidate only */
+		inner_op = dmac_inv_range;
+		outer_op = outer_inv_range;
+		break;
+	case DMA_TO_DEVICE:		/* writeback only */
+		inner_op = dmac_clean_range;
+		outer_op = outer_clean_range;
+		break;
+	case DMA_BIDIRECTIONAL:		/* writeback and invalidate */
+		inner_op = dmac_flush_range;
+		outer_op = outer_flush_range;
+		break;
+	default:
+		BUG();
+	}
+
+	if (!PageHighMem(page)) {
+		vaddr = page_address(page) + offset;
+		inner_op(vaddr, vaddr + size);
+	} else {
+		vaddr = kmap_high_get(page);
+		if (vaddr) {
+			vaddr += offset;
+			inner_op(vaddr, vaddr + size);
+			kunmap_high(page);
+		}
+	}
+
+	paddr = page_to_phys(page) + offset;
+	outer_op(paddr, paddr + size);
+}
+
+void dma_cache_maint_page(struct page *page, unsigned long offset,
+			  size_t size, int dir)
+{
+	/*
+	 * A single sg entry may refer to multiple physically contiguous
+	 * pages.  But we still need to process highmem pages individually.
+	 * If highmem is not configured then the bulk of this loop gets
+	 * optimized out.
+	 */
+	size_t left = size;
+	do {
+		size_t len = left;
+		if (PageHighMem(page) && len + offset > PAGE_SIZE) {
+			if (offset >= PAGE_SIZE) {
+				page += offset / PAGE_SIZE;
+				offset %= PAGE_SIZE;
+			}
+			len = PAGE_SIZE - offset;
+		}
+		dma_cache_maint_contiguous(page, offset, len, dir);
+		offset = 0;
+		page++;
+		left -= len;
+	} while (left);
+}
+EXPORT_SYMBOL(dma_cache_maint_page);
+
 /**
  * dma_map_sg - map a set of SG buffers for streaming mode DMA
  * @dev: valid struct device pointer, or NULL for ISA and EISA-like devices
@@ -614,7 +683,8 @@ void dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg,
 			continue;
 
 		if (!arch_is_coherent())
-			dma_cache_maint(sg_virt(s), s->length, dir);
+			dma_cache_maint_page(sg_page(s), s->offset,
+					     s->length, dir);
 	}
 }
 EXPORT_SYMBOL(dma_sync_sg_for_device);
diff --git a/arch/arm/mm/flush.c b/arch/arm/mm/flush.c
index 0fa9bf388f0..4e283481cee 100644
--- a/arch/arm/mm/flush.c
+++ b/arch/arm/mm/flush.c
@@ -192,7 +192,7 @@ void flush_dcache_page(struct page *page)
 	struct address_space *mapping = page_mapping(page);
 
 #ifndef CONFIG_SMP
-	if (mapping && !mapping_mapped(mapping))
+	if (!PageHighMem(page) && mapping && !mapping_mapped(mapping))
 		set_bit(PG_dcache_dirty, &page->flags);
 	else
 #endif
diff --git a/arch/arm/mm/highmem.c b/arch/arm/mm/highmem.c
new file mode 100644
index 00000000000..a34954d9df7
--- /dev/null
+++ b/arch/arm/mm/highmem.c
@@ -0,0 +1,116 @@
+/*
+ * arch/arm/mm/highmem.c -- ARM highmem support
+ *
+ * Author:	Nicolas Pitre
+ * Created:	september 8, 2008
+ * Copyright:	Marvell Semiconductors Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+#include <linux/highmem.h>
+#include <linux/interrupt.h>
+#include <asm/fixmap.h>
+#include <asm/cacheflush.h>
+#include <asm/tlbflush.h>
+#include "mm.h"
+
+void *kmap(struct page *page)
+{
+	might_sleep();
+	if (!PageHighMem(page))
+		return page_address(page);
+	return kmap_high(page);
+}
+EXPORT_SYMBOL(kmap);
+
+void kunmap(struct page *page)
+{
+	BUG_ON(in_interrupt());
+	if (!PageHighMem(page))
+		return;
+	kunmap_high(page);
+}
+EXPORT_SYMBOL(kunmap);
+
+void *kmap_atomic(struct page *page, enum km_type type)
+{
+	unsigned int idx;
+	unsigned long vaddr;
+
+	pagefault_disable();
+	if (!PageHighMem(page))
+		return page_address(page);
+
+	idx = type + KM_TYPE_NR * smp_processor_id();
+	vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
+#ifdef CONFIG_DEBUG_HIGHMEM
+	/*
+	 * With debugging enabled, kunmap_atomic forces that entry to 0.
+	 * Make sure it was indeed properly unmapped.
+	 */
+	BUG_ON(!pte_none(*(TOP_PTE(vaddr))));
+#endif
+	set_pte_ext(TOP_PTE(vaddr), mk_pte(page, kmap_prot), 0);
+	/*
+	 * When debugging is off, kunmap_atomic leaves the previous mapping
+	 * in place, so this TLB flush ensures the TLB is updated with the
+	 * new mapping.
+	 */
+	local_flush_tlb_kernel_page(vaddr);
+
+	return (void *)vaddr;
+}
+EXPORT_SYMBOL(kmap_atomic);
+
+void kunmap_atomic(void *kvaddr, enum km_type type)
+{
+	unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
+	unsigned int idx = type + KM_TYPE_NR * smp_processor_id();
+
+	if (kvaddr >= (void *)FIXADDR_START) {
+		__cpuc_flush_dcache_page((void *)vaddr);
+#ifdef CONFIG_DEBUG_HIGHMEM
+		BUG_ON(vaddr != __fix_to_virt(FIX_KMAP_BEGIN + idx));
+		set_pte_ext(TOP_PTE(vaddr), __pte(0), 0);
+		local_flush_tlb_kernel_page(vaddr);
+#else
+		(void) idx;  /* to kill a warning */
+#endif
+	}
+	pagefault_enable();
+}
+EXPORT_SYMBOL(kunmap_atomic);
+
+void *kmap_atomic_pfn(unsigned long pfn, enum km_type type)
+{
+	unsigned int idx;
+	unsigned long vaddr;
+
+	pagefault_disable();
+
+	idx = type + KM_TYPE_NR * smp_processor_id();
+	vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
+#ifdef CONFIG_DEBUG_HIGHMEM
+	BUG_ON(!pte_none(*(TOP_PTE(vaddr))));
+#endif
+	set_pte_ext(TOP_PTE(vaddr), pfn_pte(pfn, kmap_prot), 0);
+	local_flush_tlb_kernel_page(vaddr);
+
+	return (void *)vaddr;
+}
+
+struct page *k