Merge branch 'for-linus' of git://git.monstr.eu/linux-2.6-microblaze

* 'for-linus' of git://git.monstr.eu/linux-2.6-microblaze: (27 commits)
  microblaze: entry.S use delay slot for return handlers
  microblaze: Save current task directly
  microblaze: Simplify entry.S - save/restore r3/r4 - ret_from_trap
  microblaze: PCI early support for noMMU system
  microblaze: Fix dma alloc and free coherent dma functions
  microblaze: Add consistent code
  microblaze: pgtable.h: move consistent functions
  microblaze: Remove ancient Kconfig option for consistent mapping
  microblaze: Remove VMALLOC_VMADDR
  microblaze: Add define for ASM_LOOP
  microblaze: Preliminary support for dma drivers
  microblaze: remove trailing space in messages
  microblaze: Use generic show_mem()
  microblaze: Change temp register for cmdline
  microblaze: Preliminary support for dma drivers
  microblaze: Move cache function to cache.c
  microblaze: Add support from PREEMPT
  microblaze: Add support for Xilinx PCI host bridge
  microblaze: Enable PCI, missing files
  microblaze: Add core PCI files
  ...

8 files changed, 443 insertions, 116 deletions

diff --git a/arch/microblaze/kernel/Makefile b/arch/microblaze/kernel/Makefile
index b07594e..e51bc15 100644
--- a/arch/microblaze/kernel/Makefile
+++ b/arch/microblaze/kernel/Makefile
@@ -14,7 +14,7 @@ endif
 
 extra-y := head.o vmlinux.lds
 
-obj-y += exceptions.o \
+obj-y += dma.o exceptions.o \
 	hw_exception_handler.o init_task.o intc.o irq.o of_device.o \
 	of_platform.o process.o prom.o prom_parse.o ptrace.o \
 	setup.o signal.o sys_microblaze.o timer.o traps.o reset.o
diff --git a/arch/microblaze/kernel/asm-offsets.c b/arch/microblaze/kernel/asm-offsets.c
index 7bc7b68..0071260 100644
--- a/arch/microblaze/kernel/asm-offsets.c
+++ b/arch/microblaze/kernel/asm-offsets.c
@@ -90,6 +90,7 @@ int main(int argc, char *argv[])
 	DEFINE(TI_FLAGS, offsetof(struct thread_info, flags));
 	DEFINE(TI_ADDR_LIMIT, offsetof(struct thread_info, addr_limit));
 	DEFINE(TI_CPU_CONTEXT, offsetof(struct thread_info, cpu_context));
+	DEFINE(TI_PREEMPT_COUNT, offsetof(struct thread_info, preempt_count));
 	BLANK();
 
 	/* struct cpu_context */
diff --git a/arch/microblaze/kernel/cpu/cache.c b/arch/microblaze/kernel/cpu/cache.c
index 2a56bcc..f04d8a8 100644
--- a/arch/microblaze/kernel/cpu/cache.c
+++ b/arch/microblaze/kernel/cpu/cache.c
@@ -15,25 +15,6 @@
 #include <asm/cpuinfo.h>
 #include <asm/pvr.h>
 
-static inline void __invalidate_flush_icache(unsigned int addr)
-{
-	__asm__ __volatile__ ("wic	%0, r0;"	\
-					: : "r" (addr));
-}
-
-static inline void __flush_dcache(unsigned int addr)
-{
-	__asm__ __volatile__ ("wdc.flush	%0, r0;"	\
-					: : "r" (addr));
-}
-
-static inline void __invalidate_dcache(unsigned int baseaddr,
-						unsigned int offset)
-{
-	__asm__ __volatile__ ("wdc.clear	%0, %1;"	\
-					: : "r" (baseaddr), "r" (offset));
-}
-
 static inline void __enable_icache_msr(void)
 {
 	__asm__ __volatile__ ("	msrset	r0, %0;		\
@@ -148,9 +129,9 @@ do {									\
 	int step = -line_length;					\
 	BUG_ON(step >= 0);						\
 									\
-	__asm__ __volatile__ (" 1:      " #op " r0, %0;			\
-					bgtid   %0, 1b;			\
-					addk    %0, %0, %1;		\
+	__asm__ __volatile__ (" 1:	" #op "	r0, %0;			\
+					bgtid	%0, 1b;			\
+					addk	%0, %0, %1;		\
 					" : : "r" (len), "r" (step)	\
 					: "memory");			\
 } while (0);
@@ -162,9 +143,9 @@ do {									\
 	int count = end - start;					\
 	BUG_ON(count <= 0);						\
 									\
-	__asm__ __volatile__ (" 1:	" #op " %0, %1;			\
-					bgtid   %1, 1b;			\
-					addk    %1, %1, %2;		\
+	__asm__ __volatile__ (" 1:	" #op "	%0, %1;			\
+					bgtid	%1, 1b;			\
+					addk	%1, %1, %2;		\
 					" : : "r" (start), "r" (count),	\
 					"r" (step) : "memory");		\
 } while (0);
@@ -175,7 +156,7 @@ do {									\
 	int volatile temp;						\
 	BUG_ON(end - start <= 0);					\
 									\
-	__asm__ __volatile__ (" 1:	" #op " %1, r0;			\
+	__asm__ __volatile__ (" 1:	" #op "	%1, r0;			\
 					cmpu	%0, %1, %2;		\
 					bgtid	%0, 1b;			\
 					addk	%1, %1, %3;		\
@@ -183,10 +164,14 @@ do {									\
 					"r" (line_length) : "memory");	\
 } while (0);
 
+#define ASM_LOOP
+
 static void __flush_icache_range_msr_irq(unsigned long start, unsigned long end)
 {
 	unsigned long flags;
-
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s: start 0x%x, end 0x%x\n", __func__,
 				(unsigned int)start, (unsigned int) end);
 
@@ -196,8 +181,13 @@ static void __flush_icache_range_msr_irq(unsigned long start, unsigned long end)
 	local_irq_save(flags);
 	__disable_icache_msr();
 
+#ifdef ASM_LOOP
 	CACHE_RANGE_LOOP_1(start, end, cpuinfo.icache_line_length, wic);
-
+#else
+	for (i = start; i < end; i += cpuinfo.icache_line_length)
+		__asm__ __volatile__ ("wic	%0, r0;"	\
+				: : "r" (i));
+#endif
 	__enable_icache_msr();
 	local_irq_restore(flags);
 }
@@ -206,7 +196,9 @@ static void __flush_icache_range_nomsr_irq(unsigned long start,
 				unsigned long end)
 {
 	unsigned long flags;
-
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s: start 0x%x, end 0x%x\n", __func__,
 				(unsigned int)start, (unsigned int) end);
 
@@ -216,7 +208,13 @@ static void __flush_icache_range_nomsr_irq(unsigned long start,
 	local_irq_save(flags);
 	__disable_icache_nomsr();
 
+#ifdef ASM_LOOP
 	CACHE_RANGE_LOOP_1(start, end, cpuinfo.icache_line_length, wic);
+#else
+	for (i = start; i < end; i += cpuinfo.icache_line_length)
+		__asm__ __volatile__ ("wic	%0, r0;"	\
+				: : "r" (i));
+#endif
 
 	__enable_icache_nomsr();
 	local_irq_restore(flags);
@@ -225,25 +223,41 @@ static void __flush_icache_range_nomsr_irq(unsigned long start,
 static void __flush_icache_range_noirq(unsigned long start,
 				unsigned long end)
 {
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s: start 0x%x, end 0x%x\n", __func__,
 				(unsigned int)start, (unsigned int) end);
 
 	CACHE_LOOP_LIMITS(start, end,
 			cpuinfo.icache_line_length, cpuinfo.icache_size);
+#ifdef ASM_LOOP
 	CACHE_RANGE_LOOP_1(start, end, cpuinfo.icache_line_length, wic);
+#else
+	for (i = start; i < end; i += cpuinfo.icache_line_length)
+		__asm__ __volatile__ ("wic	%0, r0;"	\
+				: : "r" (i));
+#endif
 }
 
 static void __flush_icache_all_msr_irq(void)
 {
 	unsigned long flags;
-
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s\n", __func__);
 
 	local_irq_save(flags);
 	__disable_icache_msr();
-
+#ifdef ASM_LOOP
 	CACHE_ALL_LOOP(cpuinfo.icache_size, cpuinfo.icache_line_length, wic);
-
+#else
+	for (i = 0; i < cpuinfo.icache_size;
+		 i += cpuinfo.icache_line_length)
+			__asm__ __volatile__ ("wic	%0, r0;" \
+					: : "r" (i));
+#endif
 	__enable_icache_msr();
 	local_irq_restore(flags);
 }
@@ -251,35 +265,59 @@ static void __flush_icache_all_msr_irq(void)
 static void __flush_icache_all_nomsr_irq(void)
 {
 	unsigned long flags;
-
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s\n", __func__);
 
 	local_irq_save(flags);
 	__disable_icache_nomsr();
-
+#ifdef ASM_LOOP
 	CACHE_ALL_LOOP(cpuinfo.icache_size, cpuinfo.icache_line_length, wic);
-
+#else
+	for (i = 0; i < cpuinfo.icache_size;
+		 i += cpuinfo.icache_line_length)
+			__asm__ __volatile__ ("wic	%0, r0;" \
+					: : "r" (i));
+#endif
 	__enable_icache_nomsr();
 	local_irq_restore(flags);
 }
 
 static void __flush_icache_all_noirq(void)
 {
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s\n", __func__);
+#ifdef ASM_LOOP
 	CACHE_ALL_LOOP(cpuinfo.icache_size, cpuinfo.icache_line_length, wic);
+#else
+	for (i = 0; i < cpuinfo.icache_size;
+		 i += cpuinfo.icache_line_length)
+			__asm__ __volatile__ ("wic	%0, r0;" \
+					: : "r" (i));
+#endif
 }
 
 static void __invalidate_dcache_all_msr_irq(void)
 {
 	unsigned long flags;
-
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s\n", __func__);
 
 	local_irq_save(flags);
 	__disable_dcache_msr();
-
+#ifdef ASM_LOOP
 	CACHE_ALL_LOOP(cpuinfo.dcache_size, cpuinfo.dcache_line_length, wdc);
-
+#else
+	for (i = 0; i < cpuinfo.dcache_size;
+		 i += cpuinfo.dcache_line_length)
+			__asm__ __volatile__ ("wdc	%0, r0;" \
+					: : "r" (i));
+#endif
 	__enable_dcache_msr();
 	local_irq_restore(flags);
 }
@@ -287,60 +325,107 @@ static void __invalidate_dcache_all_msr_irq(void)
 static void __invalidate_dcache_all_nomsr_irq(void)
 {
 	unsigned long flags;
-
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s\n", __func__);
 
 	local_irq_save(flags);
 	__disable_dcache_nomsr();
-
+#ifdef ASM_LOOP
 	CACHE_ALL_LOOP(cpuinfo.dcache_size, cpuinfo.dcache_line_length, wdc);
-
+#else
+	for (i = 0; i < cpuinfo.dcache_size;
+		 i += cpuinfo.dcache_line_length)
+			__asm__ __volatile__ ("wdc	%0, r0;" \
+					: : "r" (i));
+#endif
 	__enable_dcache_nomsr();
 	local_irq_restore(flags);
 }
 
 static void __invalidate_dcache_all_noirq_wt(void)
 {
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s\n", __func__);
+#ifdef ASM_LOOP
 	CACHE_ALL_LOOP(cpuinfo.dcache_size, cpuinfo.dcache_line_length, wdc)
+#else
+	for (i = 0; i < cpuinfo.dcache_size;
+		 i += cpuinfo.dcache_line_length)
+			__asm__ __volatile__ ("wdc	%0, r0;" \
+					: : "r" (i));
+#endif
 }
 
 /* FIXME this is weird - should be only wdc but not work
  * MS: I am getting bus errors and other weird things */
 static void __invalidate_dcache_all_wb(void)
 {
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s\n", __func__);
+#ifdef ASM_LOOP
 	CACHE_ALL_LOOP2(cpuinfo.dcache_size, cpuinfo.dcache_line_length,
 					wdc.clear)
+#else
+	for (i = 0; i < cpuinfo.dcache_size;
+		 i += cpuinfo.dcache_line_length)
+			__asm__ __volatile__ ("wdc.clear	%0, r0;" \
+					: : "r" (i));
+#endif
 }
 
 static void __invalidate_dcache_range_wb(unsigned long start,
 						unsigned long end)
 {
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s: start 0x%x, end 0x%x\n", __func__,
 				(unsigned int)start, (unsigned int) end);
 
 	CACHE_LOOP_LIMITS(start, end,
 			cpuinfo.dcache_line_length, cpuinfo.dcache_size);
+#ifdef ASM_LOOP
 	CACHE_RANGE_LOOP_2(start, end, cpuinfo.dcache_line_length, wdc.clear);
+#else
+	for (i = start; i < end; i += cpuinfo.icache_line_length)
+		__asm__ __volatile__ ("wdc.clear	%0, r0;"	\
+				: : "r" (i));
+#endif
 }
 
 static void __invalidate_dcache_range_nomsr_wt(unsigned long start,
 							unsigned long end)
 {
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s: start 0x%x, end 0x%x\n", __func__,
 				(unsigned int)start, (unsigned int) end);
 	CACHE_LOOP_LIMITS(start, end,
 			cpuinfo.dcache_line_length, cpuinfo.dcache_size);
 
+#ifdef ASM_LOOP
 	CACHE_RANGE_LOOP_1(start, end, cpuinfo.dcache_line_length, wdc);
+#else
+	for (i = start; i < end; i += cpuinfo.icache_line_length)
+		__asm__ __volatile__ ("wdc	%0, r0;"	\
+				: : "r" (i));
+#endif
 }
 
 static void __invalidate_dcache_range_msr_irq_wt(unsigned long start,
 							unsigned long end)
 {
 	unsigned long flags;
-
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s: start 0x%x, end 0x%x\n", __func__,
 				(unsigned int)start, (unsigned int) end);
 	CACHE_LOOP_LIMITS(start, end,
@@ -349,7 +434,13 @@ static void __invalidate_dcache_range_msr_irq_wt(unsigned long start,
 	local_irq_save(flags);
 	__disable_dcache_msr();
 
+#ifdef ASM_LOOP
 	CACHE_RANGE_LOOP_1(start, end, cpuinfo.dcache_line_length, wdc);
+#else
+	for (i = start; i < end; i += cpuinfo.icache_line_length)
+		__asm__ __volatile__ ("wdc	%0, r0;"	\
+				: : "r" (i));
+#endif
 
 	__enable_dcache_msr();
 	local_irq_restore(flags);
@@ -359,7 +450,9 @@ static void __invalidate_dcache_range_nomsr_irq(unsigned long start,
 							unsigned long end)
 {
 	unsigned long flags;
-
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s: start 0x%x, end 0x%x\n", __func__,
 				(unsigned int)start, (unsigned int) end);
 
@@ -369,7 +462,13 @@ static void __invalidate_dcache_range_nomsr_irq(unsigned long start,
 	local_irq_save(flags);
 	__disable_dcache_nomsr();
 
+#ifdef ASM_LOOP
 	CACHE_RANGE_LOOP_1(start, end, cpuinfo.dcache_line_length, wdc);
+#else
+	for (i = start; i < end; i += cpuinfo.icache_line_length)
+		__asm__ __volatile__ ("wdc	%0, r0;"	\
+				: : "r" (i));
+#endif
 
 	__enable_dcache_nomsr();
 	local_irq_restore(flags);
@@ -377,19 +476,38 @@ static void __invalidate_dcache_range_nomsr_irq(unsigned long start,
 
 static void __flush_dcache_all_wb(void)
 {
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s\n", __func__);
+#ifdef ASM_LOOP
 	CACHE_ALL_LOOP(cpuinfo.dcache_size, cpuinfo.dcache_line_length,
 				wdc.flush);
+#else
+	for (i = 0; i < cpuinfo.dcache_size;
+		 i += cpuinfo.dcache_line_length)
+			__asm__ __volatile__ ("wdc.flush	%0, r0;" \
+					: : "r" (i));
+#endif
 }
 
 static void __flush_dcache_range_wb(unsigned long start, unsigned long end)
 {
+#ifndef ASM_LOOP
+	int i;
+#endif
 	pr_debug("%s: start 0x%x, end 0x%x\n", __func__,
 				(unsigned int)start, (unsigned int) end);
 
 	CACHE_LOOP_LIMITS(start, end,
 			cpuinfo.dcache_line_length, cpuinfo.dcache_size);
+#ifdef ASM_LOOP
 	CACHE_RANGE_LOOP_2(start, end, cpuinfo.dcache_line_length, wdc.flush);
+#else
+	for (i = start; i < end; i += cpuinfo.icache_line_length)
+		__asm__ __volatile__ ("wdc.flush	%0, r0;"	\
+				: : "r" (i));
+#endif
 }
 
 /* struct for wb caches and for wt caches */
@@ -493,7 +611,7 @@ const struct scache wt_nomsr_noirq = {
 #define CPUVER_7_20_A	0x0c
 #define CPUVER_7_20_D	0x0f
 
-#define INFO(s)	printk(KERN_INFO "cache: " s " \n");
+#define INFO(s)	printk(KERN_INFO "cache: " s "\n");
 
 void microblaze_cache_init(void)
 {
@@ -532,4 +650,9 @@ void microblaze_cache_init(void)
 			}
 		}
 	}
+	invalidate_dcache();
+	enable_dcache();
+
+	invalidate_icache();
+	enable_icache();
 }
diff --git a/arch/microblaze/kernel/dma.c b/arch/microblaze/kernel/dma.c
new file mode 100644
index 0000000..b108497
--- /dev/null
+++ b/arch/microblaze/kernel/dma.c
@@ -0,0 +1,156 @@
+/*
+ * Copyright (C) 2009-2010 PetaLogix
+ * Copyright (C) 2006 Benjamin Herrenschmidt, IBM Corporation
+ *
+ * Provide default implementations of the DMA mapping callbacks for
+ * directly mapped busses.
+ */
+
+#include <linux/device.h>
+#include <linux/dma-mapping.h>
+#include <linux/dma-debug.h>
+#include <asm/bug.h>
+#include <asm/cacheflush.h>
+
+/*
+ * Generic direct DMA implementation
+ *
+ * This implementation supports a per-device offset that can be applied if
+ * the address at which memory is visible to devices is not 0. Platform code
+ * can set archdata.dma_data to an unsigned long holding the offset. By
+ * default the offset is PCI_DRAM_OFFSET.
+ */
+static inline void __dma_sync_page(unsigned long paddr, unsigned long offset,
+				size_t size, enum dma_data_direction direction)
+{
+	switch (direction) {
+	case DMA_TO_DEVICE:
+		flush_dcache_range(paddr + offset, paddr + offset + size);
+		break;
+	case DMA_FROM_DEVICE:
+		invalidate_dcache_range(paddr + offset, paddr + offset + size);
+		break;
+	default:
+		BUG();
+	}
+}
+
+static unsigned long get_dma_direct_offset(struct device *dev)
+{
+	if (dev)
+		return (unsigned long)dev->archdata.dma_data;
+
+	return PCI_DRAM_OFFSET; /* FIXME Not sure if is correct */
+}
+
+#define NOT_COHERENT_CACHE
+
+static void *dma_direct_alloc_coherent(struct device *dev, size_t size,
+				dma_addr_t *dma_handle, gfp_t flag)
+{
+#ifdef NOT_COHERENT_CACHE
+	return consistent_alloc(flag, size, dma_handle);
+#else
+	void *ret;
+	struct page *page;
+	int node = dev_to_node(dev);
+
+	/* ignore region specifiers */
+	flag  &= ~(__GFP_HIGHMEM);
+
+	page = alloc_pages_node(node, flag, get_order(size));
+	if (page == NULL)
+		return NULL;
+	ret = page_address(page);
+	memset(ret, 0, size);
+	*dma_handle = virt_to_phys(ret) + get_dma_direct_offset(dev);
+
+	return ret;
+#endif
+}
+
+static void dma_direct_free_coherent(struct device *dev, size_t size,
+			      void *vaddr, dma_addr_t dma_handle)
+{
+#ifdef NOT_COHERENT_CACHE
+	consistent_free(vaddr);
+#else
+	free_pages((unsigned long)vaddr, get_order(size));
+#endif
+}
+
+static int dma_direct_map_sg(struct device *dev, struct scatterlist *sgl,
+			     int nents, enum dma_data_direction direction,
+			     struct dma_attrs *attrs)
+{
+	struct scatterlist *sg;
+	int i;
+
+	/* FIXME this part of code is untested */
+	for_each_sg(sgl, sg, nents, i) {
+		sg->dma_address = sg_phys(sg) + get_dma_direct_offset(dev);
+		sg->dma_length = sg->length;
+		__dma_sync_page(page_to_phys(sg_page(sg)), sg->offset,
+							sg->length, direction);
+	}
+
+	return nents;
+}
+
+static void dma_direct_unmap_sg(struct device *dev, struct scatterlist *sg,
+				int nents, enum dma_data_direction direction,
+				struct dma_attrs *attrs)
+{
+}
+
+static int dma_direct_dma_supported(struct device *dev, u64 mask)
+{
+	return 1;
+}
+
+static inline dma_addr_t dma_direct_map_page(struct device *dev,
+					     struct page *page,
+					     unsigned long offset,
+					     size_t size,
+					     enum dma_data_direction direction,
+					     struct dma_attrs *attrs)
+{
+	__dma_sync_page(page_to_phys(page), offset, size, direction);
+	return page_to_phys(page) + offset + get_dma_direct_offset(dev);
+}
+
+static inline void dma_direct_unmap_page(struct device *dev,
+					 dma_addr_t dma_address,
+					 size_t size,
+					 enum dma_data_direction direction,
+					 struct dma_attrs *attrs)
+{
+/* There is not necessary to do cache cleanup
+ *
+ * phys_to_virt is here because in __dma_sync_page is __virt_to_phys and
+ * dma_address is physical address
+ */
+	__dma_sync_page(dma_address, 0 , size, direction);
+}
+
+struct dma_map_ops dma_direct_ops = {
+	.alloc_coherent	= dma_direct_alloc_coherent,
+	.free_coherent	= dma_direct_free_coherent,
+	.map_sg		= dma_direct_map_sg,
+	.unmap_sg	= dma_direct_unmap_sg,
+	.dma_supported	= dma_direct_dma_supported,
+	.map_page	= dma_direct_map_page,
+	.unmap_page	= dma_direct_unmap_page,
+};
+EXPORT_SYMBOL(dma_direct_ops);
+
+/* Number of entries preallocated for DMA-API debugging */
+#define PREALLOC_DMA_DEBUG_ENTRIES (1 << 16)
+
+static int __init dma_init(void)
+{
+       dma_debug_init(PREALLOC_DMA_DEBUG_ENTRIES);
+
+       return 0;
+}
+fs_initcall(dma_init);
diff --git a/arch/microblaze/kernel/entry.S b/arch/microblaze/kernel/entry.S
index 3bad4ff..c0ede25 100644
--- a/arch/microblaze/kernel/entry.S
+++ b/arch/microblaze/kernel/entry.S
@@ -305,7 +305,7 @@ C_ENTRY(_user_exception):
 	swi	r11, r1, PTO+PT_R1;		/* Store user SP.  */
 	addi	r11, r0, 1;
 	swi	r11, r0, TOPHYS(PER_CPU(KM));	/* Now we're in kernel-mode.  */
-2:	lwi	r31, r0, TOPHYS(PER_CPU(CURRENT_SAVE));	/* get saved current */
+2:	lwi	CURRENT_TASK, r0, TOPHYS(PER_CPU(CURRENT_SAVE));
 	/* Save away the syscall number.  */
 	swi	r12, r1, PTO+PT_R0;
 	tovirt(r1,r1)
@@ -322,8 +322,7 @@ C_ENTRY(_user_exception):
 	rtid	r11, 0
 	nop
 3:
-	add	r11, r0, CURRENT_TASK	 /* Get current task ptr into r11 */
-	lwi	r11, r11, TS_THREAD_INFO /* get thread info */
+	lwi	r11, CURRENT_TASK, TS_THREAD_INFO /* get thread info */
 	lwi	r11, r11, TI_FLAGS	 /* get flags in thread info */
 	andi	r11, r11, _TIF_WORK_SYSCALL_MASK
 	beqi	r11, 4f
@@ -382,60 +381,50 @@ C_ENTRY(ret_from_trap):
 /* See if returning to kernel mode, if so, skip resched &c.  */
 	bnei	r11, 2f;
 
+	swi	r3, r1, PTO + PT_R3
+	swi	r4, r1, PTO + PT_R4
+
 	/* We're returning to user mode, so check for various conditions that
 	 * trigger rescheduling. */
-	# FIXME: Restructure all these flag checks.
-	add	r11, r0, CURRENT_TASK;	/* Get current task ptr into r11 */
-	lwi	r11, r11, TS_THREAD_INFO;	/* get thread info */
+	/* FIXME: Restructure all these flag checks. */
+	lwi	r11, CURRENT_TASK, TS_THREAD_INFO;	/* get thread info */
 	lwi	r11, r11, TI_FLAGS;		/* get flags in thread info */
 	andi	r11, r11, _TIF_WORK_SYSCALL_MASK
 	beqi	r11, 1f
 
-	swi	r3, r1, PTO + PT_R3
-	swi	r4, r1, PTO + PT_R4
 	brlid	r15, do_syscall_trace_leave
 	addik	r5, r1, PTO + PT_R0
-	lwi	r3, r1, PTO + PT_R3
-	lwi	r4, r1, PTO + PT_R4
 1:
-
 	/* We're returning to user mode, so check for various conditions that
 	 * trigger rescheduling. */
-	/* Get current task ptr into r11 */
-	add	r11, r0, CURRENT_TASK;	/* Get current task ptr into r11 */
-	lwi	r11, r11, TS_THREAD_INFO;	/* get thread info */
+	/* get thread info from current task */
+	lwi	r11, CURRENT_TASK, TS_THREAD_INFO;
 	lwi	r11, r11, TI_FLAGS;		/* get flags in thread info */
 	andi	r11, r11, _TIF_NEED_RESCHED;
 	beqi	r11, 5f;
 
-	swi	r3, r1, PTO + PT_R3; /* store syscall result */
-	swi	r4, r1, PTO + PT_R4;
 	bralid	r15, schedule;	/* Call scheduler */
 	nop;				/* delay slot */
-	lwi	r3, r1, PTO + PT_R3; /* restore syscall result */
-	lwi	r4, r1, PTO + PT_R4;
 
 	/* Maybe handle a signal */
-5:	add	r11, r0, CURRENT_TASK; /* Get current task ptr into r11 */
-	lwi	r11, r11, TS_THREAD_INFO;	/* get thread info */
+5:	/* get thread info from current task*/
+	lwi	r11, CURRENT_TASK, TS_THREAD_INFO;
 	lwi	r11, r11, TI_FLAGS;	/* get flags in thread info */
 	andi	r11, r11, _TIF_SIGPENDING;
 	beqi	r11, 1f;		/* Signals to handle, handle them */
 
-	swi	r3, r1, PTO + PT_R3; /* store syscall result */
-	swi	r4, r1, PTO + PT_R4;
 	la	r5, r1, PTO;		/* Arg 1: struct pt_regs *regs */
-	add	r6, r0, r0;		/* Arg 2: sigset_t *oldset */
 	addi	r7, r0, 1;		/* Arg 3: int in_syscall */
 	bralid	r15, do_signal;	/* Handle any signals */
-	nop;
+	add	r6, r0, r0;		/* Arg 2: sigset_t *oldset */
+
+/* Finally, return to user state.  */
+1:
 	lwi	r3, r1, PTO + PT_R3; /* restore syscall result */
 	lwi	r4, r1, PTO + PT_R4;
 
-/* Finally, return to user state.  */
-1:	swi	r0, r0, PER_CPU(KM);	/* Now officially in user state. */
-	add	r11, r0, CURRENT_TASK;	/* Get current task ptr into r11 */
-	swi	r11, r0, PER_CPU(CURRENT_SAVE); /* save current */
+	swi	r0, r0, PER_CPU(KM);	/* Now officially in user state. */
+	swi	CURRENT_TASK, r0, PER_CPU(CURRENT_SAVE); /* save current */
 	VM_OFF;
 	tophys(r1,r1);
 	RESTORE_REGS;
@@ -565,7 +554,7 @@ C_ENTRY(sys_rt_sigreturn_wrapper):
 	swi	r11, r1, PTO+PT_R1; /* Store user SP.  */		\
 	addi	r11, r0, 1;						\
 	swi	r11, r0, TOPHYS(PER_CPU(KM)); /* Now we're in kernel-mode.*/\
-2:	lwi	r31, r0, TOPHYS(PER_CPU(CURRENT_SAVE)); /* get saved current */\
+2:	lwi	CURRENT_TASK, r0, TOPHYS(PER_CPU(CURRENT_SAVE));	\
 	/* Save away the syscall number.  */				\
 	swi	r0, r1, PTO+PT_R0;					\
 	tovirt(r1,r1)
@@ -673,9 +662,7 @@ C_ENTRY(ret_from_exc):
 
 	/* We're returning to user mode, so check for various conditions that
 	   trigger rescheduling. */
-	/* Get current task ptr into r11 */
-	add	r11, r0, CURRENT_TASK; /* Get current task ptr into r11 */
-	lwi	r11, r11, TS_THREAD_INFO;	/* get thread info */
+	lwi	r11, CURRENT_TASK, TS_THREAD_INFO;	/* get thread info */
 	lwi	r11, r11, TI_FLAGS;	/* get flags in thread info */
 	andi	r11, r11, _TIF_NEED_RESCHED;
 	beqi	r11, 5f;
@@ -685,8 +672,7 @@ C_ENTRY(ret_from_exc):
 	nop;				/* delay slot */
 
 	/* Maybe handle a signal */
-5:	add	r11, r0, CURRENT_TASK; /* Get current task ptr into r11 */
-	lwi	r11, r11, TS_THREAD_INFO;	/* get thread info */
+5:	lwi	r11, CURRENT_TASK, TS_THREAD_INFO;	/* get thread info */
 	lwi	r11, r11, TI_FLAGS;	/* get flags in thread info */
 	andi	r11, r11, _TIF_SIGPENDING;
 	beqi	r11, 1f;		/* Signals to handle, handle them */
@@ -705,15 +691,13 @@ C_ENTRY(ret_from_exc):
 	 * store return registers separately because this macros is use
 	 * for others exceptions */
 	la	r5, r1, PTO;		/* Arg 1: struct pt_regs *regs */
-	add	r6, r0, r0;		/* Arg 2: sigset_t *oldset */
 	addi	r7, r0, 0;		/* Arg 3: int in_syscall */
 	bralid	r15, do_signal;	/* Handle any signals */
-	nop;
+	add	r6, r0, r0;		/* Arg 2: sigset_t *oldset */
 
 /* Finally, return to user state.  */
 1:	swi	r0, r0, PER_CPU(KM);	/* Now officially in user state. */
-	add	r11, r0, CURRENT_TASK; /* Get current task ptr into r11 */
-	swi	r11, r0, PER_CPU(CURRENT_SAVE); /* save current */
+	swi	CURRENT_TASK, r0, PER_CPU(CURRENT_SAVE); /* save current */
 	VM_OFF;
 	tophys(r1,r1);
 
@@ -802,7 +786,7 @@ C_ENTRY(_interrupt):
 	swi	r11, r0, TOPHYS(PER_CPU(KM));
 
 2:
-	lwi	r31, r0, TOPHYS(PER_CPU(CURRENT_SAVE));
+	lwi	CURRENT_TASK, r0, TOPHYS(PER_CPU(CURRENT_SAVE));
 	swi	r0, r1, PTO + PT_R0;
 	tovirt(r1,r1)
 	la	r5, r1, PTO;
@@ -817,8 +801,7 @@ ret_from_irq:
 	lwi	r11, r1, PTO + PT_MODE;
 	bnei	r11, 2f;
 
-	add	r11, r0, CURRENT_TASK;
-	lwi	r11, r11, TS_THREAD_INFO;
+	lwi	r11, CURRENT_TASK, TS_THREAD_INFO;
 	lwi	r11, r11, TI_FLAGS; /* MS: get flags from thread info */
 	andi	r11, r11, _TIF_NEED_RESCHED;
 	beqi	r11, 5f
@@ -826,8 +809,7 @@ ret_from_irq:
 	nop; /* delay slot */
 
     /* Maybe handle a signal */
-5:	add	r11, r0, CURRENT_TASK;
-	lwi	r11, r11, TS_THREAD_INFO; /* MS: get thread info */
+5:	lwi	r11, CURRENT_TASK, TS_THREAD_INFO; /* MS: get thread info */
 	lwi	r11, r11, TI_FLAGS; /* get flags in thread info */
 	andi	r11, r11, _TIF_SIGPENDING;
 	beqid	r11, no_intr_resched
@@ -842,8 +824,7 @@ no_intr_resched:
     /* Disable interrupts, we are now committed to the state restore */
 	disable_irq
 	swi	r0, r0, PER_CPU(KM); /* MS: Now officially in user state. */
-	add	r11, r0, CURRENT_TASK;
-	swi	r11, r0, PER_CPU(CURRENT_SAVE);
+	swi	CURRENT_TASK, r0, PER_CPU(CURRENT_SAVE);
 	VM_OFF;
 	tophys(r1,r1);
 	lwi	r3, r1, PTO + PT_R3; /* MS: restore saved r3, r4 registers */
@@ -853,7 +834,28 @@ no_intr_resched:
 	lwi	r1, r1, PT_R1 - PT_SIZE;
 	bri	6f;
 /* MS: Return to kernel state. */
-2:	VM_OFF /* MS: turn off MMU */
+2:
+#ifdef CONFIG_PREEMPT
+	lwi	r11, CURRENT_TASK, TS_THREAD_INFO;
+	/* MS: get preempt_count from thread info */
+	lwi	r5, r11, TI_PREEMPT_COUNT;
+	bgti	r5, restore;
+
+	lwi	r5, r11, TI_FLAGS;		/* get flags in thread info */
+	andi	r5, r5, _TIF_NEED_RESCHED;
+	beqi	r5, restore /* if zero jump over */
+
+preempt:
+	/* interrupts are off that's why I am calling preempt_chedule_irq */
+	bralid	r15, preempt_schedule_irq
+	nop
+	lwi	r11, CURRENT_TASK, TS_THREAD_INFO;	/* get thread info */
+	lwi	r5, r11, TI_FLAGS;		/* get flags in thread info */
+	andi	r5, r5, _TIF_NEED_RESCHED;
+	bnei	r5, preempt /* if non zero jump to resched */
+restore:
+#endif
+	VM_OFF /* MS: turn off MMU */
 	tophys(r1,r1)
 	lwi	r3, r1, PTO + PT_R3; /* MS: restore saved r3, r4 registers */
 	lwi	r4, r1, PTO + PT_R4;
@@ -915,7 +917,7 @@ C_ENTRY(_debug_exception):
 	swi	r11, r1, PTO+PT_R1; /* Store user SP.  */
 	addi	r11, r0, 1;
 	swi	r11, r0, TOPHYS(PER_CPU(KM));	/* Now we're in kernel-mode.  */
-2:	lwi	r31, r0, TOPHYS(PER_CPU(CURRENT_SAVE)); /* get saved current */
+2:	lwi	CURRENT_TASK, r0, TOPHYS(PER_CPU(CURRENT_SAVE));
 	/* Save away the syscall number.  */
 	swi	r0, r1, PTO+PT_R0;
 	tovirt(r1,r1)
@@ -935,8 +937,7 @@ dbtrap_call:	rtbd	r11, 0;
 	bnei	r11, 2f;
 
 	/* Get current task ptr into r11 */
-	add	r11, r0, CURRENT_TASK; /* Get current task ptr into r11 */
-	lwi	r11, r11, TS_THREAD_INFO;	/* get thread info */
+	lwi	r11, CURRENT_TASK, TS_THREAD_INFO;	/* get thread info */
 	lwi	r11, r11, TI_FLAGS;	/* get flags in thread info */
 	andi	r11, r11, _TIF_NEED_RESCHED;
 	beqi	r11, 5f;
@@ -949,8 +950,7 @@ dbtrap_call:	rtbd	r11, 0;
 	/* XXX m68knommu also checks TASK_STATE & TASK_COUNTER here.  */
 
 	/* Maybe handle a signal */
-5:	add	r11, r0, CURRENT_TASK; /* Get current task ptr into r11 */
-	lwi	r11, r11, TS_THREAD_INFO;	/* get thread info */
+5:	lwi	r11, CURRENT_TASK, TS_THREAD_INFO;	/* get thread info */
 	lwi	r11, r11, TI_FLAGS;	/* get flags in thread info */
 	andi	r11, r11, _TIF_SIGPENDING;
 	beqi	r11, 1f;		/* Signals to handle, handle them */
@@ -966,16 +966,14 @@ dbtrap_call:	rtbd	r11, 0;
 	   (in a possibly modified form) after do_signal returns.  */
 
 	la	r5, r1, PTO;		/* Arg 1: struct pt_regs *regs */
-	add	r6, r0, r0;		/* Arg 2: sigset_t *oldset */
 	addi  r7, r0, 0;	/* Arg 3: int in_syscall */
 	bralid	r15, do_signal;	/* Handle any signals */
-	nop;
+	add	r6, r0, r0;		/* Arg 2: sigset_t *oldset */
 
 
 /* Finally, return to user state.  */
 1:	swi	r0, r0, PER_CPU(KM);	/* Now officially in user state. */
-	add	r11, r0, CURRENT_TASK; /* Get current task ptr into r11 */
-	swi	r11, r0, PER_CPU(CURRENT_SAVE); /* save current */
+	swi	CURRENT_TASK, r0, PER_CPU(CURRENT_SAVE); /* save current */
 	VM_OFF;
 	tophys(r1,r1);
 
@@ -1007,7 +1005,7 @@ DBTRAP_return:		/* Make global symbol for debugging */
 
 ENTRY(_switch_to)
 	/* prepare return value */
-	addk	r3, r0, r31
+	addk	r3, r0, CURRENT_TASK
 
 	/* save registers in cpu_context */
 	/* use r11 and r12, volatile registers, as temp register */
@@ -1051,10 +1049,10 @@ ENTRY(_switch_to)
 	nop
 	swi	r12, r11, CC_FSR
 
-	/* update r31, the current */
-	lwi	r31, r6, TI_TASK/* give me pointer to task which will be next */
+	/* update r31, the current-give me pointer to task which will be next */
+	lwi	CURRENT_TASK, r6, TI_TASK
 	/* stored it to current_save too */
-	swi	r31, r0, PER_CPU(CURRENT_SAVE)
+	swi	CURRENT_TASK, r0, PER_CPU(CURRENT_SAVE)
 
 	/* get new process' cpu context and restore */
 	/* give me start where start context of next task */
diff --git a/arch/microblaze/kernel/head.S b/arch/microblaze/kernel/head.S
index 3091619..cb7815c 100644
--- a/arch/microblaze/kernel/head.S
+++ b/arch/microblaze/kernel/head.S
@@ -99,8 +99,8 @@ no_fdt_arg:
 	tophys(r4,r4)			/* convert to phys address */
 	ori	r3, r0, COMMAND_LINE_SIZE - 1 /* number of loops */
 _copy_command_line:
-	lbu	r7, r5, r6 /* r7=r5+r6 - r5 contain pointer to command line */
-	sb	r7, r4, r6		/* addr[r4+r6]= r7*/
+	lbu	r2, r5, r6 /* r7=r5+r6 - r5 contain pointer to command line */
+	sb	r2, r4, r6		/* addr[r4+r6]= r7*/
 	addik	r6, r6, 1		/* increment counting */
 	bgtid	r3, _copy_command_line	/* loop for all entries       */
 	addik	r3, r3, -1		/* descrement loop */
@@ -136,6 +136,11 @@ _invalidate:
 	addik	r3, r3, -1
 	/* sync */
 
+	/* Setup the kernel PID */
+	mts	rpid,r0			/* Load the kernel PID */
+	nop
+	bri	4
+
 	/*
 	 * We should still be executing code at physical address area
 	 * RAM_BASEADDR at this point. However, kernel code is at
@@ -146,10 +151,6 @@ _invalidate:
 	addik	r3,r0, CONFIG_KERNEL_START /* Load the kernel virtual address */
 	tophys(r4,r3)			/* Load the kernel physical address */
 
-	mts	rpid,r0			/* Load the kernel PID */
-	nop
-	bri	4
-
 	/*
 	 * Configure and load two entries into TLB slots 0 and 1.
 	 * In case we are pinning TLBs, these are reserved in by the
diff --git a/arch/microblaze/kernel/irq.c b/arch/microblaze/kernel/irq.c
index 0f06034..6f39e2c 100644
--- a/arch/microblaze/kernel/irq.c
+++ b/arch/microblaze/kernel/irq.c
@@ -93,3 +93,18 @@ skip:
 	}
 	return 0;
 }
+
+/* MS: There is no any advance mapping mechanism. We are using simple 32bit
+  intc without any cascades or any connection that's why mapping is 1:1 */
+unsigned int irq_create_mapping(struct irq_host *host, irq_hw_number_t hwirq)
+{
+	return hwirq;
+}
+EXPORT_SYMBOL_GPL(irq_create_mapping);
+
+unsigned int irq_create_of_mapping(struct device_node *controller,
+					u32 *intspec, unsigned int intsize)
+{
+	return intspec[0];
+}
+EXPORT_SYMBOL_GPL(irq_create_of_mapping);
diff --git a/arch/microblaze/kernel/setup.c b/arch/microblaze/kernel/setup.c
index bb8c4b9..f974ec7 100644
--- a/arch/microblaze/kernel/setup.c
+++ b/arch/microblaze/kernel/setup.c
@@ -22,7 +22,10 @@
 #include <linux/io.h>
 #include <linux/bug.h>
 #include <linux/param.h>
+#include <linux/pci.h>
 #include <linux/cache.h>
+#include <linux/of_platform.h>
+#include <linux/dma-mapping.h>
 #include <asm/cacheflush.h>
 #include <asm/entry.h>
 #include <asm/cpuinfo.h>
@@ -54,14 +57,10 @@ void __init setup_arch(char **cmdline_p)
 
 	microblaze_cache_init();
 
-	invalidate_dcache();
-	enable_dcache();
-
-	invalidate_icache();
-	enable_icache();
-
 	setup_memory();
 
+	xilinx_pci_init();
+
 #if defined(CONFIG_SELFMOD_INTC) || defined(CONFIG_SELFMOD_TIMER)
 	printk(KERN_NOTICE "Self modified code enable\n");
 #endif
@@ -188,3 +187,37 @@ static int microblaze_debugfs_init(void)
 }
 arch_initcall(microblaze_debugfs_init);
 #endif
+
+static int dflt_bus_notify(struct notifier_block *nb,
+				unsigned long action, void *data)
+{
+	struct device *dev = data;
+
+	/* We are only intereted in device addition */
+	if (action != BUS_NOTIFY_ADD_DEVICE)
+		return 0;
+
+	set_dma_ops(dev, &dma_direct_ops);
+
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block dflt_plat_bus_notifier = {
+	.notifier_call = dflt_bus_notify,
+	.priority = INT_MAX,
+};
+
+static struct notifier_block dflt_of_bus_notifier = {
+	.notifier_call = dflt_bus_notify,
+	.priority = INT_MAX,
+};
+
+static int __init setup_bus_notifier(void)
+{
+	bus_register_notifier(&platform_bus_type, &dflt_plat_bus_notifier);
+	bus_register_notifier(&of_platform_bus_type, &dflt_of_bus_notifier);
+
+	return 0;
+}
+
+arch_initcall(setup_bus_notifier);