Adding upstream version 4.19.249.upstream/4.19.249upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

7 files changed, 1330 insertions, 0 deletions

diff --git a/arch/microblaze/mm/Makefile b/arch/microblaze/mm/Makefile
new file mode 100644
index 000000000..7313bd8ac
--- /dev/null
+++ b/arch/microblaze/mm/Makefile
@@ -0,0 +1,8 @@
+#
+# Makefile
+#
+
+obj-y := consistent.o init.o
+
+obj-$(CONFIG_MMU) += pgtable.o mmu_context.o fault.o
+obj-$(CONFIG_HIGHMEM) += highmem.o
diff --git a/arch/microblaze/mm/consistent.c b/arch/microblaze/mm/consistent.c
new file mode 100644
index 000000000..c9a278ac7
--- /dev/null
+++ b/arch/microblaze/mm/consistent.c
@@ -0,0 +1,222 @@
+/*
+ * Microblaze support for cache consistent memory.
+ * Copyright (C) 2010 Michal Simek <monstr@monstr.eu>
+ * Copyright (C) 2010 PetaLogix
+ * Copyright (C) 2005 John Williams <jwilliams@itee.uq.edu.au>
+ *
+ * Based on PowerPC version derived from arch/arm/mm/consistent.c
+ * Copyright (C) 2001 Dan Malek (dmalek@jlc.net)
+ * Copyright (C) 2000 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/export.h>
+#include <linux/signal.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/ptrace.h>
+#include <linux/mman.h>
+#include <linux/mm.h>
+#include <linux/swap.h>
+#include <linux/stddef.h>
+#include <linux/vmalloc.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/bootmem.h>
+#include <linux/highmem.h>
+#include <linux/pci.h>
+#include <linux/interrupt.h>
+#include <linux/gfp.h>
+#include <linux/dma-noncoherent.h>
+
+#include <asm/pgalloc.h>
+#include <linux/io.h>
+#include <linux/hardirq.h>
+#include <linux/mmu_context.h>
+#include <asm/mmu.h>
+#include <linux/uaccess.h>
+#include <asm/pgtable.h>
+#include <asm/cpuinfo.h>
+#include <asm/tlbflush.h>
+
+#ifndef CONFIG_MMU
+/* I have to use dcache values because I can't relate on ram size */
+# define UNCACHED_SHADOW_MASK (cpuinfo.dcache_high - cpuinfo.dcache_base + 1)
+#endif
+
+/*
+ * Consistent memory allocators. Used for DMA devices that want to
+ * share uncached memory with the processor core.
+ * My crufty no-MMU approach is simple. In the HW platform we can optionally
+ * mirror the DDR up above the processor cacheable region.  So, memory accessed
+ * in this mirror region will not be cached.  It's alloced from the same
+ * pool as normal memory, but the handle we return is shifted up into the
+ * uncached region.  This will no doubt cause big problems if memory allocated
+ * here is not also freed properly. -- JW
+ */
+void *arch_dma_alloc(struct device *dev, size_t size, dma_addr_t *dma_handle,
+		gfp_t gfp, unsigned long attrs)
+{
+	unsigned long order, vaddr;
+	void *ret;
+	unsigned int i, err = 0;
+	struct page *page, *end;
+
+#ifdef CONFIG_MMU
+	phys_addr_t pa;
+	struct vm_struct *area;
+	unsigned long va;
+#endif
+
+	if (in_interrupt())
+		BUG();
+
+	/* Only allocate page size areas. */
+	size = PAGE_ALIGN(size);
+	order = get_order(size);
+
+	vaddr = __get_free_pages(gfp, order);
+	if (!vaddr)
+		return NULL;
+
+	/*
+	 * we need to ensure that there are no cachelines in use,
+	 * or worse dirty in this area.
+	 */
+	flush_dcache_range(virt_to_phys((void *)vaddr),
+					virt_to_phys((void *)vaddr) + size);
+
+#ifndef CONFIG_MMU
+	ret = (void *)vaddr;
+	/*
+	 * Here's the magic!  Note if the uncached shadow is not implemented,
+	 * it's up to the calling code to also test that condition and make
+	 * other arranegments, such as manually flushing the cache and so on.
+	 */
+# ifdef CONFIG_XILINX_UNCACHED_SHADOW
+	ret = (void *)((unsigned) ret | UNCACHED_SHADOW_MASK);
+# endif
+	if ((unsigned int)ret > cpuinfo.dcache_base &&
+				(unsigned int)ret < cpuinfo.dcache_high)
+		pr_warn("ERROR: Your cache coherent area is CACHED!!!\n");
+
+	/* dma_handle is same as physical (shadowed) address */
+	*dma_handle = (dma_addr_t)ret;
+#else
+	/* Allocate some common virtual space to map the new pages. */
+	area = get_vm_area(size, VM_ALLOC);
+	if (!area) {
+		free_pages(vaddr, order);
+		return NULL;
+	}
+	va = (unsigned long) area->addr;
+	ret = (void *)va;
+
+	/* This gives us the real physical address of the first page. */
+	*dma_handle = pa = __virt_to_phys(vaddr);
+#endif
+
+	/*
+	 * free wasted pages.  We skip the first page since we know
+	 * that it will have count = 1 and won't require freeing.
+	 * We also mark the pages in use as reserved so that
+	 * remap_page_range works.
+	 */
+	page = virt_to_page(vaddr);
+	end = page + (1 << order);
+
+	split_page(page, order);
+
+	for (i = 0; i < size && err == 0; i += PAGE_SIZE) {
+#ifdef CONFIG_MMU
+		/* MS: This is the whole magic - use cache inhibit pages */
+		err = map_page(va + i, pa + i, _PAGE_KERNEL | _PAGE_NO_CACHE);
+#endif
+
+		SetPageReserved(page);
+		page++;
+	}
+
+	/* Free the otherwise unused pages. */
+	while (page < end) {
+		__free_page(page);
+		page++;
+	}
+
+	if (err) {
+		free_pages(vaddr, order);
+		return NULL;
+	}
+
+	return ret;
+}
+
+#ifdef CONFIG_MMU
+static pte_t *consistent_virt_to_pte(void *vaddr)
+{
+	unsigned long addr = (unsigned long)vaddr;
+
+	return pte_offset_kernel(pmd_offset(pgd_offset_k(addr), addr), addr);
+}
+
+unsigned long consistent_virt_to_pfn(void *vaddr)
+{
+	pte_t *ptep = consistent_virt_to_pte(vaddr);
+
+	if (pte_none(*ptep) || !pte_present(*ptep))
+		return 0;
+
+	return pte_pfn(*ptep);
+}
+#endif
+
+/*
+ * free page(s) as defined by the above mapping.
+ */
+void arch_dma_free(struct device *dev, size_t size, void *vaddr,
+		dma_addr_t dma_addr, unsigned long attrs)
+{
+	struct page *page;
+
+	if (in_interrupt())
+		BUG();
+
+	size = PAGE_ALIGN(size);
+
+#ifndef CONFIG_MMU
+	/* Clear SHADOW_MASK bit in address, and free as per usual */
+# ifdef CONFIG_XILINX_UNCACHED_SHADOW
+	vaddr = (void *)((unsigned)vaddr & ~UNCACHED_SHADOW_MASK);
+# endif
+	page = virt_to_page(vaddr);
+
+	do {
+		__free_reserved_page(page);
+		page++;
+	} while (size -= PAGE_SIZE);
+#else
+	do {
+		pte_t *ptep = consistent_virt_to_pte(vaddr);
+		unsigned long pfn;
+
+		if (!pte_none(*ptep) && pte_present(*ptep)) {
+			pfn = pte_pfn(*ptep);
+			pte_clear(&init_mm, (unsigned int)vaddr, ptep);
+			if (pfn_valid(pfn)) {
+				page = pfn_to_page(pfn);
+				__free_reserved_page(page);
+			}
+		}
+		vaddr += PAGE_SIZE;
+	} while (size -= PAGE_SIZE);
+
+	/* flush tlb */
+	flush_tlb_all();
+#endif
+}
diff --git a/arch/microblaze/mm/fault.c b/arch/microblaze/mm/fault.c
new file mode 100644
index 000000000..202ad6a49
--- /dev/null
+++ b/arch/microblaze/mm/fault.c
@@ -0,0 +1,296 @@
+/*
+ *  arch/microblaze/mm/fault.c
+ *
+ *    Copyright (C) 2007 Xilinx, Inc.  All rights reserved.
+ *
+ *  Derived from "arch/ppc/mm/fault.c"
+ *    Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org)
+ *
+ *  Derived from "arch/i386/mm/fault.c"
+ *    Copyright (C) 1991, 1992, 1993, 1994  Linus Torvalds
+ *
+ *  Modified by Cort Dougan and Paul Mackerras.
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file COPYING in the main directory of this
+ * archive for more details.
+ *
+ */
+
+#include <linux/extable.h>
+#include <linux/signal.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/ptrace.h>
+#include <linux/mman.h>
+#include <linux/mm.h>
+#include <linux/interrupt.h>
+
+#include <asm/page.h>
+#include <asm/pgtable.h>
+#include <asm/mmu.h>
+#include <linux/mmu_context.h>
+#include <linux/uaccess.h>
+#include <asm/exceptions.h>
+
+static unsigned long pte_misses;	/* updated by do_page_fault() */
+static unsigned long pte_errors;	/* updated by do_page_fault() */
+
+/*
+ * Check whether the instruction at regs->pc is a store using
+ * an update addressing form which will update r1.
+ */
+static int store_updates_sp(struct pt_regs *regs)
+{
+	unsigned int inst;
+
+	if (get_user(inst, (unsigned int __user *)regs->pc))
+		return 0;
+	/* check for 1 in the rD field */
+	if (((inst >> 21) & 0x1f) != 1)
+		return 0;
+	/* check for store opcodes */
+	if ((inst & 0xd0000000) == 0xd0000000)
+		return 1;
+	return 0;
+}
+
+
+/*
+ * bad_page_fault is called when we have a bad access from the kernel.
+ * It is called from do_page_fault above and from some of the procedures
+ * in traps.c.
+ */
+void bad_page_fault(struct pt_regs *regs, unsigned long address, int sig)
+{
+	const struct exception_table_entry *fixup;
+/* MS: no context */
+	/* Are we prepared to handle this fault?  */
+	fixup = search_exception_tables(regs->pc);
+	if (fixup) {
+		regs->pc = fixup->fixup;
+		return;
+	}
+
+	/* kernel has accessed a bad area */
+	die("kernel access of bad area", regs, sig);
+}
+
+/*
+ * The error_code parameter is ESR for a data fault,
+ * 0 for an instruction fault.
+ */
+void do_page_fault(struct pt_regs *regs, unsigned long address,
+		   unsigned long error_code)
+{
+	struct vm_area_struct *vma;
+	struct mm_struct *mm = current->mm;
+	int code = SEGV_MAPERR;
+	int is_write = error_code & ESR_S;
+	vm_fault_t fault;
+	unsigned int flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;
+
+	regs->ear = address;
+	regs->esr = error_code;
+
+	/* On a kernel SLB miss we can only check for a valid exception entry */
+	if (unlikely(kernel_mode(regs) && (address >= TASK_SIZE))) {
+		pr_warn("kernel task_size exceed");
+		_exception(SIGSEGV, regs, code, address);
+	}
+
+	/* for instr TLB miss and instr storage exception ESR_S is undefined */
+	if ((error_code & 0x13) == 0x13 || (error_code & 0x11) == 0x11)
+		is_write = 0;
+
+	if (unlikely(faulthandler_disabled() || !mm)) {
+		if (kernel_mode(regs))
+			goto bad_area_nosemaphore;
+
+		/* faulthandler_disabled() in user mode is really bad,
+		   as is current->mm == NULL. */
+		pr_emerg("Page fault in user mode with faulthandler_disabled(), mm = %p\n",
+			 mm);
+		pr_emerg("r15 = %lx  MSR = %lx\n",
+		       regs->r15, regs->msr);
+		die("Weird page fault", regs, SIGSEGV);
+	}
+
+	if (user_mode(regs))
+		flags |= FAULT_FLAG_USER;
+
+	/* When running in the kernel we expect faults to occur only to
+	 * addresses in user space.  All other faults represent errors in the
+	 * kernel and should generate an OOPS.  Unfortunately, in the case of an
+	 * erroneous fault occurring in a code path which already holds mmap_sem
+	 * we will deadlock attempting to validate the fault against the
+	 * address space.  Luckily the kernel only validly references user
+	 * space from well defined areas of code, which are listed in the
+	 * exceptions table.
+	 *
+	 * As the vast majority of faults will be valid we will only perform
+	 * the source reference check when there is a possibility of a deadlock.
+	 * Attempt to lock the address space, if we cannot we then validate the
+	 * source.  If this is invalid we can skip the address space check,
+	 * thus avoiding the deadlock.
+	 */
+	if (unlikely(!down_read_trylock(&mm->mmap_sem))) {
+		if (kernel_mode(regs) && !search_exception_tables(regs->pc))
+			goto bad_area_nosemaphore;
+
+retry:
+		down_read(&mm->mmap_sem);
+	}
+
+	vma = find_vma(mm, address);
+	if (unlikely(!vma))
+		goto bad_area;
+
+	if (vma->vm_start <= address)
+		goto good_area;
+
+	if (unlikely(!(vma->vm_flags & VM_GROWSDOWN)))
+		goto bad_area;
+
+	if (unlikely(!is_write))
+		goto bad_area;
+
+	/*
+	 * N.B. The ABI allows programs to access up to
+	 * a few hundred bytes below the stack pointer (TBD).
+	 * The kernel signal delivery code writes up to about 1.5kB
+	 * below the stack pointer (r1) before decrementing it.
+	 * The exec code can write slightly over 640kB to the stack
+	 * before setting the user r1.  Thus we allow the stack to
+	 * expand to 1MB without further checks.
+	 */
+	if (unlikely(address + 0x100000 < vma->vm_end)) {
+
+		/* get user regs even if this fault is in kernel mode */
+		struct pt_regs *uregs = current->thread.regs;
+		if (uregs == NULL)
+			goto bad_area;
+
+		/*
+		 * A user-mode access to an address a long way below
+		 * the stack pointer is only valid if the instruction
+		 * is one which would update the stack pointer to the
+		 * address accessed if the instruction completed,
+		 * i.e. either stwu rs,n(r1) or stwux rs,r1,rb
+		 * (or the byte, halfword, float or double forms).
+		 *
+		 * If we don't check this then any write to the area
+		 * between the last mapped region and the stack will
+		 * expand the stack rather than segfaulting.
+		 */
+		if (address + 2048 < uregs->r1
+			&& (kernel_mode(regs) || !store_updates_sp(regs)))
+				goto bad_area;
+	}
+	if (expand_stack(vma, address))
+		goto bad_area;
+
+good_area:
+	code = SEGV_ACCERR;
+
+	/* a write */
+	if (unlikely(is_write)) {
+		if (unlikely(!(vma->vm_flags & VM_WRITE)))
+			goto bad_area;
+		flags |= FAULT_FLAG_WRITE;
+	/* a read */
+	} else {
+		/* protection fault */
+		if (unlikely(error_code & 0x08000000))
+			goto bad_area;
+		if (unlikely(!(vma->vm_flags & (VM_READ | VM_EXEC))))
+			goto bad_area;
+	}
+
+	/*
+	 * If for any reason at all we couldn't handle the fault,
+	 * make sure we exit gracefully rather than endlessly redo
+	 * the fault.
+	 */
+	fault = handle_mm_fault(vma, address, flags);
+
+	if ((fault & VM_FAULT_RETRY) && fatal_signal_pending(current))
+		return;
+
+	if (unlikely(fault & VM_FAULT_ERROR)) {
+		if (fault & VM_FAULT_OOM)
+			goto out_of_memory;
+		else if (fault & VM_FAULT_SIGSEGV)
+			goto bad_area;
+		else if (fault & VM_FAULT_SIGBUS)
+			goto do_sigbus;
+		BUG();
+	}
+
+	if (flags & FAULT_FLAG_ALLOW_RETRY) {
+		if (unlikely(fault & VM_FAULT_MAJOR))
+			current->maj_flt++;
+		else
+			current->min_flt++;
+		if (fault & VM_FAULT_RETRY) {
+			flags &= ~FAULT_FLAG_ALLOW_RETRY;
+			flags |= FAULT_FLAG_TRIED;
+
+			/*
+			 * No need to up_read(&mm->mmap_sem) as we would
+			 * have already released it in __lock_page_or_retry
+			 * in mm/filemap.c.
+			 */
+
+			goto retry;
+		}
+	}
+
+	up_read(&mm->mmap_sem);
+
+	/*
+	 * keep track of tlb+htab misses that are good addrs but
+	 * just need pte's created via handle_mm_fault()
+	 * -- Cort
+	 */
+	pte_misses++;
+	return;
+
+bad_area:
+	up_read(&mm->mmap_sem);
+
+bad_area_nosemaphore:
+	pte_errors++;
+
+	/* User mode accesses cause a SIGSEGV */
+	if (user_mode(regs)) {
+		_exception(SIGSEGV, regs, code, address);
+		return;
+	}
+
+	bad_page_fault(regs, address, SIGSEGV);
+	return;
+
+/*
+ * We ran out of memory, or some other thing happened to us that made
+ * us unable to handle the page fault gracefully.
+ */
+out_of_memory:
+	up_read(&mm->mmap_sem);
+	if (!user_mode(regs))
+		bad_page_fault(regs, address, SIGKILL);
+	else
+		pagefault_out_of_memory();
+	return;
+
+do_sigbus:
+	up_read(&mm->mmap_sem);
+	if (user_mode(regs)) {
+		force_sig_fault(SIGBUS, BUS_ADRERR, (void __user *)address, current);
+		return;
+	}
+	bad_page_fault(regs, address, SIGBUS);
+}
diff --git a/arch/microblaze/mm/highmem.c b/arch/microblaze/mm/highmem.c
new file mode 100644
index 000000000..d7569f77f
--- /dev/null
+++ b/arch/microblaze/mm/highmem.c
@@ -0,0 +1,89 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * highmem.c: virtual kernel memory mappings for high memory
+ *
+ * PowerPC version, stolen from the i386 version.
+ *
+ * Used in CONFIG_HIGHMEM systems for memory pages which
+ * are not addressable by direct kernel virtual addresses.
+ *
+ * Copyright (C) 1999 Gerhard Wichert, Siemens AG
+ *		      Gerhard.Wichert@pdb.siemens.de
+ *
+ *
+ * Redesigned the x86 32-bit VM architecture to deal with
+ * up to 16 Terrabyte physical memory. With current x86 CPUs
+ * we now support up to 64 Gigabytes physical RAM.
+ *
+ * Copyright (C) 1999 Ingo Molnar <mingo@redhat.com>
+ *
+ * Reworked for PowerPC by various contributors. Moved from
+ * highmem.h by Benjamin Herrenschmidt (c) 2009 IBM Corp.
+ */
+
+#include <linux/export.h>
+#include <linux/highmem.h>
+
+/*
+ * The use of kmap_atomic/kunmap_atomic is discouraged - kmap/kunmap
+ * gives a more generic (and caching) interface. But kmap_atomic can
+ * be used in IRQ contexts, so in some (very limited) cases we need
+ * it.
+ */
+#include <asm/tlbflush.h>
+
+void *kmap_atomic_prot(struct page *page, pgprot_t prot)
+{
+
+	unsigned long vaddr;
+	int idx, type;
+
+	preempt_disable();
+	pagefault_disable();
+	if (!PageHighMem(page))
+		return page_address(page);
+
+
+	type = kmap_atomic_idx_push();
+	idx = type + KM_TYPE_NR*smp_processor_id();
+	vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
+#ifdef CONFIG_DEBUG_HIGHMEM
+	BUG_ON(!pte_none(*(kmap_pte-idx)));
+#endif
+	set_pte_at(&init_mm, vaddr, kmap_pte-idx, mk_pte(page, prot));
+	local_flush_tlb_page(NULL, vaddr);
+
+	return (void *) vaddr;
+}
+EXPORT_SYMBOL(kmap_atomic_prot);
+
+void __kunmap_atomic(void *kvaddr)
+{
+	unsigned long vaddr = (unsigned long) kvaddr & PAGE_MASK;
+	int type;
+	unsigned int idx;
+
+	if (vaddr < __fix_to_virt(FIX_KMAP_END)) {
+		pagefault_enable();
+		preempt_enable();
+		return;
+	}
+
+	type = kmap_atomic_idx();
+
+	idx = type + KM_TYPE_NR * smp_processor_id();
+#ifdef CONFIG_DEBUG_HIGHMEM
+	BUG_ON(vaddr != __fix_to_virt(FIX_KMAP_BEGIN + idx));
+#endif
+	/*
+	 * force other mappings to Oops if they'll try to access
+	 * this pte without first remap it
+	 */
+	pte_clear(&init_mm, vaddr, kmap_pte-idx);
+	local_flush_tlb_page(NULL, vaddr);
+
+	kmap_atomic_idx_pop();
+	pagefault_enable();
+	preempt_enable();
+}
+EXPORT_SYMBOL(__kunmap_atomic);
diff --git a/arch/microblaze/mm/init.c b/arch/microblaze/mm/init.c
new file mode 100644
index 000000000..df6de7ccd
--- /dev/null
+++ b/arch/microblaze/mm/init.c
@@ -0,0 +1,385 @@
+/*
+ * Copyright (C) 2007-2008 Michal Simek <monstr@monstr.eu>
+ * Copyright (C) 2006 Atmark Techno, Inc.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License. See the file "COPYING" in the main directory of this archive
+ * for more details.
+ */
+
+#include <linux/bootmem.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/memblock.h>
+#include <linux/mm.h> /* mem_init */
+#include <linux/initrd.h>
+#include <linux/pagemap.h>
+#include <linux/pfn.h>
+#include <linux/slab.h>
+#include <linux/swap.h>
+#include <linux/export.h>
+
+#include <asm/page.h>
+#include <asm/mmu_context.h>
+#include <asm/pgalloc.h>
+#include <asm/sections.h>
+#include <asm/tlb.h>
+#include <asm/fixmap.h>
+
+/* Use for MMU and noMMU because of PCI generic code */
+int mem_init_done;
+
+#ifndef CONFIG_MMU
+unsigned int __page_offset;
+EXPORT_SYMBOL(__page_offset);
+#endif /* CONFIG_MMU */
+
+char *klimit = _end;
+
+/*
+ * Initialize the bootmem system and give it all the memory we
+ * have available.
+ */
+unsigned long memory_start;
+EXPORT_SYMBOL(memory_start);
+unsigned long memory_size;
+EXPORT_SYMBOL(memory_size);
+unsigned long lowmem_size;
+
+#ifdef CONFIG_HIGHMEM
+pte_t *kmap_pte;
+EXPORT_SYMBOL(kmap_pte);
+pgprot_t kmap_prot;
+EXPORT_SYMBOL(kmap_prot);
+
+static inline pte_t *virt_to_kpte(unsigned long vaddr)
+{
+	return pte_offset_kernel(pmd_offset(pgd_offset_k(vaddr),
+			vaddr), vaddr);
+}
+
+static void __init highmem_init(void)
+{
+	pr_debug("%x\n", (u32)PKMAP_BASE);
+	map_page(PKMAP_BASE, 0, 0);	/* XXX gross */
+	pkmap_page_table = virt_to_kpte(PKMAP_BASE);
+
+	kmap_pte = virt_to_kpte(__fix_to_virt(FIX_KMAP_BEGIN));
+	kmap_prot = PAGE_KERNEL;
+}
+
+static void highmem_setup(void)
+{
+	unsigned long pfn;
+
+	for (pfn = max_low_pfn; pfn < max_pfn; ++pfn) {
+		struct page *page = pfn_to_page(pfn);
+
+		/* FIXME not sure about */
+		if (!memblock_is_reserved(pfn << PAGE_SHIFT))
+			free_highmem_page(page);
+	}
+}
+#endif /* CONFIG_HIGHMEM */
+
+/*
+ * paging_init() sets up the page tables - in fact we've already done this.
+ */
+static void __init paging_init(void)
+{
+	unsigned long zones_size[MAX_NR_ZONES];
+#ifdef CONFIG_MMU
+	int idx;
+
+	/* Setup fixmaps */
+	for (idx = 0; idx < __end_of_fixed_addresses; idx++)
+		clear_fixmap(idx);
+#endif
+
+	/* Clean every zones */
+	memset(zones_size, 0, sizeof(zones_size));
+
+#ifdef CONFIG_HIGHMEM
+	highmem_init();
+
+	zones_size[ZONE_DMA] = max_low_pfn;
+	zones_size[ZONE_HIGHMEM] = max_pfn;
+#else
+	zones_size[ZONE_DMA] = max_pfn;
+#endif
+
+	/* We don't have holes in memory map */
+	free_area_init_nodes(zones_size);
+}
+
+void __init setup_memory(void)
+{
+	struct memblock_region *reg;
+
+#ifndef CONFIG_MMU
+	u32 kernel_align_start, kernel_align_size;
+
+	/* Find main memory where is the kernel */
+	for_each_memblock(memory, reg) {
+		memory_start = (u32)reg->base;
+		lowmem_size = reg->size;
+		if ((memory_start <= (u32)_text) &&
+			((u32)_text <= (memory_start + lowmem_size - 1))) {
+			memory_size = lowmem_size;
+			PAGE_OFFSET = memory_start;
+			pr_info("%s: Main mem: 0x%x, size 0x%08x\n",
+				__func__, (u32) memory_start,
+					(u32) memory_size);
+			break;
+		}
+	}
+
+	if (!memory_start || !memory_size) {
+		panic("%s: Missing memory setting 0x%08x, size=0x%08x\n",
+			__func__, (u32) memory_start, (u32) memory_size);
+	}
+
+	/* reservation of region where is the kernel */
+	kernel_align_start = PAGE_DOWN((u32)_text);
+	/* ALIGN can be remove because _end in vmlinux.lds.S is align */
+	kernel_align_size = PAGE_UP((u32)klimit) - kernel_align_start;
+	pr_info("%s: kernel addr:0x%08x-0x%08x size=0x%08x\n",
+		__func__, kernel_align_start, kernel_align_start
+			+ kernel_align_size, kernel_align_size);
+	memblock_reserve(kernel_align_start, kernel_align_size);
+#endif
+	/*
+	 * Kernel:
+	 * start: base phys address of kernel - page align
+	 * end: base phys address of kernel - page align
+	 *
+	 * min_low_pfn - the first page (mm/bootmem.c - node_boot_start)
+	 * max_low_pfn
+	 * max_mapnr - the first unused page (mm/bootmem.c - node_low_pfn)
+	 */
+
+	/* memory start is from the kernel end (aligned) to higher addr */
+	min_low_pfn = memory_start >> PAGE_SHIFT; /* minimum for allocation */
+	/* RAM is assumed contiguous */
+	max_mapnr = memory_size >> PAGE_SHIFT;
+	max_low_pfn = ((u64)memory_start + (u64)lowmem_size) >> PAGE_SHIFT;
+	max_pfn = ((u64)memory_start + (u64)memory_size) >> PAGE_SHIFT;
+
+	pr_info("%s: max_mapnr: %#lx\n", __func__, max_mapnr);
+	pr_info("%s: min_low_pfn: %#lx\n", __func__, min_low_pfn);
+	pr_info("%s: max_low_pfn: %#lx\n", __func__, max_low_pfn);
+	pr_info("%s: max_pfn: %#lx\n", __func__, max_pfn);
+
+	/* Add active regions with valid PFNs */
+	for_each_memblock(memory, reg) {
+		unsigned long start_pfn, end_pfn;
+
+		start_pfn = memblock_region_memory_base_pfn(reg);
+		end_pfn = memblock_region_memory_end_pfn(reg);
+		memblock_set_node(start_pfn << PAGE_SHIFT,
+				  (end_pfn - start_pfn) << PAGE_SHIFT,
+				  &memblock.memory, 0);
+	}
+
+	/* XXX need to clip this if using highmem? */
+	sparse_memory_present_with_active_regions(0);
+
+	paging_init();
+}
+
+#ifdef CONFIG_BLK_DEV_INITRD
+void free_initrd_mem(unsigned long start, unsigned long end)
+{
+	free_reserved_area((void *)start, (void *)end, -1, "initrd");
+}
+#endif
+
+void free_initmem(void)
+{
+	free_initmem_default(-1);
+}
+
+void __init mem_init(void)
+{
+	high_memory = (void *)__va(memory_start + lowmem_size - 1);
+
+	/* this will put all memory onto the freelists */
+	free_all_bootmem();
+#ifdef CONFIG_HIGHMEM
+	highmem_setup();
+#endif
+
+	mem_init_print_info(NULL);
+#ifdef CONFIG_MMU
+	pr_info("Kernel virtual memory layout:\n");
+	pr_info("  * 0x%08lx..0x%08lx  : fixmap\n", FIXADDR_START, FIXADDR_TOP);
+#ifdef CONFIG_HIGHMEM
+	pr_info("  * 0x%08lx..0x%08lx  : highmem PTEs\n",
+		PKMAP_BASE, PKMAP_ADDR(LAST_PKMAP));
+#endif /* CONFIG_HIGHMEM */
+	pr_info("  * 0x%08lx..0x%08lx  : early ioremap\n",
+		ioremap_bot, ioremap_base);
+	pr_info("  * 0x%08lx..0x%08lx  : vmalloc & ioremap\n",
+		(unsigned long)VMALLOC_START, VMALLOC_END);
+#endif
+	mem_init_done = 1;
+}
+
+#ifndef CONFIG_MMU
+int page_is_ram(unsigned long pfn)
+{
+	return __range_ok(pfn, 0);
+}
+#else
+int page_is_ram(unsigned long pfn)
+{
+	return pfn < max_low_pfn;
+}
+
+/*
+ * Check for command-line options that affect what MMU_init will do.
+ */
+static void mm_cmdline_setup(void)
+{
+	unsigned long maxmem = 0;
+	char *p = cmd_line;
+
+	/* Look for mem= option on command line */
+	p = strstr(cmd_line, "mem=");
+	if (p) {
+		p += 4;
+		maxmem = memparse(p, &p);
+		if (maxmem && memory_size > maxmem) {
+			memory_size = maxmem;
+			memblock.memory.regions[0].size = memory_size;
+		}
+	}
+}
+
+/*
+ * MMU_init_hw does the chip-specific initialization of the MMU hardware.
+ */
+static void __init mmu_init_hw(void)
+{
+	/*
+	 * The Zone Protection Register (ZPR) defines how protection will
+	 * be applied to every page which is a member of a given zone. At
+	 * present, we utilize only two of the zones.
+	 * The zone index bits (of ZSEL) in the PTE are used for software
+	 * indicators, except the LSB.  For user access, zone 1 is used,
+	 * for kernel access, zone 0 is used.  We set all but zone 1
+	 * to zero, allowing only kernel access as indicated in the PTE.
+	 * For zone 1, we set a 01 binary (a value of 10 will not work)
+	 * to allow user access as indicated in the PTE.  This also allows
+	 * kernel access as indicated in the PTE.
+	 */
+	__asm__ __volatile__ ("ori r11, r0, 0x10000000;" \
+			"mts rzpr, r11;"
+			: : : "r11");
+}
+
+/*
+ * MMU_init sets up the basic memory mappings for the kernel,
+ * including both RAM and possibly some I/O regions,
+ * and sets up the page tables and the MMU hardware ready to go.
+ */
+
+/* called from head.S */
+asmlinkage void __init mmu_init(void)
+{
+	unsigned int kstart, ksize;
+
+	if (!memblock.reserved.cnt) {
+		pr_emerg("Error memory count\n");
+		machine_restart(NULL);
+	}
+
+	if ((u32) memblock.memory.regions[0].size < 0x400000) {
+		pr_emerg("Memory must be greater than 4MB\n");
+		machine_restart(NULL);
+	}
+
+	if ((u32) memblock.memory.regions[0].size < kernel_tlb) {
+		pr_emerg("Kernel size is greater than memory node\n");
+		machine_restart(NULL);
+	}
+
+	/* Find main memory where the kernel is */
+	memory_start = (u32) memblock.memory.regions[0].base;
+	lowmem_size = memory_size = (u32) memblock.memory.regions[0].size;
+
+	if (lowmem_size > CONFIG_LOWMEM_SIZE) {
+		lowmem_size = CONFIG_LOWMEM_SIZE;
+#ifndef CONFIG_HIGHMEM
+		memory_size = lowmem_size;
+#endif
+	}
+
+	mm_cmdline_setup(); /* FIXME parse args from command line - not used */
+
+	/*
+	 * Map out the kernel text/data/bss from the available physical
+	 * memory.
+	 */
+	kstart = __pa(CONFIG_KERNEL_START); /* kernel start */
+	/* kernel size */
+	ksize = PAGE_ALIGN(((u32)_end - (u32)CONFIG_KERNEL_START));
+	memblock_reserve(kstart, ksize);
+
+#if defined(CONFIG_BLK_DEV_INITRD)
+	/* Remove the init RAM disk from the available memory. */
+	if (initrd_start) {
+		unsigned long size;
+		size = initrd_end - initrd_start;
+		memblock_reserve(__virt_to_phys(initrd_start), size);
+	}
+#endif /* CONFIG_BLK_DEV_INITRD */
+
+	/* Initialize the MMU hardware */
+	mmu_init_hw();
+
+	/* Map in all of RAM starting at CONFIG_KERNEL_START */
+	mapin_ram();
+
+	/* Extend vmalloc and ioremap area as big as possible */
+#ifdef CONFIG_HIGHMEM
+	ioremap_base = ioremap_bot = PKMAP_BASE;
+#else
+	ioremap_base = ioremap_bot = FIXADDR_START;
+#endif
+
+	/* Initialize the context management stuff */
+	mmu_context_init();
+
+	/* Shortly after that, the entire linear mapping will be available */
+	/* This will also cause that unflatten device tree will be allocated
+	 * inside 768MB limit */
+	memblock_set_current_limit(memory_start + lowmem_size - 1);
+}
+
+/* This is only called until mem_init is done. */
+void __init *early_get_page(void)
+{
+	/*
+	 * Mem start + kernel_tlb -> here is limit
+	 * because of mem mapping from head.S
+	 */
+	return __va(memblock_alloc_base(PAGE_SIZE, PAGE_SIZE,
+				memory_start + kernel_tlb));
+}
+
+#endif /* CONFIG_MMU */
+
+void * __ref zalloc_maybe_bootmem(size_t size, gfp_t mask)
+{
+	void *p;
+
+	if (mem_init_done)
+		p = kzalloc(size, mask);
+	else {
+		p = alloc_bootmem(size);
+		if (p)
+			memset(p, 0, size);
+	}
+	return p;
+}
diff --git a/arch/microblaze/mm/mmu_context.c b/arch/microblaze/mm/mmu_context.c
new file mode 100644
index 000000000..26ff82f4f
--- /dev/null
+++ b/arch/microblaze/mm/mmu_context.c
@@ -0,0 +1,70 @@
+/*
+ * This file contains the routines for handling the MMU.
+ *
+ *    Copyright (C) 2007 Xilinx, Inc.  All rights reserved.
+ *
+ *  Derived from arch/ppc/mm/4xx_mmu.c:
+ *  -- paulus
+ *
+ *  Derived from arch/ppc/mm/init.c:
+ *    Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org)
+ *
+ *  Modifications by Paul Mackerras (PowerMac) (paulus@cs.anu.edu.au)
+ *  and Cort Dougan (PReP) (cort@cs.nmt.edu)
+ *    Copyright (C) 1996 Paul Mackerras
+ *  Amiga/APUS changes by Jesper Skov (jskov@cygnus.co.uk).
+ *
+ *  Derived from "arch/i386/mm/init.c"
+ *    Copyright (C) 1991, 1992, 1993, 1994  Linus Torvalds
+ *
+ *  This program is free software; you can redistribute it and/or
+ *  modify it under the terms of the GNU General Public License
+ *  as published by the Free Software Foundation; either version
+ *  2 of the License, or (at your option) any later version.
+ *
+ */
+
+#include <linux/mm.h>
+#include <linux/init.h>
+
+#include <asm/tlbflush.h>
+#include <asm/mmu_context.h>
+
+mm_context_t next_mmu_context;
+unsigned long context_map[LAST_CONTEXT / BITS_PER_LONG + 1];
+atomic_t nr_free_contexts;
+struct mm_struct *context_mm[LAST_CONTEXT+1];
+
+/*
+ * Initialize the context management stuff.
+ */
+void __init mmu_context_init(void)
+{
+	/*
+	 * The use of context zero is reserved for the kernel.
+	 * This code assumes FIRST_CONTEXT < 32.
+	 */
+	context_map[0] = (1 << FIRST_CONTEXT) - 1;
+	next_mmu_context = FIRST_CONTEXT;
+	atomic_set(&nr_free_contexts, LAST_CONTEXT - FIRST_CONTEXT + 1);
+}
+
+/*
+ * Steal a context from a task that has one at the moment.
+ *
+ * This isn't an LRU system, it just frees up each context in
+ * turn (sort-of pseudo-random replacement :).  This would be the
+ * place to implement an LRU scheme if anyone were motivated to do it.
+ */
+void steal_context(void)
+{
+	struct mm_struct *mm;
+
+	/* free up context `next_mmu_context' */
+	/* if we shouldn't free context 0, don't... */
+	if (next_mmu_context < FIRST_CONTEXT)
+		next_mmu_context = FIRST_CONTEXT;
+	mm = context_mm[next_mmu_context];
+	flush_tlb_mm(mm);
+	destroy_context(mm);
+}
diff --git a/arch/microblaze/mm/pgtable.c b/arch/microblaze/mm/pgtable.c
new file mode 100644
index 000000000..7f525962c
--- /dev/null
+++ b/arch/microblaze/mm/pgtable.c
@@ -0,0 +1,260 @@
+/*
+ *  This file contains the routines setting up the linux page tables.
+ *
+ * Copyright (C) 2008 Michal Simek
+ * Copyright (C) 2008 PetaLogix
+ *
+ *    Copyright (C) 2007 Xilinx, Inc.  All rights reserved.
+ *
+ *  Derived from arch/ppc/mm/pgtable.c:
+ *    -- paulus
+ *
+ *  Derived from arch/ppc/mm/init.c:
+ *    Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org)
+ *
+ *  Modifications by Paul Mackerras (PowerMac) (paulus@cs.anu.edu.au)
+ *  and Cort Dougan (PReP) (cort@cs.nmt.edu)
+ *    Copyright (C) 1996 Paul Mackerras
+ *  Amiga/APUS changes by Jesper Skov (jskov@cygnus.co.uk).
+ *
+ *  Derived from "arch/i386/mm/init.c"
+ *    Copyright (C) 1991, 1992, 1993, 1994  Linus Torvalds
+ *
+ *  This file is subject to the terms and conditions of the GNU General
+ *  Public License.  See the file COPYING in the main directory of this
+ *  archive for more details.
+ *
+ */
+
+#include <linux/export.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/vmalloc.h>
+#include <linux/init.h>
+#include <linux/mm_types.h>
+
+#include <asm/pgtable.h>
+#include <asm/pgalloc.h>
+#include <linux/io.h>
+#include <asm/mmu.h>
+#include <asm/sections.h>
+#include <asm/fixmap.h>
+
+unsigned long ioremap_base;
+unsigned long ioremap_bot;
+EXPORT_SYMBOL(ioremap_bot);
+
+#ifndef CONFIG_SMP
+struct pgtable_cache_struct quicklists;
+#endif
+
+static void __iomem *__ioremap(phys_addr_t addr, unsigned long size,
+		unsigned long flags)
+{
+	unsigned long v, i;
+	phys_addr_t p;
+	int err;
+
+	/*
+	 * Choose an address to map it to.
+	 * Once the vmalloc system is running, we use it.
+	 * Before then, we use space going down from ioremap_base
+	 * (ioremap_bot records where we're up to).
+	 */
+	p = addr & PAGE_MASK;
+	size = PAGE_ALIGN(addr + size) - p;
+
+	/*
+	 * Don't allow anybody to remap normal RAM that we're using.
+	 * mem_init() sets high_memory so only do the check after that.
+	 *
+	 * However, allow remap of rootfs: TBD
+	 */
+
+	if (mem_init_done &&
+		p >= memory_start && p < virt_to_phys(high_memory) &&
+		!(p >= __virt_to_phys((phys_addr_t)__bss_stop) &&
+		p < __virt_to_phys((phys_addr_t)__bss_stop))) {
+		pr_warn("__ioremap(): phys addr "PTE_FMT" is RAM lr %pf\n",
+			(unsigned long)p, __builtin_return_address(0));
+		return NULL;
+	}
+
+	if (size == 0)
+		return NULL;
+
+	/*
+	 * Is it already mapped? If the whole area is mapped then we're
+	 * done, otherwise remap it since we want to keep the virt addrs for
+	 * each request contiguous.
+	 *
+	 * We make the assumption here that if the bottom and top
+	 * of the range we want are mapped then it's mapped to the
+	 * same virt address (and this is contiguous).
+	 *  -- Cort
+	 */
+
+	if (mem_init_done) {
+		struct vm_struct *area;
+		area = get_vm_area(size, VM_IOREMAP);
+		if (area == NULL)
+			return NULL;
+		v = (unsigned long) area->addr;
+	} else {
+		v = (ioremap_bot -= size);
+	}
+
+	if ((flags & _PAGE_PRESENT) == 0)
+		flags |= _PAGE_KERNEL;
+	if (flags & _PAGE_NO_CACHE)
+		flags |= _PAGE_GUARDED;
+
+	err = 0;
+	for (i = 0; i < size && err == 0; i += PAGE_SIZE)
+		err = map_page(v + i, p + i, flags);
+	if (err) {
+		if (mem_init_done)
+			vfree((void *)v);
+		return NULL;
+	}
+
+	return (void __iomem *) (v + ((unsigned long)addr & ~PAGE_MASK));
+}
+
+void __iomem *ioremap(phys_addr_t addr, unsigned long size)
+{
+	return __ioremap(addr, size, _PAGE_NO_CACHE);
+}
+EXPORT_SYMBOL(ioremap);
+
+void iounmap(volatile void __iomem *addr)
+{
+	if ((__force void *)addr > high_memory &&
+					(unsigned long) addr < ioremap_bot)
+		vfree((void *) (PAGE_MASK & (unsigned long) addr));
+}
+EXPORT_SYMBOL(iounmap);
+
+
+int map_page(unsigned long va, phys_addr_t pa, int flags)
+{
+	pmd_t *pd;
+	pte_t *pg;
+	int err = -ENOMEM;
+	/* Use upper 10 bits of VA to index the first level map */
+	pd = pmd_offset(pgd_offset_k(va), va);
+	/* Use middle 10 bits of VA to index the second-level map */
+	pg = pte_alloc_kernel(pd, va); /* from powerpc - pgtable.c */
+	/* pg = pte_alloc_kernel(&init_mm, pd, va); */
+
+	if (pg != NULL) {
+		err = 0;
+		set_pte_at(&init_mm, va, pg, pfn_pte(pa >> PAGE_SHIFT,
+				__pgprot(flags)));
+		if (unlikely(mem_init_done))
+			_tlbie(va);
+	}
+	return err;
+}
+
+/*
+ * Map in all of physical memory starting at CONFIG_KERNEL_START.
+ */
+void __init mapin_ram(void)
+{
+	unsigned long v, p, s, f;
+
+	v = CONFIG_KERNEL_START;
+	p = memory_start;
+	for (s = 0; s < lowmem_size; s += PAGE_SIZE) {
+		f = _PAGE_PRESENT | _PAGE_ACCESSED |
+				_PAGE_SHARED | _PAGE_HWEXEC;
+		if ((char *) v < _stext || (char *) v >= _etext)
+			f |= _PAGE_WRENABLE;
+		else
+			/* On the MicroBlaze, no user access
+			   forces R/W kernel access */
+			f |= _PAGE_USER;
+		map_page(v, p, f);
+		v += PAGE_SIZE;
+		p += PAGE_SIZE;
+	}
+}
+
+/* is x a power of 2? */
+#define is_power_of_2(x)	((x) != 0 && (((x) & ((x) - 1)) == 0))
+
+/* Scan the real Linux page tables and return a PTE pointer for
+ * a virtual address in a context.
+ * Returns true (1) if PTE was found, zero otherwise.  The pointer to
+ * the PTE pointer is unmodified if PTE is not found.
+ */
+static int get_pteptr(struct mm_struct *mm, unsigned long addr, pte_t **ptep)
+{
+	pgd_t	*pgd;
+	pmd_t	*pmd;
+	pte_t	*pte;
+	int     retval = 0;
+
+	pgd = pgd_offset(mm, addr & PAGE_MASK);
+	if (pgd) {
+		pmd = pmd_offset(pgd, addr & PAGE_MASK);
+		if (pmd_present(*pmd)) {
+			pte = pte_offset_kernel(pmd, addr & PAGE_MASK);
+			if (pte) {
+				retval = 1;
+				*ptep = pte;
+			}
+		}
+	}
+	return retval;
+}
+
+/* Find physical address for this virtual address.  Normally used by
+ * I/O functions, but anyone can call it.
+ */
+unsigned long iopa(unsigned long addr)
+{
+	unsigned long pa;
+
+	pte_t *pte;
+	struct mm_struct *mm;
+
+	/* Allow mapping of user addresses (within the thread)
+	 * for DMA if necessary.
+	 */
+	if (addr < TASK_SIZE)
+		mm = current->mm;
+	else
+		mm = &init_mm;
+
+	pa = 0;
+	if (get_pteptr(mm, addr, &pte))
+		pa = (pte_val(*pte) & PAGE_MASK) | (addr & ~PAGE_MASK);
+
+	return pa;
+}
+
+__ref pte_t *pte_alloc_one_kernel(struct mm_struct *mm,
+		unsigned long address)
+{
+	pte_t *pte;
+	if (mem_init_done) {
+		pte = (pte_t *)__get_free_page(GFP_KERNEL | __GFP_ZERO);
+	} else {
+		pte = (pte_t *)early_get_page();
+		if (pte)
+			clear_page(pte);
+	}
+	return pte;
+}
+
+void __set_fixmap(enum fixed_addresses idx, phys_addr_t phys, pgprot_t flags)
+{
+	unsigned long address = __fix_to_virt(idx);
+
+	if (idx >= __end_of_fixed_addresses)
+		BUG();
+
+	map_page(address, phys, pgprot_val(flags));
+}