5 files changed, 821 insertions, 10 deletions

diff --git a/arch/microblaze/mm/Makefile b/arch/microblaze/mm/Makefile
index bf9e4479a1f..6c8a924d9e2 100644
--- a/arch/microblaze/mm/Makefile
+++ b/arch/microblaze/mm/Makefile
@@ -3,3 +3,5 @@
 #
 
 obj-y := init.o
+
+obj-$(CONFIG_MMU) += pgtable.o mmu_context.o fault.o
diff --git a/arch/microblaze/mm/fault.c b/arch/microblaze/mm/fault.c
new file mode 100644
index 00000000000..5e67cd1fab4
--- /dev/null
+++ b/arch/microblaze/mm/fault.c
@@ -0,0 +1,304 @@
+/*
+ *  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/module.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 <asm/mmu_context.h>
+#include <asm/system.h>
+#include <linux/uaccess.h>
+#include <asm/exceptions.h>
+
+#if defined(CONFIG_KGDB)
+int debugger_kernel_faults = 1;
+#endif
+
+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 *)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.
+ */
+static 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 */
+#if defined(CONFIG_KGDB)
+	if (debugger_kernel_faults)
+		debugger(regs);
+#endif
+	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;
+	siginfo_t info;
+	int code = SEGV_MAPERR;
+	int is_write = error_code & ESR_S;
+	int fault;
+
+	regs->ear = address;
+	regs->esr = error_code;
+
+	/* On a kernel SLB miss we can only check for a valid exception entry */
+	if (kernel_mode(regs) && (address >= TASK_SIZE)) {
+		printk(KERN_WARNING "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 defined(CONFIG_KGDB)
+	if (debugger_fault_handler && regs->trap == 0x300) {
+		debugger_fault_handler(regs);
+		return;
+	}
+#endif /* CONFIG_KGDB */
+
+	if (in_atomic() || mm == NULL) {
+		/* FIXME */
+		if (kernel_mode(regs)) {
+			printk(KERN_EMERG
+				"Page fault in kernel mode - Oooou!!! pid %d\n",
+				current->pid);
+			_exception(SIGSEGV, regs, code, address);
+			return;
+		}
+		/* in_atomic() in user mode is really bad,
+		   as is current->mm == NULL. */
+		printk(KERN_EMERG "Page fault in user mode with "
+		       "in_atomic(), mm = %p\n", mm);
+		printk(KERN_EMERG "r15 = %lx  MSR = %lx\n",
+		       regs->r15, regs->msr);
+		die("Weird page fault", regs, SIGSEGV);
+	}
+
+	/* 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 (!down_read_trylock(&mm->mmap_sem)) {
+		if (kernel_mode(regs) && !search_exception_tables(regs->pc))
+			goto bad_area_nosemaphore;
+
+		down_read(&mm->mmap_sem);
+	}
+
+	vma = find_vma(mm, address);
+	if (!vma)
+		goto bad_area;
+
+	if (vma->vm_start <= address)
+		goto good_area;
+
+	if (!(vma->vm_flags & VM_GROWSDOWN))
+		goto bad_area;
+
+	if (!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 (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 (is_write) {
+		if (!(vma->vm_flags & VM_WRITE))
+			goto bad_area;
+	/* a read */
+	} else {
+		/* protection fault */
+		if (error_code & 0x08000000)
+			goto bad_area;
+		if (!(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.
+	 */
+survive:
+	fault = handle_mm_fault(mm, vma, address, is_write);
+	if (unlikely(fault & VM_FAULT_ERROR)) {
+		if (fault & VM_FAULT_OOM)
+			goto out_of_memory;
+		else if (fault & VM_FAULT_SIGBUS)
+			goto do_sigbus;
+		BUG();
+	}
+	if (fault & VM_FAULT_MAJOR)
+		current->maj_flt++;
+	else
+		current->min_flt++;
+	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);
+/*		info.si_signo = SIGSEGV;
+		info.si_errno = 0;
+		info.si_code = code;
+		info.si_addr = (void *) address;
+		force_sig_info(SIGSEGV, &info, current);*/
+		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:
+	if (current->pid == 1) {
+		yield();
+		down_read(&mm->mmap_sem);
+		goto survive;
+	}
+	up_read(&mm->mmap_sem);
+	printk(KERN_WARNING "VM: killing process %s\n", current->comm);
+	if (user_mode(regs))
+		do_exit(SIGKILL);
+	bad_page_fault(regs, address, SIGKILL);
+	return;
+
+do_sigbus:
+	up_read(&mm->mmap_sem);
+	if (user_mode(regs)) {
+		info.si_signo = SIGBUS;
+		info.si_errno = 0;
+		info.si_code = BUS_ADRERR;
+		info.si_addr = (void __user *)address;
+		force_sig_info(SIGBUS, &info, current);
+		return;
+	}
+	bad_page_fault(regs, address, SIGBUS);
+}
diff --git a/arch/microblaze/mm/init.c b/arch/microblaze/mm/init.c
index b0c8213cd6c..b5a701cd71e 100644
--- a/arch/microblaze/mm/init.c
+++ b/arch/microblaze/mm/init.c
@@ -23,8 +23,16 @@
 #include <asm/sections.h>
 #include <asm/tlb.h>
 
+#ifndef CONFIG_MMU
 unsigned int __page_offset;
-/* EXPORT_SYMBOL(__page_offset); */
+EXPORT_SYMBOL(__page_offset);
+
+#else
+DEFINE_PER_CPU(struct mmu_gather, mmu_gathers);
+
+int mem_init_done;
+static int init_bootmem_done;
+#endif /* CONFIG_MMU */
 
 char *klimit = _end;
 
@@ -32,28 +40,26 @@ char *klimit = _end;
  * Initialize the bootmem system and give it all the memory we
  * have available.
  */
-unsigned int memory_start;
-unsigned int memory_end; /* due to mm/nommu.c */
-unsigned int memory_size;
+unsigned long memory_start;
+unsigned long memory_end; /* due to mm/nommu.c */
+unsigned long memory_size;
 
 /*
  * paging_init() sets up the page tables - in fact we've already done this.
  */
 static void __init paging_init(void)
 {
-	int i;
 	unsigned long zones_size[MAX_NR_ZONES];
 
+	/* Clean every zones */
+	memset(zones_size, 0, sizeof(zones_size));
+
 	/*
 	 * old: we can DMA to/from any address.put all page into ZONE_DMA
 	 * We use only ZONE_NORMAL
 	 */
 	zones_size[ZONE_NORMAL] = max_mapnr;
 
-	/* every other zones are empty */
-	for (i = 1; i < MAX_NR_ZONES; i++)
-		zones_size[i] = 0;
-
 	free_area_init(zones_size);
 }
 
@@ -61,6 +67,7 @@ void __init setup_memory(void)
 {
 	int i;
 	unsigned long map_size;
+#ifndef CONFIG_MMU
 	u32 kernel_align_start, kernel_align_size;
 
 	/* Find main memory where is the kernel */
@@ -93,6 +100,7 @@ void __init setup_memory(void)
 		__func__, kernel_align_start, kernel_align_start
 			+ kernel_align_size, kernel_align_size);
 
+#endif
 	/*
 	 * Kernel:
 	 * start: base phys address of kernel - page align
@@ -121,9 +129,13 @@ void __init setup_memory(void)
 	 * for 4GB of memory, using 4kB pages), plus 1 page
 	 * (in case the address isn't page-aligned).
 	 */
+#ifndef CONFIG_MMU
 	map_size = init_bootmem_node(NODE_DATA(0), PFN_UP(TOPHYS((u32)_end)),
 					min_low_pfn, max_low_pfn);
-
+#else
+	map_size = init_bootmem_node(&contig_page_data,
+		PFN_UP(TOPHYS((u32)_end)), min_low_pfn, max_low_pfn);
+#endif
 	lmb_reserve(PFN_UP(TOPHYS((u32)_end)) << PAGE_SHIFT, map_size);
 
 	/* free bootmem is whole main memory */
@@ -137,6 +149,9 @@ void __init setup_memory(void)
 		reserve_bootmem(lmb.reserved.region[i].base,
 			lmb_size_bytes(&lmb.reserved, i) - 1, BOOTMEM_DEFAULT);
 	}
+#ifdef CONFIG_MMU
+	init_bootmem_done = 1;
+#endif
 	paging_init();
 }
 
@@ -191,11 +206,145 @@ void __init mem_init(void)
 	printk(KERN_INFO "Memory: %luk/%luk available\n",
 	       (unsigned long) nr_free_pages() << (PAGE_SHIFT-10),
 	       num_physpages << (PAGE_SHIFT-10));
+#ifdef CONFIG_MMU
+	mem_init_done = 1;
+#endif
 }
 
+#ifndef CONFIG_MMU
 /* Check against bounds of physical memory */
 int ___range_ok(unsigned long addr, unsigned long size)
 {
 	return ((addr < memory_start) ||
 		((addr + size) > memory_end));
 }
+EXPORT_SYMBOL(___range_ok);
+
+#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;
+			memory_end = memory_start + memory_size;
+			lmb.memory.region[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 (!lmb.reserved.cnt) {
+		printk(KERN_EMERG "Error memory count\n");
+		machine_restart(NULL);
+	}
+
+	if ((u32) lmb.memory.region[0].size < 0x1000000) {
+		printk(KERN_EMERG "Memory must be greater than 16MB\n");
+		machine_restart(NULL);
+	}
+	/* Find main memory where the kernel is */
+	memory_start = (u32) lmb.memory.region[0].base;
+	memory_end = (u32) lmb.memory.region[0].base +
+				(u32) lmb.memory.region[0].size;
+	memory_size = memory_end - memory_start;
+
+	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));
+	lmb_reserve(kstart, ksize);
+
+#if defined(CONFIG_BLK_DEV_INITRD)
+	/* Remove the init RAM disk from the available memory. */
+/*	if (initrd_start) {
+		mem_pieces_remove(&phys_avail, __pa(initrd_start),
+				  initrd_end - initrd_start, 1);
+	}*/
+#endif /* CONFIG_BLK_DEV_INITRD */
+
+	/* Initialize the MMU hardware */
+	mmu_init_hw();
+
+	/* Map in all of RAM starting at CONFIG_KERNEL_START */
+	mapin_ram();
+
+#ifdef HIGHMEM_START_BOOL
+	ioremap_base = HIGHMEM_START;
+#else
+	ioremap_base = 0xfe000000UL;	/* for now, could be 0xfffff000 */
+#endif /* CONFIG_HIGHMEM */
+	ioremap_bot = ioremap_base;
+
+	/* Initialize the context management stuff */
+	mmu_context_init();
+}
+
+/* This is only called until mem_init is done. */
+void __init *early_get_page(void)
+{
+	void *p;
+	if (init_bootmem_done) {
+		p = alloc_bootmem_pages(PAGE_SIZE);
+	} else {
+		/*
+		 * Mem start + 32MB -> here is limit
+		 * because of mem mapping from head.S
+		 */
+		p = __va(lmb_alloc_base(PAGE_SIZE, PAGE_SIZE,
+					memory_start + 0x2000000));
+	}
+	return p;
+}
+#endif /* CONFIG_MMU */
diff --git a/arch/microblaze/mm/mmu_context.c b/arch/microblaze/mm/mmu_context.c
new file mode 100644
index 00000000000..26ff82f4fa8
--- /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 00000000000..46c4ca5d15c
--- /dev/null
+++ b/arch/microblaze/mm/pgtable.c
@@ -0,0 +1,286 @@
+/*
+ *  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/kernel.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/vmalloc.h>
+#include <linux/init.h>
+
+#include <asm/pgtable.h>
+#include <asm/pgalloc.h>
+#include <linux/io.h>
+#include <asm/mmu.h>
+#include <asm/sections.h>
+
+#define flush_HPTE(X, va, pg)	_tlbie(va)
+
+unsigned long ioremap_base;
+unsigned long ioremap_bot;
+
+/* The maximum lowmem defaults to 768Mb, but this can be configured to
+ * another value.
+ */
+#define MAX_LOW_MEM	CONFIG_LOWMEM_SIZE
+
+#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((unsigned long)&__bss_stop) &&
+		p < virt_to_phys((unsigned long)__bss_stop))) {
+		printk(KERN_WARNING "__ioremap(): phys addr "PTE_FMT
+			" is RAM lr %p\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 = VMALLOC_VMADDR(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(void *addr)
+{
+	if (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;
+	/* spin_lock(&init_mm.page_table_lock); */
+	/* 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 (mem_init_done)
+			flush_HPTE(0, va, pmd_val(*pd));
+			/* flush_HPTE(0, va, pg); */
+
+	}
+	/* spin_unlock(&init_mm.page_table_lock); */
+	return err;
+}
+
+void __init adjust_total_lowmem(void)
+{
+/* TBD */
+#if 0
+	unsigned long max_low_mem = MAX_LOW_MEM;
+
+	if (total_lowmem > max_low_mem) {
+		total_lowmem = max_low_mem;
+#ifndef CONFIG_HIGHMEM
+		printk(KERN_INFO "Warning, memory limited to %ld Mb, use "
+				"CONFIG_HIGHMEM to reach %ld Mb\n",
+				max_low_mem >> 20, total_memory >> 20);
+		total_memory = total_lowmem;
+#endif /* CONFIG_HIGHMEM */
+	}
+#endif
+}
+
+static void show_tmem(unsigned long tmem)
+{
+	volatile unsigned long a;
+	a = a + tmem;
+}
+
+/*
+ * 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;
+	show_tmem(memory_size);
+	for (s = 0; s < memory_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))
+
+/*
+ * Set up a mapping for a block of I/O.
+ * virt, phys, size must all be page-aligned.
+ * This should only be called before ioremap is called.
+ */
+void __init io_block_mapping(unsigned long virt, phys_addr_t phys,
+			     unsigned int size, int flags)
+{
+	int i;
+
+	if (virt > CONFIG_KERNEL_START && virt < ioremap_bot)
+		ioremap_bot = ioremap_base = virt;
+
+	/* Put it in the page tables. */
+	for (i = 0; i < size; i += PAGE_SIZE)
+		map_page(virt + i, phys + i, flags);
+}
+
+/* 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;
+}