summaryrefslogtreecommitdiffstats
path: root/target/linux/coldfire/files-2.6.31/arch/m68k/mm/cf-mmu.c
diff options
context:
space:
mode:
Diffstat (limited to 'target/linux/coldfire/files-2.6.31/arch/m68k/mm/cf-mmu.c')
-rw-r--r--target/linux/coldfire/files-2.6.31/arch/m68k/mm/cf-mmu.c317
1 files changed, 317 insertions, 0 deletions
diff --git a/target/linux/coldfire/files-2.6.31/arch/m68k/mm/cf-mmu.c b/target/linux/coldfire/files-2.6.31/arch/m68k/mm/cf-mmu.c
new file mode 100644
index 000000000..8e9685086
--- /dev/null
+++ b/target/linux/coldfire/files-2.6.31/arch/m68k/mm/cf-mmu.c
@@ -0,0 +1,317 @@
+/*
+ * linux/arch/m68k/mm/cf-mmu.c
+ *
+ * Based upon linux/arch/m68k/mm/sun3mmu.c
+ * Based upon linux/arch/ppc/mm/mmu_context.c
+ *
+ * Implementations of mm routines specific to the Coldfire MMU.
+ *
+ * Copyright (c) 2008 Freescale Semiconductor, Inc.
+ * Copyright Freescale Semiconductor, Inc. 2008-2009
+ * Jason Jin Jason.Jin@freescale.com
+ * Shrek Wu B16972@freescale.com
+ */
+
+#include <linux/signal.h>
+#include <linux/sched.h>
+#include <linux/mm.h>
+#include <linux/swap.h>
+#include <linux/kernel.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/init.h>
+#ifdef CONFIG_BLK_DEV_RAM
+#include <linux/blkdev.h>
+#endif
+#include <linux/bootmem.h>
+
+#include <asm/setup.h>
+#include <asm/uaccess.h>
+#include <asm/page.h>
+#include <asm/pgtable.h>
+#include <asm/system.h>
+#include <asm/machdep.h>
+#include <asm/io.h>
+#include <asm/mmu_context.h>
+#include <asm/cf_pgalloc.h>
+
+#include <asm/coldfire.h>
+#include <asm/tlbflush.h>
+
+#define KMAPAREA(x) ((x >= VMALLOC_START) && (x < KMAP_END))
+
+#undef DEBUG
+
+#ifdef CONFIG_VDSO
+unsigned long next_mmu_context;
+#else
+mm_context_t next_mmu_context;
+#endif
+
+unsigned long context_map[LAST_CONTEXT / BITS_PER_LONG + 1];
+
+atomic_t nr_free_contexts;
+struct mm_struct *context_mm[LAST_CONTEXT+1];
+void steal_context(void);
+#ifdef CONFIG_M5445X
+void m68k_setup_node(int);
+#endif
+const char bad_pmd_string[] = "Bad pmd in pte_alloc: %08lx\n";
+
+extern unsigned long empty_bad_page_table;
+extern unsigned long empty_bad_page;
+extern unsigned long num_pages;
+#ifdef CONFIG_M5445X
+extern unsigned long availmem;
+#endif
+extern char __init_begin, __init_end;
+
+/*
+ * Free memory used for system initialization.
+ */
+void free_initmem(void)
+{
+#if 0
+ unsigned long addr;
+ unsigned long start = (unsigned long)&__init_begin;
+ unsigned long end = (unsigned long)&__init_end;
+
+ printk(KERN_INFO "free_initmem: __init_begin = 0x%lx __init_end = 0x%lx\n", start, end);
+
+ addr = (unsigned long)&__init_begin;
+ for (; addr < (unsigned long)&__init_end; addr += PAGE_SIZE) {
+ /* not currently used */
+ virt_to_page(addr)->flags &= ~(1 << PG_reserved);
+ init_page_count(virt_to_page(addr));
+ free_page(addr);
+ totalram_pages++;
+ }
+#endif
+}
+
+/*
+ * Initialize the paging system.
+ */
+void __init paging_init(void)
+{
+ pgd_t * pg_dir;
+ pte_t * pg_table;
+ int i;
+ unsigned long address;
+ unsigned long next_pgtable;
+ unsigned long zones_size[MAX_NR_ZONES];
+ unsigned long size;
+ enum zone_type zone;
+
+ /* allocate zero page */
+ empty_zero_page = (void *)alloc_bootmem_pages(PAGE_SIZE);
+ memset((void *)empty_zero_page, 0, PAGE_SIZE);
+
+ /* zero kernel page directory */
+ pg_dir = swapper_pg_dir;
+ memset(swapper_pg_dir, 0, sizeof(swapper_pg_dir));
+ /*
+ * setup page tables for PHYSRAM
+ */
+
+ /* starting loc in page directory */
+ pg_dir += PAGE_OFFSET >> PGDIR_SHIFT;
+
+ /* allocate page tables */
+ size = num_pages * sizeof(pte_t);
+ size = (size + PAGE_SIZE) & ~(PAGE_SIZE-1);
+ next_pgtable = (unsigned long)alloc_bootmem_pages(size);
+ address = PAGE_OFFSET;
+ while (address < (unsigned long)high_memory) {
+ /* setup page table in page directory */
+ pg_table = (pte_t *)next_pgtable;
+ next_pgtable += PTRS_PER_PTE * sizeof(pte_t);
+ pgd_val(*pg_dir) = (unsigned long)pg_table;
+ pg_dir++;
+
+ /* create PTEs in page table */
+ for (i=0; i<PTRS_PER_PTE; ++i, ++pg_table) {
+ pte_t pte = pfn_pte(virt_to_pfn(address), PAGE_INIT);
+ if (address >= (unsigned long)high_memory)
+ pte_val (pte) = 0;
+
+ set_pte(pg_table, pte);
+ address += PAGE_SIZE;
+ }
+ }
+
+ /*
+ * setup page tables for DMA area
+ */
+
+ /* starting loc in page directory */
+ pg_dir = swapper_pg_dir;
+ pg_dir += CONFIG_DMA_BASE >> PGDIR_SHIFT;
+
+ /* allocate page tables */
+ size = (CONFIG_DMA_SIZE >> PAGE_SHIFT) * sizeof(pte_t);
+ size = (size + PAGE_SIZE) & ~(PAGE_SIZE-1);
+ next_pgtable = (unsigned long)alloc_bootmem_pages(size);
+ address = CONFIG_DMA_BASE;
+ while (address < (CONFIG_DMA_BASE + CONFIG_DMA_SIZE)) {
+ /* setup page table in page directory */
+ pg_table = (pte_t *)next_pgtable;
+ next_pgtable += PTRS_PER_PTE * sizeof(pte_t);
+ pgd_val(*pg_dir) = (unsigned long)pg_table;
+ pg_dir++;
+
+ /* create PTEs in page table */
+ for (i=0; i<PTRS_PER_PTE; ++i, ++pg_table) {
+ pte_t pte = pfn_pte(virt_to_pfn(address), PAGE_INIT);
+ if (address >= (CONFIG_DMA_BASE + CONFIG_DMA_SIZE))
+ pte_val (pte) = 0;
+
+ set_pte(pg_table, pte);
+ address += PAGE_SIZE;
+ }
+ }
+
+ /*
+ * setup zones
+ */
+
+ current->mm = NULL;
+
+ /* clear zones */
+ for (zone = 0; zone < MAX_NR_ZONES; zone++)
+ zones_size[zone] = 0x0;
+
+ zones_size[ZONE_DMA] = CONFIG_DMA_SIZE >> PAGE_SHIFT;
+ zones_size[ZONE_NORMAL] = (((unsigned long)high_memory -
+ PAGE_OFFSET) >> PAGE_SHIFT) -
+ zones_size[ZONE_DMA];
+
+ free_area_init(zones_size);
+}
+/*
+ * Handle a missed TLB
+ */
+int cf_tlb_miss(struct pt_regs *regs, int write, int dtlb, int extension_word)
+{
+ struct mm_struct *mm;
+ pgd_t *pgd;
+ pmd_t *pmd;
+ pte_t *pte;
+ unsigned long mmuar;
+ int asid;
+ int flags;
+
+ local_save_flags(flags);
+ local_irq_disable();
+
+ mmuar = ( dtlb ) ? regs->mmuar
+ : regs->pc + (extension_word * sizeof(long));
+
+ mm = (!user_mode(regs) && KMAPAREA(mmuar)) ? &init_mm : current->mm;
+
+ if (!mm) {
+ local_irq_restore(flags);
+ return (-1);
+ }
+
+ pgd = pgd_offset(mm, mmuar);
+ if (pgd_none(*pgd)) {
+ local_irq_restore(flags);
+ return (-1);
+ }
+
+ pmd = pmd_offset(pgd, mmuar);
+ if (pmd_none(*pmd)) {
+ local_irq_restore(flags);
+ return (-1);
+ }
+
+ pte = (KMAPAREA(mmuar)) ? pte_offset_kernel(pmd, mmuar)
+ : pte_offset_map(pmd, mmuar);
+ if (pte_none(*pte) || !pte_present(*pte)) {
+ local_irq_restore(flags);
+ return (-1);
+ }
+
+ if (write) {
+ if (!pte_write(*pte)) {
+ local_irq_restore(flags);
+ return (-1);
+ }
+ set_pte(pte, pte_mkdirty(*pte));
+ }
+
+ set_pte(pte, pte_mkyoung(*pte));
+ asid = cpu_context(mm) & 0xff;
+ if (!pte_dirty(*pte) && !KMAPAREA(mmuar))
+ set_pte(pte, pte_wrprotect(*pte));
+
+ *MMUTR = (mmuar & PAGE_MASK) | (asid << CF_ASID_MMU_SHIFT)
+ | (((int)(pte->pte) & (int)CF_PAGE_MMUTR_MASK ) >> CF_PAGE_MMUTR_SHIFT)
+ | MMUTR_V;
+
+ *MMUDR = (pte_val(*pte) & PAGE_MASK)
+ | ((pte->pte) & CF_PAGE_MMUDR_MASK)
+ | MMUDR_SZ8K | MMUDR_X;
+
+ if ( dtlb )
+ *MMUOR = MMUOR_ACC | MMUOR_UAA;
+ else
+ *MMUOR = MMUOR_ITLB | MMUOR_ACC | MMUOR_UAA;
+
+ asm("nop");
+
+#ifdef DEBUG
+ printk("cf_tlb_miss: va=%lx, pa=%lx\n", (mmuar & PAGE_MASK),
+ (pte_val(*pte) & PAGE_MASK));
+#endif
+ local_irq_restore(flags);
+ return (0);
+}
+
+
+/*
+ * Context Management
+ *
+ * Based on arch/ppc/mmu_context.c
+ */
+
+/*
+ * Initialize the context management system.
+ */
+void __init mmu_context_init(void)
+{
+ /*
+ * Some processors have too few contexts to reserve one for
+ * init_mm, and require using context 0 for a normal task.
+ * Other processors reserve the use of context zero 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 is only used on 8xx and 4xx and we presently assume that
+ * they don't do SMP. If they do then thicfpgalloc.hs will have to check
+ * whether the MM we steal is in use.
+ * We also assume that this is only used on systems that don't
+ * use an MMU hash table - this is true for 8xx and 4xx.
+ * 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 was motivated to do it.
+ * -- paulus
+ */
+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);
+}