diff options
Diffstat (limited to 'arch/arm/mm')
-rw-r--r-- | arch/arm/mm/Kconfig | 30 | ||||
-rw-r--r-- | arch/arm/mm/Makefile | 2 | ||||
-rw-r--r-- | arch/arm/mm/abort-ev6.S | 5 | ||||
-rw-r--r-- | arch/arm/mm/alignment.c | 55 | ||||
-rw-r--r-- | arch/arm/mm/blockops.c | 185 | ||||
-rw-r--r-- | arch/arm/mm/cache-v6.S | 9 | ||||
-rw-r--r-- | arch/arm/mm/consistent.c | 27 | ||||
-rw-r--r-- | arch/arm/mm/copypage-v6.c | 16 | ||||
-rw-r--r-- | arch/arm/mm/fault-armv.c | 7 | ||||
-rw-r--r-- | arch/arm/mm/flush.c | 43 | ||||
-rw-r--r-- | arch/arm/mm/init.c | 497 | ||||
-rw-r--r-- | arch/arm/mm/ioremap.c | 8 | ||||
-rw-r--r-- | arch/arm/mm/mm-armv.c | 235 | ||||
-rw-r--r-- | arch/arm/mm/proc-v6.S | 35 |
14 files changed, 495 insertions, 659 deletions
diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig index db5e47d..e84fdde 100644 --- a/arch/arm/mm/Kconfig +++ b/arch/arm/mm/Kconfig @@ -102,8 +102,8 @@ config CPU_ARM922T # ARM925T config CPU_ARM925T bool "Support ARM925T processor" if ARCH_OMAP1 - depends on ARCH_OMAP1510 - default y if ARCH_OMAP1510 + depends on ARCH_OMAP15XX + default y if ARCH_OMAP15XX select CPU_32v4 select CPU_ABRT_EV4T select CPU_CACHE_V4WT @@ -120,8 +120,8 @@ config CPU_ARM925T # ARM926T config CPU_ARM926T - bool "Support ARM926T processor" if ARCH_INTEGRATOR - depends on ARCH_INTEGRATOR || ARCH_VERSATILE_PB || MACH_VERSATILE_AB || ARCH_OMAP730 || ARCH_OMAP16XX + bool "Support ARM926T processor" + depends on ARCH_INTEGRATOR || ARCH_VERSATILE_PB || MACH_VERSATILE_AB || ARCH_OMAP730 || ARCH_OMAP16XX || MACH_REALVIEW_EB default y if ARCH_VERSATILE_PB || MACH_VERSATILE_AB || ARCH_OMAP730 || ARCH_OMAP16XX select CPU_32v5 select CPU_ABRT_EV5TJ @@ -242,7 +242,7 @@ config CPU_XSCALE # ARMv6 config CPU_V6 bool "Support ARM V6 processor" - depends on ARCH_INTEGRATOR + depends on ARCH_INTEGRATOR || MACH_REALVIEW_EB || ARCH_OMAP2 select CPU_32v6 select CPU_ABRT_EV6 select CPU_CACHE_V6 @@ -250,6 +250,18 @@ config CPU_V6 select CPU_COPY_V6 select CPU_TLB_V6 +# ARMv6k +config CPU_32v6K + bool "Support ARM V6K processor extensions" if !SMP + depends on CPU_V6 + default y if SMP + help + Say Y here if your ARMv6 processor supports the 'K' extension. + This enables the kernel to use some instructions not present + on previous processors, and as such a kernel build with this + enabled will not boot on processors with do not support these + instructions. + # Figure out what processor architecture version we should be using. # This defines the compiler instruction set which depends on the machine type. config CPU_32v3 @@ -370,21 +382,21 @@ config CPU_BIG_ENDIAN config CPU_ICACHE_DISABLE bool "Disable I-Cache" - depends on CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020 + depends on CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020 || CPU_V6 help Say Y here to disable the processor instruction cache. Unless you have a reason not to or are unsure, say N. config CPU_DCACHE_DISABLE bool "Disable D-Cache" - depends on CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020 + depends on CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020 || CPU_V6 help Say Y here to disable the processor data cache. Unless you have a reason not to or are unsure, say N. config CPU_DCACHE_WRITETHROUGH bool "Force write through D-cache" - depends on (CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020) && !CPU_DCACHE_DISABLE + depends on (CPU_ARM920T || CPU_ARM922T || CPU_ARM925T || CPU_ARM926T || CPU_ARM1020 || CPU_V6) && !CPU_DCACHE_DISABLE default y if CPU_ARM925T help Say Y here to use the data cache in writethrough mode. Unless you @@ -399,7 +411,7 @@ config CPU_CACHE_ROUND_ROBIN config CPU_BPREDICT_DISABLE bool "Disable branch prediction" - depends on CPU_ARM1020 + depends on CPU_ARM1020 || CPU_V6 help Say Y here to disable branch prediction. If unsure, say N. diff --git a/arch/arm/mm/Makefile b/arch/arm/mm/Makefile index 59f47d4..ffe73ba 100644 --- a/arch/arm/mm/Makefile +++ b/arch/arm/mm/Makefile @@ -51,4 +51,4 @@ obj-$(CONFIG_CPU_ARM1026) += proc-arm1026.o obj-$(CONFIG_CPU_SA110) += proc-sa110.o obj-$(CONFIG_CPU_SA1100) += proc-sa1100.o obj-$(CONFIG_CPU_XSCALE) += proc-xscale.o -obj-$(CONFIG_CPU_V6) += proc-v6.o blockops.o +obj-$(CONFIG_CPU_V6) += proc-v6.o diff --git a/arch/arm/mm/abort-ev6.S b/arch/arm/mm/abort-ev6.S index 8f76f3d..dbd3460 100644 --- a/arch/arm/mm/abort-ev6.S +++ b/arch/arm/mm/abort-ev6.S @@ -20,6 +20,11 @@ */ .align 5 ENTRY(v6_early_abort) +#ifdef CONFIG_CPU_MPCORE + clrex +#else + strex r0, r1, [sp] @ Clear the exclusive monitor +#endif mrc p15, 0, r1, c5, c0, 0 @ get FSR mrc p15, 0, r0, c6, c0, 0 @ get FAR /* diff --git a/arch/arm/mm/alignment.c b/arch/arm/mm/alignment.c index 4b39d86..705c989 100644 --- a/arch/arm/mm/alignment.c +++ b/arch/arm/mm/alignment.c @@ -111,7 +111,7 @@ proc_alignment_read(char *page, char **start, off_t off, int count, int *eof, } static int proc_alignment_write(struct file *file, const char __user *buffer, - unsigned long count, void *data) + unsigned long count, void *data) { char mode; @@ -119,7 +119,7 @@ static int proc_alignment_write(struct file *file, const char __user *buffer, if (get_user(mode, buffer)) return -EFAULT; if (mode >= '0' && mode <= '5') - ai_usermode = mode - '0'; + ai_usermode = mode - '0'; } return count; } @@ -262,7 +262,7 @@ union offset_union { goto fault; \ } while (0) -#define put32_unaligned_check(val,addr) \ +#define put32_unaligned_check(val,addr) \ __put32_unaligned_check("strb", val, addr) #define put32t_unaligned_check(val,addr) \ @@ -306,19 +306,19 @@ do_alignment_ldrhstrh(unsigned long addr, unsigned long instr, struct pt_regs *r return TYPE_LDST; user: - if (LDST_L_BIT(instr)) { - unsigned long val; - get16t_unaligned_check(val, addr); + if (LDST_L_BIT(instr)) { + unsigned long val; + get16t_unaligned_check(val, addr); - /* signed half-word? */ - if (instr & 0x40) - val = (signed long)((signed short) val); + /* signed half-word? */ + if (instr & 0x40) + val = (signed long)((signed short) val); - regs->uregs[rd] = val; - } else - put16t_unaligned_check(regs->uregs[rd], addr); + regs->uregs[rd] = val; + } else + put16t_unaligned_check(regs->uregs[rd], addr); - return TYPE_LDST; + return TYPE_LDST; fault: return TYPE_FAULT; @@ -330,6 +330,9 @@ do_alignment_ldrdstrd(unsigned long addr, unsigned long instr, { unsigned int rd = RD_BITS(instr); + if (((rd & 1) == 1) || (rd == 14)) + goto bad; + ai_dword += 1; if (user_mode(regs)) @@ -339,11 +342,11 @@ do_alignment_ldrdstrd(unsigned long addr, unsigned long instr, unsigned long val; get32_unaligned_check(val, addr); regs->uregs[rd] = val; - get32_unaligned_check(val, addr+4); - regs->uregs[rd+1] = val; + get32_unaligned_check(val, addr + 4); + regs->uregs[rd + 1] = val; } else { put32_unaligned_check(regs->uregs[rd], addr); - put32_unaligned_check(regs->uregs[rd+1], addr+4); + put32_unaligned_check(regs->uregs[rd + 1], addr + 4); } return TYPE_LDST; @@ -353,15 +356,16 @@ do_alignment_ldrdstrd(unsigned long addr, unsigned long instr, unsigned long val; get32t_unaligned_check(val, addr); regs->uregs[rd] = val; - get32t_unaligned_check(val, addr+4); - regs->uregs[rd+1] = val; + get32t_unaligned_check(val, addr + 4); + regs->uregs[rd + 1] = val; } else { put32t_unaligned_check(regs->uregs[rd], addr); - put32t_unaligned_check(regs->uregs[rd+1], addr+4); + put32t_unaligned_check(regs->uregs[rd + 1], addr + 4); } return TYPE_LDST; - + bad: + return TYPE_ERROR; fault: return TYPE_FAULT; } @@ -439,7 +443,7 @@ do_alignment_ldmstm(unsigned long addr, unsigned long instr, struct pt_regs *reg if (LDST_P_EQ_U(instr)) /* U = P */ eaddr += 4; - /* + /* * For alignment faults on the ARM922T/ARM920T the MMU makes * the FSR (and hence addr) equal to the updated base address * of the multiple access rather than the restored value. @@ -566,7 +570,7 @@ thumb2arm(u16 tinstr) /* 6.5.1 Format 3: */ case 0x4800 >> 11: /* 7.1.28 LDR(3) */ /* NOTE: This case is not technically possible. We're - * loading 32-bit memory data via PC relative + * loading 32-bit memory data via PC relative * addressing mode. So we can and should eliminate * this case. But I'll leave it here for now. */ @@ -638,7 +642,7 @@ do_alignment(unsigned long addr, unsigned int fsr, struct pt_regs *regs) if (fault) { type = TYPE_FAULT; - goto bad_or_fault; + goto bad_or_fault; } if (user_mode(regs)) @@ -663,6 +667,8 @@ do_alignment(unsigned long addr, unsigned int fsr, struct pt_regs *regs) else if ((instr & 0x001000f0) == 0x000000d0 || /* LDRD */ (instr & 0x001000f0) == 0x000000f0) /* STRD */ handler = do_alignment_ldrdstrd; + else if ((instr & 0x01f00ff0) == 0x01000090) /* SWP */ + goto swp; else goto bad; break; @@ -733,6 +739,9 @@ do_alignment(unsigned long addr, unsigned int fsr, struct pt_regs *regs) do_bad_area(current, current->mm, addr, fsr, regs); return 0; + swp: + printk(KERN_ERR "Alignment trap: not handling swp instruction\n"); + bad: /* * Oops, we didn't handle the instruction. diff --git a/arch/arm/mm/blockops.c b/arch/arm/mm/blockops.c deleted file mode 100644 index 4f5ee2d..0000000 --- a/arch/arm/mm/blockops.c +++ /dev/null @@ -1,185 +0,0 @@ -#include <linux/kernel.h> -#include <linux/init.h> -#include <linux/errno.h> -#include <linux/mm.h> - -#include <asm/memory.h> -#include <asm/ptrace.h> -#include <asm/cacheflush.h> -#include <asm/traps.h> - -extern struct cpu_cache_fns blk_cache_fns; - -#define HARVARD_CACHE - -/* - * blk_flush_kern_dcache_page(kaddr) - * - * Ensure that the data held in the page kaddr is written back - * to the page in question. - * - * - kaddr - kernel address (guaranteed to be page aligned) - */ -static void __attribute__((naked)) -blk_flush_kern_dcache_page(void *kaddr) -{ - asm( - "add r1, r0, %0 \n\ - sub r1, r1, %1 \n\ -1: .word 0xec401f0e @ mcrr p15, 0, r0, r1, c14, 0 @ blocking \n\ - mov r0, #0 \n\ - mcr p15, 0, r0, c7, c5, 0 \n\ - mcr p15, 0, r0, c7, c10, 4 \n\ - mov pc, lr" - : - : "I" (PAGE_SIZE), "I" (L1_CACHE_BYTES)); -} - -/* - * blk_dma_inv_range(start,end) - * - * Invalidate the data cache within the specified region; we will - * be performing a DMA operation in this region and we want to - * purge old data in the cache. - * - * - start - virtual start address of region - * - end - virtual end address of region - */ -static void __attribute__((naked)) -blk_dma_inv_range_unified(unsigned long start, unsigned long end) -{ - asm( - "tst r0, %0 \n\ - mcrne p15, 0, r0, c7, c11, 1 @ clean unified line \n\ - tst r1, %0 \n\ - mcrne p15, 0, r1, c7, c15, 1 @ clean & invalidate unified line\n\ - .word 0xec401f06 @ mcrr p15, 0, r1, r0, c6, 0 @ blocking \n\ - mov r0, #0 \n\ - mcr p15, 0, r0, c7, c10, 4 @ drain write buffer \n\ - mov pc, lr" - : - : "I" (L1_CACHE_BYTES - 1)); -} - -static void __attribute__((naked)) -blk_dma_inv_range_harvard(unsigned long start, unsigned long end) -{ - asm( - "tst r0, %0 \n\ - mcrne p15, 0, r0, c7, c10, 1 @ clean D line \n\ - tst r1, %0 \n\ - mcrne p15, 0, r1, c7, c14, 1 @ clean & invalidate D line \n\ - .word 0xec401f06 @ mcrr p15, 0, r1, r0, c6, 0 @ blocking \n\ - mov r0, #0 \n\ - mcr p15, 0, r0, c7, c10, 4 @ drain write buffer \n\ - mov pc, lr" - : - : "I" (L1_CACHE_BYTES - 1)); -} - -/* - * blk_dma_clean_range(start,end) - * - start - virtual start address of region - * - end - virtual end address of region - */ -static void __attribute__((naked)) -blk_dma_clean_range(unsigned long start, unsigned long end) -{ - asm( - ".word 0xec401f0c @ mcrr p15, 0, r1, r0, c12, 0 @ blocking \n\ - mov r0, #0 \n\ - mcr p15, 0, r0, c7, c10, 4 @ drain write buffer \n\ - mov pc, lr"); -} - -/* - * blk_dma_flush_range(start,end) - * - start - virtual start address of region - * - end - virtual end address of region - */ -static void __attribute__((naked)) -blk_dma_flush_range(unsigned long start, unsigned long end) -{ - asm( - ".word 0xec401f0e @ mcrr p15, 0, r1, r0, c14, 0 @ blocking \n\ - mov pc, lr"); -} - -static int blockops_trap(struct pt_regs *regs, unsigned int instr) -{ - regs->ARM_r4 |= regs->ARM_r2; - regs->ARM_pc += 4; - return 0; -} - -static char *func[] = { - "Prefetch data range", - "Clean+Invalidate data range", - "Clean data range", - "Invalidate data range", - "Invalidate instr range" -}; - -static struct undef_hook blockops_hook __initdata = { - .instr_mask = 0x0fffffd0, - .instr_val = 0x0c401f00, - .cpsr_mask = PSR_T_BIT, - .cpsr_val = 0, - .fn = blockops_trap, -}; - -static int __init blockops_check(void) -{ - register unsigned int err asm("r4") = 0; - unsigned int err_pos = 1; - unsigned int cache_type; - int i; - - asm("mrc p15, 0, %0, c0, c0, 1" : "=r" (cache_type)); - - printk("Checking V6 block cache operations:\n"); - register_undef_hook(&blockops_hook); - - __asm__ ("mov r0, %0\n\t" - "mov r1, %1\n\t" - "mov r2, #1\n\t" - ".word 0xec401f2c @ mcrr p15, 0, r1, r0, c12, 2\n\t" - "mov r2, #2\n\t" - ".word 0xec401f0e @ mcrr p15, 0, r1, r0, c14, 0\n\t" - "mov r2, #4\n\t" - ".word 0xec401f0c @ mcrr p15, 0, r1, r0, c12, 0\n\t" - "mov r2, #8\n\t" - ".word 0xec401f06 @ mcrr p15, 0, r1, r0, c6, 0\n\t" - "mov r2, #16\n\t" - ".word 0xec401f05 @ mcrr p15, 0, r1, r0, c5, 0\n\t" - : - : "r" (PAGE_OFFSET), "r" (PAGE_OFFSET + 128) - : "r0", "r1", "r2"); - - unregister_undef_hook(&blockops_hook); - - for (i = 0; i < ARRAY_SIZE(func); i++, err_pos <<= 1) - printk("%30s: %ssupported\n", func[i], err & err_pos ? "not " : ""); - - if ((err & 8) == 0) { - printk(" --> Using %s block cache invalidate\n", - cache_type & (1 << 24) ? "harvard" : "unified"); - if (cache_type & (1 << 24)) - cpu_cache.dma_inv_range = blk_dma_inv_range_harvard; - else - cpu_cache.dma_inv_range = blk_dma_inv_range_unified; - } - if ((err & 4) == 0) { - printk(" --> Using block cache clean\n"); - cpu_cache.dma_clean_range = blk_dma_clean_range; - } - if ((err & 2) == 0) { - printk(" --> Using block cache clean+invalidate\n"); - cpu_cache.dma_flush_range = blk_dma_flush_range; - cpu_cache.flush_kern_dcache_page = blk_flush_kern_dcache_page; - } - - return 0; -} - -__initcall(blockops_check); diff --git a/arch/arm/mm/cache-v6.S b/arch/arm/mm/cache-v6.S index 85c10a7..72966d9 100644 --- a/arch/arm/mm/cache-v6.S +++ b/arch/arm/mm/cache-v6.S @@ -18,6 +18,7 @@ #define HARVARD_CACHE #define CACHE_LINE_SIZE 32 #define D_CACHE_LINE_SIZE 32 +#define BTB_FLUSH_SIZE 8 /* * v6_flush_cache_all() @@ -98,7 +99,13 @@ ENTRY(v6_coherent_user_range) mcr p15, 0, r0, c7, c5, 1 @ invalidate I line #endif mcr p15, 0, r0, c7, c5, 7 @ invalidate BTB entry - add r0, r0, #CACHE_LINE_SIZE + add r0, r0, #BTB_FLUSH_SIZE + mcr p15, 0, r0, c7, c5, 7 @ invalidate BTB entry + add r0, r0, #BTB_FLUSH_SIZE + mcr p15, 0, r0, c7, c5, 7 @ invalidate BTB entry + add r0, r0, #BTB_FLUSH_SIZE + mcr p15, 0, r0, c7, c5, 7 @ invalidate BTB entry + add r0, r0, #BTB_FLUSH_SIZE cmp r0, r1 blo 1b #ifdef HARVARD_CACHE diff --git a/arch/arm/mm/consistent.c b/arch/arm/mm/consistent.c index 26356ce..dbfe9e8 100644 --- a/arch/arm/mm/consistent.c +++ b/arch/arm/mm/consistent.c @@ -66,6 +66,7 @@ struct vm_region { unsigned long vm_start; unsigned long vm_end; struct page *vm_pages; + int vm_active; }; static struct vm_region consistent_head = { @@ -75,7 +76,7 @@ static struct vm_region consistent_head = { }; static struct vm_region * -vm_region_alloc(struct vm_region *head, size_t size, int gfp) +vm_region_alloc(struct vm_region *head, size_t size, gfp_t gfp) { unsigned long addr = head->vm_start, end = head->vm_end - size; unsigned long flags; @@ -104,6 +105,7 @@ vm_region_alloc(struct vm_region *head, size_t size, int gfp) list_add_tail(&new->vm_list, &c->vm_list); new->vm_start = addr; new->vm_end = addr + size; + new->vm_active = 1; spin_unlock_irqrestore(&consistent_lock, flags); return new; @@ -120,7 +122,7 @@ static struct vm_region *vm_region_find(struct vm_region *head, unsigned long ad struct vm_region *c; list_for_each_entry(c, &head->vm_list, vm_list) { - if (c->vm_start == addr) + if (c->vm_active && c->vm_start == addr) goto out; } c = NULL; @@ -133,7 +135,7 @@ static struct vm_region *vm_region_find(struct vm_region *head, unsigned long ad #endif static void * -__dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, int gfp, +__dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp, pgprot_t prot) { struct page *page; @@ -251,7 +253,7 @@ __dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, int gfp, * virtual and bus address for that space. */ void * -dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *handle, int gfp) +dma_alloc_coherent(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp) { return __dma_alloc(dev, size, handle, gfp, pgprot_noncached(pgprot_kernel)); @@ -263,7 +265,7 @@ EXPORT_SYMBOL(dma_alloc_coherent); * dma_alloc_coherent above. */ void * -dma_alloc_writecombine(struct device *dev, size_t size, dma_addr_t *handle, int gfp) +dma_alloc_writecombine(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp) { return __dma_alloc(dev, size, handle, gfp, pgprot_writecombine(pgprot_kernel)); @@ -319,6 +321,7 @@ EXPORT_SYMBOL(dma_mmap_writecombine); /* * free a page as defined by the above mapping. + * Must not be called with IRQs disabled. */ void dma_free_coherent(struct device *dev, size_t size, void *cpu_addr, dma_addr_t handle) { @@ -326,14 +329,18 @@ void dma_free_coherent(struct device *dev, size_t size, void *cpu_addr, dma_addr unsigned long flags, addr; pte_t *ptep; + WARN_ON(irqs_disabled()); + size = PAGE_ALIGN(size); spin_lock_irqsave(&consistent_lock, flags); - c = vm_region_find(&consistent_head, (unsigned long)cpu_addr); if (!c) goto no_area; + c->vm_active = 0; + spin_unlock_irqrestore(&consistent_lock, flags); + if ((c->vm_end - c->vm_start) != size) { printk(KERN_ERR "%s: freeing wrong coherent size (%ld != %d)\n", __func__, c->vm_end - c->vm_start, size); @@ -372,8 +379,8 @@ void dma_free_coherent(struct device *dev, size_t size, void *cpu_addr, dma_addr flush_tlb_kernel_range(c->vm_start, c->vm_end); + spin_lock_irqsave(&consistent_lock, flags); list_del(&c->vm_list); - spin_unlock_irqrestore(&consistent_lock, flags); kfree(c); @@ -397,8 +404,6 @@ static int __init consistent_init(void) pte_t *pte; int ret = 0; - spin_lock(&init_mm.page_table_lock); - do { pgd = pgd_offset(&init_mm, CONSISTENT_BASE); pmd = pmd_alloc(&init_mm, pgd, CONSISTENT_BASE); @@ -409,7 +414,7 @@ static int __init consistent_init(void) } WARN_ON(!pmd_none(*pmd)); - pte = pte_alloc_kernel(&init_mm, pmd, CONSISTENT_BASE); + pte = pte_alloc_kernel(pmd, CONSISTENT_BASE); if (!pte) { printk(KERN_ERR "%s: no pte tables\n", __func__); ret = -ENOMEM; @@ -419,8 +424,6 @@ static int __init consistent_init(void) consistent_pte = pte; } while (0); - spin_unlock(&init_mm.page_table_lock); - return ret; } diff --git a/arch/arm/mm/copypage-v6.c b/arch/arm/mm/copypage-v6.c index 27d0415..269ce69 100644 --- a/arch/arm/mm/copypage-v6.c +++ b/arch/arm/mm/copypage-v6.c @@ -22,9 +22,7 @@ #endif #define from_address (0xffff8000) -#define from_pgprot PAGE_KERNEL #define to_address (0xffffc000) -#define to_pgprot PAGE_KERNEL #define TOP_PTE(x) pte_offset_kernel(top_pmd, x) @@ -34,7 +32,7 @@ static DEFINE_SPINLOCK(v6_lock); * Copy the user page. No aliasing to deal with so we can just * attack the kernel's existing mapping of these pages. */ -void v6_copy_user_page_nonaliasing(void *kto, const void *kfrom, unsigned long vaddr) +static void v6_copy_user_page_nonaliasing(void *kto, const void *kfrom, unsigned long vaddr) { copy_page(kto, kfrom); } @@ -43,7 +41,7 @@ void v6_copy_user_page_nonaliasing(void *kto, const void *kfrom, unsigned long v * Clear the user page. No aliasing to deal with so we can just * attack the kernel's existing mapping of this page. */ -void v6_clear_user_page_nonaliasing(void *kaddr, unsigned long vaddr) +static void v6_clear_user_page_nonaliasing(void *kaddr, unsigned long vaddr) { clear_page(kaddr); } @@ -51,7 +49,7 @@ void v6_clear_user_page_nonaliasing(void *kaddr, unsigned long vaddr) /* * Copy the page, taking account of the cache colour. */ -void v6_copy_user_page_aliasing(void *kto, const void *kfrom, unsigned long vaddr) +static void v6_copy_user_page_aliasing(void *kto, const void *kfrom, unsigned long vaddr) { unsigned int offset = CACHE_COLOUR(vaddr); unsigned long from, to; @@ -72,8 +70,8 @@ void v6_copy_user_page_aliasing(void *kto, const void *kfrom, unsigned long vadd */ spin_lock(&v6_lock); - set_pte(TOP_PTE(from_address) + offset, pfn_pte(__pa(kfrom) >> PAGE_SHIFT, from_pgprot)); - set_pte(TOP_PTE(to_address) + offset, pfn_pte(__pa(kto) >> PAGE_SHIFT, to_pgprot)); + set_pte(TOP_PTE(from_address) + offset, pfn_pte(__pa(kfrom) >> PAGE_SHIFT, PAGE_KERNEL)); + set_pte(TOP_PTE(to_address) + offset, pfn_pte(__pa(kto) >> PAGE_SHIFT, PAGE_KERNEL)); from = from_address + (offset << PAGE_SHIFT); to = to_address + (offset << PAGE_SHIFT); @@ -91,7 +89,7 @@ void v6_copy_user_page_aliasing(void *kto, const void *kfrom, unsigned long vadd * so remap the kernel page into the same cache colour as the user * page. */ -void v6_clear_user_page_aliasing(void *kaddr, unsigned long vaddr) +static void v6_clear_user_page_aliasing(void *kaddr, unsigned long vaddr) { unsigned int offset = CACHE_COLOUR(vaddr); unsigned long to = to_address + (offset << PAGE_SHIFT); @@ -112,7 +110,7 @@ void v6_clear_user_page_aliasing(void *kaddr, unsigned long vaddr) */ spin_lock(&v6_lock); - set_pte(TOP_PTE(to_address) + offset, pfn_pte(__pa(kaddr) >> PAGE_SHIFT, to_pgprot)); + set_pte(TOP_PTE(to_address) + offset, pfn_pte(__pa(kaddr) >> PAGE_SHIFT, PAGE_KERNEL)); flush_tlb_kernel_page(to); clear_page((void *)to); diff --git a/arch/arm/mm/fault-armv.c b/arch/arm/mm/fault-armv.c index be4ab3d..7fc1b35 100644 --- a/arch/arm/mm/fault-armv.c +++ b/arch/arm/mm/fault-armv.c @@ -26,6 +26,11 @@ static unsigned long shared_pte_mask = L_PTE_CACHEABLE; /* * We take the easy way out of this problem - we make the * PTE uncacheable. However, we leave the write buffer on. + * + * Note that the pte lock held when calling update_mmu_cache must also + * guard the pte (somewhere else in the same mm) that we modify here. + * Therefore those configurations which might call adjust_pte (those + * without CONFIG_CPU_CACHE_VIPT) cannot support split page_table_lock. */ static int adjust_pte(struct vm_area_struct *vma, unsigned long address) { @@ -127,7 +132,7 @@ void __flush_dcache_page(struct address_space *mapping, struct page *page); * 2. If we have multiple shared mappings of the same space in * an object, we need to deal with the cache aliasing issues. * - * Note that the page_table_lock will be held. + * Note that the pte lock will be held. */ void update_mmu_cache(struct vm_area_struct *vma, unsigned long addr, pte_t pte) { diff --git a/arch/arm/mm/flush.c b/arch/arm/mm/flush.c index b0208c9..330695b 100644 --- a/arch/arm/mm/flush.c +++ b/arch/arm/mm/flush.c @@ -17,6 +17,24 @@ #ifdef CONFIG_CPU_CACHE_VIPT +#define ALIAS_FLUSH_START 0xffff4000 + +#define TOP_PTE(x) pte_offset_kernel(top_pmd, x) + +static void flush_pfn_alias(unsigned long pfn, unsigned long vaddr) +{ + unsigned long to = ALIAS_FLUSH_START + (CACHE_COLOUR(vaddr) << PAGE_SHIFT); + + set_pte(TOP_PTE(to), pfn_pte(pfn, PAGE_KERNEL)); + flush_tlb_kernel_page(to); + + asm( "mcrr p15, 0, %1, %0, c14\n" + " mcrr p15, 0, %1, %0, c5\n" + : + : "r" (to), "r" (to + PAGE_SIZE - L1_CACHE_BYTES) + : "cc"); +} + void flush_cache_mm(struct mm_struct *mm) { if (cache_is_vivt()) { @@ -67,24 +85,6 @@ void flush_cache_page(struct vm_area_struct *vma, unsigned long user_addr, unsig if (cache_is_vipt_aliasing()) flush_pfn_alias(pfn, user_addr); } - -#define ALIAS_FLUSH_START 0xffff4000 - -#define TOP_PTE(x) pte_offset_kernel(top_pmd, x) - -static void flush_pfn_alias(unsigned long pfn, unsigned long vaddr) -{ - unsigned long to = ALIAS_FLUSH_START + (CACHE_COLOUR(vaddr) << PAGE_SHIFT); - - set_pte(TOP_PTE(to), pfn_pte(pfn, PAGE_KERNEL)); - flush_tlb_kernel_page(to); - - asm( "mcrr p15, 0, %1, %0, c14\n" - " mcrr p15, 0, %1, %0, c5\n" - : - : "r" (to), "r" (to + PAGE_SIZE - L1_CACHE_BYTES) - : "cc"); -} #else #define flush_pfn_alias(pfn,vaddr) do { } while (0) #endif @@ -155,14 +155,19 @@ static void __flush_dcache_aliases(struct address_space *mapping, struct page *p * space mappings, we can be lazy and remember that we may have dirty * kernel cache lines for later. Otherwise, we assume we have * aliasing mappings. + * + * Note that we disable the lazy flush for SMP. */ void flush_dcache_page(struct page *page) { struct address_space *mapping = page_mapping(page); +#ifndef CONFIG_SMP if (mapping && !mapping_mapped(mapping)) set_bit(PG_dcache_dirty, &page->flags); - else { + else +#endif + { __flush_dcache_page(mapping, page); if (mapping && cache_is_vivt()) __flush_dcache_aliases(mapping, page); diff --git a/arch/arm/mm/init.c b/arch/arm/mm/init.c index edffa47..8b276ee 100644 --- a/arch/arm/mm/init.c +++ b/arch/arm/mm/init.c @@ -1,7 +1,7 @@ /* * linux/arch/arm/mm/init.c * - * Copyright (C) 1995-2002 Russell King + * Copyright (C) 1995-2005 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 @@ -86,14 +86,19 @@ void show_mem(void) printk("%d pages swap cached\n", cached); } -struct node_info { - unsigned int start; - unsigned int end; - int bootmap_pages; -}; +static inline pmd_t *pmd_off(pgd_t *pgd, unsigned long virt) +{ + return pmd_offset(pgd, virt); +} + +static inline pmd_t *pmd_off_k(unsigned long virt) +{ + return pmd_off(pgd_offset_k(virt), virt); +} -#define O_PFN_DOWN(x) ((x) >> PAGE_SHIFT) -#define O_PFN_UP(x) (PAGE_ALIGN(x) >> PAGE_SHIFT) +#define for_each_nodebank(iter,mi,no) \ + for (iter = 0; iter < mi->nr_banks; iter++) \ + if (mi->bank[iter].node == no) /* * FIXME: We really want to avoid allocating the bootmap bitmap @@ -106,15 +111,12 @@ find_bootmap_pfn(int node, struct meminfo *mi, unsigned int bootmap_pages) { unsigned int start_pfn, bank, bootmap_pfn; - start_pfn = O_PFN_UP(__pa(&_end)); + start_pfn = PAGE_ALIGN(__pa(&_end)) >> PAGE_SHIFT; bootmap_pfn = 0; - for (bank = 0; bank < mi->nr_banks; bank ++) { + for_each_nodebank(bank, mi, node) { unsigned int start, end; - if (mi->bank[bank].node != node) - continue; - start = mi->bank[bank].start >> PAGE_SHIFT; end = (mi->bank[bank].size + mi->bank[bank].start) >> PAGE_SHIFT; @@ -140,92 +142,6 @@ find_bootmap_pfn(int node, struct meminfo *mi, unsigned int bootmap_pages) return bootmap_pfn; } -/* - * Scan the memory info structure and pull out: - * - the end of memory - * - the number of nodes - * - the pfn range of each node - * - the number of bootmem bitmap pages - */ -static unsigned int __init -find_memend_and_nodes(struct meminfo *mi, struct node_info *np) -{ - unsigned int i, bootmem_pages = 0, memend_pfn = 0; - - for (i = 0; i < MAX_NUMNODES; i++) { - np[i].start = -1U; - np[i].end = 0; - np[i].bootmap_pages = 0; - } - - for (i = 0; i < mi->nr_banks; i++) { - unsigned long start, end; - int node; - - if (mi->bank[i].size == 0) { - /* - * Mark this bank with an invalid node number - */ - mi->bank[i].node = -1; - continue; - } - - node = mi->bank[i].node; - - /* - * Make sure we haven't exceeded the maximum number of nodes - * that we have in this configuration. If we have, we're in - * trouble. (maybe we ought to limit, instead of bugging?) - */ - if (node >= MAX_NUMNODES) - BUG(); - node_set_online(node); - - /* - * Get the start and end pfns for this bank - */ - start = mi->bank[i].start >> PAGE_SHIFT; - end = (mi->bank[i].start + mi->bank[i].size) >> PAGE_SHIFT; - - if (np[node].start > start) - np[node].start = start; - - if (np[node].end < end) - np[node].end = end; - - if (memend_pfn < end) - memend_pfn = end; - } - - /* - * Calculate the number of pages we require to - * store the bootmem bitmaps. - */ - for_each_online_node(i) { - if (np[i].end == 0) - continue; - - np[i].bootmap_pages = bootmem_bootmap_pages(np[i].end - - np[i].start); - bootmem_pages += np[i].bootmap_pages; - } - - high_memory = __va(memend_pfn << PAGE_SHIFT); - - /* - * This doesn't seem to be used by the Linux memory - * manager any more. If we can get rid of it, we - * also get rid of some of the stuff above as well. - * - * Note: max_low_pfn and max_pfn reflect the number - * of _pages_ in the system, not the maximum PFN. - */ - max_low_pfn = memend_pfn - O_PFN_DOWN(PHYS_OFFSET); - max_pfn = memend_pfn - O_PFN_DOWN(PHYS_OFFSET); - - return bootmem_pages; -} - static int __init check_initrd(struct meminfo *mi) { int initrd_node = -2; @@ -266,9 +182,8 @@ static int __init check_initrd(struct meminfo *mi) /* * Reserve the various regions of node 0 */ -static __init void reserve_node_zero(unsigned int bootmap_pfn, unsigned int bootmap_pages) +static __init void reserve_node_zero(pg_data_t *pgdat) { - pg_data_t *pgdat = NODE_DATA(0); unsigned long res_size = 0; /* @@ -289,13 +204,6 @@ static __init void reserve_node_zero(unsigned int bootmap_pfn, unsigned int boot PTRS_PER_PGD * sizeof(pgd_t)); /* - * And don't forget to reserve the allocator bitmap, - * which will be freed later. - */ - reserve_bootmem_node(pgdat, bootmap_pfn << PAGE_SHIFT, - bootmap_pages << PAGE_SHIFT); - - /* * Hmm... This should go elsewhere, but we really really need to * stop things allocating the low memory; ideally we need a better * implementation of GFP_DMA which does not assume that DMA-able @@ -324,183 +232,293 @@ static __init void reserve_node_zero(unsigned int bootmap_pfn, unsigned int boot reserve_bootmem_node(pgdat, PHYS_OFFSET, res_size); } -/* - * Register all available RAM in this node with the bootmem allocator. - */ -static inline void free_bootmem_node_bank(int node, struct meminfo *mi) +void __init build_mem_type_table(void); +void __init create_mapping(struct map_desc *md); + +static unsigned long __init +bootmem_init_node(int node, int initrd_node, struct meminfo *mi) { - pg_data_t *pgdat = NODE_DATA(node); - int bank; + unsigned long zone_size[MAX_NR_ZONES], zhole_size[MAX_NR_ZONES]; + unsigned long start_pfn, end_pfn, boot_pfn; + unsigned int boot_pages; + pg_data_t *pgdat; + int i; - for (bank = 0; bank < mi->nr_banks; bank++) - if (mi->bank[bank].node == node) - free_bootmem_node(pgdat, mi->bank[bank].start, - mi->bank[bank].size); -} + start_pfn = -1UL; + end_pfn = 0; -/* - * Initialise the bootmem allocator for all nodes. This is called - * early during the architecture specific initialisation. - */ -static void __init bootmem_init(struct meminfo *mi) -{ - struct node_info node_info[MAX_NUMNODES], *np = node_info; - unsigned int bootmap_pages, bootmap_pfn, map_pg; - int node, initrd_node; + /* + * Calculate the pfn range, and map the memory banks for this node. + */ + for_each_nodebank(i, mi, node) { + unsigned long start, end; + struct map_desc map; + + start = mi->bank[i].start >> PAGE_SHIFT; + end = (mi->bank[i].start + mi->bank[i].size) >> PAGE_SHIFT; + + if (start_pfn > start) + start_pfn = start; + if (end_pfn < end) + end_pfn = end; - bootmap_pages = find_memend_and_nodes(mi, np); - bootmap_pfn = find_bootmap_pfn(0, mi, bootmap_pages); - initrd_node = check_initrd(mi); + map.pfn = __phys_to_pfn(mi->bank[i].start); + map.virtual = __phys_to_virt(mi->bank[i].start); + map.length = mi->bank[i].size; + map.type = MT_MEMORY; - map_pg = bootmap_pfn; + create_mapping(&map); + } /* - * Initialise the bootmem nodes. - * - * What we really want to do is: - * - * unmap_all_regions_except_kernel(); - * for_each_node_in_reverse_order(node) { - * map_node(node); - * allocate_bootmem_map(node); - * init_bootmem_node(node); - * free_bootmem_node(node); - * } - * - * but this is a 2.5-type change. For now, we just set - * the nodes up in reverse order. - * - * (we could also do with rolling bootmem_init and paging_init - * into one generic "memory_init" type function). + * If there is no memory in this node, ignore it. */ - np += num_online_nodes() - 1; - for (node = num_online_nodes() - 1; node >= 0; node--, np--) { - /* - * If there are no pages in this node, ignore it. - * Note that node 0 must always have some pages. - */ - if (np->end == 0 || !node_online(node)) { - if (node == 0) - BUG(); - continue; - } + if (end_pfn == 0) + return end_pfn; - /* - * Initialise the bootmem allocator. - */ - init_bootmem_node(NODE_DATA(node), map_pg, np->start, np->end); - free_bootmem_node_bank(node, mi); - map_pg += np->bootmap_pages; + /* + * Allocate the bootmem bitmap page. + */ + boot_pages = bootmem_bootmap_pages(end_pfn - start_pfn); + boot_pfn = find_bootmap_pfn(node, mi, boot_pages); - /* - * If this is node 0, we need to reserve some areas ASAP - - * we may use bootmem on node 0 to setup the other nodes. - */ - if (node == 0) - reserve_node_zero(bootmap_pfn, bootmap_pages); - } + /* + * Initialise the bootmem allocator for this node, handing the + * memory banks over to bootmem. + */ + node_set_online(node); + pgdat = NODE_DATA(node); + init_bootmem_node(pgdat, boot_pfn, start_pfn, end_pfn); + for_each_nodebank(i, mi, node) + free_bootmem_node(pgdat, mi->bank[i].start, mi->bank[i].size); + + /* + * Reserve the bootmem bitmap for this node. + */ + reserve_bootmem_node(pgdat, boot_pfn << PAGE_SHIFT, + boot_pages << PAGE_SHIFT); #ifdef CONFIG_BLK_DEV_INITRD - if (phys_initrd_size && initrd_node >= 0) { - reserve_bootmem_node(NODE_DATA(initrd_node), phys_initrd_start, + /* + * If the initrd is in this node, reserve its memory. + */ + if (node == initrd_node) { + reserve_bootmem_node(pgdat, phys_initrd_start, phys_initrd_size); initrd_start = __phys_to_virt(phys_initrd_start); initrd_end = initrd_start + phys_initrd_size; } #endif - BUG_ON(map_pg != bootmap_pfn + bootmap_pages); + /* + * Finally, reserve any node zero regions. + */ + if (node == 0) + reserve_node_zero(pgdat); + + /* + * initialise the zones within this node. + */ + memset(zone_size, 0, sizeof(zone_size)); + memset(zhole_size, 0, sizeof(zhole_size)); + + /* + * The size of this node has already been determined. If we need + * to do anything fancy with the allocation of this memory to the + * zones, now is the time to do it. + */ + zone_size[0] = end_pfn - start_pfn; + + /* + * For each bank in this node, calculate the size of the holes. + * holes = node_size - sum(bank_sizes_in_node) + */ + zhole_size[0] = zone_size[0]; + for_each_nodebank(i, mi, node) + zhole_size[0] -= mi->bank[i].size >> PAGE_SHIFT; + + /* + * Adjust the sizes according to any special requirements for + * this machine type. + */ + arch_adjust_zones(node, zone_size, zhole_size); + + free_area_init_node(node, pgdat, zone_size, start_pfn, zhole_size); + + return end_pfn; } -/* - * paging_init() sets up the page tables, initialises the zone memory - * maps, and sets up the zero page, bad page and bad page tables. - */ -void __init paging_init(struct meminfo *mi, struct machine_desc *mdesc) +static void __init bootmem_init(struct meminfo *mi) { - void *zero_page; - int node; + unsigned long addr, memend_pfn = 0; + int node, initrd_node, i; - bootmem_init(mi); + /* + * Invalidate the node number for empty or invalid memory banks + */ + for (i = 0; i < mi->nr_banks; i++) + if (mi->bank[i].size == 0 || mi->bank[i].node >= MAX_NUMNODES) + mi->bank[i].node = -1; memcpy(&meminfo, mi, sizeof(meminfo)); /* - * allocate the zero page. Note that we count on this going ok. + * Clear out all the mappings below the kernel image. */ - zero_page = alloc_bootmem_low_pages(PAGE_SIZE); + for (addr = 0; addr < MODULE_START; addr += PGDIR_SIZE) + pmd_clear(pmd_off_k(addr)); +#ifdef CONFIG_XIP_KERNEL + /* The XIP kernel is mapped in the module area -- skip over it */ + addr = ((unsigned long)&_etext + PGDIR_SIZE - 1) & PGDIR_MASK; +#endif + for ( ; addr < PAGE_OFFSET; addr += PGDIR_SIZE) + pmd_clear(pmd_off_k(addr)); /* - * initialise the page tables. + * Clear out all the kernel space mappings, except for the first + * memory bank, up to the end of the vmalloc region. */ - memtable_init(mi); - if (mdesc->map_io) - mdesc->map_io(); - local_flush_tlb_all(); + for (addr = __phys_to_virt(mi->bank[0].start + mi->bank[0].size); + addr < VMALLOC_END; addr += PGDIR_SIZE) + pmd_clear(pmd_off_k(addr)); /* - * initialise the zones within each node + * Locate which node contains the ramdisk image, if any. */ - for_each_online_node(node) { - unsigned long zone_size[MAX_NR_ZONES]; - unsigned long zhole_size[MAX_NR_ZONES]; - struct bootmem_data *bdata; - pg_data_t *pgdat; - int i; + initrd_node = check_initrd(mi); - /* - * Initialise the zone size information. - */ - for (i = 0; i < MAX_NR_ZONES; i++) { - zone_size[i] = 0; - zhole_size[i] = 0; - } + /* + * Run through each node initialising the bootmem allocator. + */ + for_each_node(node) { + unsigned long end_pfn; - pgdat = NODE_DATA(node); - bdata = pgdat->bdata; + end_pfn = bootmem_init_node(node, initrd_node, mi); /* - * The size of this node has already been determined. - * If we need to do anything fancy with the allocation - * of this memory to the zones, now is the time to do - * it. + * Remember the highest memory PFN. */ - zone_size[0] = bdata->node_low_pfn - - (bdata->node_boot_start >> PAGE_SHIFT); + if (end_pfn > memend_pfn) + memend_pfn = end_pfn; + } - /* - * If this zone has zero size, skip it. - */ - if (!zone_size[0]) - continue; + high_memory = __va(memend_pfn << PAGE_SHIFT); - /* - * For each bank in this node, calculate the size of the - * holes. holes = node_size - sum(bank_sizes_in_node) - */ - zhole_size[0] = zone_size[0]; - for (i = 0; i < mi->nr_banks; i++) { - if (mi->bank[i].node != node) - continue; + /* + * This doesn't seem to be used by the Linux memory manager any + * more, but is used by ll_rw_block. If we can get rid of it, we + * also get rid of some of the stuff above as well. + * + * Note: max_low_pfn and max_pfn reflect the number of _pages_ in + * the system, not the maximum PFN. + */ + max_pfn = max_low_pfn = memend_pfn - PHYS_PFN_OFFSET; +} - zhole_size[0] -= mi->bank[i].size >> PAGE_SHIFT; - } +/* + * Set up device the mappings. Since we clear out the page tables for all + * mappings above VMALLOC_END, we will remove any debug device mappings. + * This means you have to be careful how you debug this function, or any + * called function. This means you can't use any function or debugging + * method which may touch any device, otherwise the kernel _will_ crash. + */ +static void __init devicemaps_init(struct machine_desc *mdesc) +{ + struct map_desc map; + unsigned long addr; + void *vectors; - /* - * Adjust the sizes according to any special - * requirements for this machine type. - */ - arch_adjust_zones(node, zone_size, zhole_size); + /* + * Allocate the vector page early. + */ + vectors = alloc_bootmem_low_pages(PAGE_SIZE); + BUG_ON(!vectors); + + for (addr = VMALLOC_END; addr; addr += PGDIR_SIZE) + pmd_clear(pmd_off_k(addr)); - free_area_init_node(node, pgdat, zone_size, - bdata->node_boot_start >> PAGE_SHIFT, zhole_size); + /* + * Map the kernel if it is XIP. + * It is always first in the modulearea. + */ +#ifdef CONFIG_XIP_KERNEL + map.pfn = __phys_to_pfn(CONFIG_XIP_PHYS_ADDR & PGDIR_MASK); + map.virtual = MODULE_START; + map.length = ((unsigned long)&_etext - map.virtual + ~PGDIR_MASK) & PGDIR_MASK; + map.type = MT_ROM; + create_mapping(&map); +#endif + + /* + * Map the cache flushing regions. + */ +#ifdef FLUSH_BASE + map.pfn = __phys_to_pfn(FLUSH_BASE_PHYS); + map.virtual = FLUSH_BASE; + map.length = PGDIR_SIZE; + map.type = MT_CACHECLEAN; + create_mapping(&map); +#endif +#ifdef FLUSH_BASE_MINICACHE + map.pfn = __phys_to_pfn(FLUSH_BASE_PHYS + PGDIR_SIZE); + map.virtual = FLUSH_BASE_MINICACHE; + map.length = PGDIR_SIZE; + map.type = MT_MINICLEAN; + create_mapping(&map); +#endif + + /* + * Create a mapping for the machine vectors at the high-vectors + * location (0xffff0000). If we aren't using high-vectors, also + * create a mapping at the low-vectors virtual address. + */ + map.pfn = __phys_to_pfn(virt_to_phys(vectors)); + map.virtual = 0xffff0000; + map.length = PAGE_SIZE; + map.type = MT_HIGH_VECTORS; + create_mapping(&map); + + if (!vectors_high()) { + map.virtual = 0; + map.type = MT_LOW_VECTORS; + create_mapping(&map); } /* - * finish off the bad pages once - * the mem_map is initialised + * Ask the machine support to map in the statically mapped devices. + */ + if (mdesc->map_io) + mdesc->map_io(); + + /* + * Finally flush the caches and tlb to ensure that we're in a + * consistent state wrt the writebuffer. This also ensures that + * any write-allocated cache lines in the vector page are written + * back. After this point, we can start to touch devices again. */ + local_flush_tlb_all(); + flush_cache_all(); +} + +/* + * paging_init() sets up the page tables, initialises the zone memory + * maps, and sets up the zero page, bad page and bad page tables. + */ +void __init paging_init(struct meminfo *mi, struct machine_desc *mdesc) +{ + void *zero_page; + + build_mem_type_table(); + bootmem_init(mi); + devicemaps_init(mdesc); + + top_pmd = pmd_off_k(0xffff0000); + + /* + * allocate the zero page. Note that we count on this going ok. + */ + zero_page = alloc_bootmem_low_pages(PAGE_SIZE); memzero(zero_page, PAGE_SIZE); empty_zero_page = virt_to_page(zero_page); flush_dcache_page(empty_zero_page); @@ -562,10 +580,7 @@ static void __init free_unused_memmap_node(int node, struct meminfo *mi) * may not be the case, especially if the user has provided the * information on the command line. */ - for (i = 0; i < mi->nr_banks; i++) { - if (mi->bank[i].size == 0 || mi->bank[i].node != node) - continue; - + for_each_nodebank(i, mi, node) { bank_start = mi->bank[i].start >> PAGE_SHIFT; if (bank_start < prev_bank_end) { printk(KERN_ERR "MEM: unordered memory banks. " diff --git a/arch/arm/mm/ioremap.c b/arch/arm/mm/ioremap.c index 7110e54..1090139 100644 --- a/arch/arm/mm/ioremap.c +++ b/arch/arm/mm/ioremap.c @@ -26,6 +26,7 @@ #include <linux/vmalloc.h> #include <asm/cacheflush.h> +#include <asm/hardware.h> #include <asm/io.h> #include <asm/tlbflush.h> @@ -74,7 +75,7 @@ remap_area_pmd(pmd_t * pmd, unsigned long address, unsigned long size, pgprot = __pgprot(L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY | L_PTE_WRITE | flags); do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, address + phys_addr, pgprot); @@ -96,7 +97,6 @@ remap_area_pages(unsigned long start, unsigned long phys_addr, phys_addr -= address; dir = pgd_offset(&init_mm, address); BUG_ON(address >= end); - spin_lock(&init_mm.page_table_lock); do { pmd_t *pmd = pmd_alloc(&init_mm, dir, address); if (!pmd) { @@ -113,7 +113,6 @@ remap_area_pages(unsigned long start, unsigned long phys_addr, dir++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); flush_cache_vmap(start, end); return err; } @@ -131,8 +130,7 @@ remap_area_pages(unsigned long start, unsigned long phys_addr, * mapping. See include/asm-arm/proc-armv/pgtable.h for more information. */ void __iomem * -__ioremap(unsigned long phys_addr, size_t size, unsigned long flags, - unsigned long align) +__ioremap(unsigned long phys_addr, size_t size, unsigned long flags) { void * addr; struct vm_struct * area; diff --git a/arch/arm/mm/mm-armv.c b/arch/arm/mm/mm-armv.c index d125a3d..9e50127 100644 --- a/arch/arm/mm/mm-armv.c +++ b/arch/arm/mm/mm-armv.c @@ -1,7 +1,7 @@ /* * linux/arch/arm/mm/mm-armv.c * - * Copyright (C) 1998-2002 Russell King + * Copyright (C) 1998-2005 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 @@ -180,11 +180,6 @@ pgd_t *get_pgd_slow(struct mm_struct *mm) if (!vectors_high()) { /* - * This lock is here just to satisfy pmd_alloc and pte_lock - */ - spin_lock(&mm->page_table_lock); - - /* * On ARM, first page must always be allocated since it * contains the machine vectors. */ @@ -201,23 +196,14 @@ pgd_t *get_pgd_slow(struct mm_struct *mm) set_pte(new_pte, *init_pte); pte_unmap_nested(init_pte); pte_unmap(new_pte); - - spin_unlock(&mm->page_table_lock); } return new_pgd; no_pte: - spin_unlock(&mm->page_table_lock); pmd_free(new_pmd); - free_pages((unsigned long)new_pgd, 2); - return NULL; - no_pmd: - spin_unlock(&mm->page_table_lock); free_pages((unsigned long)new_pgd, 2); - return NULL; - no_pgd: return NULL; } @@ -243,6 +229,7 @@ void free_pgd_slow(pgd_t *pgd) pte = pmd_page(*pmd); pmd_clear(pmd); dec_page_state(nr_page_table_pages); + pte_lock_deinit(pte); pte_free(pte); pmd_free(pmd); free: @@ -305,16 +292,6 @@ alloc_init_page(unsigned long virt, unsigned long phys, unsigned int prot_l1, pg set_pte(ptep, pfn_pte(phys >> PAGE_SHIFT, prot)); } -/* - * Clear any PGD mapping. On a two-level page table system, - * the clearance is done by the middle-level functions (pmd) - * rather than the top-level (pgd) functions. - */ -static inline void clear_mapping(unsigned long virt) -{ - pmd_clear(pmd_off_k(virt)); -} - struct mem_types { unsigned int prot_pte; unsigned int prot_l1; @@ -373,11 +350,11 @@ static struct mem_types mem_types[] __initdata = { /* * Adjust the PMD section entries according to the CPU in use. */ -static void __init build_mem_type_table(void) +void __init build_mem_type_table(void) { struct cachepolicy *cp; unsigned int cr = get_cr(); - unsigned int user_pgprot; + unsigned int user_pgprot, kern_pgprot; int cpu_arch = cpu_architecture(); int i; @@ -404,7 +381,7 @@ static void __init build_mem_type_table(void) } cp = &cache_policies[cachepolicy]; - user_pgprot = cp->pte; + kern_pgprot = user_pgprot = cp->pte; /* * ARMv6 and above have extended page tables. @@ -416,6 +393,7 @@ static void __init build_mem_type_table(void) */ mem_types[MT_MEMORY].prot_sect &= ~PMD_BIT4; mem_types[MT_ROM].prot_sect &= ~PMD_BIT4; + /* * Mark cache clean areas and XIP ROM read only * from SVC mode and no access from userspace. @@ -435,32 +413,47 @@ static void __init build_mem_type_table(void) * (iow, non-global) */ user_pgprot |= L_PTE_ASID; + +#ifdef CONFIG_SMP + /* + * Mark memory with the "shared" attribute for SMP systems + */ + user_pgprot |= L_PTE_SHARED; + kern_pgprot |= L_PTE_SHARED; + mem_types[MT_MEMORY].prot_sect |= PMD_SECT_S; +#endif } + for (i = 0; i < 16; i++) { + unsigned long v = pgprot_val(protection_map[i]); + v = (v & ~(L_PTE_BUFFERABLE|L_PTE_CACHEABLE)) | user_pgprot; + protection_map[i] = __pgprot(v); + } + + mem_types[MT_LOW_VECTORS].prot_pte |= kern_pgprot; + mem_types[MT_HIGH_VECTORS].prot_pte |= kern_pgprot; + if (cpu_arch >= CPU_ARCH_ARMv5) { - mem_types[MT_LOW_VECTORS].prot_pte |= cp->pte & PTE_CACHEABLE; - mem_types[MT_HIGH_VECTORS].prot_pte |= cp->pte & PTE_CACHEABLE; +#ifndef CONFIG_SMP + /* + * Only use write-through for non-SMP systems + */ + mem_types[MT_LOW_VECTORS].prot_pte &= ~L_PTE_BUFFERABLE; + mem_types[MT_HIGH_VECTORS].prot_pte &= ~L_PTE_BUFFERABLE; +#endif } else { - mem_types[MT_LOW_VECTORS].prot_pte |= cp->pte; - mem_types[MT_HIGH_VECTORS].prot_pte |= cp->pte; mem_types[MT_MINICLEAN].prot_sect &= ~PMD_SECT_TEX(1); } + pgprot_kernel = __pgprot(L_PTE_PRESENT | L_PTE_YOUNG | + L_PTE_DIRTY | L_PTE_WRITE | + L_PTE_EXEC | kern_pgprot); + mem_types[MT_LOW_VECTORS].prot_l1 |= ecc_mask; mem_types[MT_HIGH_VECTORS].prot_l1 |= ecc_mask; mem_types[MT_MEMORY].prot_sect |= ecc_mask | cp->pmd; mem_types[MT_ROM].prot_sect |= cp->pmd; - for (i = 0; i < 16; i++) { - unsigned long v = pgprot_val(protection_map[i]); - v = (v & ~(PTE_BUFFERABLE|PTE_CACHEABLE)) | user_pgprot; - protection_map[i] = __pgprot(v); - } - - pgprot_kernel = __pgprot(L_PTE_PRESENT | L_PTE_YOUNG | - L_PTE_DIRTY | L_PTE_WRITE | - L_PTE_EXEC | cp->pte); - switch (cp->pmd) { case PMD_SECT_WT: mem_types[MT_CACHECLEAN].prot_sect |= PMD_SECT_WT; @@ -483,25 +476,25 @@ static void __init build_mem_type_table(void) * offsets, and we take full advantage of sections and * supersections. */ -static void __init create_mapping(struct map_desc *md) +void __init create_mapping(struct map_desc *md) { unsigned long virt, length; int prot_sect, prot_l1, domain; pgprot_t prot_pte; - long off; + unsigned long off = (u32)__pfn_to_phys(md->pfn); if (md->virtual != vectors_base() && md->virtual < TASK_SIZE) { printk(KERN_WARNING "BUG: not creating mapping for " - "0x%08lx at 0x%08lx in user region\n", - md->physical, md->virtual); + "0x%08llx at 0x%08lx in user region\n", + __pfn_to_phys((u64)md->pfn), md->virtual); return; } if ((md->type == MT_DEVICE || md->type == MT_ROM) && md->virtual >= PAGE_OFFSET && md->virtual < VMALLOC_END) { - printk(KERN_WARNING "BUG: mapping for 0x%08lx at 0x%08lx " + printk(KERN_WARNING "BUG: mapping for 0x%08llx at 0x%08lx " "overlaps vmalloc space\n", - md->physical, md->virtual); + __pfn_to_phys((u64)md->pfn), md->virtual); } domain = mem_types[md->type].domain; @@ -509,15 +502,40 @@ static void __init create_mapping(struct map_desc *md) prot_l1 = mem_types[md->type].prot_l1 | PMD_DOMAIN(domain); prot_sect = mem_types[md->type].prot_sect | PMD_DOMAIN(domain); + /* + * Catch 36-bit addresses + */ + if(md->pfn >= 0x100000) { + if(domain) { + printk(KERN_ERR "MM: invalid domain in supersection " + "mapping for 0x%08llx at 0x%08lx\n", + __pfn_to_phys((u64)md->pfn), md->virtual); + return; + } + if((md->virtual | md->length | __pfn_to_phys(md->pfn)) + & ~SUPERSECTION_MASK) { + printk(KERN_ERR "MM: cannot create mapping for " + "0x%08llx at 0x%08lx invalid alignment\n", + __pfn_to_phys((u64)md->pfn), md->virtual); + return; + } + + /* + * Shift bits [35:32] of address into bits [23:20] of PMD + * (See ARMv6 spec). + */ + off |= (((md->pfn >> (32 - PAGE_SHIFT)) & 0xF) << 20); + } + virt = md->virtual; - off = md->physical - virt; + off -= virt; length = md->length; if (mem_types[md->type].prot_l1 == 0 && (virt & 0xfffff || (virt + off) & 0xfffff || (virt + length) & 0xfffff)) { printk(KERN_WARNING "BUG: map for 0x%08lx at 0x%08lx can not " "be mapped using pages, ignoring.\n", - md->physical, md->virtual); + __pfn_to_phys(md->pfn), md->virtual); return; } @@ -535,13 +553,22 @@ static void __init create_mapping(struct map_desc *md) * of the actual domain assignments in use. */ if (cpu_architecture() >= CPU_ARCH_ARMv6 && domain == 0) { - /* Align to supersection boundary */ - while ((virt & ~SUPERSECTION_MASK || (virt + off) & - ~SUPERSECTION_MASK) && length >= (PGDIR_SIZE / 2)) { - alloc_init_section(virt, virt + off, prot_sect); - - virt += (PGDIR_SIZE / 2); - length -= (PGDIR_SIZE / 2); + /* + * Align to supersection boundary if !high pages. + * High pages have already been checked for proper + * alignment above and they will fail the SUPSERSECTION_MASK + * check because of the way the address is encoded into + * offset. + */ + if (md->pfn <= 0x100000) { + while ((virt & ~SUPERSECTION_MASK || + (virt + off) & ~SUPERSECTION_MASK) && + length >= (PGDIR_SIZE / 2)) { + alloc_init_section(virt, virt + off, prot_sect); + + virt += (PGDIR_SIZE / 2); + length -= (PGDIR_SIZE / 2); + } } while (length >= SUPERSECTION_SIZE) { @@ -601,100 +628,6 @@ void setup_mm_for_reboot(char mode) } } -extern void _stext, _etext; - -/* - * Setup initial mappings. We use the page we allocated for zero page to hold - * the mappings, which will get overwritten by the vectors in traps_init(). - * The mappings must be in virtual address order. - */ -void __init memtable_init(struct meminfo *mi) -{ - struct map_desc *init_maps, *p, *q; - unsigned long address = 0; - int i; - - build_mem_type_table(); - - init_maps = p = alloc_bootmem_low_pages(PAGE_SIZE); - -#ifdef CONFIG_XIP_KERNEL - p->physical = CONFIG_XIP_PHYS_ADDR & PMD_MASK; - p->virtual = (unsigned long)&_stext & PMD_MASK; - p->length = ((unsigned long)&_etext - p->virtual + ~PMD_MASK) & PMD_MASK; - p->type = MT_ROM; - p ++; -#endif - - for (i = 0; i < mi->nr_banks; i++) { - if (mi->bank[i].size == 0) - continue; - - p->physical = mi->bank[i].start; - p->virtual = __phys_to_virt(p->physical); - p->length = mi->bank[i].size; - p->type = MT_MEMORY; - p ++; - } - -#ifdef FLUSH_BASE - p->physical = FLUSH_BASE_PHYS; - p->virtual = FLUSH_BASE; - p->length = PGDIR_SIZE; - p->type = MT_CACHECLEAN; - p ++; -#endif - -#ifdef FLUSH_BASE_MINICACHE - p->physical = FLUSH_BASE_PHYS + PGDIR_SIZE; - p->virtual = FLUSH_BASE_MINICACHE; - p->length = PGDIR_SIZE; - p->type = MT_MINICLEAN; - p ++; -#endif - - /* - * Go through the initial mappings, but clear out any - * pgdir entries that are not in the description. - */ - q = init_maps; - do { - if (address < q->virtual || q == p) { - clear_mapping(address); - address += PGDIR_SIZE; - } else { - create_mapping(q); - - address = q->virtual + q->length; - address = (address + PGDIR_SIZE - 1) & PGDIR_MASK; - - q ++; - } - } while (address != 0); - - /* - * Create a mapping for the machine vectors at the high-vectors - * location (0xffff0000). If we aren't using high-vectors, also - * create a mapping at the low-vectors virtual address. - */ - init_maps->physical = virt_to_phys(init_maps); - init_maps->virtual = 0xffff0000; - init_maps->length = PAGE_SIZE; - init_maps->type = MT_HIGH_VECTORS; - create_mapping(init_maps); - - if (!vectors_high()) { - init_maps->virtual = 0; - init_maps->type = MT_LOW_VECTORS; - create_mapping(init_maps); - } - - flush_cache_all(); - local_flush_tlb_all(); - - top_pmd = pmd_off_k(0xffff0000); -} - /* * Create the architecture specific mappings */ diff --git a/arch/arm/mm/proc-v6.S b/arch/arm/mm/proc-v6.S index caf3b19..92f3ca3 100644 --- a/arch/arm/mm/proc-v6.S +++ b/arch/arm/mm/proc-v6.S @@ -12,6 +12,7 @@ #include <linux/linkage.h> #include <asm/assembler.h> #include <asm/asm-offsets.h> +#include <asm/hardware/arm_scu.h> #include <asm/procinfo.h> #include <asm/pgtable.h> @@ -55,7 +56,14 @@ ENTRY(cpu_v6_proc_init) mov pc, lr ENTRY(cpu_v6_proc_fin) - mov pc, lr + stmfd sp!, {lr} + cpsid if @ disable interrupts + bl v6_flush_kern_cache_all + mrc p15, 0, r0, c1, c0, 0 @ ctrl register + bic r0, r0, #0x1000 @ ...i............ + bic r0, r0, #0x0006 @ .............ca. + mcr p15, 0, r0, c1, c0, 0 @ disable caches + ldmfd sp!, {pc} /* * cpu_v6_reset(loc) @@ -105,6 +113,9 @@ ENTRY(cpu_v6_dcache_clean_area) ENTRY(cpu_v6_switch_mm) mov r2, #0 ldr r1, [r1, #MM_CONTEXT_ID] @ get mm->context.id +#ifdef CONFIG_SMP + orr r0, r0, #2 @ set shared pgtable +#endif mcr p15, 0, r2, c7, c5, 6 @ flush BTAC/BTB mcr p15, 0, r2, c7, c10, 4 @ drain write buffer mcr p15, 0, r0, c2, c0, 0 @ set TTB 0 @@ -133,7 +144,7 @@ ENTRY(cpu_v6_switch_mm) ENTRY(cpu_v6_set_pte) str r1, [r0], #-2048 @ linux version - bic r2, r1, #0x000007f0 + bic r2, r1, #0x000003f0 bic r2, r2, #0x00000003 orr r2, r2, #PTE_EXT_AP0 | 2 @@ -184,6 +195,23 @@ cpu_v6_name: * - cache type register is implemented */ __v6_setup: +#ifdef CONFIG_SMP + /* Set up the SCU on core 0 only */ + mrc p15, 0, r0, c0, c0, 5 @ CPU core number + ands r0, r0, #15 + moveq r0, #0x10000000 @ SCU_BASE + orreq r0, r0, #0x00100000 + ldreq r5, [r0, #SCU_CTRL] + orreq r5, r5, #1 + streq r5, [r0, #SCU_CTRL] + +#ifndef CONFIG_CPU_DCACHE_DISABLE + mrc p15, 0, r0, c1, c0, 1 @ Enable SMP/nAMP mode + orr r0, r0, #0x20 + mcr p15, 0, r0, c1, c0, 1 +#endif +#endif + mov r0, #0 mcr p15, 0, r0, c7, c14, 0 @ clean+invalidate D cache mcr p15, 0, r0, c7, c5, 0 @ invalidate I cache @@ -191,6 +219,9 @@ __v6_setup: mcr p15, 0, r0, c7, c10, 4 @ drain write buffer mcr p15, 0, r0, c8, c7, 0 @ invalidate I + D TLBs mcr p15, 0, r0, c2, c0, 2 @ TTB control register +#ifdef CONFIG_SMP + orr r4, r4, #2 @ set shared pgtable +#endif mcr p15, 0, r4, c2, c0, 1 @ load TTB1 #ifdef CONFIG_VFP mrc p15, 0, r0, c1, c0, 2 |