aboutsummaryrefslogtreecommitdiff
path: root/arch/sh/mm
diff options
context:
space:
mode:
Diffstat (limited to 'arch/sh/mm')
-rw-r--r--arch/sh/mm/Kconfig12
-rw-r--r--arch/sh/mm/Makefile8
-rw-r--r--arch/sh/mm/alignment.c2
-rw-r--r--arch/sh/mm/cache-debugfs.c2
-rw-r--r--arch/sh/mm/cache-sh2.c4
-rw-r--r--arch/sh/mm/cache-sh2a.c6
-rw-r--r--arch/sh/mm/cache-sh4.c9
-rw-r--r--arch/sh/mm/cache-shx3.c4
-rw-r--r--arch/sh/mm/cache.c16
-rw-r--r--arch/sh/mm/consistent.c6
-rw-r--r--arch/sh/mm/fault.c517
-rw-r--r--arch/sh/mm/fault_32.c374
-rw-r--r--arch/sh/mm/fault_64.c266
-rw-r--r--arch/sh/mm/flush-sh4.c2
-rw-r--r--arch/sh/mm/init.c82
-rw-r--r--arch/sh/mm/mmap.c139
-rw-r--r--arch/sh/mm/pmb.c1
-rw-r--r--arch/sh/mm/sram.c1
-rw-r--r--arch/sh/mm/tlb-pteaex.c1
-rw-r--r--arch/sh/mm/tlb-sh3.c1
-rw-r--r--arch/sh/mm/tlb-sh4.c1
-rw-r--r--arch/sh/mm/tlb-sh5.c42
-rw-r--r--arch/sh/mm/tlbex_32.c78
-rw-r--r--arch/sh/mm/tlbex_64.c166
-rw-r--r--arch/sh/mm/tlbflush_64.c295
25 files changed, 891 insertions, 1144 deletions
diff --git a/arch/sh/mm/Kconfig b/arch/sh/mm/Kconfig
index cb8f9920f4d..dba285e8680 100644
--- a/arch/sh/mm/Kconfig
+++ b/arch/sh/mm/Kconfig
@@ -83,7 +83,7 @@ config 32BIT
config PMB
bool "Support 32-bit physical addressing through PMB"
- depends on MMU && EXPERIMENTAL && CPU_SH4A && !CPU_SH4AL_DSP
+ depends on MMU && CPU_SH4A && !CPU_SH4AL_DSP
select 32BIT
select UNCACHED_MAPPING
help
@@ -110,7 +110,8 @@ config VSYSCALL
config NUMA
bool "Non Uniform Memory Access (NUMA) Support"
- depends on MMU && SYS_SUPPORTS_NUMA && EXPERIMENTAL
+ depends on MMU && SYS_SUPPORTS_NUMA
+ select ARCH_WANT_NUMA_VARIABLE_LOCALITY
default n
help
Some SH systems have many various memories scattered around
@@ -136,13 +137,6 @@ config ARCH_SPARSEMEM_ENABLE
config ARCH_SPARSEMEM_DEFAULT
def_bool y
-config MAX_ACTIVE_REGIONS
- int
- default "6" if (CPU_SUBTYPE_SHX3 && SPARSEMEM)
- default "2" if SPARSEMEM && (CPU_SUBTYPE_SH7722 || \
- CPU_SUBTYPE_SH7785)
- default "1"
-
config ARCH_SELECT_MEMORY_MODEL
def_bool y
diff --git a/arch/sh/mm/Makefile b/arch/sh/mm/Makefile
index 2228c8cee4d..cee6b9999d8 100644
--- a/arch/sh/mm/Makefile
+++ b/arch/sh/mm/Makefile
@@ -15,8 +15,8 @@ cacheops-$(CONFIG_CPU_SHX3) += cache-shx3.o
obj-y += $(cacheops-y)
mmu-y := nommu.o extable_32.o
-mmu-$(CONFIG_MMU) := extable_$(BITS).o fault_$(BITS).o gup.o \
- ioremap.o kmap.o pgtable.o tlbflush_$(BITS).o
+mmu-$(CONFIG_MMU) := extable_$(BITS).o fault.o gup.o ioremap.o kmap.o \
+ pgtable.o tlbex_$(BITS).o tlbflush_$(BITS).o
obj-y += $(mmu-y)
@@ -44,7 +44,7 @@ obj-$(CONFIG_HAVE_SRAM_POOL) += sram.o
GCOV_PROFILE_pmb.o := n
-# Special flags for fault_64.o. This puts restrictions on the number of
+# Special flags for tlbex_64.o. This puts restrictions on the number of
# caller-save registers that the compiler can target when building this file.
# This is required because the code is called from a context in entry.S where
# very few registers have been saved in the exception handler (for speed
@@ -59,7 +59,7 @@ GCOV_PROFILE_pmb.o := n
# The resources not listed below are callee save, i.e. the compiler is free to
# use any of them and will spill them to the stack itself.
-CFLAGS_fault_64.o += -ffixed-r7 \
+CFLAGS_tlbex_64.o += -ffixed-r7 \
-ffixed-r8 -ffixed-r9 -ffixed-r10 -ffixed-r11 -ffixed-r12 \
-ffixed-r13 -ffixed-r14 -ffixed-r16 -ffixed-r17 -ffixed-r19 \
-ffixed-r20 -ffixed-r21 -ffixed-r22 -ffixed-r23 \
diff --git a/arch/sh/mm/alignment.c b/arch/sh/mm/alignment.c
index 620fa7ff9ee..ec2b2530242 100644
--- a/arch/sh/mm/alignment.c
+++ b/arch/sh/mm/alignment.c
@@ -140,7 +140,7 @@ static int alignment_proc_open(struct inode *inode, struct file *file)
static ssize_t alignment_proc_write(struct file *file,
const char __user *buffer, size_t count, loff_t *pos)
{
- int *data = PDE(file->f_path.dentry->d_inode)->data;
+ int *data = PDE_DATA(file_inode(file));
char mode;
if (count > 0) {
diff --git a/arch/sh/mm/cache-debugfs.c b/arch/sh/mm/cache-debugfs.c
index 11572519803..777e50f33c0 100644
--- a/arch/sh/mm/cache-debugfs.c
+++ b/arch/sh/mm/cache-debugfs.c
@@ -36,7 +36,7 @@ static int cache_seq_show(struct seq_file *file, void *iter)
*/
jump_to_uncached();
- ccr = __raw_readl(CCR);
+ ccr = __raw_readl(SH_CCR);
if ((ccr & CCR_CACHE_ENABLE) == 0) {
back_to_cached();
diff --git a/arch/sh/mm/cache-sh2.c b/arch/sh/mm/cache-sh2.c
index defcf719f2e..a74259f2f98 100644
--- a/arch/sh/mm/cache-sh2.c
+++ b/arch/sh/mm/cache-sh2.c
@@ -63,9 +63,9 @@ static void sh2__flush_invalidate_region(void *start, int size)
local_irq_save(flags);
jump_to_uncached();
- ccr = __raw_readl(CCR);
+ ccr = __raw_readl(SH_CCR);
ccr |= CCR_CACHE_INVALIDATE;
- __raw_writel(ccr, CCR);
+ __raw_writel(ccr, SH_CCR);
back_to_cached();
local_irq_restore(flags);
diff --git a/arch/sh/mm/cache-sh2a.c b/arch/sh/mm/cache-sh2a.c
index 949e2d3138a..ee87d081259 100644
--- a/arch/sh/mm/cache-sh2a.c
+++ b/arch/sh/mm/cache-sh2a.c
@@ -134,7 +134,8 @@ static void sh2a__flush_invalidate_region(void *start, int size)
/* If there are too many pages then just blow the cache */
if (((end - begin) >> PAGE_SHIFT) >= MAX_OCACHE_PAGES) {
- __raw_writel(__raw_readl(CCR) | CCR_OCACHE_INVALIDATE, CCR);
+ __raw_writel(__raw_readl(SH_CCR) | CCR_OCACHE_INVALIDATE,
+ SH_CCR);
} else {
for (v = begin; v < end; v += L1_CACHE_BYTES)
sh2a_invalidate_line(CACHE_OC_ADDRESS_ARRAY, v);
@@ -167,7 +168,8 @@ static void sh2a_flush_icache_range(void *args)
/* I-Cache invalidate */
/* If there are too many pages then just blow the cache */
if (((end - start) >> PAGE_SHIFT) >= MAX_ICACHE_PAGES) {
- __raw_writel(__raw_readl(CCR) | CCR_ICACHE_INVALIDATE, CCR);
+ __raw_writel(__raw_readl(SH_CCR) | CCR_ICACHE_INVALIDATE,
+ SH_CCR);
} else {
for (v = start; v < end; v += L1_CACHE_BYTES)
sh2a_invalidate_line(CACHE_IC_ADDRESS_ARRAY, v);
diff --git a/arch/sh/mm/cache-sh4.c b/arch/sh/mm/cache-sh4.c
index 92eb98633ab..51d8f7f31d1 100644
--- a/arch/sh/mm/cache-sh4.c
+++ b/arch/sh/mm/cache-sh4.c
@@ -18,6 +18,7 @@
#include <linux/highmem.h>
#include <asm/pgtable.h>
#include <asm/mmu_context.h>
+#include <asm/cache_insns.h>
#include <asm/cacheflush.h>
/*
@@ -132,9 +133,9 @@ static void flush_icache_all(void)
jump_to_uncached();
/* Flush I-cache */
- ccr = __raw_readl(CCR);
+ ccr = __raw_readl(SH_CCR);
ccr |= CCR_CACHE_ICI;
- __raw_writel(ccr, CCR);
+ __raw_writel(ccr, SH_CCR);
/*
* back_to_cached() will take care of the barrier for us, don't add
@@ -244,7 +245,7 @@ static void sh4_flush_cache_page(void *args)
if (map_coherent)
vaddr = kmap_coherent(page, address);
else
- vaddr = kmap_atomic(page, KM_USER0);
+ vaddr = kmap_atomic(page);
address = (unsigned long)vaddr;
}
@@ -259,7 +260,7 @@ static void sh4_flush_cache_page(void *args)
if (map_coherent)
kunmap_coherent(vaddr);
else
- kunmap_atomic(vaddr, KM_USER0);
+ kunmap_atomic(vaddr);
}
}
diff --git a/arch/sh/mm/cache-shx3.c b/arch/sh/mm/cache-shx3.c
index c0adbee97b5..24c58b7dc02 100644
--- a/arch/sh/mm/cache-shx3.c
+++ b/arch/sh/mm/cache-shx3.c
@@ -19,7 +19,7 @@ void __init shx3_cache_init(void)
{
unsigned int ccr;
- ccr = __raw_readl(CCR);
+ ccr = __raw_readl(SH_CCR);
/*
* If we've got cache aliases, resolve them in hardware.
@@ -40,5 +40,5 @@ void __init shx3_cache_init(void)
ccr |= CCR_CACHE_IBE;
#endif
- writel_uncached(ccr, CCR);
+ writel_uncached(ccr, SH_CCR);
}
diff --git a/arch/sh/mm/cache.c b/arch/sh/mm/cache.c
index 5a580ea0442..097c2cdd117 100644
--- a/arch/sh/mm/cache.c
+++ b/arch/sh/mm/cache.c
@@ -95,7 +95,7 @@ void copy_user_highpage(struct page *to, struct page *from,
{
void *vfrom, *vto;
- vto = kmap_atomic(to, KM_USER1);
+ vto = kmap_atomic(to);
if (boot_cpu_data.dcache.n_aliases && page_mapped(from) &&
test_bit(PG_dcache_clean, &from->flags)) {
@@ -103,16 +103,16 @@ void copy_user_highpage(struct page *to, struct page *from,
copy_page(vto, vfrom);
kunmap_coherent(vfrom);
} else {
- vfrom = kmap_atomic(from, KM_USER0);
+ vfrom = kmap_atomic(from);
copy_page(vto, vfrom);
- kunmap_atomic(vfrom, KM_USER0);
+ kunmap_atomic(vfrom);
}
if (pages_do_alias((unsigned long)vto, vaddr & PAGE_MASK) ||
(vma->vm_flags & VM_EXEC))
__flush_purge_region(vto, PAGE_SIZE);
- kunmap_atomic(vto, KM_USER1);
+ kunmap_atomic(vto);
/* Make sure this page is cleared on other CPU's too before using it */
smp_wmb();
}
@@ -120,14 +120,14 @@ EXPORT_SYMBOL(copy_user_highpage);
void clear_user_highpage(struct page *page, unsigned long vaddr)
{
- void *kaddr = kmap_atomic(page, KM_USER0);
+ void *kaddr = kmap_atomic(page);
clear_page(kaddr);
if (pages_do_alias((unsigned long)kaddr, vaddr & PAGE_MASK))
__flush_purge_region(kaddr, PAGE_SIZE);
- kunmap_atomic(kaddr, KM_USER0);
+ kunmap_atomic(kaddr);
}
EXPORT_SYMBOL(clear_user_highpage);
@@ -285,8 +285,8 @@ void __init cpu_cache_init(void)
{
unsigned int cache_disabled = 0;
-#ifdef CCR
- cache_disabled = !(__raw_readl(CCR) & CCR_CACHE_ENABLE);
+#ifdef SH_CCR
+ cache_disabled = !(__raw_readl(SH_CCR) & CCR_CACHE_ENABLE);
#endif
compute_alias(&boot_cpu_data.icache);
diff --git a/arch/sh/mm/consistent.c b/arch/sh/mm/consistent.c
index f251b5f2765..b81d9dbf9fe 100644
--- a/arch/sh/mm/consistent.c
+++ b/arch/sh/mm/consistent.c
@@ -33,7 +33,8 @@ static int __init dma_init(void)
fs_initcall(dma_init);
void *dma_generic_alloc_coherent(struct device *dev, size_t size,
- dma_addr_t *dma_handle, gfp_t gfp)
+ dma_addr_t *dma_handle, gfp_t gfp,
+ struct dma_attrs *attrs)
{
void *ret, *ret_nocache;
int order = get_order(size);
@@ -64,7 +65,8 @@ void *dma_generic_alloc_coherent(struct device *dev, size_t size,
}
void dma_generic_free_coherent(struct device *dev, size_t size,
- void *vaddr, dma_addr_t dma_handle)
+ void *vaddr, dma_addr_t dma_handle,
+ struct dma_attrs *attrs)
{
int order = get_order(size);
unsigned long pfn = dma_handle >> PAGE_SHIFT;
diff --git a/arch/sh/mm/fault.c b/arch/sh/mm/fault.c
new file mode 100644
index 00000000000..541dc610150
--- /dev/null
+++ b/arch/sh/mm/fault.c
@@ -0,0 +1,517 @@
+/*
+ * Page fault handler for SH with an MMU.
+ *
+ * Copyright (C) 1999 Niibe Yutaka
+ * Copyright (C) 2003 - 2012 Paul Mundt
+ *
+ * Based on linux/arch/i386/mm/fault.c:
+ * Copyright (C) 1995 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/mm.h>
+#include <linux/hardirq.h>
+#include <linux/kprobes.h>
+#include <linux/perf_event.h>
+#include <linux/kdebug.h>
+#include <asm/io_trapped.h>
+#include <asm/mmu_context.h>
+#include <asm/tlbflush.h>
+#include <asm/traps.h>
+
+static inline int notify_page_fault(struct pt_regs *regs, int trap)
+{
+ int ret = 0;
+
+ if (kprobes_built_in() && !user_mode(regs)) {
+ preempt_disable();
+ if (kprobe_running() && kprobe_fault_handler(regs, trap))
+ ret = 1;
+ preempt_enable();
+ }
+
+ return ret;
+}
+
+static void
+force_sig_info_fault(int si_signo, int si_code, unsigned long address,
+ struct task_struct *tsk)
+{
+ siginfo_t info;
+
+ info.si_signo = si_signo;
+ info.si_errno = 0;
+ info.si_code = si_code;
+ info.si_addr = (void __user *)address;
+
+ force_sig_info(si_signo, &info, tsk);
+}
+
+/*
+ * This is useful to dump out the page tables associated with
+ * 'addr' in mm 'mm'.
+ */
+static void show_pte(struct mm_struct *mm, unsigned long addr)
+{
+ pgd_t *pgd;
+
+ if (mm) {
+ pgd = mm->pgd;
+ } else {
+ pgd = get_TTB();
+
+ if (unlikely(!pgd))
+ pgd = swapper_pg_dir;
+ }
+
+ printk(KERN_ALERT "pgd = %p\n", pgd);
+ pgd += pgd_index(addr);
+ printk(KERN_ALERT "[%08lx] *pgd=%0*Lx", addr,
+ (u32)(sizeof(*pgd) * 2), (u64)pgd_val(*pgd));
+
+ do {
+ pud_t *pud;
+ pmd_t *pmd;
+ pte_t *pte;
+
+ if (pgd_none(*pgd))
+ break;
+
+ if (pgd_bad(*pgd)) {
+ printk("(bad)");
+ break;
+ }
+
+ pud = pud_offset(pgd, addr);
+ if (PTRS_PER_PUD != 1)
+ printk(", *pud=%0*Lx", (u32)(sizeof(*pud) * 2),
+ (u64)pud_val(*pud));
+
+ if (pud_none(*pud))
+ break;
+
+ if (pud_bad(*pud)) {
+ printk("(bad)");
+ break;
+ }
+
+ pmd = pmd_offset(pud, addr);
+ if (PTRS_PER_PMD != 1)
+ printk(", *pmd=%0*Lx", (u32)(sizeof(*pmd) * 2),
+ (u64)pmd_val(*pmd));
+
+ if (pmd_none(*pmd))
+ break;
+
+ if (pmd_bad(*pmd)) {
+ printk("(bad)");
+ break;
+ }
+
+ /* We must not map this if we have highmem enabled */
+ if (PageHighMem(pfn_to_page(pmd_val(*pmd) >> PAGE_SHIFT)))
+ break;
+
+ pte = pte_offset_kernel(pmd, addr);
+ printk(", *pte=%0*Lx", (u32)(sizeof(*pte) * 2),
+ (u64)pte_val(*pte));
+ } while (0);
+
+ printk("\n");
+}
+
+static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
+{
+ unsigned index = pgd_index(address);
+ pgd_t *pgd_k;
+ pud_t *pud, *pud_k;
+ pmd_t *pmd, *pmd_k;
+
+ pgd += index;
+ pgd_k = init_mm.pgd + index;
+
+ if (!pgd_present(*pgd_k))
+ return NULL;
+
+ pud = pud_offset(pgd, address);
+ pud_k = pud_offset(pgd_k, address);
+ if (!pud_present(*pud_k))
+ return NULL;
+
+ if (!pud_present(*pud))
+ set_pud(pud, *pud_k);
+
+ pmd = pmd_offset(pud, address);
+ pmd_k = pmd_offset(pud_k, address);
+ if (!pmd_present(*pmd_k))
+ return NULL;
+
+ if (!pmd_present(*pmd))
+ set_pmd(pmd, *pmd_k);
+ else {
+ /*
+ * The page tables are fully synchronised so there must
+ * be another reason for the fault. Return NULL here to
+ * signal that we have not taken care of the fault.
+ */
+ BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
+ return NULL;
+ }
+
+ return pmd_k;
+}
+
+#ifdef CONFIG_SH_STORE_QUEUES
+#define __FAULT_ADDR_LIMIT P3_ADDR_MAX
+#else
+#define __FAULT_ADDR_LIMIT VMALLOC_END
+#endif
+
+/*
+ * Handle a fault on the vmalloc or module mapping area
+ */
+static noinline int vmalloc_fault(unsigned long address)
+{
+ pgd_t *pgd_k;
+ pmd_t *pmd_k;
+ pte_t *pte_k;
+
+ /* Make sure we are in vmalloc/module/P3 area: */
+ if (!(address >= VMALLOC_START && address < __FAULT_ADDR_LIMIT))
+ return -1;
+
+ /*
+ * Synchronize this task's top level page-table
+ * with the 'reference' page table.
+ *
+ * Do _not_ use "current" here. We might be inside
+ * an interrupt in the middle of a task switch..
+ */
+ pgd_k = get_TTB();
+ pmd_k = vmalloc_sync_one(pgd_k, address);
+ if (!pmd_k)
+ return -1;
+
+ pte_k = pte_offset_kernel(pmd_k, address);
+ if (!pte_present(*pte_k))
+ return -1;
+
+ return 0;
+}
+
+static void
+show_fault_oops(struct pt_regs *regs, unsigned long address)
+{
+ if (!oops_may_print())
+ return;
+
+ printk(KERN_ALERT "BUG: unable to handle kernel ");
+ if (address < PAGE_SIZE)
+ printk(KERN_CONT "NULL pointer dereference");
+ else
+ printk(KERN_CONT "paging request");
+
+ printk(KERN_CONT " at %08lx\n", address);
+ printk(KERN_ALERT "PC:");
+ printk_address(regs->pc, 1);
+
+ show_pte(NULL, address);
+}
+
+static noinline void
+no_context(struct pt_regs *regs, unsigned long error_code,
+ unsigned long address)
+{
+ /* Are we prepared to handle this kernel fault? */
+ if (fixup_exception(regs))
+ return;
+
+ if (handle_trapped_io(regs, address))
+ return;
+
+ /*
+ * Oops. The kernel tried to access some bad page. We'll have to
+ * terminate things with extreme prejudice.
+ */
+ bust_spinlocks(1);
+
+ show_fault_oops(regs, address);
+
+ die("Oops", regs, error_code);
+ bust_spinlocks(0);
+ do_exit(SIGKILL);
+}
+
+static void
+__bad_area_nosemaphore(struct pt_regs *regs, unsigned long error_code,
+ unsigned long address, int si_code)
+{
+ struct task_struct *tsk = current;
+
+ /* User mode accesses just cause a SIGSEGV */
+ if (user_mode(regs)) {
+ /*
+ * It's possible to have interrupts off here:
+ */
+ local_irq_enable();
+
+ force_sig_info_fault(SIGSEGV, si_code, address, tsk);
+
+ return;
+ }
+
+ no_context(regs, error_code, address);
+}
+
+static noinline void
+bad_area_nosemaphore(struct pt_regs *regs, unsigned long error_code,
+ unsigned long address)
+{
+ __bad_area_nosemaphore(regs, error_code, address, SEGV_MAPERR);
+}
+
+static void
+__bad_area(struct pt_regs *regs, unsigned long error_code,
+ unsigned long address, int si_code)
+{
+ struct mm_struct *mm = current->mm;
+
+ /*
+ * Something tried to access memory that isn't in our memory map..
+ * Fix it, but check if it's kernel or user first..
+ */
+ up_read(&mm->mmap_sem);
+
+ __bad_area_nosemaphore(regs, error_code, address, si_code);
+}
+
+static noinline void
+bad_area(struct pt_regs *regs, unsigned long error_code, unsigned long address)
+{
+ __bad_area(regs, error_code, address, SEGV_MAPERR);
+}
+
+static noinline void
+bad_area_access_error(struct pt_regs *regs, unsigned long error_code,
+ unsigned long address)
+{
+ __bad_area(regs, error_code, address, SEGV_ACCERR);
+}
+
+static void
+do_sigbus(struct pt_regs *regs, unsigned long error_code, unsigned long address)
+{
+ struct task_struct *tsk = current;
+ struct mm_struct *mm = tsk->mm;
+
+ up_read(&mm->mmap_sem);
+
+ /* Kernel mode? Handle exceptions or die: */
+ if (!user_mode(regs))
+ no_context(regs, error_code, address);
+
+ force_sig_info_fault(SIGBUS, BUS_ADRERR, address, tsk);
+}
+
+static noinline int
+mm_fault_error(struct pt_regs *regs, unsigned long error_code,
+ unsigned long address, unsigned int fault)
+{
+ /*
+ * Pagefault was interrupted by SIGKILL. We have no reason to
+ * continue pagefault.
+ */
+ if (fatal_signal_pending(current)) {
+ if (!(fault & VM_FAULT_RETRY))
+ up_read(&current->mm->mmap_sem);
+ if (!user_mode(regs))
+ no_context(regs, error_code, address);
+ return 1;
+ }
+
+ if (!(fault & VM_FAULT_ERROR))
+ return 0;
+
+ if (fault & VM_FAULT_OOM) {
+ /* Kernel mode? Handle exceptions or die: */
+ if (!user_mode(regs)) {
+ up_read(&current->mm->mmap_sem);
+ no_context(regs, error_code, address);
+ return 1;
+ }
+ up_read(&current->mm->mmap_sem);
+
+ /*
+ * We ran out of memory, call the OOM killer, and return the
+ * userspace (which will retry the fault, or kill us if we got
+ * oom-killed):
+ */
+ pagefault_out_of_memory();
+ } else {
+ if (fault & VM_FAULT_SIGBUS)
+ do_sigbus(regs, error_code, address);
+ else
+ BUG();
+ }
+
+ return 1;
+}
+
+static inline int access_error(int error_code, struct vm_area_struct *vma)
+{
+ if (error_code & FAULT_CODE_WRITE) {
+ /* write, present and write, not present: */
+ if (unlikely(!(vma->vm_flags & VM_WRITE)))
+ return 1;
+ return 0;
+ }
+
+ /* ITLB miss on NX page */
+ if (unlikely((error_code & FAULT_CODE_ITLB) &&
+ !(vma->vm_flags & VM_EXEC)))
+ return 1;
+
+ /* read, not present: */
+ if (unlikely(!(vma->vm_flags & (VM_READ | VM_EXEC | VM_WRITE))))
+ return 1;
+
+ return 0;
+}
+
+static int fault_in_kernel_space(unsigned long address)
+{
+ return address >= TASK_SIZE;
+}
+
+/*
+ * This routine handles page faults. It determines the address,
+ * and the problem, and then passes it off to one of the appropriate
+ * routines.
+ */
+asmlinkage void __kprobes do_page_fault(struct pt_regs *regs,
+ unsigned long error_code,
+ unsigned long address)
+{
+ unsigned long vec;
+ struct task_struct *tsk;
+ struct mm_struct *mm;
+ struct vm_area_struct * vma;
+ int fault;
+ unsigned int flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;
+
+ tsk = current;
+ mm = tsk->mm;
+ vec = lookup_exception_vector();
+
+ /*
+ * We fault-in kernel-space virtual memory on-demand. The
+ * 'reference' page table is init_mm.pgd.
+ *
+ * NOTE! We MUST NOT take any locks for this case. We may
+ * be in an interrupt or a critical region, and should
+ * only copy the information from the master page table,
+ * nothing more.
+ */
+ if (unlikely(fault_in_kernel_space(address))) {
+ if (vmalloc_fault(address) >= 0)
+ return;
+ if (notify_page_fault(regs, vec))
+ return;
+
+ bad_area_nosemaphore(regs, error_code, address);
+ return;
+ }
+
+ if (unlikely(notify_page_fault(regs, vec)))
+ return;
+
+ /* Only enable interrupts if they were on before the fault */
+ if ((regs->sr & SR_IMASK) != SR_IMASK)
+ local_irq_enable();
+
+ perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
+
+ /*
+ * If we're in an interrupt, have no user context or are running
+ * in an atomic region then we must not take the fault:
+ */
+ if (unlikely(in_atomic() || !mm)) {
+ bad_area_nosemaphore(regs, error_code, address);
+ return;
+ }
+
+retry:
+ down_read(&mm->mmap_sem);
+
+ vma = find_vma(mm, address);
+ if (unlikely(!vma)) {
+ bad_area(regs, error_code, address);
+ return;
+ }
+ if (likely(vma->vm_start <= address))
+ goto good_area;
+ if (unlikely(!(vma->vm_flags & VM_GROWSDOWN))) {
+ bad_area(regs, error_code, address);
+ return;
+ }
+ if (unlikely(expand_stack(vma, address))) {
+ bad_area(regs, error_code, address);
+ return;
+ }
+
+ /*
+ * Ok, we have a good vm_area for this memory access, so
+ * we can handle it..
+ */
+good_area:
+ if (unlikely(access_error(error_code, vma))) {
+ bad_area_access_error(regs, error_code, address);
+ return;
+ }
+
+ set_thread_fault_code(error_code);
+
+ if (user_mode(regs))
+ flags |= FAULT_FLAG_USER;
+ if (error_code & FAULT_CODE_WRITE)
+ flags |= FAULT_FLAG_WRITE;
+
+ /*
+ * 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(mm, vma, address, flags);
+
+ if (unlikely(fault & (VM_FAULT_RETRY | VM_FAULT_ERROR)))
+ if (mm_fault_error(regs, error_code, address, fault))
+ return;
+
+ if (flags & FAULT_FLAG_ALLOW_RETRY) {
+ if (fault & VM_FAULT_MAJOR) {
+ tsk->maj_flt++;
+ perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1,
+ regs, address);
+ } else {
+ tsk->min_flt++;
+ perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1,
+ regs, address);
+ }
+ 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);
+}
diff --git a/arch/sh/mm/fault_32.c b/arch/sh/mm/fault_32.c
deleted file mode 100644
index 7bebd044f2a..00000000000
--- a/arch/sh/mm/fault_32.c
+++ /dev/null
@@ -1,374 +0,0 @@
-/*
- * Page fault handler for SH with an MMU.
- *
- * Copyright (C) 1999 Niibe Yutaka
- * Copyright (C) 2003 - 2009 Paul Mundt
- *
- * Based on linux/arch/i386/mm/fault.c:
- * Copyright (C) 1995 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/mm.h>
-#include <linux/hardirq.h>
-#include <linux/kprobes.h>
-#include <linux/perf_event.h>
-#include <asm/io_trapped.h>
-#include <asm/system.h>
-#include <asm/mmu_context.h>
-#include <asm/tlbflush.h>
-
-static inline int notify_page_fault(struct pt_regs *regs, int trap)
-{
- int ret = 0;
-
- if (kprobes_built_in() && !user_mode(regs)) {
- preempt_disable();
- if (kprobe_running() && kprobe_fault_handler(regs, trap))
- ret = 1;
- preempt_enable();
- }
-
- return ret;
-}
-
-static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
-{
- unsigned index = pgd_index(address);
- pgd_t *pgd_k;
- pud_t *pud, *pud_k;
- pmd_t *pmd, *pmd_k;
-
- pgd += index;
- pgd_k = init_mm.pgd + index;
-
- if (!pgd_present(*pgd_k))
- return NULL;
-
- pud = pud_offset(pgd, address);
- pud_k = pud_offset(pgd_k, address);
- if (!pud_present(*pud_k))
- return NULL;
-
- if (!pud_present(*pud))
- set_pud(pud, *pud_k);
-
- pmd = pmd_offset(pud, address);
- pmd_k = pmd_offset(pud_k, address);
- if (!pmd_present(*pmd_k))
- return NULL;
-
- if (!pmd_present(*pmd))
- set_pmd(pmd, *pmd_k);
- else {
- /*
- * The page tables are fully synchronised so there must
- * be another reason for the fault. Return NULL here to
- * signal that we have not taken care of the fault.
- */
- BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
- return NULL;
- }
-
- return pmd_k;
-}
-
-/*
- * Handle a fault on the vmalloc or module mapping area
- */
-static noinline int vmalloc_fault(unsigned long address)
-{
- pgd_t *pgd_k;
- pmd_t *pmd_k;
- pte_t *pte_k;
-
- /* Make sure we are in vmalloc/module/P3 area: */
- if (!(address >= VMALLOC_START && address < P3_ADDR_MAX))
- return -1;
-
- /*
- * Synchronize this task's top level page-table
- * with the 'reference' page table.
- *
- * Do _not_ use "current" here. We might be inside
- * an interrupt in the middle of a task switch..
- */
- pgd_k = get_TTB();
- pmd_k = vmalloc_sync_one(pgd_k, address);
- if (!pmd_k)
- return -1;
-
- pte_k = pte_offset_kernel(pmd_k, address);
- if (!pte_present(*pte_k))
- return -1;
-
- return 0;
-}
-
-static int fault_in_kernel_space(unsigned long address)
-{
- return address >= TASK_SIZE;
-}
-
-/*
- * This routine handles page faults. It determines the address,
- * and the problem, and then passes it off to one of the appropriate
- * routines.
- */
-asmlinkage void __kprobes do_page_fault(struct pt_regs *regs,
- unsigned long writeaccess,
- unsigned long address)
-{
- unsigned long vec;
- struct task_struct *tsk;
- struct mm_struct *mm;
- struct vm_area_struct * vma;
- int si_code;
- int fault;
- siginfo_t info;
-
- tsk = current;
- mm = tsk->mm;
- si_code = SEGV_MAPERR;
- vec = lookup_exception_vector();
-
- /*
- * We fault-in kernel-space virtual memory on-demand. The
- * 'reference' page table is init_mm.pgd.
- *
- * NOTE! We MUST NOT take any locks for this case. We may
- * be in an interrupt or a critical region, and should
- * only copy the information from the master page table,
- * nothing more.
- */
- if (unlikely(fault_in_kernel_space(address))) {
- if (vmalloc_fault(address) >= 0)
- return;
- if (notify_page_fault(regs, vec))
- return;
-
- goto bad_area_nosemaphore;
- }
-
- if (unlikely(notify_page_fault(regs, vec)))
- return;
-
- /* Only enable interrupts if they were on before the fault */
- if ((regs->sr & SR_IMASK) != SR_IMASK)
- local_irq_enable();
-
- perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
-
- /*
- * If we're in an interrupt, have no user context or are running
- * in an atomic region then we must not take the fault:
- */
- if (in_atomic() || !mm)
- goto no_context;
-
- 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 (expand_stack(vma, address))
- goto bad_area;
-
- /*
- * Ok, we have a good vm_area for this memory access, so
- * we can handle it..
- */
-good_area:
- si_code = SEGV_ACCERR;
- if (writeaccess) {
- if (!(vma->vm_flags & VM_WRITE))
- goto bad_area;
- } else {
- if (!(vma->vm_flags & (VM_READ | VM_EXEC | VM_WRITE)))
- 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(mm, vma, address, writeaccess ? FAULT_FLAG_WRITE : 0);
- 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) {
- tsk->maj_flt++;
- perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1,
- regs, address);
- } else {
- tsk->min_flt++;
- perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1,
- regs, address);
- }
-
- up_read(&mm->mmap_sem);
- return;
-
- /*
- * Something tried to access memory that isn't in our memory map..
- * Fix it, but check if it's kernel or user first..
- */
-bad_area:
- up_read(&mm->mmap_sem);
-
-bad_area_nosemaphore:
- if (user_mode(regs)) {
- info.si_signo = SIGSEGV;
- info.si_errno = 0;
- info.si_code = si_code;
- info.si_addr = (void *) address;
- force_sig_info(SIGSEGV, &info, tsk);
- return;
- }
-
-no_context:
- /* Are we prepared to handle this kernel fault? */
- if (fixup_exception(regs))
- return;
-
- if (handle_trapped_io(regs, address))
- return;
-/*
- * Oops. The kernel tried to access some bad page. We'll have to
- * terminate things with extreme prejudice.
- *
- */
-
- bust_spinlocks(1);
-
- if (oops_may_print()) {
- unsigned long page;
-
- if (address < PAGE_SIZE)
- printk(KERN_ALERT "Unable to handle kernel NULL "
- "pointer dereference");
- else
- printk(KERN_ALERT "Unable to handle kernel paging "
- "request");
- printk(" at virtual address %08lx\n", address);
- printk(KERN_ALERT "pc = %08lx\n", regs->pc);
- page = (unsigned long)get_TTB();
- if (page) {
- page = ((__typeof__(page) *)page)[address >> PGDIR_SHIFT];
- printk(KERN_ALERT "*pde = %08lx\n", page);
- if (page & _PAGE_PRESENT) {
- page &= PAGE_MASK;
- address &= 0x003ff000;
- page = ((__typeof__(page) *)
- __va(page))[address >>
- PAGE_SHIFT];
- printk(KERN_ALERT "*pte = %08lx\n", page);
- }
- }
- }
-
- die("Oops", regs, writeaccess);
- bust_spinlocks(0);
- do_exit(SIGKILL);
-
-/*
- * 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))
- goto no_context;
- pagefault_out_of_memory();
- return;
-
-do_sigbus:
- up_read(&mm->mmap_sem);
-
- /*
- * Send a sigbus, regardless of whether we were in kernel
- * or user mode.
- */
- info.si_signo = SIGBUS;
- info.si_errno = 0;
- info.si_code = BUS_ADRERR;
- info.si_addr = (void *)address;
- force_sig_info(SIGBUS, &info, tsk);
-
- /* Kernel mode? Handle exceptions or die */
- if (!user_mode(regs))
- goto no_context;
-}
-
-/*
- * Called with interrupts disabled.
- */
-asmlinkage int __kprobes
-handle_tlbmiss(struct pt_regs *regs, unsigned long writeaccess,
- unsigned long address)
-{
- pgd_t *pgd;
- pud_t *pud;
- pmd_t *pmd;
- pte_t *pte;
- pte_t entry;
-
- /*
- * We don't take page faults for P1, P2, and parts of P4, these
- * are always mapped, whether it be due to legacy behaviour in
- * 29-bit mode, or due to PMB configuration in 32-bit mode.
- */
- if (address >= P3SEG && address < P3_ADDR_MAX) {
- pgd = pgd_offset_k(address);
- } else {
- if (unlikely(address >= TASK_SIZE || !current->mm))
- return 1;
-
- pgd = pgd_offset(current->mm, address);
- }
-
- pud = pud_offset(pgd, address);
- if (pud_none_or_clear_bad(pud))
- return 1;
- pmd = pmd_offset(pud, address);
- if (pmd_none_or_clear_bad(pmd))
- return 1;
- pte = pte_offset_kernel(pmd, address);
- entry = *pte;
- if (unlikely(pte_none(entry) || pte_not_present(entry)))
- return 1;
- if (unlikely(writeaccess && !pte_write(entry)))
- return 1;
-
- if (writeaccess)
- entry = pte_mkdirty(entry);
- entry = pte_mkyoung(entry);
-
- set_pte(pte, entry);
-
-#if defined(CONFIG_CPU_SH4) && !defined(CONFIG_SMP)
- /*
- * SH-4 does not set MMUCR.RC to the corresponding TLB entry in
- * the case of an initial page write exception, so we need to
- * flush it in order to avoid potential TLB entry duplication.
- */
- if (writeaccess == 2)
- local_flush_tlb_one(get_asid(), address & PAGE_MASK);
-#endif
-
- update_mmu_cache(NULL, address, pte);
-
- return 0;
-}
diff --git a/arch/sh/mm/fault_64.c b/arch/sh/mm/fault_64.c
deleted file mode 100644
index 2b356cec248..00000000000
--- a/arch/sh/mm/fault_64.c
+++ /dev/null
@@ -1,266 +0,0 @@
-/*
- * The SH64 TLB miss.
- *
- * Original code from fault.c
- * Copyright (C) 2000, 2001 Paolo Alberelli
- *
- * Fast PTE->TLB refill path
- * Copyright (C) 2003 Richard.Curnow@superh.com
- *
- * IMPORTANT NOTES :
- * The do_fast_page_fault function is called from a context in entry.S
- * where very few registers have been saved. In particular, the code in
- * this file must be compiled not to use ANY caller-save registers that
- * are not part of the restricted save set. Also, it means that code in
- * this file must not make calls to functions elsewhere in the kernel, or
- * else the excepting context will see corruption in its caller-save
- * registers. Plus, the entry.S save area is non-reentrant, so this code
- * has to run with SR.BL==1, i.e. no interrupts taken inside it and panic
- * on any exception.
- *
- * 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/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/smp.h>
-#include <linux/interrupt.h>
-#include <asm/system.h>
-#include <asm/tlb.h>
-#include <asm/io.h>
-#include <asm/uaccess.h>
-#include <asm/pgalloc.h>
-#include <asm/mmu_context.h>
-#include <cpu/registers.h>
-
-/* Callable from fault.c, so not static */
-inline void __do_tlb_refill(unsigned long address,
- unsigned long long is_text_not_data, pte_t *pte)
-{
- unsigned long long ptel;
- unsigned long long pteh=0;
- struct tlb_info *tlbp;
- unsigned long long next;
-
- /* Get PTEL first */
- ptel = pte_val(*pte);
-
- /*
- * Set PTEH register
- */
- pteh = neff_sign_extend(address & MMU_VPN_MASK);
-
- /* Set the ASID. */
- pteh |= get_asid() << PTEH_ASID_SHIFT;
- pteh |= PTEH_VALID;
-
- /* Set PTEL register, set_pte has performed the sign extension */
- ptel &= _PAGE_FLAGS_HARDWARE_MASK; /* drop software flags */
-
- tlbp = is_text_not_data ? &(cpu_data->itlb) : &(cpu_data->dtlb);
- next = tlbp->next;
- __flush_tlb_slot(next);
- asm volatile ("putcfg %0,1,%2\n\n\t"
- "putcfg %0,0,%1\n"
- : : "r" (next), "r" (pteh), "r" (ptel) );
-
- next += TLB_STEP;
- if (next > tlbp->last) next = tlbp->first;
- tlbp->next = next;
-
-}
-
-static int handle_vmalloc_fault(struct mm_struct *mm,
- unsigned long protection_flags,
- unsigned long long textaccess,
- unsigned long address)
-{
- pgd_t *dir;
- pud_t *pud;
- pmd_t *pmd;
- static pte_t *pte;
- pte_t entry;
-
- dir = pgd_offset_k(address);
-
- pud = pud_offset(dir, address);
- if (pud_none_or_clear_bad(pud))
- return 0;
-
- pmd = pmd_offset(pud, address);
- if (pmd_none_or_clear_bad(pmd))
- return 0;
-
- pte = pte_offset_kernel(pmd, address);
- entry = *pte;
-
- if (pte_none(entry) || !pte_present(entry))
- return 0;
- if ((pte_val(entry) & protection_flags) != protection_flags)
- return 0;
-
- __do_tlb_refill(address, textaccess, pte);
-
- return 1;
-}
-
-static int handle_tlbmiss(struct mm_struct *mm,
- unsigned long long protection_flags,
- unsigned long long textaccess,
- unsigned long address)
-{
- pgd_t *dir;
- pud_t *pud;
- pmd_t *pmd;
- pte_t *pte;
- pte_t entry;
-
- /* NB. The PGD currently only contains a single entry - there is no
- page table tree stored for the top half of the address space since
- virtual pages in that region should never be mapped in user mode.
- (In kernel mode, the only things in that region are the 512Mb super
- page (locked in), and vmalloc (modules) + I/O device pages (handled
- by handle_vmalloc_fault), so no PGD for the upper half is required
- by kernel mode either).
-
- See how mm->pgd is allocated and initialised in pgd_alloc to see why
- the next test is necessary. - RPC */
- if (address >= (unsigned long) TASK_SIZE)
- /* upper half - never has page table entries. */
- return 0;
-
- dir = pgd_offset(mm, address);
- if (pgd_none(*dir) || !pgd_present(*dir))
- return 0;
- if (!pgd_present(*dir))
- return 0;
-
- pud = pud_offset(dir, address);
- if (pud_none(*pud) || !pud_present(*pud))
- return 0;
-
- pmd = pmd_offset(pud, address);
- if (pmd_none(*pmd) || !pmd_present(*pmd))
- return 0;
-
- pte = pte_offset_kernel(pmd, address);
- entry = *pte;
-
- if (pte_none(entry) || !pte_present(entry))
- return 0;
-
- /*
- * If the page doesn't have sufficient protection bits set to
- * service the kind of fault being handled, there's not much
- * point doing the TLB refill. Punt the fault to the general
- * handler.
- */
- if ((pte_val(entry) & protection_flags) != protection_flags)
- return 0;
-
- __do_tlb_refill(address, textaccess, pte);
-
- return 1;
-}
-
-/*
- * Put all this information into one structure so that everything is just
- * arithmetic relative to a single base address. This reduces the number
- * of movi/shori pairs needed just to load addresses of static data.
- */
-struct expevt_lookup {
- unsigned short protection_flags[8];
- unsigned char is_text_access[8];
- unsigned char is_write_access[8];
-};
-
-#define PRU (1<<9)
-#define PRW (1<<8)
-#define PRX (1<<7)
-#define PRR (1<<6)
-
-#define DIRTY (_PAGE_DIRTY | _PAGE_ACCESSED)
-#define YOUNG (_PAGE_ACCESSED)
-
-/* Sized as 8 rather than 4 to allow checking the PTE's PRU bit against whether
- the fault happened in user mode or privileged mode. */
-static struct expevt_lookup expevt_lookup_table = {
- .protection_flags = {PRX, PRX, 0, 0, PRR, PRR, PRW, PRW},
- .is_text_access = {1, 1, 0, 0, 0, 0, 0, 0}
-};
-
-/*
- This routine handles page faults that can be serviced just by refilling a
- TLB entry from an existing page table entry. (This case represents a very
- large majority of page faults.) Return 1 if the fault was successfully
- handled. Return 0 if the fault could not be handled. (This leads into the
- general fault handling in fault.c which deals with mapping file-backed
- pages, stack growth, segmentation faults, swapping etc etc)
- */
-asmlinkage int do_fast_page_fault(unsigned long long ssr_md,
- unsigned long long expevt,
- unsigned long address)
-{
- struct task_struct *tsk;
- struct mm_struct *mm;
- unsigned long long textaccess;
- unsigned long long protection_flags;
- unsigned long long index;
- unsigned long long expevt4;
-
- /* The next few lines implement a way of hashing EXPEVT into a
- * small array index which can be used to lookup parameters
- * specific to the type of TLBMISS being handled.
- *
- * Note:
- * ITLBMISS has EXPEVT==0xa40
- * RTLBMISS has EXPEVT==0x040
- * WTLBMISS has EXPEVT==0x060
- */
- expevt4 = (expevt >> 4);
- /* TODO : xor ssr_md into this expression too. Then we can check
- * that PRU is set when it needs to be. */
- index = expevt4 ^ (expevt4 >> 5);
- index &= 7;
- protection_flags = expevt_lookup_table.protection_flags[index];
- textaccess = expevt_lookup_table.is_text_access[index];
-
- /* SIM
- * Note this is now called with interrupts still disabled
- * This is to cope with being called for a missing IO port
- * address with interrupts disabled. This should be fixed as
- * soon as we have a better 'fast path' miss handler.
- *
- * Plus take care how you try and debug this stuff.
- * For example, writing debug data to a port which you
- * have just faulted on is not going to work.
- */
-
- tsk = current;
- mm = tsk->mm;
-
- if ((address >= VMALLOC_START && address < VMALLOC_END) ||
- (address >= IOBASE_VADDR && address < IOBASE_END)) {
- if (ssr_md)
- /*
- * Process-contexts can never have this address
- * range mapped
- */
- if (handle_vmalloc_fault(mm, protection_flags,
- textaccess, address))
- return 1;
- } else if (!in_interrupt() && mm) {
- if (handle_tlbmiss(mm, protection_flags, textaccess, address))
- return 1;
- }
-
- return 0;
-}
diff --git a/arch/sh/mm/flush-sh4.c b/arch/sh/mm/flush-sh4.c
index cef402678f4..0b85dd9dd3a 100644
--- a/arch/sh/mm/flush-sh4.c
+++ b/arch/sh/mm/flush-sh4.c
@@ -1,6 +1,8 @@
#include <linux/mm.h>
#include <asm/mmu_context.h>
+#include <asm/cache_insns.h>
#include <asm/cacheflush.h>
+#include <asm/traps.h>
/*
* Write back the dirty D-caches, but not invalidate them.
diff --git a/arch/sh/mm/init.c b/arch/sh/mm/init.c
index 82cc576fab1..2d089fe2cba 100644
--- a/arch/sh/mm/init.c
+++ b/arch/sh/mm/init.c
@@ -231,7 +231,7 @@ static void __init bootmem_init_one_node(unsigned int nid)
if (!p->node_spanned_pages)
return;
- end_pfn = p->node_start_pfn + p->node_spanned_pages;
+ end_pfn = pgdat_end_pfn(p);
total_pages = bootmem_bootmap_pages(p->node_spanned_pages);
@@ -407,32 +407,16 @@ unsigned int mem_init_done = 0;
void __init mem_init(void)
{
- int codesize, datasize, initsize;
- int nid;
+ pg_data_t *pgdat;
iommu_init();
- num_physpages = 0;
high_memory = NULL;
+ for_each_online_pgdat(pgdat)
+ high_memory = max_t(void *, high_memory,
+ __va(pgdat_end_pfn(pgdat) << PAGE_SHIFT));
- for_each_online_node(nid) {
- pg_data_t *pgdat = NODE_DATA(nid);
- unsigned long node_pages = 0;
- void *node_high_memory;
-
- num_physpages += pgdat->node_present_pages;
-
- if (pgdat->node_spanned_pages)
- node_pages = free_all_bootmem_node(pgdat);
-
- totalram_pages += node_pages;
-
- node_high_memory = (void *)__va((pgdat->node_start_pfn +
- pgdat->node_spanned_pages) <<
- PAGE_SHIFT);
- if (node_high_memory > high_memory)
- high_memory = node_high_memory;
- }
+ free_all_bootmem();
/* Set this up early, so we can take care of the zero page */
cpu_cache_init();
@@ -443,19 +427,8 @@ void __init mem_init(void)
vsyscall_init();
- codesize = (unsigned long) &_etext - (unsigned long) &_text;
- datasize = (unsigned long) &_edata - (unsigned long) &_etext;
- initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
-
- printk(KERN_INFO "Memory: %luk/%luk available (%dk kernel code, "
- "%dk data, %dk init)\n",
- nr_free_pages() << (PAGE_SHIFT-10),
- num_physpages << (PAGE_SHIFT-10),
- codesize >> 10,
- datasize >> 10,
- initsize >> 10);
-
- printk(KERN_INFO "virtual kernel memory layout:\n"
+ mem_init_print_info(NULL);
+ pr_info("virtual kernel memory layout:\n"
" fixmap : 0x%08lx - 0x%08lx (%4ld kB)\n"
#ifdef CONFIG_HIGHMEM
" pkmap : 0x%08lx - 0x%08lx (%4ld kB)\n"
@@ -501,31 +474,13 @@ void __init mem_init(void)
void free_initmem(void)
{
- unsigned long addr;
-
- addr = (unsigned long)(&__init_begin);
- for (; addr < (unsigned long)(&__init_end); addr += PAGE_SIZE) {
- ClearPageReserved(virt_to_page(addr));
- init_page_count(virt_to_page(addr));
- free_page(addr);
- totalram_pages++;
- }
- printk("Freeing unused kernel memory: %ldk freed\n",
- ((unsigned long)&__init_end -
- (unsigned long)&__init_begin) >> 10);
+ free_initmem_default(-1);
}
#ifdef CONFIG_BLK_DEV_INITRD
void free_initrd_mem(unsigned long start, unsigned long end)
{
- unsigned long p;
- for (p = start; p < end; p += PAGE_SIZE) {
- ClearPageReserved(virt_to_page(p));
- init_page_count(virt_to_page(p));
- free_page(p);
- totalram_pages++;
- }
- printk("Freeing initrd memory: %ldk freed\n", (end - start) >> 10);
+ free_reserved_area((void *)start, (void *)end, -1, "initrd");
}
#endif
@@ -558,4 +513,21 @@ int memory_add_physaddr_to_nid(u64 addr)
EXPORT_SYMBOL_GPL(memory_add_physaddr_to_nid);
#endif
+#ifdef CONFIG_MEMORY_HOTREMOVE
+int arch_remove_memory(u64 start, u64 size)
+{
+ unsigned long start_pfn = start >> PAGE_SHIFT;
+ unsigned long nr_pages = size >> PAGE_SHIFT;
+ struct zone *zone;
+ int ret;
+
+ zone = page_zone(pfn_to_page(start_pfn));
+ ret = __remove_pages(zone, start_pfn, nr_pages);
+ if (unlikely(ret))
+ pr_warn("%s: Failed, __remove_pages() == %d\n", __func__,
+ ret);
+
+ return ret;
+}
+#endif
#endif /* CONFIG_MEMORY_HOTPLUG */
diff --git a/arch/sh/mm/mmap.c b/arch/sh/mm/mmap.c
index afeb710ec5c..6777177807c 100644
--- a/arch/sh/mm/mmap.c
+++ b/arch/sh/mm/mmap.c
@@ -30,25 +30,13 @@ static inline unsigned long COLOUR_ALIGN(unsigned long addr,
return base + off;
}
-static inline unsigned long COLOUR_ALIGN_DOWN(unsigned long addr,
- unsigned long pgoff)
-{
- unsigned long base = addr & ~shm_align_mask;
- unsigned long off = (pgoff << PAGE_SHIFT) & shm_align_mask;
-
- if (base + off <= addr)
- return base + off;
-
- return base - off;
-}
-
unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr,
unsigned long len, unsigned long pgoff, unsigned long flags)
{
struct mm_struct *mm = current->mm;
struct vm_area_struct *vma;
- unsigned long start_addr;
int do_colour_align;
+ struct vm_unmapped_area_info info;
if (flags & MAP_FIXED) {
/* We do not accept a shared mapping if it would violate
@@ -79,47 +67,13 @@ unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr,
return addr;
}
- if (len > mm->cached_hole_size) {
- start_addr = addr = mm->free_area_cache;
- } else {
- mm->cached_hole_size = 0;
- start_addr = addr = TASK_UNMAPPED_BASE;
- }
-
-full_search:
- if (do_colour_align)
- addr = COLOUR_ALIGN(addr, pgoff);
- else
- addr = PAGE_ALIGN(mm->free_area_cache);
-
- for (vma = find_vma(mm, addr); ; vma = vma->vm_next) {
- /* At this point: (!vma || addr < vma->vm_end). */
- if (unlikely(TASK_SIZE - len < addr)) {
- /*
- * Start a new search - just in case we missed
- * some holes.
- */
- if (start_addr != TASK_UNMAPPED_BASE) {
- start_addr = addr = TASK_UNMAPPED_BASE;
- mm->cached_hole_size = 0;
- goto full_search;
- }
- return -ENOMEM;
- }
- if (likely(!vma || addr + len <= vma->vm_start)) {
- /*
- * Remember the place where we stopped the search:
- */
- mm->free_area_cache = addr + len;
- return addr;
- }
- if (addr + mm->cached_hole_size < vma->vm_start)
- mm->cached_hole_size = vma->vm_start - addr;
-
- addr = vma->vm_end;
- if (do_colour_align)
- addr = COLOUR_ALIGN(addr, pgoff);
- }
+ info.flags = 0;
+ info.length = len;
+ info.low_limit = TASK_UNMAPPED_BASE;
+ info.high_limit = TASK_SIZE;
+ info.align_mask = do_colour_align ? (PAGE_MASK & shm_align_mask) : 0;
+ info.align_offset = pgoff << PAGE_SHIFT;
+ return vm_unmapped_area(&info);
}
unsigned long
@@ -131,6 +85,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0,
struct mm_struct *mm = current->mm;
unsigned long addr = addr0;
int do_colour_align;
+ struct vm_unmapped_area_info info;
if (flags & MAP_FIXED) {
/* We do not accept a shared mapping if it would violate
@@ -162,73 +117,27 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0,
return addr;
}
- /* check if free_area_cache is useful for us */
- if (len <= mm->cached_hole_size) {
- mm->cached_hole_size = 0;
- mm->free_area_cache = mm->mmap_base;
- }
-
- /* either no address requested or can't fit in requested address hole */
- addr = mm->free_area_cache;
- if (do_colour_align) {
- unsigned long base = COLOUR_ALIGN_DOWN(addr-len, pgoff);
+ info.flags = VM_UNMAPPED_AREA_TOPDOWN;
+ info.length = len;
+ info.low_limit = PAGE_SIZE;
+ info.high_limit = mm->mmap_base;
+ info.align_mask = do_colour_align ? (PAGE_MASK & shm_align_mask) : 0;
+ info.align_offset = pgoff << PAGE_SHIFT;
+ addr = vm_unmapped_area(&info);
- addr = base + len;
- }
-
- /* make sure it can fit in the remaining address space */
- if (likely(addr > len)) {
- vma = find_vma(mm, addr-len);
- if (!vma || addr <= vma->vm_start) {
- /* remember the address as a hint for next time */
- return (mm->free_area_cache = addr-len);
- }
- }
-
- if (unlikely(mm->mmap_base < len))
- goto bottomup;
-
- addr = mm->mmap_base-len;
- if (do_colour_align)
- addr = COLOUR_ALIGN_DOWN(addr, pgoff);
-
- do {
- /*
- * Lookup failure means no vma is above this address,
- * else if new region fits below vma->vm_start,
- * return with success:
- */
- vma = find_vma(mm, addr);
- if (likely(!vma || addr+len <= vma->vm_start)) {
- /* remember the address as a hint for next time */
- return (mm->free_area_cache = addr);
- }
-
- /* remember the largest hole we saw so far */
- if (addr + mm->cached_hole_size < vma->vm_start)
- mm->cached_hole_size = vma->vm_start - addr;
-
- /* try just below the current vma->vm_start */
- addr = vma->vm_start-len;
- if (do_colour_align)
- addr = COLOUR_ALIGN_DOWN(addr, pgoff);
- } while (likely(len < vma->vm_start));
-
-bottomup:
/*
* A failed mmap() very likely causes application failure,
* so fall back to the bottom-up function here. This scenario
* can happen with large stack limits and large mmap()
* allocations.
*/
- mm->cached_hole_size = ~0UL;
- mm->free_area_cache = TASK_UNMAPPED_BASE;
- addr = arch_get_unmapped_area(filp, addr0, len, pgoff, flags);
- /*
- * Restore the topdown base:
- */
- mm->free_area_cache = mm->mmap_base;
- mm->cached_hole_size = ~0UL;
+ if (addr & ~PAGE_MASK) {
+ VM_BUG_ON(addr != -ENOMEM);
+ info.flags = 0;
+ info.low_limit = TASK_UNMAPPED_BASE;
+ info.high_limit = TASK_SIZE;
+ addr = vm_unmapped_area(&info);
+ }
return addr;
}
@@ -238,7 +147,7 @@ bottomup:
* You really shouldn't be using read() or write() on /dev/mem. This
* might go away in the future.
*/
-int valid_phys_addr_range(unsigned long addr, size_t count)
+int valid_phys_addr_range(phys_addr_t addr, size_t count)
{
if (addr < __MEMORY_START)
return 0;
diff --git a/arch/sh/mm/pmb.c b/arch/sh/mm/pmb.c
index fad52f1f681..7160c9fd6fe 100644
--- a/arch/sh/mm/pmb.c
+++ b/arch/sh/mm/pmb.c
@@ -25,7 +25,6 @@
#include <linux/vmalloc.h>
#include <asm/cacheflush.h>
#include <asm/sizes.h>
-#include <asm/system.h>
#include <asm/uaccess.h>
#include <asm/pgtable.h>
#include <asm/page.h>
diff --git a/arch/sh/mm/sram.c b/arch/sh/mm/sram.c
index bc156ec4545..2d8fa718d55 100644
--- a/arch/sh/mm/sram.c
+++ b/arch/sh/mm/sram.c
@@ -9,6 +9,7 @@
*/
#include <linux/init.h>
#include <linux/kernel.h>
+#include <linux/errno.h>
#include <asm/sram.h>
/*
diff --git a/arch/sh/mm/tlb-pteaex.c b/arch/sh/mm/tlb-pteaex.c
index b71db6af806..4db21adfe5d 100644
--- a/arch/sh/mm/tlb-pteaex.c
+++ b/arch/sh/mm/tlb-pteaex.c
@@ -12,7 +12,6 @@
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/io.h>
-#include <asm/system.h>
#include <asm/mmu_context.h>
#include <asm/cacheflush.h>
diff --git a/arch/sh/mm/tlb-sh3.c b/arch/sh/mm/tlb-sh3.c
index 7a940dbfc2e..6554fb439f0 100644
--- a/arch/sh/mm/tlb-sh3.c
+++ b/arch/sh/mm/tlb-sh3.c
@@ -20,7 +20,6 @@
#include <linux/smp.h>
#include <linux/interrupt.h>
-#include <asm/system.h>
#include <asm/io.h>
#include <asm/uaccess.h>
#include <asm/pgalloc.h>
diff --git a/arch/sh/mm/tlb-sh4.c b/arch/sh/mm/tlb-sh4.c
index cfdf7930d29..d42dd7e443d 100644
--- a/arch/sh/mm/tlb-sh4.c
+++ b/arch/sh/mm/tlb-sh4.c
@@ -11,7 +11,6 @@
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/io.h>
-#include <asm/system.h>
#include <asm/mmu_context.h>
#include <asm/cacheflush.h>
diff --git a/arch/sh/mm/tlb-sh5.c b/arch/sh/mm/tlb-sh5.c
index f27dbe1c159..e4bb2a8e0a6 100644
--- a/arch/sh/mm/tlb-sh5.c
+++ b/arch/sh/mm/tlb-sh5.c
@@ -17,7 +17,7 @@
/**
* sh64_tlb_init - Perform initial setup for the DTLB and ITLB.
*/
-int __init sh64_tlb_init(void)
+int sh64_tlb_init(void)
{
/* Assign some sane DTLB defaults */
cpu_data->dtlb.entries = 64;
@@ -182,3 +182,43 @@ void tlb_unwire_entry(void)
local_irq_restore(flags);
}
+
+void __update_tlb(struct vm_area_struct *vma, unsigned long address, pte_t pte)
+{
+ unsigned long long ptel;
+ unsigned long long pteh=0;
+ struct tlb_info *tlbp;
+ unsigned long long next;
+ unsigned int fault_code = get_thread_fault_code();
+
+ /* Get PTEL first */
+ ptel = pte.pte_low;
+
+ /*
+ * Set PTEH register
+ */
+ pteh = neff_sign_extend(address & MMU_VPN_MASK);
+
+ /* Set the ASID. */
+ pteh |= get_asid() << PTEH_ASID_SHIFT;
+ pteh |= PTEH_VALID;
+
+ /* Set PTEL register, set_pte has performed the sign extension */
+ ptel &= _PAGE_FLAGS_HARDWARE_MASK; /* drop software flags */
+
+ if (fault_code & FAULT_CODE_ITLB)
+ tlbp = &cpu_data->itlb;
+ else
+ tlbp = &cpu_data->dtlb;
+
+ next = tlbp->next;
+ __flush_tlb_slot(next);
+ asm volatile ("putcfg %0,1,%2\n\n\t"
+ "putcfg %0,0,%1\n"
+ : : "r" (next), "r" (pteh), "r" (ptel) );
+
+ next += TLB_STEP;
+ if (next > tlbp->last)
+ next = tlbp->first;
+ tlbp->next = next;
+}
diff --git a/arch/sh/mm/tlbex_32.c b/arch/sh/mm/tlbex_32.c
new file mode 100644
index 00000000000..382262dc0c4
--- /dev/null
+++ b/arch/sh/mm/tlbex_32.c
@@ -0,0 +1,78 @@
+/*
+ * TLB miss handler for SH with an MMU.
+ *
+ * Copyright (C) 1999 Niibe Yutaka
+ * Copyright (C) 2003 - 2012 Paul Mundt
+ *
+ * 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/mm.h>
+#include <linux/kprobes.h>
+#include <linux/kdebug.h>
+#include <asm/mmu_context.h>
+#include <asm/thread_info.h>
+
+/*
+ * Called with interrupts disabled.
+ */
+asmlinkage int __kprobes
+handle_tlbmiss(struct pt_regs *regs, unsigned long error_code,
+ unsigned long address)
+{
+ pgd_t *pgd;
+ pud_t *pud;
+ pmd_t *pmd;
+ pte_t *pte;
+ pte_t entry;
+
+ /*
+ * We don't take page faults for P1, P2, and parts of P4, these
+ * are always mapped, whether it be due to legacy behaviour in
+ * 29-bit mode, or due to PMB configuration in 32-bit mode.
+ */
+ if (address >= P3SEG && address < P3_ADDR_MAX) {
+ pgd = pgd_offset_k(address);
+ } else {
+ if (unlikely(address >= TASK_SIZE || !current->mm))
+ return 1;
+
+ pgd = pgd_offset(current->mm, address);
+ }
+
+ pud = pud_offset(pgd, address);
+ if (pud_none_or_clear_bad(pud))
+ return 1;
+ pmd = pmd_offset(pud, address);
+ if (pmd_none_or_clear_bad(pmd))
+ return 1;
+ pte = pte_offset_kernel(pmd, address);
+ entry = *pte;
+ if (unlikely(pte_none(entry) || pte_not_present(entry)))
+ return 1;
+ if (unlikely(error_code && !pte_write(entry)))
+ return 1;
+
+ if (error_code)
+ entry = pte_mkdirty(entry);
+ entry = pte_mkyoung(entry);
+
+ set_pte(pte, entry);
+
+#if defined(CONFIG_CPU_SH4) && !defined(CONFIG_SMP)
+ /*
+ * SH-4 does not set MMUCR.RC to the corresponding TLB entry in
+ * the case of an initial page write exception, so we need to
+ * flush it in order to avoid potential TLB entry duplication.
+ */
+ if (error_code == FAULT_CODE_INITIAL)
+ local_flush_tlb_one(get_asid(), address & PAGE_MASK);
+#endif
+
+ set_thread_fault_code(error_code);
+ update_mmu_cache(NULL, address, pte);
+
+ return 0;
+}
diff --git a/arch/sh/mm/tlbex_64.c b/arch/sh/mm/tlbex_64.c
new file mode 100644
index 00000000000..8557548fc53
--- /dev/null
+++ b/arch/sh/mm/tlbex_64.c
@@ -0,0 +1,166 @@
+/*
+ * The SH64 TLB miss.
+ *
+ * Original code from fault.c
+ * Copyright (C) 2000, 2001 Paolo Alberelli
+ *
+ * Fast PTE->TLB refill path
+ * Copyright (C) 2003 Richard.Curnow@superh.com
+ *
+ * IMPORTANT NOTES :
+ * The do_fast_page_fault function is called from a context in entry.S
+ * where very few registers have been saved. In particular, the code in
+ * this file must be compiled not to use ANY caller-save registers that
+ * are not part of the restricted save set. Also, it means that code in
+ * this file must not make calls to functions elsewhere in the kernel, or
+ * else the excepting context will see corruption in its caller-save
+ * registers. Plus, the entry.S save area is non-reentrant, so this code
+ * has to run with SR.BL==1, i.e. no interrupts taken inside it and panic
+ * on any exception.
+ *
+ * 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/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/smp.h>
+#include <linux/interrupt.h>
+#include <linux/kprobes.h>
+#include <asm/tlb.h>
+#include <asm/io.h>
+#include <asm/uaccess.h>
+#include <asm/pgalloc.h>
+#include <asm/mmu_context.h>
+
+static int handle_tlbmiss(unsigned long long protection_flags,
+ unsigned long address)
+{
+ pgd_t *pgd;
+ pud_t *pud;
+ pmd_t *pmd;
+ pte_t *pte;
+ pte_t entry;
+
+ if (is_vmalloc_addr((void *)address)) {
+ pgd = pgd_offset_k(address);
+ } else {
+ if (unlikely(address >= TASK_SIZE || !current->mm))
+ return 1;
+
+ pgd = pgd_offset(current->mm, address);
+ }
+
+ pud = pud_offset(pgd, address);
+ if (pud_none(*pud) || !pud_present(*pud))
+ return 1;
+
+ pmd = pmd_offset(pud, address);
+ if (pmd_none(*pmd) || !pmd_present(*pmd))
+ return 1;
+
+ pte = pte_offset_kernel(pmd, address);
+ entry = *pte;
+ if (pte_none(entry) || !pte_present(entry))
+ return 1;
+
+ /*
+ * If the page doesn't have sufficient protection bits set to
+ * service the kind of fault being handled, there's not much
+ * point doing the TLB refill. Punt the fault to the general
+ * handler.
+ */
+ if ((pte_val(entry) & protection_flags) != protection_flags)
+ return 1;
+
+ update_mmu_cache(NULL, address, pte);
+
+ return 0;
+}
+
+/*
+ * Put all this information into one structure so that everything is just
+ * arithmetic relative to a single base address. This reduces the number
+ * of movi/shori pairs needed just to load addresses of static data.
+ */
+struct expevt_lookup {
+ unsigned short protection_flags[8];
+ unsigned char is_text_access[8];
+ unsigned char is_write_access[8];
+};
+
+#define PRU (1<<9)
+#define PRW (1<<8)
+#define PRX (1<<7)
+#define PRR (1<<6)
+
+/* Sized as 8 rather than 4 to allow checking the PTE's PRU bit against whether
+ the fault happened in user mode or privileged mode. */
+static struct expevt_lookup expevt_lookup_table = {
+ .protection_flags = {PRX, PRX, 0, 0, PRR, PRR, PRW, PRW},
+ .is_text_access = {1, 1, 0, 0, 0, 0, 0, 0}
+};
+
+static inline unsigned int
+expevt_to_fault_code(unsigned long expevt)
+{
+ if (expevt == 0xa40)
+ return FAULT_CODE_ITLB;
+ else if (expevt == 0x060)
+ return FAULT_CODE_WRITE;
+
+ return 0;
+}
+
+/*
+ This routine handles page faults that can be serviced just by refilling a
+ TLB entry from an existing page table entry. (This case represents a very
+ large majority of page faults.) Return 1 if the fault was successfully
+ handled. Return 0 if the fault could not be handled. (This leads into the
+ general fault handling in fault.c which deals with mapping file-backed
+ pages, stack growth, segmentation faults, swapping etc etc)
+ */
+asmlinkage int __kprobes
+do_fast_page_fault(unsigned long long ssr_md, unsigned long long expevt,
+ unsigned long address)
+{
+ unsigned long long protection_flags;
+ unsigned long long index;
+ unsigned long long expevt4;
+ unsigned int fault_code;
+
+ /* The next few lines implement a way of hashing EXPEVT into a
+ * small array index which can be used to lookup parameters
+ * specific to the type of TLBMISS being handled.
+ *
+ * Note:
+ * ITLBMISS has EXPEVT==0xa40
+ * RTLBMISS has EXPEVT==0x040
+ * WTLBMISS has EXPEVT==0x060
+ */
+ expevt4 = (expevt >> 4);
+ /* TODO : xor ssr_md into this expression too. Then we can check
+ * that PRU is set when it needs to be. */
+ index = expevt4 ^ (expevt4 >> 5);
+ index &= 7;
+
+ fault_code = expevt_to_fault_code(expevt);
+
+ protection_flags = expevt_lookup_table.protection_flags[index];
+
+ if (expevt_lookup_table.is_text_access[index])
+ fault_code |= FAULT_CODE_ITLB;
+ if (!ssr_md)
+ fault_code |= FAULT_CODE_USER;
+
+ set_thread_fault_code(fault_code);
+
+ return handle_tlbmiss(protection_flags, address);
+}
diff --git a/arch/sh/mm/tlbflush_64.c b/arch/sh/mm/tlbflush_64.c
index e3430e093d4..f33fdd2558e 100644
--- a/arch/sh/mm/tlbflush_64.c
+++ b/arch/sh/mm/tlbflush_64.c
@@ -3,7 +3,7 @@
*
* Copyright (C) 2000, 2001 Paolo Alberelli
* Copyright (C) 2003 Richard Curnow (/proc/tlb, bug fixes)
- * Copyright (C) 2003 - 2009 Paul Mundt
+ * Copyright (C) 2003 - 2012 Paul Mundt
*
* 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
@@ -22,301 +22,12 @@
#include <linux/smp.h>
#include <linux/perf_event.h>
#include <linux/interrupt.h>
-#include <asm/system.h>
#include <asm/io.h>
#include <asm/tlb.h>
#include <asm/uaccess.h>
#include <asm/pgalloc.h>
#include <asm/mmu_context.h>
-extern void die(const char *,struct pt_regs *,long);
-
-#define PFLAG(val,flag) (( (val) & (flag) ) ? #flag : "" )
-#define PPROT(flag) PFLAG(pgprot_val(prot),flag)
-
-static inline void print_prots(pgprot_t prot)
-{
- printk("prot is 0x%016llx\n",pgprot_val(prot));
-
- printk("%s %s %s %s %s\n",PPROT(_PAGE_SHARED),PPROT(_PAGE_READ),
- PPROT(_PAGE_EXECUTE),PPROT(_PAGE_WRITE),PPROT(_PAGE_USER));
-}
-
-static inline void print_vma(struct vm_area_struct *vma)
-{
- printk("vma start 0x%08lx\n", vma->vm_start);
- printk("vma end 0x%08lx\n", vma->vm_end);
-
- print_prots(vma->vm_page_prot);
- printk("vm_flags 0x%08lx\n", vma->vm_flags);
-}
-
-static inline void print_task(struct task_struct *tsk)
-{
- printk("Task pid %d\n", task_pid_nr(tsk));
-}
-
-static pte_t *lookup_pte(struct mm_struct *mm, unsigned long address)
-{
- pgd_t *dir;
- pud_t *pud;
- pmd_t *pmd;
- pte_t *pte;
- pte_t entry;
-
- dir = pgd_offset(mm, address);
- if (pgd_none(*dir))
- return NULL;
-
- pud = pud_offset(dir, address);
- if (pud_none(*pud))
- return NULL;
-
- pmd = pmd_offset(pud, address);
- if (pmd_none(*pmd))
- return NULL;
-
- pte = pte_offset_kernel(pmd, address);
- entry = *pte;
- if (pte_none(entry) || !pte_present(entry))
- return NULL;
-
- return pte;
-}
-
-/*
- * This routine handles page faults. It determines the address,
- * and the problem, and then passes it off to one of the appropriate
- * routines.
- */
-asmlinkage void do_page_fault(struct pt_regs *regs, unsigned long writeaccess,
- unsigned long textaccess, unsigned long address)
-{
- struct task_struct *tsk;
- struct mm_struct *mm;
- struct vm_area_struct * vma;
- const struct exception_table_entry *fixup;
- pte_t *pte;
- int fault;
-
- /* SIM
- * Note this is now called with interrupts still disabled
- * This is to cope with being called for a missing IO port
- * address with interrupts disabled. This should be fixed as
- * soon as we have a better 'fast path' miss handler.
- *
- * Plus take care how you try and debug this stuff.
- * For example, writing debug data to a port which you
- * have just faulted on is not going to work.
- */
-
- tsk = current;
- mm = tsk->mm;
-
- /* Not an IO address, so reenable interrupts */
- local_irq_enable();
-
- perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
-
- /*
- * If we're in an interrupt or have no user
- * context, we must not take the fault..
- */
- if (in_atomic() || !mm)
- goto no_context;
-
- /* TLB misses upon some cache flushes get done under cli() */
- down_read(&mm->mmap_sem);
-
- vma = find_vma(mm, address);
-
- if (!vma) {
-#ifdef DEBUG_FAULT
- print_task(tsk);
- printk("%s:%d fault, address is 0x%08x PC %016Lx textaccess %d writeaccess %d\n",
- __func__, __LINE__,
- address,regs->pc,textaccess,writeaccess);
- show_regs(regs);
-#endif
- goto bad_area;
- }
- if (vma->vm_start <= address) {
- goto good_area;
- }
-
- if (!(vma->vm_flags & VM_GROWSDOWN)) {
-#ifdef DEBUG_FAULT
- print_task(tsk);
- printk("%s:%d fault, address is 0x%08x PC %016Lx textaccess %d writeaccess %d\n",
- __func__, __LINE__,
- address,regs->pc,textaccess,writeaccess);
- show_regs(regs);
-
- print_vma(vma);
-#endif
- goto bad_area;
- }
- if (expand_stack(vma, address)) {
-#ifdef DEBUG_FAULT
- print_task(tsk);
- printk("%s:%d fault, address is 0x%08x PC %016Lx textaccess %d writeaccess %d\n",
- __func__, __LINE__,
- address,regs->pc,textaccess,writeaccess);
- show_regs(regs);
-#endif
- goto bad_area;
- }
-/*
- * Ok, we have a good vm_area for this memory access, so
- * we can handle it..
- */
-good_area:
- if (textaccess) {
- if (!(vma->vm_flags & VM_EXEC))
- goto bad_area;
- } else {
- if (writeaccess) {
- if (!(vma->vm_flags & VM_WRITE))
- goto bad_area;
- } else {
- if (!(vma->vm_flags & VM_READ))
- 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(mm, vma, address, writeaccess ? FAULT_FLAG_WRITE : 0);
- 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) {
- tsk->maj_flt++;
- perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1,
- regs, address);
- } else {
- tsk->min_flt++;
- perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1,
- regs, address);
- }
-
- /* If we get here, the page fault has been handled. Do the TLB refill
- now from the newly-setup PTE, to avoid having to fault again right
- away on the same instruction. */
- pte = lookup_pte (mm, address);
- if (!pte) {
- /* From empirical evidence, we can get here, due to
- !pte_present(pte). (e.g. if a swap-in occurs, and the page
- is swapped back out again before the process that wanted it
- gets rescheduled?) */
- goto no_pte;
- }
-
- __do_tlb_refill(address, textaccess, pte);
-
-no_pte:
-
- up_read(&mm->mmap_sem);
- return;
-
-/*
- * Something tried to access memory that isn't in our memory map..
- * Fix it, but check if it's kernel or user first..
- */
-bad_area:
-#ifdef DEBUG_FAULT
- printk("fault:bad area\n");
-#endif
- up_read(&mm->mmap_sem);
-
- if (user_mode(regs)) {
- static int count=0;
- siginfo_t info;
- if (count < 4) {
- /* This is really to help debug faults when starting
- * usermode, so only need a few */
- count++;
- printk("user mode bad_area address=%08lx pid=%d (%s) pc=%08lx\n",
- address, task_pid_nr(current), current->comm,
- (unsigned long) regs->pc);
-#if 0
- show_regs(regs);
-#endif
- }
- if (is_global_init(tsk)) {
- panic("INIT had user mode bad_area\n");
- }
- tsk->thread.address = address;
- tsk->thread.error_code = writeaccess;
- info.si_signo = SIGSEGV;
- info.si_errno = 0;
- info.si_addr = (void *) address;
- force_sig_info(SIGSEGV, &info, tsk);
- return;
- }
-
-no_context:
-#ifdef DEBUG_FAULT
- printk("fault:No context\n");
-#endif
- /* Are we prepared to handle this kernel fault? */
- fixup = search_exception_tables(regs->pc);
- if (fixup) {
- regs->pc = fixup->fixup;
- return;
- }
-
-/*
- * Oops. The kernel tried to access some bad page. We'll have to
- * terminate things with extreme prejudice.
- *
- */
- if (address < PAGE_SIZE)
- printk(KERN_ALERT "Unable to handle kernel NULL pointer dereference");
- else
- printk(KERN_ALERT "Unable to handle kernel paging request");
- printk(" at virtual address %08lx\n", address);
- printk(KERN_ALERT "pc = %08Lx%08Lx\n", regs->pc >> 32, regs->pc & 0xffffffff);
- die("Oops", regs, writeaccess);
- do_exit(SIGKILL);
-
-/*
- * 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))
- goto no_context;
- pagefault_out_of_memory();
- return;
-
-do_sigbus:
- printk("fault:Do sigbus\n");
- up_read(&mm->mmap_sem);
-
- /*
- * Send a sigbus, regardless of whether we were in kernel
- * or user mode.
- */
- tsk->thread.address = address;
- tsk->thread.error_code = writeaccess;
- tsk->thread.trap_no = 14;
- force_sig(SIGBUS, tsk);
-
- /* Kernel mode? Handle exceptions or die */
- if (!user_mode(regs))
- goto no_context;
-}
-
void local_flush_tlb_one(unsigned long asid, unsigned long page)
{
unsigned long long match, pteh=0, lpage;
@@ -459,7 +170,3 @@ void __flush_tlb_global(void)
{
flush_tlb_all();
}
-
-void __update_tlb(struct vm_area_struct *vma, unsigned long address, pte_t pte)
-{
-}