* Loosely based on earlier PAT patchset from Eric Biederman and Andi Kleen.
*/
-#include <linux/mm.h>
+#include <linux/seq_file.h>
+#include <linux/bootmem.h>
+#include <linux/debugfs.h>
#include <linux/kernel.h>
+#include <linux/module.h>
#include <linux/gfp.h>
+#include <linux/mm.h>
#include <linux/fs.h>
-#include <linux/bootmem.h>
-#include <asm/msr.h>
-#include <asm/tlbflush.h>
+#include <asm/cacheflush.h>
#include <asm/processor.h>
-#include <asm/page.h>
+#include <asm/tlbflush.h>
#include <asm/pgtable.h>
-#include <asm/pat.h>
-#include <asm/e820.h>
-#include <asm/cacheflush.h>
#include <asm/fcntl.h>
+#include <asm/e820.h>
#include <asm/mtrr.h>
+#include <asm/page.h>
+#include <asm/msr.h>
+#include <asm/pat.h>
#include <asm/io.h>
#ifdef CONFIG_X86_PAT
int __read_mostly pat_enabled = 1;
-void __cpuinit pat_disable(char *reason)
+static inline void pat_disable(const char *reason)
{
pat_enabled = 0;
printk(KERN_INFO "%s\n", reason);
return 0;
}
early_param("nopat", nopat);
+#else
+static inline void pat_disable(const char *reason)
+{
+ (void)reason;
+}
#endif
static int debug_enable;
+
static int __init pat_debug_setup(char *str)
{
debug_enable = 1;
if (!pat_enabled)
return;
- /* Paranoia check. */
- if (!cpu_has_pat && boot_pat_state) {
- /*
- * If this happens we are on a secondary CPU, but
- * switched to PAT on the boot CPU. We have no way to
- * undo PAT.
- */
- printk(KERN_ERR "PAT enabled, "
- "but not supported by secondary CPU\n");
- BUG();
+ if (!cpu_has_pat) {
+ if (!boot_pat_state) {
+ pat_disable("PAT not supported by CPU.");
+ return;
+ } else {
+ /*
+ * If this happens we are on a secondary CPU, but
+ * switched to PAT on the boot CPU. We have no way to
+ * undo PAT.
+ */
+ printk(KERN_ERR "PAT enabled, "
+ "but not supported by secondary CPU\n");
+ BUG();
+ }
}
/* Set PWT to Write-Combining. All other bits stay the same */
*/
struct memtype {
- u64 start;
- u64 end;
- unsigned long type;
- struct list_head nd;
+ u64 start;
+ u64 end;
+ unsigned long type;
+ struct list_head nd;
};
static LIST_HEAD(memtype_list);
-static DEFINE_SPINLOCK(memtype_lock); /* protects memtype list */
+static DEFINE_SPINLOCK(memtype_lock); /* protects memtype list */
/*
* Does intersection of PAT memory type and MTRR memory type and returns
u8 mtrr_type;
mtrr_type = mtrr_type_lookup(start, end);
- if (mtrr_type == MTRR_TYPE_UNCACHABLE)
- return _PAGE_CACHE_UC;
- if (mtrr_type == MTRR_TYPE_WRCOMB)
- return _PAGE_CACHE_WC;
+ if (mtrr_type != MTRR_TYPE_WRBACK)
+ return _PAGE_CACHE_UC_MINUS;
+
+ return _PAGE_CACHE_WB;
}
return req_type;
}
-static int chk_conflict(struct memtype *new, struct memtype *entry,
- unsigned long *type)
+static int
+chk_conflict(struct memtype *new, struct memtype *entry, unsigned long *type)
{
if (new->type != entry->type) {
if (type) {
return -EBUSY;
}
+static struct memtype *cached_entry;
+static u64 cached_start;
+
+static int pat_pagerange_is_ram(unsigned long start, unsigned long end)
+{
+ int ram_page = 0, not_rampage = 0;
+ unsigned long page_nr;
+
+ for (page_nr = (start >> PAGE_SHIFT); page_nr < (end >> PAGE_SHIFT);
+ ++page_nr) {
+ /*
+ * For legacy reasons, physical address range in the legacy ISA
+ * region is tracked as non-RAM. This will allow users of
+ * /dev/mem to map portions of legacy ISA region, even when
+ * some of those portions are listed(or not even listed) with
+ * different e820 types(RAM/reserved/..)
+ */
+ if (page_nr >= (ISA_END_ADDRESS >> PAGE_SHIFT) &&
+ page_is_ram(page_nr))
+ ram_page = 1;
+ else
+ not_rampage = 1;
+
+ if (ram_page == not_rampage)
+ return -1;
+ }
+
+ return ram_page;
+}
+
+/*
+ * For RAM pages, mark the pages as non WB memory type using
+ * PageNonWB (PG_arch_1). We allow only one set_memory_uc() or
+ * set_memory_wc() on a RAM page at a time before marking it as WB again.
+ * This is ok, because only one driver will be owning the page and
+ * doing set_memory_*() calls.
+ *
+ * For now, we use PageNonWB to track that the RAM page is being mapped
+ * as non WB. In future, we will have to use one more flag
+ * (or some other mechanism in page_struct) to distinguish between
+ * UC and WC mapping.
+ */
+static int reserve_ram_pages_type(u64 start, u64 end, unsigned long req_type,
+ unsigned long *new_type)
+{
+ struct page *page;
+ u64 pfn, end_pfn;
+
+ for (pfn = (start >> PAGE_SHIFT); pfn < (end >> PAGE_SHIFT); ++pfn) {
+ page = pfn_to_page(pfn);
+ if (page_mapped(page) || PageNonWB(page))
+ goto out;
+
+ SetPageNonWB(page);
+ }
+ return 0;
+
+out:
+ end_pfn = pfn;
+ for (pfn = (start >> PAGE_SHIFT); pfn < end_pfn; ++pfn) {
+ page = pfn_to_page(pfn);
+ ClearPageNonWB(page);
+ }
+
+ return -EINVAL;
+}
+
+static int free_ram_pages_type(u64 start, u64 end)
+{
+ struct page *page;
+ u64 pfn, end_pfn;
+
+ for (pfn = (start >> PAGE_SHIFT); pfn < (end >> PAGE_SHIFT); ++pfn) {
+ page = pfn_to_page(pfn);
+ if (page_mapped(page) || !PageNonWB(page))
+ goto out;
+
+ ClearPageNonWB(page);
+ }
+ return 0;
+
+out:
+ end_pfn = pfn;
+ for (pfn = (start >> PAGE_SHIFT); pfn < end_pfn; ++pfn) {
+ page = pfn_to_page(pfn);
+ SetPageNonWB(page);
+ }
+ return -EINVAL;
+}
+
/*
* req_type typically has one of the:
* - _PAGE_CACHE_WB
* it will return a negative return value.
*/
int reserve_memtype(u64 start, u64 end, unsigned long req_type,
- unsigned long *new_type)
+ unsigned long *new_type)
{
struct memtype *new, *entry;
unsigned long actual_type;
struct list_head *where;
+ int is_range_ram;
int err = 0;
- BUG_ON(start >= end); /* end is exclusive */
+ BUG_ON(start >= end); /* end is exclusive */
if (!pat_enabled) {
/* This is identical to page table setting without PAT */
return 0;
}
- if (req_type == -1) {
- /*
- * Call mtrr_lookup to get the type hint. This is an
- * optimization for /dev/mem mmap'ers into WB memory (BIOS
- * tools and ACPI tools). Use WB request for WB memory and use
- * UC_MINUS otherwise.
- */
- u8 mtrr_type = mtrr_type_lookup(start, end);
+ /*
+ * Call mtrr_lookup to get the type hint. This is an
+ * optimization for /dev/mem mmap'ers into WB memory (BIOS
+ * tools and ACPI tools). Use WB request for WB memory and use
+ * UC_MINUS otherwise.
+ */
+ actual_type = pat_x_mtrr_type(start, end, req_type & _PAGE_CACHE_MASK);
- if (mtrr_type == MTRR_TYPE_WRBACK)
- actual_type = _PAGE_CACHE_WB;
- else
- actual_type = _PAGE_CACHE_UC_MINUS;
- } else
- actual_type = pat_x_mtrr_type(start, end,
- req_type & _PAGE_CACHE_MASK);
+ if (new_type)
+ *new_type = actual_type;
+
+ is_range_ram = pat_pagerange_is_ram(start, end);
+ if (is_range_ram == 1)
+ return reserve_ram_pages_type(start, end, req_type,
+ new_type);
+ else if (is_range_ram < 0)
+ return -EINVAL;
new = kmalloc(sizeof(struct memtype), GFP_KERNEL);
if (!new)
return -ENOMEM;
- new->start = start;
- new->end = end;
- new->type = actual_type;
-
- if (new_type)
- *new_type = actual_type;
+ new->start = start;
+ new->end = end;
+ new->type = actual_type;
spin_lock(&memtype_lock);
+ if (cached_entry && start >= cached_start)
+ entry = cached_entry;
+ else
+ entry = list_entry(&memtype_list, struct memtype, nd);
+
/* Search for existing mapping that overlaps the current range */
where = NULL;
- list_for_each_entry(entry, &memtype_list, nd) {
+ list_for_each_entry_continue(entry, &memtype_list, nd) {
if (end <= entry->start) {
where = entry->nd.prev;
+ cached_entry = list_entry(where, struct memtype, nd);
break;
} else if (start <= entry->start) { /* end > entry->start */
err = chk_conflict(new, entry, new_type);
dprintk("Overlap at 0x%Lx-0x%Lx\n",
entry->start, entry->end);
where = entry->nd.prev;
+ cached_entry = list_entry(where,
+ struct memtype, nd);
}
break;
} else if (start < entry->end) { /* start > entry->start */
if (!err) {
dprintk("Overlap at 0x%Lx-0x%Lx\n",
entry->start, entry->end);
- where = &entry->nd;
+ cached_entry = list_entry(entry->nd.prev,
+ struct memtype, nd);
+
+ /*
+ * Move to right position in the linked
+ * list to add this new entry
+ */
+ list_for_each_entry_continue(entry,
+ &memtype_list, nd) {
+ if (start <= entry->start) {
+ where = entry->nd.prev;
+ break;
+ }
+ }
}
break;
}
start, end, cattr_name(new->type), cattr_name(req_type));
kfree(new);
spin_unlock(&memtype_lock);
+
return err;
}
+ cached_start = start;
+
if (where)
list_add(&new->nd, where);
else
{
struct memtype *entry;
int err = -EINVAL;
+ int is_range_ram;
if (!pat_enabled)
return 0;
if (is_ISA_range(start, end - 1))
return 0;
+ is_range_ram = pat_pagerange_is_ram(start, end);
+ if (is_range_ram == 1)
+ return free_ram_pages_type(start, end);
+ else if (is_range_ram < 0)
+ return -EINVAL;
+
spin_lock(&memtype_lock);
list_for_each_entry(entry, &memtype_list, nd) {
if (entry->start == start && entry->end == end) {
+ if (cached_entry == entry || cached_start == start)
+ cached_entry = NULL;
+
list_del(&entry->nd);
kfree(entry);
err = 0;
}
dprintk("free_memtype request 0x%Lx-0x%Lx\n", start, end);
+
return err;
}
-/*
- * /dev/mem mmap interface. The memtype used for mapping varies:
- * - Use UC for mappings with O_SYNC flag
- * - Without O_SYNC flag, if there is any conflict in reserve_memtype,
- * inherit the memtype from existing mapping.
- * - Else use UC_MINUS memtype (for backward compatibility with existing
- * X drivers.
- */
pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
unsigned long size, pgprot_t vma_prot)
{
return vma_prot;
}
-#ifdef CONFIG_NONPROMISC_DEVMEM
-/* This check is done in drivers/char/mem.c in case of NONPROMISC_DEVMEM*/
+#ifdef CONFIG_STRICT_DEVMEM
+/* This check is done in drivers/char/mem.c in case of STRICT_DEVMEM*/
static inline int range_is_allowed(unsigned long pfn, unsigned long size)
{
return 1;
}
#else
+/* This check is needed to avoid cache aliasing when PAT is enabled */
static inline int range_is_allowed(unsigned long pfn, unsigned long size)
{
u64 from = ((u64)pfn) << PAGE_SHIFT;
u64 to = from + size;
u64 cursor = from;
+ if (!pat_enabled)
+ return 1;
+
while (cursor < to) {
if (!devmem_is_allowed(pfn)) {
printk(KERN_INFO
}
return 1;
}
-#endif /* CONFIG_NONPROMISC_DEVMEM */
+#endif /* CONFIG_STRICT_DEVMEM */
int phys_mem_access_prot_allowed(struct file *file, unsigned long pfn,
unsigned long size, pgprot_t *vma_prot)
{
- u64 offset = ((u64) pfn) << PAGE_SHIFT;
- unsigned long flags = _PAGE_CACHE_UC_MINUS;
- int retval;
+ unsigned long flags = _PAGE_CACHE_WB;
if (!range_is_allowed(pfn, size))
return 0;
if (file->f_flags & O_SYNC) {
- flags = _PAGE_CACHE_UC;
+ flags = _PAGE_CACHE_UC_MINUS;
}
#ifdef CONFIG_X86_32
}
#endif
- /*
- * With O_SYNC, we can only take UC mapping. Fail if we cannot.
- * Without O_SYNC, we want to get
- * - WB for WB-able memory and no other conflicting mappings
- * - UC_MINUS for non-WB-able memory with no other conflicting mappings
- * - Inherit from confliting mappings otherwise
- */
- if (flags != _PAGE_CACHE_UC_MINUS) {
- retval = reserve_memtype(offset, offset + size, flags, NULL);
- } else {
- retval = reserve_memtype(offset, offset + size, -1, &flags);
- }
+ *vma_prot = __pgprot((pgprot_val(*vma_prot) & ~_PAGE_CACHE_MASK) |
+ flags);
+ return 1;
+}
+
+/*
+ * Change the memory type for the physial address range in kernel identity
+ * mapping space if that range is a part of identity map.
+ */
+int kernel_map_sync_memtype(u64 base, unsigned long size, unsigned long flags)
+{
+ unsigned long id_sz;
- if (retval < 0)
+ if (!pat_enabled || base >= __pa(high_memory))
return 0;
- if (((pfn <= max_low_pfn_mapped) ||
- (pfn >= (1UL<<(32 - PAGE_SHIFT)) && pfn <= max_pfn_mapped)) &&
- ioremap_change_attr((unsigned long)__va(offset), size, flags) < 0) {
- free_memtype(offset, offset + size);
+ id_sz = (__pa(high_memory) < base + size) ?
+ __pa(high_memory) - base :
+ size;
+
+ if (ioremap_change_attr((unsigned long)__va(base), id_sz, flags) < 0) {
printk(KERN_INFO
- "%s:%d /dev/mem ioremap_change_attr failed %s for %Lx-%Lx\n",
+ "%s:%d ioremap_change_attr failed %s "
+ "for %Lx-%Lx\n",
current->comm, current->pid,
cattr_name(flags),
- offset, (unsigned long long)(offset + size));
+ base, (unsigned long long)(base + size));
+ return -EINVAL;
+ }
+ return 0;
+}
+
+/*
+ * Internal interface to reserve a range of physical memory with prot.
+ * Reserved non RAM regions only and after successful reserve_memtype,
+ * this func also keeps identity mapping (if any) in sync with this new prot.
+ */
+static int reserve_pfn_range(u64 paddr, unsigned long size, pgprot_t *vma_prot,
+ int strict_prot)
+{
+ int is_ram = 0;
+ int ret;
+ unsigned long want_flags = (pgprot_val(*vma_prot) & _PAGE_CACHE_MASK);
+ unsigned long flags = want_flags;
+
+ is_ram = pat_pagerange_is_ram(paddr, paddr + size);
+
+ /*
+ * reserve_pfn_range() doesn't support RAM pages. Maintain the current
+ * behavior with RAM pages by returning success.
+ */
+ if (is_ram != 0)
return 0;
+
+ ret = reserve_memtype(paddr, paddr + size, want_flags, &flags);
+ if (ret)
+ return ret;
+
+ if (flags != want_flags) {
+ if (strict_prot ||
+ !is_new_memtype_allowed(paddr, size, want_flags, flags)) {
+ free_memtype(paddr, paddr + size);
+ printk(KERN_ERR "%s:%d map pfn expected mapping type %s"
+ " for %Lx-%Lx, got %s\n",
+ current->comm, current->pid,
+ cattr_name(want_flags),
+ (unsigned long long)paddr,
+ (unsigned long long)(paddr + size),
+ cattr_name(flags));
+ return -EINVAL;
+ }
+ /*
+ * We allow returning different type than the one requested in
+ * non strict case.
+ */
+ *vma_prot = __pgprot((pgprot_val(*vma_prot) &
+ (~_PAGE_CACHE_MASK)) |
+ flags);
}
- *vma_prot = __pgprot((pgprot_val(*vma_prot) & ~_PAGE_CACHE_MASK) |
- flags);
- return 1;
+ if (kernel_map_sync_memtype(paddr, size, flags) < 0) {
+ free_memtype(paddr, paddr + size);
+ return -EINVAL;
+ }
+ return 0;
+}
+
+/*
+ * Internal interface to free a range of physical memory.
+ * Frees non RAM regions only.
+ */
+static void free_pfn_range(u64 paddr, unsigned long size)
+{
+ int is_ram;
+
+ is_ram = pat_pagerange_is_ram(paddr, paddr + size);
+ if (is_ram == 0)
+ free_memtype(paddr, paddr + size);
}
-void map_devmem(unsigned long pfn, unsigned long size, pgprot_t vma_prot)
+/*
+ * track_pfn_vma_copy is called when vma that is covering the pfnmap gets
+ * copied through copy_page_range().
+ *
+ * If the vma has a linear pfn mapping for the entire range, we get the prot
+ * from pte and reserve the entire vma range with single reserve_pfn_range call.
+ */
+int track_pfn_vma_copy(struct vm_area_struct *vma)
{
- u64 addr = (u64)pfn << PAGE_SHIFT;
- unsigned long flags;
- unsigned long want_flags = (pgprot_val(vma_prot) & _PAGE_CACHE_MASK);
+ resource_size_t paddr;
+ unsigned long prot;
+ unsigned long vma_size = vma->vm_end - vma->vm_start;
+ pgprot_t pgprot;
- reserve_memtype(addr, addr + size, want_flags, &flags);
- if (flags != want_flags) {
- printk(KERN_INFO
- "%s:%d /dev/mem expected mapping type %s for %Lx-%Lx, got %s\n",
- current->comm, current->pid,
- cattr_name(want_flags),
- addr, (unsigned long long)(addr + size),
- cattr_name(flags));
+ if (!pat_enabled)
+ return 0;
+
+ /*
+ * For now, only handle remap_pfn_range() vmas where
+ * is_linear_pfn_mapping() == TRUE. Handling of
+ * vm_insert_pfn() is TBD.
+ */
+ if (is_linear_pfn_mapping(vma)) {
+ /*
+ * reserve the whole chunk covered by vma. We need the
+ * starting address and protection from pte.
+ */
+ if (follow_phys(vma, vma->vm_start, 0, &prot, &paddr)) {
+ WARN_ON_ONCE(1);
+ return -EINVAL;
+ }
+ pgprot = __pgprot(prot);
+ return reserve_pfn_range(paddr, vma_size, &pgprot, 1);
}
+
+ return 0;
}
-void unmap_devmem(unsigned long pfn, unsigned long size, pgprot_t vma_prot)
+/*
+ * track_pfn_vma_new is called when a _new_ pfn mapping is being established
+ * for physical range indicated by pfn and size.
+ *
+ * prot is passed in as a parameter for the new mapping. If the vma has a
+ * linear pfn mapping for the entire range reserve the entire vma range with
+ * single reserve_pfn_range call.
+ */
+int track_pfn_vma_new(struct vm_area_struct *vma, pgprot_t *prot,
+ unsigned long pfn, unsigned long size)
{
- u64 addr = (u64)pfn << PAGE_SHIFT;
+ resource_size_t paddr;
+ unsigned long vma_size = vma->vm_end - vma->vm_start;
+
+ if (!pat_enabled)
+ return 0;
- free_memtype(addr, addr + size);
+ /*
+ * For now, only handle remap_pfn_range() vmas where
+ * is_linear_pfn_mapping() == TRUE. Handling of
+ * vm_insert_pfn() is TBD.
+ */
+ if (is_linear_pfn_mapping(vma)) {
+ /* reserve the whole chunk starting from vm_pgoff */
+ paddr = (resource_size_t)vma->vm_pgoff << PAGE_SHIFT;
+ return reserve_pfn_range(paddr, vma_size, prot, 0);
+ }
+
+ return 0;
}
+
+/*
+ * untrack_pfn_vma is called while unmapping a pfnmap for a region.
+ * untrack can be called for a specific region indicated by pfn and size or
+ * can be for the entire vma (in which case size can be zero).
+ */
+void untrack_pfn_vma(struct vm_area_struct *vma, unsigned long pfn,
+ unsigned long size)
+{
+ resource_size_t paddr;
+ unsigned long vma_size = vma->vm_end - vma->vm_start;
+
+ if (!pat_enabled)
+ return;
+
+ /*
+ * For now, only handle remap_pfn_range() vmas where
+ * is_linear_pfn_mapping() == TRUE. Handling of
+ * vm_insert_pfn() is TBD.
+ */
+ if (is_linear_pfn_mapping(vma)) {
+ /* free the whole chunk starting from vm_pgoff */
+ paddr = (resource_size_t)vma->vm_pgoff << PAGE_SHIFT;
+ free_pfn_range(paddr, vma_size);
+ return;
+ }
+}
+
+pgprot_t pgprot_writecombine(pgprot_t prot)
+{
+ if (pat_enabled)
+ return __pgprot(pgprot_val(prot) | _PAGE_CACHE_WC);
+ else
+ return pgprot_noncached(prot);
+}
+EXPORT_SYMBOL_GPL(pgprot_writecombine);
+
+#if defined(CONFIG_DEBUG_FS) && defined(CONFIG_X86_PAT)
+
+/* get Nth element of the linked list */
+static struct memtype *memtype_get_idx(loff_t pos)
+{
+ struct memtype *list_node, *print_entry;
+ int i = 1;
+
+ print_entry = kmalloc(sizeof(struct memtype), GFP_KERNEL);
+ if (!print_entry)
+ return NULL;
+
+ spin_lock(&memtype_lock);
+ list_for_each_entry(list_node, &memtype_list, nd) {
+ if (pos == i) {
+ *print_entry = *list_node;
+ spin_unlock(&memtype_lock);
+ return print_entry;
+ }
+ ++i;
+ }
+ spin_unlock(&memtype_lock);
+ kfree(print_entry);
+
+ return NULL;
+}
+
+static void *memtype_seq_start(struct seq_file *seq, loff_t *pos)
+{
+ if (*pos == 0) {
+ ++*pos;
+ seq_printf(seq, "PAT memtype list:\n");
+ }
+
+ return memtype_get_idx(*pos);
+}
+
+static void *memtype_seq_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+ ++*pos;
+ return memtype_get_idx(*pos);
+}
+
+static void memtype_seq_stop(struct seq_file *seq, void *v)
+{
+}
+
+static int memtype_seq_show(struct seq_file *seq, void *v)
+{
+ struct memtype *print_entry = (struct memtype *)v;
+
+ seq_printf(seq, "%s @ 0x%Lx-0x%Lx\n", cattr_name(print_entry->type),
+ print_entry->start, print_entry->end);
+ kfree(print_entry);
+
+ return 0;
+}
+
+static const struct seq_operations memtype_seq_ops = {
+ .start = memtype_seq_start,
+ .next = memtype_seq_next,
+ .stop = memtype_seq_stop,
+ .show = memtype_seq_show,
+};
+
+static int memtype_seq_open(struct inode *inode, struct file *file)
+{
+ return seq_open(file, &memtype_seq_ops);
+}
+
+static const struct file_operations memtype_fops = {
+ .open = memtype_seq_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = seq_release,
+};
+
+static int __init pat_memtype_list_init(void)
+{
+ debugfs_create_file("pat_memtype_list", S_IRUSR, arch_debugfs_dir,
+ NULL, &memtype_fops);
+ return 0;
+}
+
+late_initcall(pat_memtype_list_init);
+
+#endif /* CONFIG_DEBUG_FS && CONFIG_X86_PAT */