#include <linux/swapops.h>
#include <linux/mempolicy.h>
#include <linux/namei.h>
+#include <linux/ctype.h>
+#include <linux/migrate.h>
+
#include <asm/uaccess.h>
#include <asm/div64.h>
#include <asm/pgtable.h>
} while (next);
}
-static void shmem_truncate(struct inode *inode)
+static void shmem_truncate_range(struct inode *inode, loff_t start, loff_t end)
{
struct shmem_inode_info *info = SHMEM_I(inode);
unsigned long idx;
long nr_swaps_freed = 0;
int offset;
int freed;
+ int punch_hole = 0;
inode->i_ctime = inode->i_mtime = CURRENT_TIME;
- idx = (inode->i_size + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+ idx = (start + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
if (idx >= info->next_index)
return;
spin_lock(&info->lock);
info->flags |= SHMEM_TRUNCATE;
- limit = info->next_index;
- info->next_index = idx;
+ if (likely(end == (loff_t) -1)) {
+ limit = info->next_index;
+ info->next_index = idx;
+ } else {
+ limit = (end + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
+ if (limit > info->next_index)
+ limit = info->next_index;
+ punch_hole = 1;
+ }
+
topdir = info->i_indirect;
- if (topdir && idx <= SHMEM_NR_DIRECT) {
+ if (topdir && idx <= SHMEM_NR_DIRECT && !punch_hole) {
info->i_indirect = NULL;
nr_pages_to_free++;
list_add(&topdir->lru, &pages_to_free);
set_page_private(subdir, page_private(subdir) - freed);
if (offset)
spin_unlock(&info->lock);
- BUG_ON(page_private(subdir) > offset);
+ if (!punch_hole)
+ BUG_ON(page_private(subdir) > offset);
}
if (offset)
offset = 0;
- else if (subdir) {
+ else if (subdir && !page_private(subdir)) {
dir[diroff] = NULL;
nr_pages_to_free++;
list_add(&subdir->lru, &pages_to_free);
* Also, though shmem_getpage checks i_size before adding to
* cache, no recheck after: so fix the narrow window there too.
*/
- truncate_inode_pages(inode->i_mapping, inode->i_size);
+ truncate_inode_pages_range(inode->i_mapping, start, end);
}
spin_lock(&info->lock);
}
}
+static void shmem_truncate(struct inode *inode)
+{
+ shmem_truncate_range(inode, inode->i_size, (loff_t)-1);
+}
+
static int shmem_notify_change(struct dentry *dentry, struct iattr *attr)
{
struct inode *inode = dentry->d_inode;
}
#ifdef CONFIG_NUMA
+static inline int shmem_parse_mpol(char *value, int *policy, nodemask_t *policy_nodes)
+{
+ char *nodelist = strchr(value, ':');
+ int err = 1;
+
+ if (nodelist) {
+ /* NUL-terminate policy string */
+ *nodelist++ = '\0';
+ if (nodelist_parse(nodelist, *policy_nodes))
+ goto out;
+ }
+ if (!strcmp(value, "default")) {
+ *policy = MPOL_DEFAULT;
+ /* Don't allow a nodelist */
+ if (!nodelist)
+ err = 0;
+ } else if (!strcmp(value, "prefer")) {
+ *policy = MPOL_PREFERRED;
+ /* Insist on a nodelist of one node only */
+ if (nodelist) {
+ char *rest = nodelist;
+ while (isdigit(*rest))
+ rest++;
+ if (!*rest)
+ err = 0;
+ }
+ } else if (!strcmp(value, "bind")) {
+ *policy = MPOL_BIND;
+ /* Insist on a nodelist */
+ if (nodelist)
+ err = 0;
+ } else if (!strcmp(value, "interleave")) {
+ *policy = MPOL_INTERLEAVE;
+ /* Default to nodes online if no nodelist */
+ if (!nodelist)
+ *policy_nodes = node_online_map;
+ err = 0;
+ }
+out:
+ /* Restore string for error message */
+ if (nodelist)
+ *--nodelist = ':';
+ return err;
+}
+
static struct page *shmem_swapin_async(struct shared_policy *p,
swp_entry_t entry, unsigned long idx)
{
return page;
}
#else
+static inline int shmem_parse_mpol(char *value, int *policy, nodemask_t *policy_nodes)
+{
+ return 1;
+}
+
static inline struct page *
shmem_swapin(struct shmem_inode_info *info,swp_entry_t entry,unsigned long idx)
{
return retval;
}
-static int shmem_mmap(struct file *file, struct vm_area_struct *vma)
+int shmem_mmap(struct file *file, struct vm_area_struct *vma)
{
file_accessed(file);
vma->vm_ops = &shmem_vm_ops;
case S_IFREG:
inode->i_op = &shmem_inode_operations;
inode->i_fop = &shmem_file_operations;
- mpol_shared_policy_init(&info->policy);
+ mpol_shared_policy_init(&info->policy, sbinfo->policy,
+ &sbinfo->policy_nodes);
break;
case S_IFDIR:
inode->i_nlink++;
* Must not load anything in the rbtree,
* mpol_free_shared_policy will not be called.
*/
- mpol_shared_policy_init(&info->policy);
+ mpol_shared_policy_init(&info->policy, MPOL_DEFAULT,
+ NULL);
break;
}
} else if (sbinfo->max_inodes) {
if (!access_ok(VERIFY_READ, buf, count))
return -EFAULT;
- down(&inode->i_sem);
+ mutex_lock(&inode->i_mutex);
pos = *ppos;
written = 0;
if (written)
err = written;
out:
- up(&inode->i_sem);
+ mutex_unlock(&inode->i_mutex);
return err;
}
/*
* We must evaluate after, since reads (unlike writes)
- * are called without i_sem protection against truncate
+ * are called without i_mutex protection against truncate
*/
nr = PAGE_CACHE_SIZE;
i_size = i_size_read(inode);
return desc.error;
}
-static int shmem_statfs(struct super_block *sb, struct kstatfs *buf)
+static int shmem_statfs(struct dentry *dentry, struct kstatfs *buf)
{
- struct shmem_sb_info *sbinfo = SHMEM_SB(sb);
+ struct shmem_sb_info *sbinfo = SHMEM_SB(dentry->d_sb);
buf->f_type = TMPFS_MAGIC;
buf->f_bsize = PAGE_CACHE_SIZE;
if (!simple_empty(dentry))
return -ENOTEMPTY;
+ dentry->d_inode->i_nlink--;
dir->i_nlink--;
return shmem_unlink(dir, dentry);
}
.put_link = shmem_put_link,
};
-static int shmem_parse_options(char *options, int *mode, uid_t *uid, gid_t *gid, unsigned long *blocks, unsigned long *inodes)
+static int shmem_parse_options(char *options, int *mode, uid_t *uid,
+ gid_t *gid, unsigned long *blocks, unsigned long *inodes,
+ int *policy, nodemask_t *policy_nodes)
{
char *this_char, *value, *rest;
- while ((this_char = strsep(&options, ",")) != NULL) {
+ while (options != NULL) {
+ this_char = options;
+ for (;;) {
+ /*
+ * NUL-terminate this option: unfortunately,
+ * mount options form a comma-separated list,
+ * but mpol's nodelist may also contain commas.
+ */
+ options = strchr(options, ',');
+ if (options == NULL)
+ break;
+ options++;
+ if (!isdigit(*options)) {
+ options[-1] = '\0';
+ break;
+ }
+ }
if (!*this_char)
continue;
if ((value = strchr(this_char,'=')) != NULL) {
*gid = simple_strtoul(value,&rest,0);
if (*rest)
goto bad_val;
+ } else if (!strcmp(this_char,"mpol")) {
+ if (shmem_parse_mpol(value,policy,policy_nodes))
+ goto bad_val;
} else {
printk(KERN_ERR "tmpfs: Bad mount option %s\n",
this_char);
struct shmem_sb_info *sbinfo = SHMEM_SB(sb);
unsigned long max_blocks = sbinfo->max_blocks;
unsigned long max_inodes = sbinfo->max_inodes;
+ int policy = sbinfo->policy;
+ nodemask_t policy_nodes = sbinfo->policy_nodes;
unsigned long blocks;
unsigned long inodes;
int error = -EINVAL;
- if (shmem_parse_options(data, NULL, NULL, NULL,
- &max_blocks, &max_inodes))
+ if (shmem_parse_options(data, NULL, NULL, NULL, &max_blocks,
+ &max_inodes, &policy, &policy_nodes))
return error;
spin_lock(&sbinfo->stat_lock);
sbinfo->free_blocks = max_blocks - blocks;
sbinfo->max_inodes = max_inodes;
sbinfo->free_inodes = max_inodes - inodes;
+ sbinfo->policy = policy;
+ sbinfo->policy_nodes = policy_nodes;
out:
spin_unlock(&sbinfo->stat_lock);
return error;
struct shmem_sb_info *sbinfo;
unsigned long blocks = 0;
unsigned long inodes = 0;
+ int policy = MPOL_DEFAULT;
+ nodemask_t policy_nodes = node_online_map;
#ifdef CONFIG_TMPFS
/*
inodes = totalram_pages - totalhigh_pages;
if (inodes > blocks)
inodes = blocks;
- if (shmem_parse_options(data, &mode, &uid, &gid,
- &blocks, &inodes))
+ if (shmem_parse_options(data, &mode, &uid, &gid, &blocks,
+ &inodes, &policy, &policy_nodes))
return -EINVAL;
}
#else
sbinfo->free_blocks = blocks;
sbinfo->max_inodes = inodes;
sbinfo->free_inodes = inodes;
+ sbinfo->policy = policy;
+ sbinfo->policy_nodes = policy_nodes;
sb->s_fs_info = sbinfo;
sb->s_maxbytes = SHMEM_MAX_BYTES;
sb->s_blocksize_bits = PAGE_CACHE_SHIFT;
sb->s_magic = TMPFS_MAGIC;
sb->s_op = &shmem_ops;
+ sb->s_time_gran = 1;
inode = shmem_get_inode(sb, S_IFDIR | mode, 0);
if (!inode)
return err;
}
-static kmem_cache_t *shmem_inode_cachep;
+static struct kmem_cache *shmem_inode_cachep;
static struct inode *shmem_alloc_inode(struct super_block *sb)
{
kmem_cache_free(shmem_inode_cachep, SHMEM_I(inode));
}
-static void init_once(void *foo, kmem_cache_t *cachep, unsigned long flags)
+static void init_once(void *foo, struct kmem_cache *cachep,
+ unsigned long flags)
{
struct shmem_inode_info *p = (struct shmem_inode_info *) foo;
.prepare_write = shmem_prepare_write,
.commit_write = simple_commit_write,
#endif
+ .migratepage = migrate_page,
};
static struct file_operations shmem_file_operations = {
static struct inode_operations shmem_inode_operations = {
.truncate = shmem_truncate,
.setattr = shmem_notify_change,
+ .truncate_range = shmem_truncate_range,
};
static struct inode_operations shmem_dir_inode_operations = {
};
-static struct super_block *shmem_get_sb(struct file_system_type *fs_type,
- int flags, const char *dev_name, void *data)
+static int shmem_get_sb(struct file_system_type *fs_type,
+ int flags, const char *dev_name, void *data, struct vfsmount *mnt)
{
- return get_sb_nodev(fs_type, flags, data, shmem_fill_super);
+ return get_sb_nodev(fs_type, flags, data, shmem_fill_super, mnt);
}
static struct file_system_type tmpfs_fs_type = {