4b6a9316fa
Mark file system inode and similar slab caches subject to SLAB_MEM_SPREAD memory spreading. If a slab cache is marked SLAB_MEM_SPREAD, then anytime that a task that's in a cpuset with the 'memory_spread_slab' option enabled goes to allocate from such a slab cache, the allocations are spread evenly over all the memory nodes (task->mems_allowed) allowed to that task, instead of favoring allocation on the node local to the current cpu. The following inode and similar caches are marked SLAB_MEM_SPREAD: file cache ==== ===== fs/adfs/super.c adfs_inode_cache fs/affs/super.c affs_inode_cache fs/befs/linuxvfs.c befs_inode_cache fs/bfs/inode.c bfs_inode_cache fs/block_dev.c bdev_cache fs/cifs/cifsfs.c cifs_inode_cache fs/coda/inode.c coda_inode_cache fs/dquot.c dquot fs/efs/super.c efs_inode_cache fs/ext2/super.c ext2_inode_cache fs/ext2/xattr.c (fs/mbcache.c) ext2_xattr fs/ext3/super.c ext3_inode_cache fs/ext3/xattr.c (fs/mbcache.c) ext3_xattr fs/fat/cache.c fat_cache fs/fat/inode.c fat_inode_cache fs/freevxfs/vxfs_super.c vxfs_inode fs/hpfs/super.c hpfs_inode_cache fs/isofs/inode.c isofs_inode_cache fs/jffs/inode-v23.c jffs_fm fs/jffs2/super.c jffs2_i fs/jfs/super.c jfs_ip fs/minix/inode.c minix_inode_cache fs/ncpfs/inode.c ncp_inode_cache fs/nfs/direct.c nfs_direct_cache fs/nfs/inode.c nfs_inode_cache fs/ntfs/super.c ntfs_big_inode_cache_name fs/ntfs/super.c ntfs_inode_cache fs/ocfs2/dlm/dlmfs.c dlmfs_inode_cache fs/ocfs2/super.c ocfs2_inode_cache fs/proc/inode.c proc_inode_cache fs/qnx4/inode.c qnx4_inode_cache fs/reiserfs/super.c reiser_inode_cache fs/romfs/inode.c romfs_inode_cache fs/smbfs/inode.c smb_inode_cache fs/sysv/inode.c sysv_inode_cache fs/udf/super.c udf_inode_cache fs/ufs/super.c ufs_inode_cache net/socket.c sock_inode_cache net/sunrpc/rpc_pipe.c rpc_inode_cache The choice of which slab caches to so mark was quite simple. I marked those already marked SLAB_RECLAIM_ACCOUNT, except for fs/xfs, dentry_cache, inode_cache, and buffer_head, which were marked in a previous patch. Even though SLAB_RECLAIM_ACCOUNT is for a different purpose, it marks the same potentially large file system i/o related slab caches as we need for memory spreading. Given that the rule now becomes "wherever you would have used a SLAB_RECLAIM_ACCOUNT slab cache flag before (usually the inode cache), use the SLAB_MEM_SPREAD flag too", this should be easy enough to maintain. Future file system writers will just copy one of the existing file system slab cache setups and tend to get it right without thinking. Signed-off-by: Paul Jackson <pj@sgi.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
508 lines
12 KiB
C
508 lines
12 KiB
C
/*
|
|
* linux/fs/adfs/super.c
|
|
*
|
|
* Copyright (C) 1997-1999 Russell King
|
|
*
|
|
* This program is free software; you can redistribute it and/or modify
|
|
* it under the terms of the GNU General Public License version 2 as
|
|
* published by the Free Software Foundation.
|
|
*/
|
|
#include <linux/module.h>
|
|
#include <linux/errno.h>
|
|
#include <linux/fs.h>
|
|
#include <linux/adfs_fs.h>
|
|
#include <linux/slab.h>
|
|
#include <linux/time.h>
|
|
#include <linux/stat.h>
|
|
#include <linux/string.h>
|
|
#include <linux/init.h>
|
|
#include <linux/buffer_head.h>
|
|
#include <linux/vfs.h>
|
|
#include <linux/parser.h>
|
|
#include <linux/bitops.h>
|
|
|
|
#include <asm/uaccess.h>
|
|
#include <asm/system.h>
|
|
|
|
#include <stdarg.h>
|
|
|
|
#include "adfs.h"
|
|
#include "dir_f.h"
|
|
#include "dir_fplus.h"
|
|
|
|
void __adfs_error(struct super_block *sb, const char *function, const char *fmt, ...)
|
|
{
|
|
char error_buf[128];
|
|
va_list args;
|
|
|
|
va_start(args, fmt);
|
|
vsprintf(error_buf, fmt, args);
|
|
va_end(args);
|
|
|
|
printk(KERN_CRIT "ADFS-fs error (device %s)%s%s: %s\n",
|
|
sb->s_id, function ? ": " : "",
|
|
function ? function : "", error_buf);
|
|
}
|
|
|
|
static int adfs_checkdiscrecord(struct adfs_discrecord *dr)
|
|
{
|
|
int i;
|
|
|
|
/* sector size must be 256, 512 or 1024 bytes */
|
|
if (dr->log2secsize != 8 &&
|
|
dr->log2secsize != 9 &&
|
|
dr->log2secsize != 10)
|
|
return 1;
|
|
|
|
/* idlen must be at least log2secsize + 3 */
|
|
if (dr->idlen < dr->log2secsize + 3)
|
|
return 1;
|
|
|
|
/* we cannot have such a large disc that we
|
|
* are unable to represent sector offsets in
|
|
* 32 bits. This works out at 2.0 TB.
|
|
*/
|
|
if (le32_to_cpu(dr->disc_size_high) >> dr->log2secsize)
|
|
return 1;
|
|
|
|
/* idlen must be no greater than 19 v2 [1.0] */
|
|
if (dr->idlen > 19)
|
|
return 1;
|
|
|
|
/* reserved bytes should be zero */
|
|
for (i = 0; i < sizeof(dr->unused52); i++)
|
|
if (dr->unused52[i] != 0)
|
|
return 1;
|
|
|
|
return 0;
|
|
}
|
|
|
|
static unsigned char adfs_calczonecheck(struct super_block *sb, unsigned char *map)
|
|
{
|
|
unsigned int v0, v1, v2, v3;
|
|
int i;
|
|
|
|
v0 = v1 = v2 = v3 = 0;
|
|
for (i = sb->s_blocksize - 4; i; i -= 4) {
|
|
v0 += map[i] + (v3 >> 8);
|
|
v3 &= 0xff;
|
|
v1 += map[i + 1] + (v0 >> 8);
|
|
v0 &= 0xff;
|
|
v2 += map[i + 2] + (v1 >> 8);
|
|
v1 &= 0xff;
|
|
v3 += map[i + 3] + (v2 >> 8);
|
|
v2 &= 0xff;
|
|
}
|
|
v0 += v3 >> 8;
|
|
v1 += map[1] + (v0 >> 8);
|
|
v2 += map[2] + (v1 >> 8);
|
|
v3 += map[3] + (v2 >> 8);
|
|
|
|
return v0 ^ v1 ^ v2 ^ v3;
|
|
}
|
|
|
|
static int adfs_checkmap(struct super_block *sb, struct adfs_discmap *dm)
|
|
{
|
|
unsigned char crosscheck = 0, zonecheck = 1;
|
|
int i;
|
|
|
|
for (i = 0; i < ADFS_SB(sb)->s_map_size; i++) {
|
|
unsigned char *map;
|
|
|
|
map = dm[i].dm_bh->b_data;
|
|
|
|
if (adfs_calczonecheck(sb, map) != map[0]) {
|
|
adfs_error(sb, "zone %d fails zonecheck", i);
|
|
zonecheck = 0;
|
|
}
|
|
crosscheck ^= map[3];
|
|
}
|
|
if (crosscheck != 0xff)
|
|
adfs_error(sb, "crosscheck != 0xff");
|
|
return crosscheck == 0xff && zonecheck;
|
|
}
|
|
|
|
static void adfs_put_super(struct super_block *sb)
|
|
{
|
|
int i;
|
|
struct adfs_sb_info *asb = ADFS_SB(sb);
|
|
|
|
for (i = 0; i < asb->s_map_size; i++)
|
|
brelse(asb->s_map[i].dm_bh);
|
|
kfree(asb->s_map);
|
|
kfree(asb);
|
|
sb->s_fs_info = NULL;
|
|
}
|
|
|
|
enum {Opt_uid, Opt_gid, Opt_ownmask, Opt_othmask, Opt_err};
|
|
|
|
static match_table_t tokens = {
|
|
{Opt_uid, "uid=%u"},
|
|
{Opt_gid, "gid=%u"},
|
|
{Opt_ownmask, "ownmask=%o"},
|
|
{Opt_othmask, "othmask=%o"},
|
|
{Opt_err, NULL}
|
|
};
|
|
|
|
static int parse_options(struct super_block *sb, char *options)
|
|
{
|
|
char *p;
|
|
struct adfs_sb_info *asb = ADFS_SB(sb);
|
|
int option;
|
|
|
|
if (!options)
|
|
return 0;
|
|
|
|
while ((p = strsep(&options, ",")) != NULL) {
|
|
substring_t args[MAX_OPT_ARGS];
|
|
int token;
|
|
if (!*p)
|
|
continue;
|
|
|
|
token = match_token(p, tokens, args);
|
|
switch (token) {
|
|
case Opt_uid:
|
|
if (match_int(args, &option))
|
|
return -EINVAL;
|
|
asb->s_uid = option;
|
|
break;
|
|
case Opt_gid:
|
|
if (match_int(args, &option))
|
|
return -EINVAL;
|
|
asb->s_gid = option;
|
|
break;
|
|
case Opt_ownmask:
|
|
if (match_octal(args, &option))
|
|
return -EINVAL;
|
|
asb->s_owner_mask = option;
|
|
break;
|
|
case Opt_othmask:
|
|
if (match_octal(args, &option))
|
|
return -EINVAL;
|
|
asb->s_other_mask = option;
|
|
break;
|
|
default:
|
|
printk("ADFS-fs: unrecognised mount option \"%s\" "
|
|
"or missing value\n", p);
|
|
return -EINVAL;
|
|
}
|
|
}
|
|
return 0;
|
|
}
|
|
|
|
static int adfs_remount(struct super_block *sb, int *flags, char *data)
|
|
{
|
|
*flags |= MS_NODIRATIME;
|
|
return parse_options(sb, data);
|
|
}
|
|
|
|
static int adfs_statfs(struct super_block *sb, struct kstatfs *buf)
|
|
{
|
|
struct adfs_sb_info *asb = ADFS_SB(sb);
|
|
|
|
buf->f_type = ADFS_SUPER_MAGIC;
|
|
buf->f_namelen = asb->s_namelen;
|
|
buf->f_bsize = sb->s_blocksize;
|
|
buf->f_blocks = asb->s_size;
|
|
buf->f_files = asb->s_ids_per_zone * asb->s_map_size;
|
|
buf->f_bavail =
|
|
buf->f_bfree = adfs_map_free(sb);
|
|
buf->f_ffree = (long)(buf->f_bfree * buf->f_files) / (long)buf->f_blocks;
|
|
|
|
return 0;
|
|
}
|
|
|
|
static kmem_cache_t *adfs_inode_cachep;
|
|
|
|
static struct inode *adfs_alloc_inode(struct super_block *sb)
|
|
{
|
|
struct adfs_inode_info *ei;
|
|
ei = (struct adfs_inode_info *)kmem_cache_alloc(adfs_inode_cachep, SLAB_KERNEL);
|
|
if (!ei)
|
|
return NULL;
|
|
return &ei->vfs_inode;
|
|
}
|
|
|
|
static void adfs_destroy_inode(struct inode *inode)
|
|
{
|
|
kmem_cache_free(adfs_inode_cachep, ADFS_I(inode));
|
|
}
|
|
|
|
static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags)
|
|
{
|
|
struct adfs_inode_info *ei = (struct adfs_inode_info *) foo;
|
|
|
|
if ((flags & (SLAB_CTOR_VERIFY|SLAB_CTOR_CONSTRUCTOR)) ==
|
|
SLAB_CTOR_CONSTRUCTOR)
|
|
inode_init_once(&ei->vfs_inode);
|
|
}
|
|
|
|
static int init_inodecache(void)
|
|
{
|
|
adfs_inode_cachep = kmem_cache_create("adfs_inode_cache",
|
|
sizeof(struct adfs_inode_info),
|
|
0, SLAB_RECLAIM_ACCOUNT|SLAB_MEM_SPREAD,
|
|
init_once, NULL);
|
|
if (adfs_inode_cachep == NULL)
|
|
return -ENOMEM;
|
|
return 0;
|
|
}
|
|
|
|
static void destroy_inodecache(void)
|
|
{
|
|
if (kmem_cache_destroy(adfs_inode_cachep))
|
|
printk(KERN_INFO "adfs_inode_cache: not all structures were freed\n");
|
|
}
|
|
|
|
static struct super_operations adfs_sops = {
|
|
.alloc_inode = adfs_alloc_inode,
|
|
.destroy_inode = adfs_destroy_inode,
|
|
.write_inode = adfs_write_inode,
|
|
.put_super = adfs_put_super,
|
|
.statfs = adfs_statfs,
|
|
.remount_fs = adfs_remount,
|
|
};
|
|
|
|
static struct adfs_discmap *adfs_read_map(struct super_block *sb, struct adfs_discrecord *dr)
|
|
{
|
|
struct adfs_discmap *dm;
|
|
unsigned int map_addr, zone_size, nzones;
|
|
int i, zone;
|
|
struct adfs_sb_info *asb = ADFS_SB(sb);
|
|
|
|
nzones = asb->s_map_size;
|
|
zone_size = (8 << dr->log2secsize) - le16_to_cpu(dr->zone_spare);
|
|
map_addr = (nzones >> 1) * zone_size -
|
|
((nzones > 1) ? ADFS_DR_SIZE_BITS : 0);
|
|
map_addr = signed_asl(map_addr, asb->s_map2blk);
|
|
|
|
asb->s_ids_per_zone = zone_size / (asb->s_idlen + 1);
|
|
|
|
dm = kmalloc(nzones * sizeof(*dm), GFP_KERNEL);
|
|
if (dm == NULL) {
|
|
adfs_error(sb, "not enough memory");
|
|
return NULL;
|
|
}
|
|
|
|
for (zone = 0; zone < nzones; zone++, map_addr++) {
|
|
dm[zone].dm_startbit = 0;
|
|
dm[zone].dm_endbit = zone_size;
|
|
dm[zone].dm_startblk = zone * zone_size - ADFS_DR_SIZE_BITS;
|
|
dm[zone].dm_bh = sb_bread(sb, map_addr);
|
|
|
|
if (!dm[zone].dm_bh) {
|
|
adfs_error(sb, "unable to read map");
|
|
goto error_free;
|
|
}
|
|
}
|
|
|
|
/* adjust the limits for the first and last map zones */
|
|
i = zone - 1;
|
|
dm[0].dm_startblk = 0;
|
|
dm[0].dm_startbit = ADFS_DR_SIZE_BITS;
|
|
dm[i].dm_endbit = (le32_to_cpu(dr->disc_size_high) << (32 - dr->log2bpmb)) +
|
|
(le32_to_cpu(dr->disc_size) >> dr->log2bpmb) +
|
|
(ADFS_DR_SIZE_BITS - i * zone_size);
|
|
|
|
if (adfs_checkmap(sb, dm))
|
|
return dm;
|
|
|
|
adfs_error(sb, NULL, "map corrupted");
|
|
|
|
error_free:
|
|
while (--zone >= 0)
|
|
brelse(dm[zone].dm_bh);
|
|
|
|
kfree(dm);
|
|
return NULL;
|
|
}
|
|
|
|
static inline unsigned long adfs_discsize(struct adfs_discrecord *dr, int block_bits)
|
|
{
|
|
unsigned long discsize;
|
|
|
|
discsize = le32_to_cpu(dr->disc_size_high) << (32 - block_bits);
|
|
discsize |= le32_to_cpu(dr->disc_size) >> block_bits;
|
|
|
|
return discsize;
|
|
}
|
|
|
|
static int adfs_fill_super(struct super_block *sb, void *data, int silent)
|
|
{
|
|
struct adfs_discrecord *dr;
|
|
struct buffer_head *bh;
|
|
struct object_info root_obj;
|
|
unsigned char *b_data;
|
|
struct adfs_sb_info *asb;
|
|
struct inode *root;
|
|
|
|
sb->s_flags |= MS_NODIRATIME;
|
|
|
|
asb = kmalloc(sizeof(*asb), GFP_KERNEL);
|
|
if (!asb)
|
|
return -ENOMEM;
|
|
sb->s_fs_info = asb;
|
|
memset(asb, 0, sizeof(*asb));
|
|
|
|
/* set default options */
|
|
asb->s_uid = 0;
|
|
asb->s_gid = 0;
|
|
asb->s_owner_mask = S_IRWXU;
|
|
asb->s_other_mask = S_IRWXG | S_IRWXO;
|
|
|
|
if (parse_options(sb, data))
|
|
goto error;
|
|
|
|
sb_set_blocksize(sb, BLOCK_SIZE);
|
|
if (!(bh = sb_bread(sb, ADFS_DISCRECORD / BLOCK_SIZE))) {
|
|
adfs_error(sb, "unable to read superblock");
|
|
goto error;
|
|
}
|
|
|
|
b_data = bh->b_data + (ADFS_DISCRECORD % BLOCK_SIZE);
|
|
|
|
if (adfs_checkbblk(b_data)) {
|
|
if (!silent)
|
|
printk("VFS: Can't find an adfs filesystem on dev "
|
|
"%s.\n", sb->s_id);
|
|
goto error_free_bh;
|
|
}
|
|
|
|
dr = (struct adfs_discrecord *)(b_data + ADFS_DR_OFFSET);
|
|
|
|
/*
|
|
* Do some sanity checks on the ADFS disc record
|
|
*/
|
|
if (adfs_checkdiscrecord(dr)) {
|
|
if (!silent)
|
|
printk("VPS: Can't find an adfs filesystem on dev "
|
|
"%s.\n", sb->s_id);
|
|
goto error_free_bh;
|
|
}
|
|
|
|
brelse(bh);
|
|
if (sb_set_blocksize(sb, 1 << dr->log2secsize)) {
|
|
bh = sb_bread(sb, ADFS_DISCRECORD / sb->s_blocksize);
|
|
if (!bh) {
|
|
adfs_error(sb, "couldn't read superblock on "
|
|
"2nd try.");
|
|
goto error;
|
|
}
|
|
b_data = bh->b_data + (ADFS_DISCRECORD % sb->s_blocksize);
|
|
if (adfs_checkbblk(b_data)) {
|
|
adfs_error(sb, "disc record mismatch, very weird!");
|
|
goto error_free_bh;
|
|
}
|
|
dr = (struct adfs_discrecord *)(b_data + ADFS_DR_OFFSET);
|
|
} else {
|
|
if (!silent)
|
|
printk(KERN_ERR "VFS: Unsupported blocksize on dev "
|
|
"%s.\n", sb->s_id);
|
|
goto error;
|
|
}
|
|
|
|
/*
|
|
* blocksize on this device should now be set to the ADFS log2secsize
|
|
*/
|
|
|
|
sb->s_magic = ADFS_SUPER_MAGIC;
|
|
asb->s_idlen = dr->idlen;
|
|
asb->s_map_size = dr->nzones | (dr->nzones_high << 8);
|
|
asb->s_map2blk = dr->log2bpmb - dr->log2secsize;
|
|
asb->s_size = adfs_discsize(dr, sb->s_blocksize_bits);
|
|
asb->s_version = dr->format_version;
|
|
asb->s_log2sharesize = dr->log2sharesize;
|
|
|
|
asb->s_map = adfs_read_map(sb, dr);
|
|
if (!asb->s_map)
|
|
goto error_free_bh;
|
|
|
|
brelse(bh);
|
|
|
|
/*
|
|
* set up enough so that we can read an inode
|
|
*/
|
|
sb->s_op = &adfs_sops;
|
|
|
|
dr = (struct adfs_discrecord *)(asb->s_map[0].dm_bh->b_data + 4);
|
|
|
|
root_obj.parent_id = root_obj.file_id = le32_to_cpu(dr->root);
|
|
root_obj.name_len = 0;
|
|
root_obj.loadaddr = 0;
|
|
root_obj.execaddr = 0;
|
|
root_obj.size = ADFS_NEWDIR_SIZE;
|
|
root_obj.attr = ADFS_NDA_DIRECTORY | ADFS_NDA_OWNER_READ |
|
|
ADFS_NDA_OWNER_WRITE | ADFS_NDA_PUBLIC_READ;
|
|
|
|
/*
|
|
* If this is a F+ disk with variable length directories,
|
|
* get the root_size from the disc record.
|
|
*/
|
|
if (asb->s_version) {
|
|
root_obj.size = le32_to_cpu(dr->root_size);
|
|
asb->s_dir = &adfs_fplus_dir_ops;
|
|
asb->s_namelen = ADFS_FPLUS_NAME_LEN;
|
|
} else {
|
|
asb->s_dir = &adfs_f_dir_ops;
|
|
asb->s_namelen = ADFS_F_NAME_LEN;
|
|
}
|
|
|
|
root = adfs_iget(sb, &root_obj);
|
|
sb->s_root = d_alloc_root(root);
|
|
if (!sb->s_root) {
|
|
int i;
|
|
iput(root);
|
|
for (i = 0; i < asb->s_map_size; i++)
|
|
brelse(asb->s_map[i].dm_bh);
|
|
kfree(asb->s_map);
|
|
adfs_error(sb, "get root inode failed\n");
|
|
goto error;
|
|
} else
|
|
sb->s_root->d_op = &adfs_dentry_operations;
|
|
return 0;
|
|
|
|
error_free_bh:
|
|
brelse(bh);
|
|
error:
|
|
sb->s_fs_info = NULL;
|
|
kfree(asb);
|
|
return -EINVAL;
|
|
}
|
|
|
|
static struct super_block *adfs_get_sb(struct file_system_type *fs_type,
|
|
int flags, const char *dev_name, void *data)
|
|
{
|
|
return get_sb_bdev(fs_type, flags, dev_name, data, adfs_fill_super);
|
|
}
|
|
|
|
static struct file_system_type adfs_fs_type = {
|
|
.owner = THIS_MODULE,
|
|
.name = "adfs",
|
|
.get_sb = adfs_get_sb,
|
|
.kill_sb = kill_block_super,
|
|
.fs_flags = FS_REQUIRES_DEV,
|
|
};
|
|
|
|
static int __init init_adfs_fs(void)
|
|
{
|
|
int err = init_inodecache();
|
|
if (err)
|
|
goto out1;
|
|
err = register_filesystem(&adfs_fs_type);
|
|
if (err)
|
|
goto out;
|
|
return 0;
|
|
out:
|
|
destroy_inodecache();
|
|
out1:
|
|
return err;
|
|
}
|
|
|
|
static void __exit exit_adfs_fs(void)
|
|
{
|
|
unregister_filesystem(&adfs_fs_type);
|
|
destroy_inodecache();
|
|
}
|
|
|
|
module_init(init_adfs_fs)
|
|
module_exit(exit_adfs_fs)
|