xv6/fs.c
2007-08-22 06:20:10 +00:00

642 lines
14 KiB
C

// File system implementation.
//
// Four layers:
// + Blocks: allocator for raw disk blocks.
// + Files: inode allocator, reading, writing, metadata.
// + Directories: inode with special contents (list of other inodes!)
// + Names: paths like /usr/rtm/xv6/fs.c for convenient naming.
//
// Disk layout is: superblock, inodes, disk bitmap, data blocks.
//
// This file contains the low-level file system manipulation
// routines. The (higher-level) system call implementations
// are in sysfile.c.
#include "types.h"
#include "stat.h"
#include "param.h"
#include "x86.h"
#include "mmu.h"
#include "proc.h"
#include "defs.h"
#include "spinlock.h"
#include "buf.h"
#include "fs.h"
#include "fsvar.h"
#include "dev.h"
#define min(a, b) ((a) < (b) ? (a) : (b))
static void itrunc(struct inode*);
// Blocks.
// Allocate a disk block.
static uint
balloc(uint dev)
{
int b, bi, m, ninodes, size;
struct buf *bp;
struct superblock *sb;
bp = bread(dev, 1);
sb = (struct superblock*) bp->data;
size = sb->size;
ninodes = sb->ninodes;
for(b = 0; b < size; b++) {
if(b % BPB == 0) {
brelse(bp);
bp = bread(dev, BBLOCK(b, ninodes));
}
bi = b % BPB;
m = 0x1 << (bi % 8);
if((bp->data[bi/8] & m) == 0) { // is block free?
bp->data[bi/8] |= 0x1 << (bi % 8);
bwrite(bp); // mark it allocated on disk
brelse(bp);
return b;
}
}
panic("balloc: out of blocks");
}
// Free a disk block.
static void
bfree(int dev, uint b)
{
struct buf *bp;
struct superblock *sb;
int bi, m, ninodes;
bp = bread(dev, 1);
sb = (struct superblock*) bp->data;
ninodes = sb->ninodes;
brelse(bp);
bp = bread(dev, b);
memset(bp->data, 0, BSIZE);
bwrite(bp);
brelse(bp);
bp = bread(dev, BBLOCK(b, ninodes));
bi = b % BPB;
m = 0x1 << (bi % 8);
bp->data[bi/8] &= ~m;
bwrite(bp); // mark it free on disk
brelse(bp);
}
// Inodes
//
// The inodes are laid out sequentially on disk immediately after
// the superblock. The kernel keeps a cache of the in-use
// on-disk structures to provide a place for synchronizing access
// to inodes shared between multiple processes.
//
// ip->ref counts the number of references to this
// inode; references are typically kept in struct file and in cp->cwd.
// When ip->ref falls to zero, the inode is no longer cached.
// It is an error to use an inode without holding a reference to it.
//
// Inodes can be marked busy, just like bufs, meaning
// that some process has exclusive use of the inode.
// Processes are only allowed to read and write inode
// metadata and contents when holding the inode's lock.
// Because inodes locks are held during disk accesses,
// they are implemented using a flag, as in the buffer cache,
// not using spin locks. Callers are responsible for locking
// inodes before passing them to routines in this file; leaving
// this responsibility with the caller makes it possible for them
// to create arbitrarily-sized atomic operations.
//
// To give maximum control over locking to the callers,
// the routines in this file that return inode pointers
// return pointers to *unlocked* inodes. It is the callers'
// responsibility to lock them before using them.
struct {
struct spinlock lock;
struct inode inode[NINODE];
} icache;
void
iinit(void)
{
initlock(&icache.lock, "icache.lock");
}
// Find the inode with number inum on device dev
// and return the in-memory copy. h
static struct uinode*
iget(uint dev, uint inum)
{
struct inode *ip, *empty;
acquire(&icache.lock);
// Try for cached inode.
empty = 0;
for(ip = &icache.inode[0]; ip < &icache.inode[NINODE]; ip++){
if(ip->ref > 0 && ip->dev == dev && ip->inum == inum){
ip->ref++;
release(&icache.lock);
return (struct uinode*)ip;
}
if(empty == 0 && ip->ref == 0) // Remember empty slot.
empty = ip;
}
// Allocate fresh inode.
if(empty == 0)
panic("iget: no inodes");
ip = empty;
ip->dev = dev;
ip->inum = inum;
ip->ref = 1;
ip->flags = 0;
release(&icache.lock);
return (struct uinode*)ip;
}
// Increment reference count for ip.
// Returns ip to enable ip = idup(ip1) idiom.
struct uinode*
idup(struct uinode *uip)
{
struct inode *ip;
ip = (struct inode*)uip;
acquire(&icache.lock);
ip->ref++;
release(&icache.lock);
return uip;
}
// Lock the given inode.
struct inode*
ilock(struct uinode *uip)
{
struct buf *bp;
struct dinode *dip;
struct inode *ip;
ip = (struct inode*)uip;
if(ip == 0)
return 0;
if(ip->ref < 1)
panic("ilock: no refs");
acquire(&icache.lock);
while(ip->flags & I_BUSY)
sleep(ip, &icache.lock);
ip->flags |= I_BUSY;
release(&icache.lock);
if(!(ip->flags & I_VALID)){
bp = bread(ip->dev, IBLOCK(ip->inum));
dip = &((struct dinode*)(bp->data))[ip->inum % IPB];
ip->type = dip->type;
ip->major = dip->major;
ip->minor = dip->minor;
ip->nlink = dip->nlink;
ip->size = dip->size;
memmove(ip->addrs, dip->addrs, sizeof(ip->addrs));
brelse(bp);
ip->flags |= I_VALID;
if(ip->type == 0)
panic("ilock: no type");
}
return ip;
}
// Unlock the given inode.
struct uinode*
iunlock(struct inode *ip)
{
if(ip == 0)
return 0;
if(!(ip->flags & I_BUSY) || ip->ref < 1)
panic("iunlock");
acquire(&icache.lock);
ip->flags &= ~I_BUSY;
wakeup(ip);
release(&icache.lock);
return (struct uinode*)ip;
}
// Caller holds reference to unlocked ip. Drop reference.
void
iput(struct uinode *uip)
{
struct inode *ip;
ip = (struct inode*)uip;
acquire(&icache.lock);
if(ip->ref == 1 && (ip->flags & I_VALID) && ip->nlink == 0) {
// inode is no longer used: truncate and free inode.
if(ip->flags & I_BUSY)
panic("iput busy");
ip->flags |= I_BUSY;
release(&icache.lock);
// XXX convince rsc that no one will come find this inode.
itrunc(ip);
ip->type = 0;
iupdate(ip);
acquire(&icache.lock);
ip->flags &= ~I_BUSY;
}
ip->ref--;
release(&icache.lock);
}
// Allocate a new inode with the given type on device dev.
struct uinode*
ialloc(uint dev, short type)
{
int inum, ninodes;
struct buf *bp;
struct dinode *dip;
struct superblock *sb;
bp = bread(dev, 1);
sb = (struct superblock*)bp->data;
ninodes = sb->ninodes;
brelse(bp);
for(inum = 1; inum < ninodes; inum++) { // loop over inode blocks
bp = bread(dev, IBLOCK(inum));
dip = &((struct dinode*)(bp->data))[inum % IPB];
if(dip->type == 0) { // a free inode
memset(dip, 0, sizeof(*dip));
dip->type = type;
bwrite(bp); // mark it allocated on the disk
brelse(bp);
return iget(dev, inum);
}
brelse(bp);
}
panic("ialloc: no inodes");
}
// Copy inode, which has changed, from memory to disk.
void
iupdate(struct inode *ip)
{
struct buf *bp;
struct dinode *dip;
bp = bread(ip->dev, IBLOCK(ip->inum));
dip = &((struct dinode*)(bp->data))[ip->inum % IPB];
dip->type = ip->type;
dip->major = ip->major;
dip->minor = ip->minor;
dip->nlink = ip->nlink;
dip->size = ip->size;
memmove(dip->addrs, ip->addrs, sizeof(ip->addrs));
bwrite(bp);
brelse(bp);
}
// Inode contents
//
// The contents (data) associated with each inode is stored
// in a sequence of blocks on the disk. The first NDIRECT blocks
// are stored in ip->addrs[]. The next NINDIRECT blocks are
// listed in the block ip->addrs[INDIRECT].
// Return the disk block address of the nth block in inode ip.
// If there is no such block, alloc controls whether one is allocated.
static uint
bmap(struct inode *ip, uint bn, int alloc)
{
uint addr, *a;
struct buf *bp;
if(bn < NDIRECT) {
if((addr = ip->addrs[bn]) == 0) {
if(!alloc)
return -1;
ip->addrs[bn] = addr = balloc(ip->dev);
}
return addr;
}
bn -= NDIRECT;
if(bn < NINDIRECT) {
// Load indirect block, allocating if necessary.
if((addr = ip->addrs[INDIRECT]) == 0) {
if(!alloc)
return -1;
ip->addrs[INDIRECT] = addr = balloc(ip->dev);
}
bp = bread(ip->dev, addr);
a = (uint*)bp->data;
if((addr = a[bn]) == 0) {
if(!alloc) {
brelse(bp);
return -1;
}
a[bn] = addr = balloc(ip->dev);
bwrite(bp);
}
brelse(bp);
return addr;
}
panic("bmap: out of range");
}
// PAGEBREAK: 30
// Truncate inode (discard contents).
static void
itrunc(struct inode *ip)
{
int i, j;
struct buf *bp;
uint *a;
for(i = 0; i < NDIRECT; i++) {
if(ip->addrs[i]) {
bfree(ip->dev, ip->addrs[i]);
ip->addrs[i] = 0;
}
}
if(ip->addrs[INDIRECT]) {
bp = bread(ip->dev, ip->addrs[INDIRECT]);
a = (uint*)bp->data;
for(j = 0; j < NINDIRECT; j++) {
if(a[j])
bfree(ip->dev, a[j]);
}
brelse(bp);
ip->addrs[INDIRECT] = 0;
}
ip->size = 0;
iupdate(ip);
}
// Copy stat information from inode.
void
stati(struct inode *ip, struct stat *st)
{
st->dev = ip->dev;
st->ino = ip->inum;
st->type = ip->type;
st->nlink = ip->nlink;
st->size = ip->size;
}
//PAGEBREAK!
// Read data from inode.
int
readi(struct inode *ip, char *dst, uint off, uint n)
{
uint tot, m;
struct buf *bp;
if(ip->type == T_DEV) {
if(ip->major < 0 || ip->major >= NDEV || !devsw[ip->major].read)
return -1;
return devsw[ip->major].read(ip->minor, dst, n);
}
if(off + n < off)
return -1;
if(off + n > ip->size)
n = ip->size - off;
for(tot=0; tot<n; tot+=m, off+=m, dst+=m) {
bp = bread(ip->dev, bmap(ip, off/BSIZE, 0));
m = min(n - tot, BSIZE - off%BSIZE);
memmove(dst, bp->data + off%BSIZE, m);
brelse(bp);
}
return n;
}
// PAGEBREAK!
// Write data to inode.
int
writei(struct inode *ip, char *src, uint off, uint n)
{
uint tot, m;
struct buf *bp;
if(ip->type == T_DEV) {
if(ip->major < 0 || ip->major >= NDEV || !devsw[ip->major].write)
return -1;
return devsw[ip->major].write(ip->minor, src, n);
}
if(off + n < off)
return -1;
if(off + n > MAXFILE*BSIZE)
n = MAXFILE*BSIZE - off;
for(tot=0; tot<n; tot+=m, off+=m, src+=m) {
bp = bread(ip->dev, bmap(ip, off/BSIZE, 1));
m = min(n - tot, BSIZE - off%BSIZE);
memmove(bp->data + off%BSIZE, src, m);
bwrite(bp);
brelse(bp);
}
if(n > 0 && off > ip->size) {
ip->size = off;
iupdate(ip);
}
return n;
}
//PAGEBREAK!
// Directories
int
namecmp(const char *s, const char *t)
{
int i;
for(i=0; i<DIRSIZ; i++){
if(s[i] != t[i])
return s[i] - t[i];
if(s[i] == 0)
break;
}
return 0;
}
// Look for a directory entry in a directory.
// If found, set *poff to byte offset of entry.
struct uinode*
dirlookup(struct inode *dp, char *name, uint *poff)
{
uint off, inum;
struct buf *bp;
struct dirent *de;
if(dp->type != T_DIR)
return 0;
for(off = 0; off < dp->size; off += BSIZE){
bp = bread(dp->dev, bmap(dp, off / BSIZE, 0));
for(de = (struct dirent*) bp->data;
de < (struct dirent*) (bp->data + BSIZE);
de++){
if(de->inum == 0)
continue;
if(namecmp(name, de->name) == 0){
// entry matches path element
if(poff)
*poff = off + (uchar*)de - bp->data;
inum = de->inum;
brelse(bp);
return iget(dp->dev, inum);
}
}
brelse(bp);
}
return 0;
}
// Copy one name to another.
static void
namecpy(char *s, const char *t)
{
int i;
for(i=0; i<DIRSIZ && t[i]; i++)
s[i] = t[i];
for(; i<DIRSIZ; i++)
s[i] = 0;
}
// Write a new directory entry (name, ino) into the directory dp.
int
dirlink(struct inode *dp, char *name, uint ino)
{
int off;
struct dirent de;
struct uinode *ipu;
// Check that name is not present.
if((ipu = dirlookup(dp, name, 0)) != 0){
iput(ipu);
return -1;
}
// Look for an empty dirent.
for(off = 0; off < dp->size; off += sizeof(de)){
if(readi(dp, (char*)&de, off, sizeof(de)) != sizeof(de))
panic("dirwrite read");
if(de.inum == 0)
break;
}
namecpy(de.name, name);
de.inum = ino;
if(writei(dp, (char*)&de, off, sizeof(de)) != sizeof(de))
panic("dirwrite");
return 0;
}
// Paths
// Copy the next path element from path into name.
// Return a pointer to the element following the copied one.
// The returned path has no leading slashes,
// so the caller can check *path=='\0' to see if the name is the last one.
// If no name to remove, return 0.
//
// Examples:
// skipelem("a/bb/c", name) = "bb/c", setting name = "a"
// skipelem("///a/bb", name) = "b", setting name="a"
// skipelem("", name) = skipelem("////", name) = 0
//
static char*
skipelem(char *path, char *name)
{
char *s;
int len;
while(*path == '/')
path++;
if(*path == 0)
return 0;
s = path;
while(*path != '/' && *path != 0)
path++;
len = path - s;
if(len >= DIRSIZ)
memmove(name, s, DIRSIZ);
else{
memmove(name, s, len);
name[len] = 0;
}
while(*path == '/')
path++;
return path;
}
// Look up and return the inode for a path name.
// If parent is set, return the inode for the parent
// and write the final path element to name, which
// should have room for DIRSIZ bytes.
static struct uinode*
_namei(char *path, int parent, char *name)
{
struct uinode *dpu, *ipu;
struct inode *dp;
uint off;
if(*path == '/')
dpu = iget(ROOTDEV, 1);
else
dpu = idup(cp->cwd);
while((path = skipelem(path, name)) != 0){
dp = ilock(dpu);
if(dp->type != T_DIR){
iput(iunlock(dp));
return 0;
}
if(parent && *path == '\0'){
// Stop one level early.
iunlock(dp);
return dpu;
}
if((ipu = dirlookup(dp, name, &off)) == 0){
iput(iunlock(dp));
iput(ipu);
return 0;
}
iput(iunlock(dp));
dpu = ipu;
}
if(parent)
return 0;
return dpu;
}
struct uinode*
namei(char *path)
{
char name[DIRSIZ];
return _namei(path, 0, name);
}
struct uinode*
nameiparent(char *path, char *name)
{
return _namei(path, 1, name);
}