MIT6.828_HW10_Bigger file for xv6

本文主要是介绍MIT6.828_HW10_Bigger file for xv6，希望对大家解决编程问题提供一定的参考价值，需要的开发者们随着小编来一起学习吧！

MIT6.828_HW10_Bigger file for xv6

当前 xv6 文件大小限制在 140 个扇区。直接索引节点 12 个， 以及一个一级索引节点，其指向一个 sector，可包含 512/4 = 128个扇区。即总和为12+128 = 140个扇区。我们需要为xv6的文件节点添加一个二级索引节点，其包含128个一级索引节点的地址，每个一级索引节点又包含了128数据扇区。最终一个文件大小将增加到 16523 个扇区，大约8.5M。

The format of an on-disk inode is defined by struct dinode in fs.h. You’re particularly interested in NDIRECT, NINDIRECT, MAXFILE, and the addrs[] element of struct dinode. xv6 标准 inode 如下图所示。

在读写文件时，会调用 bmap()(in fs.c)。在写时，bmap 分配存储文件所需的 blocks并且如果需要存储块地址，还会分配一个 indirect block。

bmap() 处理两种页号. The bn argument is a “logical block” – a block number relative to the start of the file. The block numbers in ip->addrs[], and the argument to bread(), are disk block numbers. You can view bmap() as mapping a file’s logical block numbers into disk block numbers.

实现过程

修改 param.h 文件, 将 #define FSSIZE 1000改为

#define FSSIZE       20000

Download big.c into your xv6 directory, add it to the UPROGS list, start up xv6, and run big. It creates as big a file as xv6 will let it, and reports the resulting size. It should say 140 sectors.
做完预处理工作后运行big，应该有以下输出。

$ big
.
wrote 140 sectors
done; ok

You’ll have to have only 11 direct blocks, rather than 12, to make room for your new doubly-indirect block。 之所以这么 arrange， 可以避免我们修改 Inode 结点 addr数组的个数，索引结点总数依然是NDIRECT + 1。不过后来发现，修改NDIRECT,程序会更加清晰。即相应需要修改两处

// 1
#define NDIRECT 11
#define NINDIRECT (BSIZE / sizeof(uint))
// DOUBLE INDIRECT
#define NDINDIRECT (NINDIRECT * NINDIRECT)
#define MAXFILE (NDIRECT + NINDIRECT + NDINDIRECT)// 2
// in-memory copy of an inode
struct inode {...uint addrs[NDIRECT+ 2];
};// 3
// On-disk inode structure
struct dinode {short type;           // File type...          // Size of file (bytes)uint addrs[NDIRECT + 2];   // Data block addresses
};

buf 结构体。

struct buf {int flags;uint dev;uint blockno;struct sleeplock lock;uint refcnt;struct buf *prev; // LRU cache liststruct buf *next;struct buf *qnext; // disk queueuchar data[BSIZE];
};

最终实现代码并不难写出，仿照一级索引结点写法使用相应的函数调用。[Note]需要注意 brelse(bp)的时机。详细过程可以查看注释。

static uint
bmap(struct inode *ip, uint bn)
{uint addr, *a;struct buf *bp;struct buf *bp2;// 直接索引结点数目 bn= 0~10if(bn < NDIRECT){// 创建直接索引if((addr = ip->addrs[bn]) == 0)ip->addrs[bn] = addr = balloc(ip->dev);return addr;}bn -= NDIRECT;// #define NINDIRECT (BSIZE / sizeof(uint))  BSIZE = 512if(bn < NINDIRECT){// 一级索引// Load indirect block, allocating if necessary.if((addr = ip->addrs[NDIRECT]) == 0)ip->addrs[NDIRECT] = addr = balloc(ip->dev);bp = bread(ip->dev, addr);a = (uint*)bp->data;if((addr = a[bn]) == 0){a[bn] = addr = balloc(ip->dev);log_write(bp);}brelse(bp);return addr;}bn -= NINDIRECT;// 二级索引if (bn < NDINDIRECT) {// 根结点if((addr = ip->addrs[NDIRECT+1]) == 0)ip->addrs[NDIRECT+1] = addr = balloc(ip->dev);bp = bread(ip->dev, addr);// 指向一级索引a = (uint *)bp->data;if ((addr = a[bn/NINDIRECT]) == 0) {a[bn/NINDIRECT] = addr = balloc(ip->dev);log_write(bp);}bp2 = bread(ip->dev, addr);// 二级页表a = (uint *)bp2->data;if ((addr = a[bn%NINDIRECT]) == 0) {a[bn%NINDIRECT] = addr = balloc(ip->dev);log_write(bp2);}brelse(bp2);brelse(bp);return addr;}panic("bmap: out of range");
}

这篇关于MIT6.828_HW10_Bigger file for xv6的文章就介绍到这儿，希望我们推荐的文章对编程师们有所帮助！

---