Import big file assignment.

This commit is contained in:
Frans Kaashoek 2019-07-30 14:33:09 -04:00
parent 34980381bd
commit b70cd9ebd7

145
labs/fs.html Normal file
View file

@ -0,0 +1,145 @@
<html>
<head>
<title>Lab: file system</title>
<link rel="stylesheet" href="homework.css" type="text/css" />
</head>
<body>
<h1>Lab: file system</h1>
<p>In this lab you will add large files and <tt>mmap</tt> to the xv6 file system.
<h2>Large files</h2>
<p>In this assignment you'll increase the maximum size of an xv6
file. Currently xv6 files are limited to 268 blocks, or 268*BSIZE
bytes (BSIZE is 1024 in xv6). This limit comes from the fact that an
xv6 inode contains 12 "direct" block numbers and one "singly-indirect"
block number, which refers to a block that holds up to 256 more block
numbers, for a total of 12+256=268. You'll change the xv6 file system
code to support a "doubly-indirect" block in each inode, containing
256 addresses of singly-indirect blocks, each of which can contain up
to 256 addresses of data blocks. The result will be that a file will
be able to consist of up to 256*256+256+11 blocks (11 instead of 12,
because we will sacrifice one of the direct block numbers for the
double-indirect block).
<h3>Preliminaries</h3>
<p>Modify your Makefile's <tt>CPUS</tt> definition so that it reads:
<pre>
CPUS := 1
</pre>
<b>XXX doesn't seem to speedup things</b>
<p>Add
<pre>
QEMUEXTRA = -snapshot
</pre>
right before
<tt>QEMUOPTS</tt>
<p>
The above two steps speed up qemu tremendously when xv6
creates large files.
<p><tt>mkfs</tt> initializes the file system to have fewer
than 1000 free data blocks, too few to show off the changes
you'll make. Modify <tt>param.h</tt> to
set <tt>FSSIZE</tt> to:
<pre>
#define FSSIZE 20000 // size of file system in blocks
</pre>
<p>Download <a href="big.c">big.c</a> into your xv6 directory,
add it to the UPROGS list, start up xv6, and run <tt>big</tt>.
It creates as big a file as xv6 will let
it, and reports the resulting size. It should say 140 sectors.
<h3>What to Look At</h3>
The format of an on-disk inode is defined by <tt>struct dinode</tt>
in <tt>fs.h</tt>. You're particularly interested in <tt>NDIRECT</tt>,
<tt>NINDIRECT</tt>, <tt>MAXFILE</tt>, and the <tt>addrs[]</tt> element
of <tt>struct dinode</tt>. Look Figure 7.3 in the xv6 text for a
diagram of the standard xv6 inode.
<p>
The code that finds a file's data on disk is in <tt>bmap()</tt>
in <tt>fs.c</tt>. Have a look at it and make sure you understand
what it's doing. <tt>bmap()</tt> is called both when reading and
writing a file. When writing, <tt>bmap()</tt> allocates new
blocks as needed to hold file content, as well as allocating
an indirect block if needed to hold block addresses.
<p>
<tt>bmap()</tt> deals with two kinds of block numbers. The <tt>bn</tt>
argument is a "logical block" -- a block number relative to the start
of the file. The block numbers in <tt>ip->addrs[]</tt>, and the
argument to <tt>bread()</tt>, are disk block numbers.
You can view <tt>bmap()</tt> as mapping a file's logical
block numbers into disk block numbers.
<h3>Your Job</h3>
Modify <tt>bmap()</tt> so that it implements a doubly-indirect
block, in addition to direct blocks and a singly-indirect block.
You'll have to have only 11 direct blocks, rather than 12,
to make room for your new doubly-indirect block; you're
not allowed to change the size of an on-disk inode.
The first 11 elements of <tt>ip->addrs[]</tt> should be
direct blocks; the 12th should be a singly-indirect block
(just like the current one); the 13th should be your new
doubly-indirect block.
<p>
You don't have to modify xv6 to handle deletion of files with
doubly-indirect blocks.
<p>
If all goes well, <tt>big</tt> will now report that it
can write sectors. It will take <tt>big</tt> minutes
to finish.
<b>XXX this runs for a while!</b>
<h3>Hints</h3>
<p>
Make sure you understand <tt>bmap()</tt>. Write out a diagram of the
relationships between <tt>ip->addrs[]</tt>, the indirect block, the
doubly-indirect block and the singly-indirect blocks it points to, and
data blocks. Make sure you understand why adding a doubly-indirect
block increases the maximum file size by 256*256 blocks (really -1),
since you have to decrease the number of direct blocks by one).
<p>
Think about how you'll index the doubly-indirect block, and
the indirect blocks it points to, with the logical block
number.
<p>If you change the definition of <tt>NDIRECT</tt>, you'll
probably have to change the size of <tt>addrs[]</tt>
in <tt>struct inode</tt> in <tt>file.h</tt>. Make sure that
<tt>struct inode</tt> and <tt>struct dinode</tt> have the
same number of elements in their <tt>addrs[]</tt> arrays.
<p>If you change the definition of <tt>NDIRECT</tt>, make sure to create a
new <tt>fs.img</tt>, since <tt>mkfs</tt> uses <tt>NDIRECT</tt> too to build the
initial file systems. If you delete <tt>fs.img</tt>, <tt>make</tt> on Unix (not
xv6) will build a new one for you.
<p>If your file system gets into a bad state, perhaps by crashing,
delete <tt>fs.img</tt> (do this from Unix, not xv6). <tt>make</tt> will build a
new clean file system image for you.
<p>Don't forget to <tt>brelse()</tt> each block that you
<tt>bread()</tt>.
<p>You should allocate indirect blocks and doubly-indirect
blocks only as needed, like the original <tt>bmap()</tt>.
<h2>Memory-mapped files</h2>
</body>
</html>