CS 235 Advanced Operating Systems, Winter 2008

Lab 5: File System and Shell

Handed out Tuesday, March 4, 2008
Due Friday, March 21, 2008

Introduction

In this lab, you will implement a simple disk-based file system, client-side file descriptors, and a Unix-like command shell! The file system itself will be implemented in microkernel fashion, outside the kernel but within its own user-space environment. Other environments access the file system by making IPC requests to this special file system environment.

Getting Started

Download our reference code for lab 5 from lab5.tar.gz and untar it, then merge it into your CVS repository as you did for the previous labs. (See the CVS hints.)

The main new component for this lab is the file system server, located in the new fs directory. Scan through all the files in this directory to get a feel for what's new. Also, there are some new file system-related source files in the user and lib directories, particularly lib/file.c, and new global header files inc/fs.h and inc/fd.h. Scan through all of these files.

You should run the pingpong, primes, and forktree test cases from lab 4 again after merging in the new lab 5 code. Run, for example, make run-pingpong. This won't work initially, since the lab 5 source tries to start up the file system, and you haven't written the file system yet! So you should temporarily comment out:

• ENV_CREATE(fs_fs); in kern/init.c, and
• close_all(); in lib/exit.c.

If your lab 4 code doesn't contain any bugs, the test cases should run fine. Don't proceed until they work.

Lab Requirements

You will need to do all of the regular exercises described in the lab. If you complete a challenge problem, provide a short (e.g., one or two paragraph) description of what you did. There is no need to answer any questions this time. Place the write-up in a file called answers.txt (plain text) or answers.html (HTML format) in the top level of your lab5 directory before handing in your work to CourseWeb.

Alternately, you may skip some or all of Lab 5 in favor of working on a substantial challenge problem you define on your own. I am excited to see what you come up with! Please contact me with your challenge problem idea so I can help you push it in interesting directions.

File system preliminaries

The file system you will work with is much simpler than most "real" file systems, but it is powerful enough to provide the standard "basic" features: creating, reading, writing, and deleting files organized in a hierarchical directory structure.

We are (for the moment anyway) developing only a "single-user" operating system, so our file system doesn't support the UNIX notions of file ownership or permissions. It also currently does not support hard links, symbolic links, time stamps, or special device files like most UNIX file systems do.

On-Disk File System Structure

Most UNIX file systems divide available disk space into two main types of regions: inode regions and data regions. UNIX file systems assign one inode to each file in the file system; a file's inode holds critical meta-data about the file such as its stat attributes and pointers to its data blocks. The data regions are divided into much larger (typically 8KB or more) data blocks, within which the file system stores file data and directory meta-data. Directory entries contain file names and pointers to inodes; a file is said to be hard-linked if multiple directory entries in the file system refer to that file's inode. Since our file system will not support hard links, we do not need this level of indirection and therefore can make a convenient simplification: our file system will not use inodes at all, but instead we will simply store all of a file's (or sub-directory's) meta-data within the (one and only) directory entry describing that file.

Both files and directories logically consist of a series of data blocks, which may be scattered throughout the disk much like the pages of an environment's virtual address space can be scattered throughout physical memory. The file system allows user processes to read and write the contents of files directly, but the file system handles all modifications to directories itself as a part of performing actions such as file creation and deletion. Our file system does, however, allow user environments to read directory meta-data directly (e.g., with read and write), which means that user environments can perform directory scanning operations themselves (e.g., to implement the ls program) rather than having to rely on additional special calls to the file system. The disadvantage of this approach to directory scanning, and the reason most modern UNIX variants discourage it, is that it makes application programs dependent on the format of directory meta-data, making it difficult to change the file system's internal layout without changing or at least recompiling application programs as well.

Sectors and Blocks

Most disks cannot perform reads and writes at byte granularity, but can only perform reads and writes in units of sectors, which today are almost universally 512 bytes each. File systems actually allocate and use disk storage in units of blocks. Be wary of the distinction between the two terms: sector size is a property of the disk hardware, whereas block size is an aspect of the operating system using the disk. A file system's block size must be at least the sector size of the underlying disk, but could be greater.

The original UNIX file system used a block size of 512 bytes, the same as the sector size of the underlying disk. Most modern file systems use a larger block size, however, because storage space has gotten much cheaper and it is more efficient to manage storage at larger granularities. Our file system will use a block size of 4096 bytes, conveniently matching the processor's page size.

Superblocks

File systems typically reserve certain disk blocks, at "easy-to-find" locations on the disk such as the very start or the very end, to hold meta-data describing properties of the file system as a whole, such as the block size, disk size, any meta-data required to find the root directory, the time the file system was last mounted, the time the file system was last checked for errors, and so on. These special blocks are called superblocks.

Our file system's superblock layout is defined by struct Super in inc/fs.h. The file system superblock will always occupy block 1 on the disk; boot loaders and partition tables use Block 0, so most file systems don't use the very first disk block. Most "real" file systems maintain multiple superblocks, replicated throughout several widely-spaced regions of the disk, so that if one of them is corrupted or the disk develops a media error in that region, the other superblocks can still be found and used to access the file system.

The Block Bitmap: Managing Free Disk Blocks

In the same way that the kernel must manage the system's physical memory to ensure that a given physical page is used for only one purpose at a time, a file system must manage the blocks of storage on a disk to ensure that a given disk block is used for only one purpose at a time. In pmap.c you keep the Page structures for all free physical pages on a linked list, page_free_list, to keep track of the free physical pages. In file systems it is more common to keep track of free disk blocks using a bitmap rather than a linked list, because a bitmap is more storage-efficient than a linked list and easier to keep consistent. Searching for a free block in a bitmap can take more CPU time than simply removing the first element of a linked list, but for file systems this isn't a problem because the I/O cost of actually accessing the free block after we find it dominates for performance purposes.

To set up a free block bitmap, we reserve a contiguous region of space on the disk large enough to hold one bit for each disk block. For example, since our file system uses 4096-byte blocks, each bitmap block contains 4096*8=32768 bits, or enough bits to describe 32768 disk blocks. In other words, for every 32768 disk blocks the file system uses, we must reserve one disk block for the block bitmap. A given bit in the bitmap is set if the corresponding block is free, and clear if the corresponding block is in use. The block bitmap in our file system always starts at disk block 2, immediately after the superblock. For simplicity we will reserve enough bitmap blocks to hold one bit for each block in the entire disk, including the blocks containing the superblock and the bitmap itself. We will simply make sure that the bitmap bits corresponding to these special, "reserved" areas of the disk are always clear (marked in-use).

File Meta-data

The layout of the meta-data describing a file in our file system is described by struct File in inc/fs.h. This meta-data includes the file's name, size, type (regular file or directory), and pointers to the blocks comprising the file. Unlike in most "real" file systems, for simplicity we will use this one File structure to represent file meta-data as it appears both on disk and in memory. Some of the fields in the structure (currently, only f_dir) are only meaningful while the File structure is in memory; whenever we read a File structure from disk into memory, we clear these fields.

The block array in struct File contains space to store the block numbers of the first 10 (NDIRECT) blocks of the file, which we call the file's direct blocks. For small files up to 10*4096 = 40KB in size, this means that the block numbers of all of the file's blocks will fit directly within the File structure itself. For larger files, however, we need a place to hold the rest of the file's block numbers. For any file greater than 40KB in size, therefore, we allocate an additional disk block, called the file's indirect block, to hold up to 4096/4 = 1024 additional block numbers. Our file system therefore allows files to be up to 1034 blocks, or a bit more than four megabytes, in size. To support larger files, "real" file systems typically support double- and triple-indirect blocks as well.

Directories versus Regular Files

A File structure in our file system can represent either a regular file or a directory; these two types of "files" are distinguished by the type field in the File structure. The file system manages regular files and directory-files in exactly the same way, except that it does not interpret the contents of the data blocks associated with regular files at all, whereas the file system interprets the contents of a directory-file as a series of File structures describing the files and subdirectories within the directory.

The superblock in our file system contains a File structure (the root field in struct Super), which holds the meta-data for the file system's root directory. The contents of this directory-file is a sequence of File structures describing the files and directories located within the root directory of the file system. Any subdirectories in the root directory may in turn contain more File structures representing sub-subdirectories, and so on.

Part 1: The File System Server

Disk Access

The file system server in our operating system needs to be able to access the disk, but we have not yet implemented any disk access functionality in our kernel. Instead of taking the conventional "monolithic" operating system strategy of adding an IDE disk driver to the kernel along with the necessary system calls to allow the file system to access it, we will instead implement the IDE disk driver as part of the user-level file system environment. We will still need to modify the kernel slightly, in order to set things up so that the file system environment has the privileges it needs to implement disk access itself.

It is easy to implement disk access in user space this way as long as we rely on polling, "programmed I/O" (PIO)-based disk access and do not use disk interrupts. It is possible to implement interrupt-driven device drivers in user mode as well (the L3 and L4 kernels do this, for example), but it is more difficult since the kernel must field device interrupts and dispatch them to the correct user-mode environment.

The x86 processor uses the IOPL bits in the EFLAGS register to determine whether protected-mode code is allowed to perform special device I/O instructions, such as IN and OUT. The IOPL bits equal the minimum (i.e. numerically highest) privilege level allowed to perform IN and OUT instructions, so if those bits are 0, only the kernel can execute INs and OUTs. All of the IDE disk registers we need to access are located in the x86's I/O space (rather than memory-mapped I/O space), so to let the file system environment access the disk, all we need to do is manipulate the IOPL bits. But no other environment should be able to access I/O space.

To keep things simple, from now on we will arrange things so that the file system is always user environment 0.

(At this point, you should un-comment out ENV_CREATE(fs/fs); from kern/init.c and close_all() from lib/exit.c.)

Do you have to do anything else to ensure that this I/O privilege setting is saved and restored properly when you subsequently switch from one environment to another? Make sure you understand how this environment state is handled.

Glance through the files in the new fs directory in the source tree. The file fs/ide.c implements our minimal PIO-based disk driver. The file fs/serv.c contains the umain function for the file system server.

This lab uses the file obj/kernel.img as the image for disk 0 (typically "Drive C" under DOS/Windows) as before, and to the (new) file obj/fs.img as the image for disk 1 ("Drive D"). In this lab your file system should only ever touch disk 1; disk 0 is used only to boot the kernel. If you manage to corrupt either disk image in some way, you can reset both of them to their original, "pristine" versions simply by typing:

$ rm obj/kernel.img obj/fs.img
$ gmake

The Block Cache

The Block Bitmap

After fs_init calls read_super (which we have provided) to read and check the file system superblock, fs_init calls read_bitmap to read and perform basic validity checking on the disk's block bitmap. For speed and simplicity, our file system will always keep the entire block bitmap in memory.

File Operations

We have provided a variety of functions in fs/fs.c to implement the basic facilities you will need to interpret and manage File structures, allocate and/or find a given block of a file, scan and manage the entries of directory-files, and walk the file system from the root to resolve an absolute pathname. Read through the code in fs/fs.c and make sure you understand what each function does before proceeding.

You may notice that there are two operations conspicuously absent from this set of functions implementing "basic" file operations: namely, read and write. This is because our file server will not implement read and write operations directly on behalf of client environments, but instead will use our kernel's IPC-based page remapping functionality to pass mapped pages to file system clients, which these client environments can then read and write directly. The page mappings we pass to clients will be exactly those pages that represent in-memory file blocks in the file system's own buffer cache, fetched via file_get_block. You will see the user-space read and write in part 2.

Client/Server File System Access

Now that we have implemented the necessary functionality within the file system server itself, we must make it accessible to other environments that wish to use the file system. There are two pieces of code required to make this happen: client stubs and server stubs. Together, they form a remote procedure call, or RPC, abstraction, where we make IPC-based communication across address spaces appear as if they were ordinary C function calls within client applications.

The RPC abstraction is defined by a single RPC type, struct Fsreq, which is declared in inc/fs.h. The client places a struct Fsreq on a page of memory, fills out its fields, and passes the request to the server via ipc_send. The client then waits for the server's response. (The fsipc function in lib/file.c does this.) The server parses the arguments (checking for errors!) and responds with an ipc_send of its own. For some requests, this ipc_send comes with a page of memory; for instance, the reply to FSREQ_OPEN contains a memory page with the new file descriptor.

Part 2: File System Access from Client Environments

Client-Side File Descriptors

Although we can write applications that directly use the client-side stubs in lib/fsipc.c to communicate with the file system server and perform file operations, this approach would be inconvenient for many applications because the IPC-based file server interface is still somewhat "low-level" and does not provide conventional read/write operations. To read or write a file, the application would first have to reserve a portion of its address space, map the appropriate blocks of the file into that address region by making requests to the file server, read and/or change the appropriate portions of those mapped pages, and finally send a "close" request to the file server to ensure that the changes get written to disk. We will write library routines to perform these tasks on behalf of the application, so that the application can use conventional UNIX-style file access operations such as read, write, and seek.

The client-side code that implements these UNIX-style file operations is located in lib/fd.c and lib/file.c. lib/fd.c contains functions to allocate and manage generic Unix-like file descriptors, while lib/file.c implements file descriptors referring to files managed by the file server. We have implemented most of the functions in both of these files for you; the only ones you need to fill in are Fd::fd and Fd::unused_fd in lib/fd.c, and open and FileDev::close in lib/file.c.

The file descriptor layer defines two new virtual address regions within each application environment's address space. The first is the file descriptor table area, starting at address FDTABLE, reserves one 4KB page worth of address space for each of the up to MAXFD (currently 32) file descriptors the application can have open at once. At any given time, a particular file descriptor table page is mapped if and only if the corresponding file descriptor is in use.

The second new virtual address region is the file mapping area, starting at virtual address FILEBASE. Like the file descriptor table, the file mapping area is organized as a table indexed by file descriptor, except the "table entries" in the file mapping area consist of 4MB rather than 4KB of address space. In particular, for each of the MAXFD possible file descriptors, we reserve a fixed 4MB region in the file mapping area in which to map the contents of currently open files. Since our file server only supports files of up to 4MB in size, these client-side functions are not imposing any new restrictions by only reserving 4MB of space to map the contents of each open file.

Part 3: Spawning Processes from the File System

In this exercise, you'll extend spawn from Lab 4 to load program images from the file system as well as from kernel binary images. If spawn is passed a binary name like "/ls" that begins with a slash, it will read the program data from disk; otherwise, it will read the program data from the kernel. Luckily, this requires just a couple of changes.

Part 4: A Shell

In this part of the lab, you'll extend JOS to handle everything necessary to support a shell. We've done a lot of the work for you, but you must (1) make it possible to share file descriptors across environments, (2) clean up a couple loose ends, and (3) implement file redirection in the shell.

Sharing pages between environments

We would like to share file descriptor state across fork and spawn, but file descriptor state is kept in user-space memory. Right now, on fork, the memory will be marked copy-on-write, so the state will be duplicated rather than shared. (This means that running "(date; ls) >file" will not work properly, because even though date updates its own file offset, ls will not see the change.) On spawn, the memory will be left behind, not copied at all. (Effectively, the spawned environment starts with no open file descriptors.)

We will change both fork and spawn to know that certain regions of memory are used by the "library operating system" and should always be shared. Rather than hard-code a list of regions somewhere, we will set an otherwise-unused bit in the page table entries (just like we did with the PTE_COW bit in fork).

We have defined a new PTE_SHARE bit in inc/lib.h. If a page table entry has this bit set, then by convention, the PTE should be copied directly from parent to child in both fork and spawn. Note that this is different from marking it copy-on-write: as described in the first paragraph, we want to make sure to share updates to the page.

Use gmake run-testpteshare to check that your code is behaving properly. You should see lines that say "fork handles PTE_SHARE right" and "spawn handles PTE_SHARE right".

Use gmake run-testfdsharing to check that file descriptors are shared properly. You should see lines that say "read in child succeeded" and "read in parent succeeded".

At this point, you can use gmake run-initsh to boot into the current version of the shell, which can already do simple commands like "ls". As you progress through the lab, the shell will become more functional, and you will be able to do things like add redirections.

Pipes

Pipes and the console are both I/O stream interfaces. This means that they support reading and/or writing, but not file positions. Like Unix, JOS represents these streams using file descriptors. To support this, the file descriptor subsystem uses a simple virtual file system layer, implemented by struct Dev, so that disk files, console files, and pipes all implement the same file descriptor functions.

A pipe is a shared data buffer accessed via two file descriptors, one for writing data into the pipe and one for reading data out of it. Unix command lines like "ls | sort" use pipes. The shell creates a pipe, hooks up ls's standard output to the write end of the pipe, and hooks up sort's standard input to the read end of the pipe. As a result, ls's output is processed by sort. You may want to read the pipe manual page for background, and the pipe section of Dennis Ritchie's UNIX history paper for interesting history.

In Unix-like designs, each pipe's shared data buffer is stored in the kernel. Of course, this is not how we implement pipes on an exokernel! Your library operating system represents a pipe, including its shared buffer, by a single struct Pipe. The struct Pipe is stored on its own page to make sharing easier, and mapped into the file mapping area of both the reading and the writing file descriptor. Here's the structure:

#define PIPEBUFSIZ 32
struct Pipe {
        off_t p_rpos;                    // read position
        off_t p_wpos;                    // write position
        uint8_t p_buf[PIPEBUFSIZ];       // shared buffer
};

This is a simple lock-free queue structure. The pipe starts in this state:

p_rpos = 0 ---+
p_wpos = 0 ---|+
              VV
            +---+---+---+---+---+---+---+- ... -+---+---+---+---+
    p_buf:  |   |   |   |   |   |   |   |       |   |   |   |   |
            +---+---+---+---+---+---+---+- ... -+---+---+---+---+
              0   1   2   3   4   5   6           28  29  30  31

The bytes written to the pipe can be thought of as numbered starting from 0. The write position p_wpos gives the number of the next byte that will be written, and the read position p_rpos gives the number of the next byte to be read. After a writer writes "abc" to the pipe, it will enter this state:

p_rpos = 0 ---+
p_wpos = 3 ---|-----------+
              V           V
            +---+---+---+---+---+---+---+- ... -+---+---+---+---+
    p_buf:  | a | b | c |   |   |   |   |       |   |   |   |   |
            +---+---+---+---+---+---+---+- ... -+---+---+---+---+
              0   1   2   3   4   5   6           28  29  30  31

Since p_rpos != p_wpos, the pipe contains data. The next read from the pipe will return the next 3 characters. For example, after a read() of one byte:

p_rpos = 1 -------+
p_wpos = 3 -------|-------+
                  V       V
            +---+---+---+---+---+---+---+- ... -+---+---+---+---+
    p_buf:  |   | b | c |   |   |   |   |       |   |   |   |   |
            +---+---+---+---+---+---+---+- ... -+---+---+---+---+
              0   1   2   3   4   5   6           28  29  30  31

This data structure is safe for concurrent updates as long as there is a single reader and a single writer, since only the reader updates p_rpos and only the writer updates p_wpos.

Since the pipe buffer is not infinite, byte i is stored in pipe buffer index i % PIPEBUFSIZ. Thus, after a couple reads and writes, the pipe might enter this state:

p_rpos = 30 ----------------------------------------------+
p_wpos = 33 ------+                                       |
                  V                                       V
            +---+---+---+---+---+---+---+- ... -+---+---+---+---+
    p_buf:  | $ |   |   |   |   |   |   |       |   |   | ! | @ |
            +---+---+---+---+---+---+---+- ... -+---+---+---+---+
              0   1   2   3   4   5   6           28  29  30  31

Note that byte 32 was stored in slot 0.

If p_rpos == p_wpos, the pipe is empty. Any read call should yield until a writer adds information to the pipe. Similarly, if p_wpos - p_rpos == PIPEBUFSIZ, the pipe is full. Any write call should yield until a reader opens up some space in the pipe.

Closed Pipes

There is a catch -- maybe we are trying to read from an empty pipe but all the writers have exited. Then there is no chance that there will ever be more data in the pipe, so waiting is futile. In such a case, Unix signals end-of-file by returning 0. So will we. To detect that there are no writers left, we could put reader and writer counts into the pipe structure and update them every time we fork or spawn and every time an environment exits. This is fragile -- what if the environment doesn't exit cleanly? Instead we can use the kernel's page reference counts, which are guaranteed to be accurate.

Recall that the kernel page structures are mapped read-only in user environments. The library function pageref(void *ptr) returns the number of page table references to the page containing the virtual address ptr. It works by first examining vpt[] to find ptr's physical address, then looking up the relevant struct Page in the UPAGES array and returning its pp_ref field. So, for example, if fd is a pointer to a particular struct Fd, pageref(fd) will tell us how many different references there are to that structure.

Three pages are allocated for each pipe: the struct Fd for the reading file descriptor rfd, the struct Fd for the writing file descriptor wfd, and the struct Pipe p shared by both. The struct Pipe page is mapped once per file descriptor reference. Thus, the following equation holds: pageref(rfd) + pageref(wfd) = pageref(p). A reader can check whether there are any writers left by examining these counts. If pageref(p) == pageref(rfd), then pageref(wfd) == 0, and there are no more writers. A writer can check for readers in the same manner.

Pipe Races

File descriptor structures use shared memory that is written concurrently by multiple processes. That creepy shiver that just ran up your back is justified: this kind of situation is ripe with race conditions. Our file descriptor code contains many race conditions -- for example, if a multiple processes are reading from the same file descriptor, then updates to the file offset may get lost. But we've made one race condition, concerning pipes, particularly easy to run into.

The race is that the two calls to pageref() in _pipeisclosed might not happen atomically. If another process duplicates or closes the file descriptor page between the two calls, the comparison will be meaningless. To make it concrete, suppose that we run:

	pipe(p);
	if (fork() == 0) {
		close(p[1]);
		read(p[0], buf, sizeof(buf));
	} else {
		close(p[0]);
		write(p[1], msg, strlen(msg));
	}

The following might happen:

1. The child runs first after the fork. It closes p[1] and then tries to read from p[0]. The pipe is empty, so read checks to see whether the pipe is closed before yielding. Inside _pipeisclosed, pageref(fd) returns 2 (both the parent and the child have p[0] open), but then a clock interrupt happens.
2. Now the kernel chooses to run the parent for a little while. The parent closes p[0] and writes msg into the pipe. Msg is very long, so the write yields halfway to let a reader (the child) empty the pipe.
3. Back in the child, _pipeisclosed continues. It calls pageref(p), which returns 2 (the child has a reference associated with p[0], and the parent has a reference associated with p[1]). The counts match, so _pipeisclosed reports that the pipe is closed. Oops.

Run "gmake run-testpiperace2" to see this race in action. You should see "RACE: pipe appears closed" when the race occurs.

This race isn't that hard to fix. Comparing the counts can only be incorrect if another environment ran between when we looked up the first count and when we looked up the second count. In other words, we need to make sure that _pipeisclosed executes atomically. Since it doesn't change any variables, we can simply rerun it until it runs without being interrupted; the code is so short that it will usually not be interrupted.

But how can we tell whether our environment has been interrupted? In the uniprocessor JOS kernel, this can be simple: just check the env_runs variable in our environment structure. Each time the kernel runs an environment, it increments that environment's env_runs. Thus, user code can record env->env_runs, do its computation, and then look at env->env_runs again. If env_runs didn't change, then the environment was not interrupted. Conversely, if env_runs did change, then the environment was interrupted.

Run "gmake run-testpiperace2" to check whether the race still happens. If it's gone, you should not see "RACE: pipe appears closed", and you should see "race didn't happen". You should also see plenty of your "avoided" messages, indicating places where the race would have happened if you weren't being so careful. (The number of "avoided" messages depends on the ips value in your .bochsrc.)

The shell itself

Before running your shell, you must enable keyboard interrupts.

Run gmake xrun-initsh. This will run your kernel inside the X11 Bochs starting user/initsh, which sets up the console as file descriptors 0 and 1 (standard input and standard output), then spawns sh, the shell. Run ls and cat lorem.

Once your shell is working, you should be able to run the following commands:

echo hello world | cat
cat lorem >out
cat out
cat lorem |num
cat lorem |num |num |num |num |num
lsfd
cat script
sh <script

Note that the user library routine printf prints straight to the console, without using the file descriptor code. This is great for debugging but not great for piping into other programs. To print output to a particular file descriptor (for example, 1, standard output), use fprintf(1, "...", ...). See user/ls.c for examples.

Run gmake run-testshell to test your shell. Testshell simply feeds the above commands (also found in fs/testshell.sh) into the shell and then checks that the output matches fs/testshell.key.

This completes the course!

Back to CS 235 Advanced Operating Systems, Winter 2008