If you just want to try out busybox without installing it, download the tarball, extract it, run "make defconfig", and then run "make".
This will create a busybox binary with almost all features enabled. To try out a busybox applet, type "./busybox [appletname] [options]", for example "./busybox ls -l" or "./busybox cat LICENSE". Type "./busybox" to see a command list, and "busybox appletname --help" to see a brief usage message for a given applet.
BusyBox uses the name it was invoked under to determine which applet is being invoked. (Try "mv busybox ls" and then "./ls -l".) Installing busybox consists of creating symlinks (or hardlinks) to the busybox binary for each applet in busybox, and making sure these links are in the shell's command $PATH. The special applet name "busybox" (or with any optional suffix, such as "busybox-static") uses the first argument to determine which applet to run, as shown above.
BusyBox also has a feature called the "standalone shell", where the busybox shell runs any built-in applets before checking the command path. This feature is also enabled by "make allyesconfig", and to try it out run the command line "PATH= ./busybox ash". This will blank your command path and run busybox as your command shell, so the only commands it can find (without an explicit path such as /bin/ls) are the built-in busybox ones. This is another good way to see what's built into busybox. Note that the standalone shell requires CONFIG_BUSYBOX_EXEC_PATH to be set appropriately, depending on whether or not /proc/self/exe is available or not. If you do not have /proc, then point that config option to the location of your busybox binary, usually /bin/busybox. (So if you set it to /proc/self/exe, and happen to be able to chroot into your rootfs, you must mount /proc beforehand.)
A typical indication that you set CONFIG_BUSYBOX_EXEC_PATH to proc but forgot to mount proc is:
$ /bin/echo $PATH /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/bin/X11 $ echo $PATH /bin/sh: echo: not found
Busybox is configured similarly to the linux kernel. Create a default configuration and then run "make menuconfig" to modify it. The end result is a .config file that tells the busybox build process what features to include. So instead of "./configure; make; make install" the equivalent busybox build would be "make defconfig; make; make install".
Busybox configured with all features enabled is a little under a megabyte dynamically linked on x86. To create a smaller busybox, configure it with fewer features. Individual busybox applets cost anywhere from a few hundred bytes to tens of kilobytes. Disable unneeded applets to save, space, using menuconfig.
The most important busybox configurators are:
make defconfig - Create the maximum "sane" configuration. This enables almost all features, minus things like debugging options and features that require changes to the rest of the system to work (such as selinux or devfs device names). Use this if you want to start from a full-featured busybox and remove features until it's small enough.
make allnoconfig - Disable everything. This creates a tiny version of busybox that doesn't do anything. Start here if you know exactly what you want and would like to select only those features.
make menuconfig - Interactively modify a .config file through a multi-level menu interface. Use this after one of the previous two.
Some other configuration options are:
make oldconfig - Update an old .config file for a newer version of busybox.
make allyesconfig - Select absolutely everything. This creates a statically linked version of busybox full of debug code, with dependencies on selinux, using devfs names... This makes sure everything compiles. Whether or not the result would do anything useful is an open question.
make allbareconfig - Select all applets but disable all sub-features within each applet. More build coverage testing.
make randconfig - Create a random configuration for test purposes.
Menuconfig modifies your .config file through an interactive menu where you can enable or disable busybox features, and get help about each feature.
To build a smaller busybox binary, run "make menuconfig" and disable the features you don't need. (Or run "make allnoconfig" and then use menuconfig to add just the features you need. Don't forget to recompile with "make" once you've finished configuring.)
To build busybox with a cross-compiler, specify CROSS_COMPILE=<prefix>.
CROSS_COMPILE specifies the prefix used for all executables used during compilation. Only gcc and related binutils executables are prefixed with $(CROSS_COMPILE) in the makefiles. CROSS_COMPILE can be set on the command line:
make CROSS_COMPILE=arm-linux-uclibcgnueabi-Alternatively CROSS_COMPILE can be set in the environment. Default value for CROSS_COMPILE is not to prefix executables.
BusyBox is a package that replaces a dozen standard packages, but it is not by itself a complete bootable system. Building an entire Linux distribution from source is a bit beyond the scope of this FAQ, but it understandably keeps cropping up on the mailing list, so here are some pointers.
Start by learning how to strip a working system down to the bare essentials needed to run one or two commands, so you know what it is you actually need. An excellent practical place to do this is the Linux BootDisk Howto, or for a more theoretical approach try From PowerUp to Bash Prompt.
To learn how to build a working Linux system entirely from source code, the place to go is the Linux From Scratch project. They have an entire book of step-by-step instructions you can read online or download. Be sure to check out the other sections of their main page, including Beyond Linux From Scratch, Hardened Linux From Scratch, their Hints directory, and their LiveCD project. (They also have mailing lists which are better sources of answers to Linux-system building questions than the busybox list.)
If you want an automated yet customizable system builder which produces a BusyBox and uClibc based system, try buildroot, which is another project by the maintainer of the uClibc (Erik Andersen). Download the tarball, extract it, unset CC, make. For more instructions, see the website.
Full functionality requires Linux 2.4.x or better. (Earlier versions may still work, but are no longer regularly tested.) A large fraction of the code should run on just about anything. While the current code is fairly Linux specific, it should be fairly easy to port the majority of the code to support, say, FreeBSD or Solaris, or Mac OS X, or even Windows (if you are into that sort of thing).
BusyBox in general will build on any architecture supported by gcc. Kernel module loading for 2.4 Linux kernels is currently limited to ARM, CRIS, H8/300, x86, ia64, x86_64, m68k, MIPS, PowerPC, S390, SH3/4/5, Sparc, v850e, and x86_64 for 2.4.x kernels.
With 2.6.x kernels, module loading support should work on all architectures.
On Linux, BusyBox releases are tested against uClibc (0.9.27 or later) and glibc (2.2 or later). Both should provide full functionality with busybox, and if you find a bug we want to hear about it.
Linux-libc5 is no longer maintained (and has no known advantages over uClibc), dietlibc is known to have numerous unfixed bugs, and klibc is missing too many features to build BusyBox. If you require a small C library for Linux, the busybox developers recommend uClibc.
Some BusyBox applets have been built and run under a combination of newlib and libgloss (see this thread). This is still experimental, but may be supported in a future release.
Yes. As long as you fully comply with the generous terms of the GPL BusyBox license you can ship BusyBox as part of the software on your device.
we maintain such a list on this site!
You have not paid us a single cent and yet you still have the product of many years of our work. We are not your slaves! We work on BusyBox because we find it useful and interesting. If you go off flaming us, we will ignore you.
If you find that you need help with BusyBox, you can ask for help on the BusyBox mailing list at busybox@busybox.net.
In addition to the mailing list, Erik Andersen (andersee), Manuel Nova (mjn3), Rob Landley (landley), Mike Frysinger (SpanKY), Bernhard Fischer (blindvt), and other long-time BusyBox developers are known to hang out on the uClibc IRC channel: #uclibc on irc.freenode.net. There is a web archive of daily logs of the #uclibc IRC channel going back to 2002.
Please do not send private email to Rob, Erik, Manuel, or the other BusyBox contributors asking for private help unless you are planning on paying for consulting services.
When we answer questions on the BusyBox mailing list, it helps everyone since people with similar problems in the future will be able to get help by searching the mailing list archives. Private help is reserved as a paid service. If you need to use private communication, or if you are serious about getting timely assistance with BusyBox, you should seriously consider paying for consulting services.
Yes we are. The easy way to sponsor a new feature is to post an offer on the mailing list to see who's interested. You can also email the project's maintainer and ask them to recommend someone.
If you prefer to deal with an organization rather than an individual, Rob Landley (the current BusyBox maintainer) works for TimeSys, and Eric Andersen (the previous busybox maintainer and current uClibc maintainer) owns CodePoet Consulting. Both companies offer support contracts and handle new development, and there are plenty of other companies that do the same.
If you simply need help with using or configuring BusyBox, please submit a detailed description of your problem to the BusyBox mailing list at busybox@busybox.net. Please do not send email to individual developers asking for private help unless you are planning on paying for consulting services. When we answer questions on the BusyBox mailing list, it helps everyone, while private answers help only you...
Bug reports and new feature patches sometimes get lost when posted to the mailing list, because the developers of BusyBox are busy people and have only so much they can keep in their brains at a time. You can post a polite reminder after 2-3 days without offending anybody. If that doesn't result in a solution, please use the BusyBox Bug and Patch Tracking System to submit a detailed explanation and we'll get to it as soon as we can.
Note that bugs entered into the bug system without being mentioned on the mailing list first may languish there for months before anyone even notices them. We generally go through the bug system when preparing for new development releases, to see what fell through the cracks while we were off writing new features. (It's a fast/unreliable vs slow/reliable thing. Saves retransits, but the latency sucks.)
Variants of this one get asked a lot.
The purpose of the BusyBox mailing list is to develop and improve BusyBox, and we're happy to respond to our users' needs. But if you're coming to the list for free tech support we're going to ask you to upgrade to a current version before we try to diagnose your problem.
If you're building BusyBox 0.50 with uClibc 0.9.19 and gcc 0.9.26 there's a fairly large chance that whatever problem you're seeing has already been fixed. To get that fix, all you have to do is upgrade to a newer version. If you don't at least _try_ that, you're wasting our time.
The volunteers are happy to fix any bugs you point out in the current versions because doing so helps everybody and makes the project better. We want to make the current version work for you. But diagnosing, debugging, and backporting fixes to old versions isn't something we do for free, because it doesn't help anybody but you. The cost of volunteer tech support is using a reasonably current version of the project.
If you don't want to upgrade, you have the complete source code and thus the ability to fix it yourself, or hire a consultant to do it for you. If you got your version from a vendor who still supports the older version, they can help you. But there are limits as to what the volunteers will feel obliged to do for you.
As a rule of thumb, volunteers will generally answer polite questions about a given version for about three years after its release before it's so old we don't remember the answer off the top of our head. And if you want us to put any _effort_ into tracking it down, we want you to put in a little effort of your own by confirming it's still a problem with the current version. It's also hard for us to fix a problem of yours if we can't reproduce it because we don't have any systems running an environment that old.
A consultant will happily set up a special environment just to reproduce your problem, and you can always ask on the list if any of the developers have consulting rates.
Init is the first program that runs, so it might be that no programs are working on your new system because of a problem with your cross-compiler, kernel, console settings, shared libraries, root filesystem... To rule all that out, first build a statically linked version of the following "hello world" program with your cross compiler toolchain:
#include <stdio.h>
int main(int argc, char *argv)
{
printf("Hello world!\n");
sleep(999999999);
}
Now try to boot your device with an "init=" argument pointing to your hello world program. Did you see the hello world message? Until you do, don't bother messing with busybox init.
Once you've got it working statically linked, try getting it to work dynamically linked. Then read the FAQ entry How do I build a BusyBox-based system?, and the documentation for BusyBox init.
Configuring Busybox depends on a recent version of sed. Older distributions (Red Hat 7.2, Debian 3.0) may not come with a usable version. Luckily BusyBox can use its own sed to configure itself, although this leads to a bit of a chicken and egg problem. You can work around this by hand-configuring busybox to build with just sed, then putting that sed in your path to configure the rest of busybox with, like so:
tar xvjf sources/busybox-x.x.x.tar.bz2 cd busybox-x.x.x make allnoconfig make include/bb_config.h echo "CONFIG_SED=y" >> .config echo "#undef ENABLE_SED" >> include/bb_config.h echo "#define ENABLE_SED 1" >> include/bb_config.h make mv busybox sed export PATH=`pwd`:"$PATH"
Then you can run "make defconfig" or "make menuconfig" normally.
Job control will be turned off since your shell can not obtain a controlling terminal. This typically happens when you run your shell on /dev/console. The kernel will not provide a controlling terminal on the /dev/console device. Your should run your shell on a normal tty such as tty1 or ttyS0 and everything will work perfectly. If you REALLY want your shell to run on /dev/console, then you can hack your kernel (if you are into that sortof thing) by changing drivers/char/tty_io.c to change the lines where it sets "noctty = 1;" to instead set it to "0". I recommend you instead run your shell on a real console...
Busybox aims to be the smallest and simplest correct implementation of the standard Linux command line tools. First and foremost, this means the smallest executable size we can manage. We also want to have the simplest and cleanest implementation we can manage, be standards compliant, minimize run-time memory usage (heap and stack), run fast, and take over the world.
Busybox is like a swiss army knife: one thing with many functions. The busybox executable can act like many different programs depending on the name used to invoke it. Normal practice is to create a bunch of symlinks pointing to the busybox binary, each of which triggers a different busybox function. (See getting started in the FAQ for more information on usage, and the busybox documentation for a list of symlink names and what they do.)
The "one binary to rule them all" approach is primarily for size reasons: a single multi-purpose executable is smaller then many small files could be. This way busybox only has one set of ELF headers, it can easily share code between different apps even when statically linked, it has better packing efficiency by avoding gaps between files or compression dictionary resets, and so on.
Work is underway on new options such as "make standalone" to build separate binaries for each applet, and a "libbb.so" to make the busybox common code available as a shared library. Neither is ready yet at the time of this writing.
The directory "applets" contains the busybox startup code (applets.c and busybox.c), and several subdirectories containing the code for the individual applets.
Busybox execution starts with the main() function in applets/busybox.c, which sets the global variable applet_name to argv[0] and calls run_applet_and_exit() in applets/applets.c. That uses the applets[] array (defined in include/busybox.h and filled out in include/applets.h) to transfer control to the appropriate APPLET_main() function (such as cat_main() or sed_main()). The individual applet takes it from there.
This is why calling busybox under a different name triggers different functionality: main() looks up argv[0] in applets[] to get a function pointer to APPLET_main().
Busybox applets may also be invoked through the multiplexor applet "busybox" (see busybox_main() in applets/busybox.c), and through the standalone shell (grep for STANDALONE_SHELL in applets/shell/*.c). See getting started in the FAQ for more information on these alternate usage mechanisms, which are just different ways to reach the relevant APPLET_main() function.
The applet subdirectories (archival, console-tools, coreutils, debianutils, e2fsprogs, editors, findutils, init, loginutils, miscutils, modutils, networking, procps, shell, sysklogd, and util-linux) correspond to the configuration sub-menus in menuconfig. Each subdirectory contains the code to implement the applets in that sub-menu, as well as a Config.in file defining that configuration sub-menu (with dependencies and help text for each applet), and the makefile segment (Makefile.in) for that subdirectory.
The run-time --help is stored in usage_messages[], which is initialized at the start of applets/applets.c and gets its help text from usage.h. During the build this help text is also used to generate the BusyBox documentation (in html, txt, and man page formats) in the docs directory. See adding an applet to busybox for more information.
Most non-setup code shared between busybox applets lives in the libbb directory. It's a mess that evolved over the years without much auditing or cleanup. For anybody looking for a great project to break into busybox development with, documenting libbb would be both incredibly useful and good experience.
Common themes in libbb include allocation functions that test for failure and abort the program with an error message so the caller doesn't have to test the return value (xmalloc(), xstrdup(), etc), wrapped versions of open(), close(), read(), and write() that test for their own failures and/or retry automatically, linked list management functions (llist.c), command line argument parsing (getopt32.c), and a whole lot more.
To conserve bytes it's good to know where they're being used, and the size of the final executable isn't always a reliable indicator of the size of the components (since various structures are rounded up, so a small change may not even be visible by itself, but many small savings add up).
The busybox Makefile builds two versions of busybox, one of which (busybox_unstripped) has extra information that various analysis tools can use. (This has nothing to do with CONFIG_DEBUG, leave that off when trying to optimize for size.)
The "make bloatcheck" option uses Matt Mackall's bloat-o-meter script to compare two versions of busybox (busybox_unstripped vs busybox_old), and report which symbols changed size and by how much. To use it, first build a base version with "make baseline". (This creates busybox_old, which should have the original sizes for comparison purposes.) Then build the new version with your changes and run "make bloatcheck" to see the size differences from the old version.
The first line of output has totals: how many symbols were added or removed, how many symbols grew or shrank, the number of bytes added and number of bytes removed by these changes, and finally the total number of bytes difference between the two files. The remaining lines show each individual symbol, the old and new sizes, and the increase or decrease in size (which results are sorted by).
The "make sizes" option produces raw symbol size information for busybox_unstripped. This is the output from the "nm --size-sort" command (see "man nm" for more information), and is the information bloat-o-meter parses to produce the comparison report above. For defconfig, this is a good way to find the largest symbols in the tree (which is a good place to start when trying to shrink the code). To take a closer look at individual applets, configure busybox with just one applet (run "make allnoconfig" and then switch on a single applet with menuconfig), and then use "make sizes" to see the size of that applet's components.
The "showasm" command (in the scripts directory) produces an assembly dump of a function, providing a closer look at what changed. Try "scripts/showasm busybox_unstripped" to list available symbols, and "scripts/showasm busybox_unstripped symbolname" to see the assembly for a sepecific symbol.
To add a new applet to busybox, first pick a name for the applet and a corresponding CONFIG_NAME. Then do this:
The standard we're paying attention to is the "Shell and Utilities" portion of the Open Group Base Standards (also known as the Single Unix Specification version 3 or SUSv3). Note that paying attention isn't necessarily the same thing as following it.
SUSv3 doesn't even mention things like init, mount, tar, or losetup, nor commonly used options like echo's '-e' and '-n', or sed's '-i'. Busybox is driven by what real users actually need, not the fact the standard believes we should implement ed or sccs. For size reasons, we're unlikely to include much internationalization support beyond UTF-8, and on top of all that, our configuration menu lets developers chop out features to produce smaller but very non-standard utilities.
Also, Busybox is aimed primarily at Linux. Unix standards are interesting because Linux tries to adhere to them, but portability to dozens of platforms is only interesting in terms of offering a restricted feature set that works everywhere, not growing dozens of platform-specific extensions. Busybox should be portable to all hardware platforms Linux supports, and any other similar operating systems that are easy to do and won't require much maintenance.
In practice, standards compliance tends to be a clean-up step once an applet is otherwise finished. When polishing and testing a busybox applet, we ensure we have at least the option of full standards compliance, or else document where we (intentionally) fall short.
Busybox is a Linux project, but that doesn't mean we don't have to worry about portability. First of all, there are different hardware platforms, different C library implementations, different versions of the kernel and build toolchain... The file "include/platform.h" exists to centralize and encapsulate various platform-specific things in one place, so most busybox code doesn't have to care where it's running.
To start with, Linux runs on dozens of hardware platforms. We try to test each release on x86, x86-64, arm, power pc, and mips. (Since qemu can handle all of these, this isn't that hard.) This means we have to care about a number of portability issues like endianness, word size, and alignment, all of which belong in platform.h. That header handles conditional #includes and gives us macros we can use in the rest of our code. At some point in the future we might grow a platform.c, possibly even a platform subdirectory. As long as the applets themselves don't have to care.
On a related note, we made the "default signedness of char varies" problem go away by feeding the compiler -funsigned-char. This gives us consistent behavior on all platforms, and defaults to 8-bit clean text processing (which gets us halfway to UTF-8 support). NOMMU support is less easily separated (see the tips section later in this document), but we're working on it.
Another type of portability is build environments: we unapologetically use a number of gcc and glibc extensions (as does the Linux kernel), but these have been picked up by packages like uClibc, TCC, and Intel's C Compiler. As for gcc, we take advantage of newer compiler optimizations to get the smallest possible size, but we also regression test against an older build environment using the Red Hat 9 image at "http://busybox.net/downloads/qemu". This has a 2.4 kernel, gcc 3.2, make 3.79.1, and glibc 2.3, and is the oldest build/deployment environment we still put any effort into maintaining. (If anyone takes an interest in older kernels you're welcome to submit patches, but the effort would probably be better spent trimming down the 2.6 kernel.) Older gcc versions than that are uninteresting since we now use c99 features, although tcc might be worth a look.
We also test busybox against the current release of uClibc. Older versions of uClibc aren't very interesting (they were buggy, and uClibc wasn't really usable as a general-purpose C library before version 0.9.26 anyway).
Other unix implementations are mostly uninteresting, since Linux binaries have become the new standard for portable Unix programs. Specifically, the ubiquity of Linux was cited as the main reason the Intel Binary Compatability Standard 2 died, by the standards group organized to name a successor to ibcs2: the 86open project. That project disbanded in 1999 with the endorsement of an existing standard: Linux ELF binaries. Since then, the major players at the time (such as AIX, Solaris, and FreeBSD) have all either grown Linux support or folded.
The major exceptions are newcomer MacOS X, some embedded environments (such as newlib+libgloss) which provide a posix environment but not a full Linux environment, and environments like Cygwin that provide only partial Linux emulation. Also, some embedded Linux systems run a Linux kernel but amputate things like the /proc directory to save space.
Supporting these systems is largely a question of providing a clean subset of BusyBox's functionality -- whichever applets can easily be made to work in that environment. Annotating the configuration system to indicate which applets require which prerequisites (such as procfs) is also welcome. Other efforts to support these systems (swapping #include files to build in different environments, adding adapter code to platform.h, adding more extensive special-case supporting infrastructure such as mount's legacy mtab support) are handled on a case-by-case basis. Support that can be cleanly hidden in platform.h is reasonably attractive, and failing that support that can be cleanly separated into a separate conditionally compiled file is at least worth a look. Special-case code in the body of an applet is something we're trying to avoid.
Various things busybox uses that aren't particularly well documented elsewhere.
Password fields in /etc/passwd and /etc/shadow are in a special format. If the first character isn't '$', then it's an old DES style password. If the first character is '$' then the password is actually three fields separated by '$' characters:
$type$salt$encrypted_password
The "type" indicates which encryption algorithm to use: 1 for MD5 and 2 for SHA1.
The "salt" is a bunch of ramdom characters (generally 8) the encryption algorithm uses to perturb the password in a known and reproducible way (such as by appending the random data to the unencrypted password, or combining them with exclusive or). Salt is randomly generated when setting a password, and then the same salt value is re-used when checking the password. (Salt is thus stored unencrypted.)
The advantage of using salt is that the same cleartext password encrypted with a different salt value produces a different encrypted value. If each encrypted password uses a different salt value, an attacker is forced to do the cryptographic math all over again for each password they want to check. Without salt, they could simply produce a big dictionary of commonly used passwords ahead of time, and look up each password in a stolen password file to see if it's a known value. (Even if there are billions of possible passwords in the dictionary, checking each one is just a binary search against a file only a few gigabytes long.) With salt they can't even tell if two different users share the same password without guessing what that password is and decrypting it. They also can't precompute the attack dictionary for a specific password until they know what the salt value is.
The third field is the encrypted password (plus the salt). For md5 this is 22 bytes.
The busybox function to handle all this is pw_encrypt(clear, salt) in "libbb/pw_encrypt.c". The first argument is the clear text password to be encrypted, and the second is a string in "$type$salt$password" format, from which the "type" and "salt" fields will be extracted to produce an encrypted value. (Only the first two fields are needed, the third $ is equivalent to the end of the string.) The return value is an encrypted password in /etc/passwd format, with all three $ separated fields. It's stored in a static buffer, 128 bytes long.
So when checking an existing password, if pw_encrypt(text, old_encrypted_password) returns a string that compares identical to old_encrypted_password, you've got the right password. When setting a new password, generate a random 8 character salt string, put it in the right format with sprintf(buffer, "$%c$%s", type, salt), and feed buffer as the second argument to pw_encrypt(text,buffer).
On systems that haven't got a Memory Management Unit, fork() is unreasonably expensive to implement (and sometimes even impossible), so a less capable function called vfork() is used instead. (Using vfork() on a system with an MMU is like pounding a nail with a wrench. Not the best tool for the job, but it works.)
Busybox hides the difference between fork() and vfork() in libbb/bb_fork_exec.c. If you ever want to fork and exec, use bb_fork_exec() (which returns a pid and takes the same arguments as execve(), although in this case envp can be NULL) and don't worry about it. This description is here in case you want to know why that does what it does.
Implementing fork() depends on having a Memory Management Unit. With an MMU then you can simply set up a second set of page tables and share the physical memory via copy-on-write. So a fork() followed quickly by exec() only copies a few pages of the parent's memory, just the ones it changes before freeing them.
With a very primitive MMU (using a base pointer plus length instead of page tables, which can provide virtual addresses and protect processes from each other, but no copy on write) you can still implement fork. But it's unreasonably expensive, because you have to copy all the parent process' memory into the new process (which could easily be several megabytes per fork). And you have to do this even though that memory gets freed again as soon as the exec happens. (This is not just slow and a waste of space but causes memory usage spikes that can easily cause the system to run out of memory.)
Without even a primitive MMU, you have no virtual addresses. Every process can reach out and touch any other process' memory, because all pointers are to physical addresses with no protection. Even if you copy a process' memory to new physical addresses, all of its pointers point to the old objects in the old process. (Searching through the new copy's memory for pointers and redirect them to the new locations is not an easy problem.)
So with a primitive or missing MMU, fork() is just not a good idea.
In theory, vfork() is just a fork() that writeably shares the heap and stack rather than copying it (so what one process writes the other one sees). In practice, vfork() has to suspend the parent process until the child does exec, at which point the parent wakes up and resumes by returning from the call to vfork(). All modern kernel/libc combinations implement vfork() to put the parent to sleep until the child does its exec. There's just no other way to make it work: the parent has to know the child has done its exec() or exit() before it's safe to return from the function it's in, so it has to block until that happens. In fact without suspending the parent there's no way to even store separate copies of the return value (the pid) from the vfork() call itself: both assignments write into the same memory location.
One way to understand (and in fact implement) vfork() is this: imagine the parent does a setjmp and then continues on (pretending to be the child) until the exec() comes around, then the _exec_ does the actual fork, and the parent does a longjmp back to the original vfork call and continues on from there. (It thus becomes obvious why the child can't return, or modify local variables it doesn't want the parent to see changed when it resumes.)
Note a common mistake: the need for vfork doesn't mean you can't have two processes running at the same time. It means you can't have two processes sharing the same memory without stomping all over each other. As soon as the child calls exec(), the parent resumes.
If the child's attempt to call exec() fails, the child should call _exit() rather than a normal exit(). This avoids any atexit() code that might confuse the parent. (The parent should never call _exit(), only a vforked child that failed to exec.)
(Now in theory, a nommu system could just copy the _stack_ when it forks (which presumably is much shorter than the heap), and leave the heap shared. Even with no MMU at all In practice, you've just wound up in a multi-threaded situation and you can't do a malloc() or free() on your heap without freeing the other process' memory (and if you don't have the proper locking for being threaded, corrupting the heap if both of you try to do it at the same time and wind up stomping on each other while traversing the free memory lists). The thing about vfork is that it's a big red flag warning "there be dragons here" rather than something subtle and thus even more dangerous.)
Busybox has special functions, bb_full_read() and bb_full_write(), to check that all the data we asked for got read or written. Is this a real world consideration? Try the following:
while true; do echo hello; sleep 1; done | tee out.txt
If tee is implemented with bb_full_read(), tee doesn't display output in real time but blocks until its entire input buffer (generally a couple kilobytes) is read, then displays it all at once. In that case, we _want_ the short read, for user interface reasons. (Note that read() should never return 0 unless it has hit the end of input, and an attempt to write 0 bytes should be ignored by the OS.)
As for short writes, play around with two processes piping data to each other on the command line (cat bigfile | gzip > out.gz) and suspend and resume a few times (ctrl-z to suspend, "fg" to resume). The writer can experience short writes, which are especially dangerous because if you don't notice them you'll discard data. They can also happen when a system is under load and a fast process is piping to a slower one. (Such as an xterm waiting on x11 when the scheduler decides X is being a CPU hog with all that text console scrolling...)
So will data always be read from the far end of a pipe at the same chunk sizes it was written in? Nope. Don't rely on that. For one counterexample, see rfc 896 for Nagle's algorithm, which waits a fraction of a second or so before sending out small amounts of data through a TCP/IP connection in case more data comes in that can be merged into the same packet. (In case you were wondering why action games that use TCP/IP set TCP_NODELAY to lower the latency on their their sockets, now you know.)
The downside of standard dynamic linking is that it results in self-modifying code. Although each executable's pages are mmaped() into a process' address space from the executable file and are thus naturally shared between processes out of the page cache, the library loader (ld-linux.so.2 or ld-uClibc.so.0) writes to these pages to supply addresses for relocatable symbols. This dirties the pages, triggering copy-on-write allocation of new memory for each processes' dirtied pages.
One solution to this is Position Independent Code (PIC), a way of linking a file so all the relocations are grouped together. This dirties fewer pages (often just a single page) for each process' relocations. The down side is this results in larger executables, which take up more space on disk (and a correspondingly larger space in memory). But when many copies of the same program are running, PIC dynamic linking trades a larger disk footprint for a smaller memory footprint, by sharing more pages.
A third solution is static linking. A statically linked program has no relocations, and thus the entire executable is shared between all running instances. This tends to have a significantly larger disk footprint, but on a system with only one or two executables, shared libraries aren't much of a win anyway.
You can tell the glibc linker to display debugging information about its relocations with the environment variable "LD_DEBUG". Try "LD_DEBUG=help /bin/true" for a list of commands. Learning to interpret "LD_DEBUG=statistics cat /proc/self/statm" could be interesting.
For more on this topic, here's Rich Felker:
Dynamic linking (without fixed load addresses) fundamentally requires at least one dirty page per dso that uses symbols. Making calls (but never taking the address explicitly) to functions within the same dso does not require a dirty page by itself, but will with ELF unless you use -Bsymbolic or hidden symbols when linking.
ELF uses significant additional stack space for the kernel to pass all the ELF data structures to the newly created process image. These are located above the argument list and environment. This normally adds 1 dirty page to the process size.
The ELF dynamic linker has its own data segment, adding one or more dirty pages. I believe it also performs relocations on itself.
The ELF dynamic linker makes significant dynamic allocations to manage the global symbol table and the loaded dso's. This data is never freed. It will be needed again if libdl is used, so unconditionally freeing it is not possible, but normal programs do not use libdl. Of course with glibc all programs use libdl (due to nsswitch) so the issue was never addressed.
ELF also has the issue that segments are not page-aligned on disk. This saves up to 4k on disk, but at the expense of using an additional dirty page in most cases, due to a large portion of the first data page being filled with a duplicate copy of the last text page.
The above is just a partial list of the tiny memory penalties of ELF dynamic linking, which eventually add up to quite a bit. The smallest I've been able to get a process down to is 8 dirty pages, and the above factors seem to mostly account for it (but some were difficult to measure).
The "linux" or "asm" directories of /usr/include contain Linux kernel headers, so that the C library can talk directly to the Linux kernel. In a perfect world, applications shouldn't include these headers directly, but we don't live in a perfect world.
For example, Busybox's losetup code wants linux/loop.c because nothing else
#defines the structures to call the kernel's loopback device setup ioctls.
Attempts to cut and paste the information into a local busybox header file
proved incredibly painful, because portions of the loop_info structure vary by
architecture, namely the type __kernel_dev_t has different sizes on alpha,
arm, x86, and so on. Meaning we either #include
This is aside from the fact that the relevant type defined in
posix_types.h was renamed to __kernel_old_dev_t during the 2.5 series, so
to cut and paste the structure into our header we have to #include
The BusyBox developers spent two years trying to figure
out a clean way to do all this. There isn't one. The losetup in the
util-linux package from kernel.org isn't doing it cleanly either, they just
hide the ugliness by nesting #include files. Their mount/loop.h
#includes "my_dev_t.h", which #includes
Just because directly #including kernel headers is sometimes unavoidable doesn't me we should include them when there's a better way to do it. However, block copying information out of the kernel headers is not a better way.
The following login accounts currently exist on busybox.net. (I.E. these people can commit patches into subversion for the BusyBox, uClibc, and buildroot projects.)
aldot :Bernhard Fischer andersen :Erik Andersen - uClibc and BuildRoot maintainer. bug1 :Glenn McGrath davidm :David McCullough gkajmowi :Garrett Kajmowicz - uClibc++ maintainer jbglaw :Jan-Benedict Glaw jocke :Joakim Tjernlund landley :Rob Landley - BusyBox maintainer lethal :Paul Mundt mjn3 :Manuel Novoa III osuadmin :osuadmin pgf :Paul Fox pkj :Peter Kjellerstedt prpplague :David Anders psm :Peter S. Mazinger russ :Russ Dill sandman :Robert Griebl sjhill :Steven J. Hill solar :Ned Ludd timr :Tim Riker tobiasa :Tobias Anderberg vapier :Mike Frysinger
The following accounts used to exist on busybox.net, but don't anymore so
I can't ask /etc/passwd for their names. Rob Wentworth
aaronl :Aaron Lehmann beppu :John Beppu dwhedon :David Whedon erik :Erik Andersen gfeldman :Gennady Feldman jimg :Jim Gleason kraai :Matt Kraai markw :Mark Whitley miles :Miles Bader proski :Pavel Roskin rjune :Richard June tausq :Randolph Chung vodz :Vladimir N. Oleynik