3 Stuff that needs to be done. This is organized by who plans to get around to
4 doing it eventually, but that doesn't mean they "own" the item. If you want to
5 do one of these bounce an email off the person it's listed under to see if they
6 have any suggestions how they plan to go about it, and to minimize conflicts
7 between your work and theirs. But otherwise, all of these are fair game.
9 Rob Landley suggested these:
10 Add a libbb/platform.c
11 Implement fdprintf() for platforms that haven't got one.
12 Implement bb_realpath() that can handle NULL on non-glibc.
15 Remove obsolete _() wrapper crud for internationalization we don't do.
16 Figure out where we need utf8 support, and add it.
19 The command shell situation is a mess. We have two different
20 shells that don't really share any code, and the "standalone shell" doesn't
21 work all that well (especially not in a chroot environment), due to apps not
25 Look at the full Single Unix Specification version 3 (available online at
26 "http://www.opengroup.org/onlinepubs/009695399/nfindex.html") and
27 figure out which of our apps are compliant, and what we're missing that
28 we might actually care about.
30 Even better would be some kind of automated compliance test harness that
31 exercises each command line option and the various corner cases.
34 How much internationalization should we do?
36 The low hanging fruit is UTF-8 character set support. We should do this.
37 (Vodz pointed out the shell's cmdedit as needing work here. What else?)
39 We also have lots of hardwired english text messages. Consolidating this
40 into some kind of message table not only makes translation easier, but
41 also allows us to consolidate redundant (or close) strings.
43 We probably don't want to be bloated with locale support. (Not unless we
44 can cleanly export it from our underlying C library without having to
45 concern ourselves with it directly. Perhaps a few specific things like a
46 config option for "date" are low hanging fruit here?)
48 What level should things happen at? How much do we care about
49 internationalizing the text console when X11 and xterms are so much better
50 at it? (There's some infrastructure here we don't implement: The
51 "unicode_start" and "unicode_stop" shell scripts need "vt-is-UTF8" and a
52 --unicode option to loadkeys. That implies a real loadkeys/dumpkeys
53 implementation to replace loadkmap/dumpkmap. Plus messing with console font
54 loading. Is it worth it, or do we just say "use X"?)
56 Individual compilation of applets.
57 It would be nice if busybox had the option to compile to individual applets,
58 for people who want an alternate implementation less bloated than the gnu
59 utils (or simply with less political baggage), but without it being one big
62 Turning libbb into a real dll is another possibility, especially if libbb
63 could export some of the other library interfaces we've already more or less
64 got the code for (like zlib).
65 buildroot - Make a "dogfood" option
66 Busybox 1.1 will be capable of replacing most gnu packages for real world
67 use, such as developing software or in a live CD. It needs wider testing.
69 Busybox should now be able to replace bzip2, coreutils, e2fsprogs, file,
70 findutils, gawk, grep, inetutils, less, modutils, net-tools, patch, procps,
71 sed, shadow, sysklogd, sysvinit, tar, util-linux, and vim. The resulting
72 system should be self-hosting (I.E. able to rebuild itself from source
73 code). This means it would need (at least) binutils, gcc, and make, or
76 It would be a good "eating our own dogfood" test if buildroot had the option
77 of using a "make allyesconfig" busybox instead of the all of the above
78 packages. Anything that's wrong with the resulting system, we can fix. (It
79 would be nice to be able to upgrade busybox to be able to replace bash and
80 diffutils as well, but we're not there yet.)
82 One example of an existing system that does this already is Firmware Linux:
83 http://www.landley.net/code/firmware
85 Busybox should have a sample initramfs build script. This depends on
86 bbsh, mdev, and switch_root.
88 Write a mkdep that doesn't segfault if there's a directory it doesn't
89 have permission to read, isn't based on manually editing the output of
90 lexx and yacc, doesn't make such a mess under include/config, etc.
91 Group globals into unions of structures.
92 Go through and turn all the global and static variables into structures,
93 and have all those structures be in a big union shared between processes,
94 so busybox uses less bss. (This is a big win on nommu machines.) See
95 sed.c and mdev.c for examples.
96 Go through bugs.busybox.net and close out all of that somehow.
97 This one's open to everybody, but I'll wind up doing it...
100 Bernhard Reutner-Fischer <busybox@busybox.net> suggests to look at these:
103 Cleanup any big users
104 Collate BUFSIZ IOBUF_SIZE MY_BUF_SIZE PIPE_PROGRESS_SIZE BUFSIZE PIPESIZE
105 make bb_common_bufsiz1 configurable, size wise.
106 make pipesize configurable, size wise.
107 Use bb_common_bufsiz1 throughout applets!
113 Make sure we handle empty files properly:
114 From the patch man page:
116 you can remove a file by sending out a context diff that compares
117 the file to be deleted with an empty file dated the Epoch. The
118 file will be removed unless patch is conforming to POSIX and the
119 -E or --remove-empty-files option is not given.
122 Should have simple fuzz factor support to apply patches at an offset which
123 shouldn't take up too much space.
125 And while we're at it, a new patch filename quoting format is apparently
126 coming soon: http://marc.theaimsgroup.com/?l=git&m=112927316408690&w=2
132 stty's visible() function and catv's guts are identical. Merge them into
133 an appropriate libbb function.
136 Several duplicate users of: grep -r "1024\*1024" * -B2 -A1
137 Merge to a single size_suffixes[] in libbb.
138 Users: head tail od_bloaty hexdump and (partially as it wouldn't hurt) svlogd
141 ./busybox tail -f foo.c~ TODO
142 should not print fmt=header_fmt for subsequent date >> TODO; i.e. only
143 fmt+ if another (not the current) file did change
145 Architectural issues:
147 bb_close() with fsync()
148 We should have a bb_close() in place of normal close, with a CONFIG_ option
149 to not just check the return value of close() for an error, but fsync().
150 Close can't reliably report anything useful because if write() accepted the
151 data then it either went out to the network or it's in cache or a pipe
152 buffer. Either way, there's no guarantee it'll make it to its final
153 destination before close() gets called, so there's no guarantee that any
154 error will be reported.
156 You need to call fsync() if you care about errors that occur after write(),
157 but that can have a big performance impact. So make it a config option.
160 Lots of archivers have the same general infrastructure. The directory
161 traversal code should be factored out, and the guts of each archiver could
162 be some setup code and a series of callbacks for "add this file",
163 "add this directory", "add this symlink" and so on.
165 This could clean up tar and zip, and make it cheaper to add cpio and ar
166 write support, and possibly even cheaply add things like mkisofs or
167 mksquashfs someday, if they become relevant.
170 Several existing applets (sort, vi, less...) read
171 a whole file into memory and act on it. Use open_read_close().
174 We have a CONFIG_BUFFER mechanism that lets us select whether to do memory
175 allocation on the stack or the heap. Unfortunately, we're not using it much.
176 We need to audit our memory allocations and turn a lot of malloc/free calls
177 into RESERVE_CONFIG_BUFFER/RELEASE_CONFIG_BUFFER.
178 For a start, see e.g. make EXTRA_CFLAGS=-Wlarger-than-64
180 And while we're at it, many of the CONFIG_FEATURE_CLEAN_UP #ifdefs will be
181 optimized out by the compiler in the stack allocation case (since there's no
182 free for an alloca()), and this means that various cleanup loops that just
183 call free might also be optimized out by the compiler if written right, so
184 we can yank those #ifdefs too, and generally clean up the code.
186 Switch CONFIG_SYMBOLS to ENABLE_SYMBOLS
188 In busybox 1.0 and earlier, configuration was done by CONFIG_SYMBOLS
189 that were either defined or undefined to indicate whether the symbol was
190 selected in the .config file. They were used with #ifdefs, ala:
198 In 1.1, we have new ENABLE_SYMBOLS which are always defined (as 0 or 1),
199 meaning you can still use them for preprocessor tests by replacing
200 "#ifdef CONFIG_SYMBOL" with "#if ENABLE_SYMBOL". But more importantly, we
201 can use them as a true or false test in normal C code:
203 if (ENABLE_SYMBOL && other_test) {
207 (Optimizing away if() statements that resolve to a constant value
208 is known as "dead code elimination", an optimization so old and simple that
209 Turbo Pascal for DOS did it twenty years ago. Even modern mini-compilers
210 like the Tiny C Compiler (tcc) and the Small Device C Compiler (SDCC)
211 perform dead code elimination.)
213 Right now, busybox.h is #including both "config.h" (defining the
214 CONFIG_SYMBOLS) and "bb_config.h" (defining the ENABLE_SYMBOLS). At some
215 point in the future, it would be nice to wean ourselves off of the
216 CONFIG versions. (Among other things, some defective build environments
217 leak the Linux kernel's CONFIG_SYMBOLS into the system's standard #include
218 files. We've experienced collisions before.)
221 This is more an unresolved issue than a to-do item. More thought is needed.
223 Normally we rely on exit() to free memory, close files and unmap segments
224 for us. This makes most calls to free(), close(), and unmap() optional in
225 busybox applets that don't intend to run for very long, and optional stuff
226 can be omitted to save size.
228 The idea was raised that we could simulate fork/exit with setjmp/longjmp
229 for _really_ brainless embedded systems, or speed up the standalone shell
230 by not forking. Doing so would require a reliable FEATURE_CLEAN_UP.
231 Unfortunately, this isn't as easy as it sounds.
233 The problem is, lots of things exit(), sometimes unexpectedly (xmalloc())
234 and sometimes reliably (bb_perror_msg_and_die() or show_usage()). This
235 jumps out of the normal flow control and bypasses any cleanup code we
236 put at the end of our applets.
238 It's possible to add hooks to libbb functions like xmalloc() and xopen()
239 to add their entries to a linked list, which could be traversed and
240 freed/closed automatically. (This would need to be able to free just the
241 entries after a checkpoint to be usable for a forkless standalone shell.
242 You don't want to free the shell's own resources.)
244 Right now, FEATURE_CLEAN_UP is more or less a debugging aid, to make things
245 like valgrind happy. It's also documentation of _what_ we're trusting
246 exit() to clean up for us. But new infrastructure to auto-free stuff would
247 render the existing FEATURE_CLEAN_UP code redundant.
249 For right now, exit() handles it just fine.
253 watchdog.c could autodetect the timer duration via:
254 if(!ioctl (fd, WDIOC_GETTIMEOUT, &tmo)) timer_duration = 1 + (tmo / 2);
255 Unfortunately, that needs linux/watchdog.h and that contains unfiltered
256 kernel types on some distros, which breaks the build.
258 use bb_error_msg where appropriate: See
259 egrep "(printf.*\([[:space:]]*(stderr|2)|[^_]write.*\([[:space:]]*(stderr|2))"
261 use bb_perror_msg where appropriate: See
264 possible code duplication ingroup() and is_a_group_member()
266 Move __get_hz() to a better place and (re)use it in route.c, ash.c
269 Alot of duplication that wants cleanup.
271 in_ether duplicated in network/{interface,ifconfig}.c
273 unify progress_meter. wget, flash_eraseall, pipe_progress, fbsplash, setfiles.
275 support start-stop-daemon -d <chdir-path>
279 Replace deprecated functions.
282 vdprintf() -> similar sized functionality
285 (TODO list after discussion 11.05.2009)
288 tc/brctl seem like fairly large things to try and tackle in your timeframe,
289 and i think people have posted attempts in the past. Adding additional
290 options to ip though seems reasonable.
292 * add tests for some applets
294 * implement POSIX utilities and audit them for POSIX conformance. then
295 audit them for GNU conformance. then document all your findings in a new
296 doc/conformance.txt file while perhaps implementing some of the missing
298 you can find the latest POSIX documentation (1003.1-2008) here:
299 http://www.opengroup.org/onlinepubs/9699919799/
300 and the complete list of all utilities that POSIX covers:
301 http://www.opengroup.org/onlinepubs/9699919799/idx/utilities.html
302 The first step would to generate a file/matrix what is already archived
305 * ntpdate/ntpd (see ntpclient and openntp for examples)
309 * rpcbind (former portmap) or equivalent
310 so that we don't have to use -o nolock on nfs mounts
312 * check IPV6 compliance
314 * generate a mini example using kernel+busybox only (+libc) for example
316 * more support for advanced linux 2.6.x features, see: iotop
317 most likely there is more
319 * even more support for statistics: mpstat, iostat, powertop....
324 Unicode support uses libc multibyte functions if LOCALE_SUPPORT is on
325 (in this case, the code will also support many more encodings),
326 or uses a limited subset of re-implemented multibyte functions
327 which only understand "one byte == one char" and unicode.
328 This is useful if you build against uclibc with locale support disabled.
330 Unicode-dependent applets must call check_unicode_in_env() when they
333 Applet code may conditionalize on FEATURE_ASSUME_UNICODE
334 in order to use more efficient code if unicode support is not requested.
336 Available functions (if you need more, implement them in libbb/unicode.c
337 so that they work without LOCALE_SUPPORT too):
339 int bb_mbstrlen(str) - multibyte-aware strlen
340 size_t mbstowcs(wdest, src, n)
341 size_t wcstombs(dest, wsrc, n)
342 size_t wcrtomb(str, wc, wstate)
347 Applets which only need to align columns on screen correctly:
349 ls - already done, use source as an example
354 Applets which need to account for Unicode chars
355 while processing the output:
361 cut (-b and -c are currently the same, needs fixing)
363 These applets need to ensure that unicode input
364 is handled correctly (say, <unicode><backspace> sequence):
368 unzip (overwrite prompt)
370 Viewers/editors are more difficult (many cases to get right).
371 libbb/lineedit.c is an example how to do it:
378 Probably needs some specialized work: