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 <rob@landley.net>:
10 Add BB_NOMMU to platform.h and migrate __uClinux__ tests to that.
11 #if defined __UCLIBC__ && !defined __ARCH_USE_MMU__
12 Add a libbb/platform.c
13 Implement fdprintf() for platforms that haven't got one.
14 Implement bb_realpath() that can handle NULL on non-glibc.
17 Migrate calloc() and bb_calloc() occurrences to bb_xzalloc().
18 Remove obsolete _() wrapper crud for internationalization we don't do.
19 Figure out where we need utf8 support, and add it.
22 The command shell situation is a big mess. We have three or four different
23 shells that don't really share any code, and the "standalone shell" doesn't
24 work all that well (especially not in a chroot environment), due to apps not
26 lash is phased out. hush can be configured down to be nearly as small,
29 General cleanup (should use ENABLE_FEATURE_INIT_SYSLOG and ENABLE_FEATURE_INIT_DEBUG).
31 busybox lacks a way to update module deps when running from firmware without the
32 use of the depmod.pl (perl is to bloated for most embedded setups) and or orig
33 modutils. The orig depmod is rather pointless to have to add to a firmware image
34 in when we already have a insmod/rmmod and friends.
36 Look at the full Single Unix Specification version 3 (available online at
37 "http://www.opengroup.org/onlinepubs/009695399/nfindex.html") and
38 figure out which of our apps are compliant, and what we're missing that
39 we might actually care about.
41 Even better would be some kind of automated compliance test harness that
42 exercises each command line option and the various corner cases.
44 How much internationalization should we do?
46 The low hanging fruit is UTF-8 character set support. We should do this.
47 (Vodz pointed out the shell's cmdedit as needing work here. What else?)
49 We also have lots of hardwired english text messages. Consolidating this
50 into some kind of message table not only makes translation easier, but
51 also allows us to consolidate redundant (or close) strings.
53 We probably don't want to be bloated with locale support. (Not unless we
54 can cleanly export it from our underlying C library without having to
55 concern ourselves with it directly. Perhaps a few specific things like a
56 config option for "date" are low hanging fruit here?)
58 What level should things happen at? How much do we care about
59 internationalizing the text console when X11 and xterms are so much better
60 at it? (There's some infrastructure here we don't implement: The
61 "unicode_start" and "unicode_stop" shell scripts need "vt-is-UTF8" and a
62 --unicode option to loadkeys. That implies a real loadkeys/dumpkeys
63 implementation to replace loadkmap/dumpkmap. Plus messing with console font
64 loading. Is it worth it, or do we just say "use X"?)
66 Individual compilation of applets.
67 It would be nice if busybox had the option to compile to individual applets,
68 for people who want an alternate implementation less bloated than the gnu
69 utils (or simply with less political baggage), but without it being one big
72 Turning libbb into a real dll is another possibility, especially if libbb
73 could export some of the other library interfaces we've already more or less
74 got the code for (like zlib).
75 buildroot - Make a "dogfood" option
76 Busybox 1.1 will be capable of replacing most gnu packages for real world
77 use, such as developing software or in a live CD. It needs wider testing.
79 Busybox should now be able to replace bzip2, coreutils, e2fsprogs, file,
80 findutils, gawk, grep, inetutils, less, modutils, net-tools, patch, procps,
81 sed, shadow, sysklogd, sysvinit, tar, util-linux, and vim. The resulting
82 system should be self-hosting (I.E. able to rebuild itself from source
83 code). This means it would need (at least) binutils, gcc, and make, or
86 It would be a good "eating our own dogfood" test if buildroot had the option
87 of using a "make allyesconfig" busybox instead of the all of the above
88 packages. Anything that's wrong with the resulting system, we can fix. (It
89 would be nice to be able to upgrade busybox to be able to replace bash and
90 diffutils as well, but we're not there yet.)
92 One example of an existing system that does this already is Firmware Linux:
93 http://www.landley.net/code/firmware
95 Busybox should have a sample initramfs build script. This depends on
96 bbsh, mdev, and switch_root.
98 Write a mkdep that doesn't segfault if there's a directory it doesn't
99 have permission to read, isn't based on manually editing the output of
100 lexx and yacc, doesn't make such a mess under include/config, etc.
101 Group globals into unions of structures.
102 Go through and turn all the global and static variables into structures,
103 and have all those structures be in a big union shared between processes,
104 so busybox uses less bss. (This is a big win on nommu machines.) See
105 sed.c and mdev.c for examples.
106 Go through bugs.busybox.net and close out all of that somehow.
107 This one's open to everybody, but I'll wind up doing it...
110 Bernhard Fischer <busybox@busybox.net> suggests to look at these:
113 Cleanup any big users
115 Facilitate applet PROTOTYPES to provide means for having applets that
116 do a) not take any arguments b) need only one of argc or argv c) need
117 both argc and argv. All of these three options should go for the most
118 feature complete denominator.
119 Collate BUFSIZ IOBUF_SIZE MY_BUF_SIZE PIPE_PROGRESS_SIZE BUFSIZE PIPESIZE
120 make bb_common_bufsiz1 configurable, size wise.
121 make pipesize configurable, size wise.
122 Use bb_common_bufsiz1 throughout applets!
128 doesn't understand (), lots of susv3 stuff.
131 Make sure we handle empty files properly:
132 From the patch man page:
134 you can remove a file by sending out a context diff that compares
135 the file to be deleted with an empty file dated the Epoch. The
136 file will be removed unless patch is conforming to POSIX and the
137 -E or --remove-empty-files option is not given.
140 Should have simple fuzz factor support to apply patches at an offset which
141 shouldn't take up too much space.
143 And while we're at it, a new patch filename quoting format is apparently
144 coming soon: http://marc.theaimsgroup.com/?l=git&m=112927316408690&w=2
147 Add support for both RSS and VSIZE rather than just one or the other.
148 Or make it a build option.
151 It would be nice to have a man command. Not one that handles troff or
152 anything, just one that can handle preformatted ascii man pages, possibly
153 compressed. This could probably be a script in the extras directory that
154 calls cat/zcat/bzcat | less
156 (How doclifter might work into this is anybody's guess.)
162 stty's visible() function and catv's guts are identical. Merge them into
163 an appropriate libbb function.
166 Several duplicate users of: grep -r "1024\*1024" * -B2 -A1
167 Merge to a single size_suffixes[] in libbb.
168 Users: head tail od_bloaty hexdump and (partially as it wouldn't hurt) svlogd
171 ./busybox tail -f foo.c~ TODO
172 should not print fmt=header_fmt for subsequent date >> TODO; i.e. only
173 fmt+ if another (not the current) file did change
175 Architectural issues:
177 bb_close() with fsync()
178 We should have a bb_close() in place of normal close, with a CONFIG_ option
179 to not just check the return value of close() for an error, but fsync().
180 Close can't reliably report anything useful because if write() accepted the
181 data then it either went out to the network or it's in cache or a pipe
182 buffer. Either way, there's no guarantee it'll make it to its final
183 destination before close() gets called, so there's no guarantee that any
184 error will be reported.
186 You need to call fsync() if you care about errors that occur after write(),
187 but that can have a big performance impact. So make it a config option.
190 Lots of archivers have the same general infrastructure. The directory
191 traversal code should be factored out, and the guts of each archiver could
192 be some setup code and a series of callbacks for "add this file",
193 "add this directory", "add this symlink" and so on.
195 This could clean up tar and zip, and make it cheaper to add cpio and ar
196 write support, and possibly even cheaply add things like mkisofs or
197 mksquashfs someday, if they become relevant.
200 Several existing applets (sort, vi, less...) read
201 a whole file into memory and act on it. There might be an opportunity
202 for shared code in there that could be moved into libbb...
205 We have a CONFIG_BUFFER mechanism that lets us select whether to do memory
206 allocation on the stack or the heap. Unfortunately, we're not using it much.
207 We need to audit our memory allocations and turn a lot of malloc/free calls
208 into RESERVE_CONFIG_BUFFER/RELEASE_CONFIG_BUFFER.
209 For a start, see e.g. make EXTRA_CFLAGS=-Wlarger-than-64
211 And while we're at it, many of the CONFIG_FEATURE_CLEAN_UP #ifdefs will be
212 optimized out by the compiler in the stack allocation case (since there's no
213 free for an alloca()), and this means that various cleanup loops that just
214 call free might also be optimized out by the compiler if written right, so
215 we can yank those #ifdefs too, and generally clean up the code.
217 Switch CONFIG_SYMBOLS to ENABLE_SYMBOLS
219 In busybox 1.0 and earlier, configuration was done by CONFIG_SYMBOLS
220 that were either defined or undefined to indicate whether the symbol was
221 selected in the .config file. They were used with #ifdefs, ala:
229 In 1.1, we have new ENABLE_SYMBOLS which are always defined (as 0 or 1),
230 meaning you can still use them for preprocessor tests by replacing
231 "#ifdef CONFIG_SYMBOL" with "#if ENABLE_SYMBOL". But more importantly, we
232 can use them as a true or false test in normal C code:
234 if (ENABLE_SYMBOL && other_test) {
238 (Optimizing away if() statements that resolve to a constant value
239 is known as "dead code elimination", an optimization so old and simple that
240 Turbo Pascal for DOS did it twenty years ago. Even modern mini-compilers
241 like the Tiny C Compiler (tcc) and the Small Device C Compiler (SDCC)
242 perform dead code elimination.)
244 Right now, busybox.h is #including both "config.h" (defining the
245 CONFIG_SYMBOLS) and "bb_config.h" (defining the ENABLE_SYMBOLS). At some
246 point in the future, it would be nice to wean ourselves off of the
247 CONFIG versions. (Among other things, some defective build environments
248 leak the Linux kernel's CONFIG_SYMBOLS into the system's standard #include
249 files. We've experienced collisions before.)
252 This is more an unresolved issue than a to-do item. More thought is needed.
254 Normally we rely on exit() to free memory, close files, and unmap segments
255 for us. This makes most calls to free(), close(), and unmap() optional in
256 busybox applets that don't intend to run for very long, and optional stuff
257 can be omitted to save size.
259 The idea was raised that we could simulate fork/exit with setjmp/longjmp
260 for _really_ brainless embedded systems, or speed up the standalone shell
261 by not forking. Doing so would require a reliable FEATURE_CLEAN_UP.
262 Unfortunately, this isn't as easy as it sounds.
264 The problem is, lots of things exit(), sometimes unexpectedly (xmalloc())
265 and sometimes reliably (bb_perror_msg_and_die() or show_usage()). This
266 jumps out of the normal flow control and bypasses any cleanup code we
267 put at the end of our applets.
269 It's possible to add hooks to libbb functions like xmalloc() and xopen()
270 to add their entries to a linked list, which could be traversed and
271 freed/closed automatically. (This would need to be able to free just the
272 entries after a checkpoint to be usable for a forkless standalone shell.
273 You don't want to free the shell's own resources.)
275 Right now, FEATURE_CLEAN_UP is more or less a debugging aid, to make things
276 like valgrind happy. It's also documentation of _what_ we're trusting
277 exit() to clean up for us. But new infrastructure to auto-free stuff would
278 render the existing FEATURE_CLEAN_UP code redundant.
280 For right now, exit() handles it just fine.
285 watchdog.c could autodetect the timer duration via:
286 if(!ioctl (fd, WDIOC_GETTIMEOUT, &tmo)) timer_duration = 1 + (tmo / 2);
287 Unfortunately, that needs linux/watchdog.h and that contains unfiltered
288 kernel types on some distros, which breaks the build.
290 use bb_error_msg where appropriate: See
291 egrep "(printf.*\([[:space:]]*(stderr|2)|[^_]write.*\([[:space:]]*(stderr|2))"
293 use bb_perror_msg where appropriate: See
296 Remove superfluous fmt occurances: e.g.
297 fprintf(stderr, "%s: %s not found\n", "unalias", *argptr);
298 -> fprintf(stderr, "unalias: %s not found\n", *argptr);
300 possible code duplication ingroup() and is_a_group_member()
302 Move __get_hz() to a better place and (re)use it in route.c, ash.c, msh.c
305 Alot of duplication that wants cleanup.
311 Replace deprecated functions.
315 sigblock(), siggetmask(), sigsetmask(), sigmask() -> sigprocmask et al
317 vdprintf() -> similar sized functionality