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 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 Migrate calloc() and bb_calloc() occurrences to bb_xzalloc().
16 Remove obsolete _() wrapper crud for internationalization we don't do.
17 Figure out where we need utf8 support, and add it.
20 The command shell situation is a big mess. We have three or four different
21 shells that don't really share any code, and the "standalone shell" doesn't
22 work all that well (especially not in a chroot environment), due to apps not
23 being reentrant. I'm writing a new shell (bbsh) to unify the various
24 shells and configurably add the minimal set of bash features people
25 actually use. The hardest part is it has to configure down as small as
26 lash while providing lash's features. The rest is easy in comparison.
28 Compression-side support.
31 Unify base64 handling.
32 There's base64 encoding and decoding going on in:
33 networking/wget.c:base64enc()
34 coreutils/uudecode.c:read_base64()
35 coreutils/uuencode.c:tbl_base64[]
36 networking/httpd.c:decodeBase64()
37 And probably elsewhere. That needs to be unified into libbb functions.
39 Look at the full Single Unix Specification version 3 (available online at
40 "http://www.opengroup.org/onlinepubs/009695399/nfindex.html") and
41 figure out which of our apps are compliant, and what we're missing that
42 we might actually care about.
44 Even better would be some kind of automated compliance test harness that
45 exercises each command line option and the various corner cases.
47 How much internationalization should we do?
49 The low hanging fruit is UTF-8 character set support. We should do this.
50 (Vodz pointed out the shell's cmdedit as needing work here. What else?)
52 We also have lots of hardwired english text messages. Consolidating this
53 into some kind of message table not only makes translation easier, but
54 also allows us to consolidate redundant (or close) strings.
56 We probably don't want to be bloated with locale support. (Not unless we
57 can cleanly export it from our underlying C library without having to
58 concern ourselves with it directly. Perhaps a few specific things like a
59 config option for "date" are low hanging fruit here?)
61 What level should things happen at? How much do we care about
62 internationalizing the text console when X11 and xterms are so much better
63 at it? (There's some infrastructure here we don't implement: The
64 "unicode_start" and "unicode_stop" shell scripts need "vt-is-UTF8" and a
65 --unicode option to loadkeys. That implies a real loadkeys/dumpkeys
66 implementation to replace loadkmap/dumpkmap. Plus messing with console font
67 loading. Is it worth it, or do we just say "use X"?)
69 Individual compilation of applets.
70 It would be nice if busybox had the option to compile to individual applets,
71 for people who want an alternate implementation less bloated than the gnu
72 utils (or simply with less political baggage), but without it being one big
75 Turning libbb into a real dll is another possibility, especially if libbb
76 could export some of the other library interfaces we've already more or less
77 got the code for (like zlib).
78 buildroot - Make a "dogfood" option
79 Busybox 1.1 will be capable of replacing most gnu packages for real world
80 use, such as developing software or in a live CD. It needs wider testing.
82 Busybox should now be able to replace bzip2, coreutils, e2fsprogs, file,
83 findutils, gawk, grep, inetutils, less, modutils, net-tools, patch, procps,
84 sed, shadow, sysklogd, sysvinit, tar, util-linux, and vim. The resulting
85 system should be self-hosting (I.E. able to rebuild itself from source
86 code). This means it would need (at least) binutils, gcc, and make, or
89 It would be a good "eating our own dogfood" test if buildroot had the option
90 of using a "make allyesconfig" busybox instead of the all of the above
91 packages. Anything that's wrong with the resulting system, we can fix. (It
92 would be nice to be able to upgrade busybox to be able to replace bash and
93 diffutils as well, but we're not there yet.)
95 One example of an existing system that does this already is Firmware Linux:
96 http://www.landley.net/code/firmware
98 Busybox should have a sample initramfs build script. This depends on
99 bbsh, mdev, and switch_root.
102 Bernhard Fischer <rep.nop@anon.at>:
104 make -j is broken, -j1 is forced atm
107 Collate BUFSIZ IOBUF_SIZE MY_BUF_SIZE PIPE_PROGRESS_SIZE BUFSIZE PIPESIZE
108 Use bb_common_bufsiz1?
114 doesn't understand (), lots of susv3 stuff.
117 Make sure we handle empty files properly:
118 From the patch man page:
120 you can remove a file by sending out a context diff that compares
121 the file to be deleted with an empty file dated the Epoch. The
122 file will be removed unless patch is conforming to POSIX and the
123 -E or --remove-empty-files option is not given.
126 Should have simple fuzz factor support to apply patches at an offset which
127 shouldn't take up too much space.
129 And while we're at it, a new patch filename quoting format is apparently
130 coming soon: http://marc.theaimsgroup.com/?l=git&m=112927316408690&w=2
133 It would be nice to have a man command. Not one that handles troff or
134 anything, just one that can handle preformatted ascii man pages, possibly
135 compressed. This could probably be a script in the extras directory that
136 calls cat/zcat/bzcat | less
138 (How doclifter might work into this is anybody's guess.)
144 turn FEATURE_DEBUG_OPT into ENABLE_FEATURE_CROND_DEBUG_OPT
146 Architectural issues:
148 bb_close() with fsync()
149 We should have a bb_close() in place of normal close, with a CONFIG_ option
150 to not just check the return value of close() for an error, but fsync().
151 Close can't reliably report anything useful because if write() accepted the
152 data then it either went out to the network or it's in cache or a pipe
153 buffer. Either way, there's no guarantee it'll make it to its final
154 destination before close() gets called, so there's no guarantee that any
155 error will be reported.
157 You need to call fsync() if you care about errors that occur after write(),
158 but that can have a big performance impact. So make it a config option.
161 Lots of archivers have the same general infrastructure. The directory
162 traversal code should be factored out, and the guts of each archiver could
163 be some setup code and a series of callbacks for "add this file",
164 "add this directory", "add this symlink" and so on.
166 This could clean up tar and zip, and make it cheaper to add cpio and ar
167 write support, and possibly even cheaply add things like mkisofs or
168 mksquashfs someday, if they become relevant.
171 Several existing applets (sort, vi, less...) read
172 a whole file into memory and act on it. There might be an opportunity
173 for shared code in there that could be moved into libbb...
176 We have a CONFIG_BUFFER mechanism that lets us select whether to do memory
177 allocation on the stack or the heap. Unfortunately, we're not using it much.
178 We need to audit our memory allocations and turn a lot of malloc/free calls
179 into RESERVE_CONFIG_BUFFER/RELEASE_CONFIG_BUFFER.
180 For a start, see e.g. make CFLAGS_EXTRA=-Wlarger-than-64
182 And while we're at it, many of the CONFIG_FEATURE_CLEAN_UP #ifdefs will be
183 optimized out by the compiler in the stack allocation case (since there's no
184 free for an alloca()), and this means that various cleanup loops that just
185 call free might also be optimized out by the compiler if written right, so
186 we can yank those #ifdefs too, and generally clean up the code.
188 Switch CONFIG_SYMBOLS to ENABLE_SYMBOLS
190 In busybox 1.0 and earlier, configuration was done by CONFIG_SYMBOLS
191 that were either defined or undefined to indicate whether the symbol was
192 selected in the .config file. They were used with #ifdefs, ala:
200 In 1.1, we have new ENABLE_SYMBOLS which are always defined (as 0 or 1),
201 meaning you can still use them for preprocessor tests by replacing
202 "#ifdef CONFIG_SYMBOL" with "#if ENABLE_SYMBOL". But more importantly, we
203 can use them as a true or false test in normal C code:
205 if (ENABLE_SYMBOL && other_test) {
209 (Optimizing away if() statements that resolve to a constant value
210 is known as "dead code elimination", an optimization so old and simple that
211 Turbo Pascal for DOS did it twenty years ago. Even modern mini-compilers
212 like the Tiny C Compiler (tcc) and the Small Device C Compiler (SDCC)
213 perform dead code elimination.)
215 Right now, busybox.h is #including both "config.h" (defining the
216 CONFIG_SYMBOLS) and "bb_config.h" (defining the ENABLE_SYMBOLS). At some
217 point in the future, it would be nice to wean ourselves off of the
218 CONFIG versions. (Among other things, some defective build environments
219 leak the Linux kernel's CONFIG_SYMBOLS into the system's standard #include
220 files. We've experienced collisions before.)
223 This is more an unresolved issue than a to-do item. More thought is needed.
225 Normally we rely on exit() to free memory, close files, and unmap segments
226 for us. This makes most calls to free(), close(), and unmap() optional in
227 busybox applets that don't intend to run for very long, and optional stuff
228 can be omitted to save size.
230 The idea was raised that we could simulate fork/exit with setjmp/longjmp
231 for _really_ brainless embedded systems, or speed up the standalone shell
232 by not forking. Doing so would require a reliable FEATURE_CLEAN_UP.
233 Unfortunately, this isn't as easy as it sounds.
235 The problem is, lots of things exit(), sometimes unexpectedly (xmalloc())
236 and sometimes reliably (bb_perror_msg_and_die() or show_usage()). This
237 jumps out of the normal flow control and bypasses any cleanup code we
238 put at the end of our applets.
240 It's possible to add hooks to libbb functions like xmalloc() and bb_xopen()
241 to add their entries to a linked list, which could be traversed and
242 freed/closed automatically. (This would need to be able to free just the
243 entries after a checkpoint to be usable for a forkless standalone shell.
244 You don't want to free the shell's own resources.)
246 Right now, FEATURE_CLEAN_UP is more or less a debugging aid, to make things
247 like valgrind happy. It's also documentation of _what_ we're trusting
248 exit() to clean up for us. But new infrastructure to auto-free stuff would
249 render the existing FEATURE_CLEAN_UP code redundant.
251 For right now, exit() handles it just fine.
256 watchdog.c could autodetect the timer duration via:
257 if(!ioctl (fd, WDIOC_GETTIMEOUT, &tmo)) timer_duration = 1 + (tmo / 2);
258 Unfortunately, that needs linux/watchdog.h and that contains unfiltered
259 kernel types on some distros, which breaks the build.
264 Replace deprecated functions.
268 sigblock(), siggetmask(), sigsetmask(), sigmask() -> sigprocmask et al
270 vdprintf() -> similar sized functionality