1TODO list for every uClibc release:
2-------------------------------------------------
3    *) Test cris, i386, mips, mipsel, sh, x86_64, arm, armeb, and powerpc
4	with the latest LTP testsuite.  Fix any regressions and post LTP
5	testsuite results for each architecture on uclibc.org.
6    *) Audit header files.  Remove prototypes for all functions that
7	are not supported -- especially needed for the libm headers.
8    *) Audit header files.  When options are disabled, also disable
9	them in the include files as well by checking for the proper
10	define from include/bits/uClibc_config.h (pulled in from features.h)
11
12TODO list for the uClibc 1.0.0 release:
13-------------------------------------------------
14    *) NPTL
15       support arches: (o todo; + done)
16	o alpha
17	+ arm
18	o avr32
19	o bfin
20	o cris
21	o e1
22	o frv
23	o h8300
24	o hppa
25	+ i386
26	o i960
27	o ia64
28	o m68k
29	o microblaze
30	+ mips
31	o nios
32	o nios2
33	+ powerpc
34	+ sh
35	+ sh64
36	o sparc
37	o v850
38	o vax
39	o x86_64
40	o xtensa
41	o ...
42    *) Go through SUSv4
43       TOC: http://www.opengroup.org/onlinepubs/9699919799/xrat/contents.html
44       shell (busybox): http://www.opengroup.org/onlinepubs/9699919799/xrat/V4_xcu_chap01.html#tag_22_01_01
45       interface:
46         http://www.opengroup.org/onlinepubs/9699919799/xrat/V4_xbd_chap13.html#tag_21_13_02
47         http://www.opengroup.org/onlinepubs/9699919799/xrat/V4_xsh_chap01.html#tag_23_01_01
48    *) glob / fnmatch tests fail
49    *) regex should pass AT&T conformance tests
50    *) Finish hiding uClibc internal symbols from our exported namespace
51    *) Add system for generating minimal system call asm wrappers that abuse
52	the C ABI to minimize amount of register/stack manipulation
53    *) Explicity add some sortof CONFIG_LINUX_2_2_ABI, CONFIG_LINUX_2_4_ABI
54	and CONFIG_LINUX_2_6_ABI type options, rather than having the abi
55	depend on the selected set of kernel headers.  This will likely also
56	require that we provide a set of kernel headers (probably a derivitive
57	of Mazur's linux-libc-headers) that can supplies the 2.2, 2.4, and 2.6
58	kernel abi.
59    *) Documentation updates:
60	    *) Write a uClibc HOWTO document
61	    *) Update README document
62	    *) Update INSTALL document
63	    *) Update docs/Glibc_vs_uClibc_Differences.txt document
64		and fully document all differences between the feature
65		set of uClibc and glibc.
66	    *) Update docs/uClibc_vs_SuSv3.txt document
67	    *) Update docs/threads.txt document
68	    *) Write man pages for ldd and ldconfig utility binaries
69    *) Implement some mechanism (perhaps encoded in the .so name,
70	perhaps using an abi tag magically embedded into each object)
71	for flagging config options that break the ABI.  Options
72	such as UCLIBC_HAS_SOFT_FLOAT, UCLIBC_HAS_THREADS, UCLIBC_HAS_LOCALE,
73	and perhaps others (finalize list) produce a lib with a differing
74	ABI.  Make it so apps cannot use an ABI mis-matched uClibc.
75	This is most easily done using symbol versioning...
76    *) Implement the long double versions of math funcs on interrested
77        arches using wrappers on top of the double versions (size / precision
78	trade off where size clearly wins).
79    *) Make all small objects (>~50 bytes) into either inlines or
80	into a static library
81    *) Cleanup / rewrite sysconf.c.  It should get some information
82	from ldso (such as HZ).  Other stuff it currently just makes
83	up, which is obviously wrong.  Also bits/uClibc_clk_tck.h
84	needs to be updated at the same time to get proper HZ values.
85    *) poll emulation using select() for old 2.0.x uClinux kernels
86	in libc/sysdeps/linux/common/poll.c fails some python self-tests.
87	Of course, modern systems using the actuall poll() syscall work fine.
88    *) Cleanup/scrub all the Makefile copyright junk
89    *) Fix dlopen, for both static and dynamic cases, and make it
90	fully comply with SuSv3
91    *) check if sched_getcpu is only pulled in where it should be
92	*) remove HANDLE_OLDER_RLIMIT, NEW_GETRLIMIT (conditionalize on
93	ugetrlimit, it seems, else use the modern variant)
94	*) simplify exec*() in the light of execvpe (perhaps single internal
95	impl); USE_GNU for execvpe (and other GNU extensions, see psm).
96
97
98TODO list for AFTER the uClibc 1.0.0 release:
99-------------------------------------------------
100    *) Add support for 64bit time (to deal withhttp://2038bug.com/):
101	date -s 011903142038 sets epoch to roll over in a few seconds (2^31)
102    *) Add support for Linux 2.6.x fast vsyscalls
103    *) Enable pristine source tree builds
104    *) Fix regex so it isn't so stinking big
105    *) Fix glob so it isn't so stinking big
106    *) run 'nm -D --size-sort -t d libuClibc-0.9.26.so' and work on the
107	biggest things (i.e. stuff at the end of the list) to make
108	them smaller.
109    *) Fix dlopen/dlsym/dlclose locking to more fine grained or use RCU
110    <more wishlist items here>
111
112
113Arch specific TODO:
114-------------------------------------------------
115  alpha:
116    ldso is not implemented
117  arm:
118    update crt1 code again for nommu (http://bugs.uclibc.org/view.php?id=538)
119    recruit jbowler and integrate thumb support
120  frv:
121    needs to be updated for the new ldso changes and unification of the
122    uClibc_main funcs (__uClibc_start_main doesnt exist anymore)
123  hppa:
124    errno.c test segfaults causes segfault in clone.S (seems to be in glibc too)
125  i386:
126    add support for fast system calls
127  ia64:
128    ldso is not implemented
129  m68k:
130    ldso lazy relocation doesnt work
131  sh64:
132    ldso is reasonably broken, presently requiring additional
133    coaxing/threatening.
134  sparc:
135    need a sigaction.c since common one doesnt work (signal tests)
136    ldso needs to be updated since it's totally broken atm
137  vax:
138    pthread support, linktime warning support (implies GAS patches),
139    general touchups, testing, ldso.
140
141
142-----------------------------------------------------------------------------
143Manuel's todo:
144
145  1) Little things that need fixing:
146  ----------------------------------
147  a) Fix bug in *printf: outdigit precison bug
148  b) Check that gnu/bsd extension members tm_gmtoff and tm_zone in struct tm
149     are respected where they should be.
150  c) Implement the obstack printf funcs for glibc compat.
151  d) Implement glibc 'a' flag for scanf string conversions.
152  e) Allow use of the older non-table-based ctype functions when using
153     stub locale support. (smaller)
154  f) __drand48_iterate should be void
155  g) alphasort vs. versionsort. The former seems to be SVID, the latter GNU
156     i.e. reverse to what we currently do. The latter is unimplemented.
157  h) ponder removal/configs to turn off: __xpg_*, bsd_signal, dysize,
158     getw/putw, utimes,
159
160  2) Additional str{f|p}time issues.
161  ----------------------------------
162  a) Spacing issue wrt strptime.
163  b) Support locale specific alternate digits.  (data is in place)
164  c) Support locale era in year designations.   (data is in place)
165  d) Deal with mb format string issues in strftime.
166  e) Implement wcsftime.
167
168  3) Other locale issues (my implementation):
169  -------------------------------------------
170  a) Do a little more clean up of ctype and wctype.
171  b) Rework of the locale data organization to make using locales reasonable
172     when staticly linking.  (mmap)
173  c) Rewrite the locale data generation tools to process the text specifications
174     rather than relying on glibc.
175  d) Adapt regex lib to use my collation data and add the necessary collating
176     item tables to support SUSv3 required features.
177  e) transliteration of unsupported wchars in 8-bit locales (like glibc).
178  f) Support ISO/IEC 14652 draft locale extensions (LC_PAPER, etc).
179  g) Implement strfrom.
180  h) Shift-state codeset locale support?
181
182  4) Misc:
183  --------
184  a) Port uClibc to other OSs (including elks), or even bare metal (libgloss).
185  b) Write a space-efficient gettext substitute, to avoid storing large amounts
186     of redundant data.
187
188
189