Annotation of dietlibc/CHANGES, revision 1.59
1.34 fefe 1: 0.7.3:
2: vsnprintf(0,...) estimated the length of strings with a padding width
3: incorrectly.
1.35 fefe 4: new reboot() from Olaf.
1.36 fefe 5: added prototypes for getsockopt and setsockopt.
1.39 fefe 6: added experimental sem* and shm*
1.38 fefe 7: define EXIT_SUCCESS and EXIT_FAILURE
1.40 fefe 8: added setvbuf (only changes mode, not buffer)
1.41 fefe 9: added dietwarning.h (woohoo, linker warnings!) and used it to warn of
10: insecurity (system, sprintf) or bloat (printf and stdio).
1.42 fefe 11: added qsort() and bsearch(). Benchmark on 1000 random ints:
12: glibc: 679414 cycles qsort and 3352 cycles bsearch
13: diet libc: 300010 cycles qsort, 553 cycles bsearch
14: I have no idea why glibc is so slow.
1.43 fefe 15: added __isnan (references by libm.a from glibc).
16: added popen() and pclose(). Now the diet libc can link slsh (from
17: S-Lang!)
1.44 fefe 18: corrected usleep. It was more of an msleep until now.
1.45 fefe 19: corrected __dtostr (it looped when trying to format "0.0").
20: several fixes to stdio (thanks to S-Lang's excellent test suite!)
1.46 fefe 21: added ungetc (does it work?)
1.47 fefe 22: added inet_aton (yuck!)
1.50 fefe 23: added gethostbyname, gethostbyaddr, gethostbyname_r and gethostbyaddr_r
1.48 fefe 24: added h_errno and __h_errno_location
25: added inet_ntoa
1.49 fefe 26: added setpgrp and getpgrp, strsignal and execv (thanks, Olaf)
1.51 fefe 27: add bcopy, index, rindex macros
28: add isgraph and isxdigit
1.52 fefe 29: add strftime (ugh!!!)
1.53 olaf 30: Olaf fixed strftime, gmtime_r leap year bug.
1.54 fefe 31: Jeff Garzik contributed a spec file. I put it in contrib and hope
32: everbody knows what to do with it...
33: Olaf checked in a ton of libpthread stuff.
1.55 fefe 34: added mkstemp using /dev/urandom (harder to guess than getpid()).
1.56 fefe 35: add large file backwards compatibility support (so gzip will run on
36: kernel 2.2 and 2.0)
1.57 fefe 37: added shutdown
1.58 fefe 38: added nice
39: added pseudo-locale support for is* (they are now weak aliases,
40: isupper -> __isupper_ascii etc). So, for example, liblatin1 can
41: overwrite the default is* functions.
1.59 ! fefe 42: added liblatin1.a
1.34 fefe 43:
1.21 fefe 44: 0.7.2:
45: added several prototypes to already implemented functions and
46: <sys/reboot.h>. Thanks for Abraham for pointing this out.
1.22 fefe 47: added prototypes for process group and dup syscalls. Thanks to
48: Abraham again. He also contributed an implementation for daemon(),
49: an ugly BSD hack. I decided to diversify the lib/ directory a
50: little and now created a libugly/ directory for stuff like system()
51: and daemon() ;-)
52: added libstdio and moved all that stdio crap from lib/ there. The
53: idea is to start separating now what we will separate for shared
54: libraries anyway.
1.26 fefe 55: optimized x86 unified-syscall by Olaf.
1.27 fefe 56: added limits.h, removed __ARCHBITS from endian.h (use __WORDSIZE from
57: limits.h instead)
58: added remove.c and enough cruft to limits.h to make it compile lzo,
59: lzop and bzip2.
1.28 fefe 60: added mktime, asctime_r, asctime and ctime (needed for gzip).
61: Someone will need to implement a reader for the glibc localtime file
1.31 fefe 62: format some day (man tzfile, ugh!) :-(
1.29 fefe 63: added unlink prototype to unistd.h.
64: added gmtime and gmtime_r.
1.31 fefe 65: fixed sparc unified syscall error handling (please upgrade!).
1.30 fefe 66: removed readdir glibc compatibility cruft.
1.31 fefe 67: removed stat glibc compatibility cruft.
1.32 fefe 68: oops, my fread was broken for pipes.
1.33 fefe 69: Olaf found a bug in pipe() for sparc, too.
1.21 fefe 70:
1.9 fefe 71: 0.7.1:
1.11 fefe 72: optimized away the static arrays from __ltostr and __lltostr.
73: optimized away the static arrays (and more!) from strtoul and strtoull.
74: added system (by Olaf).
1.12 fefe 75: added tolower and toupper
1.13 fefe 76: added getopt from Olaf (~600 bytes on x86!)
1.14 fefe 77: added clone for mips (by Olaf).
78: fixed longjmp and setjmp on sparc not to require glibc include files.
1.15 olaf 79: added pause for all (by Olaf).
1.17 fefe 80: added __attribute__ ((__const__)) to ctype.h
1.16 fefe 81: added isupper and islower.
82: added putc and putchar as macros.
1.17 fefe 83: added strcasecmp and strncasecmp (thanks to Abraham vd Merwe for
84: insightful discussions and patches I partially used for the last three)
1.18 fefe 85: added getpriority and setpriority (also by Abraham)
1.19 fefe 86: oops, I broke strtoul in a recent optimization attempt.
1.20 fefe 87: oi, setjmp and longjmp were broken on other platforms, too!
88: Thanks to Sébastien Côté for reporting this.
1.9 fefe 89:
90: 0.7:
1.6 olaf 91: clone support for i386,alpha,arm and sparc by Olaf.
1.1 cvs 92: Initial dynamic loading glue for i386, also by Olaf.
93: We don't have a dynamic loader yet, so this is not yet usable.
94: added readv, sigdelset, sigaltstack, sigfillset, sigismember, usleep,
95: vsprintf, writev, all by Olaf.
1.2 fefe 96: imported readdir fix from Guillaume. Mhh, can the kernel interface
97: really be this broken?
1.3 fefe 98: Found a kludgy way to detect whether the kernel includes define struct
99: stat64 (they also define STAT64_HAS_BROKEN_ST_INO). If this does
100: not work for you, please tell me!
1.4 fefe 101: Removed bogus __seek_types enum, it's a bunch of #defines now.
1.7 fefe 102: Fixed stdio line buffering. The code was already there but I forgot
103: to mark stdio and stdout as line buffered. This should really be
104: done dynamically using isatty, though.
1.8 fefe 105: Removed superfluous "if (1)" in execvp.c
1.1 cvs 106:
107: 0.6.14:
108: *scanf did not append the 0 byte for %s.
109: added creat and changed creat to call open directly instead of open64.
110: did for fread what 0.6.10 did for fwrite.
111: repaired buffered stdio a little (line buffering is still not
112: supported).
113: [All of these bugs have been reported by Guillaume. Thanks!]
114: added __pure__ to strchr and strrchr in <string.h>
115:
116: 0.6.13:
117: included COPYING to make clear that the diet libc is covered by the
118: GNU General Public License (and _not_ the LGPL). That means that
119: you need to obtain a license from me if you want to use the diet
120: libc in a proprietary program that you want to distribute.
121: The x86 unified syscall swapped arguments #4 and #5 and noone
122: noticed until now! Thanks to Guillaume Cottenceau for reporting
123: this!
124: The malloc had a subtle bug with small allocations that could cause
125: segfault. Guillaume reported it, Olaf fixed it.
126:
127: 0.6.12:
128: added assert, statfs and fstatfs
129: added include/net/if.h so fget compiles again.
130:
131: 0.6.11:
132: added putenv.
133:
134: 0.6.10:
135: fixed fwrite. It returned the number of bytes written, not the number
136: of records. Thanks to Albert D. Cahalan for pointing this out.
137: fixed fgets. It returned EOF on empty lines. Thanks, Erik Frey.
138: added brk and sbrk. harold@nb.com.sg made me do it.
139:
140: 0.6.9:
141: fixed execvp not to return on ENOENT
142: did for {open|read|close|seek|tell}dir what I did for stat before.
143: The glibc compatibility part is currently non-functional.
144: fixed white space handling in sscanf " (" would not match " (".
145: fixed printf to accept 'l' flag.
146: fixed getpwuid (ignored last line)
147:
148: 0.6.8:
149: Olaf contributed initial thread-safe syscalls for i386, alpha, sparc,
150: mips and arm. I did the one for ppc. They can be enabled in dietfeatures.h.
151: added strlcpy and strlcat from OpenBSD.
152: added stat64, fstat64 and lstat64.
153: added endian.h and made ending of strlen.c endianness-aware
154: added string routines size tweak to dietfeatures.h. Most of the
155: string routines have been unrolled. This is up to three times as
156: fast but creates up to three times larger code. Now you can disable
157: the unrolling.
158: added sendfile.
159: I also added pread. It works on x86 and sparc but not on ppc and mips.
160: I have no idea what's going on. strace disagrees with the kernel.
161: Can anyone help? I asked the Linux kernel mailing list for help, too.
162: added a few aliases of the type __libc_open for open. nm on
163: libpthreads.so indicates that we will need them.
164:
165: 0.6.7:
166: the sources now compile without warnings with -Wall.
167: printf now prints "(null)" when %s is passed NULL. This can be
168: removed with WANT_NULL_PRINTF in dietfeatures.h
169: added vfprintf, execl, ttyname.
170: stat, lstat and fstat are now #defined to __dietstat, __dietlstat and
171: __dietfstat so they can use the normal kernel struct stat.
172: stat, lstat and fstat are now C wrappers that convert to the glibc
173: struct stat. So we can avoid that overhead for programs that use
174: dietlibc headers.
175:
176: 0.6.6:
177: changed the Makefiles so you can now set CFLAGS on the command line
178: added memccmp (analogous to memccpy) and strncmp.
179: Olaf fixed another bug in the sparc unified syscall.
180: Paul Clifford contributed a C version of his strlen.S that is much
181: more efficient than the previous strlen on all platforms! It also
182: looks like technology from Roswell. ;-}
183: added a "real" stdio. Well, almost. Please contribute!
184: added dietfeatures.h so you can remove features you don't need.
185: removed debug code from vsnprintf that changed \0 to ' ' (argh!).
186: stdio now works with simple programs and minigzip from libz.
187: stdio uses some major trickery to avoid linking the stdio, stderr and
188: stdout (and reserving space for them) when they are not used.
189:
190: 0.6.5:
191: Red Plait <redplait@ixcelerator.com> found several bugs in diet libc,
192: most of them bugs in the new header files, but also missing sigset
193: functions.
194: Paul Clifford contributed an assembly strlen.S for ARM and fixed
195: several bugs.
196: Fixed ppc/setjmp and mips/pipe.
197: Olaf contributed a new sparc unified syscall.
198:
199: 0.6.4:
200: declared is* static inline in getservent.
201: added assert_fail, strtoul, isalpha, isdigit, isalnum, isascii.
202: changed strlen to return 0 when passed NULL.
203: new, much smaller unified syscall for MIPS. Thanks to Olaf the Mad
204: Scientist who actually implemented this without access to a MIPS
205: box, just from reading the architecture manual. And it worked out
206: of the box.
207: started a set of system includes, mainly so that I can use lcc and
208: my alpha-linux cross compiler (which is unable to cross-compile
209: glibc). diet libc can now be compiled without any include files
210: from a normal libc. The includes do declare more than diet libc
211: currently delivers and they are still far from complete for real
212: applications.
213: I will only add assembly versions that are smaller _and_ faster than
214: the C version. Larger routines are only accepted if they are called
215: very often and are substantially faster.
216: added i386 assembly strchr, which is smaller and faster than the old
217: version. It is, however, larger than the version contributed by
218: proton (thanks, anyway).
219: added i386 strlen (31% faster, 14% smaller)
220: "make t" will now create a map file called "mapfile".
221:
222: 0.6.3:
223: added sys_errlist, strerror and perror
224: added isblank
225: added atol
226:
227: 0.6.2:
228: mmap for ARM didn't compile. Thanks, Paul!
229:
230: 0.6.1:
231: split mmap into the architecture specific subdirectories.
232: getenv now copes with environ==NULL, thanks Paul Clifford.
233: Paul also contributed a smaller ARM startup code.
234:
235: 0.6:
236: strcat returned the wrong result. Thanks, Dietz Pröpper.
237: strtod now understands a negative exponent (oops, thanks Bertram Barth)
238: Port to arm-linux-gnu, but on the Netwinder I use for testing the
239: __dtostr does not work (I have no idea why!)
240: The MIPS port now uses (much smaller) non-PIC code. That means that
241: the applications you link against diet libc also have to be compiled
242: non-PIC. I suggest copying the CFLAGS from the diet libc Makefile.
243: Thanks to Ralf Bächle for helping me with this!
244: I made subdirectories for the architectures and use VPATH to override
245: VPATH so that make finds the source file automatically. That should
246: simplify the sources greatly.
247: Olaf Dreesen contributed Alpha support including setjmp and longjmp!
248: ARM and MIPS now also have setjmp and longjmp
249: "compile" and "load" are now make targets. Use them for djb programs.
250: added contrib/elftrunc.c which will remove unnecessary ELF headers.
251: Again, contributed from Olaf. Great work!
252:
253: 0.5.12:
254: printf also does signed numbers.
255: If you don't use atexit, dietlibc now does not link exit, only _exit.
256: *printf now correctly returns the number of bytes written.
257: Olaf contributed experimental sscanf and vsscanf implementations.
258: If passed NULL as buffer, snprintf will not write anything but still
259: return the number of bytes it would have written.
260: Initial MIPS port! (Oh, the agony!)
261: No setjmp and longjmp support yet! Please contribute!
262: I even implemented unified syscalls for MIPS. Still, MIPS code is
263: almost twice the size of SPARC code. If anyone knows why: please
264: tell me!
265:
266: 0.5.11:
267: I implemented new unified syscalls for x86, sparc and ppc, this time based
268: on .s files and not .c files, and I moved the syscalls into
269: subdirectores to clean the dietlibc sources up a little.
270: A binary consisting of printf("%s is %d\n","olaf",23) is now
271: 2864 bytes on sparc
272: 2488 bytes on intel
273: I kludgily implemented fprintf(stdout,... and fprintf(stderr,...
274: to make a few more applications work.
275: Olaf Dreesen also implemented some unified syscall stuff for x86.
276: He discovered that with our unified syscall interface it actually
277: costs <10 bytes total to make all system calls thread safe!
278: I implemented atexit() (can register up to 4 callbacks).
279: I implemented strtod and __dtostr (the opposite). Now we can add
280: floating point support to vsnprintf and sscanf!
281:
282: 0.5.10:
283: I actually saw that I can not only merge the errno handling code of
284: the system calls on x86, I can also merge the rest (including the
285: arguments) except for setting the system call number. All those
286: system calls are now just a jump to a unified system call handler.
287: I got rid of x86openclose again. The savings are substantial:
288: chown from embutils went from 7664 to 7184 bytes!
289: If I'd move the system call wrappers to assembly language, I could
290: even reduce the alignment (does not matter since it's just a jump
291: anyway) and get rid of the "ret" behind each jump (one byte per
292: system call!)
293:
294: 0.5.9:
295: added memchr, strpbrk, strstr, strtol, isspace (hehe)
296: fixed strdup
297: fixed return values for strcpy and strcat (thanks to Norbert Berzen)
298: Olaf Dreesen contributed a strtol and initial {sn|vsn|}printf implementation
299: (no signed integers, only strings and unsigned integers (but
300: supporting octal, hex and decimal). Thanks, Olaf! (by the way:
301: that code must be wonderful, I don't understand it at all *bg*)
302: Olaf's printf does understand padding like in "%08d" and "%8d" and
303: automatically pads pointers with '0'.
304: Rewrote the x86 start code, old: 22 instructions, new: 12 instructions.
305: Thanks to proton for inspiration on this.
306: The new x86 start code also does not reference exit any more, thus saving
307: 64 bytes for executables that don't call exit explicitly.
308: [insert maniacal laughter] EVERY BYTE COUNTS!1!! ;-)
309: Moved errno and environ to start.S, saving no byte binary size but
310: speeding up compilation and slightly shortening dietlibc.a ;-)
311: For x86: joined open and close into one assembler file, sharing the
312: errno handling. This is a feasibility test and it actually saves a
313: few bytes. I think I will reimplement the _syscall[1-6] macros on
314: all platforms now to share the errno handling code for them.
315:
316: 0.5.8:
317: fixed strchr to be able to look for 0.
318:
319: added _llseek, ftruncate, getpgid, getresgid, getresuid, getsid,
320: memccpy, memmove, mprotect, setregid, setresgid, setresuid, setreuid,
321: strncpy, swapon, truncate, strtok, strtok_r, strspn, strcspn (all
322: contributed by Olaf Dreesen)
323:
324: added execvp, getcwd
325:
326: fixed __xmknod
327:
328: Note: can it be that the ftw interface really is so broken that I
329: cannot implement it without having to implement some searching data
330: structure? I included an experimental ftw implementation that will
331: not follow symlinks.
332:
333: 0.5.7:
334: oops, if_nametoindex was broken!
335:
336: 0.5.6:
337: ported to ppc-linux.
338: fixed i386 sigsetjmp (I mistyped the function name)
339: included sigjmp.c
340:
341: 0.5.5:
342: added wait, sys_siglist, longjmp/setjmp/sigsetjmp for i386 and sparc
343: actually, it wasn't fork that wasn't working for sparc, it was pipe.
344: I fixed it now.
345: dietlibc/SPARC assumes -msupersparc (does not provide div, mul, etc)
346: fixed readdir to use getdents and not the intel inline asm
347:
348: 0.5.4:
349: ported to sparclinux (sparc32 only). Beware: does not work yet.
350: added raise, abort, readlink, strcat, geteuid, geteuid, wait3, access
351:
352: 0.5.3:
353: added tcsetattr and getenv to compile e3
354:
355: 0.5.2:
356: added vhangup, tcgetattr, isatty and memcmp for fgetty
357: (http://www.fefe.de/fgetty/)
358: added localtime from uC-libc.
359:
360: 0.5.1:
361: stat and friends actually work now. Yuck! Another case of
362: translation between kernel and userland. Why can't the kernel
363: people and the libc people simply agree on a standard?
364: removed many unnecessary includes to speed up compilation.
365: added -fomit-frame-pointer and i386 compilation target to reduce code size.
366: inlined socketcall to reduce code size.
367: "load" and "compile" are examples for djb code.
368:
369: 0.5:
370: Olaf Dreesen contributed a much smaller implementation of malloc and friends.
371: each object file is now treated with "strip -x -R .note -R .comment"
372: split each system call into a separate object file
373: added ntohs, htons, alarm, if_indextoname and if_nametoindex
374:
375: 0.4.1:
376: oops, the strchr implementation was wrong. Thanks Jens Laas!
377:
378: 0.4:
379: copied opendir and friends from uC-libc.
380: added getservent and getservby* in gerservent.c
381:
382: 0.3:
383: remove readdir system call
384: added getdents system call
385: added getpwnam/getpwuid implementation in getpwnam.c
386: added a few string functions (in str*.c)
387:
LinuxTV legacy CVS <linuxtv.org/cvs>