]> git.ipfire.org Git - thirdparty/glibc.git/blob - NEWS
Add _Float128 function aliases.
[thirdparty/glibc.git] / NEWS
1 GNU C Library NEWS -- history of user-visible changes.
2 Copyright (C) 1992-2017 Free Software Foundation, Inc.
3 See the end for copying conditions.
4
5 Please send GNU C library bug reports via <http://sourceware.org/bugzilla/>
6 using `glibc' in the "product" field.
7 \f
8 Version 2.27
9
10 Major new features:
11
12 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin and tan
13 with FMA, contributed by Arjan van de Ven and H.J. Lu from Intel.
14
15 * Optimized x86-64 trunc and truncf for processors with SSE4.1.
16
17 * Optimized generic expf, exp2f, logf, log2f and powf.
18
19 * In order to support faster and safer process termination the malloc API
20 family of functions will no longer print a failure address and stack
21 backtrace after detecting heap corruption. The goal is to minimize the
22 amount of work done after corruption is detected and to avoid potential
23 security issues in continued process execution. Reducing shutdown time
24 leads to lower overall process restart latency, so there is benefit both
25 from a security and performance perspective.
26
27 * The abort function terminates the process immediately, without flushing
28 stdio streams. Previous glibc versions used to flush streams, resulting
29 in deadlocks and further data corruption. This change also affects
30 process aborts as the result of assertion failures.
31
32 * On platforms where long double has the IEEE binary128 format (aarch64,
33 alpha, mips64, s390 and sparc), the math library now implements _Float128
34 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015. These
35 are the same interfaces added in version 2.26 for some platforms where
36 this format is supported but is not the format of long double.
37
38 Deprecated and removed features, and other changes affecting compatibility:
39
40 * On GNU/Linux, the obsolete Linux constant PTRACE_SEIZE_DEVEL is no longer
41 defined by <sys/ptrace.h>.
42
43 * libm no longer supports SVID error handling (calling a user-provided
44 matherr function on error) or the _LIB_VERSION variable to control error
45 handling. (SVID error handling and the _LIB_VERSION variable still work
46 for binaries linked against older versions of the GNU C Library.) The
47 libieee.a library is no longer provided. math.h no longer defines struct
48 exception, or the macros X_TLOSS, DOMAIN, SING, OVERFLOW, UNDERFLOW,
49 TLOSS, PLOSS and HUGE.
50
51 * The libm functions pow10, pow10f and pow10l are no longer supported for
52 new programs. Programs should use the standard names exp10, exp10f and
53 exp10l for these functions instead.
54
55 * The mcontext_t type is no longer the same as struct sigcontext. On
56 platforms where it was previously the same, this changes the C++ name
57 mangling for interfaces involving this type.
58
59 * The add-ons mechanism for building additional packages at the same time as
60 glibc has been removed. The --enable-add-ons configure option is now
61 ignored.
62
63 Changes to build and runtime requirements:
64
65 [Add changes to build and runtime requirements here]
66
67 Security related changes:
68
69 CVE-2009-5064: The ldd script would sometimes run the program under
70 examination directly, without preventing code execution through the
71 dynamic linker. (The glibc project disputes that this is a security
72 vulnerability; only trusted binaries must be examined using the ldd
73 script.)
74
75 The following bugs are resolved with this release:
76
77 [The release manager will add the list generated by
78 scripts/list-fixed-bugs.py just before the release.]
79
80 \f
81 Version 2.26
82
83 Major new features:
84
85 * A per-thread cache has been added to malloc. Access to the cache requires
86 no locks and therefore significantly accelerates the fast path to allocate
87 and free small amounts of memory. Refilling an empty cache requires locking
88 the underlying arena. Performance measurements show significant gains in a
89 wide variety of user workloads. Workloads were captured using a special
90 instrumented malloc and analyzed with a malloc simulator. Contributed by
91 DJ Delorie with the help of Florian Weimer, and Carlos O'Donell.
92
93 * Unicode 10.0.0 Support: Character encoding, character type info, and
94 transliteration tables are all updated to Unicode 10.0.0, using
95 generator scripts contributed by Mike FABIAN (Red Hat).
96 These updates cause user visible changes, especially the changes in
97 wcwidth for many emoji characters cause problems when emoji sequences
98 are rendered with pango, see for example:
99 https://bugzilla.gnome.org/show_bug.cgi?id=780669#c5
100
101 * Collation of Hungarian has been overhauled and is now consistent with "The
102 Rules of Hungarian Orthography, 12th edition" (Bug 18934). Contributed by
103 Egmont Koblinger.
104
105 * Improvements to the DNS stub resolver, contributed by Florian Weimer:
106
107 - The GNU C Library will now detect when /etc/resolv.conf has been
108 modified and reload the changed configuration. The new resolver option
109 “no-reload” (RES_NORELOAD) disables this behavior.
110
111 - The GNU C Library now supports an arbitrary number of search domains
112 (configured using the “search” directive in /etc/resolv.conf);
113 previously, there was a hard limit of six domains. For backward
114 compatibility, applications that directly modify the ‘_res’ global
115 object are still limited to six search domains.
116
117 - When the “rotate” (RES_ROTATE) resolver option is active, the GNU C
118 Library will now randomly pick a name server from the configuration as a
119 starting point. (Previously, the second name server was always used.)
120
121 * The tunables feature is now enabled by default. This allows users to tweak
122 behavior of the GNU C Library using the GLIBC_TUNABLES environment variable.
123
124 * New function reallocarray, which resizes an allocated block (like realloc)
125 to the product of two sizes, with a guaranteed clean failure upon integer
126 overflow in the multiplication. Originally from OpenBSD, contributed by
127 Dennis Wölfing and Rüdiger Sonderfeld.
128
129 * New wrappers for the Linux-specific system calls preadv2 and pwritev2.
130 These are extended versions of preadv and pwritev, respectively, taking an
131 additional flags argument. The set of supported flags depends on the
132 running kernel; full support currently requires kernel 4.7 or later.
133
134 * posix_spawnattr_setflags now supports the flag POSIX_SPAWN_SETSID, to
135 create a new session ID for the spawned process. This feature is
136 scheduled to be added to the next major revision of POSIX; for the time
137 being, it is available under _GNU_SOURCE.
138
139 * errno.h is now safe to use from C-preprocessed assembly language on all
140 supported operating systems. In this context, it will only define the
141 Exxxx constants, as preprocessor macros expanding to integer literals.
142
143 * On ia64, powerpc64le, x86-32, and x86-64, the math library now implements
144 128-bit floating point as defined by ISO/IEC/IEEE 60559:2011 (IEEE
145 754-2008) and ISO/IEC TS 18661-3:2015. Contributed by Paul E. Murphy,
146 Gabriel F. T. Gomes, Tulio Magno Quites Machado Filho, and Joseph Myers.
147
148 To compile programs that use this feature, the compiler must support
149 128-bit floating point with the type name _Float128 (as defined by TS
150 18661-3) or __float128 (the nonstandard name used by GCC for C++, and for
151 C prior to version 7). _GNU_SOURCE or __STDC_WANT_IEC_60559_TYPES_EXT__
152 must be defined to make the new interfaces visible.
153
154 The new functions and macros correspond to those present for other
155 floating-point types (except for a few obsolescent interfaces not
156 supported for the new type), with F128 or f128 suffixes; for example,
157 strtof128, HUGE_VAL_F128 and cosf128. Following TS 18661-3, there are no
158 printf or scanf formats for the new type; the strfromf128 and strtof128
159 interfaces should be used instead.
160
161 Deprecated and removed features, and other changes affecting compatibility:
162
163 * The synchronization that pthread_spin_unlock performs has been changed to
164 now be equivalent to a C11 atomic store with release memory order to the
165 spin lock's memory location. Previously, several (but not all)
166 architectures used stronger synchronization (e.g., containing what is
167 often called a full barrier). This change can improve performance, but
168 may affect odd fringe uses of spin locks that depend on the previous
169 behavior (e.g., using spin locks as atomic variables to try to implement
170 Dekker's mutual exclusion algorithm).
171
172 * The port to Native Client running on ARMv7-A (--host=arm-nacl) has been
173 removed.
174
175 * Sun RPC is deprecated. The rpcgen program, librpcsvc, and Sun RPC headers
176 will only be built and installed when the GNU C Library is configured with
177 --enable-obsolete-rpc. This allows alternative RPC implementations, such
178 as TIRPC or rpcsvc-proto, to be used.
179
180 * The NIS(+) name service modules, libnss_nis, libnss_nisplus, and
181 libnss_compat, are deprecated, and will not be built or installed by
182 default.
183
184 The NIS(+) support library, libnsl, is also deprecated. By default, a
185 compatibility shared library will be built and installed, but not headers
186 or development libraries. Only a few NIS-related programs require this
187 library. (In particular, the GNU C Library has never required programs
188 that use 'gethostbyname' to be linked with libnsl.)
189
190 Replacement implementations based on TIRPC, which additionally support
191 IPv6, are available from <https://github.com/thkukuk/>. The configure
192 option --enable-obsolete-nsl will cause libnsl's headers, and the NIS(+)
193 name service modules, to be built and installed.
194
195 * The DNS stub resolver no longer performs EDNS fallback. If EDNS or DNSSEC
196 support is enabled, the configured recursive resolver must support EDNS.
197 (Responding to EDNS-enabled queries with responses which are not
198 EDNS-enabled is fine, but FORMERR responses are not.)
199
200 * res_mkquery and res_nmkquery no longer support the IQUERY opcode. DNS
201 servers have not supported this opcode for a long time.
202
203 * The _res_opcodes variable has been removed from libresolv. It had been
204 exported by accident.
205
206 * <string.h> no longer includes inline versions of any string functions,
207 as this kind of optimization is better done by the compiler. The macros
208 __USE_STRING_INLINES and __NO_STRING_INLINES no longer have any effect.
209
210 * The nonstandard header <xlocale.h> has been removed. Most programs should
211 use <locale.h> instead. If you have a specific need for the definition of
212 locale_t with no other declarations, please contact
213 libc-alpha@sourceware.org and explain.
214
215 * The obsolete header <sys/ultrasound.h> has been removed.
216
217 * The obsolete signal constant SIGUNUSED is no longer defined by <signal.h>.
218
219 * The obsolete function cfree has been removed. Applications should use
220 free instead.
221
222 * The stack_t type no longer has the name struct sigaltstack. This changes
223 the C++ name mangling for interfaces involving this type.
224
225 * The ucontext_t type no longer has the name struct ucontext. This changes
226 the C++ name mangling for interfaces involving this type.
227
228 * On M68k GNU/Linux and MIPS GNU/Linux, the fpregset_t type no longer has
229 the name struct fpregset. On Nios II GNU/Linux, the mcontext_t type no
230 longer has the name struct mcontext. On SPARC GNU/Linux, the struct
231 mc_fq, struct rwindow, struct fpq and struct fq types are no longer
232 defined in sys/ucontext.h, the mc_fpu_t type no longer has the name struct
233 mc_fpu, the gwindows_t type no longer has the name struct gwindows and the
234 fpregset_t type no longer has the name struct fpu. This changes the C++
235 name mangling for interfaces involving those types.
236
237 * On S/390 GNU/Linux, the constants defined by <sys/ptrace.h> have been
238 synced with the kernel:
239
240 - PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS and PTRACE_SETFPREGS
241 are not supported on this architecture and have been removed.
242
243 - PTRACE_SINGLEBLOCK, PTRACE_SECCOMP_GET_FILTER, PTRACE_PEEKUSR_AREA,
244 PTRACE_POKEUSR_AREA, PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE,
245 PTRACE_DISABLE_TE and PTRACE_TE_ABORT_RAND have been added.
246
247 Programs that assume the GET/SETREGS ptrace requests are universally
248 available will now fail to build, instead of malfunctioning at runtime.
249
250 Changes to build and runtime requirements:
251
252 * Linux kernel 3.2 or later is required at runtime, on all architectures
253 supported by that kernel. (This is a change from version 2.25 only for
254 x86-32 and x86-64.)
255
256 * GNU Binutils 2.25 or later is now required to build the GNU C Library.
257
258 * On most architectures, GCC 4.9 or later is required to build the GNU C
259 Library. On powerpc64le, GCC 6.2 or later is required.
260
261 Older GCC versions and non-GNU compilers are still supported when
262 compiling programs that use the GNU C Library. (We do not know exactly
263 how old, and some GNU extensions to C may be _de facto_ required. If you
264 are interested in helping us make this statement less vague, please
265 contact libc-alpha@sourceware.org.)
266
267 Security related changes:
268
269 * The DNS stub resolver limits the advertised UDP buffer size to 1200 bytes,
270 to avoid fragmentation-based spoofing attacks (CVE-2017-12132).
271
272 * LD_LIBRARY_PATH is now ignored in binaries running in privileged AT_SECURE
273 mode to guard against local privilege escalation attacks (CVE-2017-1000366).
274
275 * Avoid printing a backtrace from the __stack_chk_fail function since it is
276 called on a corrupt stack and a backtrace is unreliable on a corrupt stack
277 (CVE-2010-3192).
278
279 * A use-after-free vulnerability in clntudp_call in the Sun RPC system has been
280 fixed (CVE-2017-12133).
281
282 The following bugs are resolved with this release:
283
284 [984] network: Respond to changed resolv.conf in gethostbyname
285 [5010] network: sunrpc service cleanup causes unwanted port mapper traffic
286 [12068] localedata: sc_IT: misspelled yesexpr/day/abday/mon/abmon/date_fmt
287 fields
288 [12189] libc: __stack_chk_fail should not attempt a backtrace
289 (CVE-2010-3192)
290 [14096] time: Race condition on timezone/tst-timezone.out
291 [14172] localedata: az_IR: new locale
292 [14995] build: glibc fails to build if gold is the default linker, even if
293 ld.bfd is available
294 [15998] build: [powerpc] Set arch_minimum_kernel for powerpc LE
295 [16637] network: inet_pton function is accepting IPv6 with bad format
296 [16640] string: string/strtok.c: undefined behaviour inconsistent between
297 x86 and other generic code
298 [16875] localedata: ko_KR: fix lang_name
299 [17225] localedata: ar_SY: localized month names for May and June are
300 incorrect
301 [17297] localedata: da_DK: wrong date_fmt string
302 [18907] stdio: Incorrect order of __wur __THROW in <printf.h>
303 [18934] localedata: hu_HU: collate: fix multiple bugs and add tests
304 [18988] nptl: pthread wastes memory with mlockall(MCL_FUTURE)
305 [19066] localedata: ar_SA abbreviated day and month names are in English
306 [19569] network: resolv: Support an arbitrary number of search domains
307 [19570] network: Implement random DNS server selection in the stub
308 resolver
309 [19838] locale: localedef fails on PA-RISC
310 [19919] localedata: iso14651_t1_common: Correct the Malayalam sorting
311 order of 0D36 and 0D37
312 [19922] localedata: iso14651_t1_common: Define collation for Malayalam
313 chillu characters
314 [20098] libc: FAIL: debug/backtrace-tst on hppa
315 [20257] network: sunrpc: clntudp_call does not enforce timeout when
316 receiving data
317 [20275] localedata: locale day/abday/mon/abmon should not have trailing
318 whitespace
319 [20313] localedata: Update locale data to Unicode 9.0
320 [20424] manual: Document how to provide a malloc replacement
321 [20496] localedata: agr_PE: new language locale Awajún / Aguaruna (agr)
322 for Peru
323 [20686] locale: Add el_GR@euro to SUPPORTED.
324 [20831] dynamic-link: _dl_map_segments does not test for __mprotect
325 failures consistently
326 [21015] dynamic-link: Document and fix --enable-bind-now
327 [21016] nptl: pthread_cond support is broken on hppa
328 [21029] libc: glibc-2.23 (and later) fails to compile with -fno-omit-
329 frame-pointer on i386
330 [21049] libc: segfault in longjmp_chk() due to clobbered processor
331 register
332 [21075] libc: unused assigment to %g4 in sparc/sparc{64,32}/clone.S
333 [21088] libc: Build fails with --enable-static-nss
334 [21094] math: cosf(1.57079697) has 3 ulp error on targets where the
335 generic c code is used
336 [21109] libc: Tunables broken on big-endian
337 [21112] math: powf has large ulp errors with base close to 1 and exponent
338 around 4000
339 [21115] network: sunrpc: Use-after-free in error path in clntudp_call
340 (CVE-2017-12133)
341 [21120] malloc: glibc malloc is incompatible with GCC 7
342 [21130] math: Incorrect return from y0l (-inf) and y1l (-inf) when linking
343 with -lieee
344 [21134] math: Exception (divide by zero) not set for y0/y1 (0.0) and y0/y1
345 (-0.0) when linking with -lieee
346 [21171] math: log10, log2 and lgamma return incorrect results
347 [21179] libc: handle R_SPARC_DISP64 and R_SPARC_REGISTER relocs
348 [21182] libc: __memchr_sse2: regression in glibc-2.25 on i686
349 [21207] localedata: ce_RU: update weekdays from CLDR
350 [21209] dynamic-link: LD_HWCAP_MASK read in setuid binaries
351 [21217] localedata: Update months from CLDR-31
352 [21232] libc: miss posix_fadvise64 on MIPS64 when static linking
353 [21243] libc: support_delete_temp_file should issue warning for failed
354 remove()
355 [21244] libc: support resolv_test_start() socket fd close should be
356 checked for errors.
357 [21253] libc: localedef randomly segfaults when using -fstack-check due to
358 new posix_spawn implementation
359 [21258] dynamic-link: Branch predication in _dl_runtime_resolve_avx512_opt
360 leads to lower CPU frequency
361 [21259] libc: [alpha] termios.h missing IXANY for POSIX
362 [21261] libc: [sparc64] bits/setjmp.h namespace
363 [21267] network: [mips] bits/socket.h IOC* namespace
364 [21268] libc: [alpha] termios.h NL2, NL3 namespace
365 [21270] libc: mmap64 silently truncates large offset values
366 [21275] libc: posix_spawn always crashes on ia64 now
367 [21277] libc: [alpha] termios.h missing IUCLC for UNIX98 and older
368 [21280] math: [powerpc] logbl for POWER7 return incorrect results
369 [21289] libc: Incorrect declaration for 32-bit platforms with
370 _FILE_OFFSET_BITS=64 causes build error
371 [21295] network: GETAI(AF_UNSPEC) drops IPv6 addresses if nss module does
372 not support gethostbyname4_r
373 [21298] nptl: rwlock can deadlock on frequent reader/writer phase
374 switching
375 [21338] malloc: mallopt M_ARENA_MAX doesn't set the maximum number of
376 arenas
377 [21340] libc: Support POSIX_SPAWN_SETSID
378 [21357] libc: unwind-dw2-fde deadlock when using AddressSanitizer
379 [21359] network: ns_name_pack needs additional byte in destination buffer
380 [21361] network: resolv: Reduce advertised EDNS0 buffer size to guard
381 against fragmentation attacks (CVE-2017-12132)
382 [21369] network: resolv: Remove EDNS fallback
383 [21371] libc: Missing timespec definition when compiled with _XOPEN_SOURCE
384 and _POSIX_C_SOURCE
385 [21386] nptl: Assertion in fork for distinct parent PID is incorrect
386 [21391] dynamic-link: x86: Set dl_platform and dl_hwcap from CPU features
387 [21393] stdio: Missing dup3 error check in freopen, freopen64
388 [21396] libc: Use AVX2 memcpy/memset on Skylake server
389 [21399] localedata: Bad description for U00EC in
390 localedata/charmaps/CP1254
391 [21411] malloc: realloc documentation error
392 [21426] network: sys/socket.h uio.h namespace
393 [21428] libc: [aarch64] tst-backtrace5 testsuite failure
394 [21445] libc: signal.h bsd_signal namespace
395 [21455] network: Network headers stdint.h namespace
396 [21474] network: resolv: res_init does not use RES_DFLRETRY (2) but 4 for
397 retry value
398 [21475] network: resolv: Overlong search path is truncated mid-label
399 [21511] libc: sigstack namespace
400 [21512] libc: clone() ends up calling exit_group() through _exit() wrapper
401 [21514] libc: sysdeps/unix/sysv/linux/sys/syscall.h:31:27: fatal error:
402 bits/syscall.h: No such file or directory
403 [21517] libc: struct sigaltstack namespace
404 [21528] dynamic-link: Duplicated minimal strtoul implementations in ld.so
405 [21533] localedata: Update locale data to Unicode 10.0
406 [21537] libc:
407 ../sysdeps/unix/sysv/linux/s390/s390-32/__makecontext_ret.S:44: Error:
408 junk at end of line, first unrecognized character is `@'
409 [21538] libc: SIG_HOLD missing for XPG4
410 [21539] libc: S390: Mismatch between kernel and glibc ptrace.h with
411 request 12: PTRACE_SINGLEBLOCK vs PTRACE_GETREGS.
412 [21542] libc: Use conservative default for sysconf (_SC_NPROCESSORS_ONLN)
413 [21543] libc: sigevent namespace
414 [21548] libc: [mips] get/set/make/swap context for MIPS O32 assume wrong
415 size for general purpose registers in mcontext_t structure
416 [21550] libc: sigwait namespace
417 [21552] libc: XPG4 bsd_signal namespace
418 [21554] libc: sigpause namespace
419 [21560] libc: sys/wait.h signal.h namespace
420 [21561] libc: waitid namespace
421 [21573] nptl: GCC 7: /usr/bin/install: cannot remove
422 '/usr/include/stdlib.h': Permission denied
423 [21575] libc: sys/wait.h missing struct rusage definition
424 [21584] libc: sigaltstack etc namespace
425 [21597] libc: siginterrupt namespace
426 [21607] math: hppa: FAIL: math/test-tgmath
427 [21609] dynamic-link: Incomplete workaround for GCC __tls_get_addr ABI
428 issue on x86-64
429 [21622] libc: [tile] missing SA_* for POSIX.1:2008
430 [21624] dynamic-link: ld.so: Unsafe alloca allows local attackers to alias
431 stack and heap (CVE-2017-1000366)
432 [21625] libc: wait3 namespace
433 [21654] nss: Incorrect pointer alignment in NSS group merge result
434 construction
435 [21657] network: Parse interface zone id for node-local multicast
436 [21662] string: memcmp-avx2-movbe.S lacks saturating subtraction for
437 between_2_3
438 [21666] libc: .symver is used on common symbol
439 [21668] network: resolv: res_init cross-thread broadcast introduces race
440 conditions
441 [21687] math: tgmath.h totalorder, totalordermag return type
442 [21694] locale: Current Glibc Locale Does Not Support Tok-Pisin and Fiji
443 Hindi Locale
444 [21696] libc: Incorrect assumption of of __cpu_mask in
445 posix/sched_cpucount.c
446 [21697] libc: sysdeps/posix/spawni.c: 2 * suspicious condition ?
447 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
448 locale
449 [21707] math: ppc64le: Invalid IFUNC resolver from libgcc calls getauxval,
450 leading to relocation crash
451 [21709] libc: resolv_conf.c:552: update_from_conf: Assertion
452 `resolv_conf_matches (resp, conf)' failed.
453 [21710] localedata: Added Samoan language locale for Samoa
454 [21711] localedata: Pashto yesstr/nostr locale are missing
455 [21715] nptl: sysdeps/nptl/bits/pthreadtypes.h: typedef guard
456 __have_pthread_attr_t can cause redefinition of typedef ‘pthread_attr_t’
457 [21721] localedata: Incorrect Full Weekday names for ks_IN@devanagari
458 [21723] localedata: yesstr/nostr missing for Chinese language locale
459 [21724] localedata: yesstr and nostr are missing for Xhosa [LC_MESSAGES]
460 locale
461 [21727] localedata: yesstr and nostr are missing for Tsonga [LC_MESSAGES]
462 locale
463 [21728] localedata: New Locale for Tongan language
464 [21729] localedata: incorrect LC_NAME fields for hi_IN
465 [21733] localedata: yesstr and nostr are missing for zh_HK
466 [21734] localedata: Missing yesstr and nostr are for kw_GB
467 [21738] libc: misc/tst-preadvwritev2 and misc/tst-preadvwritev64v2 fail
468 [21741] libc: Undefined __memmove_chk_XXX and __memset_chk_XXX in libc.a
469 [21742] libc: _dl_num_cache_relocations is undefined in libc.a
470 [21743] localedata: ks_IN@devanagari: abday strings mismatch the day
471 strings
472 [21744] libc: Tests failing on --enable-tunables --enable-stack-
473 protector=all
474 [21749] localedata: Wrong abbreviated day name (“abday”) for
475 ar_JO/ar_LB/ar_SY
476 [21756] localedata: missing yesstr, nostr for nds_DE and nds_NL
477 [21757] localedata: missing yesstr, nostr for pap_AW and pap_CW
478 [21759] localedata: missing yesstr and nostr for Tigrinya
479 [21760] localedata: Fix LC_MESSAGES and LC_ADDRESS for anp_IN
480 [21766] localedata: Wrong LC_MESSAGES for om_ET Locale
481 [21767] localedata: Missing Bislama locales
482 [21768] localedata: Missing yesstr and nostr for aa_ET
483 [21770] localedata: Missing Field in li_NL
484 [21778] nptl: Robust mutex may deadlock
485 [21779] libc: MicroBlaze segfaults when loading libpthread
486 [21783] localedata: Fix int_select international_call_prefixes
487 [21784] localedata: Inconsistency in country_isbn
488 [21788] localedata: Missing Country Postal Abbreviations
489 [21794] localedata: Added-country_isbn-for-Italy
490 [21795] localedata: Add/Fix country_isbn for France
491 [21796] localedata: Added country_isbn for Republic of Korea
492 [21797] localedata: Fix inconsistency in country_isbn and missing prefixes
493 [21799] localedata: Added int_select international_call_prefixes
494 [21801] localedata: Added int_select international_call_prefixes
495 [21804] nptl: Double semicolon in thread-shared-types.h
496 [21807] localedata: LC_ADDRESS fix for pap_CW
497 [21808] localedata: Fix LC_ADDRESS for pap_AW
498 [21821] localedata: Added country_name in mai_IN
499 [21822] localedata: Fix LC_TIME for mai_IN
500 [21823] localedata: missing yesstr, nostr for sa_IN
501 [21825] localedata: Fix name_mrs for mag_IN
502 [21828] localedata: 2.26 changelog should mention user visible changes
503 with unicode 9.0
504 [21835] localedata: Added Maithili language locale for Nepal
505 [21838] localedata: Removed redundant data for the_NP
506 [21839] localedata: Fix LC_MONETARY for ta_LK
507 [21844] localedata: Fix Latin characters and Months Sequence.
508 [21848] localedata: Fix mai_NP Title Name
509
510 \f
511 Version 2.25
512
513 * The feature test macro __STDC_WANT_LIB_EXT2__, from ISO/IEC TR
514 24731-2:2010, is supported to enable declarations of functions from that
515 TR. Note that not all functions from that TR are supported by the GNU C
516 Library.
517
518 * The feature test macro __STDC_WANT_IEC_60559_BFP_EXT__, from ISO/IEC TS
519 18661-1:2014, is supported to enable declarations of functions and macros
520 from that TS. Note that not all features from that TS are supported by
521 the GNU C Library.
522
523 * The feature test macro __STDC_WANT_IEC_60559_FUNCS_EXT__, from ISO/IEC TS
524 18661-4:2015, is supported to enable declarations of functions and macros
525 from that TS. Note that most features from that TS are not supported by
526 the GNU C Library.
527
528 * The nonstandard feature selection macros _REENTRANT and _THREAD_SAFE are
529 now treated as compatibility synonyms for _POSIX_C_SOURCE=199506L.
530 Since the GNU C Library defaults to a much newer revision of POSIX, this
531 will only affect programs that specifically request an old conformance
532 mode. For instance, a program compiled with -std=c89 -D_REENTRANT will
533 see a change in the visible declarations, but a program compiled with
534 just -D_REENTRANT, or -std=c99 -D_POSIX_C_SOURCE=200809L -D_REENTRANT,
535 will not.
536
537 Some C libraries once required _REENTRANT and/or _THREAD_SAFE to be
538 defined by all multithreaded code, but glibc has not required this for
539 many years.
540
541 * The inclusion of <sys/sysmacros.h> by <sys/types.h> is deprecated. This
542 means that in a future release, the macros “major”, “minor”, and “makedev”
543 will only be available from <sys/sysmacros.h>.
544
545 These macros are not part of POSIX nor XSI, and their names frequently
546 collide with user code; see for instance glibc bug 19239 and Red Hat bug
547 130601. <stdlib.h> includes <sys/types.h> under _GNU_SOURCE, and C++ code
548 presently cannot avoid being compiled under _GNU_SOURCE, exacerbating the
549 problem.
550
551 * New <fenv.h> features from TS 18661-1:2014 are added to libm: the
552 fesetexcept, fetestexceptflag, fegetmode and fesetmode functions, the
553 femode_t type and the FE_DFL_MODE and FE_SNANS_ALWAYS_SIGNAL macros.
554
555 * Integer width macros from TS 18661-1:2014 are added to <limits.h>:
556 CHAR_WIDTH, SCHAR_WIDTH, UCHAR_WIDTH, SHRT_WIDTH, USHRT_WIDTH, INT_WIDTH,
557 UINT_WIDTH, LONG_WIDTH, ULONG_WIDTH, LLONG_WIDTH, ULLONG_WIDTH; and to
558 <stdint.h>: INT8_WIDTH, UINT8_WIDTH, INT16_WIDTH, UINT16_WIDTH,
559 INT32_WIDTH, UINT32_WIDTH, INT64_WIDTH, UINT64_WIDTH, INT_LEAST8_WIDTH,
560 UINT_LEAST8_WIDTH, INT_LEAST16_WIDTH, UINT_LEAST16_WIDTH,
561 INT_LEAST32_WIDTH, UINT_LEAST32_WIDTH, INT_LEAST64_WIDTH,
562 UINT_LEAST64_WIDTH, INT_FAST8_WIDTH, UINT_FAST8_WIDTH, INT_FAST16_WIDTH,
563 UINT_FAST16_WIDTH, INT_FAST32_WIDTH, UINT_FAST32_WIDTH, INT_FAST64_WIDTH,
564 UINT_FAST64_WIDTH, INTPTR_WIDTH, UINTPTR_WIDTH, INTMAX_WIDTH,
565 UINTMAX_WIDTH, PTRDIFF_WIDTH, SIG_ATOMIC_WIDTH, SIZE_WIDTH, WCHAR_WIDTH,
566 WINT_WIDTH.
567
568 * New <math.h> features are added from TS 18661-1:2014:
569
570 - Signaling NaN macros: SNANF, SNAN, SNANL.
571
572 - Nearest integer functions: roundeven, roundevenf, roundevenl, fromfp,
573 fromfpf, fromfpl, ufromfp, ufromfpf, ufromfpl, fromfpx, fromfpxf,
574 fromfpxl, ufromfpx, ufromfpxf, ufromfpxl.
575
576 - llogb functions: the llogb, llogbf and llogbl functions, and the
577 FP_LLOGB0 and FP_LLOGBNAN macros.
578
579 - Max-min magnitude functions: fmaxmag, fmaxmagf, fmaxmagl, fminmag,
580 fminmagf, fminmagl.
581
582 - Comparison macros: iseqsig.
583
584 - Classification macros: iscanonical, issubnormal, iszero.
585
586 - Total order functions: totalorder, totalorderf, totalorderl,
587 totalordermag, totalordermagf, totalordermagl.
588
589 - Canonicalize functions: canonicalize, canonicalizef, canonicalizel.
590
591 - NaN functions: getpayload, getpayloadf, getpayloadl, setpayload,
592 setpayloadf, setpayloadl, setpayloadsig, setpayloadsigf, setpayloadsigl.
593
594 * The functions strfromd, strfromf, and strfroml, from ISO/IEC TS 18661-1:2014,
595 are added to libc. They convert a floating-point number into string.
596
597 * Most of glibc can now be built with the stack smashing protector enabled.
598 It is recommended to build glibc with --enable-stack-protector=strong.
599 Implemented by Nick Alcock (Oracle).
600
601 * The function explicit_bzero, from OpenBSD, has been added to libc. It is
602 intended to be used instead of memset() to erase sensitive data after use;
603 the compiler will not optimize out calls to explicit_bzero even if they
604 are "unnecessary" (in the sense that no _correct_ program can observe the
605 effects of the memory clear).
606
607 * On ColdFire, MicroBlaze, Nios II and SH3, the float_t type is now defined
608 to float instead of double. This does not affect the ABI of any libraries
609 that are part of the GNU C Library, but may affect the ABI of other
610 libraries that use this type in their interfaces.
611
612 * On x86_64, when compiling with -mfpmath=387 or -mfpmath=sse+387, the
613 float_t and double_t types are now defined to long double instead of float
614 and double. These options are not the default, and this does not affect
615 the ABI of any libraries that are part of the GNU C Library, but it may
616 affect the ABI of other libraries that use this type in their interfaces,
617 if they are compiled or used with those options.
618
619 * The getentropy and getrandom functions, and the <sys/random.h> header file
620 have been added.
621
622 * The buffer size for byte-oriented stdio streams is now limited to 8192
623 bytes by default. Previously, on Linux, the default buffer size on most
624 file systems was 4096 bytes (and thus remains unchanged), except on
625 network file systems, where the buffer size was unpredictable and could be
626 as large as several megabytes.
627
628 * The <sys/quota.h> header now includes the <linux/quota.h> header. Support
629 for the Linux quota interface which predates kernel version 2.4.22 has
630 been removed.
631
632 * The malloc_get_state and malloc_set_state functions have been removed.
633 Already-existing binaries that dynamically link to these functions will
634 get a hidden implementation in which malloc_get_state is a stub. As far
635 as we know, these functions are used only by GNU Emacs and this change
636 will not adversely affect already-built Emacs executables. Any undumped
637 Emacs executables, which normally exist only during an Emacs build, should
638 be rebuilt by re-running “./configure; make” in the Emacs build tree.
639
640 * The “ip6-dotint” and “no-ip6-dotint” resolver options, and the
641 corresponding RES_NOIP6DOTINT flag from <resolv.h> have been removed.
642 “no-ip6-dotint” had already been the default, and support for the
643 “ip6-dotint” option was removed from the Internet in 2006.
644
645 * The "ip6-bytestring" resolver option and the corresponding RES_USEBSTRING
646 flag from <resolv.h> have been removed. The option relied on a
647 backwards-incompatible DNS extension which was never deployed on the
648 Internet.
649
650 * The flags RES_AAONLY, RES_PRIMARY, RES_NOCHECKNAME, RES_KEEPTSIG,
651 RES_BLAST defined in the <resolv.h> header file have been deprecated.
652 They were already unimplemented.
653
654 * The "inet6" option in /etc/resolv.conf and the RES_USE_INET6 flag for
655 _res.flags are deprecated. The flag was standardized in RFC 2133, but
656 removed again from the IETF name lookup interface specification in RFC
657 2553. Applications should use getaddrinfo instead.
658
659 * DNSSEC-related declarations and definitions have been removed from the
660 <arpa/nameser.h> header file, and libresolv will no longer attempt to
661 decode the data part of DNSSEC record types. Previous versions of glibc
662 only implemented minimal support for the previous version of DNSSEC, which
663 is incompatible with the currently deployed version.
664
665 * The resource record type classification macros ns_t_qt_p, ns_t_mrr_p,
666 ns_t_rr_p, ns_t_udp_p, ns_t_xfr_p have been removed from the
667 <arpa/nameser.h> header file because the distinction between RR types and
668 meta-RR types is not officially standardized, subject to revision, and
669 thus not suitable for encoding in a macro.
670
671 * The types res_sendhookact, res_send_qhook, re_send_rhook, and the qhook
672 and rhook members of the res_state type in <resolv.h> have been removed.
673 The glibc stub resolver did not support these hooks, but the header file
674 did not reflect that.
675
676 * For multi-arch support it is recommended to use a GCC which has
677 been built with support for GNU indirect functions. This ensures
678 that correct debugging information is generated for functions
679 selected by IFUNC resolvers. This support can either be enabled by
680 configuring GCC with '--enable-gnu-indirect-function', or by
681 enabling it by default by setting 'default_gnu_indirect_function'
682 variable for a particular architecture in the GCC source file
683 'gcc/config.gcc'.
684
685 * GDB pretty printers have been added for mutex and condition variable
686 structures in POSIX Threads. When installed and loaded in gdb these pretty
687 printers show various pthread variables in human-readable form when read
688 using the 'print' or 'display' commands in gdb.
689
690 * Tunables feature added to allow tweaking of the runtime for an application
691 program. This feature can be enabled with the '--enable-tunables' configure
692 flag. The GNU C Library manual has details on usage and README.tunables has
693 instructions on adding new tunables to the library.
694
695 * A new version of condition variables functions have been implemented in
696 the NPTL implementation of POSIX Threads to provide stronger ordering
697 guarantees.
698
699 * A new version of pthread_rwlock functions have been implemented to use a more
700 scalable algorithm primarily through not using a critical section anymore to
701 make state changes.
702
703 Security related changes:
704
705 * On ARM EABI (32-bit), generating a backtrace for execution contexts which
706 have been created with makecontext could fail to terminate due to a
707 missing .cantunwind annotation. This has been observed to lead to a hang
708 (denial of service) in some Go applications compiled with gccgo. Reported
709 by Andreas Schwab. (CVE-2016-6323)
710
711 * The DNS stub resolver functions would crash due to a NULL pointer
712 dereference when processing a query with a valid DNS question type which
713 was used internally in the implementation. The stub resolver now uses a
714 question type which is outside the range of valid question type values.
715 (CVE-2015-5180)
716
717 The following bugs are resolved with this release:
718
719 [4099] stdio: Overly agressive caching by stream i/o functions.
720 [7065] build: Support building glibc with -fstack-protector or -fstack-
721 protector-all
722 [9842] localedata: en_CA: incorrect date format
723 [13165] nptl: pthread_cond_wait() can consume a signal that was sent
724 before it started waiting
725 [14139] manual: Do not hardcode platform names in manual/libm-err-tab.pl
726 [15765] nptl: sem_open is wrongly a cancellation point
727 [16421] network: IN6_IS_ADDR_UNSPECIFIED can use undefined s6_addr32
728 [16458] libc: endian.h and netinet/in.h byte order macros return values of
729 wrong type
730 [16628] dynamic-link: Segfault after a binary without pthread dlopen()s a
731 library linked with pthread
732 [16630] nptl: Use SYSENTER for pthread_cond_broadcast/signal() (i.e. fix
733 "FIXME: Ingo" issue)
734 [16907] libc: <argp.h> compiled with --std=cXX disables __attribute__
735 [17252] libc: getrandom and getentropy syscall
736 [17730] malloc: thread-local storage is sometimes improperly free()'d
737 after being __libc_memalign()'d
738 [18241] stdio: failed fseek on memstream does not set errno and can fail
739 when it shouldnt
740 [18243] nptl: sem_wait, sem_timedwait are cancellation points shm_open is
741 not
742 [18463] nptl: pthread_cond_broadcast issue when surrounded by
743 PTHREAD_PRIO_INHERIT mutex on ARM
744 [18784] network: res_query and related function crash for special record
745 type queries (CVE-2015-5180)
746 [19380] math: strtod does not raise "inexact"
747 [19387] string: Integer overflow in memchr
748 [19390] string: Integer overflow in strncat
749 [19398] build: linknamespace tests fail with massively parallel build
750 [19402] nptl: Deadlock with robust shared mutex and asynchronous
751 termination
752 [19469] malloc: M_PERTURB in test-skeleton.c invalidates malloc tests
753 [19473] malloc: Turn malloc_get_state etc. in compatibility symbols
754 [19514] libc: [PATCH] Fix spelling errors in spelling
755 "implement"/"implementation" in several places
756 [19582] network: Deprecate RES_USE_INET6
757 [19673] manual: clog10 docs appear to be erroneous
758 [19810] dynamic-link: dlopen with both RTLD_NOLOAD and RTLD_NODELETE
759 causes a segmentation fault
760 [19826] libc: invalid pointer returned from __tls_get_addr with static
761 linking
762 [20016] network: resolv: Remove hooks support from the API
763 [20019] dynamic-link: NULL pointer dereference in libc.so.6 IFUNC due to
764 uninitialized GOT
765 [20033] math: [x86_64] vectorized math function don't call the __finite
766 versions
767 [20116] nptl: use after free in pthread_create
768 [20181] stdio: open_memstream(): writes not at end of stream corrupt data
769 [20292] dynamic-link: Comparison in elf/dl-open.c _dl_addr_inside_object
770 is always true.
771 [20311] nptl: please install proc_service.h
772 [20366] libc: Compilation errors in installed headers in strict-compliance
773 modes
774 [20370] malloc: malloc: Arena free list management is still racy
775 (incorrect fix in bug 19243)
776 [20386] libc: assert (X = 0) does not result in GCC warning
777 [20432] malloc: malloc: Minimize interface required for interposition
778 [20435] libc: Missing unwind info in __startcontext causes infinite loop
779 in _Unwind_Backtrace (CVE-2016-6323)
780 [20444] hurd: recvmsg: PF_LOCAL sockets and msg_name lead to SIGLOST
781 [20452] nptl: Addition of sysdep.o to libpthread.a breaks relinking
782 libpthread.a
783 [20455] math: [powerpc] fesetexceptflag fails to clear FE_INVALID
784 [20459] localedata: et_EE: locale has wrong {p,n}_cs_precedes value
785 [20477] network: resolv: incorrect double-checked locking related to
786 _res_hconf
787 [20478] libc: libc_ifunc macro and similar usages leads to false debug-
788 information.
789 [20495] math: x86_64 performance degradation due to AVX/SSE transition
790 penalty
791 [20497] localedata: lt_LT: LC_TIME d_fmt used is obsolete
792 [20508] dynamic-link: _dl_runtime_resolve_avx/_dl_runtime_profile_avx512
793 cause transition penalty
794 [20517] math: sparcv9 missing fdiml compat symbol
795 [20524] manual: strverscmp is inconsistent
796 [20525] libc: <sys/quota.h> should be based on kernel headers
797 [20539] math: GCC 7 -static -lm fails to link at -O3
798 [20554] libc: ld: bss-plt forced due to /usr/lib/libc_nonshared.a(ppc-
799 mcount.oS)
800 [20558] string: POSIX bcopy/bzero decls do not implement Object Size
801 Checking
802 [20591] network: Remove obsolete DNSSEC support
803 [20592] network: DNS resource record type classification macros in
804 <arpa/nameser.h> are incorrect
805 [20593] network: Update DNS RR type definitions
806 [20611] network: getaddrinfo accepts invalid numeric scope IDs
807 [20615] build: glibc build fails when using --with-cpu=power9 --enable-
808 multi-arch
809 [20629] network: libresolv: Remove support for bitstring labels
810 (RES_USEBSTRING)
811 [20647] libc: GLIBC quitting every program - glibc on Pentium-S leads to
812 assertion: "maxidx >= 2"
813 [20660] math: [arm] Use VSQRT
814 [20662] libc: checking whether x86_64-pc-linux-gnu-gcc implicitly enables
815 -fstack-protector no (32bit gcc 6.2.0 pie and ssp enable)
816 [20689] libc: Test for FMA should also check for AVX.
817 [20707] glob: gl_pathv entries not set to NULL with GLOB_DOOFFS
818 [20715] math: iszero macro breaks existing code
819 [20718] math: [powerpc] copysignl raises "invalid" for sNaN
820 [20728] libc: powerpc: Missing TOC stub in clone
821 [20729] build: glibc-2.24 fails to build for i486 with -Os
822 [20750] build: Build fails with default PIE enabled gcc-6.x
823 [20768] math: [alpha] sqrt fegetenv namespace
824 [20785] libc: binutils 2.28 fails to assemble power6/memset.S file in
825 glibc
826 [20787] math: float_t is defined as float incorrectly on x86_64 even if
827 __FLT_EVAL_METHOD__ is 2
828 [20790] network: rpcgen buffer overrun in get_prog_declaration
829 [20822] nptl: race condition in __lll_unlock_elision on powerpc
830 [20829] libc: crypt snprintf namespace
831 [20847] libc: tst-vfork3 failure
832 [20855] math: Default bits/mathdef.h has inappropriate float_t
833 [20859] math: [sh4] FP_ILOGB0 invalid
834 [20864] localedata: iconv: cp936 missing single-byte euro sign (0x80,
835 U+20AC), not same as GBK
836 [20915] dynamic-link: global-dynamic TLS broken on aarch64
837 [20916] math: pow handling of sNaN arguments
838 [20918] build: Building with --enable-nss-crypt fails tst-linkall-static.
839 [20919] math: Bad pow (qNaN, 0) result with -lieee
840 [20940] math: hypot sNaN handling
841 [20947] math: fmax, fmin sNaN handling
842 [20956] libc: debug/tst-backtrace3-6 don't work with -O3 anymore
843 [20964] network: sunrpc: Stack-based buffer overflow in getrpcport with
844 RES_USE_INET6
845 [20971] string: powerpc64/power7 memchr overflows internal pointer check
846 [20973] nptl: robust mutexes: Lost wake-ups
847 [20974] locale: bs_BA: yesexpr/noexpr regexes accept any character
848 [20978] nis: strlen on null pointer in nss_nisplus
849 [20985] nptl: robust mutexes: lowlevelrobustlock assembly on x86 blocks on
850 wrong condition
851 [21014] string: i686 memchr overflows internal pointer check
852 [21019] libc: [mips] n32 lseek incorrect on overflow
853 [21022] libc: [microblaze] __backtrace get_frame_size namespace
854 [21026] libc: [MIPS] readahead syscall is broken on n64
855 [21028] math: Fallback fesetexceptflag should always succeed
856 [21045] libc: [powerpc-nofpu] swapcontext does not restore signal mask
857 [21047] math: arm: fpu_control.h: _FPU_GETCW/_FPU_SETCW is rejected by
858 clang
859 [21053] libc: [SH] Namespace pollution from sys/ucontext.h
860 [21061] librt: [microblaze] librt lost clock_* exports
861 [21073] libc: tunables: insecure environment variables passed to
862 subprocesses with AT_SECURE
863 [21081] string: Missing vzeroupper in memset-vec-unaligned-erms.S
864 \f
865 Version 2.24
866
867 * The minimum Linux kernel version that this version of the GNU C Library
868 can be used with is 3.2, except on i[4567]86 and x86_64, where Linux
869 kernel version 2.6.32 or later suffices (on architectures that already
870 required kernel versions more recent than 3.2, those requirements remain
871 unchanged). Linux 3.2 or later kernel headers are required on all
872 architectures.
873
874 * The pap_AN locale has been deleted. This has been deprecated for a long
875 time. It has been replaced by pap_AW & pap_CW, both of which have long
876 been included in previous releases.
877
878 * The readdir_r and readdir64_r functions have been deprecated. It is
879 recommended to use readdir and readdir64 instead.
880
881 * The type “union wait” has been removed. It was deprecated in the early
882 1990s and never part of POSIX. Application code should use the int type
883 instead of “union wait”.
884
885 * A new NSS action is added to facilitate large distributed system
886 administration. The action, MERGE, allows remote user stores like LDAP
887 to be merged into local user stores like /etc/groups in order to provide
888 easy to use, updated, and managed sets of merged credentials. The new
889 action can be used by configuring it in /etc/nsswitch.conf:
890 group: files [SUCCESS=merge] nis
891 Implemented by Stephen Gallagher (Red Hat).
892
893 * The deprecated __malloc_initialize_hook variable has been removed from the
894 API.
895
896 * The long unused localedef --old-style option has been removed. It hasn't
897 done anything in over 16 years. Scripts using this option can safely
898 drop it.
899
900 * nextupl, nextup, nextupf, nextdownl, nextdown and nextdownf are added to
901 libm. They are defined by TS 18661 and IEEE754-2008. The nextup functions
902 return the next representable value in the direction of positive infinity
903 and the nextdown functions return the next representable value in the
904 direction of negative infinity. These are currently enabled as GNU
905 extensions.
906
907 Security related changes:
908
909 * An unnecessary stack copy in _nss_dns_getnetbyname_r was removed. It
910 could result in a stack overflow when getnetbyname was called with an
911 overly long name. (CVE-2016-3075)
912
913 * Previously, getaddrinfo copied large amounts of address data to the stack,
914 even after the fix for CVE-2013-4458 has been applied, potentially
915 resulting in a stack overflow. getaddrinfo now uses a heap allocation
916 instead. Reported by Michael Petlan. (CVE-2016-3706)
917
918 * The glob function suffered from a stack-based buffer overflow when it was
919 called with the GLOB_ALTDIRFUNC flag and encountered a long file name.
920 Reported by Alexander Cherepanov. (CVE-2016-1234)
921
922 * The Sun RPC UDP client could exhaust all available stack space when
923 flooded with crafted ICMP and UDP messages. Reported by Aldy Hernandez'
924 alloca plugin for GCC. (CVE-2016-4429)
925
926 * The IPv6 name server management code in libresolv could result in a memory
927 leak for each thread which is created, performs a failing naming lookup,
928 and exits. Over time, this could result in a denial of service due to
929 memory exhaustion. Reported by Matthias Schiffer. (CVE-2016-5417)
930
931 The following bugs are resolved with this release:
932
933 [1170] localedata: ne_NP: update Nepali locale definition file
934 [3629] manual: stpcpy description in string.texi refers to MS-DOG instead
935 of MS-DOS.
936 [6527] malloc: [powerpc] Malloc alignment insufficient for PowerPC
937 [6796] math: fdim() does not set errno on overflow
938 [10354] libc: posix_spawn should use vfork() in more cases than presently
939 [11213] localedata: localedata: add copyright disclaimer to locale files
940 [12143] localedata: chr_US: new Cherokee locale
941 [12450] localedata: sgs_LT: new locale
942 [12676] localedata: ln_CD: new locale
943 [13237] localedata: LC_ADDRESS.country_name: update all locales w/latest
944 CLDR data
945 [13304] math: fma, fmaf, fmal produce wrong results
946 [14259] build: --localedir arg to configure is ignored
947 [14499] nptl: Does posix_spawn invoke atfork handlers / use vfork?
948 [14750] libc: Race condition in posix_spawn vfork usage vs signal handlers
949 [14934] localedata: es_CL: wrong first weekday chilean locale
950 [15262] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of
951 romanisation
952 [15263] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of 1/0
953 and +/-
954 [15264] localedata: LC_MESSAGES.yesstr/nostr: lacking in many locales
955 [15368] nptl: raise() is not async-signal-safe
956 [15479] math: ceil, floor, round and trunc raise inexact exception
957 [15578] localedata: kk_KZ: various updates
958 [16003] localedata: pap_AN: punt old locale
959 [16137] localedata: iw_IL: punt old locale
960 [16190] localedata: eo: new esperanto locale
961 [16374] localedata: lv_LV: change currency symbol in LC_MONETARY to euro
962 [16742] malloc: race condition: pthread_atfork() called before first
963 malloc() results in unexpected locking behaviour/deadlocks
964 [16975] localedata: LC_MESSAGES.yesexpr/noexpr: revisit capitalization in
965 all locales
966 [16983] localedata: postal_fmt does not allow %l and %n modifiers
967 [17565] localedata: pt_PT: wrong (work-)week start
968 [17899] math: [powerpc] floorl returns negative zero with FE_DOWNWARD
969 [17950] build: Build fails with -msse
970 [18205] localedata: be_BY*: wrong first_weekday and first_workday
971 [18433] libc: posix_spawn does not return correctly upon failure to
972 execute
973 [18453] localedata: charmaps/IBM875: incorrect codes
974 [18712] string: bits/string2.h incompatible with -O2 -Werror=packed
975 -Wsystem-headers
976 [18896] localedata: he_IL: improvements for currency
977 [18911] localedata: ro_RO: Correcting week day name for "Tuesday" in
978 Romanian locale data
979 [18960] locale: s390: _nl_locale_subfreeres uses larl opcode on misaligned
980 symbol
981 [19056] libc: Deprecate readdir_r
982 [19133] localedata: pt_*: days & months should be lowercase in Portuguese
983 language
984 [19198] localedata: nl_NL: small improvements for Dutch locales
985 [19257] network: Per-thread memory leak in __res_vinit with IPv6
986 nameservers (CVE-2016-5417)
987 [19269] build: tst-audit4 and tst-audit10 failures with gcc-6 on non avx
988 machine
989 [19400] locale: Language missing in "iso-639.def", trivial fix in
990 description
991 [19431] malloc: Deadlock between fflush, getdelim, and fork
992 [19505] libc: Incorrect file descriptor validity checks in
993 posix_spawn_file_actions_add{open,close,dup2}
994 [19509] dynamic-link: dlsym, dlvsym do not report errors through dlerror
995 when using RTLD_NEXT
996 [19512] locale: Stale `#ifndef HAVE_BUILTIN_EXPECT' in
997 `intl/{gettextP,loadinfo}.h'
998 [19534] libc: execle, execlp may use malloc
999 [19568] localedata: *_CH: Swiss locales have inconsistent start of week
1000 [19573] network: res_nclose and __res_maybe_init disagree about name
1001 server initialization, breaking Hesiod
1002 [19575] localedata: Status of GB18030 tables
1003 [19581] localedata: sr_* date_fmt string contains additional newline
1004 [19583] string: SSSE3_Fast_Copy_Backward flag needs to be enabled for AMD
1005 Excavator core
1006 [19592] math: [ldbl-128ibm] ceill incorrect in non-default rounding modes
1007 [19593] math: [ldbl-128ibm] truncl incorrect in non-default rounding modes
1008 [19594] math: [ldbl-128ibm] roundl incorrect in non-default rounding modes
1009 [19595] math: [ldbl-128ibm] fmodl incorrect for results in subnormal
1010 double range
1011 [19602] math: [ldbl-128ibm] fmodl handling of equal arguments with low
1012 part zero incorrect
1013 [19603] math: [ldbl-128ibm] remainderl, remquol incorrect sign handling in
1014 equality tests
1015 [19610] dynamic-link: ldconfig -X removes stale symbolic links
1016 [19613] libc: s390x (64 bit) macro expansion WCOREDUMP and others
1017 [19633] locale: strfmon_l applies global locale to number formatting
1018 [19642] network: Memory leak in getnameinfo
1019 [19648] libc: test-skeleton.c: Do not set RLIMIT_DATA
1020 [19653] libc: Potential for NULL pointer dereference (CWE-476) in
1021 glibc-2.22
1022 [19654] math: [x86_64] Need testcase for BZ #19590 fix
1023 [19671] localedata: Missing Sanity Check for malloc() in 'tst-fmon.c' &
1024 'tst-numeric.c'
1025 [19674] math: [ldbl-128ibm] powl incorrect overflow handling
1026 [19677] math: [ldbl-128ibm] remainderl equality test incorrect for zero
1027 low part
1028 [19678] math: [ldbl-128ibm] nextafterl, nexttowardl incorrect sign of zero
1029 result
1030 [19679] dynamic-link: gcc-4.9.3 C++ exception handling broken due to
1031 unaligned stack
1032 [19726] locale: Converting UCS4LE to INTERNAL with iconv() does not update
1033 pointers and lengths in error-case.
1034 [19727] locale: Converting from/to UTF-xx with iconv() does not always
1035 report errors on UTF-16 surrogates values.
1036 [19755] nscd: nscd assertion failure in gc
1037 [19758] dynamic-link: Typo in EXTRA_LD_ENVVARS for x86-64
1038 [19759] libc: mempcpy shouldn't be inlined
1039 [19762] dynamic-link: HAS_CPU_FEATURE/HAS_ARCH_FEATURE are easy to misuse
1040 [19765] libc: s390 needs an optimized mempcpy
1041 [19779] glob: glob: buffer overflow with GLOB_ALTDIRFUNC due to incorrect
1042 NAME_MAX limit assumption (CVE-2016-1234)
1043 [19783] build: benchtests don't support --enable-hardcoded-path-in-tests
1044 [19787] network: Missing and incorrect truncation checks in getnameinfo
1045 [19790] math: [ldbl-128ibm] nearbyintl incorrect in non-default rounding
1046 modes
1047 [19791] network: Assertion failure in res_query.c with un-connectable name
1048 server addresses
1049 [19792] libc: MIPS: backtrace yields infinite backtrace with makecontext
1050 [19822] math: libm.so install clobbers old version
1051 [19825] network: resolv: send_vc can return uninitialized data in second
1052 response to getaddrinfo
1053 [19830] network: nss_dns: should check RDATA length against buffer length
1054 [19831] network: nss_dns: getaddrinfo returns uninitialized data when
1055 confronted with A/AAAA records of invalid size
1056 [19837] nss: nss_db: No retries for some long lines with a larger buffer
1057 [19848] math: powl(10,n) for n=-4,-5,-6,-7 is off by more than 1 ULP
1058 [19853] stdio: Printing IBM long double in decimal with high precision is
1059 sometimes incorrect
1060 [19860] build: x86_64: compile errors for tst-audit10 and tst-auditmod10b
1061 [19861] nptl: libpthread IFUNC resolver for fork can lead to crash
1062 [19862] network: resolv, nss_dns: Remove remaining logging of unexpected
1063 record types
1064 [19865] network: Assertion failure or memory leak in
1065 _nss_dns_getcanonname_r
1066 [19868] network: nss_dns: netent code does not skip over non-PTR records
1067 [19879] network: nss_dns: Stack overflow in getnetbyname implementation
1068 (CVE-2016-3075)
1069 [19881] string: Improve x86-64 memset
1070 [19907] string: Incorrect memcpy tests
1071 [19916] dynamic-link: S390: fprs/vrs are not saved/restored while
1072 resolving symbols
1073 [19925] libc: termios.h XCASE namespace
1074 [19928] string: memmove-vec-unaligned-erms.S is slow with large data size
1075 [19929] libc: limits.h NL_NMAX namespace
1076 [19931] stdio: Memory leak in vfprintf
1077 [19957] libc: clone(CLONE_VM) access invalid parent memory
1078 [19963] localedata: en_IL: New locale
1079 [19989] stdio: stdio.h cuserid namespace
1080 [19994] network: getaddrinfo does not restore RES_USE_INET6 flag in
1081 gethosts
1082 [19996] locale: langinfo.h nl_langinfo_l namespace
1083 [20005] stdio: fflush on a file opened with fmemopen resets position to 0
1084 [20010] network: getaddrinfo: Stack overflow in hostent translation
1085 (CVE-2016-3706)
1086 [20012] stdio: libio: fmemopen append mode failure
1087 [20014] stdio: stdio.h namespace for pre-threads POSIX
1088 [20017] network: resolv: Use gmtime_r instead of gmtime in p_secstodate
1089 [20023] libc: fcntl.h timespec namespace
1090 [20024] math: [x86_64] vectorized sincos trashes the stack
1091 [20031] network: nss_hesiod: Heap overflow in get_txt_records
1092 [20041] time: sys/time.h timespec namespace
1093 [20043] libc: unistd.h missing cuserid for UNIX98 and before
1094 [20044] libc: unistd.h missing pthread_atfork for UNIX98
1095 [20051] libc: ttyslot in wrong header under wrong conditions
1096 [20054] libc: gethostname not declared for XPG4
1097 [20055] libc: termios.h missing tcgetsid for XPG4
1098 [20072] dynamic-link: x86 init_cpu_features is called twice in static
1099 executable
1100 [20073] libc: sys/stat.h fchmod namespace
1101 [20074] libc: stdlib.h rand_r namespace
1102 [20076] libc: sys/stat.h missing S_IFSOCK, S_ISSOCK for XPG4
1103 [20094] libc: stdlib.h should not declare grantpt, ptsname, unlockpt for
1104 XPG3
1105 [20111] libc: struct sockaddr_storage cannot be aggregate-copied
1106 [20112] network: sunrpc: stack (frame) overflow in Sun RPC clntudp_call
1107 (CVE-2016-4429)
1108 [20115] string: Extra alignment in memset-vec-unaligned-erms.S
1109 [20119] libc: Wrong mask for processors level type from CPUID
1110 [20139] dynamic-link: Upper part of zmm is zeroed if Glibc is built with
1111 AS not supporting AVX512
1112 [20151] math: [ldbl-128/ldbl-128ibm] j0l, j1l, y0l, y1l return sNaN for
1113 sNaN argument
1114 [20153] math: [ldbl-128ibm] sqrtl (sNaN) returns sNaN
1115 [20156] math: [ldbl-128ibm] ceill, rintl etc. return sNaN for sNaN
1116 argument
1117 [20157] math: [powerpc] fabsl (sNaN) wrongly raises "invalid"
1118 [20160] math: [powerpc] ceil, rint etc. return sNaN for sNaN input
1119 [20178] libc: posix_spawn{p} should not call exit
1120 [20191] stdio: libio: vtables hardening
1121 [20195] string: FMA4 detection requires CPUID execution with register
1122 eax=0x80000001
1123 [20198] libc: quick_exit incorrectly destroys C++11 thread objects.
1124 [20205] math: [i386/x86_64] nextafterl incorrect incrementing negative
1125 subnormals
1126 [20212] math: acos (sNaN) returns sNaN
1127 [20213] math: asin (sNaN) returns sNaN
1128 [20214] network: Linux header sync with linux/in6.h and ipv6.h again.
1129 [20218] math: [i386] asinhl (sNaN) returns sNaN
1130 [20219] math: [i386] atanhl (sNaN) returns sNaN
1131 [20222] stdio: fopencookie: Mangle function pointers
1132 [20224] math: [i386] cbrtl (sNaN) returns sNaN
1133 [20225] math: ldexp, scalbn, scalbln return sNaN for sNaN input
1134 [20226] math: [i386/x86_64] expl, exp10l, expm1l return sNaN for sNaN
1135 input
1136 [20227] math: [i386/x86_64] logl (sNaN) returns sNaN
1137 [20228] math: [i386/x86_64] log10l (sNaN) returns sNaN
1138 [20229] math: [i386/x86_64] log1pl (sNaN) returns sNaN
1139 [20232] math: [ldbl-128] expm1l (sNaN) returns sNaN
1140 [20233] math: [ldbl-128ibm] expm1l (sNaN) returns sNaN
1141 [20234] math: [ldbl-128ibm] log1pl (sNaN) returns sNaN
1142 [20235] math: [i386/x86_64] log2l (sNaN) returns sNaN
1143 [20237] nss: nss_db: get*ent segfaults without preceding set*ent
1144 [20240] math: modf (sNaN) returns sNaN
1145 [20248] libc: debug/tst-longjump_chk2 calls printf from a signal handler
1146 [20250] math: frexp (sNaN) returns sNaN
1147 [20252] math: atan2 (sNaN, qNaN) fails to raise "invalid"
1148 [20255] math: [i386] fdim, fdimf return with excess range and precision /
1149 double rounding
1150 [20256] math: [i386/x86_64] fdiml returns sNaN for sNaN input
1151 [20260] string: ../sysdeps/x86/bits/string.h:1092:3: error: array
1152 subscript is below array bounds [-Werror=array-bounds]
1153 [20262] nis: _nss_nis_initgroups_dyn always returns NSS_STATUS_NOTFOUND
1154 [20263] nptl: robust mutex deadlocks if other thread requests timedlock
1155 (Only arm/linux)
1156 [20277] libc: $dp is not initialized correctly in sysdeps/hppa/start.S
1157 [20284] malloc: malloc: Corrupt arena avoidance causes unnecessary mmap
1158 fallbacks
1159 [20296] math: [i386/x86_64] scalbl returns sNaN for sNaN input, missing
1160 "invalid" exceptions
1161 [20314] nptl: make[4]: *** [/usr/include/stdlib.h] Error 1
1162 [20316] localedata: id_ID: Februari instead of Pebruari
1163 [20327] string: POWER8 strcasecmp returns incorrect result
1164 [20347] math: Failure: Test: j0_downward (0xap+0)
1165 [20348] libc: FAIL: misc/tst-preadvwritev64
1166 [20349] libc: 64-bit value is passed differently in p{readv,writev}{64}
1167 [20350] libc: There is no test for p{read,write}64
1168 [20357] math: Incorrect cos result for 1.5174239687223976
1169 [20384] build: Don't run libmvec-sincos-avx* tests on non avx machines
1170 \f
1171 Version 2.23
1172
1173 * Unicode 8.0.0 Support: Character encoding, character type info, and
1174 transliteration tables are all updated to Unicode 8.0.0, using new
1175 and/or improved generator scripts contributed by Mike FABIAN (Red Hat).
1176 These updates cause user visible changes, such as the fixes for bugs
1177 89, 16061, and 18568.
1178
1179 * sched_setaffinity, pthread_setaffinity_np no longer attempt to guess the
1180 kernel-internal CPU set size. This means that requests that change the
1181 CPU affinity which failed before (for example, an all-ones CPU mask) will
1182 now succeed. Applications that need to determine the effective CPU
1183 affinities need to call sched_getaffinity or pthread_getaffinity_np after
1184 setting it because the kernel can adjust it (and the previous size check
1185 would not detect this in the majority of cases).
1186
1187 * The fts.h header can now be used with -D_FILE_OFFSET_BITS=64. With LFS
1188 the following new symbols are used: fts64_children, fts64_close,
1189 fts64_open, fts64_read and fts64_set.
1190
1191 * getaddrinfo now detects certain invalid responses on an internal netlink
1192 socket. If such responses are received, an affected process will
1193 terminate with an error message of "Unexpected error <number> on netlink
1194 descriptor <number>" or "Unexpected netlink response of size <number> on
1195 descriptor <number>". The most likely cause for these errors is a
1196 multi-threaded application which erroneously closes and reuses the netlink
1197 file descriptor while it is used by getaddrinfo.
1198
1199 * A defect in the malloc implementation, present since glibc 2.15 (2012) or
1200 glibc 2.10 via --enable-experimental-malloc (2009), could result in the
1201 unnecessary serialization of memory allocation requests across threads.
1202 The defect is now corrected. Users should see a substantial increase in
1203 the concurent throughput of allocation requests for applications which
1204 trigger this bug. Affected applications typically create and
1205 destroy threads frequently. (Bug 19048 was reported and analyzed by
1206 Ericsson.)
1207
1208 * There is now a --disable-timezone-tools configure option for disabling the
1209 building and installing of the timezone related utilities (zic, zdump, and
1210 tzselect). This is useful for people who build the timezone data and code
1211 independent of the GNU C Library.
1212
1213 * The obsolete header <regexp.h> has been removed. Programs that require
1214 this header must be updated to use <regex.h> instead.
1215
1216 * The obsolete functions bdflush, create_module, get_kernel_syms,
1217 query_module and uselib are no longer available to newly linked binaries;
1218 the header <sys/kdaemon.h> has been removed. These functions and header
1219 were specific to systems using the Linux kernel and could not usefully be
1220 used with the GNU C Library on systems with version 2.6 or later of the
1221 Linux kernel.
1222
1223 * Optimized string, wcsmbs and memory functions for IBM z13.
1224 Implemented by Stefan Liebler.
1225
1226 * Newly linked programs that define a variable called signgam will no longer
1227 have it set by the lgamma, lgammaf and lgammal functions. Programs that
1228 require signgam to be set by those functions must ensure that they use the
1229 variable provided by the GNU C Library and declared in <math.h>, without
1230 defining their own copy.
1231
1232 * The minimum GCC version that can be used to build this version of the GNU
1233 C Library is GCC 4.7. Older GCC versions, and non-GNU compilers, can
1234 still be used to compile programs using the GNU C Library.
1235
1236 Security related changes:
1237
1238 * An out-of-bounds value in a broken-out struct tm argument to strftime no
1239 longer causes a crash. Reported by Adam Nielsen. (CVE-2015-8776)
1240
1241 * The LD_POINTER_GUARD environment variable can no longer be used to disable
1242 the pointer guard feature. It is always enabled. Previously,
1243 LD_POINTER_GUARD could be used to disable security hardening in binaries
1244 running in privileged AT_SECURE mode. Reported by Hector Marco-Gisbert.
1245 (CVE-2015-8777)
1246
1247 * An integer overflow in hcreate and hcreate_r could lead to an
1248 out-of-bounds memory access. Reported by Szabolcs Nagy. (CVE-2015-8778)
1249
1250 * The catopen function no longer has unbounded stack usage. Reported by
1251 Max. (CVE-2015-8779)
1252
1253 * The nan, nanf and nanl functions no longer have unbounded stack usage
1254 depending on the length of the string passed as an argument to the
1255 functions. Reported by Joseph Myers. (CVE-2014-9761)
1256
1257 * A stack-based buffer overflow was found in libresolv when invoked from
1258 libnss_dns, allowing specially crafted DNS responses to seize control
1259 of execution flow in the DNS client. The buffer overflow occurs in
1260 the functions send_dg (send datagram) and send_vc (send TCP) for the
1261 NSS module libnss_dns.so.2 when calling getaddrinfo with AF_UNSPEC
1262 family. The use of AF_UNSPEC triggers the low-level resolver code to
1263 send out two parallel queries for A and AAAA. A mismanagement of the
1264 buffers used for those queries could result in the response of a query
1265 writing beyond the alloca allocated buffer created by
1266 _nss_dns_gethostbyname4_r. Buffer management is simplified to remove
1267 the overflow. Thanks to the Google Security Team and Red Hat for
1268 reporting the security impact of this issue, and Robert Holiday of
1269 Ciena for reporting the related bug 18665. (CVE-2015-7547)
1270
1271 The following bugs are resolved with this release:
1272
1273 [89] localedata: Locales nb_NO and nn_NO should transliterate æøå
1274 [887] math: Math library function "logb" and "nextafter" inconsistent
1275 [2542] math: Incorrect return from float gamma (-0X1.FA471547C2FE5P+1)
1276 [2543] math: Incorrect return from float gamma (-0X1.9260DCP+1)
1277 [2558] math: Incorrect return from double gamma (-0X1.FA471547C2FE5P+1)
1278 [2898] libc: [improve] warning: the use of `mktemp' is dangerous, better
1279 use `mkstemp'
1280 [4404] localedata: German translation of "Alarm clock" is misleading
1281 [6799] math: nextafter() and nexttoward() doen't set errno on
1282 overflow/underflow errors
1283 [6803] math: scalb(), scalbln(), scalbn() do not set errno on
1284 overflow/underflow
1285 [10432] nis: _nss_nis_setnetgrent assertion failure
1286 [11460] libc: fts has no LFS support
1287 [12926] network: getaddrinfo()/make_request() may spin forever
1288 [13065] nptl: Race condition in pthread barriers
1289 [13690] nptl: pthread_mutex_unlock potentially cause invalid access
1290 [14341] dynamic-link: Dynamic linker crash when DT_JMPREL and DT_REL{,A}
1291 are not contiguous
1292 [14551] math: [ldbl-128ibm] strtold overflow handling for IBM long double
1293 [14912] libc: Rename non-installed bits/*.h headers
1294 [15002] libc: Avoid undefined behavior in posix_fallocate overflow check
1295 [15367] math: Let gcc use __builtin_isinf
1296 [15384] math: One constant fewer in ieee754/dbl-64/wordsize-64/s_finite.c
1297 [15421] math: lgamma wrongly sets signgam for ISO C
1298 [15470] math: [arm] On ARM llrintl() and llroundl() do not raise
1299 FE_INVALID with argument out of range
1300 [15491] math: [i386/x86_64] x86 nearbyint implementations wrongly clear
1301 all exceptions
1302 [15786] dynamic-link: ifunc resolver functions can smash function
1303 arguments
1304 [15918] math: Unnecessary check for equality in hypotf()
1305 [16061] localedata: Review / update transliteration data
1306 [16068] math: [i386/x86_64] x86 and x86_64 fesetenv exclude state they
1307 should include
1308 [16141] time: strptime %z offset restriction
1309 [16171] math: drem should be alias of remainder
1310 [16296] math: fegetround is pure?
1311 [16347] math: [ldbl-128ibm] ldbl-128/e_lgammal_r.c may not be suitable.
1312 [16364] libc: sleep may leave SIGCHLD blocked on sync cancellation on
1313 GNU/Linux
1314 [16399] math: [mips] lrint / llrint / lround / llround missing exceptions
1315 [16415] math: Clean up ldbl-128 / ldbl-128ibm expm1l for large positive
1316 arguments
1317 [16422] math: [powerpc] math-float, math-double failing llrint tests with
1318 "Exception "Inexact" set" on ppc32
1319 [16495] localedata: nl_NL: date_fmt: shuffle year/month around
1320 [16517] math: Missing underflow exception from tanf/tan/tanl
1321 [16519] math: Missing underflow exception from sinhf
1322 [16520] math: Missing underflow exception from tanhf
1323 [16521] math: Missing underflow exception from exp2
1324 [16620] math: [ldbl-128ibm] exp10l spurious overflows / bad directed
1325 rounding results
1326 [16734] stdio: fopen calls mmap to allocate its buffer
1327 [16961] math: nan function incorrect handling of bad sequences
1328 [16962] math: nan function unbounded stack allocation (CVE-2014-9761)
1329 [16973] localedata: Fix lang_lib/lang_term as per ISO 639-2
1330 [16985] locale: localedef: confusing error message when opening output
1331 fails
1332 [17118] math: ctanh(INFINITY + 2 * I) returns incorrect value
1333 [17197] locale: Redundant shift character in iconv conversion output at
1334 block boundary
1335 [17243] libc: trunk/posix/execl.c:53: va_args problem ?
1336 [17244] libc: trunk/sysdeps/unix/sysv/linux/semctl.c:116: va_args muxup ?
1337 [17250] dynamic-link: static linking breaks nss loading
1338 (getaddrinfo/getpwnam/etc...)
1339 [17404] libc: atomic_exchange_rel lacking a barrier on MIPS16, GCC before
1340 4.7?
1341 [17441] math: isnan() should use __builtin_isnan() in GCC
1342 [17514] nptl: Assert failure unlocking ERRORCHECK mutex after timedlock
1343 (related to lock elision)
1344 [17787] manual: Exponent on page 324 of the PDF ends prematurely
1345 [17886] time: strptime should be able to parse "Z" as a timezone with %z
1346 [17887] time: strptime should be able to parse "+01:00" style timezones
1347 [17905] libc: catopen() Multiple unbounded stack allocations
1348 (CVE-2015-8779)
1349 [18084] libc: backtrace (..., 0) dumps core on x86
1350 [18086] libc: nice() sets errno to 0 on success
1351 [18240] libc: hcreate, hcreate_r should fail with ENOMEM if element count
1352 is too large (CVE-2015-8778)
1353 [18251] dynamic-link: SONAME missing when audit modules provides path
1354 [18265] libc: add attributes for wchar string and memory functions
1355 [18370] math: csqrt missing underflows
1356 [18421] libc: [hppa] read-only segment has dynamic relocations
1357 [18472] libc: Obsolete syscall wrappers should be compat symbols
1358 [18480] libc: hppa glibc miscompilation in sched_setaffinity()
1359 [18491] localedata: Update tr_TR LC_CTYPE as part of Unicode updates
1360 [18525] localedata: Remove locale timezone information
1361 [18560] libc: [powerpc] spurious bits/ipc.h definitions
1362 [18568] localedata: Update locale data to Unicode 8.0
1363 [18589] locale: sort-test.sh fails at random
1364 [18595] math: ctan, ctanh missing underflows
1365 [18604] libc: assert macro-expands its argument
1366 [18610] math: S390: fetestexcept() reports any exception if DXC-code
1367 contains a vector instruction exception.
1368 [18611] math: j1, jn missing errno setting on underflow
1369 [18618] localedata: sync Chechen locale definitions with other *_RU
1370 locales
1371 [18647] math: powf(-0x1.000002p0, 0x1p30) returns 0 instead of +inf
1372 [18661] libc: Some x86-64 assembly codes don't align stack to 16 bytes
1373 [18665] network: In send_dg, the recvfrom function is NOT always using the
1374 buffer size of a newly created buffer (CVE-2015-7547)
1375 [18674] libc: [i386] trunk/sysdeps/i386/tst-auditmod3b.c:84: possible
1376 missing break ?
1377 [18675] libc: fpathconf(_PC_NAME_MAX) fails against large filesystems for
1378 32bit processes
1379 [18681] libc: regexp.h is obsolete and buggy, and should be desupported
1380 [18699] math: tilegx cproj() for various complex infinities does not yield
1381 infinity
1382 [18724] libc: Harden put*ent functions against data injection
1383 [18743] nptl: PowerPC: findutils testcase fails with --enable-lock-elision
1384 [18755] build: build errors with -DNDEBUG
1385 [18757] stdio: fmemopen fails to set errno on failure
1386 [18778] dynamic-link: ld.so crashes if failed dlopen causes libpthread to
1387 be forced unloaded
1388 [18781] libc: openat64 lacks O_LARGEFILE
1389 [18787] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/atomic.h:71:6:
1390 error: can’t find a register in class ‘R1_REGS’ while reloading ‘asm’
1391 [18789] math: [ldbl-128ibm] sinhl inaccurate near 0
1392 [18790] math: [ldbl-128ibm] tanhl inaccurate
1393 [18795] libc: stpncpy fortification misses buffer lengths that are
1394 statically too large
1395 [18796] build: build fails for --disable-mathvec
1396 [18803] math: hypot missing underflows
1397 [18820] stdio: fmemopen may leak memory on failure
1398 [18823] math: csqrt spurious underflows
1399 [18824] math: fma spurious underflows
1400 [18825] math: pow missing underflows
1401 [18857] math: [ldbl-128ibm] nearbyintl wrongly uses signaling comparisons
1402 [18868] nptl: pthread_barrier_init typo has in-theory-undefined behavior
1403 [18870] build: sem_open.c fails to compile with missing symbol
1404 FUTEX_SHARED
1405 [18872] stdio: Fix memory leak in printf_positional
1406 [18873] libc: posix_fallocate overflow check ineffective
1407 [18875] math: Excess precision leads incorrect libm
1408 [18877] libc: arm: mmap offset regression
1409 [18887] libc: memory corruption when using getmntent on blank lines
1410 [18918] localedata: hu_HU: change time to HH:MM:SS format
1411 [18921] libc: Regression: extraneous stat() and fstat() performed by
1412 opendir()
1413 [18928] dynamic-link: LD_POINTER_GUARD is not ignored for privileged
1414 binaries (CVE-2015-8777)
1415 [18951] math: tgamma missing underflows
1416 [18952] math: [ldbl-128/ldbl-128ibm] lgammal spurious "invalid", incorrect
1417 signgam
1418 [18953] localedata: lt_LT: change currency symbol to the euro
1419 [18956] math: powf inaccuracy
1420 [18961] math: [i386] exp missing underflows
1421 [18966] math: [i386] exp10 missing underflows
1422 [18967] math: math.h XSI POSIX namespace (gamma, isnan, scalb)
1423 [18969] build: multiple string test failures due to missing locale
1424 dependencies
1425 [18970] libc: Reference of pthread_setcancelstate in libc.a
1426 [18977] math: float / long double Bessel functions not in XSI POSIX
1427 [18980] math: i386 libm functions return with excess range and precision
1428 [18981] math: i386 scalb*, ldexp return with excess range and precision
1429 [18982] stdio: va_list and vprintf
1430 [18985] time: Passing out of range data to strftime() causes a segfault
1431 (CVE-2015-8776)
1432 [19003] math: [x86_64] fma4 version of pow inappropriate contraction
1433 [19007] libc: FAIL: elf/check-localplt with -z now and binutils 2.26
1434 [19012] locale: iconv_open leaks memory on error path
1435 [19016] math: clog, clog10 inaccuracy
1436 [19018] nptl: Mangle function pointers in tls_dtor_list
1437 [19032] math: [i386] acosh (-qNaN) spurious "invalid" exception
1438 [19046] math: ldbl-128 / ldbl-128ibm lgamma bad overflow handling
1439 [19048] malloc: malloc: arena free list can become cyclic, increasing
1440 contention
1441 [19049] math: [powerpc] erfc incorrect zero sign
1442 [19050] math: [powerpc] log* incorrect zero sign
1443 [19058] math: [x86_64] Link fail with -fopenmp and -flto
1444 [19059] math: nexttoward overflow incorrect in non-default rounding modes
1445 [19071] math: ldbl-96 lroundl incorrect just below powers of 2
1446 [19074] network: Data race in _res_hconf_reorder_addrs
1447 [19076] math: [ldbl-128ibm] log1pl (-1) wrong sign of infinity
1448 [19077] math: [ldbl-128ibm] logl (1) incorrect sign of zero result
1449 [19078] math: [ldbl-128ibm] expl overflow incorrect in non-default
1450 rounding modes
1451 [19079] math: dbl-64/wordsize-64 lround based on llround incorrect for
1452 ILP32
1453 [19085] math: ldbl-128 lrintl, lroundl missing exceptions for 32-bit long
1454 [19086] manual: posix_fallocate64 documented argument order is wrong.
1455 [19088] math: lround, llround missing exceptions close to overflow
1456 threshold
1457 [19094] math: lrint, llrint missing exceptions close to overflow threshold
1458 [19095] math: dbl-64 lrint incorrect for 64-bit long
1459 [19122] dynamic-link: Unnecessary PLT relocations in librtld.os
1460 [19124] dynamic-link: ld.so failed to build with older assmebler
1461 [19125] math: [powerpc32] llroundf, llround incorrect exceptions
1462 [19129] dynamic-link: [arm] Concurrent lazy TLSDESC resolution can crash
1463 [19134] math: [powerpc32] lround, lroundf spurious exceptions
1464 [19137] libc: i386/epoll_pwait.S doesn't support cancellation
1465 [19143] nptl: Remove CPU set size checking from sched_setaffinity,
1466 pthread_setaffinity_np
1467 [19156] math: [ldbl-128] j0l spurious underflows
1468 [19164] nptl: tst-getcpu fails with many possible CPUs
1469 [19168] math: math/test-ildoubl and math/test-ldouble failure
1470 [19174] nptl: PowerPC: TLE enabled pthread mutex performs poorly.
1471 [19178] dynamic-link: ELF_RTYPE_CLASS_EXTERN_PROTECTED_DATA confuses
1472 prelink
1473 [19181] math: [i386/x86_64] fesetenv (FE_DFL_ENV), fesetenv
1474 (FE_NOMASK_ENV) do not clear SSE exceptions
1475 [19182] malloc: malloc deadlock between ptmalloc_lock_all and
1476 _int_new_arena/reused_arena
1477 [19189] math: [ldbl-128] log1pl (-qNaN) spurious "invalid" exception
1478 [19201] math: dbl-64 remainder incorrect sign of zero result
1479 [19205] math: bits/math-finite.h conditions do not match math.h and
1480 bits/mathcalls.h
1481 [19209] math: bits/math-finite.h wrongly maps ldexp to scalbn
1482 [19211] math: lgamma functions do not set signgam for -ffinite-math-only
1483 for C99-based standards
1484 [19212] libc: features.h not -Wundef clean
1485 [19213] math: [i386/x86_64] log* (1) incorrect zero sign for -ffinite-
1486 math-only
1487 [19214] libc: Family and model identification for AMD CPU's are incorrect.
1488 [19219] libc: GLIBC build fails for ia64 with missing __nearbyintl
1489 [19228] math: [powerpc] nearbyint wrongly clears "inexact", leaves traps
1490 disabled
1491 [19235] math: [powerpc64] lround, lroundf, llround, llroundf spurious
1492 "inexact" exceptions
1493 [19238] math: [powerpc] round, roundf spurious "inexact" for integer
1494 arguments
1495 [19242] libc: strtol incorrect in Turkish locales
1496 [19243] malloc: reused_arena can pick an arena on the free list, leading
1497 to an assertion failure and reference count corruption
1498 [19253] time: tzset() ineffective when temporary TZ did not include DST
1499 rules
1500 [19266] math: strtod ("NAN(I)") incorrect in Turkish locales
1501 [19270] math: [hppa] Shared libm missing __isnanl
1502 [19285] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/mman.h: missing
1503 MAP_HUGETLB and MAP_STACK defines
1504 [19313] nptl: Wrong __cpu_mask for x32
1505 [19347] libc: grantpt: try to force a specific gid even without pt_chown
1506 [19349] math: [ldbl-128ibm] tanhl inaccurate for small arguments
1507 [19350] math: [ldbl-128ibm] sinhl spurious overflows
1508 [19351] math: [ldbl-128ibm] logl inaccurate near 1
1509 [19363] time: x32: times() return value wrongly truncates/sign extends
1510 from 32bit
1511 [19367] dynamic-link: Improve branch prediction on Silvermont
1512 [19369] network: Default domain name not reset by res_ninit when "search"
1513 / "domain" entry is removed from resolv.conf
1514 [19375] math: powerpc: incorrect results for POWER7 logb with negative
1515 subnormals
1516 [19385] localedata: bg_BG: time separator should be colon, not comma
1517 [19408] libc: linux personality syscall wrapper may erroneously return an
1518 error on 32-bit architectures
1519 [19415] libc: dladdr returns wrong names on hppa
1520 [19432] libc: iconv rejects redundant escape sequences in IBM900, IBM903,
1521 IBM905, IBM907, and IBM909
1522 [19439] math: Unix98 isinf and isnan functions conflict with C++11
1523 [19443] build: build failures with -DDEBUG
1524 [19451] build: Make check fails on test-double-vlen2
1525 [19462] libc: Glibc failed to build with -Os
1526 [19465] math: Wrong code with -Os
1527 [19466] time: time/tst-mktime2.c is compiled into an infinite loop with
1528 -Os
1529 [19467] string: Fast_Unaligned_Load needs to be enabled for Excavator core
1530 CPU's.
1531 [19475] libc: Glibc 2.22 doesn't build on sparc [PATCH]
1532 [19486] math: S390: Math tests fail with "Exception Inexact set".
1533 [19529] libc: [ARM]: FAIL: stdlib/tst-makecontext
1534 [19550] libc: [mips] mmap negative offset handling inconsistent with other
1535 architectures
1536 [19590] math: Fail to build shared objects that use libmvec.so functions.
1537 \f
1538 Version 2.22
1539
1540 * The following bugs are resolved with this release:
1541
1542 438, 2981, 4719, 6544, 6792, 11216, 12836, 13028, 13064, 13151, 13152,
1543 14094, 14113, 14292, 14841, 14906, 14958, 15319, 15467, 15790, 15969,
1544 16159, 16339, 16350, 16351, 16352, 16353, 16361, 16512, 16526, 16538,
1545 16559, 16560, 16704, 16783, 16850, 17053, 17090, 17195, 17269, 17293,
1546 17322, 17403, 17475, 17523, 17542, 17569, 17581, 17588, 17596, 17620,
1547 17621, 17628, 17631, 17692, 17711, 17715, 17776, 17779, 17792, 17833,
1548 17836, 17841, 17912, 17916, 17930, 17932, 17944, 17949, 17964, 17965,
1549 17967, 17969, 17977, 17978, 17987, 17991, 17996, 17998, 17999, 18007,
1550 18019, 18020, 18029, 18030, 18032, 18034, 18036, 18038, 18039, 18042,
1551 18043, 18046, 18047, 18049, 18068, 18078, 18080, 18093, 18100, 18104,
1552 18110, 18111, 18116, 18125, 18128, 18134, 18138, 18185, 18196, 18197,
1553 18206, 18210, 18211, 18217, 18219, 18220, 18221, 18234, 18244, 18245,
1554 18247, 18287, 18319, 18324, 18333, 18346, 18371, 18383, 18397, 18400,
1555 18409, 18410, 18412, 18418, 18422, 18434, 18444, 18457, 18468, 18469,
1556 18470, 18479, 18483, 18495, 18496, 18497, 18498, 18502, 18507, 18508,
1557 18512, 18513, 18519, 18520, 18522, 18527, 18528, 18529, 18530, 18532,
1558 18533, 18534, 18536, 18539, 18540, 18542, 18544, 18545, 18546, 18547,
1559 18549, 18553, 18557, 18558, 18569, 18583, 18585, 18586, 18592, 18593,
1560 18594, 18602, 18612, 18613, 18619, 18633, 18635, 18641, 18643, 18648,
1561 18657, 18676, 18694, 18696, 18887.
1562
1563 * Cache information can be queried via sysconf() function on s390 e.g. with
1564 _SC_LEVEL1_ICACHE_SIZE as argument.
1565
1566 * A buffer overflow in gethostbyname_r and related functions performing DNS
1567 requests has been fixed. If the NSS functions were called with a
1568 misaligned buffer, the buffer length change due to pointer alignment was
1569 not taken into account. This could result in application crashes or,
1570 potentially arbitrary code execution, using crafted, but syntactically
1571 valid DNS responses. (CVE-2015-1781)
1572
1573 * The time zone file parser has been made more robust against crafted time
1574 zone files, avoiding heap buffer overflows related to the processing of
1575 the tzh_ttisstdcnt and tzh_ttisgmtcnt fields, and a stack overflow due to
1576 large time zone data files. Overly long time zone specifiers in the TZ
1577 variable no longer result in stack overflows and crashes.
1578
1579 * A powerpc and powerpc64 optimization for TLS, similar to TLS descriptors
1580 for LD and GD on x86 and x86-64, has been implemented. You will need
1581 binutils-2.24 or later to enable this optimization.
1582
1583 * Character encoding and ctype tables were updated to Unicode 7.0.0, using
1584 new generator scripts contributed by Pravin Satpute and Mike FABIAN (Red
1585 Hat). These updates cause user visible changes, such as the fix for bug
1586 17998.
1587
1588 * CVE-2014-8121 The NSS backends shared internal state between the getXXent
1589 and getXXbyYY NSS calls for the same database, causing a denial-of-service
1590 condition in some applications.
1591
1592 * Added vector math library named libmvec with the following vectorized x86_64
1593 implementations: cos, cosf, sin, sinf, sincos, sincosf, log, logf, exp, expf,
1594 pow, powf.
1595 The library can be disabled with --disable-mathvec. Use of the functions is
1596 enabled with -fopenmp -ffast-math starting from -O1 for GCC version >= 4.9.0.
1597 Shared library libmvec.so is linked in as needed when using -lm (no need to
1598 specify -lmvec explicitly for not static builds).
1599 Visit <https://sourceware.org/glibc/wiki/libmvec> for detailed information.
1600
1601 * A new fmemopen implementation has been added with the goal of POSIX
1602 compliance. The new implementation fixes the following long-standing
1603 issues: BZ#6544, BZ#11216, BZ#12836, BZ#13151, BZ#13152, and BZ#14292. The
1604 old implementation is still present for use be by existing binaries.
1605
1606 * The 32-bit sparc sigaction ABI was inadvertently broken in the 2.20 and 2.21
1607 releases. It has been fixed to match 2.19 and older, but binaries built
1608 against 2.20 and 2.21 might need to be recompiled. See BZ#18694.
1609
1610 * Port to Native Client running on ARMv7-A (--host=arm-nacl).
1611 Contributed by Roland McGrath (Google).
1612
1613 * The header <regexp.h> is deprecated, and will be removed in a future
1614 release. Use of this header will trigger a deprecation warning.
1615 Application developers should update their code to use <regex.h> instead.
1616
1617 This header was formerly part of SUS, but was deprecated in 1994 and
1618 removed from the standard in 2001. Also, the glibc implementation
1619 leaks memory. See BZ#18681 for more details.
1620 \f
1621 Version 2.21
1622
1623 * The following bugs are resolved with this release:
1624
1625 6652, 10672, 12674, 12847, 12926, 13862, 14132, 14138, 14171, 14498,
1626 15215, 15378, 15884, 16009, 16418, 16191, 16469, 16576, 16617, 16618,
1627 16619, 16657, 16740, 16857, 17192, 17266, 17273, 17344, 17363, 17370,
1628 17371, 17411, 17460, 17475, 17485, 17501, 17506, 17508, 17522, 17555,
1629 17570, 17571, 17572, 17573, 17574, 17582, 17583, 17584, 17585, 17589,
1630 17594, 17601, 17608, 17616, 17625, 17630, 17633, 17634, 17635, 17647,
1631 17653, 17657, 17658, 17664, 17665, 17668, 17682, 17702, 17717, 17719,
1632 17722, 17723, 17724, 17725, 17732, 17733, 17744, 17745, 17746, 17747,
1633 17748, 17775, 17777, 17780, 17781, 17782, 17791, 17793, 17796, 17797,
1634 17801, 17803, 17806, 17834, 17844, 17848, 17868, 17869, 17870, 17885,
1635 17892.
1636
1637 * CVE-2015-1472 CVE-2015-1473 Under certain conditions wscanf can allocate
1638 too little memory for the to-be-scanned arguments and overflow the
1639 allocated buffer. The implementation now correctly computes the required
1640 buffer size when using malloc, and switches to malloc from alloca as
1641 intended.
1642
1643 * A new semaphore algorithm has been implemented in generic C code for all
1644 machines. Previous custom assembly implementations of semaphore were
1645 difficult to reason about or ensure that they were safe. The new version
1646 of semaphore supports machines with 64-bit or 32-bit atomic operations.
1647 The new semaphore algorithm is used by sem_init, sem_open, sem_post,
1648 sem_wait, sem_timedwait, sem_trywait, and sem_getvalue.
1649
1650 * Port to Altera Nios II has been contributed by Mentor Graphics.
1651
1652 * Optimized strcpy, stpcpy, strncpy, stpncpy, strcmp, and strncmp
1653 implementations for powerpc64/powerpc64le.
1654 Implemented by Adhemerval Zanella (IBM).
1655
1656 * Added support for TSX lock elision of pthread mutexes on powerpc32, powerpc64
1657 and powerpc64le. This may improve lock scaling of existing programs on
1658 HTM capable systems. The lock elision code is only enabled with
1659 --enable-lock-elision=yes. Also, the TSX lock elision implementation for
1660 powerpc will issue a transaction abort on every syscall to avoid side
1661 effects being visible outside transactions.
1662
1663 * Optimized strcpy, stpcpy, strchrnul and strrchr implementations for
1664 AArch64. Contributed by ARM Ltd.
1665
1666 * i386 memcpy functions optimized with SSE2 unaligned load/store.
1667
1668 * CVE-2104-7817 The wordexp function could ignore the WRDE_NOCMD flag
1669 under certain input conditions resulting in the execution of a shell for
1670 command substitution when the application did not request it. The
1671 implementation now checks WRDE_NOCMD immediately before executing the
1672 shell and returns the error WRDE_CMDSUB as expected.
1673
1674 * CVE-2012-3406 printf-style functions could run into a stack overflow when
1675 processing format strings with a large number of format specifiers.
1676
1677 * CVE-2014-9402 The nss_dns implementation of getnetbyname could run into an
1678 infinite loop if the DNS response contained a PTR record of an unexpected
1679 format.
1680
1681 * The minimum GCC version that can be used to build this version of the GNU
1682 C Library is GCC 4.6. Older GCC versions, and non-GNU compilers, can
1683 still be used to compile programs using the GNU C Library.
1684
1685 * The GNU C Library is now built with -Werror by default. This can be
1686 disabled by configuring with --disable-werror.
1687
1688 * New locales: tu_IN, bh_IN, raj_IN, ce_RU.
1689
1690 * The obsolete sigvec function has been removed. This was the original
1691 4.2BSD interface that inspired the POSIX.1 sigaction interface, which
1692 programs have been using instead for about 25 years. Of course, ABI
1693 compatibility for old binaries using sigvec remains intact.
1694
1695 * Merged gettext 0.19.3 into the intl subdirectory. This fixes building
1696 with newer versions of bison.
1697
1698 * Support for MIPS o32 FPXX, FP64A and FP64 ABI Extensions.
1699 The original MIPS o32 hard-float ABI requires an FPU where double-precision
1700 registers overlay two consecutive single-precision registers. MIPS32R2
1701 introduced a new FPU mode (FR=1) where double-precision registers extend the
1702 corresponding single-precision registers which is incompatible with the
1703 o32 hard-float ABI. The MIPS SIMD ASE and the MIPSR6 architecture both
1704 require the use of FR=1 making a transition necessary. New o32 ABI
1705 extensions enable users to migrate over time from the original o32 ABI
1706 through to the updated o32 FP64 ABI. To achieve this the dynamic linker now
1707 tracks the ABI of any loaded object and verifies that new objects are
1708 compatible. Mode transitions will also be requested as required and
1709 unsupportable objects will be rejected. The ABI checks include both soft and
1710 hard float ABIs for o32, n32 and n64.
1711
1712 GCC 5 with GNU binutils 2.25 onwards:
1713 It is strongly recommended that all o32 system libraries are built using the
1714 new o32 FPXX ABI (-mfpxx) to facilitate the transition as this is compatible
1715 with the original and all new o32 ABI extensions. Configure a MIPS GCC
1716 compiler using --with-fp-32=xx to set this by default.
1717 \f
1718 Version 2.20
1719
1720 * The following bugs are resolved with this release:
1721
1722 6804, 9894, 11505, 12994, 13347, 13651, 14308, 14770, 15119, 15132, 15347,
1723 15514, 15698, 15804, 15894, 15946, 16002, 16064, 16095, 16194, 16198,
1724 16275, 16284, 16287, 16315, 16348, 16349, 16354, 16357, 16362, 16447,
1725 16516, 16532, 16539, 16545, 16561, 16562, 16564, 16574, 16599, 16600,
1726 16609, 16610, 16611, 16613, 16619, 16623, 16629, 16632, 16634, 16639,
1727 16642, 16648, 16649, 16670, 16674, 16677, 16680, 16681, 16683, 16689,
1728 16695, 16701, 16706, 16707, 16712, 16713, 16714, 16724, 16731, 16739,
1729 16740, 16743, 16754, 16758, 16759, 16760, 16770, 16786, 16789, 16791,
1730 16796, 16799, 16800, 16815, 16823, 16824, 16831, 16838, 16839, 16849,
1731 16854, 16876, 16877, 16878, 16882, 16885, 16888, 16890, 16892, 16912,
1732 16915, 16916, 16917, 16918, 16922, 16927, 16928, 16932, 16943, 16958,
1733 16965, 16966, 16967, 16977, 16978, 16984, 16990, 16996, 17009, 17022,
1734 17031, 17042, 17048, 17050, 17058, 17061, 17062, 17069, 17075, 17078,
1735 17079, 17084, 17086, 17088, 17092, 17097, 17125, 17135, 17137, 17150,
1736 17153, 17187, 17213, 17259, 17261, 17262, 17263, 17319, 17325, 17354.
1737
1738 * Reverted change of ABI data structures for s390 and s390x:
1739 On s390 and s390x the size of struct ucontext and jmp_buf was increased in
1740 2.19. This change is reverted in 2.20. The introduced 2.19 symbol versions
1741 of getcontext, setjmp, _setjmp, __sigsetjmp, longjmp, _longjmp, siglongjmp
1742 are preserved pointing straight to the same implementation as the old ones.
1743 Given that, new callers will simply provide a too-big buffer to these
1744 functions. Any applications/libraries out there that embed jmp_buf or
1745 ucontext_t in an ABI-relevant data structure that have already been rebuilt
1746 against 2.19 headers will have to rebuilt again. This is necessary in any
1747 case to revert the breakage in their ABI caused by the glibc change.
1748
1749 * Support for file description locks is added to systems running the
1750 Linux kernel. The standard file locking interfaces are extended to
1751 operate on file descriptions, not file descriptors, via the use of
1752 F_OFD_GETLK, F_OFD_SETLK, and F_OFD_SETLKW. File description locks
1753 are associated with an open file instead of a process.
1754
1755 * Optimized strchr implementation for AArch64. Contributed by ARM Ltd.
1756
1757 * The minimum Linux kernel version that this version of the GNU C Library
1758 can be used with is 2.6.32.
1759
1760 * Running the testsuite no longer terminates as soon as a test fails.
1761 Instead, a file tests.sum (xtests.sum from "make xcheck") is generated,
1762 with PASS or FAIL lines for individual tests. A summary of the results is
1763 printed, including a list of failing lists, and "make check" exits with
1764 error status if there were any unexpected failures. "make check
1765 stop-on-test-failure=y" may be used to keep the old behavior.
1766
1767 * The am33 port, which had not worked for several years, has been removed
1768 from ports.
1769
1770 * The _BSD_SOURCE and _SVID_SOURCE feature test macros are no longer
1771 supported; they now act the same as _DEFAULT_SOURCE (but generate a
1772 warning). Except for cases where _BSD_SOURCE enabled BSD interfaces that
1773 conflicted with POSIX (support for which was removed in 2.19), the
1774 interfaces those macros enabled remain available when compiling with
1775 _GNU_SOURCE defined, with _DEFAULT_SOURCE defined, or without any feature
1776 test macros defined.
1777
1778 * Optimized strcmp implementation for ARMv7. Contributed by ARM Ltd.
1779
1780 * Added support for TX lock elision of pthread mutexes on s390 and s390x.
1781 This may improve lock scaling of existing programs on TX capable systems.
1782 The lock elision code is only built with --enable-lock-elision=yes and
1783 then requires a GCC version supporting the TX builtins. With lock elision
1784 default mutexes are elided via __builtin_tbegin, if the cpu supports
1785 transactions. By default lock elision is not enabled and the elision code
1786 is not built.
1787
1788 * CVE-2014-4043 The posix_spawn_file_actions_addopen implementation did not
1789 copy the path argument. This allowed programs to cause posix_spawn to
1790 deference a dangling pointer, or use an unexpected pathname argument if
1791 the string was modified after the posix_spawn_file_actions_addopen
1792 invocation.
1793
1794 * All supported architectures now use the main glibc sysdeps directory
1795 instead of some being in a separate "ports" directory (which was
1796 distributed separately before glibc 2.17).
1797
1798 * The NPTL implementation of POSIX pthreads is no longer an "add-on".
1799 On configurations that support it (all Linux configurations), it's now
1800 used regardless of the --enable-add-ons switch to configure. It is no
1801 longer possible to build such configurations without pthreads support.
1802
1803 * Locale names, including those obtained from environment variables (LANG
1804 and the LC_* variables), are more tightly checked for proper syntax.
1805 setlocale will now fail (with EINVAL) for locale names that are overly
1806 long, contain slashes without starting with a slash, or contain ".." path
1807 components. (CVE-2014-0475) Previously, some valid locale names were
1808 silently replaced with the "C" locale when running in AT_SECURE mode
1809 (e.g., in a SUID program). This is no longer necessary because of the
1810 additional checks.
1811
1812 * On x86-64, the dynamic linker's lazy-binding support is now compatible
1813 with application code using Intel MPX instructions. (With all previous
1814 versions, the MPX register state could be clobbered when making calls
1815 into or out of a shared library.) Note that while the new dynamic
1816 linker is compatible with all known x86 hardware whether or not it
1817 supports Intel MPX, some x86 instruction-set emulators might fail to
1818 handle the new instruction encodings. This is known to affect Valgrind
1819 versions up through 3.9 (but will be fixed in the forthcoming 3.10
1820 release), and might affect other tools that do instruction emulation.
1821
1822 * Support for loadable gconv transliteration modules has been removed.
1823 The support for transliteration modules has been non-functional for
1824 over a decade, and the removal is prompted by security defects. The
1825 normal gconv conversion modules are still supported. Transliteration
1826 with //TRANSLIT is still possible, and the //IGNORE specifier
1827 continues to be supported. (CVE-2014-5119)
1828
1829 * Decoding a crafted input sequence in the character sets IBM933, IBM935,
1830 IBM937, IBM939, IBM1364 could result in an out-of-bounds array read,
1831 resulting a denial-of-service security vulnerability in applications which
1832 use functions related to iconv. (CVE-2014-6040)
1833 \f
1834 Version 2.19
1835
1836 * The following bugs are resolved with this release:
1837
1838 156, 387, 431, 762, 832, 926, 2801, 4772, 6786, 6787, 6807, 6810, 6981,
1839 7003, 9721, 9954, 10253, 10278, 11087, 11157, 11214, 12100, 12486, 12751,
1840 12986, 13028, 13982, 13985, 14029, 14032, 14120, 14143, 14155, 14286,
1841 14547, 14699, 14752, 14782, 14876, 14910, 15004, 15048, 15073, 15089,
1842 15128, 15218, 15268, 15277, 15308, 15362, 15374, 15400, 15425, 15427,
1843 15483, 15522, 15531, 15532, 15593, 15601, 15608, 15609, 15610, 15632,
1844 15640, 15670, 15672, 15680, 15681, 15723, 15734, 15735, 15736, 15748,
1845 15749, 15754, 15760, 15763, 15764, 15797, 15799, 15825, 15843, 15844,
1846 15846, 15847, 15849, 15850, 15855, 15856, 15857, 15859, 15867, 15886,
1847 15887, 15890, 15892, 15893, 15895, 15897, 15901, 15905, 15909, 15915,
1848 15917, 15919, 15921, 15923, 15939, 15941, 15948, 15963, 15966, 15968,
1849 15985, 15988, 15997, 16032, 16034, 16036, 16037, 16038, 16041, 16046,
1850 16055, 16071, 16072, 16074, 16077, 16078, 16103, 16112, 16143, 16144,
1851 16146, 16150, 16151, 16153, 16167, 16169, 16172, 16195, 16214, 16245,
1852 16271, 16274, 16283, 16289, 16293, 16314, 16316, 16330, 16337, 16338,
1853 16356, 16365, 16366, 16369, 16372, 16375, 16379, 16381, 16384, 16385,
1854 16386, 16387, 16390, 16394, 16398, 16400, 16407, 16408, 16414, 16430,
1855 16431, 16453, 16474, 16506, 16510, 16529
1856
1857 * Slovenian translations for glibc messages have been contributed by the
1858 Translation Project's Slovenian team of translators.
1859
1860 * The public headers no longer use __unused nor __block. This change is to
1861 support compiling programs that are derived from BSD sources and use
1862 __unused internally, and to support compiling with Clang's -fblock
1863 extension which uses __block.
1864
1865 * CVE-2012-4412 The strcoll implementation caches indices and rules for
1866 large collation sequences to optimize multiple passes. This cache
1867 computation may overflow for large collation sequences and may cause a
1868 stack or buffer overflow. This is now fixed to use a slower algorithm
1869 which does not use a cache if there is an integer overflow.
1870
1871 * CVE-2012-4424 The strcoll implementation uses malloc to cache indices and
1872 rules for large collation sequences to optimize multiple passes and falls
1873 back to alloca if malloc fails, resulting in a possible stack overflow.
1874 The implementation now falls back to an uncached collation sequence lookup
1875 if malloc fails.
1876
1877 * CVE-2013-4788 The pointer guard used for pointer mangling was not
1878 initialized for static applications resulting in the security feature
1879 being disabled. The pointer guard is now correctly initialized to a
1880 random value for static applications. Existing static applications need
1881 to be recompiled to take advantage of the fix (bug 15754).
1882
1883 * CVE-2013-4237 The readdir_r function could write more than NAME_MAX bytes
1884 to the d_name member of struct dirent, or omit the terminating NUL
1885 character. (Bugzilla #14699).
1886
1887 * CVE-2013-4332 The pvalloc, valloc, memalign, posix_memalign and
1888 aligned_alloc functions could allocate too few bytes or corrupt the
1889 heap when passed very large allocation size values (Bugzilla #15855,
1890 #15856, #15857).
1891
1892 * CVE-2013-4458 Stack overflow in getaddrinfo with large number of results
1893 for AF_INET6 has been fixed (Bugzilla #16072).
1894
1895 * New locales: ak_GH, anp_IN, ar_SS, cmn_TW, hak_TW, lzh_TW, nan_TW, pap_AW,
1896 pap_CW, quz_PE, the_NP.
1897
1898 * Substantially revised locales: gd_GB, ht_HT
1899
1900 * The LC_ADDRESS field was updated to support country_car for almost all
1901 supported locales.
1902
1903 * ISO 1427 definitions were updated.
1904
1905 * ISO 3166 definitions were updated.
1906
1907 * The localedef utility now supports --big-endian and --little-endian
1908 command-line options to generate locales for a different system from that
1909 for which the C library was built.
1910
1911 * Binary locale files now only depend on the endianness of the system for
1912 which they are generated and not on other properties of that system. As a
1913 consequence, binary files generated with new localedef may be incompatible
1914 with old versions of the GNU C Library, and binary files generated with
1915 old localedef may be incompatible with this version of the GNU C Library,
1916 in the following circumstances:
1917
1918 + Locale files may be incompatible on m68k systems.
1919
1920 + Locale archive files (but not separate files for individual locales) may
1921 be incompatible on systems where plain "char" is signed.
1922
1923 * The configure option --disable-versioning has been removed. Builds with
1924 --disable-versioning had not worked for several years.
1925
1926 * ISO 639 definitions were updated for Chiga (cgg) and Chinese (gan, hak, czh,
1927 cjy, lzh, cmn, mnp, cdo, czo, cpx, wuu, hsn, yue).
1928
1929 * SystemTap probes for malloc have been introduced.
1930
1931 * SystemTap probes for slow multiple precision fallback paths of
1932 transcendental functions have been introduced.
1933
1934 * Support for powerpc64le has been added.
1935
1936 * The soft-float powerpc port now supports e500 processors.
1937
1938 * Support for STT_GNU_IFUNC symbols added for ppc32/power4+ and ppc64.
1939
1940 * A new feature test macro _DEFAULT_SOURCE is available to enable the same
1941 set of header declarations that are enabled by default, even when other
1942 feature test macros or compiler options such as -std=c99 would otherwise
1943 disable some of those declarations.
1944
1945 * The _BSD_SOURCE feature test macro no longer enables BSD interfaces that
1946 conflict with POSIX. The libbsd-compat library (which was a dummy library
1947 that did nothing) has also been removed.
1948
1949 * Preliminary documentation about Multi-Thread, Async-Signal and
1950 Async-Cancel Safety has been added.
1951
1952 * Change of ABI data structures for s390 and s390x:
1953 On s390 and s390x the size of struct ucontext and jmp_buf was increased to
1954 allow for future hardware extensions. All library functions that accept or
1955 return these structures were versioned in order to provide backwards
1956 compatibility with existing code. However, not all cases can be handled
1957 with interface versioning. If an application embeds either structure into
1958 another structure and that structure is passed to another compilation unit
1959 which expects the newer larger-sized object then any access to the new
1960 fields will result in undefined behaviour. Similarly any access to fields
1961 that were moved by the enlarging of the structures will cause undefined
1962 behaviour. To fix the undefined behaviour all objects that are part of the
1963 final in-memory image must agree on the size of structures, and this may
1964 require recompilation.
1965 \f
1966 Version 2.18
1967
1968 * The following bugs are resolved with this release:
1969
1970 2546, 2560, 5159, 6809, 7006, 10060, 10062, 10283, 10357, 10686, 11120,
1971 11561, 12310, 12387, 12492, 12515, 12723, 13550, 13889, 13951, 13988,
1972 14142, 14176, 14200, 14256, 14280, 14293, 14317, 14327, 14478, 14496,
1973 14582, 14686, 14812, 14888, 14894, 14907, 14908, 14909, 14920, 14941,
1974 14952, 14964, 14981, 14982, 14985, 14991, 14994, 14996, 15000, 15003,
1975 15006, 15007, 15014, 15020, 15022, 15023, 15036, 15054, 15055, 15062,
1976 15078, 15084, 15085, 15086, 15100, 15160, 15214, 15221, 15232, 15234,
1977 15283, 15285, 15287, 15304, 15305, 15307, 15309, 15327, 15330, 15335,
1978 15336, 15337, 15339, 15342, 15346, 15359, 15361, 15366, 15380, 15381,
1979 15394, 15395, 15405, 15406, 15409, 15416, 15418, 15419, 15423, 15424,
1980 15426, 15429, 15431, 15432, 15441, 15442, 15448, 15465, 15480, 15485,
1981 15488, 15490, 15492, 15493, 15497, 15506, 15529, 15536, 15553, 15577,
1982 15583, 15618, 15627, 15631, 15654, 15655, 15666, 15667, 15674, 15711,
1983 15755, 15759.
1984
1985 * CVE-2013-2207 Incorrectly granting access to another user's pseudo-terminal
1986 has been fixed by disabling the use of pt_chown (Bugzilla #15755).
1987 Distributions can re-enable building and using pt_chown via the new configure
1988 option `--enable-pt_chown'. Enabling the use of pt_chown carries with it
1989 considerable security risks and should only be used if the distribution
1990 understands and accepts the risks.
1991
1992 * CVE-2013-0242 Buffer overrun in regexp matcher has been fixed (Bugzilla
1993 #15078).
1994
1995 * CVE-2013-1914 Stack overflow in getaddrinfo with many results has been
1996 fixed (Bugzilla #15330).
1997
1998 * Add support for calling C++11 thread_local object destructors on thread
1999 and program exit. This needs compiler support for offloading C++11
2000 destructor calls to glibc.
2001
2002 * Improved worst case performance of libm functions with double inputs and
2003 output.
2004
2005 * Support for priority inherited mutexes in pthread condition variables on
2006 non-x86 architectures.
2007
2008 * Port to Xilinx MicroBlaze contributed by David Holsgrove.
2009
2010 * Optimized string functions for AArch64. Implemented by Marcus Shawcroft.
2011
2012 * Optimized string functions for ARM. Implemented by Will Newton and
2013 Richard Henderson.
2014
2015 * Optimized libm functions for SPARC. Implemented by David S. Miller.
2016
2017 * Support for building more of ARM glibc as Thumb-2 code. Implemented by
2018 Richard Henderson.
2019
2020 * Support for building most of MIPS glibc as MIPS16 code. Implemented by
2021 Chung-Lin Tang, Maciej W. Rozycki and Maxim Kuvyrkov.
2022
2023 * Added a benchmark framework to track performance of functions in glibc.
2024
2025 * New <math.h> macro named issignaling to check for a signaling NaN (sNaN).
2026 It is based on draft TS 18661 and currently enabled as a GNU extension.
2027
2028 * On Linux, the clock function now uses the clock_gettime system call
2029 for improved precision, rather than old times system call.
2030
2031 * Added support for version-3 tz format files. This is needed when using
2032 the tz database release 2013e or later, and affects a few unusual cases --
2033 currently only TZ='America/Godthab' for time stamps after 2037.
2034
2035 * Added new API functions pthread_getattr_default_np and
2036 pthread_setattr_default_np to get and set the default pthread
2037 attributes of a process.
2038
2039 * Added support for TSX lock elision for pthread mutexes on i386 and x86-64.
2040 This may improve lock scaling of existing programs on TSX capable systems.
2041 When the --enable-lock-elision=yes parameter is specified at configure
2042 time lock elision will be enabled by default for all PTHREAD_MUTEX_DEFAULT
2043 mutexes.
2044
2045 * Support added for AT_HWCAP2 (to coincide with Linux kernel 3.10
2046 availability). Implemented by Ryan S. Arnold.
2047
2048 * Support added for POWER8 platform. Implemented by Ryan S. Arnold.
2049 \f
2050 Version 2.17
2051
2052 * The following bugs are resolved with this release:
2053
2054 1349, 3439, 3479, 3665, 5044, 5246, 5298, 5400, 6530, 6677, 6778, 6808,
2055 9685, 9914, 10014, 10038, 10114, 10191, 10631, 10873, 11438, 11607, 11638,
2056 11741, 12140, 13013, 13412, 13542, 13601, 13603, 13604, 13629, 13679,
2057 13696, 13698, 13717, 13741, 13759, 13761, 13763, 13881, 13939, 13950,
2058 13952, 13966, 14042, 14047, 14090, 14150, 14151, 14152, 14154, 14157,
2059 14166, 14173, 14195, 14197, 14237, 14246, 14251, 14252, 14283, 14298,
2060 14303, 14307, 14328, 14331, 14336, 14337, 14347, 14349, 14368, 14376,
2061 14417, 14447, 14459, 14476, 14477, 14501, 14505, 14510, 14516, 14518,
2062 14519, 14530, 14532, 14538, 14543, 14544, 14545, 14557, 14562, 14568,
2063 14576, 14579, 14583, 14587, 14595, 14602, 14610, 14621, 14638, 14645,
2064 14648, 14652, 14660, 14661, 14669, 14672, 14683, 14694, 14716, 14719,
2065 14743, 14767, 14783, 14784, 14785, 14793, 14796, 14797, 14801, 14803,
2066 14805, 14807, 14811, 14815, 14821, 14822, 14824, 14828, 14831, 14833,
2067 14835, 14838, 14856, 14863, 14865, 14866, 14868, 14869, 14871, 14872,
2068 14879, 14889, 14893, 14898, 14914.
2069
2070 * Optimization of memcpy for MIPS.
2071
2072 * CVE-2011-4609 svc_run() produces high cpu usage when accept fails with
2073 EMFILE has been fixed (Bugzilla #14889).
2074
2075 * The add-on ports collection is now distributed in the "ports" subdirectory
2076 of the main GNU C Library distribution, rather than separately.
2077
2078 * Port to ARM AArch64 contributed by Linaro.
2079
2080 * Support for STT_GNU_IFUNC symbols added for s390 and s390x.
2081 Optimized versions of memcpy, memset, and memcmp added for System z10 and
2082 zEnterprise z196.
2083 Implemented by Andreas Krebbel.
2084
2085 * The new function secure_getenv allows secure access to the environment,
2086 returning NULL if running in a SUID/SGID process. This function replaces
2087 the internal function __secure_getenv.
2088
2089 * SystemTap static probes have been added into the dynamic linker.
2090 Implemented by Gary Benson.
2091
2092 * Optimizations of string functions strstr, strcasestr and memmem.
2093 Implemented by Maxim Kuvyrkov.
2094
2095 * The minimum Linux kernel version that this version of the GNU C Library
2096 can be used with is 2.6.16.
2097
2098 * Optimizations of string functions memchr, wcschr, wcscpy, and wcsrchr for
2099 powerpc POWER7. Implemented by Will Schmidt.
2100
2101 * New configure option --disable-nscd builds the C library such that it
2102 never attempts to contact the Name Service Caching Daemon (nscd).
2103 New configure option --disable-build-nscd avoids building nscd itself;
2104 this is the default if --disable-nscd is used.
2105
2106 * Improved support for cross-compilation, including cross-testing and
2107 bootstrap builds without a previously built glibc.
2108
2109 * Several testsuite tests are now able to test multiple IFUNC variants of an
2110 interface, rather than just testing the one that would be chooen by
2111 default.
2112
2113 * New configure options --with-bugurl and --with-pkgversion, for
2114 distributors to use to embed their bug-reporting and package version
2115 information in --help and --version output.
2116
2117 * The ttyname and ttyname_r functions on Linux now fall back to searching for
2118 the tty file descriptor in /dev/pts or /dev if /proc is not available. This
2119 allows creation of chroots without the procfs mounted on /proc.
2120
2121 * The `crypt' function now fails if passed salt bytes that violate the
2122 specification for those values. On Linux, the `crypt' function will
2123 consult /proc/sys/crypto/fips_enabled to determine if "FIPS mode" is
2124 enabled, and fail on encrypted strings using the MD5 or DES algorithm
2125 when the mode is enabled.
2126
2127 * The `clock_*' suite of functions (declared in <time.h>) is now available
2128 directly in the main C library. Previously it was necessary to link with
2129 -lrt to use these functions. This change has the effect that a
2130 single-threaded program that uses a function such as `clock_gettime' (and
2131 is not linked with -lrt) will no longer implicitly load the pthreads
2132 library at runtime and so will not suffer the overheads associated with
2133 multi-thread support in other code such as the C++ runtime library.
2134
2135 * New locales: ayc_PE, doi_IN, ia_FR, mni_IN, nhn_MX, niu_NU, niu_NZ,
2136 sat_IN, and szl_PL.
2137 \f
2138 Version 2.16
2139
2140 * The following bugs are resolved with this release:
2141
2142 174, 208, 350, 369, 411, 706, 766, 2074, 2541, 2547, 2548, 2550, 2551,
2143 2552, 2553, 2554, 2562, 2563, 2565, 2566, 2570, 2576, 2636, 2678, 3335,
2144 3440, 3748, 3768, 3866, 3868, 3906, 3976, 3992, 4026, 4108, 4596, 4822,
2145 5077, 5461, 5805, 5993, 6471, 6486, 6578, 6649, 6730, 6770, 6794, 6884,
2146 6890, 6894, 6895, 6907, 6911, 6959, 7064, 9739, 9902, 10110, 10135, 10140,
2147 10153, 10210, 10254, 10346, 10375, 10545, 10716, 10846, 10882, 11174,
2148 11261, 11322, 11365, 11451, 11494, 11521, 11677, 11837, 11959, 12047,
2149 12097, 12193, 12194, 12297, 12298, 12301, 12340, 12354, 12416, 12495,
2150 13058, 13223, 13361, 13525, 13526, 13527, 13528, 13529, 13530, 13531,
2151 13532, 13533, 13547, 13551, 13552, 13553, 13555, 13556, 13559, 13563,
2152 13566, 13576, 13579, 13583, 13592, 13594, 13613, 13618, 13637, 13656,
2153 13658, 13673, 13691, 13695, 13704, 13705, 13706, 13718, 13726, 13738,
2154 13739, 13743, 13750, 13758, 13760, 13761, 13775, 13786, 13787, 13792,
2155 13806, 13824, 13840, 13841, 13844, 13846, 13848, 13851, 13852, 13854,
2156 13871, 13872, 13873, 13879, 13882, 13883, 13884, 13885, 13886, 13892,
2157 13895, 13908, 13910, 13911, 13912, 13913, 13914, 13915, 13916, 13917,
2158 13918, 13919, 13920, 13921, 13922, 13923, 13924, 13926, 13927, 13928,
2159 13938, 13941, 13942, 13954, 13955, 13956, 13963, 13967, 13968, 13970,
2160 13973, 13979, 13983, 13986, 13996, 14012, 14027, 14033, 14034, 14036,
2161 14040, 14043, 14044, 14048, 14049, 14050, 14053, 14055, 14059, 14064,
2162 14075, 14080, 14083, 14103, 14104, 14109, 14112, 14117, 14122, 14123,
2163 14134, 14153, 14183, 14188, 14199, 14210, 14218, 14229, 14241, 14273,
2164 14277, 14278.
2165
2166 * Support for the x32 ABI on x86-64 added. The x32 target is selected by
2167 configuring glibc with:
2168 BUILD_CC='gcc' CC='gcc -mx32' CXX='g++ -mx32'
2169 Visit <http://sites.google.com/site/x32abi/> for more x32 ABI info.
2170 Implemented by H.J. Lu.
2171
2172 * ISO C11 support:
2173
2174 + define static_assert
2175
2176 + do not declare gets
2177
2178 + declare at_quick_exit and quick_exit also for ISO C11
2179
2180 + aligned_alloc. NB: The code is deliberately allows the size parameter
2181 to not be a multiple of the alignment. This is a moronic requirement
2182 in the standard but it is only a requirement on the caller, not the
2183 implementation.
2184
2185 + timespec_get added
2186
2187 + uchar.h support added
2188
2189 + CMPLX, CMPLXF, CMPLXL added
2190
2191 Implemented by Ulrich Drepper.
2192
2193 * Support for the IA-64 has been moved to ports.
2194
2195 * Remove support for anything but ELF binary format
2196
2197 * Checking versions of poll, ppoll added.
2198 Implemented by Ulrich Drepper.
2199
2200 * More generic and 64-bit performance optimizations to math functions.
2201 Implemented by Ulrich Drepper.
2202
2203 * New configure option --enable-obsolete-rpc makes the deprecated RPC
2204 headers and functions available at compile time as they were before
2205 version 2.14. This option will be removed at some time in the future
2206 after the TI-RPC library becomes fully sufficient for the needs of
2207 existing applications.
2208
2209 * Compatibility code for Linux kernel versions before 2.4 has been removed.
2210 Note that glibc is not expected to work with any Linux kernel version
2211 before 2.6.
2212
2213 * New header <sys/auxv.h> and function getauxval allowing easy access to
2214 the AT_* key-value pairs passed from the Linux kernel. The header also
2215 defines the HWCAP_* bits associated with the AT_HWCAP key.
2216
2217 * New locales: mag_IN
2218
2219 * New configure option --enable-systemtap builds SystemTap static probes
2220 into libc for setjmp and longjmp and into libpthread for various operations.
2221 So far the setjmp/longjmp probes and some of the libpthread probes are
2222 provided only for i*86 and x86_64.
2223 Implemented by Roland McGrath and Rayson Ho.
2224
2225 * Optimized expf for x86-32 and x86-64. Implemented by Liubov Dmitrieva.
2226
2227 * More optimized functions for PowerPC. Implemented by Adhemerval Zanella
2228 and Will Schmidt.
2229
2230 * More optimized functions for SPARC. Implemented by David S. Miller.
2231
2232 * Improved support for cross-compilation, especially bootstrap builds
2233 without a previously built glibc.
2234
2235 * Ports for the TILE-Gx and TILEPro families of processors. Contributed by
2236 Chris Metcalf from Tilera.
2237
2238 * Support for the old ARM ABI has been removed from ports. Only the EABI is
2239 now supported for ARM processors.
2240
2241 * The hard-float variant of the ARM EABI now uses /lib/ld-linux-armhf.so.3
2242 as the name of the dynamic linker, to distinguish it from the
2243 /lib/ld-linux.so.3 dynamic linker used for the base EABI.
2244
2245 * Support for CRIS has been removed from ports.
2246
2247 * A new class of installed header has been documented for low-level
2248 platform-specific functionality. PowerPC added the first instance with a
2249 function to provide time base register access. Contributed by Tulio
2250 Magno Quites Machado Filho.
2251
2252 * ix86 configurations now install header files that are consistent with
2253 what x86-64 configurations install. These same header files can be used
2254 for -m32, -m64, or -mx32 builds.
2255 Contributed by H.J. Lu.
2256
2257 * Math library bug fixes. A thorough audit of all open math library bugs was
2258 conducted by Joseph Myers. Significant progress was made on many math
2259 library bugs resulting in more accurate exceptions and function results.
2260 Many thanks to all those that contributed including Andreas Jaeger for his
2261 patch review and work on the x87 trigonometric instruction issues.
2262
2263 * Timezone data is no longer installed. Timezone-related binaries and scripts
2264 will continue to be installed. Users should obtain their timezone data from
2265 their distribution provider or from the tzdata package at
2266 <ftp://munnari.oz.au/pub/>.
2267 \f
2268 Version 2.15
2269
2270 * The following bugs are resolved with this release:
2271
2272 6779, 6783, 9696, 10103, 10709, 11589, 11929, 12403, 12786, 12840, 12847,
2273 12868, 12852, 12874, 12885, 12892, 12906, 12907, 12922, 12935, 12962,
2274 13007, 13021, 13061, 13062, 13067, 13068, 13085, 13088, 13090, 13092,
2275 13096, 13114, 13118, 13123, 13134, 13138, 13147, 13150, 13166, 13179,
2276 13185, 13189, 13192, 13268, 13276, 13282, 13291, 13305, 13328, 13335,
2277 13337, 13344, 13358, 13367, 13413, 13416, 13423, 13439, 13446, 13472,
2278 13484, 13506, 13515, 13523, 13524, 13538, 13540
2279
2280 * New program pldd to list loaded object of a process
2281 Implemented by Ulrich Drepper.
2282
2283 * Add nss_db support back to glibc. No more dependency on Berkeley db
2284 and support for initgroups lookups.
2285 Implemented by Ulrich Drepper.
2286
2287 * Optimized strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-32.
2288 Contributed by HJ Lu.
2289
2290 * Improved strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-64.
2291 Contributed by HJ Lu.
2292
2293 * Optimized strcat, strncat on x86-64 and optimized wcscmp, wcslen, strnlen
2294 on x86-32 and x86-64.
2295 Contributed by Liubov Dmitrieva.
2296
2297 * Optimized strchr and strrchr for SSE on x86-32.
2298 Contributed by Liubov Dmitrieva.
2299
2300 * Optimized memchr, memrchr, rawmemchr, memcmp, wmemcmp, wcschr, wcscpy
2301 for x86-64 and x86-32.
2302 Contributed by Liubov Dmitrieva.
2303
2304 * New interfaces: scandirat, scandirat64
2305 Implemented by Ulrich Drepper.
2306
2307 * Checking versions of FD_SET, FD_CLR, and FD_ISSET added.
2308 Implemented by Ulrich Drepper.
2309
2310 * nscd now also caches the netgroup database.
2311 Implemented by Ulrich Drepper.
2312
2313 * Integrate libm with gcc's -ffinite-math-only option.
2314 Implemented by Ulrich Drepper.
2315
2316 * Lots of generic, 64-bit, and x86-64-specific performance optimizations
2317 to math functions. Implemented by Ulrich Drepper.
2318
2319 * Optimized strcasecmp and strncasecmp for AVX on x86-64.
2320 Implemented by Ulrich Drepper.
2321
2322 * New Linux interfaces: process_vm_readv, process_vm_writev
2323
2324 * Optimized strcasecmp and strncasecmp for SSSE3 and SSE4.2 on x86-32.
2325 Implemented by Ulrich Drepper.
2326
2327 * Optimized nearbyint and strcasecmp for PPC.
2328 Implemented by Adhemerval Zanella.
2329
2330 * New locales: bho_IN, unm_US, es_CU, ta_LK
2331 \f
2332 Version 2.14
2333
2334 * The following bugs are resolved with this release:
2335
2336 386, 6420, 7101, 9730, 9732, 9809, 10138, 10149, 10157, 11099, 11257,
2337 11258, 11487, 11532, 11558, 11578, 11634, 11653, 11668, 11697, 11724,
2338 11781, 11799, 11820, 11837, 11857, 11884, 11892, 11895, 11901, 11945,
2339 11947, 11952, 11987, 12052, 12083, 12158, 12178, 12200, 12346, 12350,
2340 12393, 12420, 12432, 12445, 12449, 12453, 12454, 12460, 12469, 12489,
2341 12509, 12510, 12511, 12518, 12527, 12541, 12545, 12551, 12582, 12583,
2342 12587, 12597, 12601, 12611, 12625, 12626, 12631, 12650, 12653, 12655,
2343 12660, 12671, 12681, 12684, 12685, 12711, 12713, 12714, 12717, 12723,
2344 12724, 12734, 12738, 12746, 12766, 12775, 12777, 12782, 12788, 12792,
2345 12795, 12811, 12813, 12814, 12841
2346
2347 * The RPC implementation in libc is obsoleted. Old programs keep working
2348 but new programs cannot be linked with the routines in libc anymore.
2349 Programs in need of RPC functionality must be linked against TI-RPC.
2350 The TI-RPC implementation is IPv6 enabled and there are other benefits.
2351
2352 Visible changes of this change include (obviously) the inability to link
2353 programs using RPC functions without referencing the TI-RPC library and the
2354 removal of the RPC headers from the glibc headers.
2355 Implemented by Ulrich Drepper.
2356
2357 * New Linux interfaces: clock_adjtime, name_to_handle_at, open_by_handle_at,
2358 syncfs, setns, sendmmsg
2359
2360 * New locales: os_RU, bem_ZA, en_ZA, ff_SN, sw_KE, sw_TZ, lb_LU, wae_CH,
2361 yue_HK, lij_IT, mhr_RU
2362
2363 * New encodings: CP770, CP771, CP772, CP773, CP774
2364
2365 * New program sotruss to trace calls through PLTs
2366 Implemented by Ulrich Drepper.
2367
2368 * The malloc hook implementation is marked deprecated and will be removed
2369 from the default implementation in the next version. The design never
2370 worked ever since the introduction of threads. Even programs which do
2371 not create threads themselves can use multiple threads created internally.
2372 \f
2373 Version 2.13
2374
2375 * The following bugs are resolved with this release:
2376
2377 3268, 6812, 7066, 10085, 10484, 10563, 10851, 11149, 11155, 11577, 11611,
2378 11640, 11655, 11701, 11840, 11856, 11883, 11903, 11904, 11968, 11979,
2379 12005, 12037, 12067, 12077, 12078, 12092, 12093, 12107, 12108, 12113,
2380 12140, 12159, 12167, 12191, 12194, 12201, 12204, 12205, 12207, 12348,
2381 12378, 12394, 12397
2382
2383 * New Linux interfaces: prlimit, prlimit64, fanotify_init, fanotify_mark
2384
2385 * POWER7 optimizations: memset, memcmp, strncmp
2386
2387 * New optimized string functions for x86-64: strnlen (SSE2),
2388 strcasecmp (SSE2, SSSE3, SSE4.2), strncasecmp (SSE2, SSSE3, SSE4.2)
2389 Implemented by Ulrich Drepper.
2390 \f
2391 Version 2.12
2392
2393 * The following bugs are resolved with this release:
2394
2395 3662, 4457, 5553, 10162, 10401, 10414, 10554, 10824, 10864, 10915,
2396 10918, 10936, 10939, 10958, 10968, 10969, 10972, 10992, 11000, 11001,
2397 11007, 11010, 11027, 11039, 11040, 11041, 11043, 11046, 11056, 11070,
2398 11093, 11115, 11125, 11126, 11127, 11134, 11141, 11149, 11183, 11184,
2399 11185, 11186, 11187, 11188, 11189, 11190, 11191, 11192, 11193, 11194,
2400 11200, 11230, 11235, 11242, 11254, 11258, 11271, 11272, 11276, 11279,
2401 11287, 11292, 11319, 11332, 11333, 11387, 11389, 11390, 11394, 11397,
2402 11410, 11438, 11449, 11470, 11471, 11520, 11537, 11538, 11571
2403
2404 * New interfaces: pthread_getname_np, pthread_setname_np
2405
2406 * New Linux interface: recvmmsg
2407
2408 * STT_GNU_IFUNC implemented for Sparc by David Miller.
2409
2410 * The dynamic linker now recognizes supported ABI versions from the
2411 EI_ABIVERSION field in the ELF header.
2412 Implemented by Ulrich Drepper.
2413
2414 * New NIS mode selector ADJUNCT_AS_SHADOW. The passwd.adjunct.byname table
2415 will not be used to fill in password fields in the passwd.byname replies.
2416 Instead it is used to synthesize the shadow.byname table, should it be
2417 missing. This is a useful mode in some installations involving Solaris.
2418 Implemented by Ulrich Drepper.
2419
2420 * New locales: kok_IN, sq_MK, cv_RU
2421 \f
2422 Version 2.11
2423
2424 * The following bugs are resolved with this release:
2425
2426 3270, 7094, 9924, 9986, 10011, 10085, 10107, 10159, 10162, 10166, 10183,
2427 10195, 10196, 10200, 10203, 10207, 10211, 10213, 10217, 10221, 10229,
2428 10262, 10286, 10312, 10315, 10319, 10349, 10360, 10391, 10402, 10416,
2429 10418, 10422, 10425, 10446, 10448, 10452, 10540, 10553, 10560, 10564,
2430 10609, 10643, 10692, 10713, 10717, 10730, 10731, 10742, 10780, 10784,
2431 10789, 10817, 10823, 10840, 10847
2432
2433 * New interfaces: execvpe, pthread_sigqueue, mkstemps, mkstemps64, mkostemps,
2434 mkostemps64
2435 Implemented by Ulrich Drepper.
2436
2437 * Checking version of longjmp added that fails if an uninitialized stack
2438 frame would be created. Implemented by Ulrich Drepper.
2439
2440 * STT_GNU_IFUNC is now supported in static executables.
2441 Implemented by H.J. Lu.
2442
2443 * STT_GNU_IFUNC implemented for PPC by Alan Modra.
2444
2445 * New optimized string functions for x86-64: strstr, strcasestr, memcmp,
2446 strcspn, strpbrk, strspn, strcpy, stpcpy, strncpy, strcmp (SSE2, SSE4.2),
2447 strncmp (SSE2, SSE4.2), strchr (SSE4.2), strrchr (SSE4.2).
2448 Contributed by H.J. Lu.
2449
2450 strlen, rawmemchr, strcmp (SSSE3), strncmp (SSSE3).
2451 Implemented by Ulrich Drepper.
2452
2453 * New optimized string functions for x86: strlen, strcspn, strspn, strpbrk,
2454 strstr, strcasestr.
2455 Contributed by H.J. Lu.
2456
2457 * Support for fma instruction in AVX on x86-64.
2458 Implemented by H.J. Lu and Ulrich Drepper.
2459
2460 * AVX support in x86-64 auditing support in ld.so.
2461 Implemented by H.J. Lu.
2462
2463 * STB_GNU_UNIQUE support added.
2464 Implemented by Ulrich Drepper.
2465
2466 * Implement second fallback mode for optimized DNS lookup for even more
2467 broken environments. If two requests from the same port are not
2468 handled correctly close the socket and open a new one before sending
2469 the second request. The 'single-request-reopen' option in /etc/resolv.conf
2470 can be used to select this mode right away, instead of rediscovering the
2471 necessity is every process again.
2472 Implemented by Ulrich Drepper.
2473
2474 * New resolver flag RES_USE_DNSSEC to enable use of verified lookup.
2475 Implemented by Adam Tkac.
2476
2477 * Optimized iconv conversions for S390x.
2478 Implemented by Andreas Krebbel.
2479
2480 * Using condvars with PI mutexes is now more efficient due to kernel
2481 support for requeueing to PI futexes. NPTL support added for x86-64.
2482 Implemented by Ulrich Drepper.
2483
2484 * New locales: ps_AF, my_MM
2485 \f
2486 Version 2.10
2487
2488 * The following bugs are resolved with this release:
2489
2490 697, 5381, 5807, 6411, 6545, 7040, 7058, 7067, 7080, 7083, 7095, 7098,
2491 9677, 9694, 9697, 9701, 9704, 9705, 9706, 9720, 9726, 9730, 9733, 9736,
2492 9741, 9750, 9753, 9759, 9781, 9791, 9793, 9823, 9833, 9844, 9880, 9881,
2493 9891, 9893, 9895, 9913, 9916, 9920, 9948, 9955, 9957, 9963, 9985, 10052,
2494 10069, 10086, 10087, 10090, 10092, 10093, 10100, 10118, 10128
2495
2496 * New interfaces: preadv, preadv64, pwritev, pwritev64, malloc_info
2497 Implemented by Ulrich Drepper.
2498
2499 * New Linux interfaces: accept4, fallocate, fallocate64.
2500 Implemented by Ulrich Drepper.
2501
2502 * Correct declarations of string function when used in C++ code. This
2503 could lead to compile errors for invalid C++ code.
2504
2505 * XPG7/POSIX 2008 compilation environment. Many GNU-specific functions are
2506 now in POSIX.
2507
2508 * New POSIX 2008 interface: psiginfo
2509 Implemented by Ulrich Drepper.
2510
2511 * New ISO C++1x interfaces: quick_exit, at_quick_exit
2512 Implemented by Ulrich Drepper.
2513
2514 * Support for selecting between multiple function definitions at runtime
2515 using STT_GNU_IFUNC symbols. Implemented by Ulrich Drepper.
2516
2517 * The libcrypt library can now use the hash function implementations in
2518 NSS. Implemented by Ulrich Drepper.
2519
2520 * The malloc implementation can be compiled to be less memory efficient
2521 but higher performing in multi-threaded programs.
2522 Implemented by Ulrich Drepper.
2523
2524 * New locales: nan_TW@latin, ks_IN
2525
2526 * Faster strlen, strchr, strchrnul, strrchr, memchr, and rawmemchr for x86-64.
2527 Implemented by Ulrich Drepper.
2528
2529 * Extended printf hook support. It is possible to use user-defined types
2530 and extend existing format specifiers.
2531 Implemented by Ulrich Drepper.
2532
2533 * Handling for group shadow files has been added.
2534 Implemented by Ulrich Drepper.
2535
2536 * DNS IPv4-IPv6 parallel lookup now deals better with broken DNS
2537 servers (the case, e.g., for some people using the built-in DNS
2538 server in ADSL modems/routers). There is a once-per-process timeout
2539 in case of a broken server. To avoid it, users can run nscd or put
2540 'options single-request' in /etc/resolv.conf.
2541 Implemented by Ulrich Drepper.
2542 \f
2543 Version 2.9
2544
2545 * The following bugs are resolved with this release:
2546
2547 3406, 5209, 5210, 5381, 5794, 5814, 5911, 6428, 6438, 6442, 6447, 6461,
2548 6472, 6505, 6544, 6589, 6612, 6634, 6653, 6654, 6657, 6698, 6712, 6713,
2549 6719, 6723, 6724, 6740, 6763, 6771, 6790, 6791, 6817, 6824, 6839, 6843,
2550 6845, 6856, 6857, 6860, 6867, 6875, 6919, 6920, 6942, 6947, 6955, 6966,
2551 6968, 6974, 6980, 6995, 7008, 7009, 7029
2552
2553 * Unified lookup for getaddrinfo: IPv4 and IPv6 addresses are now looked
2554 up at the same time. Implemented by Ulrich Drepper.
2555
2556 * TLS descriptors for LD and GD on x86 and x86-64.
2557 Implemented by Alexandre Oliva.
2558
2559 * getaddrinfo now handles DCCP and UDPlite.
2560 Implemented by Ulrich Drepper.
2561
2562 * New fixed-size conversion macros: htobe16, htole16, be16toh, le16toh,
2563 htobe32, htole32, be32toh, le32toh, htobe64, htole64, be64toh, le64toh.
2564 Implemented by Ulrich Drepper.
2565
2566 * New implementation of memmem, strstr, and strcasestr which is O(n).
2567 Implemented by Eric Blake.
2568
2569 * New Linux interfaces: inotify_init1, dup3, epoll_create1, pipe2
2570
2571 * Implement "e" option for popen to open file descriptor with the
2572 close-on-exec flag set. Implemented by Ulrich Drepper.
2573
2574 * Implement "b" mode for fmemopen. In this mode writes writes don't
2575 implicitly add a NUL byte and seeks from the end of the buffer really
2576 use the buffer end, not the string length as the basis.
2577 Implemented by Ulrich Drepper.
2578
2579 * Many functions, exported and internal, now atomically set the close-on-exec
2580 flag when run on a sufficiently new kernel. Implemented by Ulrich Drepper.
2581
2582 * Sorting rules for some Indian languages (Kannada, Gurumukhi, Telugu,
2583 Sinhala)
2584 Implemented by Pravin Satpute.
2585
2586 * New locales: sd_IN, sd_IN@devanagari, ks_IN@devanagari, ht_HT, en_AG, nl_AW.
2587 \f
2588 Version 2.8
2589
2590 * The following bugs are resolved with this release:
2591
2592 2549, 3112, 4314, 4407, 4997, 5012, 5040, 5112, 5204, 5208, 5209, 5220,
2593 5222, 5224, 5225, 5237, 5238, 5240, 5245, 5277, 5346, 5375, 5378, 5382,
2594 5424, 5427, 5428, 5435, 5436, 5439, 5441, 5442, 5443, 5451, 5452, 5454,
2595 5455, 5463, 5464, 5465, 5475, 5477, 5478, 5520, 5541, 5545, 5553, 5600,
2596 5602, 5607, 5614, 5627, 5628, 5736, 5737, 5741, 5753, 5760, 5762, 5768,
2597 5774, 5777, 5778, 5779, 5786, 5790, 5818, 5854, 5857, 5903, 5939, 5979,
2598 5983, 5995, 5998, 6004, 6007, 6020, 6021, 6024, 6040, 6042
2599
2600 * New locales: bo_CN, bo_IN, shs_CA.
2601
2602 * New encoding: HP-ROMAN9, HP-GREEK8, HP-THAI8, HP-TURKISH8.
2603
2604 * Sorting rules for some Indian languages (Devanagari and Gujarati).
2605 Implemented by Pravin Satpute.
2606
2607 * IPV6 addresses in /etc/resolv.conf can now have a scope ID
2608
2609 * nscd caches now all timeouts for DNS entries
2610 Implemented by Ulrich Drepper.
2611
2612 * nscd is more efficient and wakes up less often.
2613 Implemented by Ulrich Drepper.
2614
2615 * More checking functions: asprintf, dprintf, obstack_printf, vasprintf,
2616 vdprintf, and obstack_vprintf.
2617 Implemented by Jakub Jelinek.
2618
2619 * Faster memset for x86-64.
2620 Implemented by Harsha Jagasia and H.J. Lu.
2621
2622 * Faster memcpy on x86.
2623 Implemented by Ulrich Drepper.
2624
2625 * ARG_MAX is not anymore constant on Linux. Use sysconf(_SC_ARG_MAX).
2626 Implemented by Ulrich Drepper.
2627
2628 * Faster sqrt and sqrtf implemention for some PPC variants.
2629 Implemented by Steven Munroe.
2630 \f
2631 Version 2.7
2632
2633 * The following bugs are resolved with this release:
2634
2635 4125, 4126, 4178, 4359, 4407, 4512, 4514, 4525, 4554, 4556, 4557, 4566,
2636 4582, 4586, 4588, 4599, 4610, 4647, 4702, 4705, 4726, 4745, 4772, 4773,
2637 4775, 4776, 4792, 4813, 4814, 4816, 4833, 4858, 4860, 4896, 4905, 4925,
2638 4936, 4937, 4938, 4941, 4946, 4963, 4972, 5010, 5028, 5043, 5058, 5063,
2639 5071, 5103, 5104, 5112, 5113, 5184, 5186
2640
2641 * More checking functions: fread, fread_unlocked, open*, mq_open.
2642 Implemented by Jakub Jelinek and Ulrich Drepper.
2643
2644 * Extend fortification to C++. Implemented by Jakub Jelinek.
2645
2646 * Implement 'm' modifier for scanf. Add stricter C99/SUS compliance
2647 by not recognizing 'a' as a modifier when those specs are requested.
2648 Implemented by Jakub Jelinek.
2649
2650 * PPC optimizations to math and string functions.
2651 Implemented by Steven Munroe.
2652
2653 * New interfaces: mkostemp, mkostemp64. Like mkstemp* but allow additional
2654 options to be passed. Implemented by Ulrich Drepper.
2655
2656 * More CPU set manipulation functions. Implemented by Ulrich Drepper.
2657
2658 * New Linux interfaces: signalfd, eventfd, eventfd_read, and eventfd_write.
2659 Implemented by Ulrich Drepper.
2660
2661 * Handle private futexes in the NPTL implementation.
2662 Implemented by Jakub Jelinek and Ulrich Drepper.
2663
2664 * Add support for O_CLOEXEC. Implement in Hurd. Use throughout libc.
2665 Implemented by Roland McGrath and Ulrich Drepper.
2666
2667 * Linux/x86-64 vDSO support. Implemented by Ulrich Drepper.
2668
2669 * SHA-256 and SHA-512 based password encryption.
2670 Implemented by Ulrich Drepper.
2671
2672 * New locales: ber_DZ, ber_MA, en_NG, fil_PH, fur_IT, fy_DE, ha_NG, ig_NG,
2673 ik_CA, iu_CA, li_BE, li_NL, nds_DE, nds_NL, pap_AN, sc_IT, tk_TM, ug_CN,
2674 yo_NG.
2675
2676 + New iconv modules: MAC-CENTRALEUROPE, ISO-8859-9E, KOI8-RU.
2677 Implemented by Ulrich Drepper.
2678 \f
2679 Version 2.6
2680
2681 * The following bugs are resolved with this release:
2682
2683 3156, 3213, 3285, 3291, 3306, 3313, 3320, 3322, 3325, 3326, 3334, 3348,
2684 3352, 3363, 3369, 3426, 3427, 3429, 3451, 3458, 3465, 3480, 3483, 3493,
2685 3514, 3515, 3559, 3632, 3664, 3673, 3674, 3745, 3747, 3818, 3840, 3842,
2686 3851, 3855, 3884, 3885, 3902, 3905, 3919, 3944, 3954, 3955, 3957, 3991,
2687 3995, 3996, 4024, 4040, 4069, 4070, 4074, 4076, 4096, 4101, 4102, 4114,
2688 4130, 4131, 4181, 4276, 4306, 4342, 4344, 4349, 4364, 4368, 4381, 4392,
2689 4403, 4405, 4406, 4411, 4438, 4439, 4455, 4456, 4465, 4512, 4514, 4586,
2690 4702, 4858
2691
2692 * New Linux interfaces: epoll_pwait, sched_getcpu.
2693
2694 * New generic interfaces: strerror_l.
2695
2696 * nscd can now cache the services database. Implemented by Ulrich Drepper.
2697 \f
2698 Version 2.5
2699
2700 * The following bugs are resolved with this release:
2701
2702 39, 68, 192, 352, 388, 531, 935, 1006, 1201, 1203, 1386, 1782, 1783, 1784,
2703 1785, 1786, 1787, 1951, 1974, 1996, 2072, 2098, 2125, 2134, 2158, 2182,
2704 2349, 2376, 2386, 2389, 2415, 2418, 2420, 2423, 2450, 2451, 2466, 2472,
2705 2473, 2477, 2498, 2499, 2501, 2502, 2505, 2507, 2508, 2509, 2517, 2526,
2706 2569, 2571, 2592, 2611, 2625, 2632, 2644, 2662, 2680, 2683, 2684, 2693,
2707 2695, 2703, 2734, 2751, 2764, 2766, 2775, 2792, 2821, 2823, 2832, 2841,
2708 2843, 2883, 2892, 2908, 2914, 2926, 2961, 2978, 2980, 2987, 2997, 2998,
2709 3013, 3018, 3034, 3040, 3044, 3123, 3124, 3137, 3138, 3143, 3155, 3189,
2710 3225, 3251, 3252, 3253, 3273
2711
2712 * For Linux, the sorting of addresses returned by getaddrinfo now also
2713 handles rules 3, 4, and 7 from RFC 3484. I.e., all rules are handled.
2714 Implemented by Ulrich Drepper.
2715
2716 * Allow system admin to configure getaddrinfo with the /etc/gai.conf file.
2717 Implemented by Ulrich Drepper.
2718
2719 * New Linux interfaces: splice, tee, sync_file_range, vmsplice.
2720
2721 * New iconv module for MIK. Contributed by Alexander Shopov.
2722
2723 * For sites with broken group and/or passwd database, the auto-propagate
2724 option of nscd can prevent creating ID lookup entries from the results
2725 of a name lookup and vice versa. This usually is no problem but some
2726 site might have problems with the default behavior.
2727 Implemented by Ulrich Drepper.
2728
2729 * Iterating over entire database in NIS can be slow. With the
2730 SETENT_BATCH_READ option in /etc/default/nss a system admin can decide
2731 to trade time for memory. The entire database will be read at once.
2732 Implemented by Ulrich Drepper.
2733
2734 * The interfaces introduced in RFC 3542 have been implemented by
2735 Ulrich Drepper.
2736
2737 * Support for the new ELF hash table format was added by Ulrich Drepper.
2738
2739 * Support for priority inheritance mutexes added by Jakub Jelinek and
2740 Ulrich Drepper.
2741
2742 * Support for priority protected mutexes added by Jakub Jelinek.
2743
2744 * New locales: nr_ZA, pa_PK, ca_AD, ca_FR, ca_IT, el_CY, tr_CY, as_IN,
2745 or_IN, csb_PL, fy_NL, sr_ME.
2746 \f
2747 Version 2.4
2748
2749 * More overflow detection functions.
2750
2751 * New iconv converters for IBM1025, IBM1122, IBM1137, IBM1153,
2752 IBM1154, IBM1155, IBM1156, IBM1157, and IBM1158 contributed by Jiro Sekiba.
2753
2754 More converters for IBM803, IBM901, IBM902, IBM921, IBM1008, IBM1097,
2755 IBM1112, IBM1123, IBM1130, IBM1140, IBM1141, IBM1142, IBM1143, IBM1144,
2756 IBM1145, IBM1146, IBM1147, IBM1148, IBM1149, IBM1166, IBM1167, IBM4517,
2757 IBM4899, IBM4909, IBM4971, IBM5347, IBM9030, IBM9066, IBM9448, IBM12712,
2758 IBM16804, IBM1364, IBM1371, IBM1388, IBM1390, and IBM1399 contributed
2759 by Masahide Washizawa.
2760
2761 * It is now possible to install audit libraries for the dynamic linker using
2762 LD_AUDIT. Implemented by Ulrich Drepper.
2763
2764 * The LinuxThreads add-on, providing pthreads on Linux 2.4 kernels, is no
2765 longer supported. The new NPTL implementation requires Linux 2.6 kernels.
2766 For a libc and libpthread that works well on Linux 2.4 kernels, we
2767 recommend using the stable 2.3 branch.
2768
2769 * The new function open_wmemstream acts like open_memstream,
2770 but uses a wchar_t wide character buffer.
2771
2772 * The new function ppoll is an improved version of poll, similar to pselect.
2773
2774 * New interfaces: fdopendir, openat, fstatat, fchmodat, fchownat,
2775 futimesat, faccessat, mkdirat, mkfifoat, mknodat,
2776 renameat, unlinkat, linkat, symlinkat, readlinkat.
2777
2778 * New Linux kernel system calls: unshare,
2779 inotify_init, inotify_add_watch, inotify_rm_watch.
2780
2781 * The euidaccess function is now also known by the alias eaccess,
2782 for compatibility with some other systems.
2783
2784 * Timezone data updated to 2006b version.
2785 \f
2786 Version 2.3.6
2787
2788 * The following bugs are resolved with this release:
2789
2790 38, 253, 549, 622, 653, 721, 758, 851, 877, 915, 934, 955, 961,
2791 1016, 1037, 1076, 1079, 1080, 1081, 1082, 1083, 1084, 1085, 1086,
2792 1087, 1088, 1090, 1091, 1092, 1093, 1094, 1095, 1096, 1097, 1098,
2793 1099, 1100, 1101, 1102, 1103, 1104, 1105, 1106, 1107, 1108, 1109,
2794 1110, 1111, 1112, 1113, 1125, 1137, 1138, 1249, 1250, 1251, 1252,
2795 1253, 1254, 1350, 1358, 1394, 1438, 1498, 1534
2796
2797 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
2798
2799 * As of this release, GCC 4 can be used to compile the C Library.
2800
2801 * Timezone data updated to 2005m version.
2802 \f
2803 Version 2.3.5
2804
2805 * The following bugs are resolved with this release:
2806
2807 284, 592, 602, 626, 633, 640, 650, 661, 671, 681, 693, 700, 710, 719,
2808 722, 723, 725, 726, 727, 728, 729, 730, 731, 732, 733, 734, 735, 736,
2809 737, 738, 739, 740, 741, 742, 743, 744, 745, 765, 767, 768, 769, 776,
2810 777, 787, 821, 822, 823, 825
2811
2812 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
2813 \f
2814 Version 2.3.4
2815
2816 * Support for RFC 3678. Real implementations exist only for Linux so far.
2817 Implemented by Ulrich Drepper.
2818
2819 * nscd can now cache entries persistently. Expiring entries are reloaded.
2820 For speedups the cache can be shared in memory with client processes.
2821 Implemented by Ulrich Drepper.
2822
2823 * nscd can now perform SELinux checks.
2824 Implemented by Matthew Rickard <mjricka@epoch.ncsc.mil>.
2825
2826 * getaddrinfo queries are now cached. Canonical name lookup is performed
2827 efficiently.
2828 Implemented by Ulrich Drepper.
2829
2830 * The nothrow function attribute is used when headers are used by gcc when
2831 compiling C code. This can avoid significant amounts of exception
2832 handling data.
2833
2834 * The malloc functions perform more error checking and are stricter when
2835 it comes to reacting on errors. The default action is to terminate
2836 the process after showing an error message. Implemented by Ulrich Drepper.
2837
2838 * Reverse lookups of IPv6 addresses does not use bit string or .ip6.int
2839 lookups anymore unless explicitly requested. Implemented by Ulrich Drepper.
2840
2841 * Namespaces in ld.so are implemented. DSOs can be loaded in separate
2842 namespaces using the new function dlmopen(). This feature is of course,
2843 like most other dynamic loading functionality, not available in statically
2844 linked applications. Implemented by Ulrich Drepper.
2845
2846 * Low-overhead boundary checking variants of string and some stdio functions
2847 were added. These are to be used in conjunction with a gcc patch by
2848 Jakub Jelinek which adds calls to these functions if possible.
2849 Implemented by Jakub Jelinek and Ulrich Drepper.
2850
2851 * Old code for several operating systems and machine architectures that
2852 have not been in working condition in a long time have been removed from
2853 the main source tree maintained by the GNU C Library's maintainers.
2854 These files are now reside in the separate `ports' source module
2855 that is usable as an add-on when building the library.
2856 \f
2857 Version 2.3.3
2858
2859 * New functions `dladdr1' and `dlinfo' in <dlfcn.h> provide more ways to
2860 interrogate the dynamic linker, compatible with the Solaris interface.
2861
2862 * ELF thread-local storage support (TLS) now works on PowerPC and PowerPC64;
2863 implemented by Paul Mackerras, Steven Munroe, and Roland McGrath.
2864
2865 * getifaddrs now uses the netlink interface on Linux to get its information.
2866 Implemented by Thorsten Kukuk.
2867
2868 * getaddrinfo now implements AI_V4MAPPED, AI_ALL, and AI_ADDRCONF.
2869 Implemented by Ulrich Drepper.
2870
2871 * support for non-executable stacks on x86 has been added. Changes mostly
2872 by Roland McGrath.
2873
2874 * regex is now much faster for multibyte locales. Changes by Jakub Jelinek
2875 and Ulrich Drepper.
2876
2877 * getaddrinfo now performs destination address selection according to
2878 RFC 3484.
2879 \f
2880 Version 2.3.2
2881
2882 * Thread-safe interfaces for many functions that access locale data
2883 were added in version 2.3, but these features were omitted from NEWS.
2884 Many functions have variants with an `_l' suffix that take a `locale_t'
2885 object as a parameter rather than consulting the current locale.
2886 The new functions `newlocale', `duplocale', and `freelocale' in <locale.h>
2887 create and maintain `locale_t' objects. Additionally, the new function
2888 `uselocale' sets "the current locale" (as used by functions not so
2889 parameterized) set for an individual thread. These features were added
2890 in version 2.3, implemented by Ulrich Drepper and Roland McGrath.
2891
2892 * The functions getresuid, getresgid, setresuid, and setresgid, which
2893 have long been available on Linux, are now declared in <unistd.h>
2894 and are now also available on the Hurd.
2895
2896 * ELF thread-local storage support (TLS) now works on x86-64.
2897
2898 * The new dynamic string token $LIB is expanded in shared library names.
2899 This normally expands to lib, but on some 64-bit platforms to lib64 instead.
2900
2901 * Aldy Hernandez contributed complete software floating point support for
2902 PowerPC machines with no FPU.
2903
2904 * fexecve is implemented on Linux.
2905
2906 * The `btowc' function should work at least twice as fast due to
2907 specialized callbacks in the iconv modules. Implemented by Bruno Haible.
2908
2909 * With approriate thread add-ons cancelable functions are now implemented
2910 in libc.so as well. No need to call the function in libpthread. This
2911 change allowed to finally disable the incorrect and expensive handling
2912 of weak definition in ld.so.
2913
2914 * Yet more PLT entries in libc.so have been removed. We finally arrived
2915 at the bare minimum. Startup times improved appropriately.
2916
2917 * Support for the new Linux/x86 system call interface was added. The
2918 AT_SYSINFO auxiliary vector entry is recognized and handled.
2919 \f
2920 Version 2.3
2921
2922 * Masahide Washizawa contributed iconv modules for IBM1163 and IBM1164
2923 charsets.
2924
2925 * iconv (the program and the interface) now accepts empty names (excluding
2926 options like //TRANSLIT) to mean "use charset of current locale".
2927
2928 * localedef can now transliterate characters in strings which are not in
2929 the provided charmap. The information from the input locale is used.
2930
2931 * Prelinking support was added for ELF targets. This requires additional
2932 tools and recent versions of the GNU binutils. Contributed by Jakub Jelinek.
2933
2934 * Read-only stdio streams now use mmap to speed up operation by eliminating
2935 copying and buffer underflows. To use add 'm' to the mode string of
2936 the fopen/fdopen/freopen call. Implemented by Ulrich Drepper.
2937
2938 * The malloc functions were completely rewritten by Wolfram Gloger based
2939 on Doug Lea's malloc-2.7.0.c.
2940
2941 * Isamu Hasegawa contributed a completely new and POSIX-conformant
2942 implementation of regex.
2943
2944 * Bruno Haible upgraded the iconv and locale implementation to support
2945 Unicode 3.2.
2946
2947 * Contents of the LC_* and LANG environment variables in the CEN style are
2948 not recognized anymore. It never was used. Change by Ulrich Drepper.
2949
2950 * The runtime (ld.so, libc, libpthread for Linux) now can handle the ELF
2951 thread-local storage (TLS) ABI on some platforms.
2952 Changes by Ulrich Drepper. SH support by Kaz Kojima.
2953
2954 * Bruno Haible contributed iconv converters for ISO-2022-JP-3, SHIFT JIS-X0213,
2955 EUC-JISX0213, and TSCII.
2956
2957 * New header <ifaddrs.h> with functions `getifaddrs' and `freeifaddrs':
2958 BSD-compatible interface for getting all network interface addresses.
2959 Implementation for IPv4 by Roland McGrath.
2960
2961 * Loading of locale data is faster due to the introduction of a locale
2962 archive. Implemented by Roland McGrath and Ulrich Drepper.
2963
2964 * Startup times are significantly reduced by not using exported functions
2965 inside the library itself. Changes by Jakub Jelinek, Roland McGrath,
2966 and Ulrich Drepper.
2967
2968 * Steven Munroe contributed a port to PowerPC64/Linux.
2969 \f
2970 Version 2.2.6
2971
2972 * The Hurd now uses the GNU libio implementation of stdio rather than the
2973 old GNU stdio implementation, and uses a new ABI (libc.so.0.3).
2974
2975 * The Hurd on x86 now has the `ioperm' function and <sys/io.h> header file
2976 with the same behavior as the Linux system call of the same name.
2977 \f
2978 Version 2.2.5
2979
2980 * Stephen Moshier implemented log2, log10, powl and cbrtl for the
2981 128-bit long double format.
2982
2983 * Masahide Washizawa contributed iconv modules for IBM1132, IBM1133, IBM1160,
2984 IBM1161, and IBM1162 charsets.
2985
2986 * Andreas Jaeger contributed a port to x86-64/Linux.
2987
2988 * Peter Bruin contributed a port to PowerPC/Hurd.
2989
2990 * libc functions using I/O streams now can handle wide-oriented streams
2991 as well.
2992
2993 * optimizations in the dynamic linker. Binaries created by recent binutils
2994 versions start up quicker due to reduced time spent on relocations.
2995
2996 * Support for use of gcc3 added by Jakub Jelinek and HJ Lu.
2997 \f
2998 Version 2.2.4
2999
3000 * Stephen Moshier implemented cosh, expm1, log1p, acos, sinh, tanh,
3001 asinh, atanh, j0 for the 128-bit long double format.
3002
3003 * Bruno Haible updated all the code handling Unicode in some form to
3004 support Unicode 3.1.
3005
3006 * Speed of regex for single-byte locales is back to previous levels.
3007 Patch by Isamu Hasegawa.
3008
3009 * Alpha, SPARC, and IA-64 now also using floating stacks.
3010
3011 * Startup time of internationalized applications greatly improved through
3012 iconv cache. Use iconvconfig to generate the cache file.
3013 Contributed by Ulrich Drepper.
3014
3015 * The IA-64 specific part of ld.so was rewritten to eliminate some pretty
3016 severe performance problems. Patch by David Mosberger.
3017
3018 * The Hurd port got a lot more functionality like AIO, various stdio
3019 extensions, etc. Mainly done by Roland McGrath.
3020
3021 * mtrace can now lookup symbols in shared libraries.
3022 \f
3023 Version 2.2.3
3024
3025 * Intel's IA-64 math library is largely integrated. It provides fast and
3026 accurate implementatations for most basic and standard math functions
3027 in float, double, and long double format.
3028
3029 * Stephen Moshier implemented j0, j1, jn, y0, y1, yn, lgamma, erf, erfc,
3030 and asin for the 96-bit long double format and asin, log, tan for the
3031 128-bit long double format.
3032
3033 * The beginning of a last-bit accurate math library by IBM Haifa were added.
3034 The basic double functions exist today. Contributed by Abraham Ziv
3035 <ziv@il.ibm.com>, Moshe Olshansky <olshansk@il.ibm.com>, Ealan Henis
3036 <ealan@il.ibm.com>, and Anna Reitman <reitman@il.ibm.com>.
3037
3038 * An asynchronous name lookup library was added. The interface is designed
3039 after POSIX AIO. The proposal was circulated beforehand to get comments.
3040 No negative ones came in. Implemented by Ulrich Drepper.
3041
3042 * Port to S390/64bit contributed by Martin Schwidefsky
3043 <schwidefsky@de.ibm.com>.
3044
3045 * David Mosberger <davidm@hpl.hp.com> implemented the setcontext family
3046 of functions for Linux/IA-64.
3047
3048 * The RPC code is now thread safe. Threads can now use the same service
3049 of different services at the same time. Patch by Eric Norum
3050 <eric.norum@usask.ca> with some help by Ulrich Drepper.
3051
3052 * Martin Schwidefsky <schwidefsky@de.ibm.com> implemented the setcontext
3053 family of functions for Linux/S390.
3054
3055 * Ulrich Drepper <drepper@redhat.com> implemented the setcontext family
3056 of functions for Linux/x86.
3057
3058 * Port to Linux/CRIS contributed by Axis Communications.
3059 \f
3060 Version 2.2.2
3061
3062 * Lots of headers were cleaned up. Using the tool in the conform/ subdir
3063 we can now check for namespace violations and missing declarations. The
3064 result is that almost all headers are now Unix-compliant (as defined in
3065 the upcoming XPG6). The negative side is that some programs might need
3066 corrections, too, if they depend on the incorrect form of the headers in
3067 previous versions which defined too many symbols and included too many
3068 other headers.
3069
3070 * regex now handles multibyte character sets correctly.
3071 Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
3072
3073 * iconv (the program) does now conform to the upcoming XPG6 and handles
3074 charmaps. Instead of the charset names the path of charmaps can be
3075 provided and the conversion happens based on this data.
3076 Contributed by Ulrich Drepper.
3077
3078 * The locale program now provides detailed information about the installed
3079 locales. While
3080
3081 locale -a
3082
3083 only lists the names of the supported locales
3084
3085 locale -a --verbose
3086
3087 provides details such as country, language, and codeset name.
3088 Contributed by Ulrich Drepper.
3089 \f
3090 Version 2.2.1
3091
3092 * The gencat program now parses the input file according to the charset
3093 selected by the LC_CTYPE category. This is important for stateful
3094 character sets. To make generating catalogs easier there is a way
3095 to overwrite the charset selected by the locale: before the first
3096 message or $ quote line the catalog can contain a line like
3097
3098 $ codeset=ISO-8859-2
3099
3100 to select the charset (ISO-8859-2 in this case).
3101
3102 Implemented by Shinya Hanataka and Ulrich Drepper.
3103
3104 * New codeset conversion modules: IBM-922 (Estonia PC codeset),
3105 IBM-1124 (Ukraine PC codeset), IBM-1129 (Vietnamese PC codeset).
3106 Contributed by Masahide Washizawa <washi@jp.ibm.com>.
3107
3108 * Optimized string functions for Alpha ev6 and ev67 by Richard Henderson
3109 <rth@redhat.com> and Rick Gorton <rick.gorton@alpha-processor.com>.
3110
3111 * The LANGUAGE environment variable is now ignored unless the locale is
3112 changed from the default "C" locale.
3113
3114 * The usual bug fixes.
3115 \f
3116 Version 2.2
3117
3118 * Greg McGary added runtime support for bounds checking using gcc's
3119 new -fbounded-pointers option. ix86 target is complete. PowerPC
3120 is in progress.
3121
3122 * Thorsten Kukuk added secure mode support to nscd.
3123
3124 * The Berkeley db libraries have been removed.
3125
3126 Related, the nss_db module is now in a separate package since it
3127 obviously requires a database library being available.
3128
3129 * Wide character I/O streams implemented by Ulrich Drepper.
3130
3131 * Functions from the extended socket API added by Ulrich Drepper.
3132
3133 * Functions feenableexcept and fedisableexcept to control the
3134 behaviour of individual exceptions have been added by Andreas Jaeger.
3135
3136 * ldconfig program added by Andreas Jaeger and Jakub Jelinek.
3137
3138 * The resolver code has been updated from bind 8.2.3-T5B which supports
3139 threads. The integration was done by Andreas Jaeger, Adam D. Bradley,
3140 and Mark Kettenis.
3141
3142 This change could in some situations effect backward compatibility. Since
3143 now `_res' is a thread-local instead of a global variable, modifying it
3144 in one thread does not have any effect in other threads.
3145
3146 The resolver library was also extended to allow IPv6 as the transport
3147 protocol for the requests. This work was done by Stig Venaas.
3148
3149 * Compatibility code for K&R C compilers has been removed from the
3150 header files. A ISO C compiler is needed to use the library
3151 (conforming to either C89 or C99 standard).
3152
3153 * Complete rewrite of the localedef program to support multibyte character
3154 sets. Implement handling of ISO 14651 and ISO 14652. Rewrite strcoll,
3155 strxfrm, wcscoll, and wcsxfrm functions. Make isw*() functions work.
3156 Implemented by Ulrich Drepper.
3157
3158 Bruno Haible significantly improved the generation and use of the data
3159 structures for the wide character tables.
3160
3161 * Plural handling in gettext implemented by Ulrich Drepper.
3162
3163 * The utmp daemon has been removed.
3164
3165 * The port to MIPS-Linux has been finished by Andreas Jaeger.
3166
3167 * A port to Hitachi SH3 and SH4 has been contributed by Kazumoto Kojima
3168 and Yutaka Niibe.
3169
3170 * POSIX clocks and timers implemented by Kaz Kylheku and Ulrich Drepper.
3171
3172 * POSIX barriers implemented by Kaz Kylheku.
3173
3174 * POSIX spawn function family implemented by Ulrich Drepper.
3175
3176 * POSIX spinlocks are now available.
3177
3178 * Timed wait functions for mutex, rwlock, and semaphores are implemented.
3179
3180 * the configure option --enable-kernel=X.Y.Z allows to strip out
3181 compatibility for kernel versions before X.Y.Z. This is currently only
3182 implemented for Linux.
3183
3184 * the sockaddr_in6 structure changed. The IPv6 working group added a new
3185 field sin6_scope_id. This means that all programs using IPv6 should be
3186 recompiled. Don't expect binary compatibility with previous glibc
3187 versions.
3188
3189 * various conversion modules for IBM character sets contributed by
3190 Masahide Washizawa.
3191
3192 * IA-64 port by Jes Sorensen and HJ Lu.
3193 \f
3194 Version 2.1.3
3195
3196 * bug fixes
3197
3198 \f
3199 Version 2.1.2
3200
3201 * bug fixes
3202
3203 \f
3204 Version 2.1.1
3205
3206 * New ISO C 9x function _Exit, imaxabs, and imaxdiv are added.
3207
3208 * New xdr functions are added; some rpc functions are now 64bit clean.
3209
3210 * Fixed a number of bugs and memory leaks (especially in NIS+ code).
3211
3212 * Fixed known incompatibilities with glibc 2.0.
3213
3214 * New functions lock64, strchrnul, rawmemchr, getutmp and getutmpx.
3215
3216 * Optimized a number of functions (especially the ELF dynamic loader).
3217
3218 * Update timezone data files.
3219
3220 * lots of charmaps corrections
3221
3222 * some new locale definitions and charmaps
3223
3224 \f
3225 Version 2.1
3226
3227 * Richard Henderson corrected size of struct timeval on Linux/Alpha to
3228 conform to POSIX member type requirements. Symbol versions have been
3229 adjusted as needed within the library, and for direct use by applications,
3230 but there is potential for problems if third-party libraries use
3231 struct timeval as part of their interface. This does not present
3232 a problem for X and other "essential" system libraries.
3233
3234 * An additional locale model to support C++ Standard Library locale
3235 model and probably more was implemented by Ulrich Drepper.
3236
3237 * Eric Youngdale and Ulrich Drepper implemented versioning of objects on
3238 symbol level.
3239
3240 * Miles Bader provided the `argp' function family to support hierachical
3241 command line argument parsing, layered on top of getopt.
3242
3243 * strtod accepts new hexadecimal floating-point format from ISO C 9X.
3244
3245 * printf has two new specifiers %a and %A to print hexadecimal flaoting-point
3246 numbers.
3247
3248 * scanf recognizes the %a and %A format for scanning floating point numbers.
3249
3250 * the new headers <stdint.h> and <inttypes.h> from ISO C 9X provides
3251 information and interfaces for the available integer types.
3252
3253 * about 130 new math functions were added to implement the ISO C9x math
3254 library.
3255
3256 * the new header <complex.h> contains definitions of the complex math
3257 functions from ISO C 9X.
3258
3259 * the new header <tgmath.h> defines generic macros to use complex or
3260 real valued functions.
3261
3262 * Thorsten Kukuk provided an implementation for NIS+, securelevel 0, 1 and 2.
3263
3264 * Andreas Jaeger provided a test suite for the math library.
3265
3266 * Mark Kettenis implemented the utmpx interface and an utmp daemon.
3267
3268 * Ulrich Drepper added character set conversion functions (iconv).
3269
3270 * Optimized string functions have been added.
3271
3272 * The localedata addon is now part of glibc.
3273
3274 * An implementation of profiling shared libraries was added by Ulrich Drepper.
3275
3276 * Thorsten Kukuk and Ulrich Drepper provided an implementation for a caching
3277 daemon for NSS (nscd).
3278
3279 Missing a better place here are some numbers on improvements. Under
3280 Linux 2.1.125 un-tar-ing the kernel sources takes
3281
3282 user system wall
3283
3284 using local files 12.19s 6.88s 22.91s
3285
3286 using NIS 13.92s 8.91s 26.34s
3287
3288 using NIS & nscd 10.37s 7.34s 25.30s
3289
3290 using NIS+ 27.57s 30.37s 640.46s
3291
3292 using NIS+ & nscd 10.25s 7.83s 26.51s
3293
3294 using NIS & old nscd [1] 13.83s 8.32s 29.60s
3295
3296 Keep in mind that non-namelookup related operations dominate above times.
3297 It was just a common complain that using NIS+ unpacking the kernel is
3298 horribly slow.
3299
3300 [1] The old nscd implementation is not available anymore since it was
3301 distributed with glibc up to version 2.0.98 and thus is now replaced.
3302
3303 * Tim Waugh provided an implementation of the POSIX.2 wordexp function family.
3304
3305 * Mark Kettenis provided a Hesiod NSS module.
3306
3307 * The ELF dynamic loader knows how to process dynamic string tokens ($ORIGIN
3308 and $PLATFORM) in RPATHs and similar strings (Ulrich Drepper).
3309
3310 * rcmd can now handle netgroups (Dick Streefland).
3311
3312 * A port to the ARM was contributed by Phil Blundell, Pat Beirne and Scott
3313 Bambrough.
3314
3315 * Support for the IPv6 protocol has been added to the socket API, as per the
3316 latest draft standards.
3317
3318 * Support for Linux 2.2 has been added.
3319
3320 * Interface changes relative to the latest 2.0.x release:
3321 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3322 addseverity NEW: Unix98
3323 alphasort64 NEW: LFS
3324 argp_err_exit_status NEW: argp, GNU ext
3325 argp_error NEW: argp, GNU ext
3326 argp_failure NEW: argp, GNU ext
3327 argp_help NEW: argp, GNU ext
3328 argp_parse NEW: argp, GNU ext
3329 argp_program_bug_address NEW: argp, GNU ext
3330 argp_program_version NEW: argp, GNU ext
3331 argp_program_version_hook NEW: argp, GNU ext
3332 argp_state_help NEW: argp, GNU ext
3333 argp_usage NEW: argp, GNU ext
3334 authdes_create NEW: Secure RPC
3335 authdes_getucred NEW: Secure RPC
3336 authdes_pk_create NEW: Secure RPC
3337 backtrace NEW: GNU ext.
3338 backtrace_symbols NEW: GNU ext.
3339 backtrace_symbols_fd NEW: GNU ext.
3340 cacos NEW: ISO C 9x
3341 cacosf NEW: ISO C 9x
3342 cacosh NEW: ISO C 9x
3343 cacoshf NEW: ISO C 9x
3344 cacoshl NEW: ISO C 9x
3345 cacosl NEW: ISO C 9x
3346 capget NEW: kernel
3347 capset NEW: kernel
3348 carg NEW: ISO C 9x
3349 cargf NEW: ISO C 9x
3350 cargl NEW: ISO C 9x
3351 casin NEW: ISO C 9x
3352 casinf NEW: ISO C 9x
3353 casinh NEW: ISO C 9x
3354 casinhf NEW: ISO C 9x
3355 casinhl NEW: ISO C 9x
3356 casinl NEW: ISO C 9x
3357 catan NEW: ISO C 9x
3358 catanf NEW: ISO C 9x
3359 catanh NEW: ISO C 9x
3360 catanhf NEW: ISO C 9x
3361 catanhl NEW: ISO C 9x
3362 catanl NEW: ISO C 9x
3363 cbc_crypt NEW: Secure RPC
3364 ccos NEW: ISO C 9x
3365 ccosf NEW: ISO C 9x
3366 ccosh NEW: ISO C 9x
3367 ccoshf NEW: ISO C 9x
3368 ccoshl NEW: ISO C 9x
3369 ccosl NEW: ISO C 9x
3370 cexp NEW: ISO C 9x
3371 cexpf NEW: ISO C 9x
3372 cexpl NEW: ISO C 9x
3373 cimag NEW: ISO C 9x
3374 cimagf NEW: ISO C 9x
3375 cimagl NEW: ISO C 9x
3376 clearerr_locked REMOVED
3377 clntunix_create NEW: sunrpc ext
3378 clog NEW: ISO C 9x
3379 clog10 NEW: ISO C 9x
3380 clog10f NEW: ISO C 9x
3381 clog10l NEW: ISO C 9x
3382 clogf NEW: ISO C 9x
3383 clogl NEW: ISO C 9x
3384 conj NEW: ISO C 9x
3385 conjf NEW: ISO C 9x
3386 conjl NEW: ISO C 9x
3387 cpow NEW: ISO C 9x
3388 cpowf NEW: ISO C 9x
3389 cpowl NEW: ISO C 9x
3390 cproj NEW: ISO C 9x
3391 cprojf NEW: ISO C 9x
3392 cprojl NEW: ISO C 9x
3393 creal NEW: ISO C 9x
3394 crealf NEW: ISO C 9x
3395 creall NEW: ISO C 9x
3396 creat64 NEW: LFS
3397 csin NEW: ISO C 9x
3398 csinf NEW: ISO C 9x
3399 csinh NEW: ISO C 9x
3400 csinhf NEW: ISO C 9x
3401 csinhl NEW: ISO C 9x
3402 csinl NEW: ISO C 9x
3403 csqrt NEW: ISO C 9x
3404 csqrtf NEW: ISO C 9x
3405 csqrtl NEW: ISO C 9x
3406 ctan NEW: ISO C 9x
3407 ctanf NEW: ISO C 9x
3408 ctanh NEW: ISO C 9x
3409 ctanhf NEW: ISO C 9x
3410 ctanhl NEW: ISO C 9x
3411 ctanl NEW: ISO C 9x
3412 des_setparity NEW: Secure RPC
3413 ecb_crypt NEW: Secure RPC
3414 endutxent NEW: Unix98
3415 exp10 NEW: ISO C 9x
3416 exp10f NEW: ISO C 9x
3417 exp10l NEW: ISO C 9x
3418 exp2 NEW: ISO C 9x
3419 exp2f NEW: ISO C 9x
3420 exp2l NEW: ISO C 9x
3421 fattach NEW: STREAMS
3422 fdetach NEW: STREAMS
3423 fdim NEW: ISO C 9x
3424 fdimf NEW: ISO C 9x
3425 fdiml NEW: ISO C 9x
3426 feclearexcept NEW: ISO C 9x
3427 fegetenv NEW: ISO C 9x
3428 fegetexceptflag NEW: ISO C 9x
3429 fegetround NEW: ISO C 9x
3430 feholdexcept NEW: ISO C 9x
3431 feof_locked REMOVED
3432 feraiseexcept NEW: ISO C 9x
3433 ferror_locked REMOVED
3434 fesetenv NEW: ISO C 9x
3435 fesetexceptflag NEW: ISO C 9x
3436 fesetround NEW: ISO C 9x
3437 fetestexcept NEW: ISO C 9x
3438 feupdateenv NEW: ISO C 9x
3439 fflush_locked REMOVED
3440 ffsl NEW: GNU ext.
3441 ffsll NEW: GNU ext.
3442 fgetpos64 NEW: LFS
3443 fgets_unlocked NEW: GNU ext.
3444 fileno_locked REMOVED
3445 fma NEW: ISO C 9x
3446 fmaf NEW: ISO C 9x
3447 fmal NEW: ISO C 9x
3448 fmax NEW: ISO C 9x
3449 fmaxf NEW: ISO C 9x
3450 fmaxl NEW: ISO C 9x
3451 fmin NEW: ISO C 9x
3452 fminf NEW: ISO C 9x
3453 fminl NEW: ISO C 9x
3454 fmtmsg NEW: Unix98
3455 fopen64 NEW: LFS
3456 fputc_locked REMOVED
3457 fputs_unlocked NEW: GNU ext.
3458 fread_unlocked NEW: GNU ext.
3459 freopen64 NEW: LFS
3460 fseeko NEW: Unix98
3461 fsetpos64 NEW: LFS
3462 fstatfs64 NEW: LFS
3463 fstatvfs NEW: Unix98
3464 fstatvfs64 NEW: LFS
3465 ftello NEW: Unix98
3466 ftello64 NEW: LFS
3467 ftruncate64 NEW: LFS
3468 ftw64 NEW: LFS
3469 fwrite_unlocked NEW: GNU ext.
3470 gai_strerror NEW: IPv6
3471 gamma_r REMOVED
3472 gammaf_r REMOVED
3473 gammal_r REMOVED
3474 getchar_locked REMOVED
3475 getdate NEW: Unix98
3476 getdate_err NEW: Unix98
3477 getdate_r NEW: GNU ext.
3478 getmsg NEW: STREAMS
3479 getnameinfo NEW: IPv6
3480 getnetname NEW: Secure RPC
3481 getpmsg NEW: STREAMS
3482 getpt NEW: Unix98 PTY
3483 getrlimit64 NEW: LFS
3484 getutxent NEW: Unix98
3485 getutxid NEW: Unix98
3486 getutxline NEW: Unix98
3487 glob64 NEW: GNU ext.
3488 globfree64 NEW: GNU ext.
3489 gnu_get_libc_release NEW: GNU ext.
3490 gnu_get_libc_version NEW: GNU ext.
3491 grantpt NEW: Unix98 PTY
3492 host2netname NEW: Secure RPC
3493 iconv NEW: iconv
3494 iconv_close NEW: iconv
3495 iconv_open NEW: iconv
3496 if_freenameindex NEW: IPv6
3497 if_indextoname NEW: IPv6
3498 if_nameindex NEW: IPv6
3499 if_nametoindex NEW: IPv6
3500 in6addr_any NEW: IPv6
3501 in6addr_loopback NEW: IPv6
3502 inet6_isipv4mapped NEW: IPv6
3503 isastream NEW: STREAMS
3504 iswblank NEW: GNU ext.
3505 key_decryptsession NEW: Secure RPC
3506 key_decryptsession_pk NEW: Secure RPC
3507 key_encryptsession NEW: Secure RPC
3508 key_encryptsession_pk NEW: Secure RPC
3509 key_gendes NEW: Secure RPC
3510 key_get_conv NEW: Secure RPC
3511 key_secretkey_is_set NEW: Secure RPC
3512 key_setnet NEW: Secure RPC
3513 key_setsecret NEW: Secure RPC
3514 llrint NEW: ISO C 9x
3515 llrintf NEW: ISO C 9x
3516 llrintl NEW: ISO C 9x
3517 llround NEW: ISO C 9x
3518 llroundf NEW: ISO C 9x
3519 llroundl NEW: ISO C 9x
3520 log2 NEW: ISO C 9x
3521 log2f NEW: ISO C 9x
3522 log2l NEW: ISO C 9x
3523 lrint NEW: ISO C 9x
3524 lrintf NEW: ISO C 9x
3525 lrintl NEW: ISO C 9x
3526 lround NEW: ISO C 9x
3527 lroundf NEW: ISO C 9x
3528 lroundl NEW: ISO C 9x
3529 lseek64 NEW: LFS
3530 makecontext NEW: Unix98
3531 mempcpy NEW: GNU ext.
3532 mmap64 NEW: LFS
3533 moncontrol REMOVED
3534 modify_ldt NEW: kernel
3535 nan NEW: ISO C 9x
3536 nanf NEW: ISO C 9x
3537 nanl NEW: ISO C 9x
3538 nearbyint NEW: ISO C 9x
3539 nearbyintf NEW: ISO C 9x
3540 nearbyintl NEW: ISO C 9x
3541 netname2host NEW: Secure RPC
3542 netname2user NEW: Secure RPC
3543 nexttoward NEW: ISO C 9x
3544 nexttowardf NEW: ISO C 9x
3545 nexttowardl NEW: ISO C 9x
3546 nftw NEW: Unix98
3547 nftw64 NEW: LFS
3548 open64 NEW: LFS
3549 passwd2des NEW: Secure RPC
3550 pow10 NEW: GNU ext.
3551 pow10f NEW: GNU ext.
3552 pow10l NEW: GNU ext.
3553 pread NEW: Unix98
3554 pread64 NEW: LFS
3555 printf_size NEW: GNU ext.
3556 printf_size_info NEW: GNU ext.
3557 profil_counter REMOVED
3558 pthread_mutexattr_getkind_np REPLACED
3559 pthread_mutexattr_setkind_np REPLACED
3560 ptsname NEW: Unix98 PTY
3561 ptsname_r NEW: Unix98 PTY
3562 putc_locked REMOVED
3563 putchar_locked REMOVED
3564 putgrent NEW: GNU ext.
3565 putmsg NEW: STREAMS
3566 putpmsg NEW: STREAMS
3567 pututxline NEW: Unix98
3568 pwrite NEW: Unix98
3569 pwrite64 NEW: LFS
3570 readdir64 NEW: LFS
3571 readdir64_r NEW: LFS
3572 remquo NEW: ISO C 9x
3573 remquof NEW: ISO C 9x
3574 remquol NEW: ISO C 9x
3575 round NEW: ISO C 9x
3576 roundf NEW: ISO C 9x
3577 roundl NEW: ISO C 9x
3578 rtime NEW: GNU ext.
3579 scalbln NEW: ISO C 9x
3580 scalblnf NEW: ISO C 9x
3581 scalblnl NEW: ISO C 9x
3582 scandir64 NEW: LFS
3583 sendfile NEW: kernel
3584 setcontext NEW: Unix98
3585 setrlimit64 NEW: LFS
3586 setutxent NEW: Unix98
3587 sighold NEW: Unix98
3588 sigignore NEW: Unix98
3589 sigqueue NEW: POSIX.1b
3590 sigrelse NEW: Unix98
3591 sigset NEW: POSIX.1b
3592 sigtimedwait NEW: POSIX.1b
3593 sigwaitinfo NEW: POSIX.1b
3594 sincos NEW: GNU ext.
3595 sincosf NEW: GNU ext.
3596 sincosl NEW: GNU ext.
3597 statfs64 NEW: LFS
3598 statvfs NEW: Unix98
3599 statvfs64 NEW: LFS
3600 strcasestr NEW: GNU ext.
3601 strtoimax NEW: ISO C 9x
3602 strtoumax NEW: ISO C 9x
3603 strverscmp NEW: GNU ext.
3604 svcauthdes_stats NEW: Secure RPC
3605 svcunix_create NEW: sunrpc etx
3606 svcunixfd_create NEW: sunrpc ext
3607 swapcontext NEW: Unix98
3608 tcgetsid NEW: Unix98 PTY
3609 tdestroy NEW: GNU ext.
3610 tgamma NEW: ISO C 9x
3611 tgammaf NEW: ISO C 9x
3612 tgammal NEW: ISO C 9x
3613 tmpfile64 NEW: LFS
3614 trunc NEW: ISO C 9x
3615 truncate64 NEW: LFS
3616 truncf NEW: ISO C 9x
3617 truncl NEW: ISO C 9x
3618 umount2 NEW: kernel
3619 unlockpt NEW: Unix98 PTY
3620 updwtmpx NEW: Unix98
3621 user2netname NEW: Secure RPC
3622 utmpxname NEW: Unix98
3623 versionsort NEW: GNU ext.
3624 versionsort64 NEW: GNU ext.
3625 waitid NEW: Unix98
3626 wcscasecmp NEW: GNU ext.
3627 wcsncasecmp NEW: GNU ext.
3628 wcsnlen NEW: GNU ext.
3629 wcstoimax NEW: ISO C 9x
3630 wcstoll NEW: ISO C 9x
3631 wcstoull NEW: ISO C 9x
3632 wcstoumax NEW: ISO C 9x
3633 wcswcs NEW: Unix98
3634 wordexp NEW: POSIX.2
3635 wordfree NEW: POSIX.2
3636 write_profiling REMOVED
3637 xdecrypt NEW: Secure RPC
3638 xdr_authdes_cred NEW: Secure RPC
3639 xdr_authdes_verf NEW: Secure RPC
3640 xdr_cryptkeyarg NEW: Secure RPC
3641 xdr_cryptkeyarg2 NEW: Secure RPC
3642 xdr_cryptkeyres NEW: Secure RPC
3643 xdr_getcredres NEW: Secure RPC
3644 xdr_key_netstarg NEW: Secure RPC
3645 xdr_key_netstres NEW: Secure RPC
3646 xdr_keybuf NEW: Secure RPC
3647 xdr_keystatus NEW: Secure RPC
3648 xdr_netnamestr NEW: Secure RPC
3649 xdr_sizeof NEW: Secure RPC
3650 xdr_unixcred NEW: sunrpc ext
3651 xencrypt NEW: Secure RPC
3652 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3653 \f
3654 Version 2.0.6
3655
3656 * more bug fixes
3657
3658 \f
3659 Version 2.0.5
3660
3661 * more bug fixes
3662
3663 * inet_ntoa is thread-safe
3664
3665 * updwtmp is moved from libutil to libc
3666
3667 * rewrite of cbrt function
3668
3669 * update of timezone data
3670 \f
3671 Version 2.0.4
3672
3673 * more bug fixes
3674 \f
3675 Version 2.0.3
3676
3677 * more bug fixes
3678 \f
3679 Version 2.0.2
3680
3681 * more bug fixes
3682
3683 * add atoll function
3684
3685 * fix complex problems in Berkeley DB code
3686
3687 * fix math functions
3688 \f
3689 Version 2.0.1
3690
3691 * fixed lots of header problems (especially Linux/GNU specific)
3692
3693 * dynamic loader preserves all registers
3694
3695 * Roland McGrath provided support for handling of auxiliary objects in
3696 the ELF dynamic loader.
3697
3698 * support for parallel builds is improved
3699 \f
3700 Version 2.0
3701
3702 * GNU extensions are no longer declared by default. To enable them you
3703 must define the macro `_GNU_SOURCE' in your program or compile with
3704 `-D_GNU_SOURCE'.
3705
3706 * The library has changed from using GNU ld symbol aliases to using weak
3707 symbols where available. The ELF object file format supports weak
3708 symbols; GNU ld also supports weak symbols in the a.out format. (There
3709 is also now support for other GNU ld extensions in ELF. Use the
3710 `--with-elf' option to configure to indicate you have ELF, and
3711 `--with-gnu-ld' if using GNU ld.) This change resulted in the deletion
3712 of many files which contained only symbol aliases, reducing the size of
3713 the source and the compiled library; many other files were renamed to
3714 less cryptic names previously occupied by the symbol alias files.
3715 There is a new header file <elf.h> for programs which operate on
3716 files in the ELF format.
3717
3718 * Converted to Autoconf version 2, so `configure' has more options.
3719 Run `configure --help' to see the details.
3720
3721 * The library can now be configured to build profiling, highly-optimized
3722 (but undebuggable), and/or shared libraries (ELF with GNU ld only). The
3723 `--enable-profile', `--enable-omitfp', and `--enable-shared' options to
3724 `configure' enable building these extra libraries. The shared library is
3725 built by default when using both ELF and GNU ld. When shared libraries
3726 are enabled, the new library `-ldl' is available for arbitrary run-time
3727 loading of shared objects; its interface is defined in <dlfcn.h>. The
3728 new header file <link.h> gives access to the internals of the run-time
3729 dynamic linker, `ld.so'. The shell script `ldd' is similar to the
3730 application of same name on other systems and it provides information
3731 about dynamically linked binaries.
3732
3733 * The C library now provides the run-time support code for profiling
3734 executables compiled with `-pg'. Programs can control the profiling code
3735 through the interface in <sys/gmon.h>. The `gmon.out' files written by
3736 the GNU C library can be read only by GNU `gprof' (from GNU binutils);
3737 the support for this file format was contributed by David Mosberger-Tang.
3738
3739 * The math code has been replaced with a math library based on fdlibm from
3740 Sun, and modified by JT Conklin and Ulrich Drepper with i387 support, by
3741 Ian Taylor with `float' functions and by Ulrich Drepper with `long double'
3742 functions. The math functions now reside in a separate library, so
3743 programs using them will need to use `-lm' their linking commands.
3744
3745 * John C. Bowman contributed optimized ix87 assembler inline functions.
3746
3747 * Ulrich Drepper has contributed support for an `/etc/nsswitch.conf'
3748 mechanism similar to that found in Solaris 2. This is now used for the
3749 group, passwd, hosts, networks, services, protocols, rpc, ethers,
3750 shadow, netgroup, publickey, and alias databases. The `nsswitch.conf'
3751 file controls what services are used for each individual database. This
3752 works by loading shared libraries with names specified in `nsswitch.conf',
3753 so service modules can be changed or added at any time without even
3754 relinking any program. Currently there are the file, db, and NIS based
3755 NSS services available.
3756
3757 * The new functions `strtoq' and `strtouq' parse integer values from
3758 strings, like `strtol' and `strtoul', but they return `long long int' and
3759 `unsigned long long int' values, respectively (64-bit quantities).
3760
3761 * The new functions `strtof' and `strtold' parse floating-point values from
3762 strings, like `strtod', but they return `float' and `long double' values,
3763 respectively (on some machines `double' and `long double' are the same).
3764
3765 * Ulrich Drepper has contributed new implementations of the floating-point
3766 printing and reading code used in the `printf' family of functions and
3767 `strtod', `strtof', and `strtold'. These new functions are perfectly
3768 accurate, and much faster than the old ones.
3769
3770 * The implementation of the POSIX locale model was completely rewritten by
3771 Ulrich Drepper. This includes the new programs `localedef' and `locale'
3772 to compile the POSIX locale definition.
3773
3774 * The former dummy implementations of the strcoll and strxfrm function are
3775 now replaced by fully functional code contributed by Ulrich Drepper. The
3776 collation information comes from the POSIX locale definitions.
3777
3778 * The new header <langinfo.h> defines an interface for accessing
3779 various locale-dependent data (using the locale chosen with `setlocale').
3780
3781 * Ulrich Drepper has contributed a new suite of functions for operation on
3782 wide-character and multibyte-character strings, in <wchar.h>;
3783 and classification and case conversion of wide characters, in <wctype.h>.
3784 These new functions are conforming to the ISO C, Amendement 1 specification.
3785
3786 * There is now a second implementation of the standard I/O library available.
3787 It comes from GNU libg++ as was written by Per Bothner, heavily modified
3788 by Hongjiu Lu and made thread safe by Ulrich Drepper.
3789
3790 * You can now use positional parameter specifications in format strings
3791 for the `printf' and `scanf' families of functions. For example,
3792 `printf ("Number %2$d, Mr %1$s\n", "Jones", 6);'' prints
3793 ``Number 6, Mr Jones''. This is mainly useful when providing different
3794 format strings for different languages, whose grammars may dictate
3795 different orderings of the values being printed. To support this
3796 feature, the interface for `register_printf_handler' has changed; see
3797 the header file <printf.h> for details.
3798
3799 * The `printf' and `scanf' families of functions now understand a new
3800 formatting flag for numeric conversions: the ' flag (e.g. %'d or %'f) says
3801 to group numbers as indicated by the locale; for `scanf' and friends, this
3802 says to accept as valid only a number with all the proper grouping
3803 separators in the right places. In the default "C" locale, numbers are
3804 not grouped; but locales for specific countries will define the usual
3805 conventions (i.e. separate thousands with `,' in the US locale).
3806
3807 * The pgrp functions have been regularized, slightly incompatibly but much
3808 less confusingly. The core functions are now `getpgid' and `setpgid',
3809 which take arguments for the PID to operate on; the POSIX.1 `getpgrp' (no
3810 argument) and BSD `setpgrp' (identical to `setpgid') functions are
3811 provided for compatibility. There is no longer an incompatible `getpgrp'
3812 with an argument declared under _BSD_SOURCE; no BSD code uses it.
3813
3814 * The new header file <fts.h> and suite of functions simplify programs that
3815 operate on directory trees. This code comes from 4.4 BSD.
3816
3817 * The resolver code has been updated from the BIND 4.9.5-P1 release.
3818 Parts of the code were heavily modified by Ulrich Drepper to fit in the
3819 NSS scheme used in glibc.
3820
3821 * There is a new malloc debugging hook `__memalign_hook'.
3822
3823 * There are new typedefs `ushort' for `unsigned short int' and `uint' for
3824 `unsigned int' in <sys/types.h>. These are for compatibility only and
3825 their use is discouraged.
3826
3827 * The `-lmcheck' library to enable standard malloc debugging hooks is now
3828 done differently, so that it works even without GNU ld.
3829
3830 * New function `euidaccess' checks allowed access to a file like `access',
3831 but using the effective IDs instead of the real IDs.
3832
3833 * The time zone data files have been updated for the latest and greatest
3834 local time conventions of the countries of the world.
3835
3836 * The new function `dirfd' extracts the file descriptor used by a DIR stream;
3837 see <dirent.h>.
3838
3839 * The new functions `ecvt', `fcvt', and `gcvt' provide an obsolete interface
3840 for formatting floating-point numbers. They are provided only for
3841 compatibility; new programs should use `sprintf' instead. There are
3842 also equivalent function for the `long double' floating-point type and
3843 all functions also exist in a reentrant form.
3844
3845 * The new auxiliary library `-lutil' from 4.4 BSD contains various
3846 functions for maintaining the login-record files (primarily of use to
3847 system programs such as `login'), and convenient functions for
3848 allocating and initializing a pseudo-terminal (pty) device.
3849
3850 * Ulrich Drepper has contributed new support for System V style
3851 shared memory and IPC on systems that support it.
3852
3853 * Ulrich Drepper has contributed several miscellaneous new functions found
3854 in System V: The `hsearch' family of functions provide an effective
3855 implementation of hash tables; `a64l' and `l64a' provide a very simple
3856 binary to ASCII mapping; `drand48' and friends provide a 48-bit random
3857 number generator.
3858
3859 * Ulrich Drepper has contributed new reentrant counterparts for the
3860 `random' and `hsearch' families of functions; `random_r', `hsearch_r', etc.
3861
3862 * Ulrich Drepper has contributed new, highly-optimized versions of several
3863 string functions for the i486/Pentium family of processors.
3864
3865 * Ulrich Drepper has updated the Linux-specific code, based largely
3866 on work done in Hongjiu Lu's version of GNU libc for Linux.
3867 The GNU library now supports Linux versions 2.0.10 and later,
3868 using the ELF object file format (i[3456]86-*-linux).
3869
3870 * Andreas Schwab has ported the C library to Linux/m68k (m68k-*-linux).
3871
3872 * David Mosberger-Tang and Richard Henderson have ported the C library
3873 to Linux/Alpha (alpha-*-linux). Richard Henderson contributed the
3874 dynamic linking support for ELF/Alpha.
3875
3876 * Richard Henderson contributed several Alpha optimized assembler function
3877 for arithmetic and string handling.
3878
3879 * Ulrich Drepper has contributed a new set of message catalog functions to
3880 support multiple languages using the <libintl.h> interface, for use with
3881 his new package GNU gettext. Translation volunteers have contributed
3882 catalogs of the library's messages in Spanish, German, and Korean.
3883
3884 * For compatibility with XPG4, Ulrich Drepper has contributed the `gencat'
3885 program and the `catgets' function for reading the catalog files it
3886 creates. (The <libintl.h> interface is preferred; we include the
3887 <nl_types.h> interface using `catgets' only for source compatibility with
3888 programs already written to use it.)
3889
3890 * New header file <values.h> gives SVID-compatible names for <limits.h>
3891 constants.
3892
3893 * Various new macros, declarations, and small header files for compatibility
3894 with 4.4 BSD.
3895
3896 * New function `group_member' is a convenient way to check if a process has
3897 a given effective group ID.
3898
3899 * When using GCC 2.7 and later, the socket functions are now declared in a
3900 special way so that passing an argument of type `struct sockaddr_in *',
3901 `struct sockaddr_ns *', or `struct sockaddr_un *' instead of the generic
3902 `struct sockaddr *' type, does not generate a type-clash warning.
3903
3904 * New function `error' declared in header file <error.h> is a convenient
3905 function for printing error messages and optionally exiting; this is the
3906 canonical function used in GNU programs. The new functions `err', `warn',
3907 and friends in header file <err.h> are the canonical 4.4 BSD interface for
3908 doing the same thing.
3909
3910 * The <glob.h> interface has several new flags from 4.4 BSD that extend the
3911 POSIX.2 `glob' function to do ~ and {...} expansion.
3912
3913 * New function `unsetenv' complements `setenv' for compatibility with 4.4 BSD.
3914 `clearenv' which is used in POSIX.9 is also available.
3915
3916 * New function `getsid' returns session ID number on systems that support it.
3917
3918 * We have incorporated the 4.4 BSD `db' library (version 1.85). New header
3919 files <db.h> and <mpool.h> provide a rich set of functions for several
3920 types of simple databases stored in memory and in files, and <ndbm.h> is
3921 an old `ndbm'-compatible interface using the `db' functions. Link with
3922 `-ldb' to get these functions.
3923
3924 * New macro `strdupa' copies a string like `strdup', but uses local stack
3925 space from `alloca' instead of dynamic heap space from `malloc'.
3926
3927 * New function `strnlen' is like `strlen' but searches only a given maximum
3928 number of characters for the null terminator. `stpncpy', `strndup' and
3929 `strndupa' are similar variants for the `stpcpy', `strdup' and `strdupa'
3930 function.
3931
3932 * New function `statfs' in header <sys/statfs.h>.
3933
3934 * The new <argz.h> and <envz.h> interfaces contributed by Miles Bader
3935 provide convenient functions for operating on blocks of null-terminated
3936 strings.
3937
3938 * A new suite of functions in <utmp.h> handle all the details of reading
3939 and writing the utmp file.
3940
3941 * An implementation of the NIS/YP(tm) based NSS service was contributed by
3942 Thorsten Kukuk.
3943
3944 * Paul Eggert and Ulrich Drepper modified the `strftime' function to be
3945 completely POSIX compliant and also implemented the extended functionality
3946 to handle alternate digit representation and alternate era date formats.
3947
3948 * Ulrich Drepper provided an implementation of the `strptime' function
3949 defined in XPG4.2 which transforms a string into a `struct tm' value.
3950
3951 * Paul Eggert provided the tzselect shell script as part of the timezone
3952 code. The shell script makes it easy to select the correct timezone
3953 specification.
3954
3955 * The implementation of the malloc family of functions is completely replaced
3956 by a new implementation by Doug Lea with many improvements by Wolfram Gloger.
3957 The implementation uses the mmap function (if available) and it is
3958 optimized for the use in multi threaded programs.
3959
3960 * Ulrich Drepper contributed a MD5 "encryption" for the crypt family of
3961 functions. This new functionality is usable by specifying a special
3962 salt string and it is compatible with implementation on *BSD systems.
3963
3964 * Lots of functions from the XPG4.2 standard were added by Ulrich Drepper:
3965 `getsubopt' to handle second level command line options, `bsd_signal'
3966 to access BSD style `signal' functionality, the obsolete `regexp' style
3967 expression matcher.
3968
3969 * the `lchown' function is available on system which support this
3970 functionality.
3971
3972 * The implementation of the shadow password handling function was contributed
3973 by Ulrich Drepper.
3974
3975 * David Mosberger-Tang changed the SunRPC implementation to be 64bit safe.
3976
3977 * POSIX.1g support was added. The <sys/select.h> header is available,
3978 `isfdtype' and `pselect' are implemented. Craig Metz contributed an
3979 implementation of `getaddrinfo'.
3980 \f
3981 Version 1.09
3982
3983 * For cross-compilation you should now set `BUILD_CC' instead of `HOST_CC'.
3984
3985 * New header file <fstab.h> and new functions `getfsspec', `getfsent' and
3986 friends, for parsing /etc/fstab. This code comes from 4.4 BSD.
3987
3988 * The new function `daemon' from 4.4 BSD is useful for server programs that
3989 want to put themselves in the background.
3990
3991 * Joel Sherrill has contributed support for several standalone boards that
3992 run without an operating system.
3993
3994 * `printf', `scanf' and friends now accept a `q' type modifier for long
3995 long int as well as `ll'. Formats using these might be `%qu' or `%lld'.
3996
3997 * All of the code taken from BSD (notably most of the math and networking
3998 routines) has been updated from the BSD 4.4-Lite release.
3999
4000 * The resolver code has been updated from the BIND-4.9.3-BETA9 release.
4001
4002 * The new functions `getdomainname' and `setdomainname' fetch or change the
4003 YP/NIS domain name. These are system calls which exist on systems which
4004 have YP (aka NIS).
4005
4006 * The time zone data files have been updated for the latest international
4007 conventions.
4008
4009 * The SunRPC programs `portmap' and `rpcinfo' are now installed in
4010 $(sbindir) (usually /usr/local/sbin) instead of $(bindir).
4011 \f
4012 Version 1.08
4013
4014 * The C library now includes support for Sun RPC, from Sun's free
4015 RPCSRC-4.0 distribution. The `portmap', `rpcinfo', and `rpcgen' programs
4016 are included. (There is still no support for YP.)
4017
4018 * Tom Quinn has contributed a port of the C library to SGI machines running
4019 Irix 4 (mips-sgi-irix4).
4020
4021 * The new `lockf' function is a simplified interface to the locking
4022 facilities of `fcntl', included for compatibility.
4023
4024 * New time functions `timegm', `timelocal', and `dysize' for compatibility.
4025
4026 * New header file <sys/timeb.h> and new function `ftime' for compatibility.
4027
4028 * New header files <poll.h> and <sys/poll.h> and new function `poll' for
4029 compatibility.
4030
4031 * The error message printed by `assert' for a failed assertion now includes
4032 the name of the program (if using GNU ld) and the name of the calling
4033 function (with versions of GCC that support this).
4034
4035 * The `psignal' function is now declared in <signal.h>, not <stdio.h>.
4036
4037 * The library now includes the <sys/mman.h> header file and memory
4038 management functions `mmap', `munmap', `mprotect', `msync', and
4039 `madvise', on systems that support those facilities.
4040
4041 * The interface for `mcheck' has changed slightly: the function called to
4042 abort the program when an allocation inconsistency is detected now takes
4043 an argument that indicates the type of failure. The new function
4044 `mprobe' lets you request a consistency check for a particular block at
4045 any time (checks are normally done only when you call `free' or `realloc'
4046 on a block).
4047
4048 * It is now possible to easily cross-compile the C library, building on one
4049 system a library to run on another machine and/or operating system. All
4050 you need to do is set the variable `HOST_CC' in `configparms' to the
4051 native compiler for programs to run on the machine you are building on (a
4052 few generator programs are used on Unix systems); set `CC' to the
4053 cross-compiler.
4054
4055 * The new function `fexecve' (only implemented on the GNU system) executes
4056 a program file given a file descriptor already open on the file.
4057 \f
4058 Version 1.07
4059
4060 * Brendan Kehoe has contributed most of a port to the DEC Alpha
4061 running OSF/1 (alpha-dec-osf1). He says it is 75% complete.
4062
4063 * You can set the variable `libprefix' in `configparms' to specify a prefix
4064 to be prepended to installed library files; this makes it easy to install
4065 the GNU C library to be linked as `-lgnuc' or whatever.
4066
4067 * The new `stpncpy' is a cross between `stpcpy' and `strncpy': It
4068 copies a limited number of characters from a string, and returns the
4069 address of the last character written.
4070
4071 * You no longer need to check for whether the installed `stddef.h' is
4072 compatible with the GNU C library. configure now checks for you.
4073
4074 * You can now define a per-stream `fileno' function to convert the
4075 stream's cookie into an integral file descriptor.
4076
4077 * ``malloc (0)'' no longer returns a null pointer. Instead, it
4078 allocates zero bytes of storage, and returns a unique pointer which
4079 you can pass to `realloc' or `free'. The behavior is undefined if
4080 you dereference this pointer.
4081
4082 * The C library now runs on Sony NEWS m68k machines running either
4083 NewsOS 3 or NewsOS 4.
4084
4085 * The new `syscall' function is a system-dependent primitive function
4086 for invoking system calls. It has the canonical behavior on Unix
4087 systems, including unreliable return values for some calls (such as
4088 `pipe', `fork' and `getppid').
4089
4090 * The error code `EWOULDBLOCK' is now obsolete; it is always defined
4091 to `EAGAIN', which is the preferred name. On systems whose kernels
4092 use two distinct codes, the C library now translates EWOULDBLOCK to
4093 EAGAIN in every system call function.
4094 \f
4095 Version 1.06
4096
4097 * The GNU C Library Reference Manual is now distributed with the library.
4098 `make dvi' will produce a DVI file of the printed manual.
4099 `make info' will produce Info files that you can read on line using C-h i
4100 in Emacs or the `info' program.
4101 Please send comments on the manual to bug-glibc-manual@gnu.org.
4102
4103 * The library now supports SVR4 on i386s (i386-unknown-sysv4).
4104
4105 * Brendan Kehoe has contributed a port to Sun SPARCs running Solaris 2.
4106
4107 * Jason Merrill has contributed a port to the Sequent Symmetry running
4108 Dynix version 3 (i386-sequent-dynix).
4109
4110 * The library has been ported to i386s running SCO 3.2.4 (also known as SCO
4111 ODT 2.0; i386-unknown-sco3.2.4) or SCO 3.2 (i386-unknown-sco3.2).
4112
4113 * New function `memory_warnings' lets you arrange to get warnings when
4114 malloc is running out of memory to allocate, like Emacs gives you.
4115
4116 * The C library now contains the relocating allocator used in Emacs 19 for
4117 its editing buffers. This allocator (ralloc) minimizes allocation
4118 overhead and fragmentation by moving allocated regions around whenever it
4119 needs to. You always refer to a ralloc'd region with a "handle" (a
4120 pointer to a pointer--an object of type `void **').
4121
4122 * There is a new `printf' format: `%m' gives you the string corresponding
4123 to the error code in `errno'.
4124
4125 * In `scanf' formats, you can now use `%as' or `%a[' to do the normal `%s'
4126 or `%[' conversion, but instead of filling in a fixed-sized buffer you
4127 pass, the `a' modifier says to fill in a `char **' you pass with a
4128 malloc'd string.
4129
4130 * The `fnmatch' function supports the new flag bits `FNM_LEADING_DIR' and
4131 `FNM_CASEFOLD'. `FNM_LEADING_DIR' lets a pattern like `foo*' match a
4132 name like `foo/bar'. `FNM_CASEFOLD' says to ignore case in matching.
4133
4134 * `mkstemp' is a traditional Unix function to atomically create and open a
4135 uniquely-named temporary file.
4136 \f
4137 Version 1.05
4138
4139 * The standard location for the file that says what the local timezone is
4140 has changed again. It is now `/usr/local/etc/localtime' (or more
4141 precisely, `${prefix}/etc/localtime') rather than `/etc/localtime'.
4142
4143 * The distribution no longer contains any files with names longer than 14
4144 characters.
4145
4146 * `struct ttyent' has two new flag bits: TTY_TRUSTED and TTY_CONSOLE.
4147 These are set by the new `trusted' and `console' keywords in `/etc/ttys'.
4148
4149 * New functions `ttyslot' and `syslog' from 4.4 BSD.
4150 \f
4151 Version 1.04
4152
4153 * The configuration process has changed quite a bit. The `configure'
4154 script is now used just like the configuration scripts for other GNU
4155 packages. The `sysdeps' directory hierarchy is much rearranged.
4156 The file `INSTALL' explains the new scheme in detail.
4157
4158 * The header files no longer need to be processed into ANSI C and
4159 traditional C versions. There is just one set of files to install, and
4160 it will work with ANSI or old C compilers (including `gcc -traditional').
4161
4162 * Brendan Kehoe and Ian Lance Taylor have ported the library to the
4163 MIPS DECStation running Ultrix 4.
4164
4165 * The Sun 4 startup code (crt0) can now properly load SunOS 4 shared libraries.
4166 Tom Quinn contributed the initial code. The GNU C library can NOT yet be
4167 made itself into a shared library.
4168
4169 * Yet further improved support for the i386, running 4.3 BSD-like systems
4170 (such as Mach 3 with the Unix single-server), or System V.
4171
4172 * New function `strncasecmp' to do case-insensitive string comparison
4173 with limited length.
4174
4175 * New function `strsep' is a reentrant alternative to `strtok'.
4176
4177 * New functions `scandir' and `alphasort' for searching directories.
4178
4179 * New function `setenv' is a better interface to `putenv'.
4180
4181 * Ian Lance Taylor has contributed an implementation of the SVID `ftw'
4182 function for traversing a directory tree.
4183
4184 * The GNU obstack package is now also part of the C library.
4185 The new function `open_obstack_stream' creates a stdio stream that
4186 writes onto an obstack; `obstack_printf' and `obstack_vprintf' do
4187 formatted output directly to an obstack.
4188
4189 * Miscellaneous new functions: reboot, nice, sigaltstack (4.4 BSD only),
4190 cfmakeraw, getusershell, getpass, swab, getttyent, seteuid, setegid.
4191
4192 * `FNM_FILE_NAME' is another name for `FNM_PATHNAME', used with `fnmatch'.
4193
4194 * The new functions `strfry' and `memfrob' do mysterious and wonderful
4195 things to your strings.
4196
4197 * There are some new test programs: test-fseek, testmb, and testrand.
4198
4199 * Some work has been done to begin porting the library to 4.4 BSD and Linux.
4200 These ports are not finished, but are a good starting place for really
4201 supporting those systems.
4202
4203 * `/etc/localtime' is now the standard location for the file that says what
4204 the local timezone is, rather than `/usr/local/lib/zoneinfo/localtime'.
4205 This follows the general principle that `/etc' is the place for all local
4206 configuration files.
4207
4208 * The C library header files now use `extern "C"' when used by the C++
4209 compiler, so the C library should now work with C++ code.
4210
4211 * The header file <bstring.h> is gone. <string.h> now declares bcopy,
4212 bcmp, bzero, and ffs. (Update: nowadays these functions are declared
4213 in <strings.h>.)
4214
4215 * Mike Haertel (of GNU e?grep and malloc fame) has written a new sorting
4216 function which uses the `merge sort' algorithm, and is said to be
4217 significantly faster than the old GNU `qsort' function. Merge sort is
4218 now the standard `qsort' function. The new algorithm can require a lot
4219 of temporary storage; so, the old sorting function is called when the
4220 required storage is not available.
4221
4222 * The C library now includes Michael Glad's Ultra Fast Crypt, which
4223 provides the Unix `crypt' function, plus some other entry points.
4224
4225 * The code and header files taken from 4.4 BSD have been updated with the
4226 latest files released from Berkeley.
4227 \f
4228 ----------------------------------------------------------------------
4229 Copyright information:
4230
4231 Copyright (C) 1992-2017 Free Software Foundation, Inc.
4232
4233 Permission is granted to anyone to make or distribute verbatim copies
4234 of this document as received, in any medium, provided that the
4235 copyright notice and this permission notice are preserved,
4236 thus giving the recipient permission to redistribute in turn.
4237
4238 Permission is granted to distribute modified versions
4239 of this document, or of portions of it,
4240 under the above conditions, provided also that they
4241 carry prominent notices stating who last changed them.
4242 \f
4243 Local variables:
4244 version-control: never
4245 fill-column: 76
4246 End: