]> git.ipfire.org Git - thirdparty/glibc.git/blob - NEWS
CVE-2018-19591: if_nametoindex: Fix descriptor for overlong name [BZ #23927]
[thirdparty/glibc.git] / NEWS
1 GNU C Library NEWS -- history of user-visible changes.
2 Copyright (C) 1992-2018 Free Software Foundation, Inc.
3 See the end for copying conditions.
4
5 Please send GNU C library bug reports via <https://sourceware.org/bugzilla/>
6 using `glibc' in the "product" field.
7 \f
8 Version 2.29
9
10 Major new features:
11
12 * A new convenience target has been added for distribution maintainers
13 to build and install all locales as directories with files. The new
14 target is run by issuing the following command in your build tree:
15 'make localedata/install-locale-files', with an optional DESTDIR
16 to set the install root if you wish to install into a non-default
17 configured location.
18
19 * Optimized generic exp, exp2, log, log2, pow, sinf, cosf, sincosf and tanf.
20
21 * The reallocarray function is now declared under _DEFAULT_SOURCE, not just
22 for _GNU_SOURCE, to match BSD environments.
23
24 * For powercp64le ABI, Transactional Lock Elision is now enabled iff kernel
25 indicates that it will abort the transaction prior to entering the kernel
26 (PPC_FEATURE2_HTM_NOSC on hwcap2). On older kernels the transaction is
27 suspended, and this caused some undefined side-effects issues by aborting
28 transactions manually. Glibc avoided it by abort transactions manually on
29 each syscall, but it lead to performance issues on newer kernels where the
30 HTM state is saved and restore lazily (the state being saved even when the
31 process actually does not use HTM).
32
33 * The posix_spawn_file_actions_addchdir_np function has been added,
34 enabling posix_spawn and posix_spawnp to run the new process in a
35 different directory. This is a GNU extension and similar to the
36 Solaris function of the same name.
37
38 Deprecated and removed features, and other changes affecting compatibility:
39
40 * The glibc.tune tunable namespace has been renamed to glibc.cpu and the
41 tunable glibc.tune.cpu has been renamed to glibc.cpu.name.
42
43 * The type of the pr_uid and pr_gid members of struct elf_prpsinfo, defined
44 in <sys/procfs.h>, has been corrected to match the type actually used by
45 the Linux kernel. This affects the size and layout of that structure on
46 MicroBlaze, MIPS (n64 ABI only), Nios II and RISC-V.
47
48 * For the MIPS n32 ABI, the type of the pr_sigpend and pr_sighold members of
49 struct elf_prstatus, and the pr_flag member of struct elf_prpsinfo,
50 defined in <sys/procfs.h>, has been corrected to match the type actually
51 used by the Linux kernel. This affects the size and layout of those
52 structures.
53
54 Changes to build and runtime requirements:
55
56 * Python 3.4 or later is required to build the GNU C Library.
57
58 Security related changes:
59
60 CVE-2018-19591: A file descriptor leak in if_nametoindex can lead to a
61 denial of service due to resource exhaustion when processing getaddrinfo
62 calls with crafted host names. Reported by Guido Vranken.
63
64 The following bugs are resolved with this release:
65
66 [The release manager will add the list generated by
67 scripts/list-fixed-bugs.py just before the release.]
68
69 \f
70 Version 2.28
71
72 Major new features:
73
74 * The localization data for ISO 14651 is updated to match the 2016
75 Edition 4 release of the standard, this matches data provided by
76 Unicode 9.0.0. This update introduces significant improvements to the
77 collation of Unicode characters. This release deviates slightly from
78 the standard in that the collation element ordering for lowercase and
79 uppercase LATIN script characters is adjusted to ensure that regular
80 expressions with ranges like [a-z] and [A-Z] don't interleave e.g. A
81 is not matched by [a-z]. With the update many locales have been
82 updated to take advantage of the new collation information. The new
83 collation information has increased the size of the compiled locale
84 archive or binary locales.
85
86 * The GNU C Library can now be compiled with support for Intel CET, AKA
87 Intel Control-flow Enforcement Technology. When the library is built
88 with --enable-cet, the resulting glibc is protected with indirect
89 branch tracking (IBT) and shadow stack (SHSTK). CET-enabled glibc is
90 compatible with all existing executables and shared libraries. This
91 feature is currently supported on i386, x86_64 and x32 with GCC 8 and
92 binutils 2.29 or later. Note that CET-enabled glibc requires CPUs
93 capable of multi-byte NOPs, like x86-64 processors as well as Intel
94 Pentium Pro or newer. NOTE: --enable-cet has been tested for i686,
95 x86_64 and x32 on non-CET processors. --enable-cet has been tested
96 for x86_64 and x32 on CET SDVs, but Intel CET support hasn't been
97 validated for i686.
98
99 * The GNU C Library now has correct support for ABSOLUTE symbols
100 (SHN_ABS-relative symbols). Previously such ABSOLUTE symbols were
101 relocated incorrectly or in some cases discarded. The GNU linker can
102 make use of the newer semantics, but it must communicate it to the
103 dynamic loader by setting the ELF file's identification (EI_ABIVERSION
104 field) to indicate such support is required.
105
106 * Unicode 11.0.0 Support: Character encoding, character type info, and
107 transliteration tables are all updated to Unicode 11.0.0, using
108 generator scripts contributed by Mike FABIAN (Red Hat).
109
110 * <math.h> functions that round their results to a narrower type are added
111 from TS 18661-1:2014 and TS 18661-3:2015:
112
113 - fadd, faddl, daddl and corresponding fMaddfN, fMaddfNx, fMxaddfN and
114 fMxaddfNx functions.
115
116 - fsub, fsubl, dsubl and corresponding fMsubfN, fMsubfNx, fMxsubfN and
117 fMxsubfNx functions.
118
119 - fmul, fmull, dmull and corresponding fMmulfN, fMmulfNx, fMxmulfN and
120 fMxmulfNx functions.
121
122 - fdiv, fdivl, ddivl and corresponding fMdivfN, fMdivfNx, fMxdivfN and
123 fMxdivfNx functions.
124
125 * Two grammatical forms of month names are now supported for the following
126 languages: Armenian, Asturian, Catalan, Czech, Kashubian, Occitan, Ossetian,
127 Scottish Gaelic, Upper Sorbian, and Walloon. The following languages now
128 support two grammatical forms in abbreviated month names: Catalan, Greek,
129 and Kashubian.
130
131 * Newly added locales: Lower Sorbian (dsb_DE) and Yakut (sah_RU) also
132 include the support for two grammatical forms of month names.
133
134 * Building and running on GNU/Hurd systems now works without out-of-tree
135 patches.
136
137 * The renameat2 function has been added, a variant of the renameat function
138 which has a flags argument. If the flags are zero, the renameat2 function
139 acts like renameat. If the flag is not zero and there is no kernel
140 support for renameat2, the function will fail with an errno value of
141 EINVAL. This is different from the existing gnulib function renameatu,
142 which performs a plain rename operation in case of a RENAME_NOREPLACE
143 flags and a non-existing destination (and therefore has a race condition
144 that can clobber the destination inadvertently).
145
146 * The statx function has been added, a variant of the fstatat64
147 function with an additional flags argument. If there is no direct
148 kernel support for statx, glibc provides basic stat support based on
149 the fstatat64 function.
150
151 * IDN domain names in getaddrinfo and getnameinfo now use the system libidn2
152 library if installed. libidn2 version 2.0.5 or later is recommended. If
153 libidn2 is not available, internationalized domain names are not encoded
154 or decoded even if the AI_IDN or NI_IDN flags are passed to getaddrinfo or
155 getnameinfo. (getaddrinfo calls with non-ASCII names and AI_IDN will fail
156 with an encoding error.) Flags which used to change the IDN encoding and
157 decoding behavior (AI_IDN_ALLOW_UNASSIGNED, AI_IDN_USE_STD3_ASCII_RULES,
158 NI_IDN_ALLOW_UNASSIGNED, NI_IDN_USE_STD3_ASCII_RULES) have been
159 deprecated. They no longer have any effect.
160
161 * Parsing of dynamic string tokens in DT_RPATH, DT_RUNPATH, DT_NEEDED,
162 DT_AUXILIARY, and DT_FILTER has been expanded to support the full
163 range of ELF gABI expressions including such constructs as
164 '$ORIGIN$ORIGIN' (if valid). For SUID/GUID applications the rules
165 have been further restricted, and where in the past a dynamic string
166 token sequence may have been interpreted as a literal string it will
167 now cause a load failure. These load failures were always considered
168 unspecified behaviour from the perspective of the dynamic loader, and
169 for safety are now load errors e.g. /foo/${ORIGIN}.so in DT_NEEDED
170 results in a load failure now.
171
172 * Support for ISO C threads (ISO/IEC 9899:2011) has been added. The
173 implementation includes all the standard functions provided by
174 <threads.h>:
175
176 - thrd_current, thrd_equal, thrd_sleep, thrd_yield, thrd_create,
177 thrd_detach, thrd_exit, and thrd_join for thread management.
178
179 - mtx_init, mtx_lock, mtx_timedlock, mtx_trylock, mtx_unlock, and
180 mtx_destroy for mutual exclusion.
181
182 - call_once for function call synchronization.
183
184 - cnd_broadcast, cnd_destroy, cnd_init, cnd_signal, cnd_timedwait, and
185 cnd_wait for conditional variables.
186
187 - tss_create, tss_delete, tss_get, and tss_set for thread-local storage.
188
189 Application developers must link against libpthread to use ISO C threads.
190
191 Deprecated and removed features, and other changes affecting compatibility:
192
193 * The nonstandard header files <libio.h> and <_G_config.h> are no longer
194 installed. Software that was using either header should be updated to
195 use standard <stdio.h> interfaces instead.
196
197 * The stdio functions 'getc' and 'putc' are no longer defined as macros.
198 This was never required by the C standard, and the macros just expanded
199 to call alternative names for the same functions. If you hoped getc and
200 putc would provide performance improvements over fgetc and fputc, instead
201 investigate using (f)getc_unlocked and (f)putc_unlocked, and, if
202 necessary, flockfile and funlockfile.
203
204 * All stdio functions now treat end-of-file as a sticky condition. If you
205 read from a file until EOF, and then the file is enlarged by another
206 process, you must call clearerr or another function with the same effect
207 (e.g. fseek, rewind) before you can read the additional data. This
208 corrects a longstanding C99 conformance bug. It is most likely to affect
209 programs that use stdio to read interactive input from a terminal.
210 (Bug #1190.)
211
212 * The macros 'major', 'minor', and 'makedev' are now only available from
213 the header <sys/sysmacros.h>; not from <sys/types.h> or various other
214 headers that happen to include <sys/types.h>. These macros are rarely
215 used, not part of POSIX nor XSI, and their names frequently collide with
216 user code; see https://sourceware.org/bugzilla/show_bug.cgi?id=19239 for
217 further explanation.
218
219 <sys/sysmacros.h> is a GNU extension. Portable programs that require
220 these macros should first include <sys/types.h>, and then include
221 <sys/sysmacros.h> if __GNU_LIBRARY__ is defined.
222
223 * The tilegx*-*-linux-gnu configurations are no longer supported.
224
225 * The obsolete function ustat is no longer available to newly linked
226 binaries; the headers <ustat.h> and <sys/ustat.h> have been removed. This
227 function has been deprecated in favor of fstatfs and statfs.
228
229 * The obsolete function nfsservctl is no longer available to newly linked
230 binaries. This function was specific to systems using the Linux kernel
231 and could not usefully be used with the GNU C Library on systems with
232 version 3.1 or later of the Linux kernel.
233
234 * The obsolete function name llseek is no longer available to newly linked
235 binaries. This function was specific to systems using the Linux kernel
236 and was not declared in a header. Programs should use the lseek64 name
237 for this function instead.
238
239 * The AI_IDN_ALLOW_UNASSIGNED and NI_IDN_ALLOW_UNASSIGNED flags for the
240 getaddrinfo and getnameinfo functions have been deprecated. The behavior
241 previously selected by them is now always enabled.
242
243 * The AI_IDN_USE_STD3_ASCII_RULES and NI_IDN_USE_STD3_ASCII_RULES flags for
244 the getaddrinfo and getnameinfo functions have been deprecated. The STD3
245 restriction (rejecting '_' in host names, among other things) has been
246 removed, for increased compatibility with non-IDN name resolution.
247
248 * The fcntl function now have a Long File Support variant named fcntl64. It
249 is added to fix some Linux Open File Description (OFD) locks usage on non
250 LFS mode. As for others *64 functions, fcntl64 semantics are analogous with
251 fcntl and LFS support is handled transparently. Also for Linux, the OFD
252 locks act as a cancellation entrypoint.
253
254 * The obsolete functions encrypt, encrypt_r, setkey, setkey_r, cbc_crypt,
255 ecb_crypt, and des_setparity are no longer available to newly linked
256 binaries, and the headers <rpc/des_crypt.h> and <rpc/rpc_des.h> are no
257 longer installed. These functions encrypted and decrypted data with the
258 DES block cipher, which is no longer considered secure. Software that
259 still uses these functions should switch to a modern cryptography library,
260 such as libgcrypt.
261
262 * Reflecting the removal of the encrypt and setkey functions above, the
263 macro _XOPEN_CRYPT is no longer defined. As a consequence, the crypt
264 function is no longer declared unless _DEFAULT_SOURCE or _GNU_SOURCE is
265 enabled.
266
267 * The obsolete function fcrypt is no longer available to newly linked
268 binaries. It was just another name for the standard function crypt,
269 and it has not appeared in any header file in many years.
270
271 * We have tentative plans to hand off maintenance of the passphrase-hashing
272 library, libcrypt, to a separate development project that will, we hope,
273 keep up better with new passphrase-hashing algorithms. We will continue
274 to declare 'crypt' in <unistd.h>, and programs that use 'crypt' or
275 'crypt_r' should not need to change at all; however, distributions will
276 need to install <crypt.h> and libcrypt from a separate project.
277
278 In this release, if the configure option --disable-crypt is used, glibc
279 will not install <crypt.h> or libcrypt, making room for the separate
280 project's versions of these files. The plan is to make this the default
281 behavior in a future release.
282
283 Changes to build and runtime requirements:
284
285 GNU make 4.0 or later is now required to build glibc.
286
287 Security related changes:
288
289 CVE-2016-6261, CVE-2016-6263, CVE-2017-14062: Various vulnerabilities have
290 been fixed by removing the glibc-internal IDNA implementation and using
291 the system-provided libidn2 library instead. Originally reported by Hanno
292 Böck and Christian Weisgerber.
293
294 CVE-2017-18269: An SSE2-based memmove implementation for the i386
295 architecture could corrupt memory. Reported by Max Horn.
296
297 CVE-2018-11236: Very long pathname arguments to realpath function could
298 result in an integer overflow and buffer overflow. Reported by Alexey
299 Izbyshev.
300
301 CVE-2018-11237: The mempcpy implementation for the Intel Xeon Phi
302 architecture could write beyond the target buffer, resulting in a buffer
303 overflow. Reported by Andreas Schwab.
304
305 The following bugs are resolved with this release:
306
307 [1190] stdio: fgetc()/fread() behaviour is not POSIX compliant
308 [6889] manual: 'PWD' mentioned but not specified
309 [13575] libc: SSIZE_MAX defined as LONG_MAX is inconsistent with ssize_t,
310 when __WORDSIZE != 64
311 [13762] regex: re_search etc. should return -2 on memory exhaustion
312 [13888] build: /tmp usage during testing
313 [13932] math: dbl-64 pow unexpectedly slow for some inputs
314 [14092] nptl: Support C11 threads
315 [14095] localedata: Review / update collation data from Unicode / ISO
316 14651
317 [14508] libc: -Wformat warnings
318 [14553] libc: Namespace pollution loff_t in sys/types.h
319 [14890] libc: Make NT_PRFPREG canonical.
320 [15105] libc: Extra PLT references with -Os
321 [15512] libc: __bswap_constant_16 not compiled when -Werror -Wsign-
322 conversion is given
323 [16335] manual: Feature test macro documentation incomplete and out of
324 date
325 [16552] libc: Unify umount implementations in terms of umount2
326 [17082] libc: htons et al.: statement-expressions prevent use on global
327 scope with -O1 and higher
328 [17343] libc: Signed integer overflow in /stdlib/random_r.c
329 [17438] localedata: pt_BR: wrong d_fmt delimiter
330 [17662] libc: please implement binding for the new renameat2 syscall
331 [17721] libc: __restrict defined as /* Ignore */ even in c11
332 [17979] libc: inconsistency between uchar.h and stdint.h
333 [18018] dynamic-link: Additional $ORIGIN handling issues (CVE-2011-0536)
334 [18023] libc: extend_alloca is broken (questionable pointer comparison,
335 horrible machine code)
336 [18124] libc: hppa: setcontext erroneously returns -1 as exit code for
337 last constant.
338 [18471] libc: llseek should be a compat symbol
339 [18473] soft-fp: [powerpc-nofpu] __sqrtsf2, __sqrtdf2 should be compat
340 symbols
341 [18991] nss: nss_files skips large entry in database
342 [19239] libc: Including stdlib.h ends up with macros major and minor being
343 defined
344 [19463] libc: linknamespace failures when compiled with -Os
345 [19485] localedata: csb_PL: Update month translations + add yesstr/nostr
346 [19527] locale: Normalized charset name not recognized by setlocale
347 [19667] string: Missing Sanity Check for malloc calls in file 'testcopy.c'
348 [19668] libc: Missing Sanity Check for malloc() in file 'tst-setcontext-
349 fpscr.c'
350 [19728] network: out of bounds stack read in libidn function
351 idna_to_ascii_4i (CVE-2016-6261)
352 [19729] network: out of bounds heap read on invalid utf-8 inputs in
353 stringprep_utf8_nfkc_normalize (CVE-2016-6263)
354 [19818] dynamic-link: Absolute (SHN_ABS) symbols incorrectly relocated by
355 the base address
356 [20079] libc: Add SHT_X86_64_UNWIND to elf.h
357 [20251] libc: 32bit programs pass garbage in struct flock for OFD locks
358 [20419] dynamic-link: files with large allocated notes crash in
359 open_verify
360 [20530] libc: bswap_16 should use __builtin_bswap16() when available
361 [20890] dynamic-link: ldconfig: fsync the files before atomic rename
362 [20980] manual: CFLAGS environment variable replaces vital options
363 [21163] regex: Assertion failure in pop_fail_stack when executing a
364 malformed regexp (CVE-2015-8985)
365 [21234] manual: use of CFLAGS makes glibc detect no optimization
366 [21269] dynamic-link: i386 sigaction sa_restorer handling is wrong
367 [21313] build: Compile Error GCC 5.4.0 MIPS with -0S
368 [21314] build: Compile Error GCC 5.2.0 MIPS with -0s
369 [21508] locale: intl/tst-gettext failure with latest msgfmt
370 [21547] localedata: Tibetan script collation broken (Dzongkha and Tibetan)
371 [21812] network: getifaddrs() returns entries with ifa_name == NULL
372 [21895] libc: ppc64 setjmp/longjmp not fully interoperable with static
373 dlopen
374 [21942] dynamic-link: _dl_dst_substitute incorrectly handles $ORIGIN: with
375 AT_SECURE=1
376 [22241] localedata: New locale: Yakut (Sakha) locale for Russia (sah_RU)
377 [22247] network: Integer overflow in the decode_digit function in
378 puny_decode.c in libidn (CVE-2017-14062)
379 [22342] nscd: NSCD not properly caching netgroup
380 [22391] nptl: Signal function clear NPTL internal symbols inconsistently
381 [22550] localedata: es_ES locale (and other es_* locales): collation
382 should treat ñ as a primary different character, sync the collation
383 for Spanish with CLDR
384 [22638] dynamic-link: sparc: static binaries are broken if glibc is built
385 by gcc configured with --enable-default-pie
386 [22639] time: year 2039 bug for localtime etc. on 64-bit platforms
387 [22644] string: memmove-sse2-unaligned on 32bit x86 produces garbage when
388 crossing 2GB threshold (CVE-2017-18269)
389 [22646] localedata: redundant data (LC_TIME) for es_CL, es_CU, es_EC and
390 es_BO
391 [22735] time: Misleading typo in time.h source comment regarding
392 CLOCKS_PER_SECOND
393 [22753] libc: preadv2/pwritev2 fallback code should handle offset=-1
394 [22761] libc: No trailing `%n' conversion specifier in FMT passed from
395 `__assert_perror_fail ()' to `__assert_fail_base ()'
396 [22766] libc: all glibc internal dlopen should use RTLD_NOW for robust
397 dlopen failures
398 [22786] libc: Stack buffer overflow in realpath() if input size is close
399 to SSIZE_MAX (CVE-2018-11236)
400 [22787] dynamic-link: _dl_check_caller returns false when libc is linked
401 through an absolute DT_NEEDED path
402 [22792] build: tcb-offsets.h dependency dropped
403 [22797] libc: pkey_get() uses non-reserved name of argument
404 [22807] libc: PTRACE_* constants missing for powerpc
405 [22818] glob: posix/tst-glob_lstat_compat failure on alpha
406 [22827] dynamic-link: RISC-V ELF64 parser mis-reads flag in ldconfig
407 [22830] malloc: malloc_stats doesn't restore cancellation state on stderr
408 [22848] localedata: ca_ES: update date definitions from CLDR
409 [22862] build: _DEFAULT_SOURCE is defined even when _ISOC11_SOURCE is
410 [22884] math: RISCV fmax/fmin handle signalling NANs incorrectly
411 [22896] localedata: Update locale data for an_ES
412 [22902] math: float128 test failures with GCC 8
413 [22918] libc: multiple common of `__nss_shadow_database'
414 [22919] libc: sparc32: backtrace yields infinite backtrace with
415 makecontext
416 [22926] libc: FTBFS on powerpcspe
417 [22932] localedata: lt_LT: Update of abbreviated month names from CLDR
418 required
419 [22937] localedata: Greek (el_GR, el_CY) locales actually need ab_alt_mon
420 [22947] libc: FAIL: misc/tst-preadvwritev2
421 [22963] localedata: cs_CZ: Add alternative month names
422 [22987] math: [powerpc/sparc] fdim inlines errno, exceptions handling
423 [22996] localedata: change LC_PAPER to en_US in es_BO locale
424 [22998] dynamic-link: execstack tests are disabled when SELinux is
425 disabled
426 [23005] network: Crash in __res_context_send after memory allocation
427 failure
428 [23007] math: strtod cannot handle -nan
429 [23024] nss: getlogin_r is performing NSS lookups when loginid isn't set
430 [23036] regex: regex equivalence class regression
431 [23037] libc: initialize msg_flags to zero for sendmmsg() calls
432 [23069] libc: sigaction broken on riscv64-linux-gnu
433 [23094] localedata: hr_HR: wrong thousands_sep and mon_thousands_sep
434 [23102] dynamic-link: Incorrect parsing of multiple consecutive $variable
435 patterns in runpath entries (e.g. $ORIGIN$ORIGIN)
436 [23137] nptl: s390: pthread_join sometimes block indefinitely (on 31bit
437 and libc build with -Os)
438 [23140] localedata: More languages need two forms of month names
439 [23145] libc: _init/_fini aren't marked as hidden
440 [23152] localedata: gd_GB: Fix typo in "May" (abbreviated)
441 [23171] math: C++ iseqsig for long double converts arguments to double
442 [23178] nscd: sudo will fail when it is run in concurrent with commands
443 that changes /etc/passwd
444 [23196] string: __mempcpy_avx512_no_vzeroupper mishandles large copies
445 (CVE-2018-11237)
446 [23206] dynamic-link: static-pie + dlopen breaks debugger interaction
447 [23208] localedata: New locale - Lower Sorbian (dsb)
448 [23233] regex: Memory leak in build_charclass_op function in file
449 posix/regcomp.c
450 [23236] stdio: Harden function pointers in _IO_str_fields
451 [23250] nptl: Offset of __private_ss differs from GCC
452 [23253] math: tgamma test suite failures on i686 with -march=x86-64
453 -mtune=generic -mfpmath=sse
454 [23259] dynamic-link: Unsubstituted ${ORIGIN} remains in DT_NEEDED for
455 AT_SECURE
456 [23264] libc: posix_spawnp wrongly executes ENOEXEC in non compat mode
457 [23266] nis: stringop-truncation warning with new gcc8.1 in nisplus-
458 parser.c
459 [23272] math: fma(INFINITY,INFIITY,0.0) should be INFINITY
460 [23277] math: nan function should not have const attribute
461 [23279] math: scanf and strtod wrong for some hex floating-point
462 [23280] math: wscanf rounds wrong; wcstod is ok for negative numbers and
463 directed rounding
464 [23290] localedata: IBM273 is not equivalent to ISO-8859-1
465 [23303] build: undefined reference to symbol
466 '__parse_hwcap_and_convert_at_platform@@GLIBC_2.23'
467 [23307] dynamic-link: Absolute symbols whose value is zero ignored in
468 lookup
469 [23313] stdio: libio vtables validation and standard file object
470 interposition
471 [23329] libc: The __libc_freeres infrastructure is not properly run across
472 DSO boundaries.
473 [23349] libc: Various glibc headers no longer compatible with
474 <linux/time.h>
475 [23351] malloc: Remove unused code related to heap dumps and malloc
476 checking
477 [23363] stdio: stdio-common/tst-printf.c has non-free license
478 [23396] regex: Regex equivalence regression in single-byte locales
479 [23422] localedata: oc_FR: More updates of locale data
480 [23442] build: New warning with GCC 8
481 [23448] libc: Out of bounds access in IBM-1390 converter
482 [23456] libc: Wrong index_cpu_LZCNT
483 [23458] build: tst-get-cpu-features-static isn't added to tests
484 [23459] libc: COMMON_CPUID_INDEX_80000001 isn't populated for Intel
485 processors
486 [23467] dynamic-link: x86/CET: A property note parser bug
487
488 \f
489 Version 2.27
490
491 Major new features:
492
493 * The GNU C Library can now be compiled with support for building static
494 PIE executables (See --enable-static-pie in INSTALL). These static PIE
495 executables are like static executables but can be loaded at any address
496 and provide additional security hardening benefits at the cost of some
497 memory and performance. When the library is built with --enable-static-pie
498 the resulting libc.a is usable with GCC 8 and above to create static PIE
499 executables using the GCC option '-static-pie'. This feature is currently
500 supported on i386, x86_64 and x32 with binutils 2.29 or later, and on
501 aarch64 with binutils 2.30 or later.
502
503 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf,
504 sinf, sincosf and tan with FMA, contributed by Arjan van de Ven and
505 H.J. Lu from Intel.
506
507 * Optimized x86-64 trunc and truncf for processors with SSE4.1.
508
509 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf.
510
511 * In order to support faster and safer process termination the malloc API
512 family of functions will no longer print a failure address and stack
513 backtrace after detecting heap corruption. The goal is to minimize the
514 amount of work done after corruption is detected and to avoid potential
515 security issues in continued process execution. Reducing shutdown time
516 leads to lower overall process restart latency, so there is benefit both
517 from a security and performance perspective.
518
519 * The abort function terminates the process immediately, without flushing
520 stdio streams. Previous glibc versions used to flush streams, resulting
521 in deadlocks and further data corruption. This change also affects
522 process aborts as the result of assertion failures.
523
524 * On platforms where long double has the IEEE binary128 format (aarch64,
525 alpha, mips64, riscv, s390 and sparc), the math library now implements
526 _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015.
527 These are the same interfaces added in version 2.26 for some platforms where
528 this format is supported but is not the format of long double.
529
530 * On platforms with support for _Float64x (aarch64, alpha, i386, ia64,
531 mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now
532 implements interfaces for that type, as defined by ISO/IEC TS
533 18661-3:2015. These are corresponding interfaces to those supported for
534 _Float128.
535
536 * The math library now implements interfaces for the _Float32, _Float64 and
537 _Float32x types, as defined by ISO/IEC TS 18661-3:2015. These are
538 corresponding interfaces to those supported for _Float128.
539
540 * glibc now implements the memfd_create and mlock2 functions on Linux.
541
542 * Support for memory protection keys was added. The <sys/mman.h> header now
543 declares the functions pkey_alloc, pkey_free, pkey_mprotect, pkey_set,
544 pkey_get.
545
546 * The copy_file_range function was added.
547
548 * Optimized memcpy, mempcpy, memmove, and memset for sparc M7.
549
550 * The ldconfig utility now processes `include' directives using the C/POSIX
551 collation ordering. Previous glibc versions used locale-specific
552 ordering, the change might break systems that relied on that.
553
554 * Support for two grammatical forms of month names has been added.
555 In a call to strftime, the "%B" and "%b" format specifiers will now
556 produce the grammatical form required when the month is used as part
557 of a complete date. New "%OB" and "%Ob" specifiers produce the form
558 required when the month is named by itself. For instance, in Greek
559 and in many Slavic and Baltic languages, "%B" will produce the month
560 in genitive case, and "%OB" will produce the month in nominative case.
561
562 In a call to strptime, "%B", "%b", "%h", "%OB", "%Ob", and "%Oh"
563 are all valid and will all accept any known form of month
564 name---standalone or complete, abbreviated or full. In a call to
565 nl_langinfo, the query constants MON_1..12 and ABMON_1..12 return
566 the strings used by "%B" and "%b", respectively. New query
567 constants ALTMON_1..12 and _NL_ABALTMON_1..12 return the strings
568 used by "%OB" and "%Ob", respectively.
569
570 In a locale definition file, use "alt_mon" and "ab_alt_mon" to
571 define the strings for %OB and %Ob, respectively; these have the
572 same syntax as "mon" and "abmon". These arrays are optional; if they
573 are not provided then they have the same content as "mon" and "abmon",
574 respectively.
575
576 These features are provided for locales which define "alt_mon" and/or
577 "ab_alt_mon" in their locale source data. This release includes such
578 alternative month name data for the following languages: Belarusian,
579 Croatian, Greek, Lithuanian, Polish, Russian, and Ukrainian.
580
581 This feature is currently a GNU extension, but it is expected to
582 be added to the next revision of POSIX, and it is also already
583 available on some BSD-derived operating systems.
584
585 This feature will cause existing statically compiled applications
586 to fail to load locales and fall back to the builtin C/POSIX locales.
587 See notes below for other changes affecting compatibility.
588
589 * Support for the RISC-V ISA running on Linux has been added. This port
590 requires at least binutils-2.30, gcc-7.3.0, and linux-4.15; and is supported
591 for the following ISA and ABI pairs:
592
593 - rv64imac lp64
594 - rv64imafdc lp64
595 - rv64imafdc lp64d
596
597 Deprecated and removed features, and other changes affecting compatibility:
598
599 * Statically compiled applications attempting to load locales compiled for the
600 GNU C Library version 2.27 will fail and fall back to the builtin C/POSIX
601 locale. The reason for this is that the addition of the new "%OB" and "%Ob",
602 support for two grammatical forms of the month names, also extends the locale
603 data binary format. Static applications needing locale support must be
604 recompiled to match the runtime and data they are deployed with. In some
605 distributions there is an upgrade window where dynamically linked applications
606 may use a new library but the old locale data and also fall back to the
607 builtin C/POSIX locales; restarting the application process is sufficient to
608 fix this.
609
610 * Support for statically linked applications which call dlopen is deprecated
611 and will be removed in a future version of glibc. Applications which call
612 dlopen need to be linked dynamically instead.
613
614 * Support for old programs which use internal stdio data structures and
615 functions is deprecated. This includes programs which use the C++ streams
616 provided by libstdc++ in GCC 2.95. Programs which use the internal
617 symbols _IO_adjust_wcolumn, _IO_default_doallocate, _IO_default_finish,
618 _IO_default_pbackfail, _IO_default_uflow, _IO_default_xsgetn,
619 _IO_default_xsputn, _IO_doallocbuf, _IO_do_write, _IO_file_attach,
620 _IO_file_close, _IO_file_close_it, _IO_file_doallocate, _IO_file_fopen,
621 _IO_file_init, _IO_file_jumps, _IO_fileno, _IO_file_open,
622 _IO_file_overflow, _IO_file_read, _IO_file_seek, _IO_file_seekoff,
623 _IO_file_setbuf, _IO_file_stat, _IO_file_sync, _IO_file_underflow,
624 _IO_file_write, _IO_file_xsputn, _IO_flockfile, _IO_flush_all,
625 _IO_flush_all_linebuffered, _IO_free_backup_area, _IO_free_wbackup_area,
626 _IO_init, _IO_init_marker, _IO_init_wmarker, _IO_iter_begin, _IO_iter_end,
627 _IO_iter_file, _IO_iter_next, _IO_least_wmarker, _IO_link_in,
628 _IO_list_all, _IO_list_lock, _IO_list_resetlock, _IO_list_unlock,
629 _IO_marker_delta, _IO_marker_difference, _IO_remove_marker, _IO_seekmark,
630 _IO_seekwmark, _IO_str_init_readonly, _IO_str_init_static,
631 _IO_str_overflow, _IO_str_pbackfail, _IO_str_seekoff, _IO_str_underflow,
632 _IO_switch_to_main_wget_area, _IO_switch_to_wget_mode,
633 _IO_unsave_wmarkers, _IO_wdefault_doallocate, _IO_wdefault_finish,
634 _IO_wdefault_pbackfail, _IO_wdefault_setbuf, _IO_wdefault_uflow,
635 _IO_wdefault_xsgetn, _IO_wdefault_xsputn, _IO_wdoallocbuf, _IO_wdo_write,
636 _IO_wfile_jumps, _IO_wfile_overflow, _IO_wfile_sync, _IO_wfile_underflow,
637 _IO_wfile_xsputn, _IO_wmarker_delta, or _IO_wsetb may stop working with a
638 future version of glibc. Unlike other symbol removals, these old
639 applications will not be supported using compatibility symbols.
640
641 * On GNU/Linux, the obsolete Linux constant PTRACE_SEIZE_DEVEL is no longer
642 defined by <sys/ptrace.h>.
643
644 * libm no longer supports SVID error handling (calling a user-provided
645 matherr function on error) or the _LIB_VERSION variable to control error
646 handling. (SVID error handling and the _LIB_VERSION variable still work
647 for binaries linked against older versions of the GNU C Library.) The
648 libieee.a library is no longer provided. math.h no longer defines struct
649 exception, or the macros X_TLOSS, DOMAIN, SING, OVERFLOW, UNDERFLOW,
650 TLOSS, PLOSS and HUGE.
651
652 * The libm functions pow10, pow10f and pow10l are no longer supported for
653 new programs. Programs should use the standard names exp10, exp10f and
654 exp10l for these functions instead.
655
656 * The mcontext_t type is no longer the same as struct sigcontext. On
657 platforms where it was previously the same, this changes the C++ name
658 mangling for interfaces involving this type.
659
660 * The add-ons mechanism for building additional packages at the same time as
661 glibc has been removed. The --enable-add-ons configure option is now
662 ignored.
663
664 * The --without-fp configure option is now ignored. Whether hardware
665 floating-point instructions are used is now configured based on whether
666 the compiler used at configure time (without any options implied by a
667 --with-cpu= configure option) uses such instructions.
668
669 * The res_hnok, res_dnok, res_mailok and res_ownok functions now check that
670 the specified string can be parsed as a domain name.
671
672 * In the malloc_info output, the <heap> element may contain another <aspace>
673 element, "subheaps", which contains the number of sub-heaps.
674
675 * The libresolv function p_secstodate is no longer supported for new
676 programs.
677
678 * The tilepro-*-linux-gnu configuration is no longer supported.
679
680 * The nonstandard header files <libio.h> and <_G_config.h> are deprecated
681 and will be removed in a future release. Software that is still using
682 either header should be updated to use standard <stdio.h> interfaces
683 instead.
684
685 libio.h was originally the header for a set of supported GNU extensions,
686 but they have not been maintained as such in many years, they are now
687 standing in the way of improvements to stdio, and we don't think there are
688 any remaining external users. _G_config.h was never intended for public
689 use, but predates the bits convention.
690
691 Changes to build and runtime requirements:
692
693 * bison version 2.7 or later is required to generate code in the 'intl'
694 subdirectory.
695
696 Security related changes:
697
698 CVE-2009-5064: The ldd script would sometimes run the program under
699 examination directly, without preventing code execution through the
700 dynamic linker. (The glibc project disputes that this is a security
701 vulnerability; only trusted binaries must be examined using the ldd
702 script.)
703
704 CVE-2017-15670: The glob function, when invoked with GLOB_TILDE,
705 suffered from a one-byte overflow during ~ operator processing (either
706 on the stack or the heap, depending on the length of the user name).
707 Reported by Tim Rühsen.
708
709 CVE-2017-15671: The glob function, when invoked with GLOB_TILDE,
710 would sometimes fail to free memory allocated during ~ operator
711 processing, leading to a memory leak and, potentially, to a denial
712 of service.
713
714 CVE-2017-15804: The glob function, when invoked with GLOB_TILDE and
715 without GLOB_NOESCAPE, could write past the end of a buffer while
716 unescaping user names. Reported by Tim Rühsen.
717
718 CVE-2017-17426: The malloc function, when called with an object size near
719 the value SIZE_MAX, would return a pointer to a buffer which is too small,
720 instead of NULL. This was a regression introduced with the new malloc
721 thread cache in glibc 2.26. Reported by Iain Buclaw.
722
723 CVE-2017-1000408: Incorrect array size computation in _dl_init_paths leads
724 to the allocation of too much memory. (This is not a security bug per se,
725 it is mentioned here only because of the CVE assignment.) Reported by
726 Qualys.
727
728 CVE-2017-1000409: Buffer overflow in _dl_init_paths due to miscomputation
729 of the number of search path components. (This is not a security
730 vulnerability per se because no trust boundary is crossed if the fix for
731 CVE-2017-1000366 has been applied, but it is mentioned here only because
732 of the CVE assignment.) Reported by Qualys.
733
734 CVE-2017-16997: Incorrect handling of RPATH or RUNPATH containing $ORIGIN
735 for AT_SECURE or SUID binaries could be used to load libraries from the
736 current directory.
737
738 CVE-2018-1000001: Buffer underflow in realpath function when getcwd function
739 succeeds without returning an absolute path due to unexpected behaviour
740 of the Linux kernel getcwd syscall. Reported by halfdog.
741
742 CVE-2018-6485: The posix_memalign and memalign functions, when called with
743 an object size near the value of SIZE_MAX, would return a pointer to a
744 buffer which is too small, instead of NULL. Reported by Jakub Wilk.
745
746 CVE-2018-6551: The malloc function, when called with an object size near
747 the value of SIZE_MAX, would return a pointer to a buffer which is too
748 small, instead of NULL.
749
750 The following bugs are resolved with this release:
751
752 [866] glob: glob should match dangling symlinks
753 [1062] glob: sysdeps/generic/glob.c merge from gnulib (part 3 of 3)
754 [2522] localedata: ca_ES@valencia: new Valencian (meridional Catalan)
755 locale
756 [5997] math: Very slow execution of sinf function
757 [10580] localedata: hr_HR: updated locale
758 [10871] locale: 'mon' array should contain both nominative and genitive
759 cases
760 [12349] localedata: eu_ES: incorrect thousands separator
761 [13605] localedata: shn_MM: new Shan locale
762 [13805] localedata: ru_RU: currency should use ',' as radix point
763 [13953] localedata: km_KH: locale update
764 [13994] localedata: mjw_IN: new locale
765 [14121] build: make writes .mo files in po directory
766 [14333] libc: Fix the race between atexit() and exit()
767 [14681] dynamic-link: _dl_get_origin leaks memory via executable link map.
768 [14925] localedata: bn_*: LC_IDENTIFICATION.language key should be
769 "Bangla"
770 [15260] localedata: LC_MESSAGES.{yes,no}{str,expr}: various errors
771 [15261] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of full-
772 width Latin characters
773 [15332] localedata: es_CU: locale update
774 [15436] stdio: Don't close or flush stdio streams on abort
775 [15537] localedata: lv_LV: invalid collation for Latvian diacritical
776 letters
777 [16148] localedata: ca_ES: incorrect thousands separator
778 [16750] dynamic-link: ldd should not try to execute the binaries
779 (CVE-2009-5064)
780 [16777] localedata: pl_PL: incorrect thousands separator in locale
781 [16905] localedata: hanzi: new collation
782 [17563] localedata: cmn_TW: add hanzi collation
783 [17750] localedata: wrong collation order of diacritics in most locales
784 [17804] libc: scandirat fails with ENOMEM because it checks for errno even
785 if malloc succeeded
786 [17956] build: Build fails on missing definitions from header file
787 nss/nss.h when Mozilla NSS is used for cryptography
788 [18203] libc: realpath() does not handle unreachable paths correctly
789 [18572] dynamic-link: [arm] Lazy TLSDESC relocation has data race
790 [18812] localedata: kab_DZ: new Kabyle Algeria locale
791 [18822] libc: Internal functions are called via PLT
792 [18858] string: _HAVE_STRING_ARCH_xxx aren't defined for i386 nor x86_64
793 [19170] libc: __gmon_start__ defined in hppa in crtn.S
794 [19574] libc: glibc should support building static PIE binaries
795 [19852] localedata: charmaps/UTF-8: incorrect wcwidth for U+3099 and
796 U+309A
797 [19971] glob: glob: Do not skip entries with zero d_ino values
798 [19982] localedata: fr.po: spelling mistake for error code EXDEV
799 [20008] localedata: km_KH: convert to translit_neutral
800 [20009] localedata: tr_TR: convert LC_CTYPE to i18n
801 [20142] math: [x86_64] Add SSE4.1 trunc, truncf
802 [20204] dynamic-link: _dl_open_hook and _dlfcn_hook hardening
803 [20482] localedata: de_CH: abbreviated weekdays should be two letters
804 [20498] localedata: miq_NI: new Mískitu / Miskito (miq) language locale
805 for Nicaragua
806 [20532] nss: getaddrinfo uses errno and h_errno without guaranteeing
807 they're set, wrong errors returned by gaih_inet when lookup functions
808 are not found.
809 [20756] localedata: [PATCH] Use Unicode wise thousands separator
810 [20826] network: posix/tst-getaddrinfo5 fails on hosts without network
811 access
812 [20952] localedata: yuw_PG: new locale
813 [21084] localedata: charmaps/IBM858: new codepage
814 [21161] manual: [PATCH] fix typo in manual/arith.texi on strtoul prototype
815 [21242] libc: assert gives pedantic warning in old gcc versions
816 [21265] dynamic-link: _dl_runtime_resolve isn't compatible with Intel C++
817 __regcall calling convention
818 [21309] math: signed integer overflow in sysdeps/ieee754/dbl-64/e_pow.c
819 [21326] libc: C99 functions are not declared for C++11 and later if
820 _GNU_SOURCE is not predefined by g++
821 [21457] libc: sys/ucontext.h namespace
822 [21530] libc: tmpfile() should be implemented using O_TMPFILE
823 [21660] math: GCC fails to compile a formula with tgmath.h
824 [21672] nptl: sys-libs/glibc on ia64 crashes on thread exit: signal
825 SIGSEGV, Segmentation fault: pthread_create.c:432: __madvise
826 (pd->stackblock, freesize - PTHREAD_STACK_MIN, MADV_DONTNEED);
827 [21684] math: tgmath.h handling of complex integers
828 [21685] math: tgmath.h handling of bit-fields
829 [21686] math: tgmath.h handling of __int128
830 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
831 locale
832 [21745] libc: [powerpc64le] Extra PLT reference with --enable-stack-
833 protector=all
834 [21750] localedata: column width of characters incompatible with classical
835 wcwidth
836 [21754] malloc: malloc: Perform as little work as possible after heap
837 consistency check failures
838 [21780] libc: hppa: p{read,write}v2 does not set ENOSUP on invalid flag
839 [21790] libc: Missing __memset_zero_constant_len_parameter in libc.so
840 [21791] string: Unused XXX_chk_XXX functions in libc.a
841 [21815] dynamic-link: FAIL: elf/tst-prelink-cmp with GCC is defaulted to
842 PIE
843 [21836] localedata: Removed redundant data (LC_MONETARY) in various Indian
844 locales
845 [21845] localedata: Added new Locale bho_NP
846 [21853] localedata: Fix abday Which looks same as day in zh_SG
847 [21854] localedata: Added New Locale en_SC
848 [21864] libc: xmalloc.o is compiled with -DMODULE_NAME=libc
849 [21871] dynamic-link: _dl_runtime_resolve_avx_opt is slower than
850 _dl_runtime_resolve_avx_slow
851 [21885] network: getaddrinfo: gethosts does not release resolver context
852 on memory allocation failure
853 [21899] libc: XPG4.2 sigaction namespace
854 [21908] dynamic-link: dynamic linker broke on ia64 (mmap2 consolidation is
855 the suspect)
856 [21913] libc: static binaries SIGSEGV in __brk when host's gcc is pie-by-
857 default (i386)
858 [21915] nss: nss_files can return with NSS_STATUS_SUCCESS and a clobbered
859 errno value, causing getaddrinfo to fail
860 [21920] localedata: Fix p_cs_precedes/n_cs_precedes for mt_MT
861 [21922] network: getaddrinfo with AF_INET/AF_INET6 returns EAI_NONAME
862 instead of EAI_NODATA
863 [21928] libc: sys/ptrace.h: remove obsolete temporary development Linux
864 constant PTRACE_SEIZE_DEVEL
865 [21930] math: C-only gcc builtins used in <math.h> isinf
866 [21932] network: Unpaired __resolv_context_get in generic get*_r
867 implementation
868 [21941] math: powerpc: Wrong register constraint for xssqrtqp in sqrtf128
869 [21944] libc: sigval namespace
870 [21951] localedata: Update hanzi collation by stroke
871 [21955] math: Wrong alignment of L(SP_RANGE)/L(SP_INF_0) in
872 sysdeps/x86_64/fpu/e_expf.S
873 [21956] libc: Stack allocation in MIPS syscall impl (ubounded stack
874 allocation in syscall loops)
875 [21959] localedata: Fix Country name for xh_ZA
876 [21960] localedata: Fix abmon for bem_ZM
877 [21966] math: AVX2 mathvec functions use FMA without checking
878 [21967] math: When 512-bit AVX2 wrapper functions in mathvec are used?
879 [21971] localedata: Added New Locale for mfe_MU
880 [21972] libc: assert macro requires operator== (int) for its argument type
881 [21973] math: [sparc] libm missing sqrtl compat symbol
882 [21974] libc: Remove __bb_init_func and __bb_exit_func
883 [21982] string: stratcliff.c: error: assuming signed overflow does not
884 occur with -O3
885 [21986] stdio: __guess_grouping is called incorrectly
886 [21987] math: [sparc32] wrong bits/long-double.h installed
887 [22019] localedata: Wrong placement of monetary symbol in el_GR (negative
888 amounts)
889 [22022] localedata: Missing country_name for mni_IN
890 [22023] localedata: Removed redundant data (LC_TIME and LC_MESSAGES) for
891 niu_NZ
892 [22025] locale: iconv: Inconsistency between pointer mangling and NULL
893 checks
894 [22026] locale: iconv_open: heap overflow on gconv_init failure
895 [22028] math: bits/math-finite.h _MSUF_ expansion namespace
896 [22035] math: [m68k] bits/math-inline.h macro namespace
897 [22038] localedata: Fix abbreviated weeks and months for Somali
898 [22044] localedata: Remove redundant data for Limburgish Language
899 [22050] malloc: Linking with -lmcheck does not hook
900 __malloc_initialize_hook correctly
901 [22051] libc: zero terminator in the middle of glibc's .eh_frame
902 [22052] malloc: malloc failed to compile with GCC 7 and -O3
903 [22070] localedata: charmaps/UTF-8: wcwidth for
904 Prepended_Concatenation_Mark codepoints set to 0 (should be 1)
905 [22074] localedata: charmaps/UTF-8: wcwidth for U+1160-U+11FF (Hangul
906 Jungseong and Jongseong) should be 0
907 [22078] nss: nss_files performance issue in multi mode
908 [22082] math: bits/math-finite.h exp10 condition
909 [22086] libc: pcprofiledump incorrect cross-endian condition
910 [22093] dynamic-link: ld.so no longer searches in .../x86_64
911 [22095] network: Name server address allocation memory leak in resolv.conf
912 parsing after OOM
913 [22096] network: __resolv_conf_attach can incorrectly free passed conf
914 object
915 [22100] localedata: om_KE: LC_TIME: copy redundant data from om_ET
916 [22101] dynamic-link: Dynamic loader must ignore "debug" shared objects
917 e.g. ET_GNU_DEBUG_*
918 [22111] malloc: malloc: per thread cache is not returned when thread exits
919 [22112] localedata: Fix LC_TELEPHONE/LC_NAME for az_AZ
920 [22134] libc: [linux] implement fexecve with execveat
921 [22142] libc: [powerpc] printf oupts a wrong value of DBL_MAX on ppc64 and
922 ppc64le
923 [22145] libc: ttyname() gives up too early in the face of namespaces
924 [22146] math: C++ build issue with float128 on x86_64
925 [22153] nptl: nptl: save error code before process termination
926 [22156] libc: [hppa,ia64,microblaze] Executable stack default
927 [22159] malloc: malloc: MALLOC_CHECK_ broken with --enable-tunables=no
928 [22161] nscd: nscd cache prune for netgroups hangs after timeout bump
929 [22165] libc: [hppa] Text relocations in libc.so
930 [22180] libc: destructor registered via __cxa_atexit is called twice
931 [22183] glob: commit 5554304f0ddd ("posix: Allow glob to match dangling
932 symlinks") cause "make" segfaults
933 [22189] math: [powerpc] math_private.h definitions of math_opt_barrier and
934 math_force_eval
935 [22207] libc: FAIL: stdlib/test-atexit-race
936 [22225] math: nearbyint arithmetic moved before feholdexcept
937 [22229] math: [sparc32] missing copysignl, fabsl, fmal compat symbols
938 [22235] math: iscanonical in C++ and float128
939 [22243] math: log2(0) and log10(0) are wrong in downward rounding without
940 the svid compat wrapper
941 [22244] math: ynf and yn are wrong without the svid compat wrapper
942 [22273] libc: Improper assert in Linux posix_spawn implementation
943 [22284] libc: -pg -pie doesn't work
944 [22292] locale: localedef exits with error 4 when it should be error 1
945 [22294] locale: Allow "" for int_currency_symbol definition in locales.
946 [22295] locale: Don't warn on non-symbolic characters in locale sources in
947 --verbose.
948 [22296] math: glibc 2.26: signbit build issue with Gcc 5.5.0 on x86_64
949 [22298] nptl: x32: lockups on recursive pthread_mutex_lock after upgrade
950 to 2.26
951 [22299] dynamic-link: Problem with $PLATFORM on x86_64 platform
952 [22320] glob: Buffer overflow in glob with GLOB_TILDE (CVE-2017-15670)
953 [22321] libc: sysconf(_SC_IOV_MAX) returns -1 on Linux
954 [22322] libc: [mips64] wrong bits/long-double.h installed
955 [22325] glob: Memory leak in glob with GLOB_TILDE (CVE-2017-15671)
956 [22332] glob: Buffer overflow in glob with GLOB_TILDE in unescaping
957 (CVE-2017-15804)
958 [22336] localedata: cs_CZ LC_COLLATE does not use i18n
959 [22343] malloc: Integer overflow in posix_memalign
960 [22347] libc: getrandom() returns the number of bytes that were copied to
961 the buffer even though the comments say "Return 0 on success and -1 on
962 failure."
963 [22353] string: sysdeps/i386/i586/strcpy.S isn't maintainable
964 [22362] libc: Installed crt1.o, crti,.o and crtn.o files are used with
965 -m32
966 [22370] dynamic-link: Incorrect note padding check
967 [22375] libc: malloc returns pointer from tcache_get when should return
968 NULL (CVE-2017-17426)
969 [22377] math: iseqsig, float128 and C++
970 [22382] localedata: Error in tpi_PG locale
971 [22387] localedata: Replace unicode sequences <Uxxxx> for characters
972 inside the ASCII printable range
973 [22402] math: [powerpc64le] __MATH_TG does not support _Float128 for
974 -mlong-double-64
975 [22403] localedata: Slash needs escaping in some locales
976 [22408] malloc: malloc_info access heaps without arena lock, ignores heaps
977 [22409] network: res_hnok does not accept some host names used on the
978 Internet
979 [22412] network: res_dnok, res_hnok should perform syntax checks
980 [22413] network: ns_name_pton ignores syntactically invalid trailing
981 backslash
982 [22415] stdio: setvbuf can lead to invalid free/segfault
983 [22432] build: Non-deterministic build
984 [22439] malloc: malloc_info should compute summary statistics for all sub-
985 heaps in an arena
986 [22442] network: if_nametoindex could report index for the wrong
987 networking interface
988 [22446] build: aliasing violation calling readlink in handle_request
989 [22447] build: unsafe call to strlen with a non-string in getlogin_r.c
990 [22457] libc: Generic preadv/pwritev incorrectly calls __posix_memalign
991 [22459] libc: FAIL: elf/check-localplt with __stack_chk_fail related to
992 __nscd_hash/__nss_hash
993 [22463] network: p_secstodate overflow handling
994 [22469] localedata: pl_PL LC_COLLATE does not use i18n
995 [22478] libc: sigwait can fail with EINTR
996 [22505] libc: ldconfig processes include directive in locale-specific
997 order
998 [22515] localedata: hsb_DE LC_COLLATE does not use copy "iso14651_t1"
999 [22517] localedata: et_EE LC_COLLATE does not use copy "iso14651_t1"
1000 [22519] localedata: is_IS LC_COLLATE does not use copy "iso14651_t1"
1001 [22524] localedata: lt_LT LC_COLLATE does not use copy "iso14651_t1"
1002 [22527] localedata: tr_TR LC_COLLATE does not use copy "iso14651_t1"
1003 [22534] localedata: Collation rules for Serbian and Bosnian should be the
1004 same as for Croatian
1005 [22561] math: [DR#471] cacosh (0 + iNaN) should return NaN +/- i pi/2
1006 [22568] math: [DR#471] ctanh (0 + iNaN), ctanh (0 + i Inf)
1007 [22577] libc: missing newline after "cannot allocate TLS data structures
1008 for initial thread"
1009 [22588] manual: manual/conf.texi: missing underscore in front of
1010 SC_SSIZE_MAX
1011 [22593] math: nextafter and nexttoward are declared with const attribute
1012 [22596] manual: manual: finite(nan) wrongly described as returning nonzero
1013 [22603] string: ia64 memchr overflows internal pointer check
1014 [22605] libc: SH clone does not set the exit code correctly
1015 [22606] dynamic-link: Incorrect array size computation in _dl_init_paths
1016 (CVE-2017-1000408)
1017 [22607] dynamic-link: Buffer Overflow in _dl_init_paths (CVE-2017-1000409)
1018 [22611] malloc: malloc/tst-realloc wrongly assumes that errno must not be
1019 modified in case of success
1020 [22614] build: gcc: error: unrecognized command line option ‘-no-pie’
1021 [22615] manual: manual: ambiguous wording about errno value in case of
1022 success
1023 [22624] libc: MIPS setjmp() saves incorrect 'o0' register in --enable-
1024 stack-protector=all
1025 [22625] dynamic-link: RPATH $ORIGIN replaced by PWD for AT_SECURE/SUID
1026 binaries or if /proc is not mounted (CVE-2017-16997)
1027 [22627] dynamic-link: $ORIGIN in $LD_LIBRARY_PATH is substituted twice
1028 [22630] build: $(no-pie-ldflag) is no longer effective
1029 [22631] math: [m68k] Bad const attributes in bits/mathinline.h
1030 [22635] nptl: pthread_self returns NULL before libpthread is loaded
1031 [22636] nptl: PTHREAD_STACK_MIN is too small on x86-64
1032 [22637] nptl: guard size is subtracted from thread stack size instead of
1033 adding it on top
1034 [22648] libc: getrlimit/setrlimit with RLIM_INFINITY broken on alpha
1035 [22657] localedata: hu_HU: Avoid double space in date
1036 [22660] math: fmax, fmin sNaN handling on alpha
1037 [22664] libc: New warning of GCC8
1038 [22665] math: alpha: ceil and floor raise inexact exceptions
1039 [22666] math: alpha: trunc raise inexact exceptions
1040 [22667] libc: makecontext lacks stack alignment on i386
1041 [22678] libc: prlimit fails for RLIM_INFINITY values on 32-bit machines
1042 [22679] libc: getcwd(3) can succeed without returning an absolute path
1043 (CVE-2018-1000001)
1044 [22685] libc: PowerPC: Static AT_SECURE binaries segfault with lock-
1045 elision and tunables
1046 [22687] math: [powerpc-nofpu] complex long double functions spurious
1047 "invalid" exception
1048 [22688] math: [powerpc-nofpu] remainderl wrong sign of zero result
1049 [22690] math: [ldbl-128ibm] lrintl, lroundl missing "invalid" exceptions
1050 [22691] math: [powerpc-nofpu] fmaxmagl, fminmagl spurious "invalid"
1051 exception
1052 [22693] math: [ldbl-128ibm] log1pl (-qNaN) spurious "invalid" exception
1053 [22697] math: [powerpc] llround spurious "inexact" exceptions on 32-bit
1054 power4
1055 [22701] nis: Incomplete removal of libnsl
1056 [22702] math: [powerpc-nofpu] nearbyintl traps with trapping "inexact"
1057 [22707] libc: Missing defines in elf.h for DF_1_STUB and DF_1_PIE.
1058 [22715] dynamic-link: FAIL: elf/tst-audit10
1059 [22719] libc: Backtrace tests fail on hppa
1060 [22742] libc: [aarch64] mcontext_t __reserved field got renamed
1061 [22743] nptl: __pthread_register_cancel corrupts stack after f81ddabffd
1062 [22765] crypt: (struct crypt_data *data)->initialized is not set to zero
1063 before the first call to crypt_r () in crypt/badsalttest.c
1064
1065 \f
1066 Version 2.26
1067
1068 Major new features:
1069
1070 * A per-thread cache has been added to malloc. Access to the cache requires
1071 no locks and therefore significantly accelerates the fast path to allocate
1072 and free small amounts of memory. Refilling an empty cache requires locking
1073 the underlying arena. Performance measurements show significant gains in a
1074 wide variety of user workloads. Workloads were captured using a special
1075 instrumented malloc and analyzed with a malloc simulator. Contributed by
1076 DJ Delorie with the help of Florian Weimer, and Carlos O'Donell.
1077
1078 * Unicode 10.0.0 Support: Character encoding, character type info, and
1079 transliteration tables are all updated to Unicode 10.0.0, using
1080 generator scripts contributed by Mike FABIAN (Red Hat).
1081 These updates cause user visible changes, especially the changes in
1082 wcwidth for many emoji characters cause problems when emoji sequences
1083 are rendered with pango, see for example:
1084 https://bugzilla.gnome.org/show_bug.cgi?id=780669#c5
1085
1086 * Collation of Hungarian has been overhauled and is now consistent with "The
1087 Rules of Hungarian Orthography, 12th edition" (Bug 18934). Contributed by
1088 Egmont Koblinger.
1089
1090 * Improvements to the DNS stub resolver, contributed by Florian Weimer:
1091
1092 - The GNU C Library will now detect when /etc/resolv.conf has been
1093 modified and reload the changed configuration. The new resolver option
1094 “no-reload” (RES_NORELOAD) disables this behavior.
1095
1096 - The GNU C Library now supports an arbitrary number of search domains
1097 (configured using the “search” directive in /etc/resolv.conf);
1098 previously, there was a hard limit of six domains. For backward
1099 compatibility, applications that directly modify the ‘_res’ global
1100 object are still limited to six search domains.
1101
1102 - When the “rotate” (RES_ROTATE) resolver option is active, the GNU C
1103 Library will now randomly pick a name server from the configuration as a
1104 starting point. (Previously, the second name server was always used.)
1105
1106 * The tunables feature is now enabled by default. This allows users to tweak
1107 behavior of the GNU C Library using the GLIBC_TUNABLES environment variable.
1108
1109 * New function reallocarray, which resizes an allocated block (like realloc)
1110 to the product of two sizes, with a guaranteed clean failure upon integer
1111 overflow in the multiplication. Originally from OpenBSD, contributed by
1112 Dennis Wölfing and Rüdiger Sonderfeld.
1113
1114 * New wrappers for the Linux-specific system calls preadv2 and pwritev2.
1115 These are extended versions of preadv and pwritev, respectively, taking an
1116 additional flags argument. The set of supported flags depends on the
1117 running kernel; full support currently requires kernel 4.7 or later.
1118
1119 * posix_spawnattr_setflags now supports the flag POSIX_SPAWN_SETSID, to
1120 create a new session ID for the spawned process. This feature is
1121 scheduled to be added to the next major revision of POSIX; for the time
1122 being, it is available under _GNU_SOURCE.
1123
1124 * errno.h is now safe to use from C-preprocessed assembly language on all
1125 supported operating systems. In this context, it will only define the
1126 Exxxx constants, as preprocessor macros expanding to integer literals.
1127
1128 * On ia64, powerpc64le, x86-32, and x86-64, the math library now implements
1129 128-bit floating point as defined by ISO/IEC/IEEE 60559:2011 (IEEE
1130 754-2008) and ISO/IEC TS 18661-3:2015. Contributed by Paul E. Murphy,
1131 Gabriel F. T. Gomes, Tulio Magno Quites Machado Filho, and Joseph Myers.
1132
1133 To compile programs that use this feature, the compiler must support
1134 128-bit floating point with the type name _Float128 (as defined by TS
1135 18661-3) or __float128 (the nonstandard name used by GCC for C++, and for
1136 C prior to version 7). _GNU_SOURCE or __STDC_WANT_IEC_60559_TYPES_EXT__
1137 must be defined to make the new interfaces visible.
1138
1139 The new functions and macros correspond to those present for other
1140 floating-point types (except for a few obsolescent interfaces not
1141 supported for the new type), with F128 or f128 suffixes; for example,
1142 strtof128, HUGE_VAL_F128 and cosf128. Following TS 18661-3, there are no
1143 printf or scanf formats for the new type; the strfromf128 and strtof128
1144 interfaces should be used instead.
1145
1146 Deprecated and removed features, and other changes affecting compatibility:
1147
1148 * The synchronization that pthread_spin_unlock performs has been changed to
1149 now be equivalent to a C11 atomic store with release memory order to the
1150 spin lock's memory location. Previously, several (but not all)
1151 architectures used stronger synchronization (e.g., containing what is
1152 often called a full barrier). This change can improve performance, but
1153 may affect odd fringe uses of spin locks that depend on the previous
1154 behavior (e.g., using spin locks as atomic variables to try to implement
1155 Dekker's mutual exclusion algorithm).
1156
1157 * The port to Native Client running on ARMv7-A (--host=arm-nacl) has been
1158 removed.
1159
1160 * Sun RPC is deprecated. The rpcgen program, librpcsvc, and Sun RPC headers
1161 will only be built and installed when the GNU C Library is configured with
1162 --enable-obsolete-rpc. This allows alternative RPC implementations, such
1163 as TIRPC or rpcsvc-proto, to be used.
1164
1165 * The NIS(+) name service modules, libnss_nis, libnss_nisplus, and
1166 libnss_compat, are deprecated, and will not be built or installed by
1167 default.
1168
1169 The NIS(+) support library, libnsl, is also deprecated. By default, a
1170 compatibility shared library will be built and installed, but not headers
1171 or development libraries. Only a few NIS-related programs require this
1172 library. (In particular, the GNU C Library has never required programs
1173 that use 'gethostbyname' to be linked with libnsl.)
1174
1175 Replacement implementations based on TIRPC, which additionally support
1176 IPv6, are available from <https://github.com/thkukuk/>. The configure
1177 option --enable-obsolete-nsl will cause libnsl's headers, and the NIS(+)
1178 name service modules, to be built and installed.
1179
1180 * The DNS stub resolver no longer performs EDNS fallback. If EDNS or DNSSEC
1181 support is enabled, the configured recursive resolver must support EDNS.
1182 (Responding to EDNS-enabled queries with responses which are not
1183 EDNS-enabled is fine, but FORMERR responses are not.)
1184
1185 * res_mkquery and res_nmkquery no longer support the IQUERY opcode. DNS
1186 servers have not supported this opcode for a long time.
1187
1188 * The _res_opcodes variable has been removed from libresolv. It had been
1189 exported by accident.
1190
1191 * <string.h> no longer includes inline versions of any string functions,
1192 as this kind of optimization is better done by the compiler. The macros
1193 __USE_STRING_INLINES and __NO_STRING_INLINES no longer have any effect.
1194
1195 * The nonstandard header <xlocale.h> has been removed. Most programs should
1196 use <locale.h> instead. If you have a specific need for the definition of
1197 locale_t with no other declarations, please contact
1198 libc-alpha@sourceware.org and explain.
1199
1200 * The obsolete header <sys/ultrasound.h> has been removed.
1201
1202 * The obsolete signal constant SIGUNUSED is no longer defined by <signal.h>.
1203
1204 * The obsolete function cfree has been removed. Applications should use
1205 free instead.
1206
1207 * The stack_t type no longer has the name struct sigaltstack. This changes
1208 the C++ name mangling for interfaces involving this type.
1209
1210 * The ucontext_t type no longer has the name struct ucontext. This changes
1211 the C++ name mangling for interfaces involving this type.
1212
1213 * On M68k GNU/Linux and MIPS GNU/Linux, the fpregset_t type no longer has
1214 the name struct fpregset. On Nios II GNU/Linux, the mcontext_t type no
1215 longer has the name struct mcontext. On SPARC GNU/Linux, the struct
1216 mc_fq, struct rwindow, struct fpq and struct fq types are no longer
1217 defined in sys/ucontext.h, the mc_fpu_t type no longer has the name struct
1218 mc_fpu, the gwindows_t type no longer has the name struct gwindows and the
1219 fpregset_t type no longer has the name struct fpu. This changes the C++
1220 name mangling for interfaces involving those types.
1221
1222 * On S/390 GNU/Linux, the constants defined by <sys/ptrace.h> have been
1223 synced with the kernel:
1224
1225 - PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS and PTRACE_SETFPREGS
1226 are not supported on this architecture and have been removed.
1227
1228 - PTRACE_SINGLEBLOCK, PTRACE_SECCOMP_GET_FILTER, PTRACE_PEEKUSR_AREA,
1229 PTRACE_POKEUSR_AREA, PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE,
1230 PTRACE_DISABLE_TE and PTRACE_TE_ABORT_RAND have been added.
1231
1232 Programs that assume the GET/SETREGS ptrace requests are universally
1233 available will now fail to build, instead of malfunctioning at runtime.
1234
1235 Changes to build and runtime requirements:
1236
1237 * Linux kernel 3.2 or later is required at runtime, on all architectures
1238 supported by that kernel. (This is a change from version 2.25 only for
1239 x86-32 and x86-64.)
1240
1241 * GNU Binutils 2.25 or later is now required to build the GNU C Library.
1242
1243 * On most architectures, GCC 4.9 or later is required to build the GNU C
1244 Library. On powerpc64le, GCC 6.2 or later is required.
1245
1246 Older GCC versions and non-GNU compilers are still supported when
1247 compiling programs that use the GNU C Library. (We do not know exactly
1248 how old, and some GNU extensions to C may be _de facto_ required. If you
1249 are interested in helping us make this statement less vague, please
1250 contact libc-alpha@sourceware.org.)
1251
1252 Security related changes:
1253
1254 * The DNS stub resolver limits the advertised UDP buffer size to 1200 bytes,
1255 to avoid fragmentation-based spoofing attacks (CVE-2017-12132).
1256
1257 * LD_LIBRARY_PATH is now ignored in binaries running in privileged AT_SECURE
1258 mode to guard against local privilege escalation attacks (CVE-2017-1000366).
1259
1260 * Avoid printing a backtrace from the __stack_chk_fail function since it is
1261 called on a corrupt stack and a backtrace is unreliable on a corrupt stack
1262 (CVE-2010-3192).
1263
1264 * A use-after-free vulnerability in clntudp_call in the Sun RPC system has been
1265 fixed (CVE-2017-12133).
1266
1267 The following bugs are resolved with this release:
1268
1269 [984] network: Respond to changed resolv.conf in gethostbyname
1270 [5010] network: sunrpc service cleanup causes unwanted port mapper traffic
1271 [12068] localedata: sc_IT: misspelled yesexpr/day/abday/mon/abmon/date_fmt
1272 fields
1273 [12189] libc: __stack_chk_fail should not attempt a backtrace
1274 (CVE-2010-3192)
1275 [14096] time: Race condition on timezone/tst-timezone.out
1276 [14172] localedata: az_IR: new locale
1277 [14995] build: glibc fails to build if gold is the default linker, even if
1278 ld.bfd is available
1279 [15998] build: [powerpc] Set arch_minimum_kernel for powerpc LE
1280 [16637] network: inet_pton function is accepting IPv6 with bad format
1281 [16640] string: string/strtok.c: undefined behaviour inconsistent between
1282 x86 and other generic code
1283 [16875] localedata: ko_KR: fix lang_name
1284 [17225] localedata: ar_SY: localized month names for May and June are
1285 incorrect
1286 [17297] localedata: da_DK: wrong date_fmt string
1287 [18907] stdio: Incorrect order of __wur __THROW in <printf.h>
1288 [18934] localedata: hu_HU: collate: fix multiple bugs and add tests
1289 [18988] nptl: pthread wastes memory with mlockall(MCL_FUTURE)
1290 [19066] localedata: ar_SA abbreviated day and month names are in English
1291 [19569] network: resolv: Support an arbitrary number of search domains
1292 [19570] network: Implement random DNS server selection in the stub
1293 resolver
1294 [19838] locale: localedef fails on PA-RISC
1295 [19919] localedata: iso14651_t1_common: Correct the Malayalam sorting
1296 order of 0D36 and 0D37
1297 [19922] localedata: iso14651_t1_common: Define collation for Malayalam
1298 chillu characters
1299 [20098] libc: FAIL: debug/backtrace-tst on hppa
1300 [20257] network: sunrpc: clntudp_call does not enforce timeout when
1301 receiving data
1302 [20275] localedata: locale day/abday/mon/abmon should not have trailing
1303 whitespace
1304 [20313] localedata: Update locale data to Unicode 9.0
1305 [20424] manual: Document how to provide a malloc replacement
1306 [20496] localedata: agr_PE: new language locale Awajún / Aguaruna (agr)
1307 for Peru
1308 [20686] locale: Add el_GR@euro to SUPPORTED.
1309 [20831] dynamic-link: _dl_map_segments does not test for __mprotect
1310 failures consistently
1311 [21015] dynamic-link: Document and fix --enable-bind-now
1312 [21016] nptl: pthread_cond support is broken on hppa
1313 [21029] libc: glibc-2.23 (and later) fails to compile with -fno-omit-
1314 frame-pointer on i386
1315 [21049] libc: segfault in longjmp_chk() due to clobbered processor
1316 register
1317 [21075] libc: unused assigment to %g4 in sparc/sparc{64,32}/clone.S
1318 [21088] libc: Build fails with --enable-static-nss
1319 [21094] math: cosf(1.57079697) has 3 ulp error on targets where the
1320 generic c code is used
1321 [21109] libc: Tunables broken on big-endian
1322 [21112] math: powf has large ulp errors with base close to 1 and exponent
1323 around 4000
1324 [21115] network: sunrpc: Use-after-free in error path in clntudp_call
1325 (CVE-2017-12133)
1326 [21120] malloc: glibc malloc is incompatible with GCC 7
1327 [21130] math: Incorrect return from y0l (-inf) and y1l (-inf) when linking
1328 with -lieee
1329 [21134] math: Exception (divide by zero) not set for y0/y1 (0.0) and y0/y1
1330 (-0.0) when linking with -lieee
1331 [21171] math: log10, log2 and lgamma return incorrect results
1332 [21179] libc: handle R_SPARC_DISP64 and R_SPARC_REGISTER relocs
1333 [21182] libc: __memchr_sse2: regression in glibc-2.25 on i686
1334 [21207] localedata: ce_RU: update weekdays from CLDR
1335 [21209] dynamic-link: LD_HWCAP_MASK read in setuid binaries
1336 [21217] localedata: Update months from CLDR-31
1337 [21232] libc: miss posix_fadvise64 on MIPS64 when static linking
1338 [21243] libc: support_delete_temp_file should issue warning for failed
1339 remove()
1340 [21244] libc: support resolv_test_start() socket fd close should be
1341 checked for errors.
1342 [21253] libc: localedef randomly segfaults when using -fstack-check due to
1343 new posix_spawn implementation
1344 [21258] dynamic-link: Branch predication in _dl_runtime_resolve_avx512_opt
1345 leads to lower CPU frequency
1346 [21259] libc: [alpha] termios.h missing IXANY for POSIX
1347 [21261] libc: [sparc64] bits/setjmp.h namespace
1348 [21267] network: [mips] bits/socket.h IOC* namespace
1349 [21268] libc: [alpha] termios.h NL2, NL3 namespace
1350 [21270] libc: mmap64 silently truncates large offset values
1351 [21275] libc: posix_spawn always crashes on ia64 now
1352 [21277] libc: [alpha] termios.h missing IUCLC for UNIX98 and older
1353 [21280] math: [powerpc] logbl for POWER7 return incorrect results
1354 [21289] libc: Incorrect declaration for 32-bit platforms with
1355 _FILE_OFFSET_BITS=64 causes build error
1356 [21295] network: GETAI(AF_UNSPEC) drops IPv6 addresses if nss module does
1357 not support gethostbyname4_r
1358 [21298] nptl: rwlock can deadlock on frequent reader/writer phase
1359 switching
1360 [21338] malloc: mallopt M_ARENA_MAX doesn't set the maximum number of
1361 arenas
1362 [21340] libc: Support POSIX_SPAWN_SETSID
1363 [21357] libc: unwind-dw2-fde deadlock when using AddressSanitizer
1364 [21359] network: ns_name_pack needs additional byte in destination buffer
1365 [21361] network: resolv: Reduce advertised EDNS0 buffer size to guard
1366 against fragmentation attacks (CVE-2017-12132)
1367 [21369] network: resolv: Remove EDNS fallback
1368 [21371] libc: Missing timespec definition when compiled with _XOPEN_SOURCE
1369 and _POSIX_C_SOURCE
1370 [21386] nptl: Assertion in fork for distinct parent PID is incorrect
1371 [21391] dynamic-link: x86: Set dl_platform and dl_hwcap from CPU features
1372 [21393] stdio: Missing dup3 error check in freopen, freopen64
1373 [21396] libc: Use AVX2 memcpy/memset on Skylake server
1374 [21399] localedata: Bad description for U00EC in
1375 localedata/charmaps/CP1254
1376 [21411] malloc: realloc documentation error
1377 [21426] network: sys/socket.h uio.h namespace
1378 [21428] libc: [aarch64] tst-backtrace5 testsuite failure
1379 [21445] libc: signal.h bsd_signal namespace
1380 [21455] network: Network headers stdint.h namespace
1381 [21474] network: resolv: res_init does not use RES_DFLRETRY (2) but 4 for
1382 retry value
1383 [21475] network: resolv: Overlong search path is truncated mid-label
1384 [21511] libc: sigstack namespace
1385 [21512] libc: clone() ends up calling exit_group() through _exit() wrapper
1386 [21514] libc: sysdeps/unix/sysv/linux/sys/syscall.h:31:27: fatal error:
1387 bits/syscall.h: No such file or directory
1388 [21517] libc: struct sigaltstack namespace
1389 [21528] dynamic-link: Duplicated minimal strtoul implementations in ld.so
1390 [21533] localedata: Update locale data to Unicode 10.0
1391 [21537] libc:
1392 ../sysdeps/unix/sysv/linux/s390/s390-32/__makecontext_ret.S:44: Error:
1393 junk at end of line, first unrecognized character is `@'
1394 [21538] libc: SIG_HOLD missing for XPG4
1395 [21539] libc: S390: Mismatch between kernel and glibc ptrace.h with
1396 request 12: PTRACE_SINGLEBLOCK vs PTRACE_GETREGS.
1397 [21542] libc: Use conservative default for sysconf (_SC_NPROCESSORS_ONLN)
1398 [21543] libc: sigevent namespace
1399 [21548] libc: [mips] get/set/make/swap context for MIPS O32 assume wrong
1400 size for general purpose registers in mcontext_t structure
1401 [21550] libc: sigwait namespace
1402 [21552] libc: XPG4 bsd_signal namespace
1403 [21554] libc: sigpause namespace
1404 [21560] libc: sys/wait.h signal.h namespace
1405 [21561] libc: waitid namespace
1406 [21573] nptl: GCC 7: /usr/bin/install: cannot remove
1407 '/usr/include/stdlib.h': Permission denied
1408 [21575] libc: sys/wait.h missing struct rusage definition
1409 [21584] libc: sigaltstack etc namespace
1410 [21597] libc: siginterrupt namespace
1411 [21607] math: hppa: FAIL: math/test-tgmath
1412 [21609] dynamic-link: Incomplete workaround for GCC __tls_get_addr ABI
1413 issue on x86-64
1414 [21622] libc: [tile] missing SA_* for POSIX.1:2008
1415 [21624] dynamic-link: ld.so: Unsafe alloca allows local attackers to alias
1416 stack and heap (CVE-2017-1000366)
1417 [21625] libc: wait3 namespace
1418 [21654] nss: Incorrect pointer alignment in NSS group merge result
1419 construction
1420 [21657] network: Parse interface zone id for node-local multicast
1421 [21662] string: memcmp-avx2-movbe.S lacks saturating subtraction for
1422 between_2_3
1423 [21666] libc: .symver is used on common symbol
1424 [21668] network: resolv: res_init cross-thread broadcast introduces race
1425 conditions
1426 [21687] math: tgmath.h totalorder, totalordermag return type
1427 [21694] locale: Current Glibc Locale Does Not Support Tok-Pisin and Fiji
1428 Hindi Locale
1429 [21696] libc: Incorrect assumption of of __cpu_mask in
1430 posix/sched_cpucount.c
1431 [21697] libc: sysdeps/posix/spawni.c: 2 * suspicious condition ?
1432 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
1433 locale
1434 [21707] math: ppc64le: Invalid IFUNC resolver from libgcc calls getauxval,
1435 leading to relocation crash
1436 [21709] libc: resolv_conf.c:552: update_from_conf: Assertion
1437 `resolv_conf_matches (resp, conf)' failed.
1438 [21710] localedata: Added Samoan language locale for Samoa
1439 [21711] localedata: Pashto yesstr/nostr locale are missing
1440 [21715] nptl: sysdeps/nptl/bits/pthreadtypes.h: typedef guard
1441 __have_pthread_attr_t can cause redefinition of typedef ‘pthread_attr_t’
1442 [21721] localedata: Incorrect Full Weekday names for ks_IN@devanagari
1443 [21723] localedata: yesstr/nostr missing for Chinese language locale
1444 [21724] localedata: yesstr and nostr are missing for Xhosa [LC_MESSAGES]
1445 locale
1446 [21727] localedata: yesstr and nostr are missing for Tsonga [LC_MESSAGES]
1447 locale
1448 [21728] localedata: New Locale for Tongan language
1449 [21729] localedata: incorrect LC_NAME fields for hi_IN
1450 [21733] localedata: yesstr and nostr are missing for zh_HK
1451 [21734] localedata: Missing yesstr and nostr are for kw_GB
1452 [21738] libc: misc/tst-preadvwritev2 and misc/tst-preadvwritev64v2 fail
1453 [21741] libc: Undefined __memmove_chk_XXX and __memset_chk_XXX in libc.a
1454 [21742] libc: _dl_num_cache_relocations is undefined in libc.a
1455 [21743] localedata: ks_IN@devanagari: abday strings mismatch the day
1456 strings
1457 [21744] libc: Tests failing on --enable-tunables --enable-stack-
1458 protector=all
1459 [21749] localedata: Wrong abbreviated day name (“abday”) for
1460 ar_JO/ar_LB/ar_SY
1461 [21756] localedata: missing yesstr, nostr for nds_DE and nds_NL
1462 [21757] localedata: missing yesstr, nostr for pap_AW and pap_CW
1463 [21759] localedata: missing yesstr and nostr for Tigrinya
1464 [21760] localedata: Fix LC_MESSAGES and LC_ADDRESS for anp_IN
1465 [21766] localedata: Wrong LC_MESSAGES for om_ET Locale
1466 [21767] localedata: Missing Bislama locales
1467 [21768] localedata: Missing yesstr and nostr for aa_ET
1468 [21770] localedata: Missing Field in li_NL
1469 [21778] nptl: Robust mutex may deadlock
1470 [21779] libc: MicroBlaze segfaults when loading libpthread
1471 [21783] localedata: Fix int_select international_call_prefixes
1472 [21784] localedata: Inconsistency in country_isbn
1473 [21788] localedata: Missing Country Postal Abbreviations
1474 [21794] localedata: Added-country_isbn-for-Italy
1475 [21795] localedata: Add/Fix country_isbn for France
1476 [21796] localedata: Added country_isbn for Republic of Korea
1477 [21797] localedata: Fix inconsistency in country_isbn and missing prefixes
1478 [21799] localedata: Added int_select international_call_prefixes
1479 [21801] localedata: Added int_select international_call_prefixes
1480 [21804] nptl: Double semicolon in thread-shared-types.h
1481 [21807] localedata: LC_ADDRESS fix for pap_CW
1482 [21808] localedata: Fix LC_ADDRESS for pap_AW
1483 [21821] localedata: Added country_name in mai_IN
1484 [21822] localedata: Fix LC_TIME for mai_IN
1485 [21823] localedata: missing yesstr, nostr for sa_IN
1486 [21825] localedata: Fix name_mrs for mag_IN
1487 [21828] localedata: 2.26 changelog should mention user visible changes
1488 with unicode 9.0
1489 [21835] localedata: Added Maithili language locale for Nepal
1490 [21838] localedata: Removed redundant data for the_NP
1491 [21839] localedata: Fix LC_MONETARY for ta_LK
1492 [21844] localedata: Fix Latin characters and Months Sequence.
1493 [21848] localedata: Fix mai_NP Title Name
1494
1495 \f
1496 Version 2.25
1497
1498 * The feature test macro __STDC_WANT_LIB_EXT2__, from ISO/IEC TR
1499 24731-2:2010, is supported to enable declarations of functions from that
1500 TR. Note that not all functions from that TR are supported by the GNU C
1501 Library.
1502
1503 * The feature test macro __STDC_WANT_IEC_60559_BFP_EXT__, from ISO/IEC TS
1504 18661-1:2014, is supported to enable declarations of functions and macros
1505 from that TS. Note that not all features from that TS are supported by
1506 the GNU C Library.
1507
1508 * The feature test macro __STDC_WANT_IEC_60559_FUNCS_EXT__, from ISO/IEC TS
1509 18661-4:2015, is supported to enable declarations of functions and macros
1510 from that TS. Note that most features from that TS are not supported by
1511 the GNU C Library.
1512
1513 * The nonstandard feature selection macros _REENTRANT and _THREAD_SAFE are
1514 now treated as compatibility synonyms for _POSIX_C_SOURCE=199506L.
1515 Since the GNU C Library defaults to a much newer revision of POSIX, this
1516 will only affect programs that specifically request an old conformance
1517 mode. For instance, a program compiled with -std=c89 -D_REENTRANT will
1518 see a change in the visible declarations, but a program compiled with
1519 just -D_REENTRANT, or -std=c99 -D_POSIX_C_SOURCE=200809L -D_REENTRANT,
1520 will not.
1521
1522 Some C libraries once required _REENTRANT and/or _THREAD_SAFE to be
1523 defined by all multithreaded code, but glibc has not required this for
1524 many years.
1525
1526 * The inclusion of <sys/sysmacros.h> by <sys/types.h> is deprecated. This
1527 means that in a future release, the macros “major”, “minor”, and “makedev”
1528 will only be available from <sys/sysmacros.h>.
1529
1530 These macros are not part of POSIX nor XSI, and their names frequently
1531 collide with user code; see for instance glibc bug 19239 and Red Hat bug
1532 130601. <stdlib.h> includes <sys/types.h> under _GNU_SOURCE, and C++ code
1533 presently cannot avoid being compiled under _GNU_SOURCE, exacerbating the
1534 problem.
1535
1536 * New <fenv.h> features from TS 18661-1:2014 are added to libm: the
1537 fesetexcept, fetestexceptflag, fegetmode and fesetmode functions, the
1538 femode_t type and the FE_DFL_MODE and FE_SNANS_ALWAYS_SIGNAL macros.
1539
1540 * Integer width macros from TS 18661-1:2014 are added to <limits.h>:
1541 CHAR_WIDTH, SCHAR_WIDTH, UCHAR_WIDTH, SHRT_WIDTH, USHRT_WIDTH, INT_WIDTH,
1542 UINT_WIDTH, LONG_WIDTH, ULONG_WIDTH, LLONG_WIDTH, ULLONG_WIDTH; and to
1543 <stdint.h>: INT8_WIDTH, UINT8_WIDTH, INT16_WIDTH, UINT16_WIDTH,
1544 INT32_WIDTH, UINT32_WIDTH, INT64_WIDTH, UINT64_WIDTH, INT_LEAST8_WIDTH,
1545 UINT_LEAST8_WIDTH, INT_LEAST16_WIDTH, UINT_LEAST16_WIDTH,
1546 INT_LEAST32_WIDTH, UINT_LEAST32_WIDTH, INT_LEAST64_WIDTH,
1547 UINT_LEAST64_WIDTH, INT_FAST8_WIDTH, UINT_FAST8_WIDTH, INT_FAST16_WIDTH,
1548 UINT_FAST16_WIDTH, INT_FAST32_WIDTH, UINT_FAST32_WIDTH, INT_FAST64_WIDTH,
1549 UINT_FAST64_WIDTH, INTPTR_WIDTH, UINTPTR_WIDTH, INTMAX_WIDTH,
1550 UINTMAX_WIDTH, PTRDIFF_WIDTH, SIG_ATOMIC_WIDTH, SIZE_WIDTH, WCHAR_WIDTH,
1551 WINT_WIDTH.
1552
1553 * New <math.h> features are added from TS 18661-1:2014:
1554
1555 - Signaling NaN macros: SNANF, SNAN, SNANL.
1556
1557 - Nearest integer functions: roundeven, roundevenf, roundevenl, fromfp,
1558 fromfpf, fromfpl, ufromfp, ufromfpf, ufromfpl, fromfpx, fromfpxf,
1559 fromfpxl, ufromfpx, ufromfpxf, ufromfpxl.
1560
1561 - llogb functions: the llogb, llogbf and llogbl functions, and the
1562 FP_LLOGB0 and FP_LLOGBNAN macros.
1563
1564 - Max-min magnitude functions: fmaxmag, fmaxmagf, fmaxmagl, fminmag,
1565 fminmagf, fminmagl.
1566
1567 - Comparison macros: iseqsig.
1568
1569 - Classification macros: iscanonical, issubnormal, iszero.
1570
1571 - Total order functions: totalorder, totalorderf, totalorderl,
1572 totalordermag, totalordermagf, totalordermagl.
1573
1574 - Canonicalize functions: canonicalize, canonicalizef, canonicalizel.
1575
1576 - NaN functions: getpayload, getpayloadf, getpayloadl, setpayload,
1577 setpayloadf, setpayloadl, setpayloadsig, setpayloadsigf, setpayloadsigl.
1578
1579 * The functions strfromd, strfromf, and strfroml, from ISO/IEC TS 18661-1:2014,
1580 are added to libc. They convert a floating-point number into string.
1581
1582 * Most of glibc can now be built with the stack smashing protector enabled.
1583 It is recommended to build glibc with --enable-stack-protector=strong.
1584 Implemented by Nick Alcock (Oracle).
1585
1586 * The function explicit_bzero, from OpenBSD, has been added to libc. It is
1587 intended to be used instead of memset() to erase sensitive data after use;
1588 the compiler will not optimize out calls to explicit_bzero even if they
1589 are "unnecessary" (in the sense that no _correct_ program can observe the
1590 effects of the memory clear).
1591
1592 * On ColdFire, MicroBlaze, Nios II and SH3, the float_t type is now defined
1593 to float instead of double. This does not affect the ABI of any libraries
1594 that are part of the GNU C Library, but may affect the ABI of other
1595 libraries that use this type in their interfaces.
1596
1597 * On x86_64, when compiling with -mfpmath=387 or -mfpmath=sse+387, the
1598 float_t and double_t types are now defined to long double instead of float
1599 and double. These options are not the default, and this does not affect
1600 the ABI of any libraries that are part of the GNU C Library, but it may
1601 affect the ABI of other libraries that use this type in their interfaces,
1602 if they are compiled or used with those options.
1603
1604 * The getentropy and getrandom functions, and the <sys/random.h> header file
1605 have been added.
1606
1607 * The buffer size for byte-oriented stdio streams is now limited to 8192
1608 bytes by default. Previously, on Linux, the default buffer size on most
1609 file systems was 4096 bytes (and thus remains unchanged), except on
1610 network file systems, where the buffer size was unpredictable and could be
1611 as large as several megabytes.
1612
1613 * The <sys/quota.h> header now includes the <linux/quota.h> header. Support
1614 for the Linux quota interface which predates kernel version 2.4.22 has
1615 been removed.
1616
1617 * The malloc_get_state and malloc_set_state functions have been removed.
1618 Already-existing binaries that dynamically link to these functions will
1619 get a hidden implementation in which malloc_get_state is a stub. As far
1620 as we know, these functions are used only by GNU Emacs and this change
1621 will not adversely affect already-built Emacs executables. Any undumped
1622 Emacs executables, which normally exist only during an Emacs build, should
1623 be rebuilt by re-running “./configure; make” in the Emacs build tree.
1624
1625 * The “ip6-dotint” and “no-ip6-dotint” resolver options, and the
1626 corresponding RES_NOIP6DOTINT flag from <resolv.h> have been removed.
1627 “no-ip6-dotint” had already been the default, and support for the
1628 “ip6-dotint” option was removed from the Internet in 2006.
1629
1630 * The "ip6-bytestring" resolver option and the corresponding RES_USEBSTRING
1631 flag from <resolv.h> have been removed. The option relied on a
1632 backwards-incompatible DNS extension which was never deployed on the
1633 Internet.
1634
1635 * The flags RES_AAONLY, RES_PRIMARY, RES_NOCHECKNAME, RES_KEEPTSIG,
1636 RES_BLAST defined in the <resolv.h> header file have been deprecated.
1637 They were already unimplemented.
1638
1639 * The "inet6" option in /etc/resolv.conf and the RES_USE_INET6 flag for
1640 _res.flags are deprecated. The flag was standardized in RFC 2133, but
1641 removed again from the IETF name lookup interface specification in RFC
1642 2553. Applications should use getaddrinfo instead.
1643
1644 * DNSSEC-related declarations and definitions have been removed from the
1645 <arpa/nameser.h> header file, and libresolv will no longer attempt to
1646 decode the data part of DNSSEC record types. Previous versions of glibc
1647 only implemented minimal support for the previous version of DNSSEC, which
1648 is incompatible with the currently deployed version.
1649
1650 * The resource record type classification macros ns_t_qt_p, ns_t_mrr_p,
1651 ns_t_rr_p, ns_t_udp_p, ns_t_xfr_p have been removed from the
1652 <arpa/nameser.h> header file because the distinction between RR types and
1653 meta-RR types is not officially standardized, subject to revision, and
1654 thus not suitable for encoding in a macro.
1655
1656 * The types res_sendhookact, res_send_qhook, re_send_rhook, and the qhook
1657 and rhook members of the res_state type in <resolv.h> have been removed.
1658 The glibc stub resolver did not support these hooks, but the header file
1659 did not reflect that.
1660
1661 * For multi-arch support it is recommended to use a GCC which has
1662 been built with support for GNU indirect functions. This ensures
1663 that correct debugging information is generated for functions
1664 selected by IFUNC resolvers. This support can either be enabled by
1665 configuring GCC with '--enable-gnu-indirect-function', or by
1666 enabling it by default by setting 'default_gnu_indirect_function'
1667 variable for a particular architecture in the GCC source file
1668 'gcc/config.gcc'.
1669
1670 * GDB pretty printers have been added for mutex and condition variable
1671 structures in POSIX Threads. When installed and loaded in gdb these pretty
1672 printers show various pthread variables in human-readable form when read
1673 using the 'print' or 'display' commands in gdb.
1674
1675 * Tunables feature added to allow tweaking of the runtime for an application
1676 program. This feature can be enabled with the '--enable-tunables' configure
1677 flag. The GNU C Library manual has details on usage and README.tunables has
1678 instructions on adding new tunables to the library.
1679
1680 * A new version of condition variables functions have been implemented in
1681 the NPTL implementation of POSIX Threads to provide stronger ordering
1682 guarantees.
1683
1684 * A new version of pthread_rwlock functions have been implemented to use a more
1685 scalable algorithm primarily through not using a critical section anymore to
1686 make state changes.
1687
1688 Security related changes:
1689
1690 * On ARM EABI (32-bit), generating a backtrace for execution contexts which
1691 have been created with makecontext could fail to terminate due to a
1692 missing .cantunwind annotation. This has been observed to lead to a hang
1693 (denial of service) in some Go applications compiled with gccgo. Reported
1694 by Andreas Schwab. (CVE-2016-6323)
1695
1696 * The DNS stub resolver functions would crash due to a NULL pointer
1697 dereference when processing a query with a valid DNS question type which
1698 was used internally in the implementation. The stub resolver now uses a
1699 question type which is outside the range of valid question type values.
1700 (CVE-2015-5180)
1701
1702 The following bugs are resolved with this release:
1703
1704 [4099] stdio: Overly agressive caching by stream i/o functions.
1705 [7065] build: Support building glibc with -fstack-protector or -fstack-
1706 protector-all
1707 [9842] localedata: en_CA: incorrect date format
1708 [13165] nptl: pthread_cond_wait() can consume a signal that was sent
1709 before it started waiting
1710 [14139] manual: Do not hardcode platform names in manual/libm-err-tab.pl
1711 [15765] nptl: sem_open is wrongly a cancellation point
1712 [16421] network: IN6_IS_ADDR_UNSPECIFIED can use undefined s6_addr32
1713 [16458] libc: endian.h and netinet/in.h byte order macros return values of
1714 wrong type
1715 [16628] dynamic-link: Segfault after a binary without pthread dlopen()s a
1716 library linked with pthread
1717 [16630] nptl: Use SYSENTER for pthread_cond_broadcast/signal() (i.e. fix
1718 "FIXME: Ingo" issue)
1719 [16907] libc: <argp.h> compiled with --std=cXX disables __attribute__
1720 [17252] libc: getrandom and getentropy syscall
1721 [17730] malloc: thread-local storage is sometimes improperly free()'d
1722 after being __libc_memalign()'d
1723 [18241] stdio: failed fseek on memstream does not set errno and can fail
1724 when it shouldnt
1725 [18243] nptl: sem_wait, sem_timedwait are cancellation points shm_open is
1726 not
1727 [18463] nptl: pthread_cond_broadcast issue when surrounded by
1728 PTHREAD_PRIO_INHERIT mutex on ARM
1729 [18784] network: res_query and related function crash for special record
1730 type queries (CVE-2015-5180)
1731 [19380] math: strtod does not raise "inexact"
1732 [19387] string: Integer overflow in memchr
1733 [19390] string: Integer overflow in strncat
1734 [19398] build: linknamespace tests fail with massively parallel build
1735 [19402] nptl: Deadlock with robust shared mutex and asynchronous
1736 termination
1737 [19469] malloc: M_PERTURB in test-skeleton.c invalidates malloc tests
1738 [19473] malloc: Turn malloc_get_state etc. in compatibility symbols
1739 [19514] libc: [PATCH] Fix spelling errors in spelling
1740 "implement"/"implementation" in several places
1741 [19582] network: Deprecate RES_USE_INET6
1742 [19673] manual: clog10 docs appear to be erroneous
1743 [19810] dynamic-link: dlopen with both RTLD_NOLOAD and RTLD_NODELETE
1744 causes a segmentation fault
1745 [19826] libc: invalid pointer returned from __tls_get_addr with static
1746 linking
1747 [20016] network: resolv: Remove hooks support from the API
1748 [20019] dynamic-link: NULL pointer dereference in libc.so.6 IFUNC due to
1749 uninitialized GOT
1750 [20033] math: [x86_64] vectorized math function don't call the __finite
1751 versions
1752 [20116] nptl: use after free in pthread_create
1753 [20181] stdio: open_memstream(): writes not at end of stream corrupt data
1754 [20292] dynamic-link: Comparison in elf/dl-open.c _dl_addr_inside_object
1755 is always true.
1756 [20311] nptl: please install proc_service.h
1757 [20366] libc: Compilation errors in installed headers in strict-compliance
1758 modes
1759 [20370] malloc: malloc: Arena free list management is still racy
1760 (incorrect fix in bug 19243)
1761 [20386] libc: assert (X = 0) does not result in GCC warning
1762 [20432] malloc: malloc: Minimize interface required for interposition
1763 [20435] libc: Missing unwind info in __startcontext causes infinite loop
1764 in _Unwind_Backtrace (CVE-2016-6323)
1765 [20444] hurd: recvmsg: PF_LOCAL sockets and msg_name lead to SIGLOST
1766 [20452] nptl: Addition of sysdep.o to libpthread.a breaks relinking
1767 libpthread.a
1768 [20455] math: [powerpc] fesetexceptflag fails to clear FE_INVALID
1769 [20459] localedata: et_EE: locale has wrong {p,n}_cs_precedes value
1770 [20477] network: resolv: incorrect double-checked locking related to
1771 _res_hconf
1772 [20478] libc: libc_ifunc macro and similar usages leads to false debug-
1773 information.
1774 [20495] math: x86_64 performance degradation due to AVX/SSE transition
1775 penalty
1776 [20497] localedata: lt_LT: LC_TIME d_fmt used is obsolete
1777 [20508] dynamic-link: _dl_runtime_resolve_avx/_dl_runtime_profile_avx512
1778 cause transition penalty
1779 [20517] math: sparcv9 missing fdiml compat symbol
1780 [20524] manual: strverscmp is inconsistent
1781 [20525] libc: <sys/quota.h> should be based on kernel headers
1782 [20539] math: GCC 7 -static -lm fails to link at -O3
1783 [20554] libc: ld: bss-plt forced due to /usr/lib/libc_nonshared.a(ppc-
1784 mcount.oS)
1785 [20558] string: POSIX bcopy/bzero decls do not implement Object Size
1786 Checking
1787 [20591] network: Remove obsolete DNSSEC support
1788 [20592] network: DNS resource record type classification macros in
1789 <arpa/nameser.h> are incorrect
1790 [20593] network: Update DNS RR type definitions
1791 [20611] network: getaddrinfo accepts invalid numeric scope IDs
1792 [20615] build: glibc build fails when using --with-cpu=power9 --enable-
1793 multi-arch
1794 [20629] network: libresolv: Remove support for bitstring labels
1795 (RES_USEBSTRING)
1796 [20647] libc: GLIBC quitting every program - glibc on Pentium-S leads to
1797 assertion: "maxidx >= 2"
1798 [20660] math: [arm] Use VSQRT
1799 [20662] libc: checking whether x86_64-pc-linux-gnu-gcc implicitly enables
1800 -fstack-protector no (32bit gcc 6.2.0 pie and ssp enable)
1801 [20689] libc: Test for FMA should also check for AVX.
1802 [20707] glob: gl_pathv entries not set to NULL with GLOB_DOOFFS
1803 [20715] math: iszero macro breaks existing code
1804 [20718] math: [powerpc] copysignl raises "invalid" for sNaN
1805 [20728] libc: powerpc: Missing TOC stub in clone
1806 [20729] build: glibc-2.24 fails to build for i486 with -Os
1807 [20750] build: Build fails with default PIE enabled gcc-6.x
1808 [20768] math: [alpha] sqrt fegetenv namespace
1809 [20785] libc: binutils 2.28 fails to assemble power6/memset.S file in
1810 glibc
1811 [20787] math: float_t is defined as float incorrectly on x86_64 even if
1812 __FLT_EVAL_METHOD__ is 2
1813 [20790] network: rpcgen buffer overrun in get_prog_declaration
1814 [20822] nptl: race condition in __lll_unlock_elision on powerpc
1815 [20829] libc: crypt snprintf namespace
1816 [20847] libc: tst-vfork3 failure
1817 [20855] math: Default bits/mathdef.h has inappropriate float_t
1818 [20859] math: [sh4] FP_ILOGB0 invalid
1819 [20864] localedata: iconv: cp936 missing single-byte euro sign (0x80,
1820 U+20AC), not same as GBK
1821 [20915] dynamic-link: global-dynamic TLS broken on aarch64
1822 [20916] math: pow handling of sNaN arguments
1823 [20918] build: Building with --enable-nss-crypt fails tst-linkall-static.
1824 [20919] math: Bad pow (qNaN, 0) result with -lieee
1825 [20940] math: hypot sNaN handling
1826 [20947] math: fmax, fmin sNaN handling
1827 [20956] libc: debug/tst-backtrace3-6 don't work with -O3 anymore
1828 [20964] network: sunrpc: Stack-based buffer overflow in getrpcport with
1829 RES_USE_INET6
1830 [20971] string: powerpc64/power7 memchr overflows internal pointer check
1831 [20973] nptl: robust mutexes: Lost wake-ups
1832 [20974] locale: bs_BA: yesexpr/noexpr regexes accept any character
1833 [20978] nis: strlen on null pointer in nss_nisplus
1834 [20985] nptl: robust mutexes: lowlevelrobustlock assembly on x86 blocks on
1835 wrong condition
1836 [21014] string: i686 memchr overflows internal pointer check
1837 [21019] libc: [mips] n32 lseek incorrect on overflow
1838 [21022] libc: [microblaze] __backtrace get_frame_size namespace
1839 [21026] libc: [MIPS] readahead syscall is broken on n64
1840 [21028] math: Fallback fesetexceptflag should always succeed
1841 [21045] libc: [powerpc-nofpu] swapcontext does not restore signal mask
1842 [21047] math: arm: fpu_control.h: _FPU_GETCW/_FPU_SETCW is rejected by
1843 clang
1844 [21053] libc: [SH] Namespace pollution from sys/ucontext.h
1845 [21061] librt: [microblaze] librt lost clock_* exports
1846 [21073] libc: tunables: insecure environment variables passed to
1847 subprocesses with AT_SECURE
1848 [21081] string: Missing vzeroupper in memset-vec-unaligned-erms.S
1849 \f
1850 Version 2.24
1851
1852 * The minimum Linux kernel version that this version of the GNU C Library
1853 can be used with is 3.2, except on i[4567]86 and x86_64, where Linux
1854 kernel version 2.6.32 or later suffices (on architectures that already
1855 required kernel versions more recent than 3.2, those requirements remain
1856 unchanged). Linux 3.2 or later kernel headers are required on all
1857 architectures.
1858
1859 * The pap_AN locale has been deleted. This has been deprecated for a long
1860 time. It has been replaced by pap_AW & pap_CW, both of which have long
1861 been included in previous releases.
1862
1863 * The readdir_r and readdir64_r functions have been deprecated. It is
1864 recommended to use readdir and readdir64 instead.
1865
1866 * The type “union wait” has been removed. It was deprecated in the early
1867 1990s and never part of POSIX. Application code should use the int type
1868 instead of “union wait”.
1869
1870 * A new NSS action is added to facilitate large distributed system
1871 administration. The action, MERGE, allows remote user stores like LDAP
1872 to be merged into local user stores like /etc/groups in order to provide
1873 easy to use, updated, and managed sets of merged credentials. The new
1874 action can be used by configuring it in /etc/nsswitch.conf:
1875 group: files [SUCCESS=merge] nis
1876 Implemented by Stephen Gallagher (Red Hat).
1877
1878 * The deprecated __malloc_initialize_hook variable has been removed from the
1879 API.
1880
1881 * The long unused localedef --old-style option has been removed. It hasn't
1882 done anything in over 16 years. Scripts using this option can safely
1883 drop it.
1884
1885 * nextupl, nextup, nextupf, nextdownl, nextdown and nextdownf are added to
1886 libm. They are defined by TS 18661 and IEEE754-2008. The nextup functions
1887 return the next representable value in the direction of positive infinity
1888 and the nextdown functions return the next representable value in the
1889 direction of negative infinity. These are currently enabled as GNU
1890 extensions.
1891
1892 Security related changes:
1893
1894 * An unnecessary stack copy in _nss_dns_getnetbyname_r was removed. It
1895 could result in a stack overflow when getnetbyname was called with an
1896 overly long name. (CVE-2016-3075)
1897
1898 * Previously, getaddrinfo copied large amounts of address data to the stack,
1899 even after the fix for CVE-2013-4458 has been applied, potentially
1900 resulting in a stack overflow. getaddrinfo now uses a heap allocation
1901 instead. Reported by Michael Petlan. (CVE-2016-3706)
1902
1903 * The glob function suffered from a stack-based buffer overflow when it was
1904 called with the GLOB_ALTDIRFUNC flag and encountered a long file name.
1905 Reported by Alexander Cherepanov. (CVE-2016-1234)
1906
1907 * The Sun RPC UDP client could exhaust all available stack space when
1908 flooded with crafted ICMP and UDP messages. Reported by Aldy Hernandez'
1909 alloca plugin for GCC. (CVE-2016-4429)
1910
1911 * The IPv6 name server management code in libresolv could result in a memory
1912 leak for each thread which is created, performs a failing naming lookup,
1913 and exits. Over time, this could result in a denial of service due to
1914 memory exhaustion. Reported by Matthias Schiffer. (CVE-2016-5417)
1915
1916 The following bugs are resolved with this release:
1917
1918 [1170] localedata: ne_NP: update Nepali locale definition file
1919 [3629] manual: stpcpy description in string.texi refers to MS-DOG instead
1920 of MS-DOS.
1921 [6527] malloc: [powerpc] Malloc alignment insufficient for PowerPC
1922 [6796] math: fdim() does not set errno on overflow
1923 [10354] libc: posix_spawn should use vfork() in more cases than presently
1924 [11213] localedata: localedata: add copyright disclaimer to locale files
1925 [12143] localedata: chr_US: new Cherokee locale
1926 [12450] localedata: sgs_LT: new locale
1927 [12676] localedata: ln_CD: new locale
1928 [13237] localedata: LC_ADDRESS.country_name: update all locales w/latest
1929 CLDR data
1930 [13304] math: fma, fmaf, fmal produce wrong results
1931 [14259] build: --localedir arg to configure is ignored
1932 [14499] nptl: Does posix_spawn invoke atfork handlers / use vfork?
1933 [14750] libc: Race condition in posix_spawn vfork usage vs signal handlers
1934 [14934] localedata: es_CL: wrong first weekday chilean locale
1935 [15262] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of
1936 romanisation
1937 [15263] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of 1/0
1938 and +/-
1939 [15264] localedata: LC_MESSAGES.yesstr/nostr: lacking in many locales
1940 [15368] nptl: raise() is not async-signal-safe
1941 [15479] math: ceil, floor, round and trunc raise inexact exception
1942 [15578] localedata: kk_KZ: various updates
1943 [16003] localedata: pap_AN: punt old locale
1944 [16137] localedata: iw_IL: punt old locale
1945 [16190] localedata: eo: new esperanto locale
1946 [16374] localedata: lv_LV: change currency symbol in LC_MONETARY to euro
1947 [16742] malloc: race condition: pthread_atfork() called before first
1948 malloc() results in unexpected locking behaviour/deadlocks
1949 [16975] localedata: LC_MESSAGES.yesexpr/noexpr: revisit capitalization in
1950 all locales
1951 [16983] localedata: postal_fmt does not allow %l and %n modifiers
1952 [17565] localedata: pt_PT: wrong (work-)week start
1953 [17899] math: [powerpc] floorl returns negative zero with FE_DOWNWARD
1954 [17950] build: Build fails with -msse
1955 [18205] localedata: be_BY*: wrong first_weekday and first_workday
1956 [18433] libc: posix_spawn does not return correctly upon failure to
1957 execute
1958 [18453] localedata: charmaps/IBM875: incorrect codes
1959 [18712] string: bits/string2.h incompatible with -O2 -Werror=packed
1960 -Wsystem-headers
1961 [18896] localedata: he_IL: improvements for currency
1962 [18911] localedata: ro_RO: Correcting week day name for "Tuesday" in
1963 Romanian locale data
1964 [18960] locale: s390: _nl_locale_subfreeres uses larl opcode on misaligned
1965 symbol
1966 [19056] libc: Deprecate readdir_r
1967 [19133] localedata: pt_*: days & months should be lowercase in Portuguese
1968 language
1969 [19198] localedata: nl_NL: small improvements for Dutch locales
1970 [19257] network: Per-thread memory leak in __res_vinit with IPv6
1971 nameservers (CVE-2016-5417)
1972 [19269] build: tst-audit4 and tst-audit10 failures with gcc-6 on non avx
1973 machine
1974 [19400] locale: Language missing in "iso-639.def", trivial fix in
1975 description
1976 [19431] malloc: Deadlock between fflush, getdelim, and fork
1977 [19505] libc: Incorrect file descriptor validity checks in
1978 posix_spawn_file_actions_add{open,close,dup2}
1979 [19509] dynamic-link: dlsym, dlvsym do not report errors through dlerror
1980 when using RTLD_NEXT
1981 [19512] locale: Stale `#ifndef HAVE_BUILTIN_EXPECT' in
1982 `intl/{gettextP,loadinfo}.h'
1983 [19534] libc: execle, execlp may use malloc
1984 [19568] localedata: *_CH: Swiss locales have inconsistent start of week
1985 [19573] network: res_nclose and __res_maybe_init disagree about name
1986 server initialization, breaking Hesiod
1987 [19575] localedata: Status of GB18030 tables
1988 [19581] localedata: sr_* date_fmt string contains additional newline
1989 [19583] string: SSSE3_Fast_Copy_Backward flag needs to be enabled for AMD
1990 Excavator core
1991 [19592] math: [ldbl-128ibm] ceill incorrect in non-default rounding modes
1992 [19593] math: [ldbl-128ibm] truncl incorrect in non-default rounding modes
1993 [19594] math: [ldbl-128ibm] roundl incorrect in non-default rounding modes
1994 [19595] math: [ldbl-128ibm] fmodl incorrect for results in subnormal
1995 double range
1996 [19602] math: [ldbl-128ibm] fmodl handling of equal arguments with low
1997 part zero incorrect
1998 [19603] math: [ldbl-128ibm] remainderl, remquol incorrect sign handling in
1999 equality tests
2000 [19610] dynamic-link: ldconfig -X removes stale symbolic links
2001 [19613] libc: s390x (64 bit) macro expansion WCOREDUMP and others
2002 [19633] locale: strfmon_l applies global locale to number formatting
2003 [19642] network: Memory leak in getnameinfo
2004 [19648] libc: test-skeleton.c: Do not set RLIMIT_DATA
2005 [19653] libc: Potential for NULL pointer dereference (CWE-476) in
2006 glibc-2.22
2007 [19654] math: [x86_64] Need testcase for BZ #19590 fix
2008 [19671] localedata: Missing Sanity Check for malloc() in 'tst-fmon.c' &
2009 'tst-numeric.c'
2010 [19674] math: [ldbl-128ibm] powl incorrect overflow handling
2011 [19677] math: [ldbl-128ibm] remainderl equality test incorrect for zero
2012 low part
2013 [19678] math: [ldbl-128ibm] nextafterl, nexttowardl incorrect sign of zero
2014 result
2015 [19679] dynamic-link: gcc-4.9.3 C++ exception handling broken due to
2016 unaligned stack
2017 [19726] locale: Converting UCS4LE to INTERNAL with iconv() does not update
2018 pointers and lengths in error-case.
2019 [19727] locale: Converting from/to UTF-xx with iconv() does not always
2020 report errors on UTF-16 surrogates values.
2021 [19755] nscd: nscd assertion failure in gc
2022 [19758] dynamic-link: Typo in EXTRA_LD_ENVVARS for x86-64
2023 [19759] libc: mempcpy shouldn't be inlined
2024 [19762] dynamic-link: HAS_CPU_FEATURE/HAS_ARCH_FEATURE are easy to misuse
2025 [19765] libc: s390 needs an optimized mempcpy
2026 [19779] glob: glob: buffer overflow with GLOB_ALTDIRFUNC due to incorrect
2027 NAME_MAX limit assumption (CVE-2016-1234)
2028 [19783] build: benchtests don't support --enable-hardcoded-path-in-tests
2029 [19787] network: Missing and incorrect truncation checks in getnameinfo
2030 [19790] math: [ldbl-128ibm] nearbyintl incorrect in non-default rounding
2031 modes
2032 [19791] network: Assertion failure in res_query.c with un-connectable name
2033 server addresses
2034 [19792] libc: MIPS: backtrace yields infinite backtrace with makecontext
2035 [19822] math: libm.so install clobbers old version
2036 [19825] network: resolv: send_vc can return uninitialized data in second
2037 response to getaddrinfo
2038 [19830] network: nss_dns: should check RDATA length against buffer length
2039 [19831] network: nss_dns: getaddrinfo returns uninitialized data when
2040 confronted with A/AAAA records of invalid size
2041 [19837] nss: nss_db: No retries for some long lines with a larger buffer
2042 [19848] math: powl(10,n) for n=-4,-5,-6,-7 is off by more than 1 ULP
2043 [19853] stdio: Printing IBM long double in decimal with high precision is
2044 sometimes incorrect
2045 [19860] build: x86_64: compile errors for tst-audit10 and tst-auditmod10b
2046 [19861] nptl: libpthread IFUNC resolver for fork can lead to crash
2047 [19862] network: resolv, nss_dns: Remove remaining logging of unexpected
2048 record types
2049 [19865] network: Assertion failure or memory leak in
2050 _nss_dns_getcanonname_r
2051 [19868] network: nss_dns: netent code does not skip over non-PTR records
2052 [19879] network: nss_dns: Stack overflow in getnetbyname implementation
2053 (CVE-2016-3075)
2054 [19881] string: Improve x86-64 memset
2055 [19907] string: Incorrect memcpy tests
2056 [19916] dynamic-link: S390: fprs/vrs are not saved/restored while
2057 resolving symbols
2058 [19925] libc: termios.h XCASE namespace
2059 [19928] string: memmove-vec-unaligned-erms.S is slow with large data size
2060 [19929] libc: limits.h NL_NMAX namespace
2061 [19931] stdio: Memory leak in vfprintf
2062 [19957] libc: clone(CLONE_VM) access invalid parent memory
2063 [19963] localedata: en_IL: New locale
2064 [19989] stdio: stdio.h cuserid namespace
2065 [19994] network: getaddrinfo does not restore RES_USE_INET6 flag in
2066 gethosts
2067 [19996] locale: langinfo.h nl_langinfo_l namespace
2068 [20005] stdio: fflush on a file opened with fmemopen resets position to 0
2069 [20010] network: getaddrinfo: Stack overflow in hostent translation
2070 (CVE-2016-3706)
2071 [20012] stdio: libio: fmemopen append mode failure
2072 [20014] stdio: stdio.h namespace for pre-threads POSIX
2073 [20017] network: resolv: Use gmtime_r instead of gmtime in p_secstodate
2074 [20023] libc: fcntl.h timespec namespace
2075 [20024] math: [x86_64] vectorized sincos trashes the stack
2076 [20031] network: nss_hesiod: Heap overflow in get_txt_records
2077 [20041] time: sys/time.h timespec namespace
2078 [20043] libc: unistd.h missing cuserid for UNIX98 and before
2079 [20044] libc: unistd.h missing pthread_atfork for UNIX98
2080 [20051] libc: ttyslot in wrong header under wrong conditions
2081 [20054] libc: gethostname not declared for XPG4
2082 [20055] libc: termios.h missing tcgetsid for XPG4
2083 [20072] dynamic-link: x86 init_cpu_features is called twice in static
2084 executable
2085 [20073] libc: sys/stat.h fchmod namespace
2086 [20074] libc: stdlib.h rand_r namespace
2087 [20076] libc: sys/stat.h missing S_IFSOCK, S_ISSOCK for XPG4
2088 [20094] libc: stdlib.h should not declare grantpt, ptsname, unlockpt for
2089 XPG3
2090 [20111] libc: struct sockaddr_storage cannot be aggregate-copied
2091 [20112] network: sunrpc: stack (frame) overflow in Sun RPC clntudp_call
2092 (CVE-2016-4429)
2093 [20115] string: Extra alignment in memset-vec-unaligned-erms.S
2094 [20119] libc: Wrong mask for processors level type from CPUID
2095 [20139] dynamic-link: Upper part of zmm is zeroed if Glibc is built with
2096 AS not supporting AVX512
2097 [20151] math: [ldbl-128/ldbl-128ibm] j0l, j1l, y0l, y1l return sNaN for
2098 sNaN argument
2099 [20153] math: [ldbl-128ibm] sqrtl (sNaN) returns sNaN
2100 [20156] math: [ldbl-128ibm] ceill, rintl etc. return sNaN for sNaN
2101 argument
2102 [20157] math: [powerpc] fabsl (sNaN) wrongly raises "invalid"
2103 [20160] math: [powerpc] ceil, rint etc. return sNaN for sNaN input
2104 [20178] libc: posix_spawn{p} should not call exit
2105 [20191] stdio: libio: vtables hardening
2106 [20195] string: FMA4 detection requires CPUID execution with register
2107 eax=0x80000001
2108 [20198] libc: quick_exit incorrectly destroys C++11 thread objects.
2109 [20205] math: [i386/x86_64] nextafterl incorrect incrementing negative
2110 subnormals
2111 [20212] math: acos (sNaN) returns sNaN
2112 [20213] math: asin (sNaN) returns sNaN
2113 [20214] network: Linux header sync with linux/in6.h and ipv6.h again.
2114 [20218] math: [i386] asinhl (sNaN) returns sNaN
2115 [20219] math: [i386] atanhl (sNaN) returns sNaN
2116 [20222] stdio: fopencookie: Mangle function pointers
2117 [20224] math: [i386] cbrtl (sNaN) returns sNaN
2118 [20225] math: ldexp, scalbn, scalbln return sNaN for sNaN input
2119 [20226] math: [i386/x86_64] expl, exp10l, expm1l return sNaN for sNaN
2120 input
2121 [20227] math: [i386/x86_64] logl (sNaN) returns sNaN
2122 [20228] math: [i386/x86_64] log10l (sNaN) returns sNaN
2123 [20229] math: [i386/x86_64] log1pl (sNaN) returns sNaN
2124 [20232] math: [ldbl-128] expm1l (sNaN) returns sNaN
2125 [20233] math: [ldbl-128ibm] expm1l (sNaN) returns sNaN
2126 [20234] math: [ldbl-128ibm] log1pl (sNaN) returns sNaN
2127 [20235] math: [i386/x86_64] log2l (sNaN) returns sNaN
2128 [20237] nss: nss_db: get*ent segfaults without preceding set*ent
2129 [20240] math: modf (sNaN) returns sNaN
2130 [20248] libc: debug/tst-longjump_chk2 calls printf from a signal handler
2131 [20250] math: frexp (sNaN) returns sNaN
2132 [20252] math: atan2 (sNaN, qNaN) fails to raise "invalid"
2133 [20255] math: [i386] fdim, fdimf return with excess range and precision /
2134 double rounding
2135 [20256] math: [i386/x86_64] fdiml returns sNaN for sNaN input
2136 [20260] string: ../sysdeps/x86/bits/string.h:1092:3: error: array
2137 subscript is below array bounds [-Werror=array-bounds]
2138 [20262] nis: _nss_nis_initgroups_dyn always returns NSS_STATUS_NOTFOUND
2139 [20263] nptl: robust mutex deadlocks if other thread requests timedlock
2140 (Only arm/linux)
2141 [20277] libc: $dp is not initialized correctly in sysdeps/hppa/start.S
2142 [20284] malloc: malloc: Corrupt arena avoidance causes unnecessary mmap
2143 fallbacks
2144 [20296] math: [i386/x86_64] scalbl returns sNaN for sNaN input, missing
2145 "invalid" exceptions
2146 [20314] nptl: make[4]: *** [/usr/include/stdlib.h] Error 1
2147 [20316] localedata: id_ID: Februari instead of Pebruari
2148 [20327] string: POWER8 strcasecmp returns incorrect result
2149 [20347] math: Failure: Test: j0_downward (0xap+0)
2150 [20348] libc: FAIL: misc/tst-preadvwritev64
2151 [20349] libc: 64-bit value is passed differently in p{readv,writev}{64}
2152 [20350] libc: There is no test for p{read,write}64
2153 [20357] math: Incorrect cos result for 1.5174239687223976
2154 [20384] build: Don't run libmvec-sincos-avx* tests on non avx machines
2155 \f
2156 Version 2.23
2157
2158 * Unicode 8.0.0 Support: Character encoding, character type info, and
2159 transliteration tables are all updated to Unicode 8.0.0, using new
2160 and/or improved generator scripts contributed by Mike FABIAN (Red Hat).
2161 These updates cause user visible changes, such as the fixes for bugs
2162 89, 16061, and 18568.
2163
2164 * sched_setaffinity, pthread_setaffinity_np no longer attempt to guess the
2165 kernel-internal CPU set size. This means that requests that change the
2166 CPU affinity which failed before (for example, an all-ones CPU mask) will
2167 now succeed. Applications that need to determine the effective CPU
2168 affinities need to call sched_getaffinity or pthread_getaffinity_np after
2169 setting it because the kernel can adjust it (and the previous size check
2170 would not detect this in the majority of cases).
2171
2172 * The fts.h header can now be used with -D_FILE_OFFSET_BITS=64. With LFS
2173 the following new symbols are used: fts64_children, fts64_close,
2174 fts64_open, fts64_read and fts64_set.
2175
2176 * getaddrinfo now detects certain invalid responses on an internal netlink
2177 socket. If such responses are received, an affected process will
2178 terminate with an error message of "Unexpected error <number> on netlink
2179 descriptor <number>" or "Unexpected netlink response of size <number> on
2180 descriptor <number>". The most likely cause for these errors is a
2181 multi-threaded application which erroneously closes and reuses the netlink
2182 file descriptor while it is used by getaddrinfo.
2183
2184 * A defect in the malloc implementation, present since glibc 2.15 (2012) or
2185 glibc 2.10 via --enable-experimental-malloc (2009), could result in the
2186 unnecessary serialization of memory allocation requests across threads.
2187 The defect is now corrected. Users should see a substantial increase in
2188 the concurent throughput of allocation requests for applications which
2189 trigger this bug. Affected applications typically create and
2190 destroy threads frequently. (Bug 19048 was reported and analyzed by
2191 Ericsson.)
2192
2193 * There is now a --disable-timezone-tools configure option for disabling the
2194 building and installing of the timezone related utilities (zic, zdump, and
2195 tzselect). This is useful for people who build the timezone data and code
2196 independent of the GNU C Library.
2197
2198 * The obsolete header <regexp.h> has been removed. Programs that require
2199 this header must be updated to use <regex.h> instead.
2200
2201 * The obsolete functions bdflush, create_module, get_kernel_syms,
2202 query_module and uselib are no longer available to newly linked binaries;
2203 the header <sys/kdaemon.h> has been removed. These functions and header
2204 were specific to systems using the Linux kernel and could not usefully be
2205 used with the GNU C Library on systems with version 2.6 or later of the
2206 Linux kernel.
2207
2208 * Optimized string, wcsmbs and memory functions for IBM z13.
2209 Implemented by Stefan Liebler.
2210
2211 * Newly linked programs that define a variable called signgam will no longer
2212 have it set by the lgamma, lgammaf and lgammal functions. Programs that
2213 require signgam to be set by those functions must ensure that they use the
2214 variable provided by the GNU C Library and declared in <math.h>, without
2215 defining their own copy.
2216
2217 * The minimum GCC version that can be used to build this version of the GNU
2218 C Library is GCC 4.7. Older GCC versions, and non-GNU compilers, can
2219 still be used to compile programs using the GNU C Library.
2220
2221 Security related changes:
2222
2223 * An out-of-bounds value in a broken-out struct tm argument to strftime no
2224 longer causes a crash. Reported by Adam Nielsen. (CVE-2015-8776)
2225
2226 * The LD_POINTER_GUARD environment variable can no longer be used to disable
2227 the pointer guard feature. It is always enabled. Previously,
2228 LD_POINTER_GUARD could be used to disable security hardening in binaries
2229 running in privileged AT_SECURE mode. Reported by Hector Marco-Gisbert.
2230 (CVE-2015-8777)
2231
2232 * An integer overflow in hcreate and hcreate_r could lead to an
2233 out-of-bounds memory access. Reported by Szabolcs Nagy. (CVE-2015-8778)
2234
2235 * The catopen function no longer has unbounded stack usage. Reported by
2236 Max. (CVE-2015-8779)
2237
2238 * The nan, nanf and nanl functions no longer have unbounded stack usage
2239 depending on the length of the string passed as an argument to the
2240 functions. Reported by Joseph Myers. (CVE-2014-9761)
2241
2242 * A stack-based buffer overflow was found in libresolv when invoked from
2243 libnss_dns, allowing specially crafted DNS responses to seize control
2244 of execution flow in the DNS client. The buffer overflow occurs in
2245 the functions send_dg (send datagram) and send_vc (send TCP) for the
2246 NSS module libnss_dns.so.2 when calling getaddrinfo with AF_UNSPEC
2247 family. The use of AF_UNSPEC triggers the low-level resolver code to
2248 send out two parallel queries for A and AAAA. A mismanagement of the
2249 buffers used for those queries could result in the response of a query
2250 writing beyond the alloca allocated buffer created by
2251 _nss_dns_gethostbyname4_r. Buffer management is simplified to remove
2252 the overflow. Thanks to the Google Security Team and Red Hat for
2253 reporting the security impact of this issue, and Robert Holiday of
2254 Ciena for reporting the related bug 18665. (CVE-2015-7547)
2255
2256 The following bugs are resolved with this release:
2257
2258 [89] localedata: Locales nb_NO and nn_NO should transliterate æøå
2259 [887] math: Math library function "logb" and "nextafter" inconsistent
2260 [2542] math: Incorrect return from float gamma (-0X1.FA471547C2FE5P+1)
2261 [2543] math: Incorrect return from float gamma (-0X1.9260DCP+1)
2262 [2558] math: Incorrect return from double gamma (-0X1.FA471547C2FE5P+1)
2263 [2898] libc: [improve] warning: the use of `mktemp' is dangerous, better
2264 use `mkstemp'
2265 [4404] localedata: German translation of "Alarm clock" is misleading
2266 [6799] math: nextafter() and nexttoward() doen't set errno on
2267 overflow/underflow errors
2268 [6803] math: scalb(), scalbln(), scalbn() do not set errno on
2269 overflow/underflow
2270 [10432] nis: _nss_nis_setnetgrent assertion failure
2271 [11460] libc: fts has no LFS support
2272 [12926] network: getaddrinfo()/make_request() may spin forever
2273 [13065] nptl: Race condition in pthread barriers
2274 [13690] nptl: pthread_mutex_unlock potentially cause invalid access
2275 [14341] dynamic-link: Dynamic linker crash when DT_JMPREL and DT_REL{,A}
2276 are not contiguous
2277 [14551] math: [ldbl-128ibm] strtold overflow handling for IBM long double
2278 [14912] libc: Rename non-installed bits/*.h headers
2279 [15002] libc: Avoid undefined behavior in posix_fallocate overflow check
2280 [15367] math: Let gcc use __builtin_isinf
2281 [15384] math: One constant fewer in ieee754/dbl-64/wordsize-64/s_finite.c
2282 [15421] math: lgamma wrongly sets signgam for ISO C
2283 [15470] math: [arm] On ARM llrintl() and llroundl() do not raise
2284 FE_INVALID with argument out of range
2285 [15491] math: [i386/x86_64] x86 nearbyint implementations wrongly clear
2286 all exceptions
2287 [15786] dynamic-link: ifunc resolver functions can smash function
2288 arguments
2289 [15918] math: Unnecessary check for equality in hypotf()
2290 [16061] localedata: Review / update transliteration data
2291 [16068] math: [i386/x86_64] x86 and x86_64 fesetenv exclude state they
2292 should include
2293 [16141] time: strptime %z offset restriction
2294 [16171] math: drem should be alias of remainder
2295 [16296] math: fegetround is pure?
2296 [16347] math: [ldbl-128ibm] ldbl-128/e_lgammal_r.c may not be suitable.
2297 [16364] libc: sleep may leave SIGCHLD blocked on sync cancellation on
2298 GNU/Linux
2299 [16399] math: [mips] lrint / llrint / lround / llround missing exceptions
2300 [16415] math: Clean up ldbl-128 / ldbl-128ibm expm1l for large positive
2301 arguments
2302 [16422] math: [powerpc] math-float, math-double failing llrint tests with
2303 "Exception "Inexact" set" on ppc32
2304 [16495] localedata: nl_NL: date_fmt: shuffle year/month around
2305 [16517] math: Missing underflow exception from tanf/tan/tanl
2306 [16519] math: Missing underflow exception from sinhf
2307 [16520] math: Missing underflow exception from tanhf
2308 [16521] math: Missing underflow exception from exp2
2309 [16620] math: [ldbl-128ibm] exp10l spurious overflows / bad directed
2310 rounding results
2311 [16734] stdio: fopen calls mmap to allocate its buffer
2312 [16961] math: nan function incorrect handling of bad sequences
2313 [16962] math: nan function unbounded stack allocation (CVE-2014-9761)
2314 [16973] localedata: Fix lang_lib/lang_term as per ISO 639-2
2315 [16985] locale: localedef: confusing error message when opening output
2316 fails
2317 [17118] math: ctanh(INFINITY + 2 * I) returns incorrect value
2318 [17197] locale: Redundant shift character in iconv conversion output at
2319 block boundary
2320 [17243] libc: trunk/posix/execl.c:53: va_args problem ?
2321 [17244] libc: trunk/sysdeps/unix/sysv/linux/semctl.c:116: va_args muxup ?
2322 [17250] dynamic-link: static linking breaks nss loading
2323 (getaddrinfo/getpwnam/etc...)
2324 [17404] libc: atomic_exchange_rel lacking a barrier on MIPS16, GCC before
2325 4.7?
2326 [17441] math: isnan() should use __builtin_isnan() in GCC
2327 [17514] nptl: Assert failure unlocking ERRORCHECK mutex after timedlock
2328 (related to lock elision)
2329 [17787] manual: Exponent on page 324 of the PDF ends prematurely
2330 [17886] time: strptime should be able to parse "Z" as a timezone with %z
2331 [17887] time: strptime should be able to parse "+01:00" style timezones
2332 [17905] libc: catopen() Multiple unbounded stack allocations
2333 (CVE-2015-8779)
2334 [18084] libc: backtrace (..., 0) dumps core on x86
2335 [18086] libc: nice() sets errno to 0 on success
2336 [18240] libc: hcreate, hcreate_r should fail with ENOMEM if element count
2337 is too large (CVE-2015-8778)
2338 [18251] dynamic-link: SONAME missing when audit modules provides path
2339 [18265] libc: add attributes for wchar string and memory functions
2340 [18370] math: csqrt missing underflows
2341 [18421] libc: [hppa] read-only segment has dynamic relocations
2342 [18472] libc: Obsolete syscall wrappers should be compat symbols
2343 [18480] libc: hppa glibc miscompilation in sched_setaffinity()
2344 [18491] localedata: Update tr_TR LC_CTYPE as part of Unicode updates
2345 [18525] localedata: Remove locale timezone information
2346 [18560] libc: [powerpc] spurious bits/ipc.h definitions
2347 [18568] localedata: Update locale data to Unicode 8.0
2348 [18589] locale: sort-test.sh fails at random
2349 [18595] math: ctan, ctanh missing underflows
2350 [18604] libc: assert macro-expands its argument
2351 [18610] math: S390: fetestexcept() reports any exception if DXC-code
2352 contains a vector instruction exception.
2353 [18611] math: j1, jn missing errno setting on underflow
2354 [18618] localedata: sync Chechen locale definitions with other *_RU
2355 locales
2356 [18647] math: powf(-0x1.000002p0, 0x1p30) returns 0 instead of +inf
2357 [18661] libc: Some x86-64 assembly codes don't align stack to 16 bytes
2358 [18665] network: In send_dg, the recvfrom function is NOT always using the
2359 buffer size of a newly created buffer (CVE-2015-7547)
2360 [18674] libc: [i386] trunk/sysdeps/i386/tst-auditmod3b.c:84: possible
2361 missing break ?
2362 [18675] libc: fpathconf(_PC_NAME_MAX) fails against large filesystems for
2363 32bit processes
2364 [18681] libc: regexp.h is obsolete and buggy, and should be desupported
2365 [18699] math: tilegx cproj() for various complex infinities does not yield
2366 infinity
2367 [18724] libc: Harden put*ent functions against data injection
2368 [18743] nptl: PowerPC: findutils testcase fails with --enable-lock-elision
2369 [18755] build: build errors with -DNDEBUG
2370 [18757] stdio: fmemopen fails to set errno on failure
2371 [18778] dynamic-link: ld.so crashes if failed dlopen causes libpthread to
2372 be forced unloaded
2373 [18781] libc: openat64 lacks O_LARGEFILE
2374 [18787] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/atomic.h:71:6:
2375 error: can’t find a register in class ‘R1_REGS’ while reloading ‘asm’
2376 [18789] math: [ldbl-128ibm] sinhl inaccurate near 0
2377 [18790] math: [ldbl-128ibm] tanhl inaccurate
2378 [18795] libc: stpncpy fortification misses buffer lengths that are
2379 statically too large
2380 [18796] build: build fails for --disable-mathvec
2381 [18803] math: hypot missing underflows
2382 [18820] stdio: fmemopen may leak memory on failure
2383 [18823] math: csqrt spurious underflows
2384 [18824] math: fma spurious underflows
2385 [18825] math: pow missing underflows
2386 [18857] math: [ldbl-128ibm] nearbyintl wrongly uses signaling comparisons
2387 [18868] nptl: pthread_barrier_init typo has in-theory-undefined behavior
2388 [18870] build: sem_open.c fails to compile with missing symbol
2389 FUTEX_SHARED
2390 [18872] stdio: Fix memory leak in printf_positional
2391 [18873] libc: posix_fallocate overflow check ineffective
2392 [18875] math: Excess precision leads incorrect libm
2393 [18877] libc: arm: mmap offset regression
2394 [18887] libc: memory corruption when using getmntent on blank lines
2395 [18918] localedata: hu_HU: change time to HH:MM:SS format
2396 [18921] libc: Regression: extraneous stat() and fstat() performed by
2397 opendir()
2398 [18928] dynamic-link: LD_POINTER_GUARD is not ignored for privileged
2399 binaries (CVE-2015-8777)
2400 [18951] math: tgamma missing underflows
2401 [18952] math: [ldbl-128/ldbl-128ibm] lgammal spurious "invalid", incorrect
2402 signgam
2403 [18953] localedata: lt_LT: change currency symbol to the euro
2404 [18956] math: powf inaccuracy
2405 [18961] math: [i386] exp missing underflows
2406 [18966] math: [i386] exp10 missing underflows
2407 [18967] math: math.h XSI POSIX namespace (gamma, isnan, scalb)
2408 [18969] build: multiple string test failures due to missing locale
2409 dependencies
2410 [18970] libc: Reference of pthread_setcancelstate in libc.a
2411 [18977] math: float / long double Bessel functions not in XSI POSIX
2412 [18980] math: i386 libm functions return with excess range and precision
2413 [18981] math: i386 scalb*, ldexp return with excess range and precision
2414 [18982] stdio: va_list and vprintf
2415 [18985] time: Passing out of range data to strftime() causes a segfault
2416 (CVE-2015-8776)
2417 [19003] math: [x86_64] fma4 version of pow inappropriate contraction
2418 [19007] libc: FAIL: elf/check-localplt with -z now and binutils 2.26
2419 [19012] locale: iconv_open leaks memory on error path
2420 [19016] math: clog, clog10 inaccuracy
2421 [19018] nptl: Mangle function pointers in tls_dtor_list
2422 [19032] math: [i386] acosh (-qNaN) spurious "invalid" exception
2423 [19046] math: ldbl-128 / ldbl-128ibm lgamma bad overflow handling
2424 [19048] malloc: malloc: arena free list can become cyclic, increasing
2425 contention
2426 [19049] math: [powerpc] erfc incorrect zero sign
2427 [19050] math: [powerpc] log* incorrect zero sign
2428 [19058] math: [x86_64] Link fail with -fopenmp and -flto
2429 [19059] math: nexttoward overflow incorrect in non-default rounding modes
2430 [19071] math: ldbl-96 lroundl incorrect just below powers of 2
2431 [19074] network: Data race in _res_hconf_reorder_addrs
2432 [19076] math: [ldbl-128ibm] log1pl (-1) wrong sign of infinity
2433 [19077] math: [ldbl-128ibm] logl (1) incorrect sign of zero result
2434 [19078] math: [ldbl-128ibm] expl overflow incorrect in non-default
2435 rounding modes
2436 [19079] math: dbl-64/wordsize-64 lround based on llround incorrect for
2437 ILP32
2438 [19085] math: ldbl-128 lrintl, lroundl missing exceptions for 32-bit long
2439 [19086] manual: posix_fallocate64 documented argument order is wrong.
2440 [19088] math: lround, llround missing exceptions close to overflow
2441 threshold
2442 [19094] math: lrint, llrint missing exceptions close to overflow threshold
2443 [19095] math: dbl-64 lrint incorrect for 64-bit long
2444 [19122] dynamic-link: Unnecessary PLT relocations in librtld.os
2445 [19124] dynamic-link: ld.so failed to build with older assmebler
2446 [19125] math: [powerpc32] llroundf, llround incorrect exceptions
2447 [19129] dynamic-link: [arm] Concurrent lazy TLSDESC resolution can crash
2448 [19134] math: [powerpc32] lround, lroundf spurious exceptions
2449 [19137] libc: i386/epoll_pwait.S doesn't support cancellation
2450 [19143] nptl: Remove CPU set size checking from sched_setaffinity,
2451 pthread_setaffinity_np
2452 [19156] math: [ldbl-128] j0l spurious underflows
2453 [19164] nptl: tst-getcpu fails with many possible CPUs
2454 [19168] math: math/test-ildoubl and math/test-ldouble failure
2455 [19174] nptl: PowerPC: TLE enabled pthread mutex performs poorly.
2456 [19178] dynamic-link: ELF_RTYPE_CLASS_EXTERN_PROTECTED_DATA confuses
2457 prelink
2458 [19181] math: [i386/x86_64] fesetenv (FE_DFL_ENV), fesetenv
2459 (FE_NOMASK_ENV) do not clear SSE exceptions
2460 [19182] malloc: malloc deadlock between ptmalloc_lock_all and
2461 _int_new_arena/reused_arena
2462 [19189] math: [ldbl-128] log1pl (-qNaN) spurious "invalid" exception
2463 [19201] math: dbl-64 remainder incorrect sign of zero result
2464 [19205] math: bits/math-finite.h conditions do not match math.h and
2465 bits/mathcalls.h
2466 [19209] math: bits/math-finite.h wrongly maps ldexp to scalbn
2467 [19211] math: lgamma functions do not set signgam for -ffinite-math-only
2468 for C99-based standards
2469 [19212] libc: features.h not -Wundef clean
2470 [19213] math: [i386/x86_64] log* (1) incorrect zero sign for -ffinite-
2471 math-only
2472 [19214] libc: Family and model identification for AMD CPU's are incorrect.
2473 [19219] libc: GLIBC build fails for ia64 with missing __nearbyintl
2474 [19228] math: [powerpc] nearbyint wrongly clears "inexact", leaves traps
2475 disabled
2476 [19235] math: [powerpc64] lround, lroundf, llround, llroundf spurious
2477 "inexact" exceptions
2478 [19238] math: [powerpc] round, roundf spurious "inexact" for integer
2479 arguments
2480 [19242] libc: strtol incorrect in Turkish locales
2481 [19243] malloc: reused_arena can pick an arena on the free list, leading
2482 to an assertion failure and reference count corruption
2483 [19253] time: tzset() ineffective when temporary TZ did not include DST
2484 rules
2485 [19266] math: strtod ("NAN(I)") incorrect in Turkish locales
2486 [19270] math: [hppa] Shared libm missing __isnanl
2487 [19285] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/mman.h: missing
2488 MAP_HUGETLB and MAP_STACK defines
2489 [19313] nptl: Wrong __cpu_mask for x32
2490 [19347] libc: grantpt: try to force a specific gid even without pt_chown
2491 [19349] math: [ldbl-128ibm] tanhl inaccurate for small arguments
2492 [19350] math: [ldbl-128ibm] sinhl spurious overflows
2493 [19351] math: [ldbl-128ibm] logl inaccurate near 1
2494 [19363] time: x32: times() return value wrongly truncates/sign extends
2495 from 32bit
2496 [19367] dynamic-link: Improve branch prediction on Silvermont
2497 [19369] network: Default domain name not reset by res_ninit when "search"
2498 / "domain" entry is removed from resolv.conf
2499 [19375] math: powerpc: incorrect results for POWER7 logb with negative
2500 subnormals
2501 [19385] localedata: bg_BG: time separator should be colon, not comma
2502 [19408] libc: linux personality syscall wrapper may erroneously return an
2503 error on 32-bit architectures
2504 [19415] libc: dladdr returns wrong names on hppa
2505 [19432] libc: iconv rejects redundant escape sequences in IBM900, IBM903,
2506 IBM905, IBM907, and IBM909
2507 [19439] math: Unix98 isinf and isnan functions conflict with C++11
2508 [19443] build: build failures with -DDEBUG
2509 [19451] build: Make check fails on test-double-vlen2
2510 [19462] libc: Glibc failed to build with -Os
2511 [19465] math: Wrong code with -Os
2512 [19466] time: time/tst-mktime2.c is compiled into an infinite loop with
2513 -Os
2514 [19467] string: Fast_Unaligned_Load needs to be enabled for Excavator core
2515 CPU's.
2516 [19475] libc: Glibc 2.22 doesn't build on sparc [PATCH]
2517 [19486] math: S390: Math tests fail with "Exception Inexact set".
2518 [19529] libc: [ARM]: FAIL: stdlib/tst-makecontext
2519 [19550] libc: [mips] mmap negative offset handling inconsistent with other
2520 architectures
2521 [19590] math: Fail to build shared objects that use libmvec.so functions.
2522 \f
2523 Version 2.22
2524
2525 * The following bugs are resolved with this release:
2526
2527 438, 2981, 4719, 6544, 6792, 11216, 12836, 13028, 13064, 13151, 13152,
2528 14094, 14113, 14292, 14841, 14906, 14958, 15319, 15467, 15790, 15969,
2529 16159, 16339, 16350, 16351, 16352, 16353, 16361, 16512, 16526, 16538,
2530 16559, 16560, 16704, 16783, 16850, 17053, 17090, 17195, 17269, 17293,
2531 17322, 17403, 17475, 17523, 17542, 17569, 17581, 17588, 17596, 17620,
2532 17621, 17628, 17631, 17692, 17711, 17715, 17776, 17779, 17792, 17833,
2533 17836, 17841, 17912, 17916, 17930, 17932, 17944, 17949, 17964, 17965,
2534 17967, 17969, 17977, 17978, 17987, 17991, 17996, 17998, 17999, 18007,
2535 18019, 18020, 18029, 18030, 18032, 18034, 18036, 18038, 18039, 18042,
2536 18043, 18046, 18047, 18049, 18068, 18078, 18080, 18093, 18100, 18104,
2537 18110, 18111, 18116, 18125, 18128, 18134, 18138, 18185, 18196, 18197,
2538 18206, 18210, 18211, 18217, 18219, 18220, 18221, 18234, 18244, 18245,
2539 18247, 18287, 18319, 18324, 18333, 18346, 18371, 18383, 18397, 18400,
2540 18409, 18410, 18412, 18418, 18422, 18434, 18444, 18457, 18468, 18469,
2541 18470, 18479, 18483, 18495, 18496, 18497, 18498, 18502, 18507, 18508,
2542 18512, 18513, 18519, 18520, 18522, 18527, 18528, 18529, 18530, 18532,
2543 18533, 18534, 18536, 18539, 18540, 18542, 18544, 18545, 18546, 18547,
2544 18549, 18553, 18557, 18558, 18569, 18583, 18585, 18586, 18592, 18593,
2545 18594, 18602, 18612, 18613, 18619, 18633, 18635, 18641, 18643, 18648,
2546 18657, 18676, 18694, 18696, 18887.
2547
2548 * Cache information can be queried via sysconf() function on s390 e.g. with
2549 _SC_LEVEL1_ICACHE_SIZE as argument.
2550
2551 * A buffer overflow in gethostbyname_r and related functions performing DNS
2552 requests has been fixed. If the NSS functions were called with a
2553 misaligned buffer, the buffer length change due to pointer alignment was
2554 not taken into account. This could result in application crashes or,
2555 potentially arbitrary code execution, using crafted, but syntactically
2556 valid DNS responses. (CVE-2015-1781)
2557
2558 * The time zone file parser has been made more robust against crafted time
2559 zone files, avoiding heap buffer overflows related to the processing of
2560 the tzh_ttisstdcnt and tzh_ttisgmtcnt fields, and a stack overflow due to
2561 large time zone data files. Overly long time zone specifiers in the TZ
2562 variable no longer result in stack overflows and crashes.
2563
2564 * A powerpc and powerpc64 optimization for TLS, similar to TLS descriptors
2565 for LD and GD on x86 and x86-64, has been implemented. You will need
2566 binutils-2.24 or later to enable this optimization.
2567
2568 * Character encoding and ctype tables were updated to Unicode 7.0.0, using
2569 new generator scripts contributed by Pravin Satpute and Mike FABIAN (Red
2570 Hat). These updates cause user visible changes, such as the fix for bug
2571 17998.
2572
2573 * CVE-2014-8121 The NSS backends shared internal state between the getXXent
2574 and getXXbyYY NSS calls for the same database, causing a denial-of-service
2575 condition in some applications.
2576
2577 * Added vector math library named libmvec with the following vectorized x86_64
2578 implementations: cos, cosf, sin, sinf, sincos, sincosf, log, logf, exp, expf,
2579 pow, powf.
2580 The library can be disabled with --disable-mathvec. Use of the functions is
2581 enabled with -fopenmp -ffast-math starting from -O1 for GCC version >= 4.9.0.
2582 Shared library libmvec.so is linked in as needed when using -lm (no need to
2583 specify -lmvec explicitly for not static builds).
2584 Visit <https://sourceware.org/glibc/wiki/libmvec> for detailed information.
2585
2586 * A new fmemopen implementation has been added with the goal of POSIX
2587 compliance. The new implementation fixes the following long-standing
2588 issues: BZ#6544, BZ#11216, BZ#12836, BZ#13151, BZ#13152, and BZ#14292. The
2589 old implementation is still present for use be by existing binaries.
2590
2591 * The 32-bit sparc sigaction ABI was inadvertently broken in the 2.20 and 2.21
2592 releases. It has been fixed to match 2.19 and older, but binaries built
2593 against 2.20 and 2.21 might need to be recompiled. See BZ#18694.
2594
2595 * Port to Native Client running on ARMv7-A (--host=arm-nacl).
2596 Contributed by Roland McGrath (Google).
2597
2598 * The header <regexp.h> is deprecated, and will be removed in a future
2599 release. Use of this header will trigger a deprecation warning.
2600 Application developers should update their code to use <regex.h> instead.
2601
2602 This header was formerly part of SUS, but was deprecated in 1994 and
2603 removed from the standard in 2001. Also, the glibc implementation
2604 leaks memory. See BZ#18681 for more details.
2605 \f
2606 Version 2.21
2607
2608 * The following bugs are resolved with this release:
2609
2610 6652, 10672, 12674, 12847, 12926, 13862, 14132, 14138, 14171, 14498,
2611 15215, 15378, 15884, 16009, 16418, 16191, 16469, 16576, 16617, 16618,
2612 16619, 16657, 16740, 16857, 17192, 17266, 17273, 17344, 17363, 17370,
2613 17371, 17411, 17460, 17475, 17485, 17501, 17506, 17508, 17522, 17555,
2614 17570, 17571, 17572, 17573, 17574, 17582, 17583, 17584, 17585, 17589,
2615 17594, 17601, 17608, 17616, 17625, 17630, 17633, 17634, 17635, 17647,
2616 17653, 17657, 17658, 17664, 17665, 17668, 17682, 17702, 17717, 17719,
2617 17722, 17723, 17724, 17725, 17732, 17733, 17744, 17745, 17746, 17747,
2618 17748, 17775, 17777, 17780, 17781, 17782, 17791, 17793, 17796, 17797,
2619 17801, 17803, 17806, 17834, 17844, 17848, 17868, 17869, 17870, 17885,
2620 17892.
2621
2622 * CVE-2015-1472 CVE-2015-1473 Under certain conditions wscanf can allocate
2623 too little memory for the to-be-scanned arguments and overflow the
2624 allocated buffer. The implementation now correctly computes the required
2625 buffer size when using malloc, and switches to malloc from alloca as
2626 intended.
2627
2628 * A new semaphore algorithm has been implemented in generic C code for all
2629 machines. Previous custom assembly implementations of semaphore were
2630 difficult to reason about or ensure that they were safe. The new version
2631 of semaphore supports machines with 64-bit or 32-bit atomic operations.
2632 The new semaphore algorithm is used by sem_init, sem_open, sem_post,
2633 sem_wait, sem_timedwait, sem_trywait, and sem_getvalue.
2634
2635 * Port to Altera Nios II has been contributed by Mentor Graphics.
2636
2637 * Optimized strcpy, stpcpy, strncpy, stpncpy, strcmp, and strncmp
2638 implementations for powerpc64/powerpc64le.
2639 Implemented by Adhemerval Zanella (IBM).
2640
2641 * Added support for TSX lock elision of pthread mutexes on powerpc32, powerpc64
2642 and powerpc64le. This may improve lock scaling of existing programs on
2643 HTM capable systems. The lock elision code is only enabled with
2644 --enable-lock-elision=yes. Also, the TSX lock elision implementation for
2645 powerpc will issue a transaction abort on every syscall to avoid side
2646 effects being visible outside transactions.
2647
2648 * Optimized strcpy, stpcpy, strchrnul and strrchr implementations for
2649 AArch64. Contributed by ARM Ltd.
2650
2651 * i386 memcpy functions optimized with SSE2 unaligned load/store.
2652
2653 * CVE-2104-7817 The wordexp function could ignore the WRDE_NOCMD flag
2654 under certain input conditions resulting in the execution of a shell for
2655 command substitution when the application did not request it. The
2656 implementation now checks WRDE_NOCMD immediately before executing the
2657 shell and returns the error WRDE_CMDSUB as expected.
2658
2659 * CVE-2012-3406 printf-style functions could run into a stack overflow when
2660 processing format strings with a large number of format specifiers.
2661
2662 * CVE-2014-9402 The nss_dns implementation of getnetbyname could run into an
2663 infinite loop if the DNS response contained a PTR record of an unexpected
2664 format.
2665
2666 * The minimum GCC version that can be used to build this version of the GNU
2667 C Library is GCC 4.6. Older GCC versions, and non-GNU compilers, can
2668 still be used to compile programs using the GNU C Library.
2669
2670 * The GNU C Library is now built with -Werror by default. This can be
2671 disabled by configuring with --disable-werror.
2672
2673 * New locales: tu_IN, bh_IN, raj_IN, ce_RU.
2674
2675 * The obsolete sigvec function has been removed. This was the original
2676 4.2BSD interface that inspired the POSIX.1 sigaction interface, which
2677 programs have been using instead for about 25 years. Of course, ABI
2678 compatibility for old binaries using sigvec remains intact.
2679
2680 * Merged gettext 0.19.3 into the intl subdirectory. This fixes building
2681 with newer versions of bison.
2682
2683 * Support for MIPS o32 FPXX, FP64A and FP64 ABI Extensions.
2684 The original MIPS o32 hard-float ABI requires an FPU where double-precision
2685 registers overlay two consecutive single-precision registers. MIPS32R2
2686 introduced a new FPU mode (FR=1) where double-precision registers extend the
2687 corresponding single-precision registers which is incompatible with the
2688 o32 hard-float ABI. The MIPS SIMD ASE and the MIPSR6 architecture both
2689 require the use of FR=1 making a transition necessary. New o32 ABI
2690 extensions enable users to migrate over time from the original o32 ABI
2691 through to the updated o32 FP64 ABI. To achieve this the dynamic linker now
2692 tracks the ABI of any loaded object and verifies that new objects are
2693 compatible. Mode transitions will also be requested as required and
2694 unsupportable objects will be rejected. The ABI checks include both soft and
2695 hard float ABIs for o32, n32 and n64.
2696
2697 GCC 5 with GNU binutils 2.25 onwards:
2698 It is strongly recommended that all o32 system libraries are built using the
2699 new o32 FPXX ABI (-mfpxx) to facilitate the transition as this is compatible
2700 with the original and all new o32 ABI extensions. Configure a MIPS GCC
2701 compiler using --with-fp-32=xx to set this by default.
2702 \f
2703 Version 2.20
2704
2705 * The following bugs are resolved with this release:
2706
2707 6804, 9894, 11505, 12994, 13347, 13651, 14308, 14770, 15119, 15132, 15347,
2708 15514, 15698, 15804, 15894, 15946, 16002, 16064, 16095, 16194, 16198,
2709 16275, 16284, 16287, 16315, 16348, 16349, 16354, 16357, 16362, 16447,
2710 16516, 16532, 16539, 16545, 16561, 16562, 16564, 16574, 16599, 16600,
2711 16609, 16610, 16611, 16613, 16619, 16623, 16629, 16632, 16634, 16639,
2712 16642, 16648, 16649, 16670, 16674, 16677, 16680, 16681, 16683, 16689,
2713 16695, 16701, 16706, 16707, 16712, 16713, 16714, 16724, 16731, 16739,
2714 16740, 16743, 16754, 16758, 16759, 16760, 16770, 16786, 16789, 16791,
2715 16796, 16799, 16800, 16815, 16823, 16824, 16831, 16838, 16839, 16849,
2716 16854, 16876, 16877, 16878, 16882, 16885, 16888, 16890, 16892, 16912,
2717 16915, 16916, 16917, 16918, 16922, 16927, 16928, 16932, 16943, 16958,
2718 16965, 16966, 16967, 16977, 16978, 16984, 16990, 16996, 17009, 17022,
2719 17031, 17042, 17048, 17050, 17058, 17061, 17062, 17069, 17075, 17078,
2720 17079, 17084, 17086, 17088, 17092, 17097, 17125, 17135, 17137, 17150,
2721 17153, 17187, 17213, 17259, 17261, 17262, 17263, 17319, 17325, 17354.
2722
2723 * Reverted change of ABI data structures for s390 and s390x:
2724 On s390 and s390x the size of struct ucontext and jmp_buf was increased in
2725 2.19. This change is reverted in 2.20. The introduced 2.19 symbol versions
2726 of getcontext, setjmp, _setjmp, __sigsetjmp, longjmp, _longjmp, siglongjmp
2727 are preserved pointing straight to the same implementation as the old ones.
2728 Given that, new callers will simply provide a too-big buffer to these
2729 functions. Any applications/libraries out there that embed jmp_buf or
2730 ucontext_t in an ABI-relevant data structure that have already been rebuilt
2731 against 2.19 headers will have to rebuilt again. This is necessary in any
2732 case to revert the breakage in their ABI caused by the glibc change.
2733
2734 * Support for file description locks is added to systems running the
2735 Linux kernel. The standard file locking interfaces are extended to
2736 operate on file descriptions, not file descriptors, via the use of
2737 F_OFD_GETLK, F_OFD_SETLK, and F_OFD_SETLKW. File description locks
2738 are associated with an open file instead of a process.
2739
2740 * Optimized strchr implementation for AArch64. Contributed by ARM Ltd.
2741
2742 * The minimum Linux kernel version that this version of the GNU C Library
2743 can be used with is 2.6.32.
2744
2745 * Running the testsuite no longer terminates as soon as a test fails.
2746 Instead, a file tests.sum (xtests.sum from "make xcheck") is generated,
2747 with PASS or FAIL lines for individual tests. A summary of the results is
2748 printed, including a list of failing lists, and "make check" exits with
2749 error status if there were any unexpected failures. "make check
2750 stop-on-test-failure=y" may be used to keep the old behavior.
2751
2752 * The am33 port, which had not worked for several years, has been removed
2753 from ports.
2754
2755 * The _BSD_SOURCE and _SVID_SOURCE feature test macros are no longer
2756 supported; they now act the same as _DEFAULT_SOURCE (but generate a
2757 warning). Except for cases where _BSD_SOURCE enabled BSD interfaces that
2758 conflicted with POSIX (support for which was removed in 2.19), the
2759 interfaces those macros enabled remain available when compiling with
2760 _GNU_SOURCE defined, with _DEFAULT_SOURCE defined, or without any feature
2761 test macros defined.
2762
2763 * Optimized strcmp implementation for ARMv7. Contributed by ARM Ltd.
2764
2765 * Added support for TX lock elision of pthread mutexes on s390 and s390x.
2766 This may improve lock scaling of existing programs on TX capable systems.
2767 The lock elision code is only built with --enable-lock-elision=yes and
2768 then requires a GCC version supporting the TX builtins. With lock elision
2769 default mutexes are elided via __builtin_tbegin, if the cpu supports
2770 transactions. By default lock elision is not enabled and the elision code
2771 is not built.
2772
2773 * CVE-2014-4043 The posix_spawn_file_actions_addopen implementation did not
2774 copy the path argument. This allowed programs to cause posix_spawn to
2775 deference a dangling pointer, or use an unexpected pathname argument if
2776 the string was modified after the posix_spawn_file_actions_addopen
2777 invocation.
2778
2779 * All supported architectures now use the main glibc sysdeps directory
2780 instead of some being in a separate "ports" directory (which was
2781 distributed separately before glibc 2.17).
2782
2783 * The NPTL implementation of POSIX pthreads is no longer an "add-on".
2784 On configurations that support it (all Linux configurations), it's now
2785 used regardless of the --enable-add-ons switch to configure. It is no
2786 longer possible to build such configurations without pthreads support.
2787
2788 * Locale names, including those obtained from environment variables (LANG
2789 and the LC_* variables), are more tightly checked for proper syntax.
2790 setlocale will now fail (with EINVAL) for locale names that are overly
2791 long, contain slashes without starting with a slash, or contain ".." path
2792 components. (CVE-2014-0475) Previously, some valid locale names were
2793 silently replaced with the "C" locale when running in AT_SECURE mode
2794 (e.g., in a SUID program). This is no longer necessary because of the
2795 additional checks.
2796
2797 * On x86-64, the dynamic linker's lazy-binding support is now compatible
2798 with application code using Intel MPX instructions. (With all previous
2799 versions, the MPX register state could be clobbered when making calls
2800 into or out of a shared library.) Note that while the new dynamic
2801 linker is compatible with all known x86 hardware whether or not it
2802 supports Intel MPX, some x86 instruction-set emulators might fail to
2803 handle the new instruction encodings. This is known to affect Valgrind
2804 versions up through 3.9 (but will be fixed in the forthcoming 3.10
2805 release), and might affect other tools that do instruction emulation.
2806
2807 * Support for loadable gconv transliteration modules has been removed.
2808 The support for transliteration modules has been non-functional for
2809 over a decade, and the removal is prompted by security defects. The
2810 normal gconv conversion modules are still supported. Transliteration
2811 with //TRANSLIT is still possible, and the //IGNORE specifier
2812 continues to be supported. (CVE-2014-5119)
2813
2814 * Decoding a crafted input sequence in the character sets IBM933, IBM935,
2815 IBM937, IBM939, IBM1364 could result in an out-of-bounds array read,
2816 resulting a denial-of-service security vulnerability in applications which
2817 use functions related to iconv. (CVE-2014-6040)
2818 \f
2819 Version 2.19
2820
2821 * The following bugs are resolved with this release:
2822
2823 156, 387, 431, 762, 832, 926, 2801, 4772, 6786, 6787, 6807, 6810, 6981,
2824 7003, 9721, 9954, 10253, 10278, 11087, 11157, 11214, 12100, 12486, 12751,
2825 12986, 13028, 13982, 13985, 14029, 14032, 14120, 14143, 14155, 14286,
2826 14547, 14699, 14752, 14782, 14876, 14910, 15004, 15048, 15073, 15089,
2827 15128, 15218, 15268, 15277, 15308, 15362, 15374, 15400, 15425, 15427,
2828 15483, 15522, 15531, 15532, 15593, 15601, 15608, 15609, 15610, 15632,
2829 15640, 15670, 15672, 15680, 15681, 15723, 15734, 15735, 15736, 15748,
2830 15749, 15754, 15760, 15763, 15764, 15797, 15799, 15825, 15843, 15844,
2831 15846, 15847, 15849, 15850, 15855, 15856, 15857, 15859, 15867, 15886,
2832 15887, 15890, 15892, 15893, 15895, 15897, 15901, 15905, 15909, 15915,
2833 15917, 15919, 15921, 15923, 15939, 15941, 15948, 15963, 15966, 15968,
2834 15985, 15988, 15997, 16032, 16034, 16036, 16037, 16038, 16041, 16046,
2835 16055, 16071, 16072, 16074, 16077, 16078, 16103, 16112, 16143, 16144,
2836 16146, 16150, 16151, 16153, 16167, 16169, 16172, 16195, 16214, 16245,
2837 16271, 16274, 16283, 16289, 16293, 16314, 16316, 16330, 16337, 16338,
2838 16356, 16365, 16366, 16369, 16372, 16375, 16379, 16381, 16384, 16385,
2839 16386, 16387, 16390, 16394, 16398, 16400, 16407, 16408, 16414, 16430,
2840 16431, 16453, 16474, 16506, 16510, 16529
2841
2842 * Slovenian translations for glibc messages have been contributed by the
2843 Translation Project's Slovenian team of translators.
2844
2845 * The public headers no longer use __unused nor __block. This change is to
2846 support compiling programs that are derived from BSD sources and use
2847 __unused internally, and to support compiling with Clang's -fblock
2848 extension which uses __block.
2849
2850 * CVE-2012-4412 The strcoll implementation caches indices and rules for
2851 large collation sequences to optimize multiple passes. This cache
2852 computation may overflow for large collation sequences and may cause a
2853 stack or buffer overflow. This is now fixed to use a slower algorithm
2854 which does not use a cache if there is an integer overflow.
2855
2856 * CVE-2012-4424 The strcoll implementation uses malloc to cache indices and
2857 rules for large collation sequences to optimize multiple passes and falls
2858 back to alloca if malloc fails, resulting in a possible stack overflow.
2859 The implementation now falls back to an uncached collation sequence lookup
2860 if malloc fails.
2861
2862 * CVE-2013-4788 The pointer guard used for pointer mangling was not
2863 initialized for static applications resulting in the security feature
2864 being disabled. The pointer guard is now correctly initialized to a
2865 random value for static applications. Existing static applications need
2866 to be recompiled to take advantage of the fix (bug 15754).
2867
2868 * CVE-2013-4237 The readdir_r function could write more than NAME_MAX bytes
2869 to the d_name member of struct dirent, or omit the terminating NUL
2870 character. (Bugzilla #14699).
2871
2872 * CVE-2013-4332 The pvalloc, valloc, memalign, posix_memalign and
2873 aligned_alloc functions could allocate too few bytes or corrupt the
2874 heap when passed very large allocation size values (Bugzilla #15855,
2875 #15856, #15857).
2876
2877 * CVE-2013-4458 Stack overflow in getaddrinfo with large number of results
2878 for AF_INET6 has been fixed (Bugzilla #16072).
2879
2880 * New locales: ak_GH, anp_IN, ar_SS, cmn_TW, hak_TW, lzh_TW, nan_TW, pap_AW,
2881 pap_CW, quz_PE, the_NP.
2882
2883 * Substantially revised locales: gd_GB, ht_HT
2884
2885 * The LC_ADDRESS field was updated to support country_car for almost all
2886 supported locales.
2887
2888 * ISO 1427 definitions were updated.
2889
2890 * ISO 3166 definitions were updated.
2891
2892 * The localedef utility now supports --big-endian and --little-endian
2893 command-line options to generate locales for a different system from that
2894 for which the C library was built.
2895
2896 * Binary locale files now only depend on the endianness of the system for
2897 which they are generated and not on other properties of that system. As a
2898 consequence, binary files generated with new localedef may be incompatible
2899 with old versions of the GNU C Library, and binary files generated with
2900 old localedef may be incompatible with this version of the GNU C Library,
2901 in the following circumstances:
2902
2903 + Locale files may be incompatible on m68k systems.
2904
2905 + Locale archive files (but not separate files for individual locales) may
2906 be incompatible on systems where plain "char" is signed.
2907
2908 * The configure option --disable-versioning has been removed. Builds with
2909 --disable-versioning had not worked for several years.
2910
2911 * ISO 639 definitions were updated for Chiga (cgg) and Chinese (gan, hak, czh,
2912 cjy, lzh, cmn, mnp, cdo, czo, cpx, wuu, hsn, yue).
2913
2914 * SystemTap probes for malloc have been introduced.
2915
2916 * SystemTap probes for slow multiple precision fallback paths of
2917 transcendental functions have been introduced.
2918
2919 * Support for powerpc64le has been added.
2920
2921 * The soft-float powerpc port now supports e500 processors.
2922
2923 * Support for STT_GNU_IFUNC symbols added for ppc32/power4+ and ppc64.
2924
2925 * A new feature test macro _DEFAULT_SOURCE is available to enable the same
2926 set of header declarations that are enabled by default, even when other
2927 feature test macros or compiler options such as -std=c99 would otherwise
2928 disable some of those declarations.
2929
2930 * The _BSD_SOURCE feature test macro no longer enables BSD interfaces that
2931 conflict with POSIX. The libbsd-compat library (which was a dummy library
2932 that did nothing) has also been removed.
2933
2934 * Preliminary documentation about Multi-Thread, Async-Signal and
2935 Async-Cancel Safety has been added.
2936
2937 * Change of ABI data structures for s390 and s390x:
2938 On s390 and s390x the size of struct ucontext and jmp_buf was increased to
2939 allow for future hardware extensions. All library functions that accept or
2940 return these structures were versioned in order to provide backwards
2941 compatibility with existing code. However, not all cases can be handled
2942 with interface versioning. If an application embeds either structure into
2943 another structure and that structure is passed to another compilation unit
2944 which expects the newer larger-sized object then any access to the new
2945 fields will result in undefined behaviour. Similarly any access to fields
2946 that were moved by the enlarging of the structures will cause undefined
2947 behaviour. To fix the undefined behaviour all objects that are part of the
2948 final in-memory image must agree on the size of structures, and this may
2949 require recompilation.
2950 \f
2951 Version 2.18
2952
2953 * The following bugs are resolved with this release:
2954
2955 2546, 2560, 5159, 6809, 7006, 10060, 10062, 10283, 10357, 10686, 11120,
2956 11561, 12310, 12387, 12492, 12515, 12723, 13550, 13889, 13951, 13988,
2957 14142, 14176, 14200, 14256, 14280, 14293, 14317, 14327, 14478, 14496,
2958 14582, 14686, 14812, 14888, 14894, 14907, 14908, 14909, 14920, 14941,
2959 14952, 14964, 14981, 14982, 14985, 14991, 14994, 14996, 15000, 15003,
2960 15006, 15007, 15014, 15020, 15022, 15023, 15036, 15054, 15055, 15062,
2961 15078, 15084, 15085, 15086, 15100, 15160, 15214, 15221, 15232, 15234,
2962 15283, 15285, 15287, 15304, 15305, 15307, 15309, 15327, 15330, 15335,
2963 15336, 15337, 15339, 15342, 15346, 15359, 15361, 15366, 15380, 15381,
2964 15394, 15395, 15405, 15406, 15409, 15416, 15418, 15419, 15423, 15424,
2965 15426, 15429, 15431, 15432, 15441, 15442, 15448, 15465, 15480, 15485,
2966 15488, 15490, 15492, 15493, 15497, 15506, 15529, 15536, 15553, 15577,
2967 15583, 15618, 15627, 15631, 15654, 15655, 15666, 15667, 15674, 15711,
2968 15755, 15759.
2969
2970 * CVE-2013-2207 Incorrectly granting access to another user's pseudo-terminal
2971 has been fixed by disabling the use of pt_chown (Bugzilla #15755).
2972 Distributions can re-enable building and using pt_chown via the new configure
2973 option `--enable-pt_chown'. Enabling the use of pt_chown carries with it
2974 considerable security risks and should only be used if the distribution
2975 understands and accepts the risks.
2976
2977 * CVE-2013-0242 Buffer overrun in regexp matcher has been fixed (Bugzilla
2978 #15078).
2979
2980 * CVE-2013-1914 Stack overflow in getaddrinfo with many results has been
2981 fixed (Bugzilla #15330).
2982
2983 * Add support for calling C++11 thread_local object destructors on thread
2984 and program exit. This needs compiler support for offloading C++11
2985 destructor calls to glibc.
2986
2987 * Improved worst case performance of libm functions with double inputs and
2988 output.
2989
2990 * Support for priority inherited mutexes in pthread condition variables on
2991 non-x86 architectures.
2992
2993 * Port to Xilinx MicroBlaze contributed by David Holsgrove.
2994
2995 * Optimized string functions for AArch64. Implemented by Marcus Shawcroft.
2996
2997 * Optimized string functions for ARM. Implemented by Will Newton and
2998 Richard Henderson.
2999
3000 * Optimized libm functions for SPARC. Implemented by David S. Miller.
3001
3002 * Support for building more of ARM glibc as Thumb-2 code. Implemented by
3003 Richard Henderson.
3004
3005 * Support for building most of MIPS glibc as MIPS16 code. Implemented by
3006 Chung-Lin Tang, Maciej W. Rozycki and Maxim Kuvyrkov.
3007
3008 * Added a benchmark framework to track performance of functions in glibc.
3009
3010 * New <math.h> macro named issignaling to check for a signaling NaN (sNaN).
3011 It is based on draft TS 18661 and currently enabled as a GNU extension.
3012
3013 * On Linux, the clock function now uses the clock_gettime system call
3014 for improved precision, rather than old times system call.
3015
3016 * Added support for version-3 tz format files. This is needed when using
3017 the tz database release 2013e or later, and affects a few unusual cases --
3018 currently only TZ='America/Godthab' for time stamps after 2037.
3019
3020 * Added new API functions pthread_getattr_default_np and
3021 pthread_setattr_default_np to get and set the default pthread
3022 attributes of a process.
3023
3024 * Added support for TSX lock elision for pthread mutexes on i386 and x86-64.
3025 This may improve lock scaling of existing programs on TSX capable systems.
3026 When the --enable-lock-elision=yes parameter is specified at configure
3027 time lock elision will be enabled by default for all PTHREAD_MUTEX_DEFAULT
3028 mutexes.
3029
3030 * Support added for AT_HWCAP2 (to coincide with Linux kernel 3.10
3031 availability). Implemented by Ryan S. Arnold.
3032
3033 * Support added for POWER8 platform. Implemented by Ryan S. Arnold.
3034 \f
3035 Version 2.17
3036
3037 * The following bugs are resolved with this release:
3038
3039 1349, 3439, 3479, 3665, 5044, 5246, 5298, 5400, 6530, 6677, 6778, 6808,
3040 9685, 9914, 10014, 10038, 10114, 10191, 10631, 10873, 11438, 11607, 11638,
3041 11741, 12140, 13013, 13412, 13542, 13601, 13603, 13604, 13629, 13679,
3042 13696, 13698, 13717, 13741, 13759, 13761, 13763, 13881, 13939, 13950,
3043 13952, 13966, 14042, 14047, 14090, 14150, 14151, 14152, 14154, 14157,
3044 14166, 14173, 14195, 14197, 14237, 14246, 14251, 14252, 14283, 14298,
3045 14303, 14307, 14328, 14331, 14336, 14337, 14347, 14349, 14368, 14376,
3046 14417, 14447, 14459, 14476, 14477, 14501, 14505, 14510, 14516, 14518,
3047 14519, 14530, 14532, 14538, 14543, 14544, 14545, 14557, 14562, 14568,
3048 14576, 14579, 14583, 14587, 14595, 14602, 14610, 14621, 14638, 14645,
3049 14648, 14652, 14660, 14661, 14669, 14672, 14683, 14694, 14716, 14719,
3050 14743, 14767, 14783, 14784, 14785, 14793, 14796, 14797, 14801, 14803,
3051 14805, 14807, 14811, 14815, 14821, 14822, 14824, 14828, 14831, 14833,
3052 14835, 14838, 14856, 14863, 14865, 14866, 14868, 14869, 14871, 14872,
3053 14879, 14889, 14893, 14898, 14914.
3054
3055 * Optimization of memcpy for MIPS.
3056
3057 * CVE-2011-4609 svc_run() produces high cpu usage when accept fails with
3058 EMFILE has been fixed (Bugzilla #14889).
3059
3060 * The add-on ports collection is now distributed in the "ports" subdirectory
3061 of the main GNU C Library distribution, rather than separately.
3062
3063 * Port to ARM AArch64 contributed by Linaro.
3064
3065 * Support for STT_GNU_IFUNC symbols added for s390 and s390x.
3066 Optimized versions of memcpy, memset, and memcmp added for System z10 and
3067 zEnterprise z196.
3068 Implemented by Andreas Krebbel.
3069
3070 * The new function secure_getenv allows secure access to the environment,
3071 returning NULL if running in a SUID/SGID process. This function replaces
3072 the internal function __secure_getenv.
3073
3074 * SystemTap static probes have been added into the dynamic linker.
3075 Implemented by Gary Benson.
3076
3077 * Optimizations of string functions strstr, strcasestr and memmem.
3078 Implemented by Maxim Kuvyrkov.
3079
3080 * The minimum Linux kernel version that this version of the GNU C Library
3081 can be used with is 2.6.16.
3082
3083 * Optimizations of string functions memchr, wcschr, wcscpy, and wcsrchr for
3084 powerpc POWER7. Implemented by Will Schmidt.
3085
3086 * New configure option --disable-nscd builds the C library such that it
3087 never attempts to contact the Name Service Caching Daemon (nscd).
3088 New configure option --disable-build-nscd avoids building nscd itself;
3089 this is the default if --disable-nscd is used.
3090
3091 * Improved support for cross-compilation, including cross-testing and
3092 bootstrap builds without a previously built glibc.
3093
3094 * Several testsuite tests are now able to test multiple IFUNC variants of an
3095 interface, rather than just testing the one that would be chooen by
3096 default.
3097
3098 * New configure options --with-bugurl and --with-pkgversion, for
3099 distributors to use to embed their bug-reporting and package version
3100 information in --help and --version output.
3101
3102 * The ttyname and ttyname_r functions on Linux now fall back to searching for
3103 the tty file descriptor in /dev/pts or /dev if /proc is not available. This
3104 allows creation of chroots without the procfs mounted on /proc.
3105
3106 * The `crypt' function now fails if passed salt bytes that violate the
3107 specification for those values. On Linux, the `crypt' function will
3108 consult /proc/sys/crypto/fips_enabled to determine if "FIPS mode" is
3109 enabled, and fail on encrypted strings using the MD5 or DES algorithm
3110 when the mode is enabled.
3111
3112 * The `clock_*' suite of functions (declared in <time.h>) is now available
3113 directly in the main C library. Previously it was necessary to link with
3114 -lrt to use these functions. This change has the effect that a
3115 single-threaded program that uses a function such as `clock_gettime' (and
3116 is not linked with -lrt) will no longer implicitly load the pthreads
3117 library at runtime and so will not suffer the overheads associated with
3118 multi-thread support in other code such as the C++ runtime library.
3119
3120 * New locales: ayc_PE, doi_IN, ia_FR, mni_IN, nhn_MX, niu_NU, niu_NZ,
3121 sat_IN, and szl_PL.
3122 \f
3123 Version 2.16
3124
3125 * The following bugs are resolved with this release:
3126
3127 174, 208, 350, 369, 411, 706, 766, 2074, 2541, 2547, 2548, 2550, 2551,
3128 2552, 2553, 2554, 2562, 2563, 2565, 2566, 2570, 2576, 2636, 2678, 3335,
3129 3440, 3748, 3768, 3866, 3868, 3906, 3976, 3992, 4026, 4108, 4596, 4822,
3130 5077, 5461, 5805, 5993, 6471, 6486, 6578, 6649, 6730, 6770, 6794, 6884,
3131 6890, 6894, 6895, 6907, 6911, 6959, 7064, 9739, 9902, 10110, 10135, 10140,
3132 10153, 10210, 10254, 10346, 10375, 10545, 10716, 10846, 10882, 11174,
3133 11261, 11322, 11365, 11451, 11494, 11521, 11677, 11837, 11959, 12047,
3134 12097, 12193, 12194, 12297, 12298, 12301, 12340, 12354, 12416, 12495,
3135 13058, 13223, 13361, 13525, 13526, 13527, 13528, 13529, 13530, 13531,
3136 13532, 13533, 13547, 13551, 13552, 13553, 13555, 13556, 13559, 13563,
3137 13566, 13576, 13579, 13583, 13592, 13594, 13613, 13618, 13637, 13656,
3138 13658, 13673, 13691, 13695, 13704, 13705, 13706, 13718, 13726, 13738,
3139 13739, 13743, 13750, 13758, 13760, 13761, 13775, 13786, 13787, 13792,
3140 13806, 13824, 13840, 13841, 13844, 13846, 13848, 13851, 13852, 13854,
3141 13871, 13872, 13873, 13879, 13882, 13883, 13884, 13885, 13886, 13892,
3142 13895, 13908, 13910, 13911, 13912, 13913, 13914, 13915, 13916, 13917,
3143 13918, 13919, 13920, 13921, 13922, 13923, 13924, 13926, 13927, 13928,
3144 13938, 13941, 13942, 13954, 13955, 13956, 13963, 13967, 13968, 13970,
3145 13973, 13979, 13983, 13986, 13996, 14012, 14027, 14033, 14034, 14036,
3146 14040, 14043, 14044, 14048, 14049, 14050, 14053, 14055, 14059, 14064,
3147 14075, 14080, 14083, 14103, 14104, 14109, 14112, 14117, 14122, 14123,
3148 14134, 14153, 14183, 14188, 14199, 14210, 14218, 14229, 14241, 14273,
3149 14277, 14278.
3150
3151 * Support for the x32 ABI on x86-64 added. The x32 target is selected by
3152 configuring glibc with:
3153 BUILD_CC='gcc' CC='gcc -mx32' CXX='g++ -mx32'
3154 Visit <http://sites.google.com/site/x32abi/> for more x32 ABI info.
3155 Implemented by H.J. Lu.
3156
3157 * ISO C11 support:
3158
3159 + define static_assert
3160
3161 + do not declare gets
3162
3163 + declare at_quick_exit and quick_exit also for ISO C11
3164
3165 + aligned_alloc. NB: The code is deliberately allows the size parameter
3166 to not be a multiple of the alignment. This is a moronic requirement
3167 in the standard but it is only a requirement on the caller, not the
3168 implementation.
3169
3170 + timespec_get added
3171
3172 + uchar.h support added
3173
3174 + CMPLX, CMPLXF, CMPLXL added
3175
3176 Implemented by Ulrich Drepper.
3177
3178 * Support for the IA-64 has been moved to ports.
3179
3180 * Remove support for anything but ELF binary format
3181
3182 * Checking versions of poll, ppoll added.
3183 Implemented by Ulrich Drepper.
3184
3185 * More generic and 64-bit performance optimizations to math functions.
3186 Implemented by Ulrich Drepper.
3187
3188 * New configure option --enable-obsolete-rpc makes the deprecated RPC
3189 headers and functions available at compile time as they were before
3190 version 2.14. This option will be removed at some time in the future
3191 after the TI-RPC library becomes fully sufficient for the needs of
3192 existing applications.
3193
3194 * Compatibility code for Linux kernel versions before 2.4 has been removed.
3195 Note that glibc is not expected to work with any Linux kernel version
3196 before 2.6.
3197
3198 * New header <sys/auxv.h> and function getauxval allowing easy access to
3199 the AT_* key-value pairs passed from the Linux kernel. The header also
3200 defines the HWCAP_* bits associated with the AT_HWCAP key.
3201
3202 * New locales: mag_IN
3203
3204 * New configure option --enable-systemtap builds SystemTap static probes
3205 into libc for setjmp and longjmp and into libpthread for various operations.
3206 So far the setjmp/longjmp probes and some of the libpthread probes are
3207 provided only for i*86 and x86_64.
3208 Implemented by Roland McGrath and Rayson Ho.
3209
3210 * Optimized expf for x86-32 and x86-64. Implemented by Liubov Dmitrieva.
3211
3212 * More optimized functions for PowerPC. Implemented by Adhemerval Zanella
3213 and Will Schmidt.
3214
3215 * More optimized functions for SPARC. Implemented by David S. Miller.
3216
3217 * Improved support for cross-compilation, especially bootstrap builds
3218 without a previously built glibc.
3219
3220 * Ports for the TILE-Gx and TILEPro families of processors. Contributed by
3221 Chris Metcalf from Tilera.
3222
3223 * Support for the old ARM ABI has been removed from ports. Only the EABI is
3224 now supported for ARM processors.
3225
3226 * The hard-float variant of the ARM EABI now uses /lib/ld-linux-armhf.so.3
3227 as the name of the dynamic linker, to distinguish it from the
3228 /lib/ld-linux.so.3 dynamic linker used for the base EABI.
3229
3230 * Support for CRIS has been removed from ports.
3231
3232 * A new class of installed header has been documented for low-level
3233 platform-specific functionality. PowerPC added the first instance with a
3234 function to provide time base register access. Contributed by Tulio
3235 Magno Quites Machado Filho.
3236
3237 * ix86 configurations now install header files that are consistent with
3238 what x86-64 configurations install. These same header files can be used
3239 for -m32, -m64, or -mx32 builds.
3240 Contributed by H.J. Lu.
3241
3242 * Math library bug fixes. A thorough audit of all open math library bugs was
3243 conducted by Joseph Myers. Significant progress was made on many math
3244 library bugs resulting in more accurate exceptions and function results.
3245 Many thanks to all those that contributed including Andreas Jaeger for his
3246 patch review and work on the x87 trigonometric instruction issues.
3247
3248 * Timezone data is no longer installed. Timezone-related binaries and scripts
3249 will continue to be installed. Users should obtain their timezone data from
3250 their distribution provider or from the tzdata package at
3251 <ftp://munnari.oz.au/pub/>.
3252 \f
3253 Version 2.15
3254
3255 * The following bugs are resolved with this release:
3256
3257 6779, 6783, 9696, 10103, 10709, 11589, 11929, 12403, 12786, 12840, 12847,
3258 12868, 12852, 12874, 12885, 12892, 12906, 12907, 12922, 12935, 12962,
3259 13007, 13021, 13061, 13062, 13067, 13068, 13085, 13088, 13090, 13092,
3260 13096, 13114, 13118, 13123, 13134, 13138, 13147, 13150, 13166, 13179,
3261 13185, 13189, 13192, 13268, 13276, 13282, 13291, 13305, 13328, 13335,
3262 13337, 13344, 13358, 13367, 13413, 13416, 13423, 13439, 13446, 13472,
3263 13484, 13506, 13515, 13523, 13524, 13538, 13540
3264
3265 * New program pldd to list loaded object of a process
3266 Implemented by Ulrich Drepper.
3267
3268 * Add nss_db support back to glibc. No more dependency on Berkeley db
3269 and support for initgroups lookups.
3270 Implemented by Ulrich Drepper.
3271
3272 * Optimized strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-32.
3273 Contributed by HJ Lu.
3274
3275 * Improved strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-64.
3276 Contributed by HJ Lu.
3277
3278 * Optimized strcat, strncat on x86-64 and optimized wcscmp, wcslen, strnlen
3279 on x86-32 and x86-64.
3280 Contributed by Liubov Dmitrieva.
3281
3282 * Optimized strchr and strrchr for SSE on x86-32.
3283 Contributed by Liubov Dmitrieva.
3284
3285 * Optimized memchr, memrchr, rawmemchr, memcmp, wmemcmp, wcschr, wcscpy
3286 for x86-64 and x86-32.
3287 Contributed by Liubov Dmitrieva.
3288
3289 * New interfaces: scandirat, scandirat64
3290 Implemented by Ulrich Drepper.
3291
3292 * Checking versions of FD_SET, FD_CLR, and FD_ISSET added.
3293 Implemented by Ulrich Drepper.
3294
3295 * nscd now also caches the netgroup database.
3296 Implemented by Ulrich Drepper.
3297
3298 * Integrate libm with gcc's -ffinite-math-only option.
3299 Implemented by Ulrich Drepper.
3300
3301 * Lots of generic, 64-bit, and x86-64-specific performance optimizations
3302 to math functions. Implemented by Ulrich Drepper.
3303
3304 * Optimized strcasecmp and strncasecmp for AVX on x86-64.
3305 Implemented by Ulrich Drepper.
3306
3307 * New Linux interfaces: process_vm_readv, process_vm_writev
3308
3309 * Optimized strcasecmp and strncasecmp for SSSE3 and SSE4.2 on x86-32.
3310 Implemented by Ulrich Drepper.
3311
3312 * Optimized nearbyint and strcasecmp for PPC.
3313 Implemented by Adhemerval Zanella.
3314
3315 * New locales: bho_IN, unm_US, es_CU, ta_LK
3316 \f
3317 Version 2.14
3318
3319 * The following bugs are resolved with this release:
3320
3321 386, 6420, 7101, 9730, 9732, 9809, 10138, 10149, 10157, 11099, 11257,
3322 11258, 11487, 11532, 11558, 11578, 11634, 11653, 11668, 11697, 11724,
3323 11781, 11799, 11820, 11837, 11857, 11884, 11892, 11895, 11901, 11945,
3324 11947, 11952, 11987, 12052, 12083, 12158, 12178, 12200, 12346, 12350,
3325 12393, 12420, 12432, 12445, 12449, 12453, 12454, 12460, 12469, 12489,
3326 12509, 12510, 12511, 12518, 12527, 12541, 12545, 12551, 12582, 12583,
3327 12587, 12597, 12601, 12611, 12625, 12626, 12631, 12650, 12653, 12655,
3328 12660, 12671, 12681, 12684, 12685, 12711, 12713, 12714, 12717, 12723,
3329 12724, 12734, 12738, 12746, 12766, 12775, 12777, 12782, 12788, 12792,
3330 12795, 12811, 12813, 12814, 12841
3331
3332 * The RPC implementation in libc is obsoleted. Old programs keep working
3333 but new programs cannot be linked with the routines in libc anymore.
3334 Programs in need of RPC functionality must be linked against TI-RPC.
3335 The TI-RPC implementation is IPv6 enabled and there are other benefits.
3336
3337 Visible changes of this change include (obviously) the inability to link
3338 programs using RPC functions without referencing the TI-RPC library and the
3339 removal of the RPC headers from the glibc headers.
3340 Implemented by Ulrich Drepper.
3341
3342 * New Linux interfaces: clock_adjtime, name_to_handle_at, open_by_handle_at,
3343 syncfs, setns, sendmmsg
3344
3345 * New locales: os_RU, bem_ZA, en_ZA, ff_SN, sw_KE, sw_TZ, lb_LU, wae_CH,
3346 yue_HK, lij_IT, mhr_RU
3347
3348 * New encodings: CP770, CP771, CP772, CP773, CP774
3349
3350 * New program sotruss to trace calls through PLTs
3351 Implemented by Ulrich Drepper.
3352
3353 * The malloc hook implementation is marked deprecated and will be removed
3354 from the default implementation in the next version. The design never
3355 worked ever since the introduction of threads. Even programs which do
3356 not create threads themselves can use multiple threads created internally.
3357 \f
3358 Version 2.13
3359
3360 * The following bugs are resolved with this release:
3361
3362 3268, 6812, 7066, 10085, 10484, 10563, 10851, 11149, 11155, 11577, 11611,
3363 11640, 11655, 11701, 11840, 11856, 11883, 11903, 11904, 11968, 11979,
3364 12005, 12037, 12067, 12077, 12078, 12092, 12093, 12107, 12108, 12113,
3365 12140, 12159, 12167, 12191, 12194, 12201, 12204, 12205, 12207, 12348,
3366 12378, 12394, 12397
3367
3368 * New Linux interfaces: prlimit, prlimit64, fanotify_init, fanotify_mark
3369
3370 * POWER7 optimizations: memset, memcmp, strncmp
3371
3372 * New optimized string functions for x86-64: strnlen (SSE2),
3373 strcasecmp (SSE2, SSSE3, SSE4.2), strncasecmp (SSE2, SSSE3, SSE4.2)
3374 Implemented by Ulrich Drepper.
3375 \f
3376 Version 2.12
3377
3378 * The following bugs are resolved with this release:
3379
3380 3662, 4457, 5553, 10162, 10401, 10414, 10554, 10824, 10864, 10915,
3381 10918, 10936, 10939, 10958, 10968, 10969, 10972, 10992, 11000, 11001,
3382 11007, 11010, 11027, 11039, 11040, 11041, 11043, 11046, 11056, 11070,
3383 11093, 11115, 11125, 11126, 11127, 11134, 11141, 11149, 11183, 11184,
3384 11185, 11186, 11187, 11188, 11189, 11190, 11191, 11192, 11193, 11194,
3385 11200, 11230, 11235, 11242, 11254, 11258, 11271, 11272, 11276, 11279,
3386 11287, 11292, 11319, 11332, 11333, 11387, 11389, 11390, 11394, 11397,
3387 11410, 11438, 11449, 11470, 11471, 11520, 11537, 11538, 11571
3388
3389 * New interfaces: pthread_getname_np, pthread_setname_np
3390
3391 * New Linux interface: recvmmsg
3392
3393 * STT_GNU_IFUNC implemented for Sparc by David Miller.
3394
3395 * The dynamic linker now recognizes supported ABI versions from the
3396 EI_ABIVERSION field in the ELF header.
3397 Implemented by Ulrich Drepper.
3398
3399 * New NIS mode selector ADJUNCT_AS_SHADOW. The passwd.adjunct.byname table
3400 will not be used to fill in password fields in the passwd.byname replies.
3401 Instead it is used to synthesize the shadow.byname table, should it be
3402 missing. This is a useful mode in some installations involving Solaris.
3403 Implemented by Ulrich Drepper.
3404
3405 * New locales: kok_IN, sq_MK, cv_RU
3406 \f
3407 Version 2.11
3408
3409 * The following bugs are resolved with this release:
3410
3411 3270, 7094, 9924, 9986, 10011, 10085, 10107, 10159, 10162, 10166, 10183,
3412 10195, 10196, 10200, 10203, 10207, 10211, 10213, 10217, 10221, 10229,
3413 10262, 10286, 10312, 10315, 10319, 10349, 10360, 10391, 10402, 10416,
3414 10418, 10422, 10425, 10446, 10448, 10452, 10540, 10553, 10560, 10564,
3415 10609, 10643, 10692, 10713, 10717, 10730, 10731, 10742, 10780, 10784,
3416 10789, 10817, 10823, 10840, 10847
3417
3418 * New interfaces: execvpe, pthread_sigqueue, mkstemps, mkstemps64, mkostemps,
3419 mkostemps64
3420 Implemented by Ulrich Drepper.
3421
3422 * Checking version of longjmp added that fails if an uninitialized stack
3423 frame would be created. Implemented by Ulrich Drepper.
3424
3425 * STT_GNU_IFUNC is now supported in static executables.
3426 Implemented by H.J. Lu.
3427
3428 * STT_GNU_IFUNC implemented for PPC by Alan Modra.
3429
3430 * New optimized string functions for x86-64: strstr, strcasestr, memcmp,
3431 strcspn, strpbrk, strspn, strcpy, stpcpy, strncpy, strcmp (SSE2, SSE4.2),
3432 strncmp (SSE2, SSE4.2), strchr (SSE4.2), strrchr (SSE4.2).
3433 Contributed by H.J. Lu.
3434
3435 strlen, rawmemchr, strcmp (SSSE3), strncmp (SSSE3).
3436 Implemented by Ulrich Drepper.
3437
3438 * New optimized string functions for x86: strlen, strcspn, strspn, strpbrk,
3439 strstr, strcasestr.
3440 Contributed by H.J. Lu.
3441
3442 * Support for fma instruction in AVX on x86-64.
3443 Implemented by H.J. Lu and Ulrich Drepper.
3444
3445 * AVX support in x86-64 auditing support in ld.so.
3446 Implemented by H.J. Lu.
3447
3448 * STB_GNU_UNIQUE support added.
3449 Implemented by Ulrich Drepper.
3450
3451 * Implement second fallback mode for optimized DNS lookup for even more
3452 broken environments. If two requests from the same port are not
3453 handled correctly close the socket and open a new one before sending
3454 the second request. The 'single-request-reopen' option in /etc/resolv.conf
3455 can be used to select this mode right away, instead of rediscovering the
3456 necessity is every process again.
3457 Implemented by Ulrich Drepper.
3458
3459 * New resolver flag RES_USE_DNSSEC to enable use of verified lookup.
3460 Implemented by Adam Tkac.
3461
3462 * Optimized iconv conversions for S390x.
3463 Implemented by Andreas Krebbel.
3464
3465 * Using condvars with PI mutexes is now more efficient due to kernel
3466 support for requeueing to PI futexes. NPTL support added for x86-64.
3467 Implemented by Ulrich Drepper.
3468
3469 * New locales: ps_AF, my_MM
3470 \f
3471 Version 2.10
3472
3473 * The following bugs are resolved with this release:
3474
3475 697, 5381, 5807, 6411, 6545, 7040, 7058, 7067, 7080, 7083, 7095, 7098,
3476 9677, 9694, 9697, 9701, 9704, 9705, 9706, 9720, 9726, 9730, 9733, 9736,
3477 9741, 9750, 9753, 9759, 9781, 9791, 9793, 9823, 9833, 9844, 9880, 9881,
3478 9891, 9893, 9895, 9913, 9916, 9920, 9948, 9955, 9957, 9963, 9985, 10052,
3479 10069, 10086, 10087, 10090, 10092, 10093, 10100, 10118, 10128
3480
3481 * New interfaces: preadv, preadv64, pwritev, pwritev64, malloc_info
3482 Implemented by Ulrich Drepper.
3483
3484 * New Linux interfaces: accept4, fallocate, fallocate64.
3485 Implemented by Ulrich Drepper.
3486
3487 * Correct declarations of string function when used in C++ code. This
3488 could lead to compile errors for invalid C++ code.
3489
3490 * XPG7/POSIX 2008 compilation environment. Many GNU-specific functions are
3491 now in POSIX.
3492
3493 * New POSIX 2008 interface: psiginfo
3494 Implemented by Ulrich Drepper.
3495
3496 * New ISO C++1x interfaces: quick_exit, at_quick_exit
3497 Implemented by Ulrich Drepper.
3498
3499 * Support for selecting between multiple function definitions at runtime
3500 using STT_GNU_IFUNC symbols. Implemented by Ulrich Drepper.
3501
3502 * The libcrypt library can now use the hash function implementations in
3503 NSS. Implemented by Ulrich Drepper.
3504
3505 * The malloc implementation can be compiled to be less memory efficient
3506 but higher performing in multi-threaded programs.
3507 Implemented by Ulrich Drepper.
3508
3509 * New locales: nan_TW@latin, ks_IN
3510
3511 * Faster strlen, strchr, strchrnul, strrchr, memchr, and rawmemchr for x86-64.
3512 Implemented by Ulrich Drepper.
3513
3514 * Extended printf hook support. It is possible to use user-defined types
3515 and extend existing format specifiers.
3516 Implemented by Ulrich Drepper.
3517
3518 * Handling for group shadow files has been added.
3519 Implemented by Ulrich Drepper.
3520
3521 * DNS IPv4-IPv6 parallel lookup now deals better with broken DNS
3522 servers (the case, e.g., for some people using the built-in DNS
3523 server in ADSL modems/routers). There is a once-per-process timeout
3524 in case of a broken server. To avoid it, users can run nscd or put
3525 'options single-request' in /etc/resolv.conf.
3526 Implemented by Ulrich Drepper.
3527 \f
3528 Version 2.9
3529
3530 * The following bugs are resolved with this release:
3531
3532 3406, 5209, 5210, 5381, 5794, 5814, 5911, 6428, 6438, 6442, 6447, 6461,
3533 6472, 6505, 6544, 6589, 6612, 6634, 6653, 6654, 6657, 6698, 6712, 6713,
3534 6719, 6723, 6724, 6740, 6763, 6771, 6790, 6791, 6817, 6824, 6839, 6843,
3535 6845, 6856, 6857, 6860, 6867, 6875, 6919, 6920, 6942, 6947, 6955, 6966,
3536 6968, 6974, 6980, 6995, 7008, 7009, 7029
3537
3538 * Unified lookup for getaddrinfo: IPv4 and IPv6 addresses are now looked
3539 up at the same time. Implemented by Ulrich Drepper.
3540
3541 * TLS descriptors for LD and GD on x86 and x86-64.
3542 Implemented by Alexandre Oliva.
3543
3544 * getaddrinfo now handles DCCP and UDPlite.
3545 Implemented by Ulrich Drepper.
3546
3547 * New fixed-size conversion macros: htobe16, htole16, be16toh, le16toh,
3548 htobe32, htole32, be32toh, le32toh, htobe64, htole64, be64toh, le64toh.
3549 Implemented by Ulrich Drepper.
3550
3551 * New implementation of memmem, strstr, and strcasestr which is O(n).
3552 Implemented by Eric Blake.
3553
3554 * New Linux interfaces: inotify_init1, dup3, epoll_create1, pipe2
3555
3556 * Implement "e" option for popen to open file descriptor with the
3557 close-on-exec flag set. Implemented by Ulrich Drepper.
3558
3559 * Implement "b" mode for fmemopen. In this mode writes writes don't
3560 implicitly add a NUL byte and seeks from the end of the buffer really
3561 use the buffer end, not the string length as the basis.
3562 Implemented by Ulrich Drepper.
3563
3564 * Many functions, exported and internal, now atomically set the close-on-exec
3565 flag when run on a sufficiently new kernel. Implemented by Ulrich Drepper.
3566
3567 * Sorting rules for some Indian languages (Kannada, Gurumukhi, Telugu,
3568 Sinhala)
3569 Implemented by Pravin Satpute.
3570
3571 * New locales: sd_IN, sd_IN@devanagari, ks_IN@devanagari, ht_HT, en_AG, nl_AW.
3572 \f
3573 Version 2.8
3574
3575 * The following bugs are resolved with this release:
3576
3577 2549, 3112, 4314, 4407, 4997, 5012, 5040, 5112, 5204, 5208, 5209, 5220,
3578 5222, 5224, 5225, 5237, 5238, 5240, 5245, 5277, 5346, 5375, 5378, 5382,
3579 5424, 5427, 5428, 5435, 5436, 5439, 5441, 5442, 5443, 5451, 5452, 5454,
3580 5455, 5463, 5464, 5465, 5475, 5477, 5478, 5520, 5541, 5545, 5553, 5600,
3581 5602, 5607, 5614, 5627, 5628, 5736, 5737, 5741, 5753, 5760, 5762, 5768,
3582 5774, 5777, 5778, 5779, 5786, 5790, 5818, 5854, 5857, 5903, 5939, 5979,
3583 5983, 5995, 5998, 6004, 6007, 6020, 6021, 6024, 6040, 6042
3584
3585 * New locales: bo_CN, bo_IN, shs_CA.
3586
3587 * New encoding: HP-ROMAN9, HP-GREEK8, HP-THAI8, HP-TURKISH8.
3588
3589 * Sorting rules for some Indian languages (Devanagari and Gujarati).
3590 Implemented by Pravin Satpute.
3591
3592 * IPV6 addresses in /etc/resolv.conf can now have a scope ID
3593
3594 * nscd caches now all timeouts for DNS entries
3595 Implemented by Ulrich Drepper.
3596
3597 * nscd is more efficient and wakes up less often.
3598 Implemented by Ulrich Drepper.
3599
3600 * More checking functions: asprintf, dprintf, obstack_printf, vasprintf,
3601 vdprintf, and obstack_vprintf.
3602 Implemented by Jakub Jelinek.
3603
3604 * Faster memset for x86-64.
3605 Implemented by Harsha Jagasia and H.J. Lu.
3606
3607 * Faster memcpy on x86.
3608 Implemented by Ulrich Drepper.
3609
3610 * ARG_MAX is not anymore constant on Linux. Use sysconf(_SC_ARG_MAX).
3611 Implemented by Ulrich Drepper.
3612
3613 * Faster sqrt and sqrtf implemention for some PPC variants.
3614 Implemented by Steven Munroe.
3615 \f
3616 Version 2.7
3617
3618 * The following bugs are resolved with this release:
3619
3620 4125, 4126, 4178, 4359, 4407, 4512, 4514, 4525, 4554, 4556, 4557, 4566,
3621 4582, 4586, 4588, 4599, 4610, 4647, 4702, 4705, 4726, 4745, 4772, 4773,
3622 4775, 4776, 4792, 4813, 4814, 4816, 4833, 4858, 4860, 4896, 4905, 4925,
3623 4936, 4937, 4938, 4941, 4946, 4963, 4972, 5010, 5028, 5043, 5058, 5063,
3624 5071, 5103, 5104, 5112, 5113, 5184, 5186
3625
3626 * More checking functions: fread, fread_unlocked, open*, mq_open.
3627 Implemented by Jakub Jelinek and Ulrich Drepper.
3628
3629 * Extend fortification to C++. Implemented by Jakub Jelinek.
3630
3631 * Implement 'm' modifier for scanf. Add stricter C99/SUS compliance
3632 by not recognizing 'a' as a modifier when those specs are requested.
3633 Implemented by Jakub Jelinek.
3634
3635 * PPC optimizations to math and string functions.
3636 Implemented by Steven Munroe.
3637
3638 * New interfaces: mkostemp, mkostemp64. Like mkstemp* but allow additional
3639 options to be passed. Implemented by Ulrich Drepper.
3640
3641 * More CPU set manipulation functions. Implemented by Ulrich Drepper.
3642
3643 * New Linux interfaces: signalfd, eventfd, eventfd_read, and eventfd_write.
3644 Implemented by Ulrich Drepper.
3645
3646 * Handle private futexes in the NPTL implementation.
3647 Implemented by Jakub Jelinek and Ulrich Drepper.
3648
3649 * Add support for O_CLOEXEC. Implement in Hurd. Use throughout libc.
3650 Implemented by Roland McGrath and Ulrich Drepper.
3651
3652 * Linux/x86-64 vDSO support. Implemented by Ulrich Drepper.
3653
3654 * SHA-256 and SHA-512 based password encryption.
3655 Implemented by Ulrich Drepper.
3656
3657 * New locales: ber_DZ, ber_MA, en_NG, fil_PH, fur_IT, fy_DE, ha_NG, ig_NG,
3658 ik_CA, iu_CA, li_BE, li_NL, nds_DE, nds_NL, pap_AN, sc_IT, tk_TM, ug_CN,
3659 yo_NG.
3660
3661 + New iconv modules: MAC-CENTRALEUROPE, ISO-8859-9E, KOI8-RU.
3662 Implemented by Ulrich Drepper.
3663 \f
3664 Version 2.6
3665
3666 * The following bugs are resolved with this release:
3667
3668 3156, 3213, 3285, 3291, 3306, 3313, 3320, 3322, 3325, 3326, 3334, 3348,
3669 3352, 3363, 3369, 3426, 3427, 3429, 3451, 3458, 3465, 3480, 3483, 3493,
3670 3514, 3515, 3559, 3632, 3664, 3673, 3674, 3745, 3747, 3818, 3840, 3842,
3671 3851, 3855, 3884, 3885, 3902, 3905, 3919, 3944, 3954, 3955, 3957, 3991,
3672 3995, 3996, 4024, 4040, 4069, 4070, 4074, 4076, 4096, 4101, 4102, 4114,
3673 4130, 4131, 4181, 4276, 4306, 4342, 4344, 4349, 4364, 4368, 4381, 4392,
3674 4403, 4405, 4406, 4411, 4438, 4439, 4455, 4456, 4465, 4512, 4514, 4586,
3675 4702, 4858
3676
3677 * New Linux interfaces: epoll_pwait, sched_getcpu.
3678
3679 * New generic interfaces: strerror_l.
3680
3681 * nscd can now cache the services database. Implemented by Ulrich Drepper.
3682 \f
3683 Version 2.5
3684
3685 * The following bugs are resolved with this release:
3686
3687 39, 68, 192, 352, 388, 531, 935, 1006, 1201, 1203, 1386, 1782, 1783, 1784,
3688 1785, 1786, 1787, 1951, 1974, 1996, 2072, 2098, 2125, 2134, 2158, 2182,
3689 2349, 2376, 2386, 2389, 2415, 2418, 2420, 2423, 2450, 2451, 2466, 2472,
3690 2473, 2477, 2498, 2499, 2501, 2502, 2505, 2507, 2508, 2509, 2517, 2526,
3691 2569, 2571, 2592, 2611, 2625, 2632, 2644, 2662, 2680, 2683, 2684, 2693,
3692 2695, 2703, 2734, 2751, 2764, 2766, 2775, 2792, 2821, 2823, 2832, 2841,
3693 2843, 2883, 2892, 2908, 2914, 2926, 2961, 2978, 2980, 2987, 2997, 2998,
3694 3013, 3018, 3034, 3040, 3044, 3123, 3124, 3137, 3138, 3143, 3155, 3189,
3695 3225, 3251, 3252, 3253, 3273
3696
3697 * For Linux, the sorting of addresses returned by getaddrinfo now also
3698 handles rules 3, 4, and 7 from RFC 3484. I.e., all rules are handled.
3699 Implemented by Ulrich Drepper.
3700
3701 * Allow system admin to configure getaddrinfo with the /etc/gai.conf file.
3702 Implemented by Ulrich Drepper.
3703
3704 * New Linux interfaces: splice, tee, sync_file_range, vmsplice.
3705
3706 * New iconv module for MIK. Contributed by Alexander Shopov.
3707
3708 * For sites with broken group and/or passwd database, the auto-propagate
3709 option of nscd can prevent creating ID lookup entries from the results
3710 of a name lookup and vice versa. This usually is no problem but some
3711 site might have problems with the default behavior.
3712 Implemented by Ulrich Drepper.
3713
3714 * Iterating over entire database in NIS can be slow. With the
3715 SETENT_BATCH_READ option in /etc/default/nss a system admin can decide
3716 to trade time for memory. The entire database will be read at once.
3717 Implemented by Ulrich Drepper.
3718
3719 * The interfaces introduced in RFC 3542 have been implemented by
3720 Ulrich Drepper.
3721
3722 * Support for the new ELF hash table format was added by Ulrich Drepper.
3723
3724 * Support for priority inheritance mutexes added by Jakub Jelinek and
3725 Ulrich Drepper.
3726
3727 * Support for priority protected mutexes added by Jakub Jelinek.
3728
3729 * New locales: nr_ZA, pa_PK, ca_AD, ca_FR, ca_IT, el_CY, tr_CY, as_IN,
3730 or_IN, csb_PL, fy_NL, sr_ME.
3731 \f
3732 Version 2.4
3733
3734 * More overflow detection functions.
3735
3736 * New iconv converters for IBM1025, IBM1122, IBM1137, IBM1153,
3737 IBM1154, IBM1155, IBM1156, IBM1157, and IBM1158 contributed by Jiro Sekiba.
3738
3739 More converters for IBM803, IBM901, IBM902, IBM921, IBM1008, IBM1097,
3740 IBM1112, IBM1123, IBM1130, IBM1140, IBM1141, IBM1142, IBM1143, IBM1144,
3741 IBM1145, IBM1146, IBM1147, IBM1148, IBM1149, IBM1166, IBM1167, IBM4517,
3742 IBM4899, IBM4909, IBM4971, IBM5347, IBM9030, IBM9066, IBM9448, IBM12712,
3743 IBM16804, IBM1364, IBM1371, IBM1388, IBM1390, and IBM1399 contributed
3744 by Masahide Washizawa.
3745
3746 * It is now possible to install audit libraries for the dynamic linker using
3747 LD_AUDIT. Implemented by Ulrich Drepper.
3748
3749 * The LinuxThreads add-on, providing pthreads on Linux 2.4 kernels, is no
3750 longer supported. The new NPTL implementation requires Linux 2.6 kernels.
3751 For a libc and libpthread that works well on Linux 2.4 kernels, we
3752 recommend using the stable 2.3 branch.
3753
3754 * The new function open_wmemstream acts like open_memstream,
3755 but uses a wchar_t wide character buffer.
3756
3757 * The new function ppoll is an improved version of poll, similar to pselect.
3758
3759 * New interfaces: fdopendir, openat, fstatat, fchmodat, fchownat,
3760 futimesat, faccessat, mkdirat, mkfifoat, mknodat,
3761 renameat, unlinkat, linkat, symlinkat, readlinkat.
3762
3763 * New Linux kernel system calls: unshare,
3764 inotify_init, inotify_add_watch, inotify_rm_watch.
3765
3766 * The euidaccess function is now also known by the alias eaccess,
3767 for compatibility with some other systems.
3768
3769 * Timezone data updated to 2006b version.
3770 \f
3771 Version 2.3.6
3772
3773 * The following bugs are resolved with this release:
3774
3775 38, 253, 549, 622, 653, 721, 758, 851, 877, 915, 934, 955, 961,
3776 1016, 1037, 1076, 1079, 1080, 1081, 1082, 1083, 1084, 1085, 1086,
3777 1087, 1088, 1090, 1091, 1092, 1093, 1094, 1095, 1096, 1097, 1098,
3778 1099, 1100, 1101, 1102, 1103, 1104, 1105, 1106, 1107, 1108, 1109,
3779 1110, 1111, 1112, 1113, 1125, 1137, 1138, 1249, 1250, 1251, 1252,
3780 1253, 1254, 1350, 1358, 1394, 1438, 1498, 1534
3781
3782 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
3783
3784 * As of this release, GCC 4 can be used to compile the C Library.
3785
3786 * Timezone data updated to 2005m version.
3787 \f
3788 Version 2.3.5
3789
3790 * The following bugs are resolved with this release:
3791
3792 284, 592, 602, 626, 633, 640, 650, 661, 671, 681, 693, 700, 710, 719,
3793 722, 723, 725, 726, 727, 728, 729, 730, 731, 732, 733, 734, 735, 736,
3794 737, 738, 739, 740, 741, 742, 743, 744, 745, 765, 767, 768, 769, 776,
3795 777, 787, 821, 822, 823, 825
3796
3797 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
3798 \f
3799 Version 2.3.4
3800
3801 * Support for RFC 3678. Real implementations exist only for Linux so far.
3802 Implemented by Ulrich Drepper.
3803
3804 * nscd can now cache entries persistently. Expiring entries are reloaded.
3805 For speedups the cache can be shared in memory with client processes.
3806 Implemented by Ulrich Drepper.
3807
3808 * nscd can now perform SELinux checks.
3809 Implemented by Matthew Rickard <mjricka@epoch.ncsc.mil>.
3810
3811 * getaddrinfo queries are now cached. Canonical name lookup is performed
3812 efficiently.
3813 Implemented by Ulrich Drepper.
3814
3815 * The nothrow function attribute is used when headers are used by gcc when
3816 compiling C code. This can avoid significant amounts of exception
3817 handling data.
3818
3819 * The malloc functions perform more error checking and are stricter when
3820 it comes to reacting on errors. The default action is to terminate
3821 the process after showing an error message. Implemented by Ulrich Drepper.
3822
3823 * Reverse lookups of IPv6 addresses does not use bit string or .ip6.int
3824 lookups anymore unless explicitly requested. Implemented by Ulrich Drepper.
3825
3826 * Namespaces in ld.so are implemented. DSOs can be loaded in separate
3827 namespaces using the new function dlmopen(). This feature is of course,
3828 like most other dynamic loading functionality, not available in statically
3829 linked applications. Implemented by Ulrich Drepper.
3830
3831 * Low-overhead boundary checking variants of string and some stdio functions
3832 were added. These are to be used in conjunction with a gcc patch by
3833 Jakub Jelinek which adds calls to these functions if possible.
3834 Implemented by Jakub Jelinek and Ulrich Drepper.
3835
3836 * Old code for several operating systems and machine architectures that
3837 have not been in working condition in a long time have been removed from
3838 the main source tree maintained by the GNU C Library's maintainers.
3839 These files are now reside in the separate `ports' source module
3840 that is usable as an add-on when building the library.
3841 \f
3842 Version 2.3.3
3843
3844 * New functions `dladdr1' and `dlinfo' in <dlfcn.h> provide more ways to
3845 interrogate the dynamic linker, compatible with the Solaris interface.
3846
3847 * ELF thread-local storage support (TLS) now works on PowerPC and PowerPC64;
3848 implemented by Paul Mackerras, Steven Munroe, and Roland McGrath.
3849
3850 * getifaddrs now uses the netlink interface on Linux to get its information.
3851 Implemented by Thorsten Kukuk.
3852
3853 * getaddrinfo now implements AI_V4MAPPED, AI_ALL, and AI_ADDRCONF.
3854 Implemented by Ulrich Drepper.
3855
3856 * support for non-executable stacks on x86 has been added. Changes mostly
3857 by Roland McGrath.
3858
3859 * regex is now much faster for multibyte locales. Changes by Jakub Jelinek
3860 and Ulrich Drepper.
3861
3862 * getaddrinfo now performs destination address selection according to
3863 RFC 3484.
3864 \f
3865 Version 2.3.2
3866
3867 * Thread-safe interfaces for many functions that access locale data
3868 were added in version 2.3, but these features were omitted from NEWS.
3869 Many functions have variants with an `_l' suffix that take a `locale_t'
3870 object as a parameter rather than consulting the current locale.
3871 The new functions `newlocale', `duplocale', and `freelocale' in <locale.h>
3872 create and maintain `locale_t' objects. Additionally, the new function
3873 `uselocale' sets "the current locale" (as used by functions not so
3874 parameterized) set for an individual thread. These features were added
3875 in version 2.3, implemented by Ulrich Drepper and Roland McGrath.
3876
3877 * The functions getresuid, getresgid, setresuid, and setresgid, which
3878 have long been available on Linux, are now declared in <unistd.h>
3879 and are now also available on the Hurd.
3880
3881 * ELF thread-local storage support (TLS) now works on x86-64.
3882
3883 * The new dynamic string token $LIB is expanded in shared library names.
3884 This normally expands to lib, but on some 64-bit platforms to lib64 instead.
3885
3886 * Aldy Hernandez contributed complete software floating point support for
3887 PowerPC machines with no FPU.
3888
3889 * fexecve is implemented on Linux.
3890
3891 * The `btowc' function should work at least twice as fast due to
3892 specialized callbacks in the iconv modules. Implemented by Bruno Haible.
3893
3894 * With approriate thread add-ons cancelable functions are now implemented
3895 in libc.so as well. No need to call the function in libpthread. This
3896 change allowed to finally disable the incorrect and expensive handling
3897 of weak definition in ld.so.
3898
3899 * Yet more PLT entries in libc.so have been removed. We finally arrived
3900 at the bare minimum. Startup times improved appropriately.
3901
3902 * Support for the new Linux/x86 system call interface was added. The
3903 AT_SYSINFO auxiliary vector entry is recognized and handled.
3904 \f
3905 Version 2.3
3906
3907 * Masahide Washizawa contributed iconv modules for IBM1163 and IBM1164
3908 charsets.
3909
3910 * iconv (the program and the interface) now accepts empty names (excluding
3911 options like //TRANSLIT) to mean "use charset of current locale".
3912
3913 * localedef can now transliterate characters in strings which are not in
3914 the provided charmap. The information from the input locale is used.
3915
3916 * Prelinking support was added for ELF targets. This requires additional
3917 tools and recent versions of the GNU binutils. Contributed by Jakub Jelinek.
3918
3919 * Read-only stdio streams now use mmap to speed up operation by eliminating
3920 copying and buffer underflows. To use add 'm' to the mode string of
3921 the fopen/fdopen/freopen call. Implemented by Ulrich Drepper.
3922
3923 * The malloc functions were completely rewritten by Wolfram Gloger based
3924 on Doug Lea's malloc-2.7.0.c.
3925
3926 * Isamu Hasegawa contributed a completely new and POSIX-conformant
3927 implementation of regex.
3928
3929 * Bruno Haible upgraded the iconv and locale implementation to support
3930 Unicode 3.2.
3931
3932 * Contents of the LC_* and LANG environment variables in the CEN style are
3933 not recognized anymore. It never was used. Change by Ulrich Drepper.
3934
3935 * The runtime (ld.so, libc, libpthread for Linux) now can handle the ELF
3936 thread-local storage (TLS) ABI on some platforms.
3937 Changes by Ulrich Drepper. SH support by Kaz Kojima.
3938
3939 * Bruno Haible contributed iconv converters for ISO-2022-JP-3, SHIFT JIS-X0213,
3940 EUC-JISX0213, and TSCII.
3941
3942 * New header <ifaddrs.h> with functions `getifaddrs' and `freeifaddrs':
3943 BSD-compatible interface for getting all network interface addresses.
3944 Implementation for IPv4 by Roland McGrath.
3945
3946 * Loading of locale data is faster due to the introduction of a locale
3947 archive. Implemented by Roland McGrath and Ulrich Drepper.
3948
3949 * Startup times are significantly reduced by not using exported functions
3950 inside the library itself. Changes by Jakub Jelinek, Roland McGrath,
3951 and Ulrich Drepper.
3952
3953 * Steven Munroe contributed a port to PowerPC64/Linux.
3954 \f
3955 Version 2.2.6
3956
3957 * The Hurd now uses the GNU libio implementation of stdio rather than the
3958 old GNU stdio implementation, and uses a new ABI (libc.so.0.3).
3959
3960 * The Hurd on x86 now has the `ioperm' function and <sys/io.h> header file
3961 with the same behavior as the Linux system call of the same name.
3962 \f
3963 Version 2.2.5
3964
3965 * Stephen Moshier implemented log2, log10, powl and cbrtl for the
3966 128-bit long double format.
3967
3968 * Masahide Washizawa contributed iconv modules for IBM1132, IBM1133, IBM1160,
3969 IBM1161, and IBM1162 charsets.
3970
3971 * Andreas Jaeger contributed a port to x86-64/Linux.
3972
3973 * Peter Bruin contributed a port to PowerPC/Hurd.
3974
3975 * libc functions using I/O streams now can handle wide-oriented streams
3976 as well.
3977
3978 * optimizations in the dynamic linker. Binaries created by recent binutils
3979 versions start up quicker due to reduced time spent on relocations.
3980
3981 * Support for use of gcc3 added by Jakub Jelinek and HJ Lu.
3982 \f
3983 Version 2.2.4
3984
3985 * Stephen Moshier implemented cosh, expm1, log1p, acos, sinh, tanh,
3986 asinh, atanh, j0 for the 128-bit long double format.
3987
3988 * Bruno Haible updated all the code handling Unicode in some form to
3989 support Unicode 3.1.
3990
3991 * Speed of regex for single-byte locales is back to previous levels.
3992 Patch by Isamu Hasegawa.
3993
3994 * Alpha, SPARC, and IA-64 now also using floating stacks.
3995
3996 * Startup time of internationalized applications greatly improved through
3997 iconv cache. Use iconvconfig to generate the cache file.
3998 Contributed by Ulrich Drepper.
3999
4000 * The IA-64 specific part of ld.so was rewritten to eliminate some pretty
4001 severe performance problems. Patch by David Mosberger.
4002
4003 * The Hurd port got a lot more functionality like AIO, various stdio
4004 extensions, etc. Mainly done by Roland McGrath.
4005
4006 * mtrace can now lookup symbols in shared libraries.
4007 \f
4008 Version 2.2.3
4009
4010 * Intel's IA-64 math library is largely integrated. It provides fast and
4011 accurate implementatations for most basic and standard math functions
4012 in float, double, and long double format.
4013
4014 * Stephen Moshier implemented j0, j1, jn, y0, y1, yn, lgamma, erf, erfc,
4015 and asin for the 96-bit long double format and asin, log, tan for the
4016 128-bit long double format.
4017
4018 * The beginning of a last-bit accurate math library by IBM Haifa were added.
4019 The basic double functions exist today. Contributed by Abraham Ziv
4020 <ziv@il.ibm.com>, Moshe Olshansky <olshansk@il.ibm.com>, Ealan Henis
4021 <ealan@il.ibm.com>, and Anna Reitman <reitman@il.ibm.com>.
4022
4023 * An asynchronous name lookup library was added. The interface is designed
4024 after POSIX AIO. The proposal was circulated beforehand to get comments.
4025 No negative ones came in. Implemented by Ulrich Drepper.
4026
4027 * Port to S390/64bit contributed by Martin Schwidefsky
4028 <schwidefsky@de.ibm.com>.
4029
4030 * David Mosberger <davidm@hpl.hp.com> implemented the setcontext family
4031 of functions for Linux/IA-64.
4032
4033 * The RPC code is now thread safe. Threads can now use the same service
4034 of different services at the same time. Patch by Eric Norum
4035 <eric.norum@usask.ca> with some help by Ulrich Drepper.
4036
4037 * Martin Schwidefsky <schwidefsky@de.ibm.com> implemented the setcontext
4038 family of functions for Linux/S390.
4039
4040 * Ulrich Drepper <drepper@redhat.com> implemented the setcontext family
4041 of functions for Linux/x86.
4042
4043 * Port to Linux/CRIS contributed by Axis Communications.
4044 \f
4045 Version 2.2.2
4046
4047 * Lots of headers were cleaned up. Using the tool in the conform/ subdir
4048 we can now check for namespace violations and missing declarations. The
4049 result is that almost all headers are now Unix-compliant (as defined in
4050 the upcoming XPG6). The negative side is that some programs might need
4051 corrections, too, if they depend on the incorrect form of the headers in
4052 previous versions which defined too many symbols and included too many
4053 other headers.
4054
4055 * regex now handles multibyte character sets correctly.
4056 Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
4057
4058 * iconv (the program) does now conform to the upcoming XPG6 and handles
4059 charmaps. Instead of the charset names the path of charmaps can be
4060 provided and the conversion happens based on this data.
4061 Contributed by Ulrich Drepper.
4062
4063 * The locale program now provides detailed information about the installed
4064 locales. While
4065
4066 locale -a
4067
4068 only lists the names of the supported locales
4069
4070 locale -a --verbose
4071
4072 provides details such as country, language, and codeset name.
4073 Contributed by Ulrich Drepper.
4074 \f
4075 Version 2.2.1
4076
4077 * The gencat program now parses the input file according to the charset
4078 selected by the LC_CTYPE category. This is important for stateful
4079 character sets. To make generating catalogs easier there is a way
4080 to overwrite the charset selected by the locale: before the first
4081 message or $ quote line the catalog can contain a line like
4082
4083 $ codeset=ISO-8859-2
4084
4085 to select the charset (ISO-8859-2 in this case).
4086
4087 Implemented by Shinya Hanataka and Ulrich Drepper.
4088
4089 * New codeset conversion modules: IBM-922 (Estonia PC codeset),
4090 IBM-1124 (Ukraine PC codeset), IBM-1129 (Vietnamese PC codeset).
4091 Contributed by Masahide Washizawa <washi@jp.ibm.com>.
4092
4093 * Optimized string functions for Alpha ev6 and ev67 by Richard Henderson
4094 <rth@redhat.com> and Rick Gorton <rick.gorton@alpha-processor.com>.
4095
4096 * The LANGUAGE environment variable is now ignored unless the locale is
4097 changed from the default "C" locale.
4098
4099 * The usual bug fixes.
4100 \f
4101 Version 2.2
4102
4103 * Greg McGary added runtime support for bounds checking using gcc's
4104 new -fbounded-pointers option. ix86 target is complete. PowerPC
4105 is in progress.
4106
4107 * Thorsten Kukuk added secure mode support to nscd.
4108
4109 * The Berkeley db libraries have been removed.
4110
4111 Related, the nss_db module is now in a separate package since it
4112 obviously requires a database library being available.
4113
4114 * Wide character I/O streams implemented by Ulrich Drepper.
4115
4116 * Functions from the extended socket API added by Ulrich Drepper.
4117
4118 * Functions feenableexcept and fedisableexcept to control the
4119 behaviour of individual exceptions have been added by Andreas Jaeger.
4120
4121 * ldconfig program added by Andreas Jaeger and Jakub Jelinek.
4122
4123 * The resolver code has been updated from bind 8.2.3-T5B which supports
4124 threads. The integration was done by Andreas Jaeger, Adam D. Bradley,
4125 and Mark Kettenis.
4126
4127 This change could in some situations effect backward compatibility. Since
4128 now `_res' is a thread-local instead of a global variable, modifying it
4129 in one thread does not have any effect in other threads.
4130
4131 The resolver library was also extended to allow IPv6 as the transport
4132 protocol for the requests. This work was done by Stig Venaas.
4133
4134 * Compatibility code for K&R C compilers has been removed from the
4135 header files. A ISO C compiler is needed to use the library
4136 (conforming to either C89 or C99 standard).
4137
4138 * Complete rewrite of the localedef program to support multibyte character
4139 sets. Implement handling of ISO 14651 and ISO 14652. Rewrite strcoll,
4140 strxfrm, wcscoll, and wcsxfrm functions. Make isw*() functions work.
4141 Implemented by Ulrich Drepper.
4142
4143 Bruno Haible significantly improved the generation and use of the data
4144 structures for the wide character tables.
4145
4146 * Plural handling in gettext implemented by Ulrich Drepper.
4147
4148 * The utmp daemon has been removed.
4149
4150 * The port to MIPS-Linux has been finished by Andreas Jaeger.
4151
4152 * A port to Hitachi SH3 and SH4 has been contributed by Kazumoto Kojima
4153 and Yutaka Niibe.
4154
4155 * POSIX clocks and timers implemented by Kaz Kylheku and Ulrich Drepper.
4156
4157 * POSIX barriers implemented by Kaz Kylheku.
4158
4159 * POSIX spawn function family implemented by Ulrich Drepper.
4160
4161 * POSIX spinlocks are now available.
4162
4163 * Timed wait functions for mutex, rwlock, and semaphores are implemented.
4164
4165 * the configure option --enable-kernel=X.Y.Z allows to strip out
4166 compatibility for kernel versions before X.Y.Z. This is currently only
4167 implemented for Linux.
4168
4169 * the sockaddr_in6 structure changed. The IPv6 working group added a new
4170 field sin6_scope_id. This means that all programs using IPv6 should be
4171 recompiled. Don't expect binary compatibility with previous glibc
4172 versions.
4173
4174 * various conversion modules for IBM character sets contributed by
4175 Masahide Washizawa.
4176
4177 * IA-64 port by Jes Sorensen and HJ Lu.
4178 \f
4179 Version 2.1.3
4180
4181 * bug fixes
4182
4183 \f
4184 Version 2.1.2
4185
4186 * bug fixes
4187
4188 \f
4189 Version 2.1.1
4190
4191 * New ISO C 9x function _Exit, imaxabs, and imaxdiv are added.
4192
4193 * New xdr functions are added; some rpc functions are now 64bit clean.
4194
4195 * Fixed a number of bugs and memory leaks (especially in NIS+ code).
4196
4197 * Fixed known incompatibilities with glibc 2.0.
4198
4199 * New functions lock64, strchrnul, rawmemchr, getutmp and getutmpx.
4200
4201 * Optimized a number of functions (especially the ELF dynamic loader).
4202
4203 * Update timezone data files.
4204
4205 * lots of charmaps corrections
4206
4207 * some new locale definitions and charmaps
4208
4209 \f
4210 Version 2.1
4211
4212 * Richard Henderson corrected size of struct timeval on Linux/Alpha to
4213 conform to POSIX member type requirements. Symbol versions have been
4214 adjusted as needed within the library, and for direct use by applications,
4215 but there is potential for problems if third-party libraries use
4216 struct timeval as part of their interface. This does not present
4217 a problem for X and other "essential" system libraries.
4218
4219 * An additional locale model to support C++ Standard Library locale
4220 model and probably more was implemented by Ulrich Drepper.
4221
4222 * Eric Youngdale and Ulrich Drepper implemented versioning of objects on
4223 symbol level.
4224
4225 * Miles Bader provided the `argp' function family to support hierachical
4226 command line argument parsing, layered on top of getopt.
4227
4228 * strtod accepts new hexadecimal floating-point format from ISO C 9X.
4229
4230 * printf has two new specifiers %a and %A to print hexadecimal flaoting-point
4231 numbers.
4232
4233 * scanf recognizes the %a and %A format for scanning floating point numbers.
4234
4235 * the new headers <stdint.h> and <inttypes.h> from ISO C 9X provides
4236 information and interfaces for the available integer types.
4237
4238 * about 130 new math functions were added to implement the ISO C9x math
4239 library.
4240
4241 * the new header <complex.h> contains definitions of the complex math
4242 functions from ISO C 9X.
4243
4244 * the new header <tgmath.h> defines generic macros to use complex or
4245 real valued functions.
4246
4247 * Thorsten Kukuk provided an implementation for NIS+, securelevel 0, 1 and 2.
4248
4249 * Andreas Jaeger provided a test suite for the math library.
4250
4251 * Mark Kettenis implemented the utmpx interface and an utmp daemon.
4252
4253 * Ulrich Drepper added character set conversion functions (iconv).
4254
4255 * Optimized string functions have been added.
4256
4257 * The localedata addon is now part of glibc.
4258
4259 * An implementation of profiling shared libraries was added by Ulrich Drepper.
4260
4261 * Thorsten Kukuk and Ulrich Drepper provided an implementation for a caching
4262 daemon for NSS (nscd).
4263
4264 Missing a better place here are some numbers on improvements. Under
4265 Linux 2.1.125 un-tar-ing the kernel sources takes
4266
4267 user system wall
4268
4269 using local files 12.19s 6.88s 22.91s
4270
4271 using NIS 13.92s 8.91s 26.34s
4272
4273 using NIS & nscd 10.37s 7.34s 25.30s
4274
4275 using NIS+ 27.57s 30.37s 640.46s
4276
4277 using NIS+ & nscd 10.25s 7.83s 26.51s
4278
4279 using NIS & old nscd [1] 13.83s 8.32s 29.60s
4280
4281 Keep in mind that non-namelookup related operations dominate above times.
4282 It was just a common complain that using NIS+ unpacking the kernel is
4283 horribly slow.
4284
4285 [1] The old nscd implementation is not available anymore since it was
4286 distributed with glibc up to version 2.0.98 and thus is now replaced.
4287
4288 * Tim Waugh provided an implementation of the POSIX.2 wordexp function family.
4289
4290 * Mark Kettenis provided a Hesiod NSS module.
4291
4292 * The ELF dynamic loader knows how to process dynamic string tokens ($ORIGIN
4293 and $PLATFORM) in RPATHs and similar strings (Ulrich Drepper).
4294
4295 * rcmd can now handle netgroups (Dick Streefland).
4296
4297 * A port to the ARM was contributed by Phil Blundell, Pat Beirne and Scott
4298 Bambrough.
4299
4300 * Support for the IPv6 protocol has been added to the socket API, as per the
4301 latest draft standards.
4302
4303 * Support for Linux 2.2 has been added.
4304
4305 * Interface changes relative to the latest 2.0.x release:
4306 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4307 addseverity NEW: Unix98
4308 alphasort64 NEW: LFS
4309 argp_err_exit_status NEW: argp, GNU ext
4310 argp_error NEW: argp, GNU ext
4311 argp_failure NEW: argp, GNU ext
4312 argp_help NEW: argp, GNU ext
4313 argp_parse NEW: argp, GNU ext
4314 argp_program_bug_address NEW: argp, GNU ext
4315 argp_program_version NEW: argp, GNU ext
4316 argp_program_version_hook NEW: argp, GNU ext
4317 argp_state_help NEW: argp, GNU ext
4318 argp_usage NEW: argp, GNU ext
4319 authdes_create NEW: Secure RPC
4320 authdes_getucred NEW: Secure RPC
4321 authdes_pk_create NEW: Secure RPC
4322 backtrace NEW: GNU ext.
4323 backtrace_symbols NEW: GNU ext.
4324 backtrace_symbols_fd NEW: GNU ext.
4325 cacos NEW: ISO C 9x
4326 cacosf NEW: ISO C 9x
4327 cacosh NEW: ISO C 9x
4328 cacoshf NEW: ISO C 9x
4329 cacoshl NEW: ISO C 9x
4330 cacosl NEW: ISO C 9x
4331 capget NEW: kernel
4332 capset NEW: kernel
4333 carg NEW: ISO C 9x
4334 cargf NEW: ISO C 9x
4335 cargl NEW: ISO C 9x
4336 casin NEW: ISO C 9x
4337 casinf NEW: ISO C 9x
4338 casinh NEW: ISO C 9x
4339 casinhf NEW: ISO C 9x
4340 casinhl NEW: ISO C 9x
4341 casinl NEW: ISO C 9x
4342 catan NEW: ISO C 9x
4343 catanf NEW: ISO C 9x
4344 catanh NEW: ISO C 9x
4345 catanhf NEW: ISO C 9x
4346 catanhl NEW: ISO C 9x
4347 catanl NEW: ISO C 9x
4348 cbc_crypt NEW: Secure RPC
4349 ccos NEW: ISO C 9x
4350 ccosf NEW: ISO C 9x
4351 ccosh NEW: ISO C 9x
4352 ccoshf NEW: ISO C 9x
4353 ccoshl NEW: ISO C 9x
4354 ccosl NEW: ISO C 9x
4355 cexp NEW: ISO C 9x
4356 cexpf NEW: ISO C 9x
4357 cexpl NEW: ISO C 9x
4358 cimag NEW: ISO C 9x
4359 cimagf NEW: ISO C 9x
4360 cimagl NEW: ISO C 9x
4361 clearerr_locked REMOVED
4362 clntunix_create NEW: sunrpc ext
4363 clog NEW: ISO C 9x
4364 clog10 NEW: ISO C 9x
4365 clog10f NEW: ISO C 9x
4366 clog10l NEW: ISO C 9x
4367 clogf NEW: ISO C 9x
4368 clogl NEW: ISO C 9x
4369 conj NEW: ISO C 9x
4370 conjf NEW: ISO C 9x
4371 conjl NEW: ISO C 9x
4372 cpow NEW: ISO C 9x
4373 cpowf NEW: ISO C 9x
4374 cpowl NEW: ISO C 9x
4375 cproj NEW: ISO C 9x
4376 cprojf NEW: ISO C 9x
4377 cprojl NEW: ISO C 9x
4378 creal NEW: ISO C 9x
4379 crealf NEW: ISO C 9x
4380 creall NEW: ISO C 9x
4381 creat64 NEW: LFS
4382 csin NEW: ISO C 9x
4383 csinf NEW: ISO C 9x
4384 csinh NEW: ISO C 9x
4385 csinhf NEW: ISO C 9x
4386 csinhl NEW: ISO C 9x
4387 csinl NEW: ISO C 9x
4388 csqrt NEW: ISO C 9x
4389 csqrtf NEW: ISO C 9x
4390 csqrtl NEW: ISO C 9x
4391 ctan NEW: ISO C 9x
4392 ctanf NEW: ISO C 9x
4393 ctanh NEW: ISO C 9x
4394 ctanhf NEW: ISO C 9x
4395 ctanhl NEW: ISO C 9x
4396 ctanl NEW: ISO C 9x
4397 des_setparity NEW: Secure RPC
4398 ecb_crypt NEW: Secure RPC
4399 endutxent NEW: Unix98
4400 exp10 NEW: ISO C 9x
4401 exp10f NEW: ISO C 9x
4402 exp10l NEW: ISO C 9x
4403 exp2 NEW: ISO C 9x
4404 exp2f NEW: ISO C 9x
4405 exp2l NEW: ISO C 9x
4406 fattach NEW: STREAMS
4407 fdetach NEW: STREAMS
4408 fdim NEW: ISO C 9x
4409 fdimf NEW: ISO C 9x
4410 fdiml NEW: ISO C 9x
4411 feclearexcept NEW: ISO C 9x
4412 fegetenv NEW: ISO C 9x
4413 fegetexceptflag NEW: ISO C 9x
4414 fegetround NEW: ISO C 9x
4415 feholdexcept NEW: ISO C 9x
4416 feof_locked REMOVED
4417 feraiseexcept NEW: ISO C 9x
4418 ferror_locked REMOVED
4419 fesetenv NEW: ISO C 9x
4420 fesetexceptflag NEW: ISO C 9x
4421 fesetround NEW: ISO C 9x
4422 fetestexcept NEW: ISO C 9x
4423 feupdateenv NEW: ISO C 9x
4424 fflush_locked REMOVED
4425 ffsl NEW: GNU ext.
4426 ffsll NEW: GNU ext.
4427 fgetpos64 NEW: LFS
4428 fgets_unlocked NEW: GNU ext.
4429 fileno_locked REMOVED
4430 fma NEW: ISO C 9x
4431 fmaf NEW: ISO C 9x
4432 fmal NEW: ISO C 9x
4433 fmax NEW: ISO C 9x
4434 fmaxf NEW: ISO C 9x
4435 fmaxl NEW: ISO C 9x
4436 fmin NEW: ISO C 9x
4437 fminf NEW: ISO C 9x
4438 fminl NEW: ISO C 9x
4439 fmtmsg NEW: Unix98
4440 fopen64 NEW: LFS
4441 fputc_locked REMOVED
4442 fputs_unlocked NEW: GNU ext.
4443 fread_unlocked NEW: GNU ext.
4444 freopen64 NEW: LFS
4445 fseeko NEW: Unix98
4446 fsetpos64 NEW: LFS
4447 fstatfs64 NEW: LFS
4448 fstatvfs NEW: Unix98
4449 fstatvfs64 NEW: LFS
4450 ftello NEW: Unix98
4451 ftello64 NEW: LFS
4452 ftruncate64 NEW: LFS
4453 ftw64 NEW: LFS
4454 fwrite_unlocked NEW: GNU ext.
4455 gai_strerror NEW: IPv6
4456 gamma_r REMOVED
4457 gammaf_r REMOVED
4458 gammal_r REMOVED
4459 getchar_locked REMOVED
4460 getdate NEW: Unix98
4461 getdate_err NEW: Unix98
4462 getdate_r NEW: GNU ext.
4463 getmsg NEW: STREAMS
4464 getnameinfo NEW: IPv6
4465 getnetname NEW: Secure RPC
4466 getpmsg NEW: STREAMS
4467 getpt NEW: Unix98 PTY
4468 getrlimit64 NEW: LFS
4469 getutxent NEW: Unix98
4470 getutxid NEW: Unix98
4471 getutxline NEW: Unix98
4472 glob64 NEW: GNU ext.
4473 globfree64 NEW: GNU ext.
4474 gnu_get_libc_release NEW: GNU ext.
4475 gnu_get_libc_version NEW: GNU ext.
4476 grantpt NEW: Unix98 PTY
4477 host2netname NEW: Secure RPC
4478 iconv NEW: iconv
4479 iconv_close NEW: iconv
4480 iconv_open NEW: iconv
4481 if_freenameindex NEW: IPv6
4482 if_indextoname NEW: IPv6
4483 if_nameindex NEW: IPv6
4484 if_nametoindex NEW: IPv6
4485 in6addr_any NEW: IPv6
4486 in6addr_loopback NEW: IPv6
4487 inet6_isipv4mapped NEW: IPv6
4488 isastream NEW: STREAMS
4489 iswblank NEW: GNU ext.
4490 key_decryptsession NEW: Secure RPC
4491 key_decryptsession_pk NEW: Secure RPC
4492 key_encryptsession NEW: Secure RPC
4493 key_encryptsession_pk NEW: Secure RPC
4494 key_gendes NEW: Secure RPC
4495 key_get_conv NEW: Secure RPC
4496 key_secretkey_is_set NEW: Secure RPC
4497 key_setnet NEW: Secure RPC
4498 key_setsecret NEW: Secure RPC
4499 llrint NEW: ISO C 9x
4500 llrintf NEW: ISO C 9x
4501 llrintl NEW: ISO C 9x
4502 llround NEW: ISO C 9x
4503 llroundf NEW: ISO C 9x
4504 llroundl NEW: ISO C 9x
4505 log2 NEW: ISO C 9x
4506 log2f NEW: ISO C 9x
4507 log2l NEW: ISO C 9x
4508 lrint NEW: ISO C 9x
4509 lrintf NEW: ISO C 9x
4510 lrintl NEW: ISO C 9x
4511 lround NEW: ISO C 9x
4512 lroundf NEW: ISO C 9x
4513 lroundl NEW: ISO C 9x
4514 lseek64 NEW: LFS
4515 makecontext NEW: Unix98
4516 mempcpy NEW: GNU ext.
4517 mmap64 NEW: LFS
4518 moncontrol REMOVED
4519 modify_ldt NEW: kernel
4520 nan NEW: ISO C 9x
4521 nanf NEW: ISO C 9x
4522 nanl NEW: ISO C 9x
4523 nearbyint NEW: ISO C 9x
4524 nearbyintf NEW: ISO C 9x
4525 nearbyintl NEW: ISO C 9x
4526 netname2host NEW: Secure RPC
4527 netname2user NEW: Secure RPC
4528 nexttoward NEW: ISO C 9x
4529 nexttowardf NEW: ISO C 9x
4530 nexttowardl NEW: ISO C 9x
4531 nftw NEW: Unix98
4532 nftw64 NEW: LFS
4533 open64 NEW: LFS
4534 passwd2des NEW: Secure RPC
4535 pow10 NEW: GNU ext.
4536 pow10f NEW: GNU ext.
4537 pow10l NEW: GNU ext.
4538 pread NEW: Unix98
4539 pread64 NEW: LFS
4540 printf_size NEW: GNU ext.
4541 printf_size_info NEW: GNU ext.
4542 profil_counter REMOVED
4543 pthread_mutexattr_getkind_np REPLACED
4544 pthread_mutexattr_setkind_np REPLACED
4545 ptsname NEW: Unix98 PTY
4546 ptsname_r NEW: Unix98 PTY
4547 putc_locked REMOVED
4548 putchar_locked REMOVED
4549 putgrent NEW: GNU ext.
4550 putmsg NEW: STREAMS
4551 putpmsg NEW: STREAMS
4552 pututxline NEW: Unix98
4553 pwrite NEW: Unix98
4554 pwrite64 NEW: LFS
4555 readdir64 NEW: LFS
4556 readdir64_r NEW: LFS
4557 remquo NEW: ISO C 9x
4558 remquof NEW: ISO C 9x
4559 remquol NEW: ISO C 9x
4560 round NEW: ISO C 9x
4561 roundf NEW: ISO C 9x
4562 roundl NEW: ISO C 9x
4563 rtime NEW: GNU ext.
4564 scalbln NEW: ISO C 9x
4565 scalblnf NEW: ISO C 9x
4566 scalblnl NEW: ISO C 9x
4567 scandir64 NEW: LFS
4568 sendfile NEW: kernel
4569 setcontext NEW: Unix98
4570 setrlimit64 NEW: LFS
4571 setutxent NEW: Unix98
4572 sighold NEW: Unix98
4573 sigignore NEW: Unix98
4574 sigqueue NEW: POSIX.1b
4575 sigrelse NEW: Unix98
4576 sigset NEW: POSIX.1b
4577 sigtimedwait NEW: POSIX.1b
4578 sigwaitinfo NEW: POSIX.1b
4579 sincos NEW: GNU ext.
4580 sincosf NEW: GNU ext.
4581 sincosl NEW: GNU ext.
4582 statfs64 NEW: LFS
4583 statvfs NEW: Unix98
4584 statvfs64 NEW: LFS
4585 strcasestr NEW: GNU ext.
4586 strtoimax NEW: ISO C 9x
4587 strtoumax NEW: ISO C 9x
4588 strverscmp NEW: GNU ext.
4589 svcauthdes_stats NEW: Secure RPC
4590 svcunix_create NEW: sunrpc etx
4591 svcunixfd_create NEW: sunrpc ext
4592 swapcontext NEW: Unix98
4593 tcgetsid NEW: Unix98 PTY
4594 tdestroy NEW: GNU ext.
4595 tgamma NEW: ISO C 9x
4596 tgammaf NEW: ISO C 9x
4597 tgammal NEW: ISO C 9x
4598 tmpfile64 NEW: LFS
4599 trunc NEW: ISO C 9x
4600 truncate64 NEW: LFS
4601 truncf NEW: ISO C 9x
4602 truncl NEW: ISO C 9x
4603 umount2 NEW: kernel
4604 unlockpt NEW: Unix98 PTY
4605 updwtmpx NEW: Unix98
4606 user2netname NEW: Secure RPC
4607 utmpxname NEW: Unix98
4608 versionsort NEW: GNU ext.
4609 versionsort64 NEW: GNU ext.
4610 waitid NEW: Unix98
4611 wcscasecmp NEW: GNU ext.
4612 wcsncasecmp NEW: GNU ext.
4613 wcsnlen NEW: GNU ext.
4614 wcstoimax NEW: ISO C 9x
4615 wcstoll NEW: ISO C 9x
4616 wcstoull NEW: ISO C 9x
4617 wcstoumax NEW: ISO C 9x
4618 wcswcs NEW: Unix98
4619 wordexp NEW: POSIX.2
4620 wordfree NEW: POSIX.2
4621 write_profiling REMOVED
4622 xdecrypt NEW: Secure RPC
4623 xdr_authdes_cred NEW: Secure RPC
4624 xdr_authdes_verf NEW: Secure RPC
4625 xdr_cryptkeyarg NEW: Secure RPC
4626 xdr_cryptkeyarg2 NEW: Secure RPC
4627 xdr_cryptkeyres NEW: Secure RPC
4628 xdr_getcredres NEW: Secure RPC
4629 xdr_key_netstarg NEW: Secure RPC
4630 xdr_key_netstres NEW: Secure RPC
4631 xdr_keybuf NEW: Secure RPC
4632 xdr_keystatus NEW: Secure RPC
4633 xdr_netnamestr NEW: Secure RPC
4634 xdr_sizeof NEW: Secure RPC
4635 xdr_unixcred NEW: sunrpc ext
4636 xencrypt NEW: Secure RPC
4637 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4638 \f
4639 Version 2.0.6
4640
4641 * more bug fixes
4642
4643 \f
4644 Version 2.0.5
4645
4646 * more bug fixes
4647
4648 * inet_ntoa is thread-safe
4649
4650 * updwtmp is moved from libutil to libc
4651
4652 * rewrite of cbrt function
4653
4654 * update of timezone data
4655 \f
4656 Version 2.0.4
4657
4658 * more bug fixes
4659 \f
4660 Version 2.0.3
4661
4662 * more bug fixes
4663 \f
4664 Version 2.0.2
4665
4666 * more bug fixes
4667
4668 * add atoll function
4669
4670 * fix complex problems in Berkeley DB code
4671
4672 * fix math functions
4673 \f
4674 Version 2.0.1
4675
4676 * fixed lots of header problems (especially Linux/GNU specific)
4677
4678 * dynamic loader preserves all registers
4679
4680 * Roland McGrath provided support for handling of auxiliary objects in
4681 the ELF dynamic loader.
4682
4683 * support for parallel builds is improved
4684 \f
4685 Version 2.0
4686
4687 * GNU extensions are no longer declared by default. To enable them you
4688 must define the macro `_GNU_SOURCE' in your program or compile with
4689 `-D_GNU_SOURCE'.
4690
4691 * The library has changed from using GNU ld symbol aliases to using weak
4692 symbols where available. The ELF object file format supports weak
4693 symbols; GNU ld also supports weak symbols in the a.out format. (There
4694 is also now support for other GNU ld extensions in ELF. Use the
4695 `--with-elf' option to configure to indicate you have ELF, and
4696 `--with-gnu-ld' if using GNU ld.) This change resulted in the deletion
4697 of many files which contained only symbol aliases, reducing the size of
4698 the source and the compiled library; many other files were renamed to
4699 less cryptic names previously occupied by the symbol alias files.
4700 There is a new header file <elf.h> for programs which operate on
4701 files in the ELF format.
4702
4703 * Converted to Autoconf version 2, so `configure' has more options.
4704 Run `configure --help' to see the details.
4705
4706 * The library can now be configured to build profiling, highly-optimized
4707 (but undebuggable), and/or shared libraries (ELF with GNU ld only). The
4708 `--enable-profile', `--enable-omitfp', and `--enable-shared' options to
4709 `configure' enable building these extra libraries. The shared library is
4710 built by default when using both ELF and GNU ld. When shared libraries
4711 are enabled, the new library `-ldl' is available for arbitrary run-time
4712 loading of shared objects; its interface is defined in <dlfcn.h>. The
4713 new header file <link.h> gives access to the internals of the run-time
4714 dynamic linker, `ld.so'. The shell script `ldd' is similar to the
4715 application of same name on other systems and it provides information
4716 about dynamically linked binaries.
4717
4718 * The C library now provides the run-time support code for profiling
4719 executables compiled with `-pg'. Programs can control the profiling code
4720 through the interface in <sys/gmon.h>. The `gmon.out' files written by
4721 the GNU C library can be read only by GNU `gprof' (from GNU binutils);
4722 the support for this file format was contributed by David Mosberger-Tang.
4723
4724 * The math code has been replaced with a math library based on fdlibm from
4725 Sun, and modified by JT Conklin and Ulrich Drepper with i387 support, by
4726 Ian Taylor with `float' functions and by Ulrich Drepper with `long double'
4727 functions. The math functions now reside in a separate library, so
4728 programs using them will need to use `-lm' their linking commands.
4729
4730 * John C. Bowman contributed optimized ix87 assembler inline functions.
4731
4732 * Ulrich Drepper has contributed support for an `/etc/nsswitch.conf'
4733 mechanism similar to that found in Solaris 2. This is now used for the
4734 group, passwd, hosts, networks, services, protocols, rpc, ethers,
4735 shadow, netgroup, publickey, and alias databases. The `nsswitch.conf'
4736 file controls what services are used for each individual database. This
4737 works by loading shared libraries with names specified in `nsswitch.conf',
4738 so service modules can be changed or added at any time without even
4739 relinking any program. Currently there are the file, db, and NIS based
4740 NSS services available.
4741
4742 * The new functions `strtoq' and `strtouq' parse integer values from
4743 strings, like `strtol' and `strtoul', but they return `long long int' and
4744 `unsigned long long int' values, respectively (64-bit quantities).
4745
4746 * The new functions `strtof' and `strtold' parse floating-point values from
4747 strings, like `strtod', but they return `float' and `long double' values,
4748 respectively (on some machines `double' and `long double' are the same).
4749
4750 * Ulrich Drepper has contributed new implementations of the floating-point
4751 printing and reading code used in the `printf' family of functions and
4752 `strtod', `strtof', and `strtold'. These new functions are perfectly
4753 accurate, and much faster than the old ones.
4754
4755 * The implementation of the POSIX locale model was completely rewritten by
4756 Ulrich Drepper. This includes the new programs `localedef' and `locale'
4757 to compile the POSIX locale definition.
4758
4759 * The former dummy implementations of the strcoll and strxfrm function are
4760 now replaced by fully functional code contributed by Ulrich Drepper. The
4761 collation information comes from the POSIX locale definitions.
4762
4763 * The new header <langinfo.h> defines an interface for accessing
4764 various locale-dependent data (using the locale chosen with `setlocale').
4765
4766 * Ulrich Drepper has contributed a new suite of functions for operation on
4767 wide-character and multibyte-character strings, in <wchar.h>;
4768 and classification and case conversion of wide characters, in <wctype.h>.
4769 These new functions are conforming to the ISO C, Amendement 1 specification.
4770
4771 * There is now a second implementation of the standard I/O library available.
4772 It comes from GNU libg++ as was written by Per Bothner, heavily modified
4773 by Hongjiu Lu and made thread safe by Ulrich Drepper.
4774
4775 * You can now use positional parameter specifications in format strings
4776 for the `printf' and `scanf' families of functions. For example,
4777 `printf ("Number %2$d, Mr %1$s\n", "Jones", 6);'' prints
4778 ``Number 6, Mr Jones''. This is mainly useful when providing different
4779 format strings for different languages, whose grammars may dictate
4780 different orderings of the values being printed. To support this
4781 feature, the interface for `register_printf_handler' has changed; see
4782 the header file <printf.h> for details.
4783
4784 * The `printf' and `scanf' families of functions now understand a new
4785 formatting flag for numeric conversions: the ' flag (e.g. %'d or %'f) says
4786 to group numbers as indicated by the locale; for `scanf' and friends, this
4787 says to accept as valid only a number with all the proper grouping
4788 separators in the right places. In the default "C" locale, numbers are
4789 not grouped; but locales for specific countries will define the usual
4790 conventions (i.e. separate thousands with `,' in the US locale).
4791
4792 * The pgrp functions have been regularized, slightly incompatibly but much
4793 less confusingly. The core functions are now `getpgid' and `setpgid',
4794 which take arguments for the PID to operate on; the POSIX.1 `getpgrp' (no
4795 argument) and BSD `setpgrp' (identical to `setpgid') functions are
4796 provided for compatibility. There is no longer an incompatible `getpgrp'
4797 with an argument declared under _BSD_SOURCE; no BSD code uses it.
4798
4799 * The new header file <fts.h> and suite of functions simplify programs that
4800 operate on directory trees. This code comes from 4.4 BSD.
4801
4802 * The resolver code has been updated from the BIND 4.9.5-P1 release.
4803 Parts of the code were heavily modified by Ulrich Drepper to fit in the
4804 NSS scheme used in glibc.
4805
4806 * There is a new malloc debugging hook `__memalign_hook'.
4807
4808 * There are new typedefs `ushort' for `unsigned short int' and `uint' for
4809 `unsigned int' in <sys/types.h>. These are for compatibility only and
4810 their use is discouraged.
4811
4812 * The `-lmcheck' library to enable standard malloc debugging hooks is now
4813 done differently, so that it works even without GNU ld.
4814
4815 * New function `euidaccess' checks allowed access to a file like `access',
4816 but using the effective IDs instead of the real IDs.
4817
4818 * The time zone data files have been updated for the latest and greatest
4819 local time conventions of the countries of the world.
4820
4821 * The new function `dirfd' extracts the file descriptor used by a DIR stream;
4822 see <dirent.h>.
4823
4824 * The new functions `ecvt', `fcvt', and `gcvt' provide an obsolete interface
4825 for formatting floating-point numbers. They are provided only for
4826 compatibility; new programs should use `sprintf' instead. There are
4827 also equivalent function for the `long double' floating-point type and
4828 all functions also exist in a reentrant form.
4829
4830 * The new auxiliary library `-lutil' from 4.4 BSD contains various
4831 functions for maintaining the login-record files (primarily of use to
4832 system programs such as `login'), and convenient functions for
4833 allocating and initializing a pseudo-terminal (pty) device.
4834
4835 * Ulrich Drepper has contributed new support for System V style
4836 shared memory and IPC on systems that support it.
4837
4838 * Ulrich Drepper has contributed several miscellaneous new functions found
4839 in System V: The `hsearch' family of functions provide an effective
4840 implementation of hash tables; `a64l' and `l64a' provide a very simple
4841 binary to ASCII mapping; `drand48' and friends provide a 48-bit random
4842 number generator.
4843
4844 * Ulrich Drepper has contributed new reentrant counterparts for the
4845 `random' and `hsearch' families of functions; `random_r', `hsearch_r', etc.
4846
4847 * Ulrich Drepper has contributed new, highly-optimized versions of several
4848 string functions for the i486/Pentium family of processors.
4849
4850 * Ulrich Drepper has updated the Linux-specific code, based largely
4851 on work done in Hongjiu Lu's version of GNU libc for Linux.
4852 The GNU library now supports Linux versions 2.0.10 and later,
4853 using the ELF object file format (i[3456]86-*-linux).
4854
4855 * Andreas Schwab has ported the C library to Linux/m68k (m68k-*-linux).
4856
4857 * David Mosberger-Tang and Richard Henderson have ported the C library
4858 to Linux/Alpha (alpha-*-linux). Richard Henderson contributed the
4859 dynamic linking support for ELF/Alpha.
4860
4861 * Richard Henderson contributed several Alpha optimized assembler function
4862 for arithmetic and string handling.
4863
4864 * Ulrich Drepper has contributed a new set of message catalog functions to
4865 support multiple languages using the <libintl.h> interface, for use with
4866 his new package GNU gettext. Translation volunteers have contributed
4867 catalogs of the library's messages in Spanish, German, and Korean.
4868
4869 * For compatibility with XPG4, Ulrich Drepper has contributed the `gencat'
4870 program and the `catgets' function for reading the catalog files it
4871 creates. (The <libintl.h> interface is preferred; we include the
4872 <nl_types.h> interface using `catgets' only for source compatibility with
4873 programs already written to use it.)
4874
4875 * New header file <values.h> gives SVID-compatible names for <limits.h>
4876 constants.
4877
4878 * Various new macros, declarations, and small header files for compatibility
4879 with 4.4 BSD.
4880
4881 * New function `group_member' is a convenient way to check if a process has
4882 a given effective group ID.
4883
4884 * When using GCC 2.7 and later, the socket functions are now declared in a
4885 special way so that passing an argument of type `struct sockaddr_in *',
4886 `struct sockaddr_ns *', or `struct sockaddr_un *' instead of the generic
4887 `struct sockaddr *' type, does not generate a type-clash warning.
4888
4889 * New function `error' declared in header file <error.h> is a convenient
4890 function for printing error messages and optionally exiting; this is the
4891 canonical function used in GNU programs. The new functions `err', `warn',
4892 and friends in header file <err.h> are the canonical 4.4 BSD interface for
4893 doing the same thing.
4894
4895 * The <glob.h> interface has several new flags from 4.4 BSD that extend the
4896 POSIX.2 `glob' function to do ~ and {...} expansion.
4897
4898 * New function `unsetenv' complements `setenv' for compatibility with 4.4 BSD.
4899 `clearenv' which is used in POSIX.9 is also available.
4900
4901 * New function `getsid' returns session ID number on systems that support it.
4902
4903 * We have incorporated the 4.4 BSD `db' library (version 1.85). New header
4904 files <db.h> and <mpool.h> provide a rich set of functions for several
4905 types of simple databases stored in memory and in files, and <ndbm.h> is
4906 an old `ndbm'-compatible interface using the `db' functions. Link with
4907 `-ldb' to get these functions.
4908
4909 * New macro `strdupa' copies a string like `strdup', but uses local stack
4910 space from `alloca' instead of dynamic heap space from `malloc'.
4911
4912 * New function `strnlen' is like `strlen' but searches only a given maximum
4913 number of characters for the null terminator. `stpncpy', `strndup' and
4914 `strndupa' are similar variants for the `stpcpy', `strdup' and `strdupa'
4915 function.
4916
4917 * New function `statfs' in header <sys/statfs.h>.
4918
4919 * The new <argz.h> and <envz.h> interfaces contributed by Miles Bader
4920 provide convenient functions for operating on blocks of null-terminated
4921 strings.
4922
4923 * A new suite of functions in <utmp.h> handle all the details of reading
4924 and writing the utmp file.
4925
4926 * An implementation of the NIS/YP(tm) based NSS service was contributed by
4927 Thorsten Kukuk.
4928
4929 * Paul Eggert and Ulrich Drepper modified the `strftime' function to be
4930 completely POSIX compliant and also implemented the extended functionality
4931 to handle alternate digit representation and alternate era date formats.
4932
4933 * Ulrich Drepper provided an implementation of the `strptime' function
4934 defined in XPG4.2 which transforms a string into a `struct tm' value.
4935
4936 * Paul Eggert provided the tzselect shell script as part of the timezone
4937 code. The shell script makes it easy to select the correct timezone
4938 specification.
4939
4940 * The implementation of the malloc family of functions is completely replaced
4941 by a new implementation by Doug Lea with many improvements by Wolfram Gloger.
4942 The implementation uses the mmap function (if available) and it is
4943 optimized for the use in multi threaded programs.
4944
4945 * Ulrich Drepper contributed a MD5 "encryption" for the crypt family of
4946 functions. This new functionality is usable by specifying a special
4947 salt string and it is compatible with implementation on *BSD systems.
4948
4949 * Lots of functions from the XPG4.2 standard were added by Ulrich Drepper:
4950 `getsubopt' to handle second level command line options, `bsd_signal'
4951 to access BSD style `signal' functionality, the obsolete `regexp' style
4952 expression matcher.
4953
4954 * the `lchown' function is available on system which support this
4955 functionality.
4956
4957 * The implementation of the shadow password handling function was contributed
4958 by Ulrich Drepper.
4959
4960 * David Mosberger-Tang changed the SunRPC implementation to be 64bit safe.
4961
4962 * POSIX.1g support was added. The <sys/select.h> header is available,
4963 `isfdtype' and `pselect' are implemented. Craig Metz contributed an
4964 implementation of `getaddrinfo'.
4965 \f
4966 Version 1.09
4967
4968 * For cross-compilation you should now set `BUILD_CC' instead of `HOST_CC'.
4969
4970 * New header file <fstab.h> and new functions `getfsspec', `getfsent' and
4971 friends, for parsing /etc/fstab. This code comes from 4.4 BSD.
4972
4973 * The new function `daemon' from 4.4 BSD is useful for server programs that
4974 want to put themselves in the background.
4975
4976 * Joel Sherrill has contributed support for several standalone boards that
4977 run without an operating system.
4978
4979 * `printf', `scanf' and friends now accept a `q' type modifier for long
4980 long int as well as `ll'. Formats using these might be `%qu' or `%lld'.
4981
4982 * All of the code taken from BSD (notably most of the math and networking
4983 routines) has been updated from the BSD 4.4-Lite release.
4984
4985 * The resolver code has been updated from the BIND-4.9.3-BETA9 release.
4986
4987 * The new functions `getdomainname' and `setdomainname' fetch or change the
4988 YP/NIS domain name. These are system calls which exist on systems which
4989 have YP (aka NIS).
4990
4991 * The time zone data files have been updated for the latest international
4992 conventions.
4993
4994 * The SunRPC programs `portmap' and `rpcinfo' are now installed in
4995 $(sbindir) (usually /usr/local/sbin) instead of $(bindir).
4996 \f
4997 Version 1.08
4998
4999 * The C library now includes support for Sun RPC, from Sun's free
5000 RPCSRC-4.0 distribution. The `portmap', `rpcinfo', and `rpcgen' programs
5001 are included. (There is still no support for YP.)
5002
5003 * Tom Quinn has contributed a port of the C library to SGI machines running
5004 Irix 4 (mips-sgi-irix4).
5005
5006 * The new `lockf' function is a simplified interface to the locking
5007 facilities of `fcntl', included for compatibility.
5008
5009 * New time functions `timegm', `timelocal', and `dysize' for compatibility.
5010
5011 * New header file <sys/timeb.h> and new function `ftime' for compatibility.
5012
5013 * New header files <poll.h> and <sys/poll.h> and new function `poll' for
5014 compatibility.
5015
5016 * The error message printed by `assert' for a failed assertion now includes
5017 the name of the program (if using GNU ld) and the name of the calling
5018 function (with versions of GCC that support this).
5019
5020 * The `psignal' function is now declared in <signal.h>, not <stdio.h>.
5021
5022 * The library now includes the <sys/mman.h> header file and memory
5023 management functions `mmap', `munmap', `mprotect', `msync', and
5024 `madvise', on systems that support those facilities.
5025
5026 * The interface for `mcheck' has changed slightly: the function called to
5027 abort the program when an allocation inconsistency is detected now takes
5028 an argument that indicates the type of failure. The new function
5029 `mprobe' lets you request a consistency check for a particular block at
5030 any time (checks are normally done only when you call `free' or `realloc'
5031 on a block).
5032
5033 * It is now possible to easily cross-compile the C library, building on one
5034 system a library to run on another machine and/or operating system. All
5035 you need to do is set the variable `HOST_CC' in `configparms' to the
5036 native compiler for programs to run on the machine you are building on (a
5037 few generator programs are used on Unix systems); set `CC' to the
5038 cross-compiler.
5039
5040 * The new function `fexecve' (only implemented on the GNU system) executes
5041 a program file given a file descriptor already open on the file.
5042 \f
5043 Version 1.07
5044
5045 * Brendan Kehoe has contributed most of a port to the DEC Alpha
5046 running OSF/1 (alpha-dec-osf1). He says it is 75% complete.
5047
5048 * You can set the variable `libprefix' in `configparms' to specify a prefix
5049 to be prepended to installed library files; this makes it easy to install
5050 the GNU C library to be linked as `-lgnuc' or whatever.
5051
5052 * The new `stpncpy' is a cross between `stpcpy' and `strncpy': It
5053 copies a limited number of characters from a string, and returns the
5054 address of the last character written.
5055
5056 * You no longer need to check for whether the installed `stddef.h' is
5057 compatible with the GNU C library. configure now checks for you.
5058
5059 * You can now define a per-stream `fileno' function to convert the
5060 stream's cookie into an integral file descriptor.
5061
5062 * ``malloc (0)'' no longer returns a null pointer. Instead, it
5063 allocates zero bytes of storage, and returns a unique pointer which
5064 you can pass to `realloc' or `free'. The behavior is undefined if
5065 you dereference this pointer.
5066
5067 * The C library now runs on Sony NEWS m68k machines running either
5068 NewsOS 3 or NewsOS 4.
5069
5070 * The new `syscall' function is a system-dependent primitive function
5071 for invoking system calls. It has the canonical behavior on Unix
5072 systems, including unreliable return values for some calls (such as
5073 `pipe', `fork' and `getppid').
5074
5075 * The error code `EWOULDBLOCK' is now obsolete; it is always defined
5076 to `EAGAIN', which is the preferred name. On systems whose kernels
5077 use two distinct codes, the C library now translates EWOULDBLOCK to
5078 EAGAIN in every system call function.
5079 \f
5080 Version 1.06
5081
5082 * The GNU C Library Reference Manual is now distributed with the library.
5083 `make dvi' will produce a DVI file of the printed manual.
5084 `make info' will produce Info files that you can read on line using C-h i
5085 in Emacs or the `info' program.
5086 Please send comments on the manual to bug-glibc-manual@gnu.org.
5087
5088 * The library now supports SVR4 on i386s (i386-unknown-sysv4).
5089
5090 * Brendan Kehoe has contributed a port to Sun SPARCs running Solaris 2.
5091
5092 * Jason Merrill has contributed a port to the Sequent Symmetry running
5093 Dynix version 3 (i386-sequent-dynix).
5094
5095 * The library has been ported to i386s running SCO 3.2.4 (also known as SCO
5096 ODT 2.0; i386-unknown-sco3.2.4) or SCO 3.2 (i386-unknown-sco3.2).
5097
5098 * New function `memory_warnings' lets you arrange to get warnings when
5099 malloc is running out of memory to allocate, like Emacs gives you.
5100
5101 * The C library now contains the relocating allocator used in Emacs 19 for
5102 its editing buffers. This allocator (ralloc) minimizes allocation
5103 overhead and fragmentation by moving allocated regions around whenever it
5104 needs to. You always refer to a ralloc'd region with a "handle" (a
5105 pointer to a pointer--an object of type `void **').
5106
5107 * There is a new `printf' format: `%m' gives you the string corresponding
5108 to the error code in `errno'.
5109
5110 * In `scanf' formats, you can now use `%as' or `%a[' to do the normal `%s'
5111 or `%[' conversion, but instead of filling in a fixed-sized buffer you
5112 pass, the `a' modifier says to fill in a `char **' you pass with a
5113 malloc'd string.
5114
5115 * The `fnmatch' function supports the new flag bits `FNM_LEADING_DIR' and
5116 `FNM_CASEFOLD'. `FNM_LEADING_DIR' lets a pattern like `foo*' match a
5117 name like `foo/bar'. `FNM_CASEFOLD' says to ignore case in matching.
5118
5119 * `mkstemp' is a traditional Unix function to atomically create and open a
5120 uniquely-named temporary file.
5121 \f
5122 Version 1.05
5123
5124 * The standard location for the file that says what the local timezone is
5125 has changed again. It is now `/usr/local/etc/localtime' (or more
5126 precisely, `${prefix}/etc/localtime') rather than `/etc/localtime'.
5127
5128 * The distribution no longer contains any files with names longer than 14
5129 characters.
5130
5131 * `struct ttyent' has two new flag bits: TTY_TRUSTED and TTY_CONSOLE.
5132 These are set by the new `trusted' and `console' keywords in `/etc/ttys'.
5133
5134 * New functions `ttyslot' and `syslog' from 4.4 BSD.
5135 \f
5136 Version 1.04
5137
5138 * The configuration process has changed quite a bit. The `configure'
5139 script is now used just like the configuration scripts for other GNU
5140 packages. The `sysdeps' directory hierarchy is much rearranged.
5141 The file `INSTALL' explains the new scheme in detail.
5142
5143 * The header files no longer need to be processed into ANSI C and
5144 traditional C versions. There is just one set of files to install, and
5145 it will work with ANSI or old C compilers (including `gcc -traditional').
5146
5147 * Brendan Kehoe and Ian Lance Taylor have ported the library to the
5148 MIPS DECStation running Ultrix 4.
5149
5150 * The Sun 4 startup code (crt0) can now properly load SunOS 4 shared libraries.
5151 Tom Quinn contributed the initial code. The GNU C library can NOT yet be
5152 made itself into a shared library.
5153
5154 * Yet further improved support for the i386, running 4.3 BSD-like systems
5155 (such as Mach 3 with the Unix single-server), or System V.
5156
5157 * New function `strncasecmp' to do case-insensitive string comparison
5158 with limited length.
5159
5160 * New function `strsep' is a reentrant alternative to `strtok'.
5161
5162 * New functions `scandir' and `alphasort' for searching directories.
5163
5164 * New function `setenv' is a better interface to `putenv'.
5165
5166 * Ian Lance Taylor has contributed an implementation of the SVID `ftw'
5167 function for traversing a directory tree.
5168
5169 * The GNU obstack package is now also part of the C library.
5170 The new function `open_obstack_stream' creates a stdio stream that
5171 writes onto an obstack; `obstack_printf' and `obstack_vprintf' do
5172 formatted output directly to an obstack.
5173
5174 * Miscellaneous new functions: reboot, nice, sigaltstack (4.4 BSD only),
5175 cfmakeraw, getusershell, getpass, swab, getttyent, seteuid, setegid.
5176
5177 * `FNM_FILE_NAME' is another name for `FNM_PATHNAME', used with `fnmatch'.
5178
5179 * The new functions `strfry' and `memfrob' do mysterious and wonderful
5180 things to your strings.
5181
5182 * There are some new test programs: test-fseek, testmb, and testrand.
5183
5184 * Some work has been done to begin porting the library to 4.4 BSD and Linux.
5185 These ports are not finished, but are a good starting place for really
5186 supporting those systems.
5187
5188 * `/etc/localtime' is now the standard location for the file that says what
5189 the local timezone is, rather than `/usr/local/lib/zoneinfo/localtime'.
5190 This follows the general principle that `/etc' is the place for all local
5191 configuration files.
5192
5193 * The C library header files now use `extern "C"' when used by the C++
5194 compiler, so the C library should now work with C++ code.
5195
5196 * The header file <bstring.h> is gone. <string.h> now declares bcopy,
5197 bcmp, bzero, and ffs. (Update: nowadays these functions are declared
5198 in <strings.h>.)
5199
5200 * Mike Haertel (of GNU e?grep and malloc fame) has written a new sorting
5201 function which uses the `merge sort' algorithm, and is said to be
5202 significantly faster than the old GNU `qsort' function. Merge sort is
5203 now the standard `qsort' function. The new algorithm can require a lot
5204 of temporary storage; so, the old sorting function is called when the
5205 required storage is not available.
5206
5207 * The C library now includes Michael Glad's Ultra Fast Crypt, which
5208 provides the Unix `crypt' function, plus some other entry points.
5209
5210 * The code and header files taken from 4.4 BSD have been updated with the
5211 latest files released from Berkeley.
5212 \f
5213 ----------------------------------------------------------------------
5214 Copyright information:
5215
5216 Copyright (C) 1992-2018 Free Software Foundation, Inc.
5217
5218 Permission is granted to anyone to make or distribute verbatim copies
5219 of this document as received, in any medium, provided that the
5220 copyright notice and this permission notice are preserved,
5221 thus giving the recipient permission to redistribute in turn.
5222
5223 Permission is granted to distribute modified versions
5224 of this document, or of portions of it,
5225 under the above conditions, provided also that they
5226 carry prominent notices stating who last changed them.
5227 \f
5228 Local variables:
5229 version-control: never
5230 fill-column: 76
5231 End: