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