]> git.ipfire.org Git - thirdparty/glibc.git/blob - NEWS
Remove __get_clockfreq
[thirdparty/glibc.git] / NEWS
1 GNU C Library NEWS -- history of user-visible changes.
2 Copyright (C) 1992-2019 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 2.30
9
10 Major new features:
11
12 * Unicode 12.0.0 Support: Character encoding, character type info, and
13 transliteration tables are all updated to Unicode 12.0.0, using
14 generator scripts contributed by Mike FABIAN (Red Hat).
15
16 * The dynamic linker accepts the --preload argument to preload shared
17 objects, in addition to the LD_PRELOAD environment variable.
18
19 * On Linux, the gettid function has been added.
20
21 Deprecated and removed features, and other changes affecting compatibility:
22
23 * The functions clock_gettime, clock_getres, clock_settime,
24 clock_getcpuclockid, clock_nanosleep were removed from the librt library
25 for new applications (on architectures which had them). Instead, the
26 definitions in libc will be used automatically, which have been available
27 since glibc 2.17.
28
29 * The obsolete and never-implemented XSI STREAMS header files <stropts.h>
30 and <sys/stropts.h> have been removed.
31
32 Changes to build and runtime requirements:
33
34 * GCC 6.2 or later is required to build the GNU C Library.
35
36 Older GCC versions and non-GNU compilers are still supported when
37 compiling programs that use the GNU C Library.
38
39 Security related changes:
40
41 CVE-2019-7309: x86-64 memcmp used signed Jcc instructions to check
42 size. For x86-64, memcmp on an object size larger than SSIZE_MAX
43 has undefined behavior. On x32, the size_t argument may be passed
44 in the lower 32 bits of the 64-bit RDX register with non-zero upper
45 32 bits. When it happened with the sign bit of RDX register set,
46 memcmp gave the wrong result since it treated the size argument as
47 zero. Reported by H.J. Lu.
48
49 CVE-2019-9169: Attempted case-insensitive regular-expression match
50 via proceed_next_node in posix/regexec.c leads to heap-based buffer
51 over-read. Reported by Hongxu Chen.
52
53 The following bugs are resolved with this release:
54
55 [The release manager will add the list generated by
56 scripts/list-fixed-bugs.py just before the release.]
57
58 \f
59 Version 2.29
60
61 Major new features:
62
63 * The getcpu wrapper function has been added, which returns the currently
64 used CPU and NUMA node. This function is Linux-specific.
65
66 * A new convenience target has been added for distribution maintainers
67 to build and install all locales as directories with files. The new
68 target is run by issuing the following command in your build tree:
69 'make localedata/install-locale-files', with an optional DESTDIR
70 to set the install root if you wish to install into a non-default
71 configured location.
72
73 * Optimized generic exp, exp2, log, log2, pow, sinf, cosf, sincosf and tanf.
74
75 * The reallocarray function is now declared under _DEFAULT_SOURCE, not just
76 for _GNU_SOURCE, to match BSD environments.
77
78 * For powercp64le ABI, Transactional Lock Elision is now enabled iff kernel
79 indicates that it will abort the transaction prior to entering the kernel
80 (PPC_FEATURE2_HTM_NOSC on hwcap2). On older kernels the transaction is
81 suspended, and this caused some undefined side-effects issues by aborting
82 transactions manually. Glibc avoided it by abort transactions manually on
83 each syscall, but it lead to performance issues on newer kernels where the
84 HTM state is saved and restore lazily (the state being saved even when the
85 process actually does not use HTM).
86
87 * The functions posix_spawn_file_actions_addchdir_np and
88 posix_spawn_file_actions_addfchdir_np have been added, enabling
89 posix_spawn and posix_spawnp to run the new process in a different
90 directory. These functions are GNU extensions. The function
91 posix_spawn_file_actions_addchdir_np is similar to the Solaris function
92 of the same name.
93
94 * The popen and system do not run atfork handlers anymore (BZ#17490).
95 Although it is a possible POSIX violation, the POSIX rationale in
96 pthread_atfork documentation regarding atfork handlers is to handle
97 inconsistent mutex state after a fork call in a multi-threaded process.
98 In both popen and system there is no direct access to user-defined mutexes.
99
100 * Support for the C-SKY ABIV2 running on Linux has been added. This port
101 requires at least binutils-2.32, gcc-9.0, and linux-4.20. Two ABIs are
102 supported:
103 - C-SKY ABIV2 soft-float little-endian
104 - C-SKY ABIV2 hard-float little-endian
105
106 * strftime's default formatting of a locale's alternative year (%Ey)
107 has been changed to zero-pad the year to a minimum of two digits,
108 like "%y". This improves the display of Japanese era years during
109 the first nine years of a new era, and is expected to be harmless
110 for all other locales (only Japanese locales regularly have
111 alternative year numbers less than 10). Zero-padding can be
112 overridden with the '_' or '-' flags (which are GNU extensions).
113
114 * As a GNU extension, the '_' and '-' flags can now be applied to
115 "%EY" to control how the year number is formatted; they have the
116 same effect that they would on "%Ey".
117
118 Deprecated and removed features, and other changes affecting compatibility:
119
120 * The glibc.tune tunable namespace has been renamed to glibc.cpu and the
121 tunable glibc.tune.cpu has been renamed to glibc.cpu.name.
122
123 * The type of the pr_uid and pr_gid members of struct elf_prpsinfo, defined
124 in <sys/procfs.h>, has been corrected to match the type actually used by
125 the Linux kernel. This affects the size and layout of that structure on
126 MicroBlaze, MIPS (n64 ABI only), Nios II and RISC-V.
127
128 * For the MIPS n32 ABI, the type of the pr_sigpend and pr_sighold members of
129 struct elf_prstatus, and the pr_flag member of struct elf_prpsinfo,
130 defined in <sys/procfs.h>, has been corrected to match the type actually
131 used by the Linux kernel. This affects the size and layout of those
132 structures.
133
134 * An archaic GNU extension to scanf, under which '%as', '%aS', and '%a[...]'
135 meant to scan a string and allocate space for it with malloc, is now
136 restricted to programs compiled in C89 or C++98 mode with _GNU_SOURCE
137 defined. This extension conflicts with C99's use of '%a' to scan a
138 hexadecimal floating-point number, which is now available to programs
139 compiled as C99 or C++11 or higher, regardless of _GNU_SOURCE.
140
141 POSIX.1-2008 includes the feature of allocating a buffer for string input
142 with malloc, using the modifier letter 'm' instead. Programs using
143 '%as', '%aS', or '%a[...]' with the old GNU meaning should change to
144 '%ms', '%mS', or '%m[...]' respectively. Programs that wish to use the
145 C99 '%a' no longer need to avoid _GNU_SOURCE.
146
147 GCC's -Wformat warnings can detect most uses of this extension, as long
148 as all functions that call vscanf, vfscanf, or vsscanf are annotated with
149 __attribute__ ((format (scanf, ...))).
150
151 Changes to build and runtime requirements:
152
153 * Python 3.4 or later is required to build the GNU C Library.
154
155 * On most architectures, GCC 5 or later is required to build the GNU C
156 Library. (On powerpc64le, GCC 6.2 or later is still required, as before.)
157
158 Older GCC versions and non-GNU compilers are still supported when
159 compiling programs that use the GNU C Library.
160
161 Security related changes:
162
163 CVE-2018-19591: A file descriptor leak in if_nametoindex can lead to a
164 denial of service due to resource exhaustion when processing getaddrinfo
165 calls with crafted host names. Reported by Guido Vranken.
166
167 CVE-2019-6488: On x32, the size_t parameter may be passed in the lower
168 32 bits of a 64-bit register with with non-zero upper 32 bit. When it
169 happened, accessing the 32-bit size_t value as the full 64-bit register
170 in the assembly string/memory functions would cause a buffer overflow.
171 Reported by H.J. Lu.
172
173 CVE-2016-10739: The getaddrinfo function could successfully parse IPv4
174 addresses with arbitrary trailing characters, potentially leading to data
175 or command injection issues in applications.
176
177 The following bugs are resolved with this release:
178
179 [10425] localedata: it_IT/it_CH: LC_TIME format is wrong
180 [10496] localedata: 12h time representation in multiple locales faulty
181 [10797] localedata: it_IT locale numeric does not have a separator for
182 thousands
183 [11319] libc: dprintf doesn't handle errors properly
184 [16346] time: mktime: potentially unsafe use of localtime_offset
185 [17248] build: glibc should not sort CFLAGS (support gcc plugins and
186 --param options)
187 [17405] libc: Implement posix_spawn_file_actions_addchdir_np,
188 posix_spawn_file_actions_addfchdir_np
189 [17426] localedata: Indian locales: set the correct date format
190 [17490] stdio: popen should not invoke atfork handlers
191 [17783] libc: TIOCSER_TEMT conditions inconsistent
192 [18040] regex: use-after-free in regexec/get_subexp
193 [18093] libc: Corrupted aux-cache causes ldconfig to segfault
194 [20018] network: getaddrinfo should reject IP addresses with trailing
195 characters (CVE-2016-10739)
196 [20209] localedata: Spelling mistake for Sunday in Greenlandic kl_GL
197 [20271] libc: Missing "\n" in __libc_fatal calls
198 [20480] dynamic-link: Patch: ifunc not executable, crashes sudo qemu
199 [20544] libc: RFE: atexit, __cxa_atexit, on_exit should assert function
200 pointer argument is non-NULL
201 [21037] stdio: open_memstream and freopen
202 [21286] libc: bits/siginfo.h is missing enum definition for TRAP_HWBKPT
203 [21716] time: Crash in glibc's mktime in low-memory situations
204 [22834] stdio: Subprocess forked by popen may crash in Linux when
205 multithreads call popen
206 [22927] network: crash in vn_gai_enqueue_request if requests_tail was NULL
207 and pthread_create fails.
208 [23032] hurd: sysdeps/htl/pt-barrier-init.c:39: bad call to memcmp ?
209 [23125] libc: riscv64: endless loop when throwing an exception from a
210 constructor
211 [23275] nptl: Race in pthread_mutex_lock while promoting to
212 PTHREAD_MUTEX_ELISION_NP.
213 [23400] libc: stdlib/test-bz22786.c creates temporary files in glibc
214 source tree
215 [23479] math: [mips] bits/fenv.h should not define some macros for soft-
216 float
217 [23490] libc: sysdeps/unix/sysv/linux/x86/tst-cet-property-2.c:49: off by
218 one error
219 [23497] libc: readdir64@GLIBC_2.1 cannot parse the kernel directory stream
220 [23509] dynamic-link: CET enabled glibc is incompatible with the older
221 linker
222 [23520] nscd: nscd: Use-after-free in addgetnetgrentX and its callers
223 [23521] nss: get_next_alias nss_files file stream leak
224 [23538] nptl: Hang in pthread_cond_broadcast
225 [23562] libc: Wrong type for si_band in Linux-specific siginfo_t
226 [23578] regex: Invalid memory access if regex pattern contains NUL byte
227 [23579] libc: Errors misreported in preadv2
228 [23597] build: support/test-container.c doesn't work with different
229 filesystems
230 [23603] time: mktime signed integer overflow on large timestamps
231 [23606] libc: Missing ENDBR32 in sysdeps/i386/start.S
232 [23614] libc: powerpc: missing CFI register information in __mpn_*
233 functions
234 [23637] string: Generic strstr/strcasestr fails with huge needles
235 [23640] libc: no way to easily clear FD_CLOEXEC in
236 posix_spawn_file_actions_adddup2()
237 [23649] libc: [microblaze/mips/nios2/riscv] sys/procfs.h pr_uid, pr_gid
238 have wrong type
239 [23656] libc: [mips n32] sys/procfs.h pr_sigpend, pr_sighold, pr_flag have
240 wrong type
241 [23679] libc: gethostid: Missing NULL check for gethostbyname_r result
242 [23689] libc: Bug in documentation for rusage.ru_ixrss in
243 bits/types/struct_rusage.h
244 [23690] dynamic-link: Segfault in _dl_profile_fixup with a high number of
245 threads
246 [23707] dynamic-link: Missing unwind info in sysdeps/powerpc/powerpc32/dl-
247 start.S
248 [23709] string: glibc 2.25 lacks sse2 optimized strstr()
249 [23716] dynamic-link: _dl_runtime_resolve_shstk isn't selected properly
250 [23717] libc: glibc: stdlib/tst-setcontext9 test suite failure on
251 powerpc64le
252 [23724] localedata: Albanian date formats are incorrect
253 [23735] math: libnldbl_nonshared.a references internal libm symbols
254 [23740] localedata: kl_GL: Month names and date formats need update
255 [23744] regex: regex refactorings to remove BE, avoid duplication
256 [23745] time: mktime fix for Gnulib + coreutils
257 [23758] time: Improve the width of alternate representation for year in
258 strftime
259 [23783] libc: [mips] Missing CMSPAR bits/termios.h
260 [23789] time: mktime does not set errno on failure
261 [23791] localedata: Wrong monetary format for ca_ES locale
262 [23793] locale: c32rtomb and mbrtoc32 should not alias wcrtomb and mbrtowc
263 [23794] locale: c16rtomb does not handle surrogate pairs
264 [23821] libc: si_band in siginfo_t has wrong type long int on sparc64
265 [23822] math: ia64 static libm.a is missing exp2f, log2f and powf symbols
266 [23836] time: time/tst-mktime2 test failure on Arm (32-bit)
267 [23848] libc: [sparc] Some socket syscalls wrongly assumed to be present
268 [23861] nptl: rdlock stalls indefinitely on an unlocked pthread rwlock
269 [23862] libc: [sh] missing kernel-features.h undefines
270 [23864] libc: [riscv] missing kernel-features.h undefines
271 [23867] libc: [arm/microblaze] __ASSUME_MLOCK2 incorrect
272 [23907] malloc: Incorrect double-free malloc tcache check disregards
273 tcache size
274 [23913] libc: off-by-one in function maybe_script_execute in
275 sysdeps/posix/spawni.c
276 [23915] libc: [arm] __ASSUME_COPY_FILE_RANGE incorrect
277 [23923] locale: Add --no-hard-links option to localedef
278 [23927] network: Linux if_nametoindex() does not close descriptor
279 (CVE-2018-19591)
280 [23961] math: powf can overflow to inf without setting errno in non-
281 nearest rounding mode
282 [23967] libc: [2.28 Regression]: New sigaction implementation breaks m68k
283 [23972] libc: __old_getdents64 uses wrong d_off value on overflow
284 [23993] libc: glibc 2.29 doesn't build with gcc 4.9
285 [23995] localedata: Remove execution flags from localedata/locales/bi_VU
286 [24011] localedata: Fixed small type in comment for locale bs_BA
287 [24018] libc: gettext() may return NULL
288 [24022] build: riscv build failure with Linux kernel 4.20-rc7
289 [24023] build: [2.29 Regression] FAIL: elf/check-localplt
290 [24024] string: strerror() might set errno to ENOMEM due to -fno-math-
291 error
292 [24027] malloc: glibc: realloc() ncopies 32-bit integer overflow
293 [24034] libc: tst-cancel21-static fails with SIGBUS on pre-ARMv7 when
294 using GCC 8
295 [24046] localedata: en_US locale doesn't define date_fmt
296 [24063] manual: @var{errno} should be @code{errno}
297 [24066] soft-fp: Inconsistent _FP_W_TYPE_SIZE check
298 [24088] libc: VSCR field is not being correctly read in ucontext_t on
299 ppc64le
300 [24096] time: Specifying '_' or '-' flag for "%EY" does not produce the
301 expected result
302 [24097] string: Can't use 64-bit register for size_t in assembly codes for
303 x32 (CVE-2019-6488)
304 [24110] hurd: SS_DISABLE never set in stack_t value returned by
305 sigaltstack
306 [24112] network: Do not send DNS queries for non-host names (where all
307 answers will be rejected)
308 [24130] libc: alpha __remqu corrupts $f3 register
309
310 \f
311 Version 2.28
312
313 Major new features:
314
315 * The localization data for ISO 14651 is updated to match the 2016
316 Edition 4 release of the standard, this matches data provided by
317 Unicode 9.0.0. This update introduces significant improvements to the
318 collation of Unicode characters. This release deviates slightly from
319 the standard in that the collation element ordering for lowercase and
320 uppercase LATIN script characters is adjusted to ensure that regular
321 expressions with ranges like [a-z] and [A-Z] don't interleave e.g. A
322 is not matched by [a-z]. With the update many locales have been
323 updated to take advantage of the new collation information. The new
324 collation information has increased the size of the compiled locale
325 archive or binary locales.
326
327 * The GNU C Library can now be compiled with support for Intel CET, AKA
328 Intel Control-flow Enforcement Technology. When the library is built
329 with --enable-cet, the resulting glibc is protected with indirect
330 branch tracking (IBT) and shadow stack (SHSTK). CET-enabled glibc is
331 compatible with all existing executables and shared libraries. This
332 feature is currently supported on i386, x86_64 and x32 with GCC 8 and
333 binutils 2.29 or later. Note that CET-enabled glibc requires CPUs
334 capable of multi-byte NOPs, like x86-64 processors as well as Intel
335 Pentium Pro or newer. NOTE: --enable-cet has been tested for i686,
336 x86_64 and x32 on non-CET processors. --enable-cet has been tested
337 for x86_64 and x32 on CET SDVs, but Intel CET support hasn't been
338 validated for i686.
339
340 * The GNU C Library now has correct support for ABSOLUTE symbols
341 (SHN_ABS-relative symbols). Previously such ABSOLUTE symbols were
342 relocated incorrectly or in some cases discarded. The GNU linker can
343 make use of the newer semantics, but it must communicate it to the
344 dynamic loader by setting the ELF file's identification (EI_ABIVERSION
345 field) to indicate such support is required.
346
347 * Unicode 11.0.0 Support: Character encoding, character type info, and
348 transliteration tables are all updated to Unicode 11.0.0, using
349 generator scripts contributed by Mike FABIAN (Red Hat).
350
351 * <math.h> functions that round their results to a narrower type are added
352 from TS 18661-1:2014 and TS 18661-3:2015:
353
354 - fadd, faddl, daddl and corresponding fMaddfN, fMaddfNx, fMxaddfN and
355 fMxaddfNx functions.
356
357 - fsub, fsubl, dsubl and corresponding fMsubfN, fMsubfNx, fMxsubfN and
358 fMxsubfNx functions.
359
360 - fmul, fmull, dmull and corresponding fMmulfN, fMmulfNx, fMxmulfN and
361 fMxmulfNx functions.
362
363 - fdiv, fdivl, ddivl and corresponding fMdivfN, fMdivfNx, fMxdivfN and
364 fMxdivfNx functions.
365
366 * Two grammatical forms of month names are now supported for the following
367 languages: Armenian, Asturian, Catalan, Czech, Kashubian, Occitan, Ossetian,
368 Scottish Gaelic, Upper Sorbian, and Walloon. The following languages now
369 support two grammatical forms in abbreviated month names: Catalan, Greek,
370 and Kashubian.
371
372 * Newly added locales: Lower Sorbian (dsb_DE) and Yakut (sah_RU) also
373 include the support for two grammatical forms of month names.
374
375 * Building and running on GNU/Hurd systems now works without out-of-tree
376 patches.
377
378 * The renameat2 function has been added, a variant of the renameat function
379 which has a flags argument. If the flags are zero, the renameat2 function
380 acts like renameat. If the flag is not zero and there is no kernel
381 support for renameat2, the function will fail with an errno value of
382 EINVAL. This is different from the existing gnulib function renameatu,
383 which performs a plain rename operation in case of a RENAME_NOREPLACE
384 flags and a non-existing destination (and therefore has a race condition
385 that can clobber the destination inadvertently).
386
387 * The statx function has been added, a variant of the fstatat64
388 function with an additional flags argument. If there is no direct
389 kernel support for statx, glibc provides basic stat support based on
390 the fstatat64 function.
391
392 * IDN domain names in getaddrinfo and getnameinfo now use the system libidn2
393 library if installed. libidn2 version 2.0.5 or later is recommended. If
394 libidn2 is not available, internationalized domain names are not encoded
395 or decoded even if the AI_IDN or NI_IDN flags are passed to getaddrinfo or
396 getnameinfo. (getaddrinfo calls with non-ASCII names and AI_IDN will fail
397 with an encoding error.) Flags which used to change the IDN encoding and
398 decoding behavior (AI_IDN_ALLOW_UNASSIGNED, AI_IDN_USE_STD3_ASCII_RULES,
399 NI_IDN_ALLOW_UNASSIGNED, NI_IDN_USE_STD3_ASCII_RULES) have been
400 deprecated. They no longer have any effect.
401
402 * Parsing of dynamic string tokens in DT_RPATH, DT_RUNPATH, DT_NEEDED,
403 DT_AUXILIARY, and DT_FILTER has been expanded to support the full
404 range of ELF gABI expressions including such constructs as
405 '$ORIGIN$ORIGIN' (if valid). For SUID/GUID applications the rules
406 have been further restricted, and where in the past a dynamic string
407 token sequence may have been interpreted as a literal string it will
408 now cause a load failure. These load failures were always considered
409 unspecified behaviour from the perspective of the dynamic loader, and
410 for safety are now load errors e.g. /foo/${ORIGIN}.so in DT_NEEDED
411 results in a load failure now.
412
413 * Support for ISO C threads (ISO/IEC 9899:2011) has been added. The
414 implementation includes all the standard functions provided by
415 <threads.h>:
416
417 - thrd_current, thrd_equal, thrd_sleep, thrd_yield, thrd_create,
418 thrd_detach, thrd_exit, and thrd_join for thread management.
419
420 - mtx_init, mtx_lock, mtx_timedlock, mtx_trylock, mtx_unlock, and
421 mtx_destroy for mutual exclusion.
422
423 - call_once for function call synchronization.
424
425 - cnd_broadcast, cnd_destroy, cnd_init, cnd_signal, cnd_timedwait, and
426 cnd_wait for conditional variables.
427
428 - tss_create, tss_delete, tss_get, and tss_set for thread-local storage.
429
430 Application developers must link against libpthread to use ISO C threads.
431
432 Deprecated and removed features, and other changes affecting compatibility:
433
434 * The nonstandard header files <libio.h> and <_G_config.h> are no longer
435 installed. Software that was using either header should be updated to
436 use standard <stdio.h> interfaces instead.
437
438 * The stdio functions 'getc' and 'putc' are no longer defined as macros.
439 This was never required by the C standard, and the macros just expanded
440 to call alternative names for the same functions. If you hoped getc and
441 putc would provide performance improvements over fgetc and fputc, instead
442 investigate using (f)getc_unlocked and (f)putc_unlocked, and, if
443 necessary, flockfile and funlockfile.
444
445 * All stdio functions now treat end-of-file as a sticky condition. If you
446 read from a file until EOF, and then the file is enlarged by another
447 process, you must call clearerr or another function with the same effect
448 (e.g. fseek, rewind) before you can read the additional data. This
449 corrects a longstanding C99 conformance bug. It is most likely to affect
450 programs that use stdio to read interactive input from a terminal.
451 (Bug #1190.)
452
453 * The macros 'major', 'minor', and 'makedev' are now only available from
454 the header <sys/sysmacros.h>; not from <sys/types.h> or various other
455 headers that happen to include <sys/types.h>. These macros are rarely
456 used, not part of POSIX nor XSI, and their names frequently collide with
457 user code; see https://sourceware.org/bugzilla/show_bug.cgi?id=19239 for
458 further explanation.
459
460 <sys/sysmacros.h> is a GNU extension. Portable programs that require
461 these macros should first include <sys/types.h>, and then include
462 <sys/sysmacros.h> if __GNU_LIBRARY__ is defined.
463
464 * The tilegx*-*-linux-gnu configurations are no longer supported.
465
466 * The obsolete function ustat is no longer available to newly linked
467 binaries; the headers <ustat.h> and <sys/ustat.h> have been removed. This
468 function has been deprecated in favor of fstatfs and statfs.
469
470 * The obsolete function nfsservctl is no longer available to newly linked
471 binaries. This function was specific to systems using the Linux kernel
472 and could not usefully be used with the GNU C Library on systems with
473 version 3.1 or later of the Linux kernel.
474
475 * The obsolete function name llseek is no longer available to newly linked
476 binaries. This function was specific to systems using the Linux kernel
477 and was not declared in a header. Programs should use the lseek64 name
478 for this function instead.
479
480 * The AI_IDN_ALLOW_UNASSIGNED and NI_IDN_ALLOW_UNASSIGNED flags for the
481 getaddrinfo and getnameinfo functions have been deprecated. The behavior
482 previously selected by them is now always enabled.
483
484 * The AI_IDN_USE_STD3_ASCII_RULES and NI_IDN_USE_STD3_ASCII_RULES flags for
485 the getaddrinfo and getnameinfo functions have been deprecated. The STD3
486 restriction (rejecting '_' in host names, among other things) has been
487 removed, for increased compatibility with non-IDN name resolution.
488
489 * The fcntl function now have a Long File Support variant named fcntl64. It
490 is added to fix some Linux Open File Description (OFD) locks usage on non
491 LFS mode. As for others *64 functions, fcntl64 semantics are analogous with
492 fcntl and LFS support is handled transparently. Also for Linux, the OFD
493 locks act as a cancellation entrypoint.
494
495 * The obsolete functions encrypt, encrypt_r, setkey, setkey_r, cbc_crypt,
496 ecb_crypt, and des_setparity are no longer available to newly linked
497 binaries, and the headers <rpc/des_crypt.h> and <rpc/rpc_des.h> are no
498 longer installed. These functions encrypted and decrypted data with the
499 DES block cipher, which is no longer considered secure. Software that
500 still uses these functions should switch to a modern cryptography library,
501 such as libgcrypt.
502
503 * Reflecting the removal of the encrypt and setkey functions above, the
504 macro _XOPEN_CRYPT is no longer defined. As a consequence, the crypt
505 function is no longer declared unless _DEFAULT_SOURCE or _GNU_SOURCE is
506 enabled.
507
508 * The obsolete function fcrypt is no longer available to newly linked
509 binaries. It was just another name for the standard function crypt,
510 and it has not appeared in any header file in many years.
511
512 * We have tentative plans to hand off maintenance of the passphrase-hashing
513 library, libcrypt, to a separate development project that will, we hope,
514 keep up better with new passphrase-hashing algorithms. We will continue
515 to declare 'crypt' in <unistd.h>, and programs that use 'crypt' or
516 'crypt_r' should not need to change at all; however, distributions will
517 need to install <crypt.h> and libcrypt from a separate project.
518
519 In this release, if the configure option --disable-crypt is used, glibc
520 will not install <crypt.h> or libcrypt, making room for the separate
521 project's versions of these files. The plan is to make this the default
522 behavior in a future release.
523
524 Changes to build and runtime requirements:
525
526 GNU make 4.0 or later is now required to build glibc.
527
528 Security related changes:
529
530 CVE-2016-6261, CVE-2016-6263, CVE-2017-14062: Various vulnerabilities have
531 been fixed by removing the glibc-internal IDNA implementation and using
532 the system-provided libidn2 library instead. Originally reported by Hanno
533 Böck and Christian Weisgerber.
534
535 CVE-2017-18269: An SSE2-based memmove implementation for the i386
536 architecture could corrupt memory. Reported by Max Horn.
537
538 CVE-2018-11236: Very long pathname arguments to realpath function could
539 result in an integer overflow and buffer overflow. Reported by Alexey
540 Izbyshev.
541
542 CVE-2018-11237: The mempcpy implementation for the Intel Xeon Phi
543 architecture could write beyond the target buffer, resulting in a buffer
544 overflow. Reported by Andreas Schwab.
545
546 The following bugs are resolved with this release:
547
548 [1190] stdio: fgetc()/fread() behaviour is not POSIX compliant
549 [6889] manual: 'PWD' mentioned but not specified
550 [13575] libc: SSIZE_MAX defined as LONG_MAX is inconsistent with ssize_t,
551 when __WORDSIZE != 64
552 [13762] regex: re_search etc. should return -2 on memory exhaustion
553 [13888] build: /tmp usage during testing
554 [13932] math: dbl-64 pow unexpectedly slow for some inputs
555 [14092] nptl: Support C11 threads
556 [14095] localedata: Review / update collation data from Unicode / ISO
557 14651
558 [14508] libc: -Wformat warnings
559 [14553] libc: Namespace pollution loff_t in sys/types.h
560 [14890] libc: Make NT_PRFPREG canonical.
561 [15105] libc: Extra PLT references with -Os
562 [15512] libc: __bswap_constant_16 not compiled when -Werror -Wsign-
563 conversion is given
564 [16335] manual: Feature test macro documentation incomplete and out of
565 date
566 [16552] libc: Unify umount implementations in terms of umount2
567 [17082] libc: htons et al.: statement-expressions prevent use on global
568 scope with -O1 and higher
569 [17343] libc: Signed integer overflow in /stdlib/random_r.c
570 [17438] localedata: pt_BR: wrong d_fmt delimiter
571 [17662] libc: please implement binding for the new renameat2 syscall
572 [17721] libc: __restrict defined as /* Ignore */ even in c11
573 [17979] libc: inconsistency between uchar.h and stdint.h
574 [18018] dynamic-link: Additional $ORIGIN handling issues (CVE-2011-0536)
575 [18023] libc: extend_alloca is broken (questionable pointer comparison,
576 horrible machine code)
577 [18124] libc: hppa: setcontext erroneously returns -1 as exit code for
578 last constant.
579 [18471] libc: llseek should be a compat symbol
580 [18473] soft-fp: [powerpc-nofpu] __sqrtsf2, __sqrtdf2 should be compat
581 symbols
582 [18991] nss: nss_files skips large entry in database
583 [19239] libc: Including stdlib.h ends up with macros major and minor being
584 defined
585 [19463] libc: linknamespace failures when compiled with -Os
586 [19485] localedata: csb_PL: Update month translations + add yesstr/nostr
587 [19527] locale: Normalized charset name not recognized by setlocale
588 [19667] string: Missing Sanity Check for malloc calls in file 'testcopy.c'
589 [19668] libc: Missing Sanity Check for malloc() in file 'tst-setcontext-
590 fpscr.c'
591 [19728] network: out of bounds stack read in libidn function
592 idna_to_ascii_4i (CVE-2016-6261)
593 [19729] network: out of bounds heap read on invalid utf-8 inputs in
594 stringprep_utf8_nfkc_normalize (CVE-2016-6263)
595 [19818] dynamic-link: Absolute (SHN_ABS) symbols incorrectly relocated by
596 the base address
597 [20079] libc: Add SHT_X86_64_UNWIND to elf.h
598 [20251] libc: 32bit programs pass garbage in struct flock for OFD locks
599 [20419] dynamic-link: files with large allocated notes crash in
600 open_verify
601 [20530] libc: bswap_16 should use __builtin_bswap16() when available
602 [20890] dynamic-link: ldconfig: fsync the files before atomic rename
603 [20980] manual: CFLAGS environment variable replaces vital options
604 [21163] regex: Assertion failure in pop_fail_stack when executing a
605 malformed regexp (CVE-2015-8985)
606 [21234] manual: use of CFLAGS makes glibc detect no optimization
607 [21269] dynamic-link: i386 sigaction sa_restorer handling is wrong
608 [21313] build: Compile Error GCC 5.4.0 MIPS with -0S
609 [21314] build: Compile Error GCC 5.2.0 MIPS with -0s
610 [21508] locale: intl/tst-gettext failure with latest msgfmt
611 [21547] localedata: Tibetan script collation broken (Dzongkha and Tibetan)
612 [21812] network: getifaddrs() returns entries with ifa_name == NULL
613 [21895] libc: ppc64 setjmp/longjmp not fully interoperable with static
614 dlopen
615 [21942] dynamic-link: _dl_dst_substitute incorrectly handles $ORIGIN: with
616 AT_SECURE=1
617 [22241] localedata: New locale: Yakut (Sakha) locale for Russia (sah_RU)
618 [22247] network: Integer overflow in the decode_digit function in
619 puny_decode.c in libidn (CVE-2017-14062)
620 [22342] nscd: NSCD not properly caching netgroup
621 [22391] nptl: Signal function clear NPTL internal symbols inconsistently
622 [22550] localedata: es_ES locale (and other es_* locales): collation
623 should treat ñ as a primary different character, sync the collation
624 for Spanish with CLDR
625 [22638] dynamic-link: sparc: static binaries are broken if glibc is built
626 by gcc configured with --enable-default-pie
627 [22639] time: year 2039 bug for localtime etc. on 64-bit platforms
628 [22644] string: memmove-sse2-unaligned on 32bit x86 produces garbage when
629 crossing 2GB threshold (CVE-2017-18269)
630 [22646] localedata: redundant data (LC_TIME) for es_CL, es_CU, es_EC and
631 es_BO
632 [22735] time: Misleading typo in time.h source comment regarding
633 CLOCKS_PER_SECOND
634 [22753] libc: preadv2/pwritev2 fallback code should handle offset=-1
635 [22761] libc: No trailing `%n' conversion specifier in FMT passed from
636 `__assert_perror_fail ()' to `__assert_fail_base ()'
637 [22766] libc: all glibc internal dlopen should use RTLD_NOW for robust
638 dlopen failures
639 [22786] libc: Stack buffer overflow in realpath() if input size is close
640 to SSIZE_MAX (CVE-2018-11236)
641 [22787] dynamic-link: _dl_check_caller returns false when libc is linked
642 through an absolute DT_NEEDED path
643 [22792] build: tcb-offsets.h dependency dropped
644 [22797] libc: pkey_get() uses non-reserved name of argument
645 [22807] libc: PTRACE_* constants missing for powerpc
646 [22818] glob: posix/tst-glob_lstat_compat failure on alpha
647 [22827] dynamic-link: RISC-V ELF64 parser mis-reads flag in ldconfig
648 [22830] malloc: malloc_stats doesn't restore cancellation state on stderr
649 [22848] localedata: ca_ES: update date definitions from CLDR
650 [22862] build: _DEFAULT_SOURCE is defined even when _ISOC11_SOURCE is
651 [22884] math: RISCV fmax/fmin handle signalling NANs incorrectly
652 [22896] localedata: Update locale data for an_ES
653 [22902] math: float128 test failures with GCC 8
654 [22918] libc: multiple common of `__nss_shadow_database'
655 [22919] libc: sparc32: backtrace yields infinite backtrace with
656 makecontext
657 [22926] libc: FTBFS on powerpcspe
658 [22932] localedata: lt_LT: Update of abbreviated month names from CLDR
659 required
660 [22937] localedata: Greek (el_GR, el_CY) locales actually need ab_alt_mon
661 [22947] libc: FAIL: misc/tst-preadvwritev2
662 [22963] localedata: cs_CZ: Add alternative month names
663 [22987] math: [powerpc/sparc] fdim inlines errno, exceptions handling
664 [22996] localedata: change LC_PAPER to en_US in es_BO locale
665 [22998] dynamic-link: execstack tests are disabled when SELinux is
666 disabled
667 [23005] network: Crash in __res_context_send after memory allocation
668 failure
669 [23007] math: strtod cannot handle -nan
670 [23024] nss: getlogin_r is performing NSS lookups when loginid isn't set
671 [23036] regex: regex equivalence class regression
672 [23037] libc: initialize msg_flags to zero for sendmmsg() calls
673 [23069] libc: sigaction broken on riscv64-linux-gnu
674 [23094] localedata: hr_HR: wrong thousands_sep and mon_thousands_sep
675 [23102] dynamic-link: Incorrect parsing of multiple consecutive $variable
676 patterns in runpath entries (e.g. $ORIGIN$ORIGIN)
677 [23137] nptl: s390: pthread_join sometimes block indefinitely (on 31bit
678 and libc build with -Os)
679 [23140] localedata: More languages need two forms of month names
680 [23145] libc: _init/_fini aren't marked as hidden
681 [23152] localedata: gd_GB: Fix typo in "May" (abbreviated)
682 [23171] math: C++ iseqsig for long double converts arguments to double
683 [23178] nscd: sudo will fail when it is run in concurrent with commands
684 that changes /etc/passwd
685 [23196] string: __mempcpy_avx512_no_vzeroupper mishandles large copies
686 (CVE-2018-11237)
687 [23206] dynamic-link: static-pie + dlopen breaks debugger interaction
688 [23208] localedata: New locale - Lower Sorbian (dsb)
689 [23233] regex: Memory leak in build_charclass_op function in file
690 posix/regcomp.c
691 [23236] stdio: Harden function pointers in _IO_str_fields
692 [23250] nptl: Offset of __private_ss differs from GCC
693 [23253] math: tgamma test suite failures on i686 with -march=x86-64
694 -mtune=generic -mfpmath=sse
695 [23259] dynamic-link: Unsubstituted ${ORIGIN} remains in DT_NEEDED for
696 AT_SECURE
697 [23264] libc: posix_spawnp wrongly executes ENOEXEC in non compat mode
698 [23266] nis: stringop-truncation warning with new gcc8.1 in nisplus-
699 parser.c
700 [23272] math: fma(INFINITY,INFIITY,0.0) should be INFINITY
701 [23277] math: nan function should not have const attribute
702 [23279] math: scanf and strtod wrong for some hex floating-point
703 [23280] math: wscanf rounds wrong; wcstod is ok for negative numbers and
704 directed rounding
705 [23290] localedata: IBM273 is not equivalent to ISO-8859-1
706 [23303] build: undefined reference to symbol
707 '__parse_hwcap_and_convert_at_platform@@GLIBC_2.23'
708 [23307] dynamic-link: Absolute symbols whose value is zero ignored in
709 lookup
710 [23313] stdio: libio vtables validation and standard file object
711 interposition
712 [23329] libc: The __libc_freeres infrastructure is not properly run across
713 DSO boundaries.
714 [23349] libc: Various glibc headers no longer compatible with
715 <linux/time.h>
716 [23351] malloc: Remove unused code related to heap dumps and malloc
717 checking
718 [23363] stdio: stdio-common/tst-printf.c has non-free license
719 [23396] regex: Regex equivalence regression in single-byte locales
720 [23422] localedata: oc_FR: More updates of locale data
721 [23442] build: New warning with GCC 8
722 [23448] libc: Out of bounds access in IBM-1390 converter
723 [23456] libc: Wrong index_cpu_LZCNT
724 [23458] build: tst-get-cpu-features-static isn't added to tests
725 [23459] libc: COMMON_CPUID_INDEX_80000001 isn't populated for Intel
726 processors
727 [23467] dynamic-link: x86/CET: A property note parser bug
728
729 \f
730 Version 2.27
731
732 Major new features:
733
734 * The GNU C Library can now be compiled with support for building static
735 PIE executables (See --enable-static-pie in INSTALL). These static PIE
736 executables are like static executables but can be loaded at any address
737 and provide additional security hardening benefits at the cost of some
738 memory and performance. When the library is built with --enable-static-pie
739 the resulting libc.a is usable with GCC 8 and above to create static PIE
740 executables using the GCC option '-static-pie'. This feature is currently
741 supported on i386, x86_64 and x32 with binutils 2.29 or later, and on
742 aarch64 with binutils 2.30 or later.
743
744 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf,
745 sinf, sincosf and tan with FMA, contributed by Arjan van de Ven and
746 H.J. Lu from Intel.
747
748 * Optimized x86-64 trunc and truncf for processors with SSE4.1.
749
750 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf.
751
752 * In order to support faster and safer process termination the malloc API
753 family of functions will no longer print a failure address and stack
754 backtrace after detecting heap corruption. The goal is to minimize the
755 amount of work done after corruption is detected and to avoid potential
756 security issues in continued process execution. Reducing shutdown time
757 leads to lower overall process restart latency, so there is benefit both
758 from a security and performance perspective.
759
760 * The abort function terminates the process immediately, without flushing
761 stdio streams. Previous glibc versions used to flush streams, resulting
762 in deadlocks and further data corruption. This change also affects
763 process aborts as the result of assertion failures.
764
765 * On platforms where long double has the IEEE binary128 format (aarch64,
766 alpha, mips64, riscv, s390 and sparc), the math library now implements
767 _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015.
768 These are the same interfaces added in version 2.26 for some platforms where
769 this format is supported but is not the format of long double.
770
771 * On platforms with support for _Float64x (aarch64, alpha, i386, ia64,
772 mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now
773 implements interfaces for that type, as defined by ISO/IEC TS
774 18661-3:2015. These are corresponding interfaces to those supported for
775 _Float128.
776
777 * The math library now implements interfaces for the _Float32, _Float64 and
778 _Float32x types, as defined by ISO/IEC TS 18661-3:2015. These are
779 corresponding interfaces to those supported for _Float128.
780
781 * glibc now implements the memfd_create and mlock2 functions on Linux.
782
783 * Support for memory protection keys was added. The <sys/mman.h> header now
784 declares the functions pkey_alloc, pkey_free, pkey_mprotect, pkey_set,
785 pkey_get.
786
787 * The copy_file_range function was added.
788
789 * Optimized memcpy, mempcpy, memmove, and memset for sparc M7.
790
791 * The ldconfig utility now processes `include' directives using the C/POSIX
792 collation ordering. Previous glibc versions used locale-specific
793 ordering, the change might break systems that relied on that.
794
795 * Support for two grammatical forms of month names has been added.
796 In a call to strftime, the "%B" and "%b" format specifiers will now
797 produce the grammatical form required when the month is used as part
798 of a complete date. New "%OB" and "%Ob" specifiers produce the form
799 required when the month is named by itself. For instance, in Greek
800 and in many Slavic and Baltic languages, "%B" will produce the month
801 in genitive case, and "%OB" will produce the month in nominative case.
802
803 In a call to strptime, "%B", "%b", "%h", "%OB", "%Ob", and "%Oh"
804 are all valid and will all accept any known form of month
805 name---standalone or complete, abbreviated or full. In a call to
806 nl_langinfo, the query constants MON_1..12 and ABMON_1..12 return
807 the strings used by "%B" and "%b", respectively. New query
808 constants ALTMON_1..12 and _NL_ABALTMON_1..12 return the strings
809 used by "%OB" and "%Ob", respectively.
810
811 In a locale definition file, use "alt_mon" and "ab_alt_mon" to
812 define the strings for %OB and %Ob, respectively; these have the
813 same syntax as "mon" and "abmon". These arrays are optional; if they
814 are not provided then they have the same content as "mon" and "abmon",
815 respectively.
816
817 These features are provided for locales which define "alt_mon" and/or
818 "ab_alt_mon" in their locale source data. This release includes such
819 alternative month name data for the following languages: Belarusian,
820 Croatian, Greek, Lithuanian, Polish, Russian, and Ukrainian.
821
822 This feature is currently a GNU extension, but it is expected to
823 be added to the next revision of POSIX, and it is also already
824 available on some BSD-derived operating systems.
825
826 This feature will cause existing statically compiled applications
827 to fail to load locales and fall back to the builtin C/POSIX locales.
828 See notes below for other changes affecting compatibility.
829
830 * Support for the RISC-V ISA running on Linux has been added. This port
831 requires at least binutils-2.30, gcc-7.3.0, and linux-4.15; and is supported
832 for the following ISA and ABI pairs:
833
834 - rv64imac lp64
835 - rv64imafdc lp64
836 - rv64imafdc lp64d
837
838 Deprecated and removed features, and other changes affecting compatibility:
839
840 * Statically compiled applications attempting to load locales compiled for the
841 GNU C Library version 2.27 will fail and fall back to the builtin C/POSIX
842 locale. The reason for this is that the addition of the new "%OB" and "%Ob",
843 support for two grammatical forms of the month names, also extends the locale
844 data binary format. Static applications needing locale support must be
845 recompiled to match the runtime and data they are deployed with. In some
846 distributions there is an upgrade window where dynamically linked applications
847 may use a new library but the old locale data and also fall back to the
848 builtin C/POSIX locales; restarting the application process is sufficient to
849 fix this.
850
851 * Support for statically linked applications which call dlopen is deprecated
852 and will be removed in a future version of glibc. Applications which call
853 dlopen need to be linked dynamically instead.
854
855 * Support for old programs which use internal stdio data structures and
856 functions is deprecated. This includes programs which use the C++ streams
857 provided by libstdc++ in GCC 2.95. Programs which use the internal
858 symbols _IO_adjust_wcolumn, _IO_default_doallocate, _IO_default_finish,
859 _IO_default_pbackfail, _IO_default_uflow, _IO_default_xsgetn,
860 _IO_default_xsputn, _IO_doallocbuf, _IO_do_write, _IO_file_attach,
861 _IO_file_close, _IO_file_close_it, _IO_file_doallocate, _IO_file_fopen,
862 _IO_file_init, _IO_file_jumps, _IO_fileno, _IO_file_open,
863 _IO_file_overflow, _IO_file_read, _IO_file_seek, _IO_file_seekoff,
864 _IO_file_setbuf, _IO_file_stat, _IO_file_sync, _IO_file_underflow,
865 _IO_file_write, _IO_file_xsputn, _IO_flockfile, _IO_flush_all,
866 _IO_flush_all_linebuffered, _IO_free_backup_area, _IO_free_wbackup_area,
867 _IO_init, _IO_init_marker, _IO_init_wmarker, _IO_iter_begin, _IO_iter_end,
868 _IO_iter_file, _IO_iter_next, _IO_least_wmarker, _IO_link_in,
869 _IO_list_all, _IO_list_lock, _IO_list_resetlock, _IO_list_unlock,
870 _IO_marker_delta, _IO_marker_difference, _IO_remove_marker, _IO_seekmark,
871 _IO_seekwmark, _IO_str_init_readonly, _IO_str_init_static,
872 _IO_str_overflow, _IO_str_pbackfail, _IO_str_seekoff, _IO_str_underflow,
873 _IO_switch_to_main_wget_area, _IO_switch_to_wget_mode,
874 _IO_unsave_wmarkers, _IO_wdefault_doallocate, _IO_wdefault_finish,
875 _IO_wdefault_pbackfail, _IO_wdefault_setbuf, _IO_wdefault_uflow,
876 _IO_wdefault_xsgetn, _IO_wdefault_xsputn, _IO_wdoallocbuf, _IO_wdo_write,
877 _IO_wfile_jumps, _IO_wfile_overflow, _IO_wfile_sync, _IO_wfile_underflow,
878 _IO_wfile_xsputn, _IO_wmarker_delta, or _IO_wsetb may stop working with a
879 future version of glibc. Unlike other symbol removals, these old
880 applications will not be supported using compatibility symbols.
881
882 * On GNU/Linux, the obsolete Linux constant PTRACE_SEIZE_DEVEL is no longer
883 defined by <sys/ptrace.h>.
884
885 * libm no longer supports SVID error handling (calling a user-provided
886 matherr function on error) or the _LIB_VERSION variable to control error
887 handling. (SVID error handling and the _LIB_VERSION variable still work
888 for binaries linked against older versions of the GNU C Library.) The
889 libieee.a library is no longer provided. math.h no longer defines struct
890 exception, or the macros X_TLOSS, DOMAIN, SING, OVERFLOW, UNDERFLOW,
891 TLOSS, PLOSS and HUGE.
892
893 * The libm functions pow10, pow10f and pow10l are no longer supported for
894 new programs. Programs should use the standard names exp10, exp10f and
895 exp10l for these functions instead.
896
897 * The mcontext_t type is no longer the same as struct sigcontext. On
898 platforms where it was previously the same, this changes the C++ name
899 mangling for interfaces involving this type.
900
901 * The add-ons mechanism for building additional packages at the same time as
902 glibc has been removed. The --enable-add-ons configure option is now
903 ignored.
904
905 * The --without-fp configure option is now ignored. Whether hardware
906 floating-point instructions are used is now configured based on whether
907 the compiler used at configure time (without any options implied by a
908 --with-cpu= configure option) uses such instructions.
909
910 * The res_hnok, res_dnok, res_mailok and res_ownok functions now check that
911 the specified string can be parsed as a domain name.
912
913 * In the malloc_info output, the <heap> element may contain another <aspace>
914 element, "subheaps", which contains the number of sub-heaps.
915
916 * The libresolv function p_secstodate is no longer supported for new
917 programs.
918
919 * The tilepro-*-linux-gnu configuration is no longer supported.
920
921 * The nonstandard header files <libio.h> and <_G_config.h> are deprecated
922 and will be removed in a future release. Software that is still using
923 either header should be updated to use standard <stdio.h> interfaces
924 instead.
925
926 libio.h was originally the header for a set of supported GNU extensions,
927 but they have not been maintained as such in many years, they are now
928 standing in the way of improvements to stdio, and we don't think there are
929 any remaining external users. _G_config.h was never intended for public
930 use, but predates the bits convention.
931
932 Changes to build and runtime requirements:
933
934 * bison version 2.7 or later is required to generate code in the 'intl'
935 subdirectory.
936
937 Security related changes:
938
939 CVE-2009-5064: The ldd script would sometimes run the program under
940 examination directly, without preventing code execution through the
941 dynamic linker. (The glibc project disputes that this is a security
942 vulnerability; only trusted binaries must be examined using the ldd
943 script.)
944
945 CVE-2017-15670: The glob function, when invoked with GLOB_TILDE,
946 suffered from a one-byte overflow during ~ operator processing (either
947 on the stack or the heap, depending on the length of the user name).
948 Reported by Tim Rühsen.
949
950 CVE-2017-15671: The glob function, when invoked with GLOB_TILDE,
951 would sometimes fail to free memory allocated during ~ operator
952 processing, leading to a memory leak and, potentially, to a denial
953 of service.
954
955 CVE-2017-15804: The glob function, when invoked with GLOB_TILDE and
956 without GLOB_NOESCAPE, could write past the end of a buffer while
957 unescaping user names. Reported by Tim Rühsen.
958
959 CVE-2017-17426: The malloc function, when called with an object size near
960 the value SIZE_MAX, would return a pointer to a buffer which is too small,
961 instead of NULL. This was a regression introduced with the new malloc
962 thread cache in glibc 2.26. Reported by Iain Buclaw.
963
964 CVE-2017-1000408: Incorrect array size computation in _dl_init_paths leads
965 to the allocation of too much memory. (This is not a security bug per se,
966 it is mentioned here only because of the CVE assignment.) Reported by
967 Qualys.
968
969 CVE-2017-1000409: Buffer overflow in _dl_init_paths due to miscomputation
970 of the number of search path components. (This is not a security
971 vulnerability per se because no trust boundary is crossed if the fix for
972 CVE-2017-1000366 has been applied, but it is mentioned here only because
973 of the CVE assignment.) Reported by Qualys.
974
975 CVE-2017-16997: Incorrect handling of RPATH or RUNPATH containing $ORIGIN
976 for AT_SECURE or SUID binaries could be used to load libraries from the
977 current directory.
978
979 CVE-2018-1000001: Buffer underflow in realpath function when getcwd function
980 succeeds without returning an absolute path due to unexpected behaviour
981 of the Linux kernel getcwd syscall. Reported by halfdog.
982
983 CVE-2018-6485: The posix_memalign and memalign functions, when called with
984 an object size near the value of SIZE_MAX, would return a pointer to a
985 buffer which is too small, instead of NULL. Reported by Jakub Wilk.
986
987 CVE-2018-6551: The malloc function, when called with an object size near
988 the value of SIZE_MAX, would return a pointer to a buffer which is too
989 small, instead of NULL.
990
991 The following bugs are resolved with this release:
992
993 [866] glob: glob should match dangling symlinks
994 [1062] glob: sysdeps/generic/glob.c merge from gnulib (part 3 of 3)
995 [2522] localedata: ca_ES@valencia: new Valencian (meridional Catalan)
996 locale
997 [5997] math: Very slow execution of sinf function
998 [10580] localedata: hr_HR: updated locale
999 [10871] locale: 'mon' array should contain both nominative and genitive
1000 cases
1001 [12349] localedata: eu_ES: incorrect thousands separator
1002 [13605] localedata: shn_MM: new Shan locale
1003 [13805] localedata: ru_RU: currency should use ',' as radix point
1004 [13953] localedata: km_KH: locale update
1005 [13994] localedata: mjw_IN: new locale
1006 [14121] build: make writes .mo files in po directory
1007 [14333] libc: Fix the race between atexit() and exit()
1008 [14681] dynamic-link: _dl_get_origin leaks memory via executable link map.
1009 [14925] localedata: bn_*: LC_IDENTIFICATION.language key should be
1010 "Bangla"
1011 [15260] localedata: LC_MESSAGES.{yes,no}{str,expr}: various errors
1012 [15261] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of full-
1013 width Latin characters
1014 [15332] localedata: es_CU: locale update
1015 [15436] stdio: Don't close or flush stdio streams on abort
1016 [15537] localedata: lv_LV: invalid collation for Latvian diacritical
1017 letters
1018 [16148] localedata: ca_ES: incorrect thousands separator
1019 [16750] dynamic-link: ldd should not try to execute the binaries
1020 (CVE-2009-5064)
1021 [16777] localedata: pl_PL: incorrect thousands separator in locale
1022 [16905] localedata: hanzi: new collation
1023 [17563] localedata: cmn_TW: add hanzi collation
1024 [17750] localedata: wrong collation order of diacritics in most locales
1025 [17804] libc: scandirat fails with ENOMEM because it checks for errno even
1026 if malloc succeeded
1027 [17956] build: Build fails on missing definitions from header file
1028 nss/nss.h when Mozilla NSS is used for cryptography
1029 [18203] libc: realpath() does not handle unreachable paths correctly
1030 [18572] dynamic-link: [arm] Lazy TLSDESC relocation has data race
1031 [18812] localedata: kab_DZ: new Kabyle Algeria locale
1032 [18822] libc: Internal functions are called via PLT
1033 [18858] string: _HAVE_STRING_ARCH_xxx aren't defined for i386 nor x86_64
1034 [19170] libc: __gmon_start__ defined in hppa in crtn.S
1035 [19574] libc: glibc should support building static PIE binaries
1036 [19852] localedata: charmaps/UTF-8: incorrect wcwidth for U+3099 and
1037 U+309A
1038 [19971] glob: glob: Do not skip entries with zero d_ino values
1039 [19982] localedata: fr.po: spelling mistake for error code EXDEV
1040 [20008] localedata: km_KH: convert to translit_neutral
1041 [20009] localedata: tr_TR: convert LC_CTYPE to i18n
1042 [20142] math: [x86_64] Add SSE4.1 trunc, truncf
1043 [20204] dynamic-link: _dl_open_hook and _dlfcn_hook hardening
1044 [20482] localedata: de_CH: abbreviated weekdays should be two letters
1045 [20498] localedata: miq_NI: new Mískitu / Miskito (miq) language locale
1046 for Nicaragua
1047 [20532] nss: getaddrinfo uses errno and h_errno without guaranteeing
1048 they're set, wrong errors returned by gaih_inet when lookup functions
1049 are not found.
1050 [20756] localedata: [PATCH] Use Unicode wise thousands separator
1051 [20826] network: posix/tst-getaddrinfo5 fails on hosts without network
1052 access
1053 [20952] localedata: yuw_PG: new locale
1054 [21084] localedata: charmaps/IBM858: new codepage
1055 [21161] manual: [PATCH] fix typo in manual/arith.texi on strtoul prototype
1056 [21242] libc: assert gives pedantic warning in old gcc versions
1057 [21265] dynamic-link: _dl_runtime_resolve isn't compatible with Intel C++
1058 __regcall calling convention
1059 [21309] math: signed integer overflow in sysdeps/ieee754/dbl-64/e_pow.c
1060 [21326] libc: C99 functions are not declared for C++11 and later if
1061 _GNU_SOURCE is not predefined by g++
1062 [21457] libc: sys/ucontext.h namespace
1063 [21530] libc: tmpfile() should be implemented using O_TMPFILE
1064 [21660] math: GCC fails to compile a formula with tgmath.h
1065 [21672] nptl: sys-libs/glibc on ia64 crashes on thread exit: signal
1066 SIGSEGV, Segmentation fault: pthread_create.c:432: __madvise
1067 (pd->stackblock, freesize - PTHREAD_STACK_MIN, MADV_DONTNEED);
1068 [21684] math: tgmath.h handling of complex integers
1069 [21685] math: tgmath.h handling of bit-fields
1070 [21686] math: tgmath.h handling of __int128
1071 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
1072 locale
1073 [21745] libc: [powerpc64le] Extra PLT reference with --enable-stack-
1074 protector=all
1075 [21750] localedata: column width of characters incompatible with classical
1076 wcwidth
1077 [21754] malloc: malloc: Perform as little work as possible after heap
1078 consistency check failures
1079 [21780] libc: hppa: p{read,write}v2 does not set ENOSUP on invalid flag
1080 [21790] libc: Missing __memset_zero_constant_len_parameter in libc.so
1081 [21791] string: Unused XXX_chk_XXX functions in libc.a
1082 [21815] dynamic-link: FAIL: elf/tst-prelink-cmp with GCC is defaulted to
1083 PIE
1084 [21836] localedata: Removed redundant data (LC_MONETARY) in various Indian
1085 locales
1086 [21845] localedata: Added new Locale bho_NP
1087 [21853] localedata: Fix abday Which looks same as day in zh_SG
1088 [21854] localedata: Added New Locale en_SC
1089 [21864] libc: xmalloc.o is compiled with -DMODULE_NAME=libc
1090 [21871] dynamic-link: _dl_runtime_resolve_avx_opt is slower than
1091 _dl_runtime_resolve_avx_slow
1092 [21885] network: getaddrinfo: gethosts does not release resolver context
1093 on memory allocation failure
1094 [21899] libc: XPG4.2 sigaction namespace
1095 [21908] dynamic-link: dynamic linker broke on ia64 (mmap2 consolidation is
1096 the suspect)
1097 [21913] libc: static binaries SIGSEGV in __brk when host's gcc is pie-by-
1098 default (i386)
1099 [21915] nss: nss_files can return with NSS_STATUS_SUCCESS and a clobbered
1100 errno value, causing getaddrinfo to fail
1101 [21920] localedata: Fix p_cs_precedes/n_cs_precedes for mt_MT
1102 [21922] network: getaddrinfo with AF_INET/AF_INET6 returns EAI_NONAME
1103 instead of EAI_NODATA
1104 [21928] libc: sys/ptrace.h: remove obsolete temporary development Linux
1105 constant PTRACE_SEIZE_DEVEL
1106 [21930] math: C-only gcc builtins used in <math.h> isinf
1107 [21932] network: Unpaired __resolv_context_get in generic get*_r
1108 implementation
1109 [21941] math: powerpc: Wrong register constraint for xssqrtqp in sqrtf128
1110 [21944] libc: sigval namespace
1111 [21951] localedata: Update hanzi collation by stroke
1112 [21955] math: Wrong alignment of L(SP_RANGE)/L(SP_INF_0) in
1113 sysdeps/x86_64/fpu/e_expf.S
1114 [21956] libc: Stack allocation in MIPS syscall impl (ubounded stack
1115 allocation in syscall loops)
1116 [21959] localedata: Fix Country name for xh_ZA
1117 [21960] localedata: Fix abmon for bem_ZM
1118 [21966] math: AVX2 mathvec functions use FMA without checking
1119 [21967] math: When 512-bit AVX2 wrapper functions in mathvec are used?
1120 [21971] localedata: Added New Locale for mfe_MU
1121 [21972] libc: assert macro requires operator== (int) for its argument type
1122 [21973] math: [sparc] libm missing sqrtl compat symbol
1123 [21974] libc: Remove __bb_init_func and __bb_exit_func
1124 [21982] string: stratcliff.c: error: assuming signed overflow does not
1125 occur with -O3
1126 [21986] stdio: __guess_grouping is called incorrectly
1127 [21987] math: [sparc32] wrong bits/long-double.h installed
1128 [22019] localedata: Wrong placement of monetary symbol in el_GR (negative
1129 amounts)
1130 [22022] localedata: Missing country_name for mni_IN
1131 [22023] localedata: Removed redundant data (LC_TIME and LC_MESSAGES) for
1132 niu_NZ
1133 [22025] locale: iconv: Inconsistency between pointer mangling and NULL
1134 checks
1135 [22026] locale: iconv_open: heap overflow on gconv_init failure
1136 [22028] math: bits/math-finite.h _MSUF_ expansion namespace
1137 [22035] math: [m68k] bits/math-inline.h macro namespace
1138 [22038] localedata: Fix abbreviated weeks and months for Somali
1139 [22044] localedata: Remove redundant data for Limburgish Language
1140 [22050] malloc: Linking with -lmcheck does not hook
1141 __malloc_initialize_hook correctly
1142 [22051] libc: zero terminator in the middle of glibc's .eh_frame
1143 [22052] malloc: malloc failed to compile with GCC 7 and -O3
1144 [22070] localedata: charmaps/UTF-8: wcwidth for
1145 Prepended_Concatenation_Mark codepoints set to 0 (should be 1)
1146 [22074] localedata: charmaps/UTF-8: wcwidth for U+1160-U+11FF (Hangul
1147 Jungseong and Jongseong) should be 0
1148 [22078] nss: nss_files performance issue in multi mode
1149 [22082] math: bits/math-finite.h exp10 condition
1150 [22086] libc: pcprofiledump incorrect cross-endian condition
1151 [22093] dynamic-link: ld.so no longer searches in .../x86_64
1152 [22095] network: Name server address allocation memory leak in resolv.conf
1153 parsing after OOM
1154 [22096] network: __resolv_conf_attach can incorrectly free passed conf
1155 object
1156 [22100] localedata: om_KE: LC_TIME: copy redundant data from om_ET
1157 [22101] dynamic-link: Dynamic loader must ignore "debug" shared objects
1158 e.g. ET_GNU_DEBUG_*
1159 [22111] malloc: malloc: per thread cache is not returned when thread exits
1160 [22112] localedata: Fix LC_TELEPHONE/LC_NAME for az_AZ
1161 [22134] libc: [linux] implement fexecve with execveat
1162 [22142] libc: [powerpc] printf oupts a wrong value of DBL_MAX on ppc64 and
1163 ppc64le
1164 [22145] libc: ttyname() gives up too early in the face of namespaces
1165 [22146] math: C++ build issue with float128 on x86_64
1166 [22153] nptl: nptl: save error code before process termination
1167 [22156] libc: [hppa,ia64,microblaze] Executable stack default
1168 [22159] malloc: malloc: MALLOC_CHECK_ broken with --enable-tunables=no
1169 [22161] nscd: nscd cache prune for netgroups hangs after timeout bump
1170 [22165] libc: [hppa] Text relocations in libc.so
1171 [22180] libc: destructor registered via __cxa_atexit is called twice
1172 [22183] glob: commit 5554304f0ddd ("posix: Allow glob to match dangling
1173 symlinks") cause "make" segfaults
1174 [22189] math: [powerpc] math_private.h definitions of math_opt_barrier and
1175 math_force_eval
1176 [22207] libc: FAIL: stdlib/test-atexit-race
1177 [22225] math: nearbyint arithmetic moved before feholdexcept
1178 [22229] math: [sparc32] missing copysignl, fabsl, fmal compat symbols
1179 [22235] math: iscanonical in C++ and float128
1180 [22243] math: log2(0) and log10(0) are wrong in downward rounding without
1181 the svid compat wrapper
1182 [22244] math: ynf and yn are wrong without the svid compat wrapper
1183 [22273] libc: Improper assert in Linux posix_spawn implementation
1184 [22284] libc: -pg -pie doesn't work
1185 [22292] locale: localedef exits with error 4 when it should be error 1
1186 [22294] locale: Allow "" for int_currency_symbol definition in locales.
1187 [22295] locale: Don't warn on non-symbolic characters in locale sources in
1188 --verbose.
1189 [22296] math: glibc 2.26: signbit build issue with Gcc 5.5.0 on x86_64
1190 [22298] nptl: x32: lockups on recursive pthread_mutex_lock after upgrade
1191 to 2.26
1192 [22299] dynamic-link: Problem with $PLATFORM on x86_64 platform
1193 [22320] glob: Buffer overflow in glob with GLOB_TILDE (CVE-2017-15670)
1194 [22321] libc: sysconf(_SC_IOV_MAX) returns -1 on Linux
1195 [22322] libc: [mips64] wrong bits/long-double.h installed
1196 [22325] glob: Memory leak in glob with GLOB_TILDE (CVE-2017-15671)
1197 [22332] glob: Buffer overflow in glob with GLOB_TILDE in unescaping
1198 (CVE-2017-15804)
1199 [22336] localedata: cs_CZ LC_COLLATE does not use i18n
1200 [22343] malloc: Integer overflow in posix_memalign
1201 [22347] libc: getrandom() returns the number of bytes that were copied to
1202 the buffer even though the comments say "Return 0 on success and -1 on
1203 failure."
1204 [22353] string: sysdeps/i386/i586/strcpy.S isn't maintainable
1205 [22362] libc: Installed crt1.o, crti,.o and crtn.o files are used with
1206 -m32
1207 [22370] dynamic-link: Incorrect note padding check
1208 [22375] libc: malloc returns pointer from tcache_get when should return
1209 NULL (CVE-2017-17426)
1210 [22377] math: iseqsig, float128 and C++
1211 [22382] localedata: Error in tpi_PG locale
1212 [22387] localedata: Replace unicode sequences <Uxxxx> for characters
1213 inside the ASCII printable range
1214 [22402] math: [powerpc64le] __MATH_TG does not support _Float128 for
1215 -mlong-double-64
1216 [22403] localedata: Slash needs escaping in some locales
1217 [22408] malloc: malloc_info access heaps without arena lock, ignores heaps
1218 [22409] network: res_hnok does not accept some host names used on the
1219 Internet
1220 [22412] network: res_dnok, res_hnok should perform syntax checks
1221 [22413] network: ns_name_pton ignores syntactically invalid trailing
1222 backslash
1223 [22415] stdio: setvbuf can lead to invalid free/segfault
1224 [22432] build: Non-deterministic build
1225 [22439] malloc: malloc_info should compute summary statistics for all sub-
1226 heaps in an arena
1227 [22442] network: if_nametoindex could report index for the wrong
1228 networking interface
1229 [22446] build: aliasing violation calling readlink in handle_request
1230 [22447] build: unsafe call to strlen with a non-string in getlogin_r.c
1231 [22457] libc: Generic preadv/pwritev incorrectly calls __posix_memalign
1232 [22459] libc: FAIL: elf/check-localplt with __stack_chk_fail related to
1233 __nscd_hash/__nss_hash
1234 [22463] network: p_secstodate overflow handling
1235 [22469] localedata: pl_PL LC_COLLATE does not use i18n
1236 [22478] libc: sigwait can fail with EINTR
1237 [22505] libc: ldconfig processes include directive in locale-specific
1238 order
1239 [22515] localedata: hsb_DE LC_COLLATE does not use copy "iso14651_t1"
1240 [22517] localedata: et_EE LC_COLLATE does not use copy "iso14651_t1"
1241 [22519] localedata: is_IS LC_COLLATE does not use copy "iso14651_t1"
1242 [22524] localedata: lt_LT LC_COLLATE does not use copy "iso14651_t1"
1243 [22527] localedata: tr_TR LC_COLLATE does not use copy "iso14651_t1"
1244 [22534] localedata: Collation rules for Serbian and Bosnian should be the
1245 same as for Croatian
1246 [22561] math: [DR#471] cacosh (0 + iNaN) should return NaN +/- i pi/2
1247 [22568] math: [DR#471] ctanh (0 + iNaN), ctanh (0 + i Inf)
1248 [22577] libc: missing newline after "cannot allocate TLS data structures
1249 for initial thread"
1250 [22588] manual: manual/conf.texi: missing underscore in front of
1251 SC_SSIZE_MAX
1252 [22593] math: nextafter and nexttoward are declared with const attribute
1253 [22596] manual: manual: finite(nan) wrongly described as returning nonzero
1254 [22603] string: ia64 memchr overflows internal pointer check
1255 [22605] libc: SH clone does not set the exit code correctly
1256 [22606] dynamic-link: Incorrect array size computation in _dl_init_paths
1257 (CVE-2017-1000408)
1258 [22607] dynamic-link: Buffer Overflow in _dl_init_paths (CVE-2017-1000409)
1259 [22611] malloc: malloc/tst-realloc wrongly assumes that errno must not be
1260 modified in case of success
1261 [22614] build: gcc: error: unrecognized command line option ‘-no-pie’
1262 [22615] manual: manual: ambiguous wording about errno value in case of
1263 success
1264 [22624] libc: MIPS setjmp() saves incorrect 'o0' register in --enable-
1265 stack-protector=all
1266 [22625] dynamic-link: RPATH $ORIGIN replaced by PWD for AT_SECURE/SUID
1267 binaries or if /proc is not mounted (CVE-2017-16997)
1268 [22627] dynamic-link: $ORIGIN in $LD_LIBRARY_PATH is substituted twice
1269 [22630] build: $(no-pie-ldflag) is no longer effective
1270 [22631] math: [m68k] Bad const attributes in bits/mathinline.h
1271 [22635] nptl: pthread_self returns NULL before libpthread is loaded
1272 [22636] nptl: PTHREAD_STACK_MIN is too small on x86-64
1273 [22637] nptl: guard size is subtracted from thread stack size instead of
1274 adding it on top
1275 [22648] libc: getrlimit/setrlimit with RLIM_INFINITY broken on alpha
1276 [22657] localedata: hu_HU: Avoid double space in date
1277 [22660] math: fmax, fmin sNaN handling on alpha
1278 [22664] libc: New warning of GCC8
1279 [22665] math: alpha: ceil and floor raise inexact exceptions
1280 [22666] math: alpha: trunc raise inexact exceptions
1281 [22667] libc: makecontext lacks stack alignment on i386
1282 [22678] libc: prlimit fails for RLIM_INFINITY values on 32-bit machines
1283 [22679] libc: getcwd(3) can succeed without returning an absolute path
1284 (CVE-2018-1000001)
1285 [22685] libc: PowerPC: Static AT_SECURE binaries segfault with lock-
1286 elision and tunables
1287 [22687] math: [powerpc-nofpu] complex long double functions spurious
1288 "invalid" exception
1289 [22688] math: [powerpc-nofpu] remainderl wrong sign of zero result
1290 [22690] math: [ldbl-128ibm] lrintl, lroundl missing "invalid" exceptions
1291 [22691] math: [powerpc-nofpu] fmaxmagl, fminmagl spurious "invalid"
1292 exception
1293 [22693] math: [ldbl-128ibm] log1pl (-qNaN) spurious "invalid" exception
1294 [22697] math: [powerpc] llround spurious "inexact" exceptions on 32-bit
1295 power4
1296 [22701] nis: Incomplete removal of libnsl
1297 [22702] math: [powerpc-nofpu] nearbyintl traps with trapping "inexact"
1298 [22707] libc: Missing defines in elf.h for DF_1_STUB and DF_1_PIE.
1299 [22715] dynamic-link: FAIL: elf/tst-audit10
1300 [22719] libc: Backtrace tests fail on hppa
1301 [22742] libc: [aarch64] mcontext_t __reserved field got renamed
1302 [22743] nptl: __pthread_register_cancel corrupts stack after f81ddabffd
1303 [22765] crypt: (struct crypt_data *data)->initialized is not set to zero
1304 before the first call to crypt_r () in crypt/badsalttest.c
1305
1306 \f
1307 Version 2.26
1308
1309 Major new features:
1310
1311 * A per-thread cache has been added to malloc. Access to the cache requires
1312 no locks and therefore significantly accelerates the fast path to allocate
1313 and free small amounts of memory. Refilling an empty cache requires locking
1314 the underlying arena. Performance measurements show significant gains in a
1315 wide variety of user workloads. Workloads were captured using a special
1316 instrumented malloc and analyzed with a malloc simulator. Contributed by
1317 DJ Delorie with the help of Florian Weimer, and Carlos O'Donell.
1318
1319 * Unicode 10.0.0 Support: Character encoding, character type info, and
1320 transliteration tables are all updated to Unicode 10.0.0, using
1321 generator scripts contributed by Mike FABIAN (Red Hat).
1322 These updates cause user visible changes, especially the changes in
1323 wcwidth for many emoji characters cause problems when emoji sequences
1324 are rendered with pango, see for example:
1325 https://bugzilla.gnome.org/show_bug.cgi?id=780669#c5
1326
1327 * Collation of Hungarian has been overhauled and is now consistent with "The
1328 Rules of Hungarian Orthography, 12th edition" (Bug 18934). Contributed by
1329 Egmont Koblinger.
1330
1331 * Improvements to the DNS stub resolver, contributed by Florian Weimer:
1332
1333 - The GNU C Library will now detect when /etc/resolv.conf has been
1334 modified and reload the changed configuration. The new resolver option
1335 “no-reload” (RES_NORELOAD) disables this behavior.
1336
1337 - The GNU C Library now supports an arbitrary number of search domains
1338 (configured using the “search” directive in /etc/resolv.conf);
1339 previously, there was a hard limit of six domains. For backward
1340 compatibility, applications that directly modify the ‘_res’ global
1341 object are still limited to six search domains.
1342
1343 - When the “rotate” (RES_ROTATE) resolver option is active, the GNU C
1344 Library will now randomly pick a name server from the configuration as a
1345 starting point. (Previously, the second name server was always used.)
1346
1347 * The tunables feature is now enabled by default. This allows users to tweak
1348 behavior of the GNU C Library using the GLIBC_TUNABLES environment variable.
1349
1350 * New function reallocarray, which resizes an allocated block (like realloc)
1351 to the product of two sizes, with a guaranteed clean failure upon integer
1352 overflow in the multiplication. Originally from OpenBSD, contributed by
1353 Dennis Wölfing and Rüdiger Sonderfeld.
1354
1355 * New wrappers for the Linux-specific system calls preadv2 and pwritev2.
1356 These are extended versions of preadv and pwritev, respectively, taking an
1357 additional flags argument. The set of supported flags depends on the
1358 running kernel; full support currently requires kernel 4.7 or later.
1359
1360 * posix_spawnattr_setflags now supports the flag POSIX_SPAWN_SETSID, to
1361 create a new session ID for the spawned process. This feature is
1362 scheduled to be added to the next major revision of POSIX; for the time
1363 being, it is available under _GNU_SOURCE.
1364
1365 * errno.h is now safe to use from C-preprocessed assembly language on all
1366 supported operating systems. In this context, it will only define the
1367 Exxxx constants, as preprocessor macros expanding to integer literals.
1368
1369 * On ia64, powerpc64le, x86-32, and x86-64, the math library now implements
1370 128-bit floating point as defined by ISO/IEC/IEEE 60559:2011 (IEEE
1371 754-2008) and ISO/IEC TS 18661-3:2015. Contributed by Paul E. Murphy,
1372 Gabriel F. T. Gomes, Tulio Magno Quites Machado Filho, and Joseph Myers.
1373
1374 To compile programs that use this feature, the compiler must support
1375 128-bit floating point with the type name _Float128 (as defined by TS
1376 18661-3) or __float128 (the nonstandard name used by GCC for C++, and for
1377 C prior to version 7). _GNU_SOURCE or __STDC_WANT_IEC_60559_TYPES_EXT__
1378 must be defined to make the new interfaces visible.
1379
1380 The new functions and macros correspond to those present for other
1381 floating-point types (except for a few obsolescent interfaces not
1382 supported for the new type), with F128 or f128 suffixes; for example,
1383 strtof128, HUGE_VAL_F128 and cosf128. Following TS 18661-3, there are no
1384 printf or scanf formats for the new type; the strfromf128 and strtof128
1385 interfaces should be used instead.
1386
1387 Deprecated and removed features, and other changes affecting compatibility:
1388
1389 * The synchronization that pthread_spin_unlock performs has been changed to
1390 now be equivalent to a C11 atomic store with release memory order to the
1391 spin lock's memory location. Previously, several (but not all)
1392 architectures used stronger synchronization (e.g., containing what is
1393 often called a full barrier). This change can improve performance, but
1394 may affect odd fringe uses of spin locks that depend on the previous
1395 behavior (e.g., using spin locks as atomic variables to try to implement
1396 Dekker's mutual exclusion algorithm).
1397
1398 * The port to Native Client running on ARMv7-A (--host=arm-nacl) has been
1399 removed.
1400
1401 * Sun RPC is deprecated. The rpcgen program, librpcsvc, and Sun RPC headers
1402 will only be built and installed when the GNU C Library is configured with
1403 --enable-obsolete-rpc. This allows alternative RPC implementations, such
1404 as TIRPC or rpcsvc-proto, to be used.
1405
1406 * The NIS(+) name service modules, libnss_nis, libnss_nisplus, and
1407 libnss_compat, are deprecated, and will not be built or installed by
1408 default.
1409
1410 The NIS(+) support library, libnsl, is also deprecated. By default, a
1411 compatibility shared library will be built and installed, but not headers
1412 or development libraries. Only a few NIS-related programs require this
1413 library. (In particular, the GNU C Library has never required programs
1414 that use 'gethostbyname' to be linked with libnsl.)
1415
1416 Replacement implementations based on TIRPC, which additionally support
1417 IPv6, are available from <https://github.com/thkukuk/>. The configure
1418 option --enable-obsolete-nsl will cause libnsl's headers, and the NIS(+)
1419 name service modules, to be built and installed.
1420
1421 * The DNS stub resolver no longer performs EDNS fallback. If EDNS or DNSSEC
1422 support is enabled, the configured recursive resolver must support EDNS.
1423 (Responding to EDNS-enabled queries with responses which are not
1424 EDNS-enabled is fine, but FORMERR responses are not.)
1425
1426 * res_mkquery and res_nmkquery no longer support the IQUERY opcode. DNS
1427 servers have not supported this opcode for a long time.
1428
1429 * The _res_opcodes variable has been removed from libresolv. It had been
1430 exported by accident.
1431
1432 * <string.h> no longer includes inline versions of any string functions,
1433 as this kind of optimization is better done by the compiler. The macros
1434 __USE_STRING_INLINES and __NO_STRING_INLINES no longer have any effect.
1435
1436 * The nonstandard header <xlocale.h> has been removed. Most programs should
1437 use <locale.h> instead. If you have a specific need for the definition of
1438 locale_t with no other declarations, please contact
1439 libc-alpha@sourceware.org and explain.
1440
1441 * The obsolete header <sys/ultrasound.h> has been removed.
1442
1443 * The obsolete signal constant SIGUNUSED is no longer defined by <signal.h>.
1444
1445 * The obsolete function cfree has been removed. Applications should use
1446 free instead.
1447
1448 * The stack_t type no longer has the name struct sigaltstack. This changes
1449 the C++ name mangling for interfaces involving this type.
1450
1451 * The ucontext_t type no longer has the name struct ucontext. This changes
1452 the C++ name mangling for interfaces involving this type.
1453
1454 * On M68k GNU/Linux and MIPS GNU/Linux, the fpregset_t type no longer has
1455 the name struct fpregset. On Nios II GNU/Linux, the mcontext_t type no
1456 longer has the name struct mcontext. On SPARC GNU/Linux, the struct
1457 mc_fq, struct rwindow, struct fpq and struct fq types are no longer
1458 defined in sys/ucontext.h, the mc_fpu_t type no longer has the name struct
1459 mc_fpu, the gwindows_t type no longer has the name struct gwindows and the
1460 fpregset_t type no longer has the name struct fpu. This changes the C++
1461 name mangling for interfaces involving those types.
1462
1463 * On S/390 GNU/Linux, the constants defined by <sys/ptrace.h> have been
1464 synced with the kernel:
1465
1466 - PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS and PTRACE_SETFPREGS
1467 are not supported on this architecture and have been removed.
1468
1469 - PTRACE_SINGLEBLOCK, PTRACE_SECCOMP_GET_FILTER, PTRACE_PEEKUSR_AREA,
1470 PTRACE_POKEUSR_AREA, PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE,
1471 PTRACE_DISABLE_TE and PTRACE_TE_ABORT_RAND have been added.
1472
1473 Programs that assume the GET/SETREGS ptrace requests are universally
1474 available will now fail to build, instead of malfunctioning at runtime.
1475
1476 Changes to build and runtime requirements:
1477
1478 * Linux kernel 3.2 or later is required at runtime, on all architectures
1479 supported by that kernel. (This is a change from version 2.25 only for
1480 x86-32 and x86-64.)
1481
1482 * GNU Binutils 2.25 or later is now required to build the GNU C Library.
1483
1484 * On most architectures, GCC 4.9 or later is required to build the GNU C
1485 Library. On powerpc64le, GCC 6.2 or later is required.
1486
1487 Older GCC versions and non-GNU compilers are still supported when
1488 compiling programs that use the GNU C Library. (We do not know exactly
1489 how old, and some GNU extensions to C may be _de facto_ required. If you
1490 are interested in helping us make this statement less vague, please
1491 contact libc-alpha@sourceware.org.)
1492
1493 Security related changes:
1494
1495 * The DNS stub resolver limits the advertised UDP buffer size to 1200 bytes,
1496 to avoid fragmentation-based spoofing attacks (CVE-2017-12132).
1497
1498 * LD_LIBRARY_PATH is now ignored in binaries running in privileged AT_SECURE
1499 mode to guard against local privilege escalation attacks (CVE-2017-1000366).
1500
1501 * Avoid printing a backtrace from the __stack_chk_fail function since it is
1502 called on a corrupt stack and a backtrace is unreliable on a corrupt stack
1503 (CVE-2010-3192).
1504
1505 * A use-after-free vulnerability in clntudp_call in the Sun RPC system has been
1506 fixed (CVE-2017-12133).
1507
1508 The following bugs are resolved with this release:
1509
1510 [984] network: Respond to changed resolv.conf in gethostbyname
1511 [5010] network: sunrpc service cleanup causes unwanted port mapper traffic
1512 [12068] localedata: sc_IT: misspelled yesexpr/day/abday/mon/abmon/date_fmt
1513 fields
1514 [12189] libc: __stack_chk_fail should not attempt a backtrace
1515 (CVE-2010-3192)
1516 [14096] time: Race condition on timezone/tst-timezone.out
1517 [14172] localedata: az_IR: new locale
1518 [14995] build: glibc fails to build if gold is the default linker, even if
1519 ld.bfd is available
1520 [15998] build: [powerpc] Set arch_minimum_kernel for powerpc LE
1521 [16637] network: inet_pton function is accepting IPv6 with bad format
1522 [16640] string: string/strtok.c: undefined behaviour inconsistent between
1523 x86 and other generic code
1524 [16875] localedata: ko_KR: fix lang_name
1525 [17225] localedata: ar_SY: localized month names for May and June are
1526 incorrect
1527 [17297] localedata: da_DK: wrong date_fmt string
1528 [18907] stdio: Incorrect order of __wur __THROW in <printf.h>
1529 [18934] localedata: hu_HU: collate: fix multiple bugs and add tests
1530 [18988] nptl: pthread wastes memory with mlockall(MCL_FUTURE)
1531 [19066] localedata: ar_SA abbreviated day and month names are in English
1532 [19569] network: resolv: Support an arbitrary number of search domains
1533 [19570] network: Implement random DNS server selection in the stub
1534 resolver
1535 [19838] locale: localedef fails on PA-RISC
1536 [19919] localedata: iso14651_t1_common: Correct the Malayalam sorting
1537 order of 0D36 and 0D37
1538 [19922] localedata: iso14651_t1_common: Define collation for Malayalam
1539 chillu characters
1540 [20098] libc: FAIL: debug/backtrace-tst on hppa
1541 [20257] network: sunrpc: clntudp_call does not enforce timeout when
1542 receiving data
1543 [20275] localedata: locale day/abday/mon/abmon should not have trailing
1544 whitespace
1545 [20313] localedata: Update locale data to Unicode 9.0
1546 [20424] manual: Document how to provide a malloc replacement
1547 [20496] localedata: agr_PE: new language locale Awajún / Aguaruna (agr)
1548 for Peru
1549 [20686] locale: Add el_GR@euro to SUPPORTED.
1550 [20831] dynamic-link: _dl_map_segments does not test for __mprotect
1551 failures consistently
1552 [21015] dynamic-link: Document and fix --enable-bind-now
1553 [21016] nptl: pthread_cond support is broken on hppa
1554 [21029] libc: glibc-2.23 (and later) fails to compile with -fno-omit-
1555 frame-pointer on i386
1556 [21049] libc: segfault in longjmp_chk() due to clobbered processor
1557 register
1558 [21075] libc: unused assigment to %g4 in sparc/sparc{64,32}/clone.S
1559 [21088] libc: Build fails with --enable-static-nss
1560 [21094] math: cosf(1.57079697) has 3 ulp error on targets where the
1561 generic c code is used
1562 [21109] libc: Tunables broken on big-endian
1563 [21112] math: powf has large ulp errors with base close to 1 and exponent
1564 around 4000
1565 [21115] network: sunrpc: Use-after-free in error path in clntudp_call
1566 (CVE-2017-12133)
1567 [21120] malloc: glibc malloc is incompatible with GCC 7
1568 [21130] math: Incorrect return from y0l (-inf) and y1l (-inf) when linking
1569 with -lieee
1570 [21134] math: Exception (divide by zero) not set for y0/y1 (0.0) and y0/y1
1571 (-0.0) when linking with -lieee
1572 [21171] math: log10, log2 and lgamma return incorrect results
1573 [21179] libc: handle R_SPARC_DISP64 and R_SPARC_REGISTER relocs
1574 [21182] libc: __memchr_sse2: regression in glibc-2.25 on i686
1575 [21207] localedata: ce_RU: update weekdays from CLDR
1576 [21209] dynamic-link: LD_HWCAP_MASK read in setuid binaries
1577 [21217] localedata: Update months from CLDR-31
1578 [21232] libc: miss posix_fadvise64 on MIPS64 when static linking
1579 [21243] libc: support_delete_temp_file should issue warning for failed
1580 remove()
1581 [21244] libc: support resolv_test_start() socket fd close should be
1582 checked for errors.
1583 [21253] libc: localedef randomly segfaults when using -fstack-check due to
1584 new posix_spawn implementation
1585 [21258] dynamic-link: Branch predication in _dl_runtime_resolve_avx512_opt
1586 leads to lower CPU frequency
1587 [21259] libc: [alpha] termios.h missing IXANY for POSIX
1588 [21261] libc: [sparc64] bits/setjmp.h namespace
1589 [21267] network: [mips] bits/socket.h IOC* namespace
1590 [21268] libc: [alpha] termios.h NL2, NL3 namespace
1591 [21270] libc: mmap64 silently truncates large offset values
1592 [21275] libc: posix_spawn always crashes on ia64 now
1593 [21277] libc: [alpha] termios.h missing IUCLC for UNIX98 and older
1594 [21280] math: [powerpc] logbl for POWER7 return incorrect results
1595 [21289] libc: Incorrect declaration for 32-bit platforms with
1596 _FILE_OFFSET_BITS=64 causes build error
1597 [21295] network: GETAI(AF_UNSPEC) drops IPv6 addresses if nss module does
1598 not support gethostbyname4_r
1599 [21298] nptl: rwlock can deadlock on frequent reader/writer phase
1600 switching
1601 [21338] malloc: mallopt M_ARENA_MAX doesn't set the maximum number of
1602 arenas
1603 [21340] libc: Support POSIX_SPAWN_SETSID
1604 [21357] libc: unwind-dw2-fde deadlock when using AddressSanitizer
1605 [21359] network: ns_name_pack needs additional byte in destination buffer
1606 [21361] network: resolv: Reduce advertised EDNS0 buffer size to guard
1607 against fragmentation attacks (CVE-2017-12132)
1608 [21369] network: resolv: Remove EDNS fallback
1609 [21371] libc: Missing timespec definition when compiled with _XOPEN_SOURCE
1610 and _POSIX_C_SOURCE
1611 [21386] nptl: Assertion in fork for distinct parent PID is incorrect
1612 [21391] dynamic-link: x86: Set dl_platform and dl_hwcap from CPU features
1613 [21393] stdio: Missing dup3 error check in freopen, freopen64
1614 [21396] libc: Use AVX2 memcpy/memset on Skylake server
1615 [21399] localedata: Bad description for U00EC in
1616 localedata/charmaps/CP1254
1617 [21411] malloc: realloc documentation error
1618 [21426] network: sys/socket.h uio.h namespace
1619 [21428] libc: [aarch64] tst-backtrace5 testsuite failure
1620 [21445] libc: signal.h bsd_signal namespace
1621 [21455] network: Network headers stdint.h namespace
1622 [21474] network: resolv: res_init does not use RES_DFLRETRY (2) but 4 for
1623 retry value
1624 [21475] network: resolv: Overlong search path is truncated mid-label
1625 [21511] libc: sigstack namespace
1626 [21512] libc: clone() ends up calling exit_group() through _exit() wrapper
1627 [21514] libc: sysdeps/unix/sysv/linux/sys/syscall.h:31:27: fatal error:
1628 bits/syscall.h: No such file or directory
1629 [21517] libc: struct sigaltstack namespace
1630 [21528] dynamic-link: Duplicated minimal strtoul implementations in ld.so
1631 [21533] localedata: Update locale data to Unicode 10.0
1632 [21537] libc:
1633 ../sysdeps/unix/sysv/linux/s390/s390-32/__makecontext_ret.S:44: Error:
1634 junk at end of line, first unrecognized character is `@'
1635 [21538] libc: SIG_HOLD missing for XPG4
1636 [21539] libc: S390: Mismatch between kernel and glibc ptrace.h with
1637 request 12: PTRACE_SINGLEBLOCK vs PTRACE_GETREGS.
1638 [21542] libc: Use conservative default for sysconf (_SC_NPROCESSORS_ONLN)
1639 [21543] libc: sigevent namespace
1640 [21548] libc: [mips] get/set/make/swap context for MIPS O32 assume wrong
1641 size for general purpose registers in mcontext_t structure
1642 [21550] libc: sigwait namespace
1643 [21552] libc: XPG4 bsd_signal namespace
1644 [21554] libc: sigpause namespace
1645 [21560] libc: sys/wait.h signal.h namespace
1646 [21561] libc: waitid namespace
1647 [21573] nptl: GCC 7: /usr/bin/install: cannot remove
1648 '/usr/include/stdlib.h': Permission denied
1649 [21575] libc: sys/wait.h missing struct rusage definition
1650 [21584] libc: sigaltstack etc namespace
1651 [21597] libc: siginterrupt namespace
1652 [21607] math: hppa: FAIL: math/test-tgmath
1653 [21609] dynamic-link: Incomplete workaround for GCC __tls_get_addr ABI
1654 issue on x86-64
1655 [21622] libc: [tile] missing SA_* for POSIX.1:2008
1656 [21624] dynamic-link: ld.so: Unsafe alloca allows local attackers to alias
1657 stack and heap (CVE-2017-1000366)
1658 [21625] libc: wait3 namespace
1659 [21654] nss: Incorrect pointer alignment in NSS group merge result
1660 construction
1661 [21657] network: Parse interface zone id for node-local multicast
1662 [21662] string: memcmp-avx2-movbe.S lacks saturating subtraction for
1663 between_2_3
1664 [21666] libc: .symver is used on common symbol
1665 [21668] network: resolv: res_init cross-thread broadcast introduces race
1666 conditions
1667 [21687] math: tgmath.h totalorder, totalordermag return type
1668 [21694] locale: Current Glibc Locale Does Not Support Tok-Pisin and Fiji
1669 Hindi Locale
1670 [21696] libc: Incorrect assumption of of __cpu_mask in
1671 posix/sched_cpucount.c
1672 [21697] libc: sysdeps/posix/spawni.c: 2 * suspicious condition ?
1673 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
1674 locale
1675 [21707] math: ppc64le: Invalid IFUNC resolver from libgcc calls getauxval,
1676 leading to relocation crash
1677 [21709] libc: resolv_conf.c:552: update_from_conf: Assertion
1678 `resolv_conf_matches (resp, conf)' failed.
1679 [21710] localedata: Added Samoan language locale for Samoa
1680 [21711] localedata: Pashto yesstr/nostr locale are missing
1681 [21715] nptl: sysdeps/nptl/bits/pthreadtypes.h: typedef guard
1682 __have_pthread_attr_t can cause redefinition of typedef ‘pthread_attr_t’
1683 [21721] localedata: Incorrect Full Weekday names for ks_IN@devanagari
1684 [21723] localedata: yesstr/nostr missing for Chinese language locale
1685 [21724] localedata: yesstr and nostr are missing for Xhosa [LC_MESSAGES]
1686 locale
1687 [21727] localedata: yesstr and nostr are missing for Tsonga [LC_MESSAGES]
1688 locale
1689 [21728] localedata: New Locale for Tongan language
1690 [21729] localedata: incorrect LC_NAME fields for hi_IN
1691 [21733] localedata: yesstr and nostr are missing for zh_HK
1692 [21734] localedata: Missing yesstr and nostr are for kw_GB
1693 [21738] libc: misc/tst-preadvwritev2 and misc/tst-preadvwritev64v2 fail
1694 [21741] libc: Undefined __memmove_chk_XXX and __memset_chk_XXX in libc.a
1695 [21742] libc: _dl_num_cache_relocations is undefined in libc.a
1696 [21743] localedata: ks_IN@devanagari: abday strings mismatch the day
1697 strings
1698 [21744] libc: Tests failing on --enable-tunables --enable-stack-
1699 protector=all
1700 [21749] localedata: Wrong abbreviated day name (“abday”) for
1701 ar_JO/ar_LB/ar_SY
1702 [21756] localedata: missing yesstr, nostr for nds_DE and nds_NL
1703 [21757] localedata: missing yesstr, nostr for pap_AW and pap_CW
1704 [21759] localedata: missing yesstr and nostr for Tigrinya
1705 [21760] localedata: Fix LC_MESSAGES and LC_ADDRESS for anp_IN
1706 [21766] localedata: Wrong LC_MESSAGES for om_ET Locale
1707 [21767] localedata: Missing Bislama locales
1708 [21768] localedata: Missing yesstr and nostr for aa_ET
1709 [21770] localedata: Missing Field in li_NL
1710 [21778] nptl: Robust mutex may deadlock
1711 [21779] libc: MicroBlaze segfaults when loading libpthread
1712 [21783] localedata: Fix int_select international_call_prefixes
1713 [21784] localedata: Inconsistency in country_isbn
1714 [21788] localedata: Missing Country Postal Abbreviations
1715 [21794] localedata: Added-country_isbn-for-Italy
1716 [21795] localedata: Add/Fix country_isbn for France
1717 [21796] localedata: Added country_isbn for Republic of Korea
1718 [21797] localedata: Fix inconsistency in country_isbn and missing prefixes
1719 [21799] localedata: Added int_select international_call_prefixes
1720 [21801] localedata: Added int_select international_call_prefixes
1721 [21804] nptl: Double semicolon in thread-shared-types.h
1722 [21807] localedata: LC_ADDRESS fix for pap_CW
1723 [21808] localedata: Fix LC_ADDRESS for pap_AW
1724 [21821] localedata: Added country_name in mai_IN
1725 [21822] localedata: Fix LC_TIME for mai_IN
1726 [21823] localedata: missing yesstr, nostr for sa_IN
1727 [21825] localedata: Fix name_mrs for mag_IN
1728 [21828] localedata: 2.26 changelog should mention user visible changes
1729 with unicode 9.0
1730 [21835] localedata: Added Maithili language locale for Nepal
1731 [21838] localedata: Removed redundant data for the_NP
1732 [21839] localedata: Fix LC_MONETARY for ta_LK
1733 [21844] localedata: Fix Latin characters and Months Sequence.
1734 [21848] localedata: Fix mai_NP Title Name
1735
1736 \f
1737 Version 2.25
1738
1739 * The feature test macro __STDC_WANT_LIB_EXT2__, from ISO/IEC TR
1740 24731-2:2010, is supported to enable declarations of functions from that
1741 TR. Note that not all functions from that TR are supported by the GNU C
1742 Library.
1743
1744 * The feature test macro __STDC_WANT_IEC_60559_BFP_EXT__, from ISO/IEC TS
1745 18661-1:2014, is supported to enable declarations of functions and macros
1746 from that TS. Note that not all features from that TS are supported by
1747 the GNU C Library.
1748
1749 * The feature test macro __STDC_WANT_IEC_60559_FUNCS_EXT__, from ISO/IEC TS
1750 18661-4:2015, is supported to enable declarations of functions and macros
1751 from that TS. Note that most features from that TS are not supported by
1752 the GNU C Library.
1753
1754 * The nonstandard feature selection macros _REENTRANT and _THREAD_SAFE are
1755 now treated as compatibility synonyms for _POSIX_C_SOURCE=199506L.
1756 Since the GNU C Library defaults to a much newer revision of POSIX, this
1757 will only affect programs that specifically request an old conformance
1758 mode. For instance, a program compiled with -std=c89 -D_REENTRANT will
1759 see a change in the visible declarations, but a program compiled with
1760 just -D_REENTRANT, or -std=c99 -D_POSIX_C_SOURCE=200809L -D_REENTRANT,
1761 will not.
1762
1763 Some C libraries once required _REENTRANT and/or _THREAD_SAFE to be
1764 defined by all multithreaded code, but glibc has not required this for
1765 many years.
1766
1767 * The inclusion of <sys/sysmacros.h> by <sys/types.h> is deprecated. This
1768 means that in a future release, the macros “major”, “minor”, and “makedev”
1769 will only be available from <sys/sysmacros.h>.
1770
1771 These macros are not part of POSIX nor XSI, and their names frequently
1772 collide with user code; see for instance glibc bug 19239 and Red Hat bug
1773 130601. <stdlib.h> includes <sys/types.h> under _GNU_SOURCE, and C++ code
1774 presently cannot avoid being compiled under _GNU_SOURCE, exacerbating the
1775 problem.
1776
1777 * New <fenv.h> features from TS 18661-1:2014 are added to libm: the
1778 fesetexcept, fetestexceptflag, fegetmode and fesetmode functions, the
1779 femode_t type and the FE_DFL_MODE and FE_SNANS_ALWAYS_SIGNAL macros.
1780
1781 * Integer width macros from TS 18661-1:2014 are added to <limits.h>:
1782 CHAR_WIDTH, SCHAR_WIDTH, UCHAR_WIDTH, SHRT_WIDTH, USHRT_WIDTH, INT_WIDTH,
1783 UINT_WIDTH, LONG_WIDTH, ULONG_WIDTH, LLONG_WIDTH, ULLONG_WIDTH; and to
1784 <stdint.h>: INT8_WIDTH, UINT8_WIDTH, INT16_WIDTH, UINT16_WIDTH,
1785 INT32_WIDTH, UINT32_WIDTH, INT64_WIDTH, UINT64_WIDTH, INT_LEAST8_WIDTH,
1786 UINT_LEAST8_WIDTH, INT_LEAST16_WIDTH, UINT_LEAST16_WIDTH,
1787 INT_LEAST32_WIDTH, UINT_LEAST32_WIDTH, INT_LEAST64_WIDTH,
1788 UINT_LEAST64_WIDTH, INT_FAST8_WIDTH, UINT_FAST8_WIDTH, INT_FAST16_WIDTH,
1789 UINT_FAST16_WIDTH, INT_FAST32_WIDTH, UINT_FAST32_WIDTH, INT_FAST64_WIDTH,
1790 UINT_FAST64_WIDTH, INTPTR_WIDTH, UINTPTR_WIDTH, INTMAX_WIDTH,
1791 UINTMAX_WIDTH, PTRDIFF_WIDTH, SIG_ATOMIC_WIDTH, SIZE_WIDTH, WCHAR_WIDTH,
1792 WINT_WIDTH.
1793
1794 * New <math.h> features are added from TS 18661-1:2014:
1795
1796 - Signaling NaN macros: SNANF, SNAN, SNANL.
1797
1798 - Nearest integer functions: roundeven, roundevenf, roundevenl, fromfp,
1799 fromfpf, fromfpl, ufromfp, ufromfpf, ufromfpl, fromfpx, fromfpxf,
1800 fromfpxl, ufromfpx, ufromfpxf, ufromfpxl.
1801
1802 - llogb functions: the llogb, llogbf and llogbl functions, and the
1803 FP_LLOGB0 and FP_LLOGBNAN macros.
1804
1805 - Max-min magnitude functions: fmaxmag, fmaxmagf, fmaxmagl, fminmag,
1806 fminmagf, fminmagl.
1807
1808 - Comparison macros: iseqsig.
1809
1810 - Classification macros: iscanonical, issubnormal, iszero.
1811
1812 - Total order functions: totalorder, totalorderf, totalorderl,
1813 totalordermag, totalordermagf, totalordermagl.
1814
1815 - Canonicalize functions: canonicalize, canonicalizef, canonicalizel.
1816
1817 - NaN functions: getpayload, getpayloadf, getpayloadl, setpayload,
1818 setpayloadf, setpayloadl, setpayloadsig, setpayloadsigf, setpayloadsigl.
1819
1820 * The functions strfromd, strfromf, and strfroml, from ISO/IEC TS 18661-1:2014,
1821 are added to libc. They convert a floating-point number into string.
1822
1823 * Most of glibc can now be built with the stack smashing protector enabled.
1824 It is recommended to build glibc with --enable-stack-protector=strong.
1825 Implemented by Nick Alcock (Oracle).
1826
1827 * The function explicit_bzero, from OpenBSD, has been added to libc. It is
1828 intended to be used instead of memset() to erase sensitive data after use;
1829 the compiler will not optimize out calls to explicit_bzero even if they
1830 are "unnecessary" (in the sense that no _correct_ program can observe the
1831 effects of the memory clear).
1832
1833 * On ColdFire, MicroBlaze, Nios II and SH3, the float_t type is now defined
1834 to float instead of double. This does not affect the ABI of any libraries
1835 that are part of the GNU C Library, but may affect the ABI of other
1836 libraries that use this type in their interfaces.
1837
1838 * On x86_64, when compiling with -mfpmath=387 or -mfpmath=sse+387, the
1839 float_t and double_t types are now defined to long double instead of float
1840 and double. These options are not the default, and this does not affect
1841 the ABI of any libraries that are part of the GNU C Library, but it may
1842 affect the ABI of other libraries that use this type in their interfaces,
1843 if they are compiled or used with those options.
1844
1845 * The getentropy and getrandom functions, and the <sys/random.h> header file
1846 have been added.
1847
1848 * The buffer size for byte-oriented stdio streams is now limited to 8192
1849 bytes by default. Previously, on Linux, the default buffer size on most
1850 file systems was 4096 bytes (and thus remains unchanged), except on
1851 network file systems, where the buffer size was unpredictable and could be
1852 as large as several megabytes.
1853
1854 * The <sys/quota.h> header now includes the <linux/quota.h> header. Support
1855 for the Linux quota interface which predates kernel version 2.4.22 has
1856 been removed.
1857
1858 * The malloc_get_state and malloc_set_state functions have been removed.
1859 Already-existing binaries that dynamically link to these functions will
1860 get a hidden implementation in which malloc_get_state is a stub. As far
1861 as we know, these functions are used only by GNU Emacs and this change
1862 will not adversely affect already-built Emacs executables. Any undumped
1863 Emacs executables, which normally exist only during an Emacs build, should
1864 be rebuilt by re-running “./configure; make” in the Emacs build tree.
1865
1866 * The “ip6-dotint” and “no-ip6-dotint” resolver options, and the
1867 corresponding RES_NOIP6DOTINT flag from <resolv.h> have been removed.
1868 “no-ip6-dotint” had already been the default, and support for the
1869 “ip6-dotint” option was removed from the Internet in 2006.
1870
1871 * The "ip6-bytestring" resolver option and the corresponding RES_USEBSTRING
1872 flag from <resolv.h> have been removed. The option relied on a
1873 backwards-incompatible DNS extension which was never deployed on the
1874 Internet.
1875
1876 * The flags RES_AAONLY, RES_PRIMARY, RES_NOCHECKNAME, RES_KEEPTSIG,
1877 RES_BLAST defined in the <resolv.h> header file have been deprecated.
1878 They were already unimplemented.
1879
1880 * The "inet6" option in /etc/resolv.conf and the RES_USE_INET6 flag for
1881 _res.flags are deprecated. The flag was standardized in RFC 2133, but
1882 removed again from the IETF name lookup interface specification in RFC
1883 2553. Applications should use getaddrinfo instead.
1884
1885 * DNSSEC-related declarations and definitions have been removed from the
1886 <arpa/nameser.h> header file, and libresolv will no longer attempt to
1887 decode the data part of DNSSEC record types. Previous versions of glibc
1888 only implemented minimal support for the previous version of DNSSEC, which
1889 is incompatible with the currently deployed version.
1890
1891 * The resource record type classification macros ns_t_qt_p, ns_t_mrr_p,
1892 ns_t_rr_p, ns_t_udp_p, ns_t_xfr_p have been removed from the
1893 <arpa/nameser.h> header file because the distinction between RR types and
1894 meta-RR types is not officially standardized, subject to revision, and
1895 thus not suitable for encoding in a macro.
1896
1897 * The types res_sendhookact, res_send_qhook, re_send_rhook, and the qhook
1898 and rhook members of the res_state type in <resolv.h> have been removed.
1899 The glibc stub resolver did not support these hooks, but the header file
1900 did not reflect that.
1901
1902 * For multi-arch support it is recommended to use a GCC which has
1903 been built with support for GNU indirect functions. This ensures
1904 that correct debugging information is generated for functions
1905 selected by IFUNC resolvers. This support can either be enabled by
1906 configuring GCC with '--enable-gnu-indirect-function', or by
1907 enabling it by default by setting 'default_gnu_indirect_function'
1908 variable for a particular architecture in the GCC source file
1909 'gcc/config.gcc'.
1910
1911 * GDB pretty printers have been added for mutex and condition variable
1912 structures in POSIX Threads. When installed and loaded in gdb these pretty
1913 printers show various pthread variables in human-readable form when read
1914 using the 'print' or 'display' commands in gdb.
1915
1916 * Tunables feature added to allow tweaking of the runtime for an application
1917 program. This feature can be enabled with the '--enable-tunables' configure
1918 flag. The GNU C Library manual has details on usage and README.tunables has
1919 instructions on adding new tunables to the library.
1920
1921 * A new version of condition variables functions have been implemented in
1922 the NPTL implementation of POSIX Threads to provide stronger ordering
1923 guarantees.
1924
1925 * A new version of pthread_rwlock functions have been implemented to use a more
1926 scalable algorithm primarily through not using a critical section anymore to
1927 make state changes.
1928
1929 Security related changes:
1930
1931 * On ARM EABI (32-bit), generating a backtrace for execution contexts which
1932 have been created with makecontext could fail to terminate due to a
1933 missing .cantunwind annotation. This has been observed to lead to a hang
1934 (denial of service) in some Go applications compiled with gccgo. Reported
1935 by Andreas Schwab. (CVE-2016-6323)
1936
1937 * The DNS stub resolver functions would crash due to a NULL pointer
1938 dereference when processing a query with a valid DNS question type which
1939 was used internally in the implementation. The stub resolver now uses a
1940 question type which is outside the range of valid question type values.
1941 (CVE-2015-5180)
1942
1943 The following bugs are resolved with this release:
1944
1945 [4099] stdio: Overly agressive caching by stream i/o functions.
1946 [7065] build: Support building glibc with -fstack-protector or -fstack-
1947 protector-all
1948 [9842] localedata: en_CA: incorrect date format
1949 [13165] nptl: pthread_cond_wait() can consume a signal that was sent
1950 before it started waiting
1951 [14139] manual: Do not hardcode platform names in manual/libm-err-tab.pl
1952 [15765] nptl: sem_open is wrongly a cancellation point
1953 [16421] network: IN6_IS_ADDR_UNSPECIFIED can use undefined s6_addr32
1954 [16458] libc: endian.h and netinet/in.h byte order macros return values of
1955 wrong type
1956 [16628] dynamic-link: Segfault after a binary without pthread dlopen()s a
1957 library linked with pthread
1958 [16630] nptl: Use SYSENTER for pthread_cond_broadcast/signal() (i.e. fix
1959 "FIXME: Ingo" issue)
1960 [16907] libc: <argp.h> compiled with --std=cXX disables __attribute__
1961 [17252] libc: getrandom and getentropy syscall
1962 [17730] malloc: thread-local storage is sometimes improperly free()'d
1963 after being __libc_memalign()'d
1964 [18241] stdio: failed fseek on memstream does not set errno and can fail
1965 when it shouldnt
1966 [18243] nptl: sem_wait, sem_timedwait are cancellation points shm_open is
1967 not
1968 [18463] nptl: pthread_cond_broadcast issue when surrounded by
1969 PTHREAD_PRIO_INHERIT mutex on ARM
1970 [18784] network: res_query and related function crash for special record
1971 type queries (CVE-2015-5180)
1972 [19380] math: strtod does not raise "inexact"
1973 [19387] string: Integer overflow in memchr
1974 [19390] string: Integer overflow in strncat
1975 [19398] build: linknamespace tests fail with massively parallel build
1976 [19402] nptl: Deadlock with robust shared mutex and asynchronous
1977 termination
1978 [19469] malloc: M_PERTURB in test-skeleton.c invalidates malloc tests
1979 [19473] malloc: Turn malloc_get_state etc. in compatibility symbols
1980 [19514] libc: [PATCH] Fix spelling errors in spelling
1981 "implement"/"implementation" in several places
1982 [19582] network: Deprecate RES_USE_INET6
1983 [19673] manual: clog10 docs appear to be erroneous
1984 [19810] dynamic-link: dlopen with both RTLD_NOLOAD and RTLD_NODELETE
1985 causes a segmentation fault
1986 [19826] libc: invalid pointer returned from __tls_get_addr with static
1987 linking
1988 [20016] network: resolv: Remove hooks support from the API
1989 [20019] dynamic-link: NULL pointer dereference in libc.so.6 IFUNC due to
1990 uninitialized GOT
1991 [20033] math: [x86_64] vectorized math function don't call the __finite
1992 versions
1993 [20116] nptl: use after free in pthread_create
1994 [20181] stdio: open_memstream(): writes not at end of stream corrupt data
1995 [20292] dynamic-link: Comparison in elf/dl-open.c _dl_addr_inside_object
1996 is always true.
1997 [20311] nptl: please install proc_service.h
1998 [20366] libc: Compilation errors in installed headers in strict-compliance
1999 modes
2000 [20370] malloc: malloc: Arena free list management is still racy
2001 (incorrect fix in bug 19243)
2002 [20386] libc: assert (X = 0) does not result in GCC warning
2003 [20432] malloc: malloc: Minimize interface required for interposition
2004 [20435] libc: Missing unwind info in __startcontext causes infinite loop
2005 in _Unwind_Backtrace (CVE-2016-6323)
2006 [20444] hurd: recvmsg: PF_LOCAL sockets and msg_name lead to SIGLOST
2007 [20452] nptl: Addition of sysdep.o to libpthread.a breaks relinking
2008 libpthread.a
2009 [20455] math: [powerpc] fesetexceptflag fails to clear FE_INVALID
2010 [20459] localedata: et_EE: locale has wrong {p,n}_cs_precedes value
2011 [20477] network: resolv: incorrect double-checked locking related to
2012 _res_hconf
2013 [20478] libc: libc_ifunc macro and similar usages leads to false debug-
2014 information.
2015 [20495] math: x86_64 performance degradation due to AVX/SSE transition
2016 penalty
2017 [20497] localedata: lt_LT: LC_TIME d_fmt used is obsolete
2018 [20508] dynamic-link: _dl_runtime_resolve_avx/_dl_runtime_profile_avx512
2019 cause transition penalty
2020 [20517] math: sparcv9 missing fdiml compat symbol
2021 [20524] manual: strverscmp is inconsistent
2022 [20525] libc: <sys/quota.h> should be based on kernel headers
2023 [20539] math: GCC 7 -static -lm fails to link at -O3
2024 [20554] libc: ld: bss-plt forced due to /usr/lib/libc_nonshared.a(ppc-
2025 mcount.oS)
2026 [20558] string: POSIX bcopy/bzero decls do not implement Object Size
2027 Checking
2028 [20591] network: Remove obsolete DNSSEC support
2029 [20592] network: DNS resource record type classification macros in
2030 <arpa/nameser.h> are incorrect
2031 [20593] network: Update DNS RR type definitions
2032 [20611] network: getaddrinfo accepts invalid numeric scope IDs
2033 [20615] build: glibc build fails when using --with-cpu=power9 --enable-
2034 multi-arch
2035 [20629] network: libresolv: Remove support for bitstring labels
2036 (RES_USEBSTRING)
2037 [20647] libc: GLIBC quitting every program - glibc on Pentium-S leads to
2038 assertion: "maxidx >= 2"
2039 [20660] math: [arm] Use VSQRT
2040 [20662] libc: checking whether x86_64-pc-linux-gnu-gcc implicitly enables
2041 -fstack-protector no (32bit gcc 6.2.0 pie and ssp enable)
2042 [20689] libc: Test for FMA should also check for AVX.
2043 [20707] glob: gl_pathv entries not set to NULL with GLOB_DOOFFS
2044 [20715] math: iszero macro breaks existing code
2045 [20718] math: [powerpc] copysignl raises "invalid" for sNaN
2046 [20728] libc: powerpc: Missing TOC stub in clone
2047 [20729] build: glibc-2.24 fails to build for i486 with -Os
2048 [20750] build: Build fails with default PIE enabled gcc-6.x
2049 [20768] math: [alpha] sqrt fegetenv namespace
2050 [20785] libc: binutils 2.28 fails to assemble power6/memset.S file in
2051 glibc
2052 [20787] math: float_t is defined as float incorrectly on x86_64 even if
2053 __FLT_EVAL_METHOD__ is 2
2054 [20790] network: rpcgen buffer overrun in get_prog_declaration
2055 [20822] nptl: race condition in __lll_unlock_elision on powerpc
2056 [20829] libc: crypt snprintf namespace
2057 [20847] libc: tst-vfork3 failure
2058 [20855] math: Default bits/mathdef.h has inappropriate float_t
2059 [20859] math: [sh4] FP_ILOGB0 invalid
2060 [20864] localedata: iconv: cp936 missing single-byte euro sign (0x80,
2061 U+20AC), not same as GBK
2062 [20915] dynamic-link: global-dynamic TLS broken on aarch64
2063 [20916] math: pow handling of sNaN arguments
2064 [20918] build: Building with --enable-nss-crypt fails tst-linkall-static.
2065 [20919] math: Bad pow (qNaN, 0) result with -lieee
2066 [20940] math: hypot sNaN handling
2067 [20947] math: fmax, fmin sNaN handling
2068 [20956] libc: debug/tst-backtrace3-6 don't work with -O3 anymore
2069 [20964] network: sunrpc: Stack-based buffer overflow in getrpcport with
2070 RES_USE_INET6
2071 [20971] string: powerpc64/power7 memchr overflows internal pointer check
2072 [20973] nptl: robust mutexes: Lost wake-ups
2073 [20974] locale: bs_BA: yesexpr/noexpr regexes accept any character
2074 [20978] nis: strlen on null pointer in nss_nisplus
2075 [20985] nptl: robust mutexes: lowlevelrobustlock assembly on x86 blocks on
2076 wrong condition
2077 [21014] string: i686 memchr overflows internal pointer check
2078 [21019] libc: [mips] n32 lseek incorrect on overflow
2079 [21022] libc: [microblaze] __backtrace get_frame_size namespace
2080 [21026] libc: [MIPS] readahead syscall is broken on n64
2081 [21028] math: Fallback fesetexceptflag should always succeed
2082 [21045] libc: [powerpc-nofpu] swapcontext does not restore signal mask
2083 [21047] math: arm: fpu_control.h: _FPU_GETCW/_FPU_SETCW is rejected by
2084 clang
2085 [21053] libc: [SH] Namespace pollution from sys/ucontext.h
2086 [21061] librt: [microblaze] librt lost clock_* exports
2087 [21073] libc: tunables: insecure environment variables passed to
2088 subprocesses with AT_SECURE
2089 [21081] string: Missing vzeroupper in memset-vec-unaligned-erms.S
2090 \f
2091 Version 2.24
2092
2093 * The minimum Linux kernel version that this version of the GNU C Library
2094 can be used with is 3.2, except on i[4567]86 and x86_64, where Linux
2095 kernel version 2.6.32 or later suffices (on architectures that already
2096 required kernel versions more recent than 3.2, those requirements remain
2097 unchanged). Linux 3.2 or later kernel headers are required on all
2098 architectures.
2099
2100 * The pap_AN locale has been deleted. This has been deprecated for a long
2101 time. It has been replaced by pap_AW & pap_CW, both of which have long
2102 been included in previous releases.
2103
2104 * The readdir_r and readdir64_r functions have been deprecated. It is
2105 recommended to use readdir and readdir64 instead.
2106
2107 * The type “union wait” has been removed. It was deprecated in the early
2108 1990s and never part of POSIX. Application code should use the int type
2109 instead of “union wait”.
2110
2111 * A new NSS action is added to facilitate large distributed system
2112 administration. The action, MERGE, allows remote user stores like LDAP
2113 to be merged into local user stores like /etc/groups in order to provide
2114 easy to use, updated, and managed sets of merged credentials. The new
2115 action can be used by configuring it in /etc/nsswitch.conf:
2116 group: files [SUCCESS=merge] nis
2117 Implemented by Stephen Gallagher (Red Hat).
2118
2119 * The deprecated __malloc_initialize_hook variable has been removed from the
2120 API.
2121
2122 * The long unused localedef --old-style option has been removed. It hasn't
2123 done anything in over 16 years. Scripts using this option can safely
2124 drop it.
2125
2126 * nextupl, nextup, nextupf, nextdownl, nextdown and nextdownf are added to
2127 libm. They are defined by TS 18661 and IEEE754-2008. The nextup functions
2128 return the next representable value in the direction of positive infinity
2129 and the nextdown functions return the next representable value in the
2130 direction of negative infinity. These are currently enabled as GNU
2131 extensions.
2132
2133 Security related changes:
2134
2135 * An unnecessary stack copy in _nss_dns_getnetbyname_r was removed. It
2136 could result in a stack overflow when getnetbyname was called with an
2137 overly long name. (CVE-2016-3075)
2138
2139 * Previously, getaddrinfo copied large amounts of address data to the stack,
2140 even after the fix for CVE-2013-4458 has been applied, potentially
2141 resulting in a stack overflow. getaddrinfo now uses a heap allocation
2142 instead. Reported by Michael Petlan. (CVE-2016-3706)
2143
2144 * The glob function suffered from a stack-based buffer overflow when it was
2145 called with the GLOB_ALTDIRFUNC flag and encountered a long file name.
2146 Reported by Alexander Cherepanov. (CVE-2016-1234)
2147
2148 * The Sun RPC UDP client could exhaust all available stack space when
2149 flooded with crafted ICMP and UDP messages. Reported by Aldy Hernandez'
2150 alloca plugin for GCC. (CVE-2016-4429)
2151
2152 * The IPv6 name server management code in libresolv could result in a memory
2153 leak for each thread which is created, performs a failing naming lookup,
2154 and exits. Over time, this could result in a denial of service due to
2155 memory exhaustion. Reported by Matthias Schiffer. (CVE-2016-5417)
2156
2157 The following bugs are resolved with this release:
2158
2159 [1170] localedata: ne_NP: update Nepali locale definition file
2160 [3629] manual: stpcpy description in string.texi refers to MS-DOG instead
2161 of MS-DOS.
2162 [6527] malloc: [powerpc] Malloc alignment insufficient for PowerPC
2163 [6796] math: fdim() does not set errno on overflow
2164 [10354] libc: posix_spawn should use vfork() in more cases than presently
2165 [11213] localedata: localedata: add copyright disclaimer to locale files
2166 [12143] localedata: chr_US: new Cherokee locale
2167 [12450] localedata: sgs_LT: new locale
2168 [12676] localedata: ln_CD: new locale
2169 [13237] localedata: LC_ADDRESS.country_name: update all locales w/latest
2170 CLDR data
2171 [13304] math: fma, fmaf, fmal produce wrong results
2172 [14259] build: --localedir arg to configure is ignored
2173 [14499] nptl: Does posix_spawn invoke atfork handlers / use vfork?
2174 [14750] libc: Race condition in posix_spawn vfork usage vs signal handlers
2175 [14934] localedata: es_CL: wrong first weekday chilean locale
2176 [15262] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of
2177 romanisation
2178 [15263] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of 1/0
2179 and +/-
2180 [15264] localedata: LC_MESSAGES.yesstr/nostr: lacking in many locales
2181 [15368] nptl: raise() is not async-signal-safe
2182 [15479] math: ceil, floor, round and trunc raise inexact exception
2183 [15578] localedata: kk_KZ: various updates
2184 [16003] localedata: pap_AN: punt old locale
2185 [16137] localedata: iw_IL: punt old locale
2186 [16190] localedata: eo: new esperanto locale
2187 [16374] localedata: lv_LV: change currency symbol in LC_MONETARY to euro
2188 [16742] malloc: race condition: pthread_atfork() called before first
2189 malloc() results in unexpected locking behaviour/deadlocks
2190 [16975] localedata: LC_MESSAGES.yesexpr/noexpr: revisit capitalization in
2191 all locales
2192 [16983] localedata: postal_fmt does not allow %l and %n modifiers
2193 [17565] localedata: pt_PT: wrong (work-)week start
2194 [17899] math: [powerpc] floorl returns negative zero with FE_DOWNWARD
2195 [17950] build: Build fails with -msse
2196 [18205] localedata: be_BY*: wrong first_weekday and first_workday
2197 [18433] libc: posix_spawn does not return correctly upon failure to
2198 execute
2199 [18453] localedata: charmaps/IBM875: incorrect codes
2200 [18712] string: bits/string2.h incompatible with -O2 -Werror=packed
2201 -Wsystem-headers
2202 [18896] localedata: he_IL: improvements for currency
2203 [18911] localedata: ro_RO: Correcting week day name for "Tuesday" in
2204 Romanian locale data
2205 [18960] locale: s390: _nl_locale_subfreeres uses larl opcode on misaligned
2206 symbol
2207 [19056] libc: Deprecate readdir_r
2208 [19133] localedata: pt_*: days & months should be lowercase in Portuguese
2209 language
2210 [19198] localedata: nl_NL: small improvements for Dutch locales
2211 [19257] network: Per-thread memory leak in __res_vinit with IPv6
2212 nameservers (CVE-2016-5417)
2213 [19269] build: tst-audit4 and tst-audit10 failures with gcc-6 on non avx
2214 machine
2215 [19400] locale: Language missing in "iso-639.def", trivial fix in
2216 description
2217 [19431] malloc: Deadlock between fflush, getdelim, and fork
2218 [19505] libc: Incorrect file descriptor validity checks in
2219 posix_spawn_file_actions_add{open,close,dup2}
2220 [19509] dynamic-link: dlsym, dlvsym do not report errors through dlerror
2221 when using RTLD_NEXT
2222 [19512] locale: Stale `#ifndef HAVE_BUILTIN_EXPECT' in
2223 `intl/{gettextP,loadinfo}.h'
2224 [19534] libc: execle, execlp may use malloc
2225 [19568] localedata: *_CH: Swiss locales have inconsistent start of week
2226 [19573] network: res_nclose and __res_maybe_init disagree about name
2227 server initialization, breaking Hesiod
2228 [19575] localedata: Status of GB18030 tables
2229 [19581] localedata: sr_* date_fmt string contains additional newline
2230 [19583] string: SSSE3_Fast_Copy_Backward flag needs to be enabled for AMD
2231 Excavator core
2232 [19592] math: [ldbl-128ibm] ceill incorrect in non-default rounding modes
2233 [19593] math: [ldbl-128ibm] truncl incorrect in non-default rounding modes
2234 [19594] math: [ldbl-128ibm] roundl incorrect in non-default rounding modes
2235 [19595] math: [ldbl-128ibm] fmodl incorrect for results in subnormal
2236 double range
2237 [19602] math: [ldbl-128ibm] fmodl handling of equal arguments with low
2238 part zero incorrect
2239 [19603] math: [ldbl-128ibm] remainderl, remquol incorrect sign handling in
2240 equality tests
2241 [19610] dynamic-link: ldconfig -X removes stale symbolic links
2242 [19613] libc: s390x (64 bit) macro expansion WCOREDUMP and others
2243 [19633] locale: strfmon_l applies global locale to number formatting
2244 [19642] network: Memory leak in getnameinfo
2245 [19648] libc: test-skeleton.c: Do not set RLIMIT_DATA
2246 [19653] libc: Potential for NULL pointer dereference (CWE-476) in
2247 glibc-2.22
2248 [19654] math: [x86_64] Need testcase for BZ #19590 fix
2249 [19671] localedata: Missing Sanity Check for malloc() in 'tst-fmon.c' &
2250 'tst-numeric.c'
2251 [19674] math: [ldbl-128ibm] powl incorrect overflow handling
2252 [19677] math: [ldbl-128ibm] remainderl equality test incorrect for zero
2253 low part
2254 [19678] math: [ldbl-128ibm] nextafterl, nexttowardl incorrect sign of zero
2255 result
2256 [19679] dynamic-link: gcc-4.9.3 C++ exception handling broken due to
2257 unaligned stack
2258 [19726] locale: Converting UCS4LE to INTERNAL with iconv() does not update
2259 pointers and lengths in error-case.
2260 [19727] locale: Converting from/to UTF-xx with iconv() does not always
2261 report errors on UTF-16 surrogates values.
2262 [19755] nscd: nscd assertion failure in gc
2263 [19758] dynamic-link: Typo in EXTRA_LD_ENVVARS for x86-64
2264 [19759] libc: mempcpy shouldn't be inlined
2265 [19762] dynamic-link: HAS_CPU_FEATURE/HAS_ARCH_FEATURE are easy to misuse
2266 [19765] libc: s390 needs an optimized mempcpy
2267 [19779] glob: glob: buffer overflow with GLOB_ALTDIRFUNC due to incorrect
2268 NAME_MAX limit assumption (CVE-2016-1234)
2269 [19783] build: benchtests don't support --enable-hardcoded-path-in-tests
2270 [19787] network: Missing and incorrect truncation checks in getnameinfo
2271 [19790] math: [ldbl-128ibm] nearbyintl incorrect in non-default rounding
2272 modes
2273 [19791] network: Assertion failure in res_query.c with un-connectable name
2274 server addresses
2275 [19792] libc: MIPS: backtrace yields infinite backtrace with makecontext
2276 [19822] math: libm.so install clobbers old version
2277 [19825] network: resolv: send_vc can return uninitialized data in second
2278 response to getaddrinfo
2279 [19830] network: nss_dns: should check RDATA length against buffer length
2280 [19831] network: nss_dns: getaddrinfo returns uninitialized data when
2281 confronted with A/AAAA records of invalid size
2282 [19837] nss: nss_db: No retries for some long lines with a larger buffer
2283 [19848] math: powl(10,n) for n=-4,-5,-6,-7 is off by more than 1 ULP
2284 [19853] stdio: Printing IBM long double in decimal with high precision is
2285 sometimes incorrect
2286 [19860] build: x86_64: compile errors for tst-audit10 and tst-auditmod10b
2287 [19861] nptl: libpthread IFUNC resolver for fork can lead to crash
2288 [19862] network: resolv, nss_dns: Remove remaining logging of unexpected
2289 record types
2290 [19865] network: Assertion failure or memory leak in
2291 _nss_dns_getcanonname_r
2292 [19868] network: nss_dns: netent code does not skip over non-PTR records
2293 [19879] network: nss_dns: Stack overflow in getnetbyname implementation
2294 (CVE-2016-3075)
2295 [19881] string: Improve x86-64 memset
2296 [19907] string: Incorrect memcpy tests
2297 [19916] dynamic-link: S390: fprs/vrs are not saved/restored while
2298 resolving symbols
2299 [19925] libc: termios.h XCASE namespace
2300 [19928] string: memmove-vec-unaligned-erms.S is slow with large data size
2301 [19929] libc: limits.h NL_NMAX namespace
2302 [19931] stdio: Memory leak in vfprintf
2303 [19957] libc: clone(CLONE_VM) access invalid parent memory
2304 [19963] localedata: en_IL: New locale
2305 [19989] stdio: stdio.h cuserid namespace
2306 [19994] network: getaddrinfo does not restore RES_USE_INET6 flag in
2307 gethosts
2308 [19996] locale: langinfo.h nl_langinfo_l namespace
2309 [20005] stdio: fflush on a file opened with fmemopen resets position to 0
2310 [20010] network: getaddrinfo: Stack overflow in hostent translation
2311 (CVE-2016-3706)
2312 [20012] stdio: libio: fmemopen append mode failure
2313 [20014] stdio: stdio.h namespace for pre-threads POSIX
2314 [20017] network: resolv: Use gmtime_r instead of gmtime in p_secstodate
2315 [20023] libc: fcntl.h timespec namespace
2316 [20024] math: [x86_64] vectorized sincos trashes the stack
2317 [20031] network: nss_hesiod: Heap overflow in get_txt_records
2318 [20041] time: sys/time.h timespec namespace
2319 [20043] libc: unistd.h missing cuserid for UNIX98 and before
2320 [20044] libc: unistd.h missing pthread_atfork for UNIX98
2321 [20051] libc: ttyslot in wrong header under wrong conditions
2322 [20054] libc: gethostname not declared for XPG4
2323 [20055] libc: termios.h missing tcgetsid for XPG4
2324 [20072] dynamic-link: x86 init_cpu_features is called twice in static
2325 executable
2326 [20073] libc: sys/stat.h fchmod namespace
2327 [20074] libc: stdlib.h rand_r namespace
2328 [20076] libc: sys/stat.h missing S_IFSOCK, S_ISSOCK for XPG4
2329 [20094] libc: stdlib.h should not declare grantpt, ptsname, unlockpt for
2330 XPG3
2331 [20111] libc: struct sockaddr_storage cannot be aggregate-copied
2332 [20112] network: sunrpc: stack (frame) overflow in Sun RPC clntudp_call
2333 (CVE-2016-4429)
2334 [20115] string: Extra alignment in memset-vec-unaligned-erms.S
2335 [20119] libc: Wrong mask for processors level type from CPUID
2336 [20139] dynamic-link: Upper part of zmm is zeroed if Glibc is built with
2337 AS not supporting AVX512
2338 [20151] math: [ldbl-128/ldbl-128ibm] j0l, j1l, y0l, y1l return sNaN for
2339 sNaN argument
2340 [20153] math: [ldbl-128ibm] sqrtl (sNaN) returns sNaN
2341 [20156] math: [ldbl-128ibm] ceill, rintl etc. return sNaN for sNaN
2342 argument
2343 [20157] math: [powerpc] fabsl (sNaN) wrongly raises "invalid"
2344 [20160] math: [powerpc] ceil, rint etc. return sNaN for sNaN input
2345 [20178] libc: posix_spawn{p} should not call exit
2346 [20191] stdio: libio: vtables hardening
2347 [20195] string: FMA4 detection requires CPUID execution with register
2348 eax=0x80000001
2349 [20198] libc: quick_exit incorrectly destroys C++11 thread objects.
2350 [20205] math: [i386/x86_64] nextafterl incorrect incrementing negative
2351 subnormals
2352 [20212] math: acos (sNaN) returns sNaN
2353 [20213] math: asin (sNaN) returns sNaN
2354 [20214] network: Linux header sync with linux/in6.h and ipv6.h again.
2355 [20218] math: [i386] asinhl (sNaN) returns sNaN
2356 [20219] math: [i386] atanhl (sNaN) returns sNaN
2357 [20222] stdio: fopencookie: Mangle function pointers
2358 [20224] math: [i386] cbrtl (sNaN) returns sNaN
2359 [20225] math: ldexp, scalbn, scalbln return sNaN for sNaN input
2360 [20226] math: [i386/x86_64] expl, exp10l, expm1l return sNaN for sNaN
2361 input
2362 [20227] math: [i386/x86_64] logl (sNaN) returns sNaN
2363 [20228] math: [i386/x86_64] log10l (sNaN) returns sNaN
2364 [20229] math: [i386/x86_64] log1pl (sNaN) returns sNaN
2365 [20232] math: [ldbl-128] expm1l (sNaN) returns sNaN
2366 [20233] math: [ldbl-128ibm] expm1l (sNaN) returns sNaN
2367 [20234] math: [ldbl-128ibm] log1pl (sNaN) returns sNaN
2368 [20235] math: [i386/x86_64] log2l (sNaN) returns sNaN
2369 [20237] nss: nss_db: get*ent segfaults without preceding set*ent
2370 [20240] math: modf (sNaN) returns sNaN
2371 [20248] libc: debug/tst-longjump_chk2 calls printf from a signal handler
2372 [20250] math: frexp (sNaN) returns sNaN
2373 [20252] math: atan2 (sNaN, qNaN) fails to raise "invalid"
2374 [20255] math: [i386] fdim, fdimf return with excess range and precision /
2375 double rounding
2376 [20256] math: [i386/x86_64] fdiml returns sNaN for sNaN input
2377 [20260] string: ../sysdeps/x86/bits/string.h:1092:3: error: array
2378 subscript is below array bounds [-Werror=array-bounds]
2379 [20262] nis: _nss_nis_initgroups_dyn always returns NSS_STATUS_NOTFOUND
2380 [20263] nptl: robust mutex deadlocks if other thread requests timedlock
2381 (Only arm/linux)
2382 [20277] libc: $dp is not initialized correctly in sysdeps/hppa/start.S
2383 [20284] malloc: malloc: Corrupt arena avoidance causes unnecessary mmap
2384 fallbacks
2385 [20296] math: [i386/x86_64] scalbl returns sNaN for sNaN input, missing
2386 "invalid" exceptions
2387 [20314] nptl: make[4]: *** [/usr/include/stdlib.h] Error 1
2388 [20316] localedata: id_ID: Februari instead of Pebruari
2389 [20327] string: POWER8 strcasecmp returns incorrect result
2390 [20347] math: Failure: Test: j0_downward (0xap+0)
2391 [20348] libc: FAIL: misc/tst-preadvwritev64
2392 [20349] libc: 64-bit value is passed differently in p{readv,writev}{64}
2393 [20350] libc: There is no test for p{read,write}64
2394 [20357] math: Incorrect cos result for 1.5174239687223976
2395 [20384] build: Don't run libmvec-sincos-avx* tests on non avx machines
2396 \f
2397 Version 2.23
2398
2399 * Unicode 8.0.0 Support: Character encoding, character type info, and
2400 transliteration tables are all updated to Unicode 8.0.0, using new
2401 and/or improved generator scripts contributed by Mike FABIAN (Red Hat).
2402 These updates cause user visible changes, such as the fixes for bugs
2403 89, 16061, and 18568.
2404
2405 * sched_setaffinity, pthread_setaffinity_np no longer attempt to guess the
2406 kernel-internal CPU set size. This means that requests that change the
2407 CPU affinity which failed before (for example, an all-ones CPU mask) will
2408 now succeed. Applications that need to determine the effective CPU
2409 affinities need to call sched_getaffinity or pthread_getaffinity_np after
2410 setting it because the kernel can adjust it (and the previous size check
2411 would not detect this in the majority of cases).
2412
2413 * The fts.h header can now be used with -D_FILE_OFFSET_BITS=64. With LFS
2414 the following new symbols are used: fts64_children, fts64_close,
2415 fts64_open, fts64_read and fts64_set.
2416
2417 * getaddrinfo now detects certain invalid responses on an internal netlink
2418 socket. If such responses are received, an affected process will
2419 terminate with an error message of "Unexpected error <number> on netlink
2420 descriptor <number>" or "Unexpected netlink response of size <number> on
2421 descriptor <number>". The most likely cause for these errors is a
2422 multi-threaded application which erroneously closes and reuses the netlink
2423 file descriptor while it is used by getaddrinfo.
2424
2425 * A defect in the malloc implementation, present since glibc 2.15 (2012) or
2426 glibc 2.10 via --enable-experimental-malloc (2009), could result in the
2427 unnecessary serialization of memory allocation requests across threads.
2428 The defect is now corrected. Users should see a substantial increase in
2429 the concurent throughput of allocation requests for applications which
2430 trigger this bug. Affected applications typically create and
2431 destroy threads frequently. (Bug 19048 was reported and analyzed by
2432 Ericsson.)
2433
2434 * There is now a --disable-timezone-tools configure option for disabling the
2435 building and installing of the timezone related utilities (zic, zdump, and
2436 tzselect). This is useful for people who build the timezone data and code
2437 independent of the GNU C Library.
2438
2439 * The obsolete header <regexp.h> has been removed. Programs that require
2440 this header must be updated to use <regex.h> instead.
2441
2442 * The obsolete functions bdflush, create_module, get_kernel_syms,
2443 query_module and uselib are no longer available to newly linked binaries;
2444 the header <sys/kdaemon.h> has been removed. These functions and header
2445 were specific to systems using the Linux kernel and could not usefully be
2446 used with the GNU C Library on systems with version 2.6 or later of the
2447 Linux kernel.
2448
2449 * Optimized string, wcsmbs and memory functions for IBM z13.
2450 Implemented by Stefan Liebler.
2451
2452 * Newly linked programs that define a variable called signgam will no longer
2453 have it set by the lgamma, lgammaf and lgammal functions. Programs that
2454 require signgam to be set by those functions must ensure that they use the
2455 variable provided by the GNU C Library and declared in <math.h>, without
2456 defining their own copy.
2457
2458 * The minimum GCC version that can be used to build this version of the GNU
2459 C Library is GCC 4.7. Older GCC versions, and non-GNU compilers, can
2460 still be used to compile programs using the GNU C Library.
2461
2462 Security related changes:
2463
2464 * An out-of-bounds value in a broken-out struct tm argument to strftime no
2465 longer causes a crash. Reported by Adam Nielsen. (CVE-2015-8776)
2466
2467 * The LD_POINTER_GUARD environment variable can no longer be used to disable
2468 the pointer guard feature. It is always enabled. Previously,
2469 LD_POINTER_GUARD could be used to disable security hardening in binaries
2470 running in privileged AT_SECURE mode. Reported by Hector Marco-Gisbert.
2471 (CVE-2015-8777)
2472
2473 * An integer overflow in hcreate and hcreate_r could lead to an
2474 out-of-bounds memory access. Reported by Szabolcs Nagy. (CVE-2015-8778)
2475
2476 * The catopen function no longer has unbounded stack usage. Reported by
2477 Max. (CVE-2015-8779)
2478
2479 * The nan, nanf and nanl functions no longer have unbounded stack usage
2480 depending on the length of the string passed as an argument to the
2481 functions. Reported by Joseph Myers. (CVE-2014-9761)
2482
2483 * A stack-based buffer overflow was found in libresolv when invoked from
2484 libnss_dns, allowing specially crafted DNS responses to seize control
2485 of execution flow in the DNS client. The buffer overflow occurs in
2486 the functions send_dg (send datagram) and send_vc (send TCP) for the
2487 NSS module libnss_dns.so.2 when calling getaddrinfo with AF_UNSPEC
2488 family. The use of AF_UNSPEC triggers the low-level resolver code to
2489 send out two parallel queries for A and AAAA. A mismanagement of the
2490 buffers used for those queries could result in the response of a query
2491 writing beyond the alloca allocated buffer created by
2492 _nss_dns_gethostbyname4_r. Buffer management is simplified to remove
2493 the overflow. Thanks to the Google Security Team and Red Hat for
2494 reporting the security impact of this issue, and Robert Holiday of
2495 Ciena for reporting the related bug 18665. (CVE-2015-7547)
2496
2497 The following bugs are resolved with this release:
2498
2499 [89] localedata: Locales nb_NO and nn_NO should transliterate æøå
2500 [887] math: Math library function "logb" and "nextafter" inconsistent
2501 [2542] math: Incorrect return from float gamma (-0X1.FA471547C2FE5P+1)
2502 [2543] math: Incorrect return from float gamma (-0X1.9260DCP+1)
2503 [2558] math: Incorrect return from double gamma (-0X1.FA471547C2FE5P+1)
2504 [2898] libc: [improve] warning: the use of `mktemp' is dangerous, better
2505 use `mkstemp'
2506 [4404] localedata: German translation of "Alarm clock" is misleading
2507 [6799] math: nextafter() and nexttoward() doen't set errno on
2508 overflow/underflow errors
2509 [6803] math: scalb(), scalbln(), scalbn() do not set errno on
2510 overflow/underflow
2511 [10432] nis: _nss_nis_setnetgrent assertion failure
2512 [11460] libc: fts has no LFS support
2513 [12926] network: getaddrinfo()/make_request() may spin forever
2514 [13065] nptl: Race condition in pthread barriers
2515 [13690] nptl: pthread_mutex_unlock potentially cause invalid access
2516 [14341] dynamic-link: Dynamic linker crash when DT_JMPREL and DT_REL{,A}
2517 are not contiguous
2518 [14551] math: [ldbl-128ibm] strtold overflow handling for IBM long double
2519 [14912] libc: Rename non-installed bits/*.h headers
2520 [15002] libc: Avoid undefined behavior in posix_fallocate overflow check
2521 [15367] math: Let gcc use __builtin_isinf
2522 [15384] math: One constant fewer in ieee754/dbl-64/wordsize-64/s_finite.c
2523 [15421] math: lgamma wrongly sets signgam for ISO C
2524 [15470] math: [arm] On ARM llrintl() and llroundl() do not raise
2525 FE_INVALID with argument out of range
2526 [15491] math: [i386/x86_64] x86 nearbyint implementations wrongly clear
2527 all exceptions
2528 [15786] dynamic-link: ifunc resolver functions can smash function
2529 arguments
2530 [15918] math: Unnecessary check for equality in hypotf()
2531 [16061] localedata: Review / update transliteration data
2532 [16068] math: [i386/x86_64] x86 and x86_64 fesetenv exclude state they
2533 should include
2534 [16141] time: strptime %z offset restriction
2535 [16171] math: drem should be alias of remainder
2536 [16296] math: fegetround is pure?
2537 [16347] math: [ldbl-128ibm] ldbl-128/e_lgammal_r.c may not be suitable.
2538 [16364] libc: sleep may leave SIGCHLD blocked on sync cancellation on
2539 GNU/Linux
2540 [16399] math: [mips] lrint / llrint / lround / llround missing exceptions
2541 [16415] math: Clean up ldbl-128 / ldbl-128ibm expm1l for large positive
2542 arguments
2543 [16422] math: [powerpc] math-float, math-double failing llrint tests with
2544 "Exception "Inexact" set" on ppc32
2545 [16495] localedata: nl_NL: date_fmt: shuffle year/month around
2546 [16517] math: Missing underflow exception from tanf/tan/tanl
2547 [16519] math: Missing underflow exception from sinhf
2548 [16520] math: Missing underflow exception from tanhf
2549 [16521] math: Missing underflow exception from exp2
2550 [16620] math: [ldbl-128ibm] exp10l spurious overflows / bad directed
2551 rounding results
2552 [16734] stdio: fopen calls mmap to allocate its buffer
2553 [16961] math: nan function incorrect handling of bad sequences
2554 [16962] math: nan function unbounded stack allocation (CVE-2014-9761)
2555 [16973] localedata: Fix lang_lib/lang_term as per ISO 639-2
2556 [16985] locale: localedef: confusing error message when opening output
2557 fails
2558 [17118] math: ctanh(INFINITY + 2 * I) returns incorrect value
2559 [17197] locale: Redundant shift character in iconv conversion output at
2560 block boundary
2561 [17243] libc: trunk/posix/execl.c:53: va_args problem ?
2562 [17244] libc: trunk/sysdeps/unix/sysv/linux/semctl.c:116: va_args muxup ?
2563 [17250] dynamic-link: static linking breaks nss loading
2564 (getaddrinfo/getpwnam/etc...)
2565 [17404] libc: atomic_exchange_rel lacking a barrier on MIPS16, GCC before
2566 4.7?
2567 [17441] math: isnan() should use __builtin_isnan() in GCC
2568 [17514] nptl: Assert failure unlocking ERRORCHECK mutex after timedlock
2569 (related to lock elision)
2570 [17787] manual: Exponent on page 324 of the PDF ends prematurely
2571 [17886] time: strptime should be able to parse "Z" as a timezone with %z
2572 [17887] time: strptime should be able to parse "+01:00" style timezones
2573 [17905] libc: catopen() Multiple unbounded stack allocations
2574 (CVE-2015-8779)
2575 [18084] libc: backtrace (..., 0) dumps core on x86
2576 [18086] libc: nice() sets errno to 0 on success
2577 [18240] libc: hcreate, hcreate_r should fail with ENOMEM if element count
2578 is too large (CVE-2015-8778)
2579 [18251] dynamic-link: SONAME missing when audit modules provides path
2580 [18265] libc: add attributes for wchar string and memory functions
2581 [18370] math: csqrt missing underflows
2582 [18421] libc: [hppa] read-only segment has dynamic relocations
2583 [18472] libc: Obsolete syscall wrappers should be compat symbols
2584 [18480] libc: hppa glibc miscompilation in sched_setaffinity()
2585 [18491] localedata: Update tr_TR LC_CTYPE as part of Unicode updates
2586 [18525] localedata: Remove locale timezone information
2587 [18560] libc: [powerpc] spurious bits/ipc.h definitions
2588 [18568] localedata: Update locale data to Unicode 8.0
2589 [18589] locale: sort-test.sh fails at random
2590 [18595] math: ctan, ctanh missing underflows
2591 [18604] libc: assert macro-expands its argument
2592 [18610] math: S390: fetestexcept() reports any exception if DXC-code
2593 contains a vector instruction exception.
2594 [18611] math: j1, jn missing errno setting on underflow
2595 [18618] localedata: sync Chechen locale definitions with other *_RU
2596 locales
2597 [18647] math: powf(-0x1.000002p0, 0x1p30) returns 0 instead of +inf
2598 [18661] libc: Some x86-64 assembly codes don't align stack to 16 bytes
2599 [18665] network: In send_dg, the recvfrom function is NOT always using the
2600 buffer size of a newly created buffer (CVE-2015-7547)
2601 [18674] libc: [i386] trunk/sysdeps/i386/tst-auditmod3b.c:84: possible
2602 missing break ?
2603 [18675] libc: fpathconf(_PC_NAME_MAX) fails against large filesystems for
2604 32bit processes
2605 [18681] libc: regexp.h is obsolete and buggy, and should be desupported
2606 [18699] math: tilegx cproj() for various complex infinities does not yield
2607 infinity
2608 [18724] libc: Harden put*ent functions against data injection
2609 [18743] nptl: PowerPC: findutils testcase fails with --enable-lock-elision
2610 [18755] build: build errors with -DNDEBUG
2611 [18757] stdio: fmemopen fails to set errno on failure
2612 [18778] dynamic-link: ld.so crashes if failed dlopen causes libpthread to
2613 be forced unloaded
2614 [18781] libc: openat64 lacks O_LARGEFILE
2615 [18787] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/atomic.h:71:6:
2616 error: can’t find a register in class ‘R1_REGS’ while reloading ‘asm’
2617 [18789] math: [ldbl-128ibm] sinhl inaccurate near 0
2618 [18790] math: [ldbl-128ibm] tanhl inaccurate
2619 [18795] libc: stpncpy fortification misses buffer lengths that are
2620 statically too large
2621 [18796] build: build fails for --disable-mathvec
2622 [18803] math: hypot missing underflows
2623 [18820] stdio: fmemopen may leak memory on failure
2624 [18823] math: csqrt spurious underflows
2625 [18824] math: fma spurious underflows
2626 [18825] math: pow missing underflows
2627 [18857] math: [ldbl-128ibm] nearbyintl wrongly uses signaling comparisons
2628 [18868] nptl: pthread_barrier_init typo has in-theory-undefined behavior
2629 [18870] build: sem_open.c fails to compile with missing symbol
2630 FUTEX_SHARED
2631 [18872] stdio: Fix memory leak in printf_positional
2632 [18873] libc: posix_fallocate overflow check ineffective
2633 [18875] math: Excess precision leads incorrect libm
2634 [18877] libc: arm: mmap offset regression
2635 [18887] libc: memory corruption when using getmntent on blank lines
2636 [18918] localedata: hu_HU: change time to HH:MM:SS format
2637 [18921] libc: Regression: extraneous stat() and fstat() performed by
2638 opendir()
2639 [18928] dynamic-link: LD_POINTER_GUARD is not ignored for privileged
2640 binaries (CVE-2015-8777)
2641 [18951] math: tgamma missing underflows
2642 [18952] math: [ldbl-128/ldbl-128ibm] lgammal spurious "invalid", incorrect
2643 signgam
2644 [18953] localedata: lt_LT: change currency symbol to the euro
2645 [18956] math: powf inaccuracy
2646 [18961] math: [i386] exp missing underflows
2647 [18966] math: [i386] exp10 missing underflows
2648 [18967] math: math.h XSI POSIX namespace (gamma, isnan, scalb)
2649 [18969] build: multiple string test failures due to missing locale
2650 dependencies
2651 [18970] libc: Reference of pthread_setcancelstate in libc.a
2652 [18977] math: float / long double Bessel functions not in XSI POSIX
2653 [18980] math: i386 libm functions return with excess range and precision
2654 [18981] math: i386 scalb*, ldexp return with excess range and precision
2655 [18982] stdio: va_list and vprintf
2656 [18985] time: Passing out of range data to strftime() causes a segfault
2657 (CVE-2015-8776)
2658 [19003] math: [x86_64] fma4 version of pow inappropriate contraction
2659 [19007] libc: FAIL: elf/check-localplt with -z now and binutils 2.26
2660 [19012] locale: iconv_open leaks memory on error path
2661 [19016] math: clog, clog10 inaccuracy
2662 [19018] nptl: Mangle function pointers in tls_dtor_list
2663 [19032] math: [i386] acosh (-qNaN) spurious "invalid" exception
2664 [19046] math: ldbl-128 / ldbl-128ibm lgamma bad overflow handling
2665 [19048] malloc: malloc: arena free list can become cyclic, increasing
2666 contention
2667 [19049] math: [powerpc] erfc incorrect zero sign
2668 [19050] math: [powerpc] log* incorrect zero sign
2669 [19058] math: [x86_64] Link fail with -fopenmp and -flto
2670 [19059] math: nexttoward overflow incorrect in non-default rounding modes
2671 [19071] math: ldbl-96 lroundl incorrect just below powers of 2
2672 [19074] network: Data race in _res_hconf_reorder_addrs
2673 [19076] math: [ldbl-128ibm] log1pl (-1) wrong sign of infinity
2674 [19077] math: [ldbl-128ibm] logl (1) incorrect sign of zero result
2675 [19078] math: [ldbl-128ibm] expl overflow incorrect in non-default
2676 rounding modes
2677 [19079] math: dbl-64/wordsize-64 lround based on llround incorrect for
2678 ILP32
2679 [19085] math: ldbl-128 lrintl, lroundl missing exceptions for 32-bit long
2680 [19086] manual: posix_fallocate64 documented argument order is wrong.
2681 [19088] math: lround, llround missing exceptions close to overflow
2682 threshold
2683 [19094] math: lrint, llrint missing exceptions close to overflow threshold
2684 [19095] math: dbl-64 lrint incorrect for 64-bit long
2685 [19122] dynamic-link: Unnecessary PLT relocations in librtld.os
2686 [19124] dynamic-link: ld.so failed to build with older assmebler
2687 [19125] math: [powerpc32] llroundf, llround incorrect exceptions
2688 [19129] dynamic-link: [arm] Concurrent lazy TLSDESC resolution can crash
2689 [19134] math: [powerpc32] lround, lroundf spurious exceptions
2690 [19137] libc: i386/epoll_pwait.S doesn't support cancellation
2691 [19143] nptl: Remove CPU set size checking from sched_setaffinity,
2692 pthread_setaffinity_np
2693 [19156] math: [ldbl-128] j0l spurious underflows
2694 [19164] nptl: tst-getcpu fails with many possible CPUs
2695 [19168] math: math/test-ildoubl and math/test-ldouble failure
2696 [19174] nptl: PowerPC: TLE enabled pthread mutex performs poorly.
2697 [19178] dynamic-link: ELF_RTYPE_CLASS_EXTERN_PROTECTED_DATA confuses
2698 prelink
2699 [19181] math: [i386/x86_64] fesetenv (FE_DFL_ENV), fesetenv
2700 (FE_NOMASK_ENV) do not clear SSE exceptions
2701 [19182] malloc: malloc deadlock between ptmalloc_lock_all and
2702 _int_new_arena/reused_arena
2703 [19189] math: [ldbl-128] log1pl (-qNaN) spurious "invalid" exception
2704 [19201] math: dbl-64 remainder incorrect sign of zero result
2705 [19205] math: bits/math-finite.h conditions do not match math.h and
2706 bits/mathcalls.h
2707 [19209] math: bits/math-finite.h wrongly maps ldexp to scalbn
2708 [19211] math: lgamma functions do not set signgam for -ffinite-math-only
2709 for C99-based standards
2710 [19212] libc: features.h not -Wundef clean
2711 [19213] math: [i386/x86_64] log* (1) incorrect zero sign for -ffinite-
2712 math-only
2713 [19214] libc: Family and model identification for AMD CPU's are incorrect.
2714 [19219] libc: GLIBC build fails for ia64 with missing __nearbyintl
2715 [19228] math: [powerpc] nearbyint wrongly clears "inexact", leaves traps
2716 disabled
2717 [19235] math: [powerpc64] lround, lroundf, llround, llroundf spurious
2718 "inexact" exceptions
2719 [19238] math: [powerpc] round, roundf spurious "inexact" for integer
2720 arguments
2721 [19242] libc: strtol incorrect in Turkish locales
2722 [19243] malloc: reused_arena can pick an arena on the free list, leading
2723 to an assertion failure and reference count corruption
2724 [19253] time: tzset() ineffective when temporary TZ did not include DST
2725 rules
2726 [19266] math: strtod ("NAN(I)") incorrect in Turkish locales
2727 [19270] math: [hppa] Shared libm missing __isnanl
2728 [19285] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/mman.h: missing
2729 MAP_HUGETLB and MAP_STACK defines
2730 [19313] nptl: Wrong __cpu_mask for x32
2731 [19347] libc: grantpt: try to force a specific gid even without pt_chown
2732 [19349] math: [ldbl-128ibm] tanhl inaccurate for small arguments
2733 [19350] math: [ldbl-128ibm] sinhl spurious overflows
2734 [19351] math: [ldbl-128ibm] logl inaccurate near 1
2735 [19363] time: x32: times() return value wrongly truncates/sign extends
2736 from 32bit
2737 [19367] dynamic-link: Improve branch prediction on Silvermont
2738 [19369] network: Default domain name not reset by res_ninit when "search"
2739 / "domain" entry is removed from resolv.conf
2740 [19375] math: powerpc: incorrect results for POWER7 logb with negative
2741 subnormals
2742 [19385] localedata: bg_BG: time separator should be colon, not comma
2743 [19408] libc: linux personality syscall wrapper may erroneously return an
2744 error on 32-bit architectures
2745 [19415] libc: dladdr returns wrong names on hppa
2746 [19432] libc: iconv rejects redundant escape sequences in IBM900, IBM903,
2747 IBM905, IBM907, and IBM909
2748 [19439] math: Unix98 isinf and isnan functions conflict with C++11
2749 [19443] build: build failures with -DDEBUG
2750 [19451] build: Make check fails on test-double-vlen2
2751 [19462] libc: Glibc failed to build with -Os
2752 [19465] math: Wrong code with -Os
2753 [19466] time: time/tst-mktime2.c is compiled into an infinite loop with
2754 -Os
2755 [19467] string: Fast_Unaligned_Load needs to be enabled for Excavator core
2756 CPU's.
2757 [19475] libc: Glibc 2.22 doesn't build on sparc [PATCH]
2758 [19486] math: S390: Math tests fail with "Exception Inexact set".
2759 [19529] libc: [ARM]: FAIL: stdlib/tst-makecontext
2760 [19550] libc: [mips] mmap negative offset handling inconsistent with other
2761 architectures
2762 [19590] math: Fail to build shared objects that use libmvec.so functions.
2763 \f
2764 Version 2.22
2765
2766 * The following bugs are resolved with this release:
2767
2768 438, 2981, 4719, 6544, 6792, 11216, 12836, 13028, 13064, 13151, 13152,
2769 14094, 14113, 14292, 14841, 14906, 14958, 15319, 15467, 15790, 15969,
2770 16159, 16339, 16350, 16351, 16352, 16353, 16361, 16512, 16526, 16538,
2771 16559, 16560, 16704, 16783, 16850, 17053, 17090, 17195, 17269, 17293,
2772 17322, 17403, 17475, 17523, 17542, 17569, 17581, 17588, 17596, 17620,
2773 17621, 17628, 17631, 17692, 17711, 17715, 17776, 17779, 17792, 17833,
2774 17836, 17841, 17912, 17916, 17930, 17932, 17944, 17949, 17964, 17965,
2775 17967, 17969, 17977, 17978, 17987, 17991, 17996, 17998, 17999, 18007,
2776 18019, 18020, 18029, 18030, 18032, 18034, 18036, 18038, 18039, 18042,
2777 18043, 18046, 18047, 18049, 18068, 18078, 18080, 18093, 18100, 18104,
2778 18110, 18111, 18116, 18125, 18128, 18134, 18138, 18185, 18196, 18197,
2779 18206, 18210, 18211, 18217, 18219, 18220, 18221, 18234, 18244, 18245,
2780 18247, 18287, 18319, 18324, 18333, 18346, 18371, 18383, 18397, 18400,
2781 18409, 18410, 18412, 18418, 18422, 18434, 18444, 18457, 18468, 18469,
2782 18470, 18479, 18483, 18495, 18496, 18497, 18498, 18502, 18507, 18508,
2783 18512, 18513, 18519, 18520, 18522, 18527, 18528, 18529, 18530, 18532,
2784 18533, 18534, 18536, 18539, 18540, 18542, 18544, 18545, 18546, 18547,
2785 18549, 18553, 18557, 18558, 18569, 18583, 18585, 18586, 18592, 18593,
2786 18594, 18602, 18612, 18613, 18619, 18633, 18635, 18641, 18643, 18648,
2787 18657, 18676, 18694, 18696, 18887.
2788
2789 * Cache information can be queried via sysconf() function on s390 e.g. with
2790 _SC_LEVEL1_ICACHE_SIZE as argument.
2791
2792 * A buffer overflow in gethostbyname_r and related functions performing DNS
2793 requests has been fixed. If the NSS functions were called with a
2794 misaligned buffer, the buffer length change due to pointer alignment was
2795 not taken into account. This could result in application crashes or,
2796 potentially arbitrary code execution, using crafted, but syntactically
2797 valid DNS responses. (CVE-2015-1781)
2798
2799 * The time zone file parser has been made more robust against crafted time
2800 zone files, avoiding heap buffer overflows related to the processing of
2801 the tzh_ttisstdcnt and tzh_ttisgmtcnt fields, and a stack overflow due to
2802 large time zone data files. Overly long time zone specifiers in the TZ
2803 variable no longer result in stack overflows and crashes.
2804
2805 * A powerpc and powerpc64 optimization for TLS, similar to TLS descriptors
2806 for LD and GD on x86 and x86-64, has been implemented. You will need
2807 binutils-2.24 or later to enable this optimization.
2808
2809 * Character encoding and ctype tables were updated to Unicode 7.0.0, using
2810 new generator scripts contributed by Pravin Satpute and Mike FABIAN (Red
2811 Hat). These updates cause user visible changes, such as the fix for bug
2812 17998.
2813
2814 * CVE-2014-8121 The NSS backends shared internal state between the getXXent
2815 and getXXbyYY NSS calls for the same database, causing a denial-of-service
2816 condition in some applications.
2817
2818 * Added vector math library named libmvec with the following vectorized x86_64
2819 implementations: cos, cosf, sin, sinf, sincos, sincosf, log, logf, exp, expf,
2820 pow, powf.
2821 The library can be disabled with --disable-mathvec. Use of the functions is
2822 enabled with -fopenmp -ffast-math starting from -O1 for GCC version >= 4.9.0.
2823 Shared library libmvec.so is linked in as needed when using -lm (no need to
2824 specify -lmvec explicitly for not static builds).
2825 Visit <https://sourceware.org/glibc/wiki/libmvec> for detailed information.
2826
2827 * A new fmemopen implementation has been added with the goal of POSIX
2828 compliance. The new implementation fixes the following long-standing
2829 issues: BZ#6544, BZ#11216, BZ#12836, BZ#13151, BZ#13152, and BZ#14292. The
2830 old implementation is still present for use be by existing binaries.
2831
2832 * The 32-bit sparc sigaction ABI was inadvertently broken in the 2.20 and 2.21
2833 releases. It has been fixed to match 2.19 and older, but binaries built
2834 against 2.20 and 2.21 might need to be recompiled. See BZ#18694.
2835
2836 * Port to Native Client running on ARMv7-A (--host=arm-nacl).
2837 Contributed by Roland McGrath (Google).
2838
2839 * The header <regexp.h> is deprecated, and will be removed in a future
2840 release. Use of this header will trigger a deprecation warning.
2841 Application developers should update their code to use <regex.h> instead.
2842
2843 This header was formerly part of SUS, but was deprecated in 1994 and
2844 removed from the standard in 2001. Also, the glibc implementation
2845 leaks memory. See BZ#18681 for more details.
2846 \f
2847 Version 2.21
2848
2849 * The following bugs are resolved with this release:
2850
2851 6652, 10672, 12674, 12847, 12926, 13862, 14132, 14138, 14171, 14498,
2852 15215, 15378, 15884, 16009, 16418, 16191, 16469, 16576, 16617, 16618,
2853 16619, 16657, 16740, 16857, 17192, 17266, 17273, 17344, 17363, 17370,
2854 17371, 17411, 17460, 17475, 17485, 17501, 17506, 17508, 17522, 17555,
2855 17570, 17571, 17572, 17573, 17574, 17582, 17583, 17584, 17585, 17589,
2856 17594, 17601, 17608, 17616, 17625, 17630, 17633, 17634, 17635, 17647,
2857 17653, 17657, 17658, 17664, 17665, 17668, 17682, 17702, 17717, 17719,
2858 17722, 17723, 17724, 17725, 17732, 17733, 17744, 17745, 17746, 17747,
2859 17748, 17775, 17777, 17780, 17781, 17782, 17791, 17793, 17796, 17797,
2860 17801, 17803, 17806, 17834, 17844, 17848, 17868, 17869, 17870, 17885,
2861 17892.
2862
2863 * CVE-2015-1472 CVE-2015-1473 Under certain conditions wscanf can allocate
2864 too little memory for the to-be-scanned arguments and overflow the
2865 allocated buffer. The implementation now correctly computes the required
2866 buffer size when using malloc, and switches to malloc from alloca as
2867 intended.
2868
2869 * A new semaphore algorithm has been implemented in generic C code for all
2870 machines. Previous custom assembly implementations of semaphore were
2871 difficult to reason about or ensure that they were safe. The new version
2872 of semaphore supports machines with 64-bit or 32-bit atomic operations.
2873 The new semaphore algorithm is used by sem_init, sem_open, sem_post,
2874 sem_wait, sem_timedwait, sem_trywait, and sem_getvalue.
2875
2876 * Port to Altera Nios II has been contributed by Mentor Graphics.
2877
2878 * Optimized strcpy, stpcpy, strncpy, stpncpy, strcmp, and strncmp
2879 implementations for powerpc64/powerpc64le.
2880 Implemented by Adhemerval Zanella (IBM).
2881
2882 * Added support for TSX lock elision of pthread mutexes on powerpc32, powerpc64
2883 and powerpc64le. This may improve lock scaling of existing programs on
2884 HTM capable systems. The lock elision code is only enabled with
2885 --enable-lock-elision=yes. Also, the TSX lock elision implementation for
2886 powerpc will issue a transaction abort on every syscall to avoid side
2887 effects being visible outside transactions.
2888
2889 * Optimized strcpy, stpcpy, strchrnul and strrchr implementations for
2890 AArch64. Contributed by ARM Ltd.
2891
2892 * i386 memcpy functions optimized with SSE2 unaligned load/store.
2893
2894 * CVE-2104-7817 The wordexp function could ignore the WRDE_NOCMD flag
2895 under certain input conditions resulting in the execution of a shell for
2896 command substitution when the application did not request it. The
2897 implementation now checks WRDE_NOCMD immediately before executing the
2898 shell and returns the error WRDE_CMDSUB as expected.
2899
2900 * CVE-2012-3406 printf-style functions could run into a stack overflow when
2901 processing format strings with a large number of format specifiers.
2902
2903 * CVE-2014-9402 The nss_dns implementation of getnetbyname could run into an
2904 infinite loop if the DNS response contained a PTR record of an unexpected
2905 format.
2906
2907 * The minimum GCC version that can be used to build this version of the GNU
2908 C Library is GCC 4.6. Older GCC versions, and non-GNU compilers, can
2909 still be used to compile programs using the GNU C Library.
2910
2911 * The GNU C Library is now built with -Werror by default. This can be
2912 disabled by configuring with --disable-werror.
2913
2914 * New locales: tu_IN, bh_IN, raj_IN, ce_RU.
2915
2916 * The obsolete sigvec function has been removed. This was the original
2917 4.2BSD interface that inspired the POSIX.1 sigaction interface, which
2918 programs have been using instead for about 25 years. Of course, ABI
2919 compatibility for old binaries using sigvec remains intact.
2920
2921 * Merged gettext 0.19.3 into the intl subdirectory. This fixes building
2922 with newer versions of bison.
2923
2924 * Support for MIPS o32 FPXX, FP64A and FP64 ABI Extensions.
2925 The original MIPS o32 hard-float ABI requires an FPU where double-precision
2926 registers overlay two consecutive single-precision registers. MIPS32R2
2927 introduced a new FPU mode (FR=1) where double-precision registers extend the
2928 corresponding single-precision registers which is incompatible with the
2929 o32 hard-float ABI. The MIPS SIMD ASE and the MIPSR6 architecture both
2930 require the use of FR=1 making a transition necessary. New o32 ABI
2931 extensions enable users to migrate over time from the original o32 ABI
2932 through to the updated o32 FP64 ABI. To achieve this the dynamic linker now
2933 tracks the ABI of any loaded object and verifies that new objects are
2934 compatible. Mode transitions will also be requested as required and
2935 unsupportable objects will be rejected. The ABI checks include both soft and
2936 hard float ABIs for o32, n32 and n64.
2937
2938 GCC 5 with GNU binutils 2.25 onwards:
2939 It is strongly recommended that all o32 system libraries are built using the
2940 new o32 FPXX ABI (-mfpxx) to facilitate the transition as this is compatible
2941 with the original and all new o32 ABI extensions. Configure a MIPS GCC
2942 compiler using --with-fp-32=xx to set this by default.
2943 \f
2944 Version 2.20
2945
2946 * The following bugs are resolved with this release:
2947
2948 6804, 9894, 11505, 12994, 13347, 13651, 14308, 14770, 15119, 15132, 15347,
2949 15514, 15698, 15804, 15894, 15946, 16002, 16064, 16095, 16194, 16198,
2950 16275, 16284, 16287, 16315, 16348, 16349, 16354, 16357, 16362, 16447,
2951 16516, 16532, 16539, 16545, 16561, 16562, 16564, 16574, 16599, 16600,
2952 16609, 16610, 16611, 16613, 16619, 16623, 16629, 16632, 16634, 16639,
2953 16642, 16648, 16649, 16670, 16674, 16677, 16680, 16681, 16683, 16689,
2954 16695, 16701, 16706, 16707, 16712, 16713, 16714, 16724, 16731, 16739,
2955 16740, 16743, 16754, 16758, 16759, 16760, 16770, 16786, 16789, 16791,
2956 16796, 16799, 16800, 16815, 16823, 16824, 16831, 16838, 16839, 16849,
2957 16854, 16876, 16877, 16878, 16882, 16885, 16888, 16890, 16892, 16912,
2958 16915, 16916, 16917, 16918, 16922, 16927, 16928, 16932, 16943, 16958,
2959 16965, 16966, 16967, 16977, 16978, 16984, 16990, 16996, 17009, 17022,
2960 17031, 17042, 17048, 17050, 17058, 17061, 17062, 17069, 17075, 17078,
2961 17079, 17084, 17086, 17088, 17092, 17097, 17125, 17135, 17137, 17150,
2962 17153, 17187, 17213, 17259, 17261, 17262, 17263, 17319, 17325, 17354.
2963
2964 * Reverted change of ABI data structures for s390 and s390x:
2965 On s390 and s390x the size of struct ucontext and jmp_buf was increased in
2966 2.19. This change is reverted in 2.20. The introduced 2.19 symbol versions
2967 of getcontext, setjmp, _setjmp, __sigsetjmp, longjmp, _longjmp, siglongjmp
2968 are preserved pointing straight to the same implementation as the old ones.
2969 Given that, new callers will simply provide a too-big buffer to these
2970 functions. Any applications/libraries out there that embed jmp_buf or
2971 ucontext_t in an ABI-relevant data structure that have already been rebuilt
2972 against 2.19 headers will have to rebuilt again. This is necessary in any
2973 case to revert the breakage in their ABI caused by the glibc change.
2974
2975 * Support for file description locks is added to systems running the
2976 Linux kernel. The standard file locking interfaces are extended to
2977 operate on file descriptions, not file descriptors, via the use of
2978 F_OFD_GETLK, F_OFD_SETLK, and F_OFD_SETLKW. File description locks
2979 are associated with an open file instead of a process.
2980
2981 * Optimized strchr implementation for AArch64. Contributed by ARM Ltd.
2982
2983 * The minimum Linux kernel version that this version of the GNU C Library
2984 can be used with is 2.6.32.
2985
2986 * Running the testsuite no longer terminates as soon as a test fails.
2987 Instead, a file tests.sum (xtests.sum from "make xcheck") is generated,
2988 with PASS or FAIL lines for individual tests. A summary of the results is
2989 printed, including a list of failing lists, and "make check" exits with
2990 error status if there were any unexpected failures. "make check
2991 stop-on-test-failure=y" may be used to keep the old behavior.
2992
2993 * The am33 port, which had not worked for several years, has been removed
2994 from ports.
2995
2996 * The _BSD_SOURCE and _SVID_SOURCE feature test macros are no longer
2997 supported; they now act the same as _DEFAULT_SOURCE (but generate a
2998 warning). Except for cases where _BSD_SOURCE enabled BSD interfaces that
2999 conflicted with POSIX (support for which was removed in 2.19), the
3000 interfaces those macros enabled remain available when compiling with
3001 _GNU_SOURCE defined, with _DEFAULT_SOURCE defined, or without any feature
3002 test macros defined.
3003
3004 * Optimized strcmp implementation for ARMv7. Contributed by ARM Ltd.
3005
3006 * Added support for TX lock elision of pthread mutexes on s390 and s390x.
3007 This may improve lock scaling of existing programs on TX capable systems.
3008 The lock elision code is only built with --enable-lock-elision=yes and
3009 then requires a GCC version supporting the TX builtins. With lock elision
3010 default mutexes are elided via __builtin_tbegin, if the cpu supports
3011 transactions. By default lock elision is not enabled and the elision code
3012 is not built.
3013
3014 * CVE-2014-4043 The posix_spawn_file_actions_addopen implementation did not
3015 copy the path argument. This allowed programs to cause posix_spawn to
3016 deference a dangling pointer, or use an unexpected pathname argument if
3017 the string was modified after the posix_spawn_file_actions_addopen
3018 invocation.
3019
3020 * All supported architectures now use the main glibc sysdeps directory
3021 instead of some being in a separate "ports" directory (which was
3022 distributed separately before glibc 2.17).
3023
3024 * The NPTL implementation of POSIX pthreads is no longer an "add-on".
3025 On configurations that support it (all Linux configurations), it's now
3026 used regardless of the --enable-add-ons switch to configure. It is no
3027 longer possible to build such configurations without pthreads support.
3028
3029 * Locale names, including those obtained from environment variables (LANG
3030 and the LC_* variables), are more tightly checked for proper syntax.
3031 setlocale will now fail (with EINVAL) for locale names that are overly
3032 long, contain slashes without starting with a slash, or contain ".." path
3033 components. (CVE-2014-0475) Previously, some valid locale names were
3034 silently replaced with the "C" locale when running in AT_SECURE mode
3035 (e.g., in a SUID program). This is no longer necessary because of the
3036 additional checks.
3037
3038 * On x86-64, the dynamic linker's lazy-binding support is now compatible
3039 with application code using Intel MPX instructions. (With all previous
3040 versions, the MPX register state could be clobbered when making calls
3041 into or out of a shared library.) Note that while the new dynamic
3042 linker is compatible with all known x86 hardware whether or not it
3043 supports Intel MPX, some x86 instruction-set emulators might fail to
3044 handle the new instruction encodings. This is known to affect Valgrind
3045 versions up through 3.9 (but will be fixed in the forthcoming 3.10
3046 release), and might affect other tools that do instruction emulation.
3047
3048 * Support for loadable gconv transliteration modules has been removed.
3049 The support for transliteration modules has been non-functional for
3050 over a decade, and the removal is prompted by security defects. The
3051 normal gconv conversion modules are still supported. Transliteration
3052 with //TRANSLIT is still possible, and the //IGNORE specifier
3053 continues to be supported. (CVE-2014-5119)
3054
3055 * Decoding a crafted input sequence in the character sets IBM933, IBM935,
3056 IBM937, IBM939, IBM1364 could result in an out-of-bounds array read,
3057 resulting a denial-of-service security vulnerability in applications which
3058 use functions related to iconv. (CVE-2014-6040)
3059 \f
3060 Version 2.19
3061
3062 * The following bugs are resolved with this release:
3063
3064 156, 387, 431, 762, 832, 926, 2801, 4772, 6786, 6787, 6807, 6810, 6981,
3065 7003, 9721, 9954, 10253, 10278, 11087, 11157, 11214, 12100, 12486, 12751,
3066 12986, 13028, 13982, 13985, 14029, 14032, 14120, 14143, 14155, 14286,
3067 14547, 14699, 14752, 14782, 14876, 14910, 15004, 15048, 15073, 15089,
3068 15128, 15218, 15268, 15277, 15308, 15362, 15374, 15400, 15425, 15427,
3069 15483, 15522, 15531, 15532, 15593, 15601, 15608, 15609, 15610, 15632,
3070 15640, 15670, 15672, 15680, 15681, 15723, 15734, 15735, 15736, 15748,
3071 15749, 15754, 15760, 15763, 15764, 15797, 15799, 15825, 15843, 15844,
3072 15846, 15847, 15849, 15850, 15855, 15856, 15857, 15859, 15867, 15886,
3073 15887, 15890, 15892, 15893, 15895, 15897, 15901, 15905, 15909, 15915,
3074 15917, 15919, 15921, 15923, 15939, 15941, 15948, 15963, 15966, 15968,
3075 15985, 15988, 15997, 16032, 16034, 16036, 16037, 16038, 16041, 16046,
3076 16055, 16071, 16072, 16074, 16077, 16078, 16103, 16112, 16143, 16144,
3077 16146, 16150, 16151, 16153, 16167, 16169, 16172, 16195, 16214, 16245,
3078 16271, 16274, 16283, 16289, 16293, 16314, 16316, 16330, 16337, 16338,
3079 16356, 16365, 16366, 16369, 16372, 16375, 16379, 16381, 16384, 16385,
3080 16386, 16387, 16390, 16394, 16398, 16400, 16407, 16408, 16414, 16430,
3081 16431, 16453, 16474, 16506, 16510, 16529
3082
3083 * Slovenian translations for glibc messages have been contributed by the
3084 Translation Project's Slovenian team of translators.
3085
3086 * The public headers no longer use __unused nor __block. This change is to
3087 support compiling programs that are derived from BSD sources and use
3088 __unused internally, and to support compiling with Clang's -fblock
3089 extension which uses __block.
3090
3091 * CVE-2012-4412 The strcoll implementation caches indices and rules for
3092 large collation sequences to optimize multiple passes. This cache
3093 computation may overflow for large collation sequences and may cause a
3094 stack or buffer overflow. This is now fixed to use a slower algorithm
3095 which does not use a cache if there is an integer overflow.
3096
3097 * CVE-2012-4424 The strcoll implementation uses malloc to cache indices and
3098 rules for large collation sequences to optimize multiple passes and falls
3099 back to alloca if malloc fails, resulting in a possible stack overflow.
3100 The implementation now falls back to an uncached collation sequence lookup
3101 if malloc fails.
3102
3103 * CVE-2013-4788 The pointer guard used for pointer mangling was not
3104 initialized for static applications resulting in the security feature
3105 being disabled. The pointer guard is now correctly initialized to a
3106 random value for static applications. Existing static applications need
3107 to be recompiled to take advantage of the fix (bug 15754).
3108
3109 * CVE-2013-4237 The readdir_r function could write more than NAME_MAX bytes
3110 to the d_name member of struct dirent, or omit the terminating NUL
3111 character. (Bugzilla #14699).
3112
3113 * CVE-2013-4332 The pvalloc, valloc, memalign, posix_memalign and
3114 aligned_alloc functions could allocate too few bytes or corrupt the
3115 heap when passed very large allocation size values (Bugzilla #15855,
3116 #15856, #15857).
3117
3118 * CVE-2013-4458 Stack overflow in getaddrinfo with large number of results
3119 for AF_INET6 has been fixed (Bugzilla #16072).
3120
3121 * New locales: ak_GH, anp_IN, ar_SS, cmn_TW, hak_TW, lzh_TW, nan_TW, pap_AW,
3122 pap_CW, quz_PE, the_NP.
3123
3124 * Substantially revised locales: gd_GB, ht_HT
3125
3126 * The LC_ADDRESS field was updated to support country_car for almost all
3127 supported locales.
3128
3129 * ISO 1427 definitions were updated.
3130
3131 * ISO 3166 definitions were updated.
3132
3133 * The localedef utility now supports --big-endian and --little-endian
3134 command-line options to generate locales for a different system from that
3135 for which the C library was built.
3136
3137 * Binary locale files now only depend on the endianness of the system for
3138 which they are generated and not on other properties of that system. As a
3139 consequence, binary files generated with new localedef may be incompatible
3140 with old versions of the GNU C Library, and binary files generated with
3141 old localedef may be incompatible with this version of the GNU C Library,
3142 in the following circumstances:
3143
3144 + Locale files may be incompatible on m68k systems.
3145
3146 + Locale archive files (but not separate files for individual locales) may
3147 be incompatible on systems where plain "char" is signed.
3148
3149 * The configure option --disable-versioning has been removed. Builds with
3150 --disable-versioning had not worked for several years.
3151
3152 * ISO 639 definitions were updated for Chiga (cgg) and Chinese (gan, hak, czh,
3153 cjy, lzh, cmn, mnp, cdo, czo, cpx, wuu, hsn, yue).
3154
3155 * SystemTap probes for malloc have been introduced.
3156
3157 * SystemTap probes for slow multiple precision fallback paths of
3158 transcendental functions have been introduced.
3159
3160 * Support for powerpc64le has been added.
3161
3162 * The soft-float powerpc port now supports e500 processors.
3163
3164 * Support for STT_GNU_IFUNC symbols added for ppc32/power4+ and ppc64.
3165
3166 * A new feature test macro _DEFAULT_SOURCE is available to enable the same
3167 set of header declarations that are enabled by default, even when other
3168 feature test macros or compiler options such as -std=c99 would otherwise
3169 disable some of those declarations.
3170
3171 * The _BSD_SOURCE feature test macro no longer enables BSD interfaces that
3172 conflict with POSIX. The libbsd-compat library (which was a dummy library
3173 that did nothing) has also been removed.
3174
3175 * Preliminary documentation about Multi-Thread, Async-Signal and
3176 Async-Cancel Safety has been added.
3177
3178 * Change of ABI data structures for s390 and s390x:
3179 On s390 and s390x the size of struct ucontext and jmp_buf was increased to
3180 allow for future hardware extensions. All library functions that accept or
3181 return these structures were versioned in order to provide backwards
3182 compatibility with existing code. However, not all cases can be handled
3183 with interface versioning. If an application embeds either structure into
3184 another structure and that structure is passed to another compilation unit
3185 which expects the newer larger-sized object then any access to the new
3186 fields will result in undefined behaviour. Similarly any access to fields
3187 that were moved by the enlarging of the structures will cause undefined
3188 behaviour. To fix the undefined behaviour all objects that are part of the
3189 final in-memory image must agree on the size of structures, and this may
3190 require recompilation.
3191 \f
3192 Version 2.18
3193
3194 * The following bugs are resolved with this release:
3195
3196 2546, 2560, 5159, 6809, 7006, 10060, 10062, 10283, 10357, 10686, 11120,
3197 11561, 12310, 12387, 12492, 12515, 12723, 13550, 13889, 13951, 13988,
3198 14142, 14176, 14200, 14256, 14280, 14293, 14317, 14327, 14478, 14496,
3199 14582, 14686, 14812, 14888, 14894, 14907, 14908, 14909, 14920, 14941,
3200 14952, 14964, 14981, 14982, 14985, 14991, 14994, 14996, 15000, 15003,
3201 15006, 15007, 15014, 15020, 15022, 15023, 15036, 15054, 15055, 15062,
3202 15078, 15084, 15085, 15086, 15100, 15160, 15214, 15221, 15232, 15234,
3203 15283, 15285, 15287, 15304, 15305, 15307, 15309, 15327, 15330, 15335,
3204 15336, 15337, 15339, 15342, 15346, 15359, 15361, 15366, 15380, 15381,
3205 15394, 15395, 15405, 15406, 15409, 15416, 15418, 15419, 15423, 15424,
3206 15426, 15429, 15431, 15432, 15441, 15442, 15448, 15465, 15480, 15485,
3207 15488, 15490, 15492, 15493, 15497, 15506, 15529, 15536, 15553, 15577,
3208 15583, 15618, 15627, 15631, 15654, 15655, 15666, 15667, 15674, 15711,
3209 15755, 15759.
3210
3211 * CVE-2013-2207 Incorrectly granting access to another user's pseudo-terminal
3212 has been fixed by disabling the use of pt_chown (Bugzilla #15755).
3213 Distributions can re-enable building and using pt_chown via the new configure
3214 option `--enable-pt_chown'. Enabling the use of pt_chown carries with it
3215 considerable security risks and should only be used if the distribution
3216 understands and accepts the risks.
3217
3218 * CVE-2013-0242 Buffer overrun in regexp matcher has been fixed (Bugzilla
3219 #15078).
3220
3221 * CVE-2013-1914 Stack overflow in getaddrinfo with many results has been
3222 fixed (Bugzilla #15330).
3223
3224 * Add support for calling C++11 thread_local object destructors on thread
3225 and program exit. This needs compiler support for offloading C++11
3226 destructor calls to glibc.
3227
3228 * Improved worst case performance of libm functions with double inputs and
3229 output.
3230
3231 * Support for priority inherited mutexes in pthread condition variables on
3232 non-x86 architectures.
3233
3234 * Port to Xilinx MicroBlaze contributed by David Holsgrove.
3235
3236 * Optimized string functions for AArch64. Implemented by Marcus Shawcroft.
3237
3238 * Optimized string functions for ARM. Implemented by Will Newton and
3239 Richard Henderson.
3240
3241 * Optimized libm functions for SPARC. Implemented by David S. Miller.
3242
3243 * Support for building more of ARM glibc as Thumb-2 code. Implemented by
3244 Richard Henderson.
3245
3246 * Support for building most of MIPS glibc as MIPS16 code. Implemented by
3247 Chung-Lin Tang, Maciej W. Rozycki and Maxim Kuvyrkov.
3248
3249 * Added a benchmark framework to track performance of functions in glibc.
3250
3251 * New <math.h> macro named issignaling to check for a signaling NaN (sNaN).
3252 It is based on draft TS 18661 and currently enabled as a GNU extension.
3253
3254 * On Linux, the clock function now uses the clock_gettime system call
3255 for improved precision, rather than old times system call.
3256
3257 * Added support for version-3 tz format files. This is needed when using
3258 the tz database release 2013e or later, and affects a few unusual cases --
3259 currently only TZ='America/Godthab' for time stamps after 2037.
3260
3261 * Added new API functions pthread_getattr_default_np and
3262 pthread_setattr_default_np to get and set the default pthread
3263 attributes of a process.
3264
3265 * Added support for TSX lock elision for pthread mutexes on i386 and x86-64.
3266 This may improve lock scaling of existing programs on TSX capable systems.
3267 When the --enable-lock-elision=yes parameter is specified at configure
3268 time lock elision will be enabled by default for all PTHREAD_MUTEX_DEFAULT
3269 mutexes.
3270
3271 * Support added for AT_HWCAP2 (to coincide with Linux kernel 3.10
3272 availability). Implemented by Ryan S. Arnold.
3273
3274 * Support added for POWER8 platform. Implemented by Ryan S. Arnold.
3275 \f
3276 Version 2.17
3277
3278 * The following bugs are resolved with this release:
3279
3280 1349, 3439, 3479, 3665, 5044, 5246, 5298, 5400, 6530, 6677, 6778, 6808,
3281 9685, 9914, 10014, 10038, 10114, 10191, 10631, 10873, 11438, 11607, 11638,
3282 11741, 12140, 13013, 13412, 13542, 13601, 13603, 13604, 13629, 13679,
3283 13696, 13698, 13717, 13741, 13759, 13761, 13763, 13881, 13939, 13950,
3284 13952, 13966, 14042, 14047, 14090, 14150, 14151, 14152, 14154, 14157,
3285 14166, 14173, 14195, 14197, 14237, 14246, 14251, 14252, 14283, 14298,
3286 14303, 14307, 14328, 14331, 14336, 14337, 14347, 14349, 14368, 14376,
3287 14417, 14447, 14459, 14476, 14477, 14501, 14505, 14510, 14516, 14518,
3288 14519, 14530, 14532, 14538, 14543, 14544, 14545, 14557, 14562, 14568,
3289 14576, 14579, 14583, 14587, 14595, 14602, 14610, 14621, 14638, 14645,
3290 14648, 14652, 14660, 14661, 14669, 14672, 14683, 14694, 14716, 14719,
3291 14743, 14767, 14783, 14784, 14785, 14793, 14796, 14797, 14801, 14803,
3292 14805, 14807, 14811, 14815, 14821, 14822, 14824, 14828, 14831, 14833,
3293 14835, 14838, 14856, 14863, 14865, 14866, 14868, 14869, 14871, 14872,
3294 14879, 14889, 14893, 14898, 14914.
3295
3296 * Optimization of memcpy for MIPS.
3297
3298 * CVE-2011-4609 svc_run() produces high cpu usage when accept fails with
3299 EMFILE has been fixed (Bugzilla #14889).
3300
3301 * The add-on ports collection is now distributed in the "ports" subdirectory
3302 of the main GNU C Library distribution, rather than separately.
3303
3304 * Port to ARM AArch64 contributed by Linaro.
3305
3306 * Support for STT_GNU_IFUNC symbols added for s390 and s390x.
3307 Optimized versions of memcpy, memset, and memcmp added for System z10 and
3308 zEnterprise z196.
3309 Implemented by Andreas Krebbel.
3310
3311 * The new function secure_getenv allows secure access to the environment,
3312 returning NULL if running in a SUID/SGID process. This function replaces
3313 the internal function __secure_getenv.
3314
3315 * SystemTap static probes have been added into the dynamic linker.
3316 Implemented by Gary Benson.
3317
3318 * Optimizations of string functions strstr, strcasestr and memmem.
3319 Implemented by Maxim Kuvyrkov.
3320
3321 * The minimum Linux kernel version that this version of the GNU C Library
3322 can be used with is 2.6.16.
3323
3324 * Optimizations of string functions memchr, wcschr, wcscpy, and wcsrchr for
3325 powerpc POWER7. Implemented by Will Schmidt.
3326
3327 * New configure option --disable-nscd builds the C library such that it
3328 never attempts to contact the Name Service Caching Daemon (nscd).
3329 New configure option --disable-build-nscd avoids building nscd itself;
3330 this is the default if --disable-nscd is used.
3331
3332 * Improved support for cross-compilation, including cross-testing and
3333 bootstrap builds without a previously built glibc.
3334
3335 * Several testsuite tests are now able to test multiple IFUNC variants of an
3336 interface, rather than just testing the one that would be chooen by
3337 default.
3338
3339 * New configure options --with-bugurl and --with-pkgversion, for
3340 distributors to use to embed their bug-reporting and package version
3341 information in --help and --version output.
3342
3343 * The ttyname and ttyname_r functions on Linux now fall back to searching for
3344 the tty file descriptor in /dev/pts or /dev if /proc is not available. This
3345 allows creation of chroots without the procfs mounted on /proc.
3346
3347 * The `crypt' function now fails if passed salt bytes that violate the
3348 specification for those values. On Linux, the `crypt' function will
3349 consult /proc/sys/crypto/fips_enabled to determine if "FIPS mode" is
3350 enabled, and fail on encrypted strings using the MD5 or DES algorithm
3351 when the mode is enabled.
3352
3353 * The `clock_*' suite of functions (declared in <time.h>) is now available
3354 directly in the main C library. Previously it was necessary to link with
3355 -lrt to use these functions. This change has the effect that a
3356 single-threaded program that uses a function such as `clock_gettime' (and
3357 is not linked with -lrt) will no longer implicitly load the pthreads
3358 library at runtime and so will not suffer the overheads associated with
3359 multi-thread support in other code such as the C++ runtime library.
3360
3361 * New locales: ayc_PE, doi_IN, ia_FR, mni_IN, nhn_MX, niu_NU, niu_NZ,
3362 sat_IN, and szl_PL.
3363 \f
3364 Version 2.16
3365
3366 * The following bugs are resolved with this release:
3367
3368 174, 208, 350, 369, 411, 706, 766, 2074, 2541, 2547, 2548, 2550, 2551,
3369 2552, 2553, 2554, 2562, 2563, 2565, 2566, 2570, 2576, 2636, 2678, 3335,
3370 3440, 3748, 3768, 3866, 3868, 3906, 3976, 3992, 4026, 4108, 4596, 4822,
3371 5077, 5461, 5805, 5993, 6471, 6486, 6578, 6649, 6730, 6770, 6794, 6884,
3372 6890, 6894, 6895, 6907, 6911, 6959, 7064, 9739, 9902, 10110, 10135, 10140,
3373 10153, 10210, 10254, 10346, 10375, 10545, 10716, 10846, 10882, 11174,
3374 11261, 11322, 11365, 11451, 11494, 11521, 11677, 11837, 11959, 12047,
3375 12097, 12193, 12194, 12297, 12298, 12301, 12340, 12354, 12416, 12495,
3376 13058, 13223, 13361, 13525, 13526, 13527, 13528, 13529, 13530, 13531,
3377 13532, 13533, 13547, 13551, 13552, 13553, 13555, 13556, 13559, 13563,
3378 13566, 13576, 13579, 13583, 13592, 13594, 13613, 13618, 13637, 13656,
3379 13658, 13673, 13691, 13695, 13704, 13705, 13706, 13718, 13726, 13738,
3380 13739, 13743, 13750, 13758, 13760, 13761, 13775, 13786, 13787, 13792,
3381 13806, 13824, 13840, 13841, 13844, 13846, 13848, 13851, 13852, 13854,
3382 13871, 13872, 13873, 13879, 13882, 13883, 13884, 13885, 13886, 13892,
3383 13895, 13908, 13910, 13911, 13912, 13913, 13914, 13915, 13916, 13917,
3384 13918, 13919, 13920, 13921, 13922, 13923, 13924, 13926, 13927, 13928,
3385 13938, 13941, 13942, 13954, 13955, 13956, 13963, 13967, 13968, 13970,
3386 13973, 13979, 13983, 13986, 13996, 14012, 14027, 14033, 14034, 14036,
3387 14040, 14043, 14044, 14048, 14049, 14050, 14053, 14055, 14059, 14064,
3388 14075, 14080, 14083, 14103, 14104, 14109, 14112, 14117, 14122, 14123,
3389 14134, 14153, 14183, 14188, 14199, 14210, 14218, 14229, 14241, 14273,
3390 14277, 14278.
3391
3392 * Support for the x32 ABI on x86-64 added. The x32 target is selected by
3393 configuring glibc with:
3394 BUILD_CC='gcc' CC='gcc -mx32' CXX='g++ -mx32'
3395 Visit <http://sites.google.com/site/x32abi/> for more x32 ABI info.
3396 Implemented by H.J. Lu.
3397
3398 * ISO C11 support:
3399
3400 + define static_assert
3401
3402 + do not declare gets
3403
3404 + declare at_quick_exit and quick_exit also for ISO C11
3405
3406 + aligned_alloc. NB: The code is deliberately allows the size parameter
3407 to not be a multiple of the alignment. This is a moronic requirement
3408 in the standard but it is only a requirement on the caller, not the
3409 implementation.
3410
3411 + timespec_get added
3412
3413 + uchar.h support added
3414
3415 + CMPLX, CMPLXF, CMPLXL added
3416
3417 Implemented by Ulrich Drepper.
3418
3419 * Support for the IA-64 has been moved to ports.
3420
3421 * Remove support for anything but ELF binary format
3422
3423 * Checking versions of poll, ppoll added.
3424 Implemented by Ulrich Drepper.
3425
3426 * More generic and 64-bit performance optimizations to math functions.
3427 Implemented by Ulrich Drepper.
3428
3429 * New configure option --enable-obsolete-rpc makes the deprecated RPC
3430 headers and functions available at compile time as they were before
3431 version 2.14. This option will be removed at some time in the future
3432 after the TI-RPC library becomes fully sufficient for the needs of
3433 existing applications.
3434
3435 * Compatibility code for Linux kernel versions before 2.4 has been removed.
3436 Note that glibc is not expected to work with any Linux kernel version
3437 before 2.6.
3438
3439 * New header <sys/auxv.h> and function getauxval allowing easy access to
3440 the AT_* key-value pairs passed from the Linux kernel. The header also
3441 defines the HWCAP_* bits associated with the AT_HWCAP key.
3442
3443 * New locales: mag_IN
3444
3445 * New configure option --enable-systemtap builds SystemTap static probes
3446 into libc for setjmp and longjmp and into libpthread for various operations.
3447 So far the setjmp/longjmp probes and some of the libpthread probes are
3448 provided only for i*86 and x86_64.
3449 Implemented by Roland McGrath and Rayson Ho.
3450
3451 * Optimized expf for x86-32 and x86-64. Implemented by Liubov Dmitrieva.
3452
3453 * More optimized functions for PowerPC. Implemented by Adhemerval Zanella
3454 and Will Schmidt.
3455
3456 * More optimized functions for SPARC. Implemented by David S. Miller.
3457
3458 * Improved support for cross-compilation, especially bootstrap builds
3459 without a previously built glibc.
3460
3461 * Ports for the TILE-Gx and TILEPro families of processors. Contributed by
3462 Chris Metcalf from Tilera.
3463
3464 * Support for the old ARM ABI has been removed from ports. Only the EABI is
3465 now supported for ARM processors.
3466
3467 * The hard-float variant of the ARM EABI now uses /lib/ld-linux-armhf.so.3
3468 as the name of the dynamic linker, to distinguish it from the
3469 /lib/ld-linux.so.3 dynamic linker used for the base EABI.
3470
3471 * Support for CRIS has been removed from ports.
3472
3473 * A new class of installed header has been documented for low-level
3474 platform-specific functionality. PowerPC added the first instance with a
3475 function to provide time base register access. Contributed by Tulio
3476 Magno Quites Machado Filho.
3477
3478 * ix86 configurations now install header files that are consistent with
3479 what x86-64 configurations install. These same header files can be used
3480 for -m32, -m64, or -mx32 builds.
3481 Contributed by H.J. Lu.
3482
3483 * Math library bug fixes. A thorough audit of all open math library bugs was
3484 conducted by Joseph Myers. Significant progress was made on many math
3485 library bugs resulting in more accurate exceptions and function results.
3486 Many thanks to all those that contributed including Andreas Jaeger for his
3487 patch review and work on the x87 trigonometric instruction issues.
3488
3489 * Timezone data is no longer installed. Timezone-related binaries and scripts
3490 will continue to be installed. Users should obtain their timezone data from
3491 their distribution provider or from the tzdata package at
3492 <ftp://munnari.oz.au/pub/>.
3493 \f
3494 Version 2.15
3495
3496 * The following bugs are resolved with this release:
3497
3498 6779, 6783, 9696, 10103, 10709, 11589, 11929, 12403, 12786, 12840, 12847,
3499 12868, 12852, 12874, 12885, 12892, 12906, 12907, 12922, 12935, 12962,
3500 13007, 13021, 13061, 13062, 13067, 13068, 13085, 13088, 13090, 13092,
3501 13096, 13114, 13118, 13123, 13134, 13138, 13147, 13150, 13166, 13179,
3502 13185, 13189, 13192, 13268, 13276, 13282, 13291, 13305, 13328, 13335,
3503 13337, 13344, 13358, 13367, 13413, 13416, 13423, 13439, 13446, 13472,
3504 13484, 13506, 13515, 13523, 13524, 13538, 13540
3505
3506 * New program pldd to list loaded object of a process
3507 Implemented by Ulrich Drepper.
3508
3509 * Add nss_db support back to glibc. No more dependency on Berkeley db
3510 and support for initgroups lookups.
3511 Implemented by Ulrich Drepper.
3512
3513 * Optimized strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-32.
3514 Contributed by HJ Lu.
3515
3516 * Improved strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-64.
3517 Contributed by HJ Lu.
3518
3519 * Optimized strcat, strncat on x86-64 and optimized wcscmp, wcslen, strnlen
3520 on x86-32 and x86-64.
3521 Contributed by Liubov Dmitrieva.
3522
3523 * Optimized strchr and strrchr for SSE on x86-32.
3524 Contributed by Liubov Dmitrieva.
3525
3526 * Optimized memchr, memrchr, rawmemchr, memcmp, wmemcmp, wcschr, wcscpy
3527 for x86-64 and x86-32.
3528 Contributed by Liubov Dmitrieva.
3529
3530 * New interfaces: scandirat, scandirat64
3531 Implemented by Ulrich Drepper.
3532
3533 * Checking versions of FD_SET, FD_CLR, and FD_ISSET added.
3534 Implemented by Ulrich Drepper.
3535
3536 * nscd now also caches the netgroup database.
3537 Implemented by Ulrich Drepper.
3538
3539 * Integrate libm with gcc's -ffinite-math-only option.
3540 Implemented by Ulrich Drepper.
3541
3542 * Lots of generic, 64-bit, and x86-64-specific performance optimizations
3543 to math functions. Implemented by Ulrich Drepper.
3544
3545 * Optimized strcasecmp and strncasecmp for AVX on x86-64.
3546 Implemented by Ulrich Drepper.
3547
3548 * New Linux interfaces: process_vm_readv, process_vm_writev
3549
3550 * Optimized strcasecmp and strncasecmp for SSSE3 and SSE4.2 on x86-32.
3551 Implemented by Ulrich Drepper.
3552
3553 * Optimized nearbyint and strcasecmp for PPC.
3554 Implemented by Adhemerval Zanella.
3555
3556 * New locales: bho_IN, unm_US, es_CU, ta_LK
3557 \f
3558 Version 2.14
3559
3560 * The following bugs are resolved with this release:
3561
3562 386, 6420, 7101, 9730, 9732, 9809, 10138, 10149, 10157, 11099, 11257,
3563 11258, 11487, 11532, 11558, 11578, 11634, 11653, 11668, 11697, 11724,
3564 11781, 11799, 11820, 11837, 11857, 11884, 11892, 11895, 11901, 11945,
3565 11947, 11952, 11987, 12052, 12083, 12158, 12178, 12200, 12346, 12350,
3566 12393, 12420, 12432, 12445, 12449, 12453, 12454, 12460, 12469, 12489,
3567 12509, 12510, 12511, 12518, 12527, 12541, 12545, 12551, 12582, 12583,
3568 12587, 12597, 12601, 12611, 12625, 12626, 12631, 12650, 12653, 12655,
3569 12660, 12671, 12681, 12684, 12685, 12711, 12713, 12714, 12717, 12723,
3570 12724, 12734, 12738, 12746, 12766, 12775, 12777, 12782, 12788, 12792,
3571 12795, 12811, 12813, 12814, 12841
3572
3573 * The RPC implementation in libc is obsoleted. Old programs keep working
3574 but new programs cannot be linked with the routines in libc anymore.
3575 Programs in need of RPC functionality must be linked against TI-RPC.
3576 The TI-RPC implementation is IPv6 enabled and there are other benefits.
3577
3578 Visible changes of this change include (obviously) the inability to link
3579 programs using RPC functions without referencing the TI-RPC library and the
3580 removal of the RPC headers from the glibc headers.
3581 Implemented by Ulrich Drepper.
3582
3583 * New Linux interfaces: clock_adjtime, name_to_handle_at, open_by_handle_at,
3584 syncfs, setns, sendmmsg
3585
3586 * New locales: os_RU, bem_ZA, en_ZA, ff_SN, sw_KE, sw_TZ, lb_LU, wae_CH,
3587 yue_HK, lij_IT, mhr_RU
3588
3589 * New encodings: CP770, CP771, CP772, CP773, CP774
3590
3591 * New program sotruss to trace calls through PLTs
3592 Implemented by Ulrich Drepper.
3593
3594 * The malloc hook implementation is marked deprecated and will be removed
3595 from the default implementation in the next version. The design never
3596 worked ever since the introduction of threads. Even programs which do
3597 not create threads themselves can use multiple threads created internally.
3598 \f
3599 Version 2.13
3600
3601 * The following bugs are resolved with this release:
3602
3603 3268, 6812, 7066, 10085, 10484, 10563, 10851, 11149, 11155, 11577, 11611,
3604 11640, 11655, 11701, 11840, 11856, 11883, 11903, 11904, 11968, 11979,
3605 12005, 12037, 12067, 12077, 12078, 12092, 12093, 12107, 12108, 12113,
3606 12140, 12159, 12167, 12191, 12194, 12201, 12204, 12205, 12207, 12348,
3607 12378, 12394, 12397
3608
3609 * New Linux interfaces: prlimit, prlimit64, fanotify_init, fanotify_mark
3610
3611 * POWER7 optimizations: memset, memcmp, strncmp
3612
3613 * New optimized string functions for x86-64: strnlen (SSE2),
3614 strcasecmp (SSE2, SSSE3, SSE4.2), strncasecmp (SSE2, SSSE3, SSE4.2)
3615 Implemented by Ulrich Drepper.
3616 \f
3617 Version 2.12
3618
3619 * The following bugs are resolved with this release:
3620
3621 3662, 4457, 5553, 10162, 10401, 10414, 10554, 10824, 10864, 10915,
3622 10918, 10936, 10939, 10958, 10968, 10969, 10972, 10992, 11000, 11001,
3623 11007, 11010, 11027, 11039, 11040, 11041, 11043, 11046, 11056, 11070,
3624 11093, 11115, 11125, 11126, 11127, 11134, 11141, 11149, 11183, 11184,
3625 11185, 11186, 11187, 11188, 11189, 11190, 11191, 11192, 11193, 11194,
3626 11200, 11230, 11235, 11242, 11254, 11258, 11271, 11272, 11276, 11279,
3627 11287, 11292, 11319, 11332, 11333, 11387, 11389, 11390, 11394, 11397,
3628 11410, 11438, 11449, 11470, 11471, 11520, 11537, 11538, 11571
3629
3630 * New interfaces: pthread_getname_np, pthread_setname_np
3631
3632 * New Linux interface: recvmmsg
3633
3634 * STT_GNU_IFUNC implemented for Sparc by David Miller.
3635
3636 * The dynamic linker now recognizes supported ABI versions from the
3637 EI_ABIVERSION field in the ELF header.
3638 Implemented by Ulrich Drepper.
3639
3640 * New NIS mode selector ADJUNCT_AS_SHADOW. The passwd.adjunct.byname table
3641 will not be used to fill in password fields in the passwd.byname replies.
3642 Instead it is used to synthesize the shadow.byname table, should it be
3643 missing. This is a useful mode in some installations involving Solaris.
3644 Implemented by Ulrich Drepper.
3645
3646 * New locales: kok_IN, sq_MK, cv_RU
3647 \f
3648 Version 2.11
3649
3650 * The following bugs are resolved with this release:
3651
3652 3270, 7094, 9924, 9986, 10011, 10085, 10107, 10159, 10162, 10166, 10183,
3653 10195, 10196, 10200, 10203, 10207, 10211, 10213, 10217, 10221, 10229,
3654 10262, 10286, 10312, 10315, 10319, 10349, 10360, 10391, 10402, 10416,
3655 10418, 10422, 10425, 10446, 10448, 10452, 10540, 10553, 10560, 10564,
3656 10609, 10643, 10692, 10713, 10717, 10730, 10731, 10742, 10780, 10784,
3657 10789, 10817, 10823, 10840, 10847
3658
3659 * New interfaces: execvpe, pthread_sigqueue, mkstemps, mkstemps64, mkostemps,
3660 mkostemps64
3661 Implemented by Ulrich Drepper.
3662
3663 * Checking version of longjmp added that fails if an uninitialized stack
3664 frame would be created. Implemented by Ulrich Drepper.
3665
3666 * STT_GNU_IFUNC is now supported in static executables.
3667 Implemented by H.J. Lu.
3668
3669 * STT_GNU_IFUNC implemented for PPC by Alan Modra.
3670
3671 * New optimized string functions for x86-64: strstr, strcasestr, memcmp,
3672 strcspn, strpbrk, strspn, strcpy, stpcpy, strncpy, strcmp (SSE2, SSE4.2),
3673 strncmp (SSE2, SSE4.2), strchr (SSE4.2), strrchr (SSE4.2).
3674 Contributed by H.J. Lu.
3675
3676 strlen, rawmemchr, strcmp (SSSE3), strncmp (SSSE3).
3677 Implemented by Ulrich Drepper.
3678
3679 * New optimized string functions for x86: strlen, strcspn, strspn, strpbrk,
3680 strstr, strcasestr.
3681 Contributed by H.J. Lu.
3682
3683 * Support for fma instruction in AVX on x86-64.
3684 Implemented by H.J. Lu and Ulrich Drepper.
3685
3686 * AVX support in x86-64 auditing support in ld.so.
3687 Implemented by H.J. Lu.
3688
3689 * STB_GNU_UNIQUE support added.
3690 Implemented by Ulrich Drepper.
3691
3692 * Implement second fallback mode for optimized DNS lookup for even more
3693 broken environments. If two requests from the same port are not
3694 handled correctly close the socket and open a new one before sending
3695 the second request. The 'single-request-reopen' option in /etc/resolv.conf
3696 can be used to select this mode right away, instead of rediscovering the
3697 necessity is every process again.
3698 Implemented by Ulrich Drepper.
3699
3700 * New resolver flag RES_USE_DNSSEC to enable use of verified lookup.
3701 Implemented by Adam Tkac.
3702
3703 * Optimized iconv conversions for S390x.
3704 Implemented by Andreas Krebbel.
3705
3706 * Using condvars with PI mutexes is now more efficient due to kernel
3707 support for requeueing to PI futexes. NPTL support added for x86-64.
3708 Implemented by Ulrich Drepper.
3709
3710 * New locales: ps_AF, my_MM
3711 \f
3712 Version 2.10
3713
3714 * The following bugs are resolved with this release:
3715
3716 697, 5381, 5807, 6411, 6545, 7040, 7058, 7067, 7080, 7083, 7095, 7098,
3717 9677, 9694, 9697, 9701, 9704, 9705, 9706, 9720, 9726, 9730, 9733, 9736,
3718 9741, 9750, 9753, 9759, 9781, 9791, 9793, 9823, 9833, 9844, 9880, 9881,
3719 9891, 9893, 9895, 9913, 9916, 9920, 9948, 9955, 9957, 9963, 9985, 10052,
3720 10069, 10086, 10087, 10090, 10092, 10093, 10100, 10118, 10128
3721
3722 * New interfaces: preadv, preadv64, pwritev, pwritev64, malloc_info
3723 Implemented by Ulrich Drepper.
3724
3725 * New Linux interfaces: accept4, fallocate, fallocate64.
3726 Implemented by Ulrich Drepper.
3727
3728 * Correct declarations of string function when used in C++ code. This
3729 could lead to compile errors for invalid C++ code.
3730
3731 * XPG7/POSIX 2008 compilation environment. Many GNU-specific functions are
3732 now in POSIX.
3733
3734 * New POSIX 2008 interface: psiginfo
3735 Implemented by Ulrich Drepper.
3736
3737 * New ISO C++1x interfaces: quick_exit, at_quick_exit
3738 Implemented by Ulrich Drepper.
3739
3740 * Support for selecting between multiple function definitions at runtime
3741 using STT_GNU_IFUNC symbols. Implemented by Ulrich Drepper.
3742
3743 * The libcrypt library can now use the hash function implementations in
3744 NSS. Implemented by Ulrich Drepper.
3745
3746 * The malloc implementation can be compiled to be less memory efficient
3747 but higher performing in multi-threaded programs.
3748 Implemented by Ulrich Drepper.
3749
3750 * New locales: nan_TW@latin, ks_IN
3751
3752 * Faster strlen, strchr, strchrnul, strrchr, memchr, and rawmemchr for x86-64.
3753 Implemented by Ulrich Drepper.
3754
3755 * Extended printf hook support. It is possible to use user-defined types
3756 and extend existing format specifiers.
3757 Implemented by Ulrich Drepper.
3758
3759 * Handling for group shadow files has been added.
3760 Implemented by Ulrich Drepper.
3761
3762 * DNS IPv4-IPv6 parallel lookup now deals better with broken DNS
3763 servers (the case, e.g., for some people using the built-in DNS
3764 server in ADSL modems/routers). There is a once-per-process timeout
3765 in case of a broken server. To avoid it, users can run nscd or put
3766 'options single-request' in /etc/resolv.conf.
3767 Implemented by Ulrich Drepper.
3768 \f
3769 Version 2.9
3770
3771 * The following bugs are resolved with this release:
3772
3773 3406, 5209, 5210, 5381, 5794, 5814, 5911, 6428, 6438, 6442, 6447, 6461,
3774 6472, 6505, 6544, 6589, 6612, 6634, 6653, 6654, 6657, 6698, 6712, 6713,
3775 6719, 6723, 6724, 6740, 6763, 6771, 6790, 6791, 6817, 6824, 6839, 6843,
3776 6845, 6856, 6857, 6860, 6867, 6875, 6919, 6920, 6942, 6947, 6955, 6966,
3777 6968, 6974, 6980, 6995, 7008, 7009, 7029
3778
3779 * Unified lookup for getaddrinfo: IPv4 and IPv6 addresses are now looked
3780 up at the same time. Implemented by Ulrich Drepper.
3781
3782 * TLS descriptors for LD and GD on x86 and x86-64.
3783 Implemented by Alexandre Oliva.
3784
3785 * getaddrinfo now handles DCCP and UDPlite.
3786 Implemented by Ulrich Drepper.
3787
3788 * New fixed-size conversion macros: htobe16, htole16, be16toh, le16toh,
3789 htobe32, htole32, be32toh, le32toh, htobe64, htole64, be64toh, le64toh.
3790 Implemented by Ulrich Drepper.
3791
3792 * New implementation of memmem, strstr, and strcasestr which is O(n).
3793 Implemented by Eric Blake.
3794
3795 * New Linux interfaces: inotify_init1, dup3, epoll_create1, pipe2
3796
3797 * Implement "e" option for popen to open file descriptor with the
3798 close-on-exec flag set. Implemented by Ulrich Drepper.
3799
3800 * Implement "b" mode for fmemopen. In this mode writes writes don't
3801 implicitly add a NUL byte and seeks from the end of the buffer really
3802 use the buffer end, not the string length as the basis.
3803 Implemented by Ulrich Drepper.
3804
3805 * Many functions, exported and internal, now atomically set the close-on-exec
3806 flag when run on a sufficiently new kernel. Implemented by Ulrich Drepper.
3807
3808 * Sorting rules for some Indian languages (Kannada, Gurumukhi, Telugu,
3809 Sinhala)
3810 Implemented by Pravin Satpute.
3811
3812 * New locales: sd_IN, sd_IN@devanagari, ks_IN@devanagari, ht_HT, en_AG, nl_AW.
3813 \f
3814 Version 2.8
3815
3816 * The following bugs are resolved with this release:
3817
3818 2549, 3112, 4314, 4407, 4997, 5012, 5040, 5112, 5204, 5208, 5209, 5220,
3819 5222, 5224, 5225, 5237, 5238, 5240, 5245, 5277, 5346, 5375, 5378, 5382,
3820 5424, 5427, 5428, 5435, 5436, 5439, 5441, 5442, 5443, 5451, 5452, 5454,
3821 5455, 5463, 5464, 5465, 5475, 5477, 5478, 5520, 5541, 5545, 5553, 5600,
3822 5602, 5607, 5614, 5627, 5628, 5736, 5737, 5741, 5753, 5760, 5762, 5768,
3823 5774, 5777, 5778, 5779, 5786, 5790, 5818, 5854, 5857, 5903, 5939, 5979,
3824 5983, 5995, 5998, 6004, 6007, 6020, 6021, 6024, 6040, 6042
3825
3826 * New locales: bo_CN, bo_IN, shs_CA.
3827
3828 * New encoding: HP-ROMAN9, HP-GREEK8, HP-THAI8, HP-TURKISH8.
3829
3830 * Sorting rules for some Indian languages (Devanagari and Gujarati).
3831 Implemented by Pravin Satpute.
3832
3833 * IPV6 addresses in /etc/resolv.conf can now have a scope ID
3834
3835 * nscd caches now all timeouts for DNS entries
3836 Implemented by Ulrich Drepper.
3837
3838 * nscd is more efficient and wakes up less often.
3839 Implemented by Ulrich Drepper.
3840
3841 * More checking functions: asprintf, dprintf, obstack_printf, vasprintf,
3842 vdprintf, and obstack_vprintf.
3843 Implemented by Jakub Jelinek.
3844
3845 * Faster memset for x86-64.
3846 Implemented by Harsha Jagasia and H.J. Lu.
3847
3848 * Faster memcpy on x86.
3849 Implemented by Ulrich Drepper.
3850
3851 * ARG_MAX is not anymore constant on Linux. Use sysconf(_SC_ARG_MAX).
3852 Implemented by Ulrich Drepper.
3853
3854 * Faster sqrt and sqrtf implemention for some PPC variants.
3855 Implemented by Steven Munroe.
3856 \f
3857 Version 2.7
3858
3859 * The following bugs are resolved with this release:
3860
3861 4125, 4126, 4178, 4359, 4407, 4512, 4514, 4525, 4554, 4556, 4557, 4566,
3862 4582, 4586, 4588, 4599, 4610, 4647, 4702, 4705, 4726, 4745, 4772, 4773,
3863 4775, 4776, 4792, 4813, 4814, 4816, 4833, 4858, 4860, 4896, 4905, 4925,
3864 4936, 4937, 4938, 4941, 4946, 4963, 4972, 5010, 5028, 5043, 5058, 5063,
3865 5071, 5103, 5104, 5112, 5113, 5184, 5186
3866
3867 * More checking functions: fread, fread_unlocked, open*, mq_open.
3868 Implemented by Jakub Jelinek and Ulrich Drepper.
3869
3870 * Extend fortification to C++. Implemented by Jakub Jelinek.
3871
3872 * Implement 'm' modifier for scanf. Add stricter C99/SUS compliance
3873 by not recognizing 'a' as a modifier when those specs are requested.
3874 Implemented by Jakub Jelinek.
3875
3876 * PPC optimizations to math and string functions.
3877 Implemented by Steven Munroe.
3878
3879 * New interfaces: mkostemp, mkostemp64. Like mkstemp* but allow additional
3880 options to be passed. Implemented by Ulrich Drepper.
3881
3882 * More CPU set manipulation functions. Implemented by Ulrich Drepper.
3883
3884 * New Linux interfaces: signalfd, eventfd, eventfd_read, and eventfd_write.
3885 Implemented by Ulrich Drepper.
3886
3887 * Handle private futexes in the NPTL implementation.
3888 Implemented by Jakub Jelinek and Ulrich Drepper.
3889
3890 * Add support for O_CLOEXEC. Implement in Hurd. Use throughout libc.
3891 Implemented by Roland McGrath and Ulrich Drepper.
3892
3893 * Linux/x86-64 vDSO support. Implemented by Ulrich Drepper.
3894
3895 * SHA-256 and SHA-512 based password encryption.
3896 Implemented by Ulrich Drepper.
3897
3898 * New locales: ber_DZ, ber_MA, en_NG, fil_PH, fur_IT, fy_DE, ha_NG, ig_NG,
3899 ik_CA, iu_CA, li_BE, li_NL, nds_DE, nds_NL, pap_AN, sc_IT, tk_TM, ug_CN,
3900 yo_NG.
3901
3902 + New iconv modules: MAC-CENTRALEUROPE, ISO-8859-9E, KOI8-RU.
3903 Implemented by Ulrich Drepper.
3904 \f
3905 Version 2.6
3906
3907 * The following bugs are resolved with this release:
3908
3909 3156, 3213, 3285, 3291, 3306, 3313, 3320, 3322, 3325, 3326, 3334, 3348,
3910 3352, 3363, 3369, 3426, 3427, 3429, 3451, 3458, 3465, 3480, 3483, 3493,
3911 3514, 3515, 3559, 3632, 3664, 3673, 3674, 3745, 3747, 3818, 3840, 3842,
3912 3851, 3855, 3884, 3885, 3902, 3905, 3919, 3944, 3954, 3955, 3957, 3991,
3913 3995, 3996, 4024, 4040, 4069, 4070, 4074, 4076, 4096, 4101, 4102, 4114,
3914 4130, 4131, 4181, 4276, 4306, 4342, 4344, 4349, 4364, 4368, 4381, 4392,
3915 4403, 4405, 4406, 4411, 4438, 4439, 4455, 4456, 4465, 4512, 4514, 4586,
3916 4702, 4858
3917
3918 * New Linux interfaces: epoll_pwait, sched_getcpu.
3919
3920 * New generic interfaces: strerror_l.
3921
3922 * nscd can now cache the services database. Implemented by Ulrich Drepper.
3923 \f
3924 Version 2.5
3925
3926 * The following bugs are resolved with this release:
3927
3928 39, 68, 192, 352, 388, 531, 935, 1006, 1201, 1203, 1386, 1782, 1783, 1784,
3929 1785, 1786, 1787, 1951, 1974, 1996, 2072, 2098, 2125, 2134, 2158, 2182,
3930 2349, 2376, 2386, 2389, 2415, 2418, 2420, 2423, 2450, 2451, 2466, 2472,
3931 2473, 2477, 2498, 2499, 2501, 2502, 2505, 2507, 2508, 2509, 2517, 2526,
3932 2569, 2571, 2592, 2611, 2625, 2632, 2644, 2662, 2680, 2683, 2684, 2693,
3933 2695, 2703, 2734, 2751, 2764, 2766, 2775, 2792, 2821, 2823, 2832, 2841,
3934 2843, 2883, 2892, 2908, 2914, 2926, 2961, 2978, 2980, 2987, 2997, 2998,
3935 3013, 3018, 3034, 3040, 3044, 3123, 3124, 3137, 3138, 3143, 3155, 3189,
3936 3225, 3251, 3252, 3253, 3273
3937
3938 * For Linux, the sorting of addresses returned by getaddrinfo now also
3939 handles rules 3, 4, and 7 from RFC 3484. I.e., all rules are handled.
3940 Implemented by Ulrich Drepper.
3941
3942 * Allow system admin to configure getaddrinfo with the /etc/gai.conf file.
3943 Implemented by Ulrich Drepper.
3944
3945 * New Linux interfaces: splice, tee, sync_file_range, vmsplice.
3946
3947 * New iconv module for MIK. Contributed by Alexander Shopov.
3948
3949 * For sites with broken group and/or passwd database, the auto-propagate
3950 option of nscd can prevent creating ID lookup entries from the results
3951 of a name lookup and vice versa. This usually is no problem but some
3952 site might have problems with the default behavior.
3953 Implemented by Ulrich Drepper.
3954
3955 * Iterating over entire database in NIS can be slow. With the
3956 SETENT_BATCH_READ option in /etc/default/nss a system admin can decide
3957 to trade time for memory. The entire database will be read at once.
3958 Implemented by Ulrich Drepper.
3959
3960 * The interfaces introduced in RFC 3542 have been implemented by
3961 Ulrich Drepper.
3962
3963 * Support for the new ELF hash table format was added by Ulrich Drepper.
3964
3965 * Support for priority inheritance mutexes added by Jakub Jelinek and
3966 Ulrich Drepper.
3967
3968 * Support for priority protected mutexes added by Jakub Jelinek.
3969
3970 * New locales: nr_ZA, pa_PK, ca_AD, ca_FR, ca_IT, el_CY, tr_CY, as_IN,
3971 or_IN, csb_PL, fy_NL, sr_ME.
3972 \f
3973 Version 2.4
3974
3975 * More overflow detection functions.
3976
3977 * New iconv converters for IBM1025, IBM1122, IBM1137, IBM1153,
3978 IBM1154, IBM1155, IBM1156, IBM1157, and IBM1158 contributed by Jiro Sekiba.
3979
3980 More converters for IBM803, IBM901, IBM902, IBM921, IBM1008, IBM1097,
3981 IBM1112, IBM1123, IBM1130, IBM1140, IBM1141, IBM1142, IBM1143, IBM1144,
3982 IBM1145, IBM1146, IBM1147, IBM1148, IBM1149, IBM1166, IBM1167, IBM4517,
3983 IBM4899, IBM4909, IBM4971, IBM5347, IBM9030, IBM9066, IBM9448, IBM12712,
3984 IBM16804, IBM1364, IBM1371, IBM1388, IBM1390, and IBM1399 contributed
3985 by Masahide Washizawa.
3986
3987 * It is now possible to install audit libraries for the dynamic linker using
3988 LD_AUDIT. Implemented by Ulrich Drepper.
3989
3990 * The LinuxThreads add-on, providing pthreads on Linux 2.4 kernels, is no
3991 longer supported. The new NPTL implementation requires Linux 2.6 kernels.
3992 For a libc and libpthread that works well on Linux 2.4 kernels, we
3993 recommend using the stable 2.3 branch.
3994
3995 * The new function open_wmemstream acts like open_memstream,
3996 but uses a wchar_t wide character buffer.
3997
3998 * The new function ppoll is an improved version of poll, similar to pselect.
3999
4000 * New interfaces: fdopendir, openat, fstatat, fchmodat, fchownat,
4001 futimesat, faccessat, mkdirat, mkfifoat, mknodat,
4002 renameat, unlinkat, linkat, symlinkat, readlinkat.
4003
4004 * New Linux kernel system calls: unshare,
4005 inotify_init, inotify_add_watch, inotify_rm_watch.
4006
4007 * The euidaccess function is now also known by the alias eaccess,
4008 for compatibility with some other systems.
4009
4010 * Timezone data updated to 2006b version.
4011 \f
4012 Version 2.3.6
4013
4014 * The following bugs are resolved with this release:
4015
4016 38, 253, 549, 622, 653, 721, 758, 851, 877, 915, 934, 955, 961,
4017 1016, 1037, 1076, 1079, 1080, 1081, 1082, 1083, 1084, 1085, 1086,
4018 1087, 1088, 1090, 1091, 1092, 1093, 1094, 1095, 1096, 1097, 1098,
4019 1099, 1100, 1101, 1102, 1103, 1104, 1105, 1106, 1107, 1108, 1109,
4020 1110, 1111, 1112, 1113, 1125, 1137, 1138, 1249, 1250, 1251, 1252,
4021 1253, 1254, 1350, 1358, 1394, 1438, 1498, 1534
4022
4023 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
4024
4025 * As of this release, GCC 4 can be used to compile the C Library.
4026
4027 * Timezone data updated to 2005m version.
4028 \f
4029 Version 2.3.5
4030
4031 * The following bugs are resolved with this release:
4032
4033 284, 592, 602, 626, 633, 640, 650, 661, 671, 681, 693, 700, 710, 719,
4034 722, 723, 725, 726, 727, 728, 729, 730, 731, 732, 733, 734, 735, 736,
4035 737, 738, 739, 740, 741, 742, 743, 744, 745, 765, 767, 768, 769, 776,
4036 777, 787, 821, 822, 823, 825
4037
4038 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
4039 \f
4040 Version 2.3.4
4041
4042 * Support for RFC 3678. Real implementations exist only for Linux so far.
4043 Implemented by Ulrich Drepper.
4044
4045 * nscd can now cache entries persistently. Expiring entries are reloaded.
4046 For speedups the cache can be shared in memory with client processes.
4047 Implemented by Ulrich Drepper.
4048
4049 * nscd can now perform SELinux checks.
4050 Implemented by Matthew Rickard <mjricka@epoch.ncsc.mil>.
4051
4052 * getaddrinfo queries are now cached. Canonical name lookup is performed
4053 efficiently.
4054 Implemented by Ulrich Drepper.
4055
4056 * The nothrow function attribute is used when headers are used by gcc when
4057 compiling C code. This can avoid significant amounts of exception
4058 handling data.
4059
4060 * The malloc functions perform more error checking and are stricter when
4061 it comes to reacting on errors. The default action is to terminate
4062 the process after showing an error message. Implemented by Ulrich Drepper.
4063
4064 * Reverse lookups of IPv6 addresses does not use bit string or .ip6.int
4065 lookups anymore unless explicitly requested. Implemented by Ulrich Drepper.
4066
4067 * Namespaces in ld.so are implemented. DSOs can be loaded in separate
4068 namespaces using the new function dlmopen(). This feature is of course,
4069 like most other dynamic loading functionality, not available in statically
4070 linked applications. Implemented by Ulrich Drepper.
4071
4072 * Low-overhead boundary checking variants of string and some stdio functions
4073 were added. These are to be used in conjunction with a gcc patch by
4074 Jakub Jelinek which adds calls to these functions if possible.
4075 Implemented by Jakub Jelinek and Ulrich Drepper.
4076
4077 * Old code for several operating systems and machine architectures that
4078 have not been in working condition in a long time have been removed from
4079 the main source tree maintained by the GNU C Library's maintainers.
4080 These files are now reside in the separate `ports' source module
4081 that is usable as an add-on when building the library.
4082 \f
4083 Version 2.3.3
4084
4085 * New functions `dladdr1' and `dlinfo' in <dlfcn.h> provide more ways to
4086 interrogate the dynamic linker, compatible with the Solaris interface.
4087
4088 * ELF thread-local storage support (TLS) now works on PowerPC and PowerPC64;
4089 implemented by Paul Mackerras, Steven Munroe, and Roland McGrath.
4090
4091 * getifaddrs now uses the netlink interface on Linux to get its information.
4092 Implemented by Thorsten Kukuk.
4093
4094 * getaddrinfo now implements AI_V4MAPPED, AI_ALL, and AI_ADDRCONF.
4095 Implemented by Ulrich Drepper.
4096
4097 * support for non-executable stacks on x86 has been added. Changes mostly
4098 by Roland McGrath.
4099
4100 * regex is now much faster for multibyte locales. Changes by Jakub Jelinek
4101 and Ulrich Drepper.
4102
4103 * getaddrinfo now performs destination address selection according to
4104 RFC 3484.
4105 \f
4106 Version 2.3.2
4107
4108 * Thread-safe interfaces for many functions that access locale data
4109 were added in version 2.3, but these features were omitted from NEWS.
4110 Many functions have variants with an `_l' suffix that take a `locale_t'
4111 object as a parameter rather than consulting the current locale.
4112 The new functions `newlocale', `duplocale', and `freelocale' in <locale.h>
4113 create and maintain `locale_t' objects. Additionally, the new function
4114 `uselocale' sets "the current locale" (as used by functions not so
4115 parameterized) set for an individual thread. These features were added
4116 in version 2.3, implemented by Ulrich Drepper and Roland McGrath.
4117
4118 * The functions getresuid, getresgid, setresuid, and setresgid, which
4119 have long been available on Linux, are now declared in <unistd.h>
4120 and are now also available on the Hurd.
4121
4122 * ELF thread-local storage support (TLS) now works on x86-64.
4123
4124 * The new dynamic string token $LIB is expanded in shared library names.
4125 This normally expands to lib, but on some 64-bit platforms to lib64 instead.
4126
4127 * Aldy Hernandez contributed complete software floating point support for
4128 PowerPC machines with no FPU.
4129
4130 * fexecve is implemented on Linux.
4131
4132 * The `btowc' function should work at least twice as fast due to
4133 specialized callbacks in the iconv modules. Implemented by Bruno Haible.
4134
4135 * With approriate thread add-ons cancelable functions are now implemented
4136 in libc.so as well. No need to call the function in libpthread. This
4137 change allowed to finally disable the incorrect and expensive handling
4138 of weak definition in ld.so.
4139
4140 * Yet more PLT entries in libc.so have been removed. We finally arrived
4141 at the bare minimum. Startup times improved appropriately.
4142
4143 * Support for the new Linux/x86 system call interface was added. The
4144 AT_SYSINFO auxiliary vector entry is recognized and handled.
4145 \f
4146 Version 2.3
4147
4148 * Masahide Washizawa contributed iconv modules for IBM1163 and IBM1164
4149 charsets.
4150
4151 * iconv (the program and the interface) now accepts empty names (excluding
4152 options like //TRANSLIT) to mean "use charset of current locale".
4153
4154 * localedef can now transliterate characters in strings which are not in
4155 the provided charmap. The information from the input locale is used.
4156
4157 * Prelinking support was added for ELF targets. This requires additional
4158 tools and recent versions of the GNU binutils. Contributed by Jakub Jelinek.
4159
4160 * Read-only stdio streams now use mmap to speed up operation by eliminating
4161 copying and buffer underflows. To use add 'm' to the mode string of
4162 the fopen/fdopen/freopen call. Implemented by Ulrich Drepper.
4163
4164 * The malloc functions were completely rewritten by Wolfram Gloger based
4165 on Doug Lea's malloc-2.7.0.c.
4166
4167 * Isamu Hasegawa contributed a completely new and POSIX-conformant
4168 implementation of regex.
4169
4170 * Bruno Haible upgraded the iconv and locale implementation to support
4171 Unicode 3.2.
4172
4173 * Contents of the LC_* and LANG environment variables in the CEN style are
4174 not recognized anymore. It never was used. Change by Ulrich Drepper.
4175
4176 * The runtime (ld.so, libc, libpthread for Linux) now can handle the ELF
4177 thread-local storage (TLS) ABI on some platforms.
4178 Changes by Ulrich Drepper. SH support by Kaz Kojima.
4179
4180 * Bruno Haible contributed iconv converters for ISO-2022-JP-3, SHIFT JIS-X0213,
4181 EUC-JISX0213, and TSCII.
4182
4183 * New header <ifaddrs.h> with functions `getifaddrs' and `freeifaddrs':
4184 BSD-compatible interface for getting all network interface addresses.
4185 Implementation for IPv4 by Roland McGrath.
4186
4187 * Loading of locale data is faster due to the introduction of a locale
4188 archive. Implemented by Roland McGrath and Ulrich Drepper.
4189
4190 * Startup times are significantly reduced by not using exported functions
4191 inside the library itself. Changes by Jakub Jelinek, Roland McGrath,
4192 and Ulrich Drepper.
4193
4194 * Steven Munroe contributed a port to PowerPC64/Linux.
4195 \f
4196 Version 2.2.6
4197
4198 * The Hurd now uses the GNU libio implementation of stdio rather than the
4199 old GNU stdio implementation, and uses a new ABI (libc.so.0.3).
4200
4201 * The Hurd on x86 now has the `ioperm' function and <sys/io.h> header file
4202 with the same behavior as the Linux system call of the same name.
4203 \f
4204 Version 2.2.5
4205
4206 * Stephen Moshier implemented log2, log10, powl and cbrtl for the
4207 128-bit long double format.
4208
4209 * Masahide Washizawa contributed iconv modules for IBM1132, IBM1133, IBM1160,
4210 IBM1161, and IBM1162 charsets.
4211
4212 * Andreas Jaeger contributed a port to x86-64/Linux.
4213
4214 * Peter Bruin contributed a port to PowerPC/Hurd.
4215
4216 * libc functions using I/O streams now can handle wide-oriented streams
4217 as well.
4218
4219 * optimizations in the dynamic linker. Binaries created by recent binutils
4220 versions start up quicker due to reduced time spent on relocations.
4221
4222 * Support for use of gcc3 added by Jakub Jelinek and HJ Lu.
4223 \f
4224 Version 2.2.4
4225
4226 * Stephen Moshier implemented cosh, expm1, log1p, acos, sinh, tanh,
4227 asinh, atanh, j0 for the 128-bit long double format.
4228
4229 * Bruno Haible updated all the code handling Unicode in some form to
4230 support Unicode 3.1.
4231
4232 * Speed of regex for single-byte locales is back to previous levels.
4233 Patch by Isamu Hasegawa.
4234
4235 * Alpha, SPARC, and IA-64 now also using floating stacks.
4236
4237 * Startup time of internationalized applications greatly improved through
4238 iconv cache. Use iconvconfig to generate the cache file.
4239 Contributed by Ulrich Drepper.
4240
4241 * The IA-64 specific part of ld.so was rewritten to eliminate some pretty
4242 severe performance problems. Patch by David Mosberger.
4243
4244 * The Hurd port got a lot more functionality like AIO, various stdio
4245 extensions, etc. Mainly done by Roland McGrath.
4246
4247 * mtrace can now lookup symbols in shared libraries.
4248 \f
4249 Version 2.2.3
4250
4251 * Intel's IA-64 math library is largely integrated. It provides fast and
4252 accurate implementatations for most basic and standard math functions
4253 in float, double, and long double format.
4254
4255 * Stephen Moshier implemented j0, j1, jn, y0, y1, yn, lgamma, erf, erfc,
4256 and asin for the 96-bit long double format and asin, log, tan for the
4257 128-bit long double format.
4258
4259 * The beginning of a last-bit accurate math library by IBM Haifa were added.
4260 The basic double functions exist today. Contributed by Abraham Ziv
4261 <ziv@il.ibm.com>, Moshe Olshansky <olshansk@il.ibm.com>, Ealan Henis
4262 <ealan@il.ibm.com>, and Anna Reitman <reitman@il.ibm.com>.
4263
4264 * An asynchronous name lookup library was added. The interface is designed
4265 after POSIX AIO. The proposal was circulated beforehand to get comments.
4266 No negative ones came in. Implemented by Ulrich Drepper.
4267
4268 * Port to S390/64bit contributed by Martin Schwidefsky
4269 <schwidefsky@de.ibm.com>.
4270
4271 * David Mosberger <davidm@hpl.hp.com> implemented the setcontext family
4272 of functions for Linux/IA-64.
4273
4274 * The RPC code is now thread safe. Threads can now use the same service
4275 of different services at the same time. Patch by Eric Norum
4276 <eric.norum@usask.ca> with some help by Ulrich Drepper.
4277
4278 * Martin Schwidefsky <schwidefsky@de.ibm.com> implemented the setcontext
4279 family of functions for Linux/S390.
4280
4281 * Ulrich Drepper <drepper@redhat.com> implemented the setcontext family
4282 of functions for Linux/x86.
4283
4284 * Port to Linux/CRIS contributed by Axis Communications.
4285 \f
4286 Version 2.2.2
4287
4288 * Lots of headers were cleaned up. Using the tool in the conform/ subdir
4289 we can now check for namespace violations and missing declarations. The
4290 result is that almost all headers are now Unix-compliant (as defined in
4291 the upcoming XPG6). The negative side is that some programs might need
4292 corrections, too, if they depend on the incorrect form of the headers in
4293 previous versions which defined too many symbols and included too many
4294 other headers.
4295
4296 * regex now handles multibyte character sets correctly.
4297 Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
4298
4299 * iconv (the program) does now conform to the upcoming XPG6 and handles
4300 charmaps. Instead of the charset names the path of charmaps can be
4301 provided and the conversion happens based on this data.
4302 Contributed by Ulrich Drepper.
4303
4304 * The locale program now provides detailed information about the installed
4305 locales. While
4306
4307 locale -a
4308
4309 only lists the names of the supported locales
4310
4311 locale -a --verbose
4312
4313 provides details such as country, language, and codeset name.
4314 Contributed by Ulrich Drepper.
4315 \f
4316 Version 2.2.1
4317
4318 * The gencat program now parses the input file according to the charset
4319 selected by the LC_CTYPE category. This is important for stateful
4320 character sets. To make generating catalogs easier there is a way
4321 to overwrite the charset selected by the locale: before the first
4322 message or $ quote line the catalog can contain a line like
4323
4324 $ codeset=ISO-8859-2
4325
4326 to select the charset (ISO-8859-2 in this case).
4327
4328 Implemented by Shinya Hanataka and Ulrich Drepper.
4329
4330 * New codeset conversion modules: IBM-922 (Estonia PC codeset),
4331 IBM-1124 (Ukraine PC codeset), IBM-1129 (Vietnamese PC codeset).
4332 Contributed by Masahide Washizawa <washi@jp.ibm.com>.
4333
4334 * Optimized string functions for Alpha ev6 and ev67 by Richard Henderson
4335 <rth@redhat.com> and Rick Gorton <rick.gorton@alpha-processor.com>.
4336
4337 * The LANGUAGE environment variable is now ignored unless the locale is
4338 changed from the default "C" locale.
4339
4340 * The usual bug fixes.
4341 \f
4342 Version 2.2
4343
4344 * Greg McGary added runtime support for bounds checking using gcc's
4345 new -fbounded-pointers option. ix86 target is complete. PowerPC
4346 is in progress.
4347
4348 * Thorsten Kukuk added secure mode support to nscd.
4349
4350 * The Berkeley db libraries have been removed.
4351
4352 Related, the nss_db module is now in a separate package since it
4353 obviously requires a database library being available.
4354
4355 * Wide character I/O streams implemented by Ulrich Drepper.
4356
4357 * Functions from the extended socket API added by Ulrich Drepper.
4358
4359 * Functions feenableexcept and fedisableexcept to control the
4360 behaviour of individual exceptions have been added by Andreas Jaeger.
4361
4362 * ldconfig program added by Andreas Jaeger and Jakub Jelinek.
4363
4364 * The resolver code has been updated from bind 8.2.3-T5B which supports
4365 threads. The integration was done by Andreas Jaeger, Adam D. Bradley,
4366 and Mark Kettenis.
4367
4368 This change could in some situations effect backward compatibility. Since
4369 now `_res' is a thread-local instead of a global variable, modifying it
4370 in one thread does not have any effect in other threads.
4371
4372 The resolver library was also extended to allow IPv6 as the transport
4373 protocol for the requests. This work was done by Stig Venaas.
4374
4375 * Compatibility code for K&R C compilers has been removed from the
4376 header files. A ISO C compiler is needed to use the library
4377 (conforming to either C89 or C99 standard).
4378
4379 * Complete rewrite of the localedef program to support multibyte character
4380 sets. Implement handling of ISO 14651 and ISO 14652. Rewrite strcoll,
4381 strxfrm, wcscoll, and wcsxfrm functions. Make isw*() functions work.
4382 Implemented by Ulrich Drepper.
4383
4384 Bruno Haible significantly improved the generation and use of the data
4385 structures for the wide character tables.
4386
4387 * Plural handling in gettext implemented by Ulrich Drepper.
4388
4389 * The utmp daemon has been removed.
4390
4391 * The port to MIPS-Linux has been finished by Andreas Jaeger.
4392
4393 * A port to Hitachi SH3 and SH4 has been contributed by Kazumoto Kojima
4394 and Yutaka Niibe.
4395
4396 * POSIX clocks and timers implemented by Kaz Kylheku and Ulrich Drepper.
4397
4398 * POSIX barriers implemented by Kaz Kylheku.
4399
4400 * POSIX spawn function family implemented by Ulrich Drepper.
4401
4402 * POSIX spinlocks are now available.
4403
4404 * Timed wait functions for mutex, rwlock, and semaphores are implemented.
4405
4406 * the configure option --enable-kernel=X.Y.Z allows to strip out
4407 compatibility for kernel versions before X.Y.Z. This is currently only
4408 implemented for Linux.
4409
4410 * the sockaddr_in6 structure changed. The IPv6 working group added a new
4411 field sin6_scope_id. This means that all programs using IPv6 should be
4412 recompiled. Don't expect binary compatibility with previous glibc
4413 versions.
4414
4415 * various conversion modules for IBM character sets contributed by
4416 Masahide Washizawa.
4417
4418 * IA-64 port by Jes Sorensen and HJ Lu.
4419 \f
4420 Version 2.1.3
4421
4422 * bug fixes
4423
4424 \f
4425 Version 2.1.2
4426
4427 * bug fixes
4428
4429 \f
4430 Version 2.1.1
4431
4432 * New ISO C 9x function _Exit, imaxabs, and imaxdiv are added.
4433
4434 * New xdr functions are added; some rpc functions are now 64bit clean.
4435
4436 * Fixed a number of bugs and memory leaks (especially in NIS+ code).
4437
4438 * Fixed known incompatibilities with glibc 2.0.
4439
4440 * New functions lock64, strchrnul, rawmemchr, getutmp and getutmpx.
4441
4442 * Optimized a number of functions (especially the ELF dynamic loader).
4443
4444 * Update timezone data files.
4445
4446 * lots of charmaps corrections
4447
4448 * some new locale definitions and charmaps
4449
4450 \f
4451 Version 2.1
4452
4453 * Richard Henderson corrected size of struct timeval on Linux/Alpha to
4454 conform to POSIX member type requirements. Symbol versions have been
4455 adjusted as needed within the library, and for direct use by applications,
4456 but there is potential for problems if third-party libraries use
4457 struct timeval as part of their interface. This does not present
4458 a problem for X and other "essential" system libraries.
4459
4460 * An additional locale model to support C++ Standard Library locale
4461 model and probably more was implemented by Ulrich Drepper.
4462
4463 * Eric Youngdale and Ulrich Drepper implemented versioning of objects on
4464 symbol level.
4465
4466 * Miles Bader provided the `argp' function family to support hierachical
4467 command line argument parsing, layered on top of getopt.
4468
4469 * strtod accepts new hexadecimal floating-point format from ISO C 9X.
4470
4471 * printf has two new specifiers %a and %A to print hexadecimal flaoting-point
4472 numbers.
4473
4474 * scanf recognizes the %a and %A format for scanning floating point numbers.
4475
4476 * the new headers <stdint.h> and <inttypes.h> from ISO C 9X provides
4477 information and interfaces for the available integer types.
4478
4479 * about 130 new math functions were added to implement the ISO C9x math
4480 library.
4481
4482 * the new header <complex.h> contains definitions of the complex math
4483 functions from ISO C 9X.
4484
4485 * the new header <tgmath.h> defines generic macros to use complex or
4486 real valued functions.
4487
4488 * Thorsten Kukuk provided an implementation for NIS+, securelevel 0, 1 and 2.
4489
4490 * Andreas Jaeger provided a test suite for the math library.
4491
4492 * Mark Kettenis implemented the utmpx interface and an utmp daemon.
4493
4494 * Ulrich Drepper added character set conversion functions (iconv).
4495
4496 * Optimized string functions have been added.
4497
4498 * The localedata addon is now part of glibc.
4499
4500 * An implementation of profiling shared libraries was added by Ulrich Drepper.
4501
4502 * Thorsten Kukuk and Ulrich Drepper provided an implementation for a caching
4503 daemon for NSS (nscd).
4504
4505 Missing a better place here are some numbers on improvements. Under
4506 Linux 2.1.125 un-tar-ing the kernel sources takes
4507
4508 user system wall
4509
4510 using local files 12.19s 6.88s 22.91s
4511
4512 using NIS 13.92s 8.91s 26.34s
4513
4514 using NIS & nscd 10.37s 7.34s 25.30s
4515
4516 using NIS+ 27.57s 30.37s 640.46s
4517
4518 using NIS+ & nscd 10.25s 7.83s 26.51s
4519
4520 using NIS & old nscd [1] 13.83s 8.32s 29.60s
4521
4522 Keep in mind that non-namelookup related operations dominate above times.
4523 It was just a common complain that using NIS+ unpacking the kernel is
4524 horribly slow.
4525
4526 [1] The old nscd implementation is not available anymore since it was
4527 distributed with glibc up to version 2.0.98 and thus is now replaced.
4528
4529 * Tim Waugh provided an implementation of the POSIX.2 wordexp function family.
4530
4531 * Mark Kettenis provided a Hesiod NSS module.
4532
4533 * The ELF dynamic loader knows how to process dynamic string tokens ($ORIGIN
4534 and $PLATFORM) in RPATHs and similar strings (Ulrich Drepper).
4535
4536 * rcmd can now handle netgroups (Dick Streefland).
4537
4538 * A port to the ARM was contributed by Phil Blundell, Pat Beirne and Scott
4539 Bambrough.
4540
4541 * Support for the IPv6 protocol has been added to the socket API, as per the
4542 latest draft standards.
4543
4544 * Support for Linux 2.2 has been added.
4545
4546 * Interface changes relative to the latest 2.0.x release:
4547 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4548 addseverity NEW: Unix98
4549 alphasort64 NEW: LFS
4550 argp_err_exit_status NEW: argp, GNU ext
4551 argp_error NEW: argp, GNU ext
4552 argp_failure NEW: argp, GNU ext
4553 argp_help NEW: argp, GNU ext
4554 argp_parse NEW: argp, GNU ext
4555 argp_program_bug_address NEW: argp, GNU ext
4556 argp_program_version NEW: argp, GNU ext
4557 argp_program_version_hook NEW: argp, GNU ext
4558 argp_state_help NEW: argp, GNU ext
4559 argp_usage NEW: argp, GNU ext
4560 authdes_create NEW: Secure RPC
4561 authdes_getucred NEW: Secure RPC
4562 authdes_pk_create NEW: Secure RPC
4563 backtrace NEW: GNU ext.
4564 backtrace_symbols NEW: GNU ext.
4565 backtrace_symbols_fd NEW: GNU ext.
4566 cacos NEW: ISO C 9x
4567 cacosf NEW: ISO C 9x
4568 cacosh NEW: ISO C 9x
4569 cacoshf NEW: ISO C 9x
4570 cacoshl NEW: ISO C 9x
4571 cacosl NEW: ISO C 9x
4572 capget NEW: kernel
4573 capset NEW: kernel
4574 carg NEW: ISO C 9x
4575 cargf NEW: ISO C 9x
4576 cargl NEW: ISO C 9x
4577 casin NEW: ISO C 9x
4578 casinf NEW: ISO C 9x
4579 casinh NEW: ISO C 9x
4580 casinhf NEW: ISO C 9x
4581 casinhl NEW: ISO C 9x
4582 casinl NEW: ISO C 9x
4583 catan NEW: ISO C 9x
4584 catanf NEW: ISO C 9x
4585 catanh NEW: ISO C 9x
4586 catanhf NEW: ISO C 9x
4587 catanhl NEW: ISO C 9x
4588 catanl NEW: ISO C 9x
4589 cbc_crypt NEW: Secure RPC
4590 ccos NEW: ISO C 9x
4591 ccosf NEW: ISO C 9x
4592 ccosh NEW: ISO C 9x
4593 ccoshf NEW: ISO C 9x
4594 ccoshl NEW: ISO C 9x
4595 ccosl NEW: ISO C 9x
4596 cexp NEW: ISO C 9x
4597 cexpf NEW: ISO C 9x
4598 cexpl NEW: ISO C 9x
4599 cimag NEW: ISO C 9x
4600 cimagf NEW: ISO C 9x
4601 cimagl NEW: ISO C 9x
4602 clearerr_locked REMOVED
4603 clntunix_create NEW: sunrpc ext
4604 clog NEW: ISO C 9x
4605 clog10 NEW: ISO C 9x
4606 clog10f NEW: ISO C 9x
4607 clog10l NEW: ISO C 9x
4608 clogf NEW: ISO C 9x
4609 clogl NEW: ISO C 9x
4610 conj NEW: ISO C 9x
4611 conjf NEW: ISO C 9x
4612 conjl NEW: ISO C 9x
4613 cpow NEW: ISO C 9x
4614 cpowf NEW: ISO C 9x
4615 cpowl NEW: ISO C 9x
4616 cproj NEW: ISO C 9x
4617 cprojf NEW: ISO C 9x
4618 cprojl NEW: ISO C 9x
4619 creal NEW: ISO C 9x
4620 crealf NEW: ISO C 9x
4621 creall NEW: ISO C 9x
4622 creat64 NEW: LFS
4623 csin NEW: ISO C 9x
4624 csinf NEW: ISO C 9x
4625 csinh NEW: ISO C 9x
4626 csinhf NEW: ISO C 9x
4627 csinhl NEW: ISO C 9x
4628 csinl NEW: ISO C 9x
4629 csqrt NEW: ISO C 9x
4630 csqrtf NEW: ISO C 9x
4631 csqrtl NEW: ISO C 9x
4632 ctan NEW: ISO C 9x
4633 ctanf NEW: ISO C 9x
4634 ctanh NEW: ISO C 9x
4635 ctanhf NEW: ISO C 9x
4636 ctanhl NEW: ISO C 9x
4637 ctanl NEW: ISO C 9x
4638 des_setparity NEW: Secure RPC
4639 ecb_crypt NEW: Secure RPC
4640 endutxent NEW: Unix98
4641 exp10 NEW: ISO C 9x
4642 exp10f NEW: ISO C 9x
4643 exp10l NEW: ISO C 9x
4644 exp2 NEW: ISO C 9x
4645 exp2f NEW: ISO C 9x
4646 exp2l NEW: ISO C 9x
4647 fattach NEW: STREAMS
4648 fdetach NEW: STREAMS
4649 fdim NEW: ISO C 9x
4650 fdimf NEW: ISO C 9x
4651 fdiml NEW: ISO C 9x
4652 feclearexcept NEW: ISO C 9x
4653 fegetenv NEW: ISO C 9x
4654 fegetexceptflag NEW: ISO C 9x
4655 fegetround NEW: ISO C 9x
4656 feholdexcept NEW: ISO C 9x
4657 feof_locked REMOVED
4658 feraiseexcept NEW: ISO C 9x
4659 ferror_locked REMOVED
4660 fesetenv NEW: ISO C 9x
4661 fesetexceptflag NEW: ISO C 9x
4662 fesetround NEW: ISO C 9x
4663 fetestexcept NEW: ISO C 9x
4664 feupdateenv NEW: ISO C 9x
4665 fflush_locked REMOVED
4666 ffsl NEW: GNU ext.
4667 ffsll NEW: GNU ext.
4668 fgetpos64 NEW: LFS
4669 fgets_unlocked NEW: GNU ext.
4670 fileno_locked REMOVED
4671 fma NEW: ISO C 9x
4672 fmaf NEW: ISO C 9x
4673 fmal NEW: ISO C 9x
4674 fmax NEW: ISO C 9x
4675 fmaxf NEW: ISO C 9x
4676 fmaxl NEW: ISO C 9x
4677 fmin NEW: ISO C 9x
4678 fminf NEW: ISO C 9x
4679 fminl NEW: ISO C 9x
4680 fmtmsg NEW: Unix98
4681 fopen64 NEW: LFS
4682 fputc_locked REMOVED
4683 fputs_unlocked NEW: GNU ext.
4684 fread_unlocked NEW: GNU ext.
4685 freopen64 NEW: LFS
4686 fseeko NEW: Unix98
4687 fsetpos64 NEW: LFS
4688 fstatfs64 NEW: LFS
4689 fstatvfs NEW: Unix98
4690 fstatvfs64 NEW: LFS
4691 ftello NEW: Unix98
4692 ftello64 NEW: LFS
4693 ftruncate64 NEW: LFS
4694 ftw64 NEW: LFS
4695 fwrite_unlocked NEW: GNU ext.
4696 gai_strerror NEW: IPv6
4697 gamma_r REMOVED
4698 gammaf_r REMOVED
4699 gammal_r REMOVED
4700 getchar_locked REMOVED
4701 getdate NEW: Unix98
4702 getdate_err NEW: Unix98
4703 getdate_r NEW: GNU ext.
4704 getmsg NEW: STREAMS
4705 getnameinfo NEW: IPv6
4706 getnetname NEW: Secure RPC
4707 getpmsg NEW: STREAMS
4708 getpt NEW: Unix98 PTY
4709 getrlimit64 NEW: LFS
4710 getutxent NEW: Unix98
4711 getutxid NEW: Unix98
4712 getutxline NEW: Unix98
4713 glob64 NEW: GNU ext.
4714 globfree64 NEW: GNU ext.
4715 gnu_get_libc_release NEW: GNU ext.
4716 gnu_get_libc_version NEW: GNU ext.
4717 grantpt NEW: Unix98 PTY
4718 host2netname NEW: Secure RPC
4719 iconv NEW: iconv
4720 iconv_close NEW: iconv
4721 iconv_open NEW: iconv
4722 if_freenameindex NEW: IPv6
4723 if_indextoname NEW: IPv6
4724 if_nameindex NEW: IPv6
4725 if_nametoindex NEW: IPv6
4726 in6addr_any NEW: IPv6
4727 in6addr_loopback NEW: IPv6
4728 inet6_isipv4mapped NEW: IPv6
4729 isastream NEW: STREAMS
4730 iswblank NEW: GNU ext.
4731 key_decryptsession NEW: Secure RPC
4732 key_decryptsession_pk NEW: Secure RPC
4733 key_encryptsession NEW: Secure RPC
4734 key_encryptsession_pk NEW: Secure RPC
4735 key_gendes NEW: Secure RPC
4736 key_get_conv NEW: Secure RPC
4737 key_secretkey_is_set NEW: Secure RPC
4738 key_setnet NEW: Secure RPC
4739 key_setsecret NEW: Secure RPC
4740 llrint NEW: ISO C 9x
4741 llrintf NEW: ISO C 9x
4742 llrintl NEW: ISO C 9x
4743 llround NEW: ISO C 9x
4744 llroundf NEW: ISO C 9x
4745 llroundl NEW: ISO C 9x
4746 log2 NEW: ISO C 9x
4747 log2f NEW: ISO C 9x
4748 log2l NEW: ISO C 9x
4749 lrint NEW: ISO C 9x
4750 lrintf NEW: ISO C 9x
4751 lrintl NEW: ISO C 9x
4752 lround NEW: ISO C 9x
4753 lroundf NEW: ISO C 9x
4754 lroundl NEW: ISO C 9x
4755 lseek64 NEW: LFS
4756 makecontext NEW: Unix98
4757 mempcpy NEW: GNU ext.
4758 mmap64 NEW: LFS
4759 moncontrol REMOVED
4760 modify_ldt NEW: kernel
4761 nan NEW: ISO C 9x
4762 nanf NEW: ISO C 9x
4763 nanl NEW: ISO C 9x
4764 nearbyint NEW: ISO C 9x
4765 nearbyintf NEW: ISO C 9x
4766 nearbyintl NEW: ISO C 9x
4767 netname2host NEW: Secure RPC
4768 netname2user NEW: Secure RPC
4769 nexttoward NEW: ISO C 9x
4770 nexttowardf NEW: ISO C 9x
4771 nexttowardl NEW: ISO C 9x
4772 nftw NEW: Unix98
4773 nftw64 NEW: LFS
4774 open64 NEW: LFS
4775 passwd2des NEW: Secure RPC
4776 pow10 NEW: GNU ext.
4777 pow10f NEW: GNU ext.
4778 pow10l NEW: GNU ext.
4779 pread NEW: Unix98
4780 pread64 NEW: LFS
4781 printf_size NEW: GNU ext.
4782 printf_size_info NEW: GNU ext.
4783 profil_counter REMOVED
4784 pthread_mutexattr_getkind_np REPLACED
4785 pthread_mutexattr_setkind_np REPLACED
4786 ptsname NEW: Unix98 PTY
4787 ptsname_r NEW: Unix98 PTY
4788 putc_locked REMOVED
4789 putchar_locked REMOVED
4790 putgrent NEW: GNU ext.
4791 putmsg NEW: STREAMS
4792 putpmsg NEW: STREAMS
4793 pututxline NEW: Unix98
4794 pwrite NEW: Unix98
4795 pwrite64 NEW: LFS
4796 readdir64 NEW: LFS
4797 readdir64_r NEW: LFS
4798 remquo NEW: ISO C 9x
4799 remquof NEW: ISO C 9x
4800 remquol NEW: ISO C 9x
4801 round NEW: ISO C 9x
4802 roundf NEW: ISO C 9x
4803 roundl NEW: ISO C 9x
4804 rtime NEW: GNU ext.
4805 scalbln NEW: ISO C 9x
4806 scalblnf NEW: ISO C 9x
4807 scalblnl NEW: ISO C 9x
4808 scandir64 NEW: LFS
4809 sendfile NEW: kernel
4810 setcontext NEW: Unix98
4811 setrlimit64 NEW: LFS
4812 setutxent NEW: Unix98
4813 sighold NEW: Unix98
4814 sigignore NEW: Unix98
4815 sigqueue NEW: POSIX.1b
4816 sigrelse NEW: Unix98
4817 sigset NEW: POSIX.1b
4818 sigtimedwait NEW: POSIX.1b
4819 sigwaitinfo NEW: POSIX.1b
4820 sincos NEW: GNU ext.
4821 sincosf NEW: GNU ext.
4822 sincosl NEW: GNU ext.
4823 statfs64 NEW: LFS
4824 statvfs NEW: Unix98
4825 statvfs64 NEW: LFS
4826 strcasestr NEW: GNU ext.
4827 strtoimax NEW: ISO C 9x
4828 strtoumax NEW: ISO C 9x
4829 strverscmp NEW: GNU ext.
4830 svcauthdes_stats NEW: Secure RPC
4831 svcunix_create NEW: sunrpc etx
4832 svcunixfd_create NEW: sunrpc ext
4833 swapcontext NEW: Unix98
4834 tcgetsid NEW: Unix98 PTY
4835 tdestroy NEW: GNU ext.
4836 tgamma NEW: ISO C 9x
4837 tgammaf NEW: ISO C 9x
4838 tgammal NEW: ISO C 9x
4839 tmpfile64 NEW: LFS
4840 trunc NEW: ISO C 9x
4841 truncate64 NEW: LFS
4842 truncf NEW: ISO C 9x
4843 truncl NEW: ISO C 9x
4844 umount2 NEW: kernel
4845 unlockpt NEW: Unix98 PTY
4846 updwtmpx NEW: Unix98
4847 user2netname NEW: Secure RPC
4848 utmpxname NEW: Unix98
4849 versionsort NEW: GNU ext.
4850 versionsort64 NEW: GNU ext.
4851 waitid NEW: Unix98
4852 wcscasecmp NEW: GNU ext.
4853 wcsncasecmp NEW: GNU ext.
4854 wcsnlen NEW: GNU ext.
4855 wcstoimax NEW: ISO C 9x
4856 wcstoll NEW: ISO C 9x
4857 wcstoull NEW: ISO C 9x
4858 wcstoumax NEW: ISO C 9x
4859 wcswcs NEW: Unix98
4860 wordexp NEW: POSIX.2
4861 wordfree NEW: POSIX.2
4862 write_profiling REMOVED
4863 xdecrypt NEW: Secure RPC
4864 xdr_authdes_cred NEW: Secure RPC
4865 xdr_authdes_verf NEW: Secure RPC
4866 xdr_cryptkeyarg NEW: Secure RPC
4867 xdr_cryptkeyarg2 NEW: Secure RPC
4868 xdr_cryptkeyres NEW: Secure RPC
4869 xdr_getcredres NEW: Secure RPC
4870 xdr_key_netstarg NEW: Secure RPC
4871 xdr_key_netstres NEW: Secure RPC
4872 xdr_keybuf NEW: Secure RPC
4873 xdr_keystatus NEW: Secure RPC
4874 xdr_netnamestr NEW: Secure RPC
4875 xdr_sizeof NEW: Secure RPC
4876 xdr_unixcred NEW: sunrpc ext
4877 xencrypt NEW: Secure RPC
4878 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4879 \f
4880 Version 2.0.6
4881
4882 * more bug fixes
4883
4884 \f
4885 Version 2.0.5
4886
4887 * more bug fixes
4888
4889 * inet_ntoa is thread-safe
4890
4891 * updwtmp is moved from libutil to libc
4892
4893 * rewrite of cbrt function
4894
4895 * update of timezone data
4896 \f
4897 Version 2.0.4
4898
4899 * more bug fixes
4900 \f
4901 Version 2.0.3
4902
4903 * more bug fixes
4904 \f
4905 Version 2.0.2
4906
4907 * more bug fixes
4908
4909 * add atoll function
4910
4911 * fix complex problems in Berkeley DB code
4912
4913 * fix math functions
4914 \f
4915 Version 2.0.1
4916
4917 * fixed lots of header problems (especially Linux/GNU specific)
4918
4919 * dynamic loader preserves all registers
4920
4921 * Roland McGrath provided support for handling of auxiliary objects in
4922 the ELF dynamic loader.
4923
4924 * support for parallel builds is improved
4925 \f
4926 Version 2.0
4927
4928 * GNU extensions are no longer declared by default. To enable them you
4929 must define the macro `_GNU_SOURCE' in your program or compile with
4930 `-D_GNU_SOURCE'.
4931
4932 * The library has changed from using GNU ld symbol aliases to using weak
4933 symbols where available. The ELF object file format supports weak
4934 symbols; GNU ld also supports weak symbols in the a.out format. (There
4935 is also now support for other GNU ld extensions in ELF. Use the
4936 `--with-elf' option to configure to indicate you have ELF, and
4937 `--with-gnu-ld' if using GNU ld.) This change resulted in the deletion
4938 of many files which contained only symbol aliases, reducing the size of
4939 the source and the compiled library; many other files were renamed to
4940 less cryptic names previously occupied by the symbol alias files.
4941 There is a new header file <elf.h> for programs which operate on
4942 files in the ELF format.
4943
4944 * Converted to Autoconf version 2, so `configure' has more options.
4945 Run `configure --help' to see the details.
4946
4947 * The library can now be configured to build profiling, highly-optimized
4948 (but undebuggable), and/or shared libraries (ELF with GNU ld only). The
4949 `--enable-profile', `--enable-omitfp', and `--enable-shared' options to
4950 `configure' enable building these extra libraries. The shared library is
4951 built by default when using both ELF and GNU ld. When shared libraries
4952 are enabled, the new library `-ldl' is available for arbitrary run-time
4953 loading of shared objects; its interface is defined in <dlfcn.h>. The
4954 new header file <link.h> gives access to the internals of the run-time
4955 dynamic linker, `ld.so'. The shell script `ldd' is similar to the
4956 application of same name on other systems and it provides information
4957 about dynamically linked binaries.
4958
4959 * The C library now provides the run-time support code for profiling
4960 executables compiled with `-pg'. Programs can control the profiling code
4961 through the interface in <sys/gmon.h>. The `gmon.out' files written by
4962 the GNU C library can be read only by GNU `gprof' (from GNU binutils);
4963 the support for this file format was contributed by David Mosberger-Tang.
4964
4965 * The math code has been replaced with a math library based on fdlibm from
4966 Sun, and modified by JT Conklin and Ulrich Drepper with i387 support, by
4967 Ian Taylor with `float' functions and by Ulrich Drepper with `long double'
4968 functions. The math functions now reside in a separate library, so
4969 programs using them will need to use `-lm' their linking commands.
4970
4971 * John C. Bowman contributed optimized ix87 assembler inline functions.
4972
4973 * Ulrich Drepper has contributed support for an `/etc/nsswitch.conf'
4974 mechanism similar to that found in Solaris 2. This is now used for the
4975 group, passwd, hosts, networks, services, protocols, rpc, ethers,
4976 shadow, netgroup, publickey, and alias databases. The `nsswitch.conf'
4977 file controls what services are used for each individual database. This
4978 works by loading shared libraries with names specified in `nsswitch.conf',
4979 so service modules can be changed or added at any time without even
4980 relinking any program. Currently there are the file, db, and NIS based
4981 NSS services available.
4982
4983 * The new functions `strtoq' and `strtouq' parse integer values from
4984 strings, like `strtol' and `strtoul', but they return `long long int' and
4985 `unsigned long long int' values, respectively (64-bit quantities).
4986
4987 * The new functions `strtof' and `strtold' parse floating-point values from
4988 strings, like `strtod', but they return `float' and `long double' values,
4989 respectively (on some machines `double' and `long double' are the same).
4990
4991 * Ulrich Drepper has contributed new implementations of the floating-point
4992 printing and reading code used in the `printf' family of functions and
4993 `strtod', `strtof', and `strtold'. These new functions are perfectly
4994 accurate, and much faster than the old ones.
4995
4996 * The implementation of the POSIX locale model was completely rewritten by
4997 Ulrich Drepper. This includes the new programs `localedef' and `locale'
4998 to compile the POSIX locale definition.
4999
5000 * The former dummy implementations of the strcoll and strxfrm function are
5001 now replaced by fully functional code contributed by Ulrich Drepper. The
5002 collation information comes from the POSIX locale definitions.
5003
5004 * The new header <langinfo.h> defines an interface for accessing
5005 various locale-dependent data (using the locale chosen with `setlocale').
5006
5007 * Ulrich Drepper has contributed a new suite of functions for operation on
5008 wide-character and multibyte-character strings, in <wchar.h>;
5009 and classification and case conversion of wide characters, in <wctype.h>.
5010 These new functions are conforming to the ISO C, Amendement 1 specification.
5011
5012 * There is now a second implementation of the standard I/O library available.
5013 It comes from GNU libg++ as was written by Per Bothner, heavily modified
5014 by Hongjiu Lu and made thread safe by Ulrich Drepper.
5015
5016 * You can now use positional parameter specifications in format strings
5017 for the `printf' and `scanf' families of functions. For example,
5018 `printf ("Number %2$d, Mr %1$s\n", "Jones", 6);'' prints
5019 ``Number 6, Mr Jones''. This is mainly useful when providing different
5020 format strings for different languages, whose grammars may dictate
5021 different orderings of the values being printed. To support this
5022 feature, the interface for `register_printf_handler' has changed; see
5023 the header file <printf.h> for details.
5024
5025 * The `printf' and `scanf' families of functions now understand a new
5026 formatting flag for numeric conversions: the ' flag (e.g. %'d or %'f) says
5027 to group numbers as indicated by the locale; for `scanf' and friends, this
5028 says to accept as valid only a number with all the proper grouping
5029 separators in the right places. In the default "C" locale, numbers are
5030 not grouped; but locales for specific countries will define the usual
5031 conventions (i.e. separate thousands with `,' in the US locale).
5032
5033 * The pgrp functions have been regularized, slightly incompatibly but much
5034 less confusingly. The core functions are now `getpgid' and `setpgid',
5035 which take arguments for the PID to operate on; the POSIX.1 `getpgrp' (no
5036 argument) and BSD `setpgrp' (identical to `setpgid') functions are
5037 provided for compatibility. There is no longer an incompatible `getpgrp'
5038 with an argument declared under _BSD_SOURCE; no BSD code uses it.
5039
5040 * The new header file <fts.h> and suite of functions simplify programs that
5041 operate on directory trees. This code comes from 4.4 BSD.
5042
5043 * The resolver code has been updated from the BIND 4.9.5-P1 release.
5044 Parts of the code were heavily modified by Ulrich Drepper to fit in the
5045 NSS scheme used in glibc.
5046
5047 * There is a new malloc debugging hook `__memalign_hook'.
5048
5049 * There are new typedefs `ushort' for `unsigned short int' and `uint' for
5050 `unsigned int' in <sys/types.h>. These are for compatibility only and
5051 their use is discouraged.
5052
5053 * The `-lmcheck' library to enable standard malloc debugging hooks is now
5054 done differently, so that it works even without GNU ld.
5055
5056 * New function `euidaccess' checks allowed access to a file like `access',
5057 but using the effective IDs instead of the real IDs.
5058
5059 * The time zone data files have been updated for the latest and greatest
5060 local time conventions of the countries of the world.
5061
5062 * The new function `dirfd' extracts the file descriptor used by a DIR stream;
5063 see <dirent.h>.
5064
5065 * The new functions `ecvt', `fcvt', and `gcvt' provide an obsolete interface
5066 for formatting floating-point numbers. They are provided only for
5067 compatibility; new programs should use `sprintf' instead. There are
5068 also equivalent function for the `long double' floating-point type and
5069 all functions also exist in a reentrant form.
5070
5071 * The new auxiliary library `-lutil' from 4.4 BSD contains various
5072 functions for maintaining the login-record files (primarily of use to
5073 system programs such as `login'), and convenient functions for
5074 allocating and initializing a pseudo-terminal (pty) device.
5075
5076 * Ulrich Drepper has contributed new support for System V style
5077 shared memory and IPC on systems that support it.
5078
5079 * Ulrich Drepper has contributed several miscellaneous new functions found
5080 in System V: The `hsearch' family of functions provide an effective
5081 implementation of hash tables; `a64l' and `l64a' provide a very simple
5082 binary to ASCII mapping; `drand48' and friends provide a 48-bit random
5083 number generator.
5084
5085 * Ulrich Drepper has contributed new reentrant counterparts for the
5086 `random' and `hsearch' families of functions; `random_r', `hsearch_r', etc.
5087
5088 * Ulrich Drepper has contributed new, highly-optimized versions of several
5089 string functions for the i486/Pentium family of processors.
5090
5091 * Ulrich Drepper has updated the Linux-specific code, based largely
5092 on work done in Hongjiu Lu's version of GNU libc for Linux.
5093 The GNU library now supports Linux versions 2.0.10 and later,
5094 using the ELF object file format (i[3456]86-*-linux).
5095
5096 * Andreas Schwab has ported the C library to Linux/m68k (m68k-*-linux).
5097
5098 * David Mosberger-Tang and Richard Henderson have ported the C library
5099 to Linux/Alpha (alpha-*-linux). Richard Henderson contributed the
5100 dynamic linking support for ELF/Alpha.
5101
5102 * Richard Henderson contributed several Alpha optimized assembler function
5103 for arithmetic and string handling.
5104
5105 * Ulrich Drepper has contributed a new set of message catalog functions to
5106 support multiple languages using the <libintl.h> interface, for use with
5107 his new package GNU gettext. Translation volunteers have contributed
5108 catalogs of the library's messages in Spanish, German, and Korean.
5109
5110 * For compatibility with XPG4, Ulrich Drepper has contributed the `gencat'
5111 program and the `catgets' function for reading the catalog files it
5112 creates. (The <libintl.h> interface is preferred; we include the
5113 <nl_types.h> interface using `catgets' only for source compatibility with
5114 programs already written to use it.)
5115
5116 * New header file <values.h> gives SVID-compatible names for <limits.h>
5117 constants.
5118
5119 * Various new macros, declarations, and small header files for compatibility
5120 with 4.4 BSD.
5121
5122 * New function `group_member' is a convenient way to check if a process has
5123 a given effective group ID.
5124
5125 * When using GCC 2.7 and later, the socket functions are now declared in a
5126 special way so that passing an argument of type `struct sockaddr_in *',
5127 `struct sockaddr_ns *', or `struct sockaddr_un *' instead of the generic
5128 `struct sockaddr *' type, does not generate a type-clash warning.
5129
5130 * New function `error' declared in header file <error.h> is a convenient
5131 function for printing error messages and optionally exiting; this is the
5132 canonical function used in GNU programs. The new functions `err', `warn',
5133 and friends in header file <err.h> are the canonical 4.4 BSD interface for
5134 doing the same thing.
5135
5136 * The <glob.h> interface has several new flags from 4.4 BSD that extend the
5137 POSIX.2 `glob' function to do ~ and {...} expansion.
5138
5139 * New function `unsetenv' complements `setenv' for compatibility with 4.4 BSD.
5140 `clearenv' which is used in POSIX.9 is also available.
5141
5142 * New function `getsid' returns session ID number on systems that support it.
5143
5144 * We have incorporated the 4.4 BSD `db' library (version 1.85). New header
5145 files <db.h> and <mpool.h> provide a rich set of functions for several
5146 types of simple databases stored in memory and in files, and <ndbm.h> is
5147 an old `ndbm'-compatible interface using the `db' functions. Link with
5148 `-ldb' to get these functions.
5149
5150 * New macro `strdupa' copies a string like `strdup', but uses local stack
5151 space from `alloca' instead of dynamic heap space from `malloc'.
5152
5153 * New function `strnlen' is like `strlen' but searches only a given maximum
5154 number of characters for the null terminator. `stpncpy', `strndup' and
5155 `strndupa' are similar variants for the `stpcpy', `strdup' and `strdupa'
5156 function.
5157
5158 * New function `statfs' in header <sys/statfs.h>.
5159
5160 * The new <argz.h> and <envz.h> interfaces contributed by Miles Bader
5161 provide convenient functions for operating on blocks of null-terminated
5162 strings.
5163
5164 * A new suite of functions in <utmp.h> handle all the details of reading
5165 and writing the utmp file.
5166
5167 * An implementation of the NIS/YP(tm) based NSS service was contributed by
5168 Thorsten Kukuk.
5169
5170 * Paul Eggert and Ulrich Drepper modified the `strftime' function to be
5171 completely POSIX compliant and also implemented the extended functionality
5172 to handle alternate digit representation and alternate era date formats.
5173
5174 * Ulrich Drepper provided an implementation of the `strptime' function
5175 defined in XPG4.2 which transforms a string into a `struct tm' value.
5176
5177 * Paul Eggert provided the tzselect shell script as part of the timezone
5178 code. The shell script makes it easy to select the correct timezone
5179 specification.
5180
5181 * The implementation of the malloc family of functions is completely replaced
5182 by a new implementation by Doug Lea with many improvements by Wolfram Gloger.
5183 The implementation uses the mmap function (if available) and it is
5184 optimized for the use in multi threaded programs.
5185
5186 * Ulrich Drepper contributed a MD5 "encryption" for the crypt family of
5187 functions. This new functionality is usable by specifying a special
5188 salt string and it is compatible with implementation on *BSD systems.
5189
5190 * Lots of functions from the XPG4.2 standard were added by Ulrich Drepper:
5191 `getsubopt' to handle second level command line options, `bsd_signal'
5192 to access BSD style `signal' functionality, the obsolete `regexp' style
5193 expression matcher.
5194
5195 * the `lchown' function is available on system which support this
5196 functionality.
5197
5198 * The implementation of the shadow password handling function was contributed
5199 by Ulrich Drepper.
5200
5201 * David Mosberger-Tang changed the SunRPC implementation to be 64bit safe.
5202
5203 * POSIX.1g support was added. The <sys/select.h> header is available,
5204 `isfdtype' and `pselect' are implemented. Craig Metz contributed an
5205 implementation of `getaddrinfo'.
5206 \f
5207 Version 1.09
5208
5209 * For cross-compilation you should now set `BUILD_CC' instead of `HOST_CC'.
5210
5211 * New header file <fstab.h> and new functions `getfsspec', `getfsent' and
5212 friends, for parsing /etc/fstab. This code comes from 4.4 BSD.
5213
5214 * The new function `daemon' from 4.4 BSD is useful for server programs that
5215 want to put themselves in the background.
5216
5217 * Joel Sherrill has contributed support for several standalone boards that
5218 run without an operating system.
5219
5220 * `printf', `scanf' and friends now accept a `q' type modifier for long
5221 long int as well as `ll'. Formats using these might be `%qu' or `%lld'.
5222
5223 * All of the code taken from BSD (notably most of the math and networking
5224 routines) has been updated from the BSD 4.4-Lite release.
5225
5226 * The resolver code has been updated from the BIND-4.9.3-BETA9 release.
5227
5228 * The new functions `getdomainname' and `setdomainname' fetch or change the
5229 YP/NIS domain name. These are system calls which exist on systems which
5230 have YP (aka NIS).
5231
5232 * The time zone data files have been updated for the latest international
5233 conventions.
5234
5235 * The SunRPC programs `portmap' and `rpcinfo' are now installed in
5236 $(sbindir) (usually /usr/local/sbin) instead of $(bindir).
5237 \f
5238 Version 1.08
5239
5240 * The C library now includes support for Sun RPC, from Sun's free
5241 RPCSRC-4.0 distribution. The `portmap', `rpcinfo', and `rpcgen' programs
5242 are included. (There is still no support for YP.)
5243
5244 * Tom Quinn has contributed a port of the C library to SGI machines running
5245 Irix 4 (mips-sgi-irix4).
5246
5247 * The new `lockf' function is a simplified interface to the locking
5248 facilities of `fcntl', included for compatibility.
5249
5250 * New time functions `timegm', `timelocal', and `dysize' for compatibility.
5251
5252 * New header file <sys/timeb.h> and new function `ftime' for compatibility.
5253
5254 * New header files <poll.h> and <sys/poll.h> and new function `poll' for
5255 compatibility.
5256
5257 * The error message printed by `assert' for a failed assertion now includes
5258 the name of the program (if using GNU ld) and the name of the calling
5259 function (with versions of GCC that support this).
5260
5261 * The `psignal' function is now declared in <signal.h>, not <stdio.h>.
5262
5263 * The library now includes the <sys/mman.h> header file and memory
5264 management functions `mmap', `munmap', `mprotect', `msync', and
5265 `madvise', on systems that support those facilities.
5266
5267 * The interface for `mcheck' has changed slightly: the function called to
5268 abort the program when an allocation inconsistency is detected now takes
5269 an argument that indicates the type of failure. The new function
5270 `mprobe' lets you request a consistency check for a particular block at
5271 any time (checks are normally done only when you call `free' or `realloc'
5272 on a block).
5273
5274 * It is now possible to easily cross-compile the C library, building on one
5275 system a library to run on another machine and/or operating system. All
5276 you need to do is set the variable `HOST_CC' in `configparms' to the
5277 native compiler for programs to run on the machine you are building on (a
5278 few generator programs are used on Unix systems); set `CC' to the
5279 cross-compiler.
5280
5281 * The new function `fexecve' (only implemented on the GNU system) executes
5282 a program file given a file descriptor already open on the file.
5283 \f
5284 Version 1.07
5285
5286 * Brendan Kehoe has contributed most of a port to the DEC Alpha
5287 running OSF/1 (alpha-dec-osf1). He says it is 75% complete.
5288
5289 * You can set the variable `libprefix' in `configparms' to specify a prefix
5290 to be prepended to installed library files; this makes it easy to install
5291 the GNU C library to be linked as `-lgnuc' or whatever.
5292
5293 * The new `stpncpy' is a cross between `stpcpy' and `strncpy': It
5294 copies a limited number of characters from a string, and returns the
5295 address of the last character written.
5296
5297 * You no longer need to check for whether the installed `stddef.h' is
5298 compatible with the GNU C library. configure now checks for you.
5299
5300 * You can now define a per-stream `fileno' function to convert the
5301 stream's cookie into an integral file descriptor.
5302
5303 * ``malloc (0)'' no longer returns a null pointer. Instead, it
5304 allocates zero bytes of storage, and returns a unique pointer which
5305 you can pass to `realloc' or `free'. The behavior is undefined if
5306 you dereference this pointer.
5307
5308 * The C library now runs on Sony NEWS m68k machines running either
5309 NewsOS 3 or NewsOS 4.
5310
5311 * The new `syscall' function is a system-dependent primitive function
5312 for invoking system calls. It has the canonical behavior on Unix
5313 systems, including unreliable return values for some calls (such as
5314 `pipe', `fork' and `getppid').
5315
5316 * The error code `EWOULDBLOCK' is now obsolete; it is always defined
5317 to `EAGAIN', which is the preferred name. On systems whose kernels
5318 use two distinct codes, the C library now translates EWOULDBLOCK to
5319 EAGAIN in every system call function.
5320 \f
5321 Version 1.06
5322
5323 * The GNU C Library Reference Manual is now distributed with the library.
5324 `make dvi' will produce a DVI file of the printed manual.
5325 `make info' will produce Info files that you can read on line using C-h i
5326 in Emacs or the `info' program.
5327 Please send comments on the manual to bug-glibc-manual@gnu.org.
5328
5329 * The library now supports SVR4 on i386s (i386-unknown-sysv4).
5330
5331 * Brendan Kehoe has contributed a port to Sun SPARCs running Solaris 2.
5332
5333 * Jason Merrill has contributed a port to the Sequent Symmetry running
5334 Dynix version 3 (i386-sequent-dynix).
5335
5336 * The library has been ported to i386s running SCO 3.2.4 (also known as SCO
5337 ODT 2.0; i386-unknown-sco3.2.4) or SCO 3.2 (i386-unknown-sco3.2).
5338
5339 * New function `memory_warnings' lets you arrange to get warnings when
5340 malloc is running out of memory to allocate, like Emacs gives you.
5341
5342 * The C library now contains the relocating allocator used in Emacs 19 for
5343 its editing buffers. This allocator (ralloc) minimizes allocation
5344 overhead and fragmentation by moving allocated regions around whenever it
5345 needs to. You always refer to a ralloc'd region with a "handle" (a
5346 pointer to a pointer--an object of type `void **').
5347
5348 * There is a new `printf' format: `%m' gives you the string corresponding
5349 to the error code in `errno'.
5350
5351 * In `scanf' formats, you can now use `%as' or `%a[' to do the normal `%s'
5352 or `%[' conversion, but instead of filling in a fixed-sized buffer you
5353 pass, the `a' modifier says to fill in a `char **' you pass with a
5354 malloc'd string.
5355
5356 * The `fnmatch' function supports the new flag bits `FNM_LEADING_DIR' and
5357 `FNM_CASEFOLD'. `FNM_LEADING_DIR' lets a pattern like `foo*' match a
5358 name like `foo/bar'. `FNM_CASEFOLD' says to ignore case in matching.
5359
5360 * `mkstemp' is a traditional Unix function to atomically create and open a
5361 uniquely-named temporary file.
5362 \f
5363 Version 1.05
5364
5365 * The standard location for the file that says what the local timezone is
5366 has changed again. It is now `/usr/local/etc/localtime' (or more
5367 precisely, `${prefix}/etc/localtime') rather than `/etc/localtime'.
5368
5369 * The distribution no longer contains any files with names longer than 14
5370 characters.
5371
5372 * `struct ttyent' has two new flag bits: TTY_TRUSTED and TTY_CONSOLE.
5373 These are set by the new `trusted' and `console' keywords in `/etc/ttys'.
5374
5375 * New functions `ttyslot' and `syslog' from 4.4 BSD.
5376 \f
5377 Version 1.04
5378
5379 * The configuration process has changed quite a bit. The `configure'
5380 script is now used just like the configuration scripts for other GNU
5381 packages. The `sysdeps' directory hierarchy is much rearranged.
5382 The file `INSTALL' explains the new scheme in detail.
5383
5384 * The header files no longer need to be processed into ANSI C and
5385 traditional C versions. There is just one set of files to install, and
5386 it will work with ANSI or old C compilers (including `gcc -traditional').
5387
5388 * Brendan Kehoe and Ian Lance Taylor have ported the library to the
5389 MIPS DECStation running Ultrix 4.
5390
5391 * The Sun 4 startup code (crt0) can now properly load SunOS 4 shared libraries.
5392 Tom Quinn contributed the initial code. The GNU C library can NOT yet be
5393 made itself into a shared library.
5394
5395 * Yet further improved support for the i386, running 4.3 BSD-like systems
5396 (such as Mach 3 with the Unix single-server), or System V.
5397
5398 * New function `strncasecmp' to do case-insensitive string comparison
5399 with limited length.
5400
5401 * New function `strsep' is a reentrant alternative to `strtok'.
5402
5403 * New functions `scandir' and `alphasort' for searching directories.
5404
5405 * New function `setenv' is a better interface to `putenv'.
5406
5407 * Ian Lance Taylor has contributed an implementation of the SVID `ftw'
5408 function for traversing a directory tree.
5409
5410 * The GNU obstack package is now also part of the C library.
5411 The new function `open_obstack_stream' creates a stdio stream that
5412 writes onto an obstack; `obstack_printf' and `obstack_vprintf' do
5413 formatted output directly to an obstack.
5414
5415 * Miscellaneous new functions: reboot, nice, sigaltstack (4.4 BSD only),
5416 cfmakeraw, getusershell, getpass, swab, getttyent, seteuid, setegid.
5417
5418 * `FNM_FILE_NAME' is another name for `FNM_PATHNAME', used with `fnmatch'.
5419
5420 * The new functions `strfry' and `memfrob' do mysterious and wonderful
5421 things to your strings.
5422
5423 * There are some new test programs: test-fseek, testmb, and testrand.
5424
5425 * Some work has been done to begin porting the library to 4.4 BSD and Linux.
5426 These ports are not finished, but are a good starting place for really
5427 supporting those systems.
5428
5429 * `/etc/localtime' is now the standard location for the file that says what
5430 the local timezone is, rather than `/usr/local/lib/zoneinfo/localtime'.
5431 This follows the general principle that `/etc' is the place for all local
5432 configuration files.
5433
5434 * The C library header files now use `extern "C"' when used by the C++
5435 compiler, so the C library should now work with C++ code.
5436
5437 * The header file <bstring.h> is gone. <string.h> now declares bcopy,
5438 bcmp, bzero, and ffs. (Update: nowadays these functions are declared
5439 in <strings.h>.)
5440
5441 * Mike Haertel (of GNU e?grep and malloc fame) has written a new sorting
5442 function which uses the `merge sort' algorithm, and is said to be
5443 significantly faster than the old GNU `qsort' function. Merge sort is
5444 now the standard `qsort' function. The new algorithm can require a lot
5445 of temporary storage; so, the old sorting function is called when the
5446 required storage is not available.
5447
5448 * The C library now includes Michael Glad's Ultra Fast Crypt, which
5449 provides the Unix `crypt' function, plus some other entry points.
5450
5451 * The code and header files taken from 4.4 BSD have been updated with the
5452 latest files released from Berkeley.
5453 \f
5454 ----------------------------------------------------------------------
5455 Copyright information:
5456
5457 Copyright (C) 1992-2019 Free Software Foundation, Inc.
5458
5459 Permission is granted to anyone to make or distribute verbatim copies
5460 of this document as received, in any medium, provided that the
5461 copyright notice and this permission notice are preserved,
5462 thus giving the recipient permission to redistribute in turn.
5463
5464 Permission is granted to distribute modified versions
5465 of this document, or of portions of it,
5466 under the above conditions, provided also that they
5467 carry prominent notices stating who last changed them.
5468 \f
5469 Local variables:
5470 version-control: never
5471 fill-column: 76
5472 End: