]> git.ipfire.org Git - thirdparty/glibc.git/blob - NEWS
x86-64: Require BMI2 for __strlen_evex and __strnlen_evex
[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 2.29.1
9
10 Major new features:
11
12 * The entry for the new Japanese era has been added for ja_JP locale.
13
14 Deprecated and removed features, and other changes affecting compatibility:
15
16 * The copy_file_range function fails with ENOSYS if the kernel does not
17 support the system call of the same name. Previously, user space
18 emulation was performed, but its behavior did not match the kernel
19 behavior, which was deemed too confusing. Applications which use the
20 copy_file_range function will have to be run on kernels which implement
21 the copy_file_range system call. Support for most architectures was added
22 in version 4.5 of the mainline Linux kernel.
23
24 The following bugs are resolved with this release:
25
26 [16573] malloc: Set and reset all hooks for tracing
27 [18035] Fix pldd hang
28 [20019] NULL pointer dereference in libc.so.6 IFUNC due to uninitialized GOT
29 [20568] Fix crash in _IO_wfile_sync
30 [24155] x32 memcmp can treat positive length as 0 (if sign bit in RDX is set) (CVE-2019-7309)
31 [24164] Systemtap probes need to use "nr" constraint on 32-bit Arm
32 [24161] __run_fork_handlers self-deadlocks in malloc/tst-mallocfork2
33 [24211] Use-after-free in Systemtap probe in pthread_join
34 [24228] old x86 applications that use legacy libio crash on exit
35 [24476] dlfcn: Guard __dlerror_main_freeres with __libc_once_get (once)
36 [24744] io: Remove the copy_file_range emulation
37 [24794] Partial test suite run builds corrupt test-in-container testroot
38 [24986] alpha: new getegid, geteuid and getppid syscalls used
39 unconditionally
40 [25203] libio: Disable vtable validation for pre-2.1 interposed handles
41 [25204] Ignore LD_PREFER_MAP_32BIT_EXEC for SUID programs
42 [25225] ld.so fails to link on x86 if GCC defaults to -fcf-protection
43 [25232] No const correctness for strchr et al. for Clang++
44 [25933] Off by one error in __strncmp_avx2
45 [25976] nss_compat: internal_end*ent may clobber errno, hiding ERANGE
46 [27130] "rep movsb" performance issue
47 [27177] GLIBC_TUNABLES=glibc.cpu.x86_ibt=on:glibc.cpu.x86_shstk=on doesn't work
48 [27457] vzeroupper use in AVX2 multiarch string functions cause HTM aborts
49 [28755] overflow bug in wcsncmp_avx2 and wcsncmp_evex
50
51 Security related changes:
52
53 CVE-2019-7309: x86-64 memcmp used signed Jcc instructions to check
54 size. For x86-64, memcmp on an object size larger than SSIZE_MAX
55 has undefined behavior. On x32, the size_t argument may be passed
56 in the lower 32 bits of the 64-bit RDX register with non-zero upper
57 32 bits. When it happened with the sign bit of RDX register set,
58 memcmp gave the wrong result since it treated the size argument as
59 zero. Reported by H.J. Lu.
60
61 CVE-2019-9169: Attempted case-insensitive regular-expression match
62 via proceed_next_node in posix/regexec.c leads to heap-based buffer
63 over-read. Reported by Hongxu Chen.
64
65 CVE-2019-19126: ld.so failed to ignore the LD_PREFER_MAP_32BIT_EXEC
66 environment variable during program execution after a security
67 transition, allowing local attackers to restrict the possible mapping
68 addresses for loaded libraries and thus bypass ASLR for a setuid
69 program. Reported by Marcin Koƛcielnicki.
70
71 \f
72 Version 2.29
73
74 Major new features:
75
76 * The getcpu wrapper function has been added, which returns the currently
77 used CPU and NUMA node. This function is Linux-specific.
78
79 * A new convenience target has been added for distribution maintainers
80 to build and install all locales as directories with files. The new
81 target is run by issuing the following command in your build tree:
82 'make localedata/install-locale-files', with an optional DESTDIR
83 to set the install root if you wish to install into a non-default
84 configured location.
85
86 * Optimized generic exp, exp2, log, log2, pow, sinf, cosf, sincosf and tanf.
87
88 * The reallocarray function is now declared under _DEFAULT_SOURCE, not just
89 for _GNU_SOURCE, to match BSD environments.
90
91 * For powercp64le ABI, Transactional Lock Elision is now enabled iff kernel
92 indicates that it will abort the transaction prior to entering the kernel
93 (PPC_FEATURE2_HTM_NOSC on hwcap2). On older kernels the transaction is
94 suspended, and this caused some undefined side-effects issues by aborting
95 transactions manually. Glibc avoided it by abort transactions manually on
96 each syscall, but it lead to performance issues on newer kernels where the
97 HTM state is saved and restore lazily (the state being saved even when the
98 process actually does not use HTM).
99
100 * The functions posix_spawn_file_actions_addchdir_np and
101 posix_spawn_file_actions_addfchdir_np have been added, enabling
102 posix_spawn and posix_spawnp to run the new process in a different
103 directory. These functions are GNU extensions. The function
104 posix_spawn_file_actions_addchdir_np is similar to the Solaris function
105 of the same name.
106
107 * The popen and system do not run atfork handlers anymore (BZ#17490).
108 Although it is a possible POSIX violation, the POSIX rationale in
109 pthread_atfork documentation regarding atfork handlers is to handle
110 inconsistent mutex state after a fork call in a multi-threaded process.
111 In both popen and system there is no direct access to user-defined mutexes.
112
113 * Support for the C-SKY ABIV2 running on Linux has been added. This port
114 requires at least binutils-2.32, gcc-9.0, and linux-4.20. Two ABIs are
115 supported:
116 - C-SKY ABIV2 soft-float little-endian
117 - C-SKY ABIV2 hard-float little-endian
118
119 * strftime's default formatting of a locale's alternative year (%Ey)
120 has been changed to zero-pad the year to a minimum of two digits,
121 like "%y". This improves the display of Japanese era years during
122 the first nine years of a new era, and is expected to be harmless
123 for all other locales (only Japanese locales regularly have
124 alternative year numbers less than 10). Zero-padding can be
125 overridden with the '_' or '-' flags (which are GNU extensions).
126
127 * As a GNU extension, the '_' and '-' flags can now be applied to
128 "%EY" to control how the year number is formatted; they have the
129 same effect that they would on "%Ey".
130
131 Deprecated and removed features, and other changes affecting compatibility:
132
133 * The glibc.tune tunable namespace has been renamed to glibc.cpu and the
134 tunable glibc.tune.cpu has been renamed to glibc.cpu.name.
135
136 * The type of the pr_uid and pr_gid members of struct elf_prpsinfo, defined
137 in <sys/procfs.h>, has been corrected to match the type actually used by
138 the Linux kernel. This affects the size and layout of that structure on
139 MicroBlaze, MIPS (n64 ABI only), Nios II and RISC-V.
140
141 * For the MIPS n32 ABI, the type of the pr_sigpend and pr_sighold members of
142 struct elf_prstatus, and the pr_flag member of struct elf_prpsinfo,
143 defined in <sys/procfs.h>, has been corrected to match the type actually
144 used by the Linux kernel. This affects the size and layout of those
145 structures.
146
147 * An archaic GNU extension to scanf, under which '%as', '%aS', and '%a[...]'
148 meant to scan a string and allocate space for it with malloc, is now
149 restricted to programs compiled in C89 or C++98 mode with _GNU_SOURCE
150 defined. This extension conflicts with C99's use of '%a' to scan a
151 hexadecimal floating-point number, which is now available to programs
152 compiled as C99 or C++11 or higher, regardless of _GNU_SOURCE.
153
154 POSIX.1-2008 includes the feature of allocating a buffer for string input
155 with malloc, using the modifier letter 'm' instead. Programs using
156 '%as', '%aS', or '%a[...]' with the old GNU meaning should change to
157 '%ms', '%mS', or '%m[...]' respectively. Programs that wish to use the
158 C99 '%a' no longer need to avoid _GNU_SOURCE.
159
160 GCC's -Wformat warnings can detect most uses of this extension, as long
161 as all functions that call vscanf, vfscanf, or vsscanf are annotated with
162 __attribute__ ((format (scanf, ...))).
163
164 Changes to build and runtime requirements:
165
166 * Python 3.4 or later is required to build the GNU C Library.
167
168 * On most architectures, GCC 5 or later is required to build the GNU C
169 Library. (On powerpc64le, GCC 6.2 or later is still required, as before.)
170
171 Older GCC versions and non-GNU compilers are still supported when
172 compiling programs that use the GNU C Library.
173
174 Security related changes:
175
176 CVE-2018-19591: A file descriptor leak in if_nametoindex can lead to a
177 denial of service due to resource exhaustion when processing getaddrinfo
178 calls with crafted host names. Reported by Guido Vranken.
179
180 CVE-2019-6488: On x32, the size_t parameter may be passed in the lower
181 32 bits of a 64-bit register with with non-zero upper 32 bit. When it
182 happened, accessing the 32-bit size_t value as the full 64-bit register
183 in the assembly string/memory functions would cause a buffer overflow.
184 Reported by H.J. Lu.
185
186 CVE-2016-10739: The getaddrinfo function could successfully parse IPv4
187 addresses with arbitrary trailing characters, potentially leading to data
188 or command injection issues in applications.
189
190 The following bugs are resolved with this release:
191
192 [10425] localedata: it_IT/it_CH: LC_TIME format is wrong
193 [10496] localedata: 12h time representation in multiple locales faulty
194 [10797] localedata: it_IT locale numeric does not have a separator for
195 thousands
196 [11319] libc: dprintf doesn't handle errors properly
197 [16346] time: mktime: potentially unsafe use of localtime_offset
198 [17248] build: glibc should not sort CFLAGS (support gcc plugins and
199 --param options)
200 [17405] libc: Implement posix_spawn_file_actions_addchdir_np,
201 posix_spawn_file_actions_addfchdir_np
202 [17426] localedata: Indian locales: set the correct date format
203 [17490] stdio: popen should not invoke atfork handlers
204 [17783] libc: TIOCSER_TEMT conditions inconsistent
205 [18040] regex: use-after-free in regexec/get_subexp
206 [18093] libc: Corrupted aux-cache causes ldconfig to segfault
207 [20018] network: getaddrinfo should reject IP addresses with trailing
208 characters (CVE-2016-10739)
209 [20209] localedata: Spelling mistake for Sunday in Greenlandic kl_GL
210 [20271] libc: Missing "\n" in __libc_fatal calls
211 [20480] dynamic-link: Patch: ifunc not executable, crashes sudo qemu
212 [20544] libc: RFE: atexit, __cxa_atexit, on_exit should assert function
213 pointer argument is non-NULL
214 [21037] stdio: open_memstream and freopen
215 [21286] libc: bits/siginfo.h is missing enum definition for TRAP_HWBKPT
216 [21716] time: Crash in glibc's mktime in low-memory situations
217 [22834] stdio: Subprocess forked by popen may crash in Linux when
218 multithreads call popen
219 [22927] network: crash in vn_gai_enqueue_request if requests_tail was NULL
220 and pthread_create fails.
221 [23032] hurd: sysdeps/htl/pt-barrier-init.c:39: bad call to memcmp ?
222 [23125] libc: riscv64: endless loop when throwing an exception from a
223 constructor
224 [23275] nptl: Race in pthread_mutex_lock while promoting to
225 PTHREAD_MUTEX_ELISION_NP.
226 [23400] libc: stdlib/test-bz22786.c creates temporary files in glibc
227 source tree
228 [23479] math: [mips] bits/fenv.h should not define some macros for soft-
229 float
230 [23490] libc: sysdeps/unix/sysv/linux/x86/tst-cet-property-2.c:49: off by
231 one error
232 [23497] libc: readdir64@GLIBC_2.1 cannot parse the kernel directory stream
233 [23509] dynamic-link: CET enabled glibc is incompatible with the older
234 linker
235 [23520] nscd: nscd: Use-after-free in addgetnetgrentX and its callers
236 [23521] nss: get_next_alias nss_files file stream leak
237 [23538] nptl: Hang in pthread_cond_broadcast
238 [23562] libc: Wrong type for si_band in Linux-specific siginfo_t
239 [23578] regex: Invalid memory access if regex pattern contains NUL byte
240 [23579] libc: Errors misreported in preadv2
241 [23597] build: support/test-container.c doesn't work with different
242 filesystems
243 [23603] time: mktime signed integer overflow on large timestamps
244 [23606] libc: Missing ENDBR32 in sysdeps/i386/start.S
245 [23614] libc: powerpc: missing CFI register information in __mpn_*
246 functions
247 [23637] string: Generic strstr/strcasestr fails with huge needles
248 [23640] libc: no way to easily clear FD_CLOEXEC in
249 posix_spawn_file_actions_adddup2()
250 [23649] libc: [microblaze/mips/nios2/riscv] sys/procfs.h pr_uid, pr_gid
251 have wrong type
252 [23656] libc: [mips n32] sys/procfs.h pr_sigpend, pr_sighold, pr_flag have
253 wrong type
254 [23679] libc: gethostid: Missing NULL check for gethostbyname_r result
255 [23689] libc: Bug in documentation for rusage.ru_ixrss in
256 bits/types/struct_rusage.h
257 [23690] dynamic-link: Segfault in _dl_profile_fixup with a high number of
258 threads
259 [23707] dynamic-link: Missing unwind info in sysdeps/powerpc/powerpc32/dl-
260 start.S
261 [23709] string: glibc 2.25 lacks sse2 optimized strstr()
262 [23716] dynamic-link: _dl_runtime_resolve_shstk isn't selected properly
263 [23717] libc: glibc: stdlib/tst-setcontext9 test suite failure on
264 powerpc64le
265 [23724] localedata: Albanian date formats are incorrect
266 [23735] math: libnldbl_nonshared.a references internal libm symbols
267 [23740] localedata: kl_GL: Month names and date formats need update
268 [23744] regex: regex refactorings to remove BE, avoid duplication
269 [23745] time: mktime fix for Gnulib + coreutils
270 [23758] time: Improve the width of alternate representation for year in
271 strftime
272 [23783] libc: [mips] Missing CMSPAR bits/termios.h
273 [23789] time: mktime does not set errno on failure
274 [23791] localedata: Wrong monetary format for ca_ES locale
275 [23793] locale: c32rtomb and mbrtoc32 should not alias wcrtomb and mbrtowc
276 [23794] locale: c16rtomb does not handle surrogate pairs
277 [23821] libc: si_band in siginfo_t has wrong type long int on sparc64
278 [23822] math: ia64 static libm.a is missing exp2f, log2f and powf symbols
279 [23836] time: time/tst-mktime2 test failure on Arm (32-bit)
280 [23848] libc: [sparc] Some socket syscalls wrongly assumed to be present
281 [23861] nptl: rdlock stalls indefinitely on an unlocked pthread rwlock
282 [23862] libc: [sh] missing kernel-features.h undefines
283 [23864] libc: [riscv] missing kernel-features.h undefines
284 [23867] libc: [arm/microblaze] __ASSUME_MLOCK2 incorrect
285 [23907] malloc: Incorrect double-free malloc tcache check disregards
286 tcache size
287 [23913] libc: off-by-one in function maybe_script_execute in
288 sysdeps/posix/spawni.c
289 [23915] libc: [arm] __ASSUME_COPY_FILE_RANGE incorrect
290 [23923] locale: Add --no-hard-links option to localedef
291 [23927] network: Linux if_nametoindex() does not close descriptor
292 (CVE-2018-19591)
293 [23961] math: powf can overflow to inf without setting errno in non-
294 nearest rounding mode
295 [23967] libc: [2.28 Regression]: New sigaction implementation breaks m68k
296 [23972] libc: __old_getdents64 uses wrong d_off value on overflow
297 [23993] libc: glibc 2.29 doesn't build with gcc 4.9
298 [23995] localedata: Remove execution flags from localedata/locales/bi_VU
299 [24011] localedata: Fixed small type in comment for locale bs_BA
300 [24018] libc: gettext() may return NULL
301 [24022] build: riscv build failure with Linux kernel 4.20-rc7
302 [24023] build: [2.29 Regression] FAIL: elf/check-localplt
303 [24024] string: strerror() might set errno to ENOMEM due to -fno-math-
304 error
305 [24027] malloc: glibc: realloc() ncopies 32-bit integer overflow
306 [24034] libc: tst-cancel21-static fails with SIGBUS on pre-ARMv7 when
307 using GCC 8
308 [24046] localedata: en_US locale doesn't define date_fmt
309 [24063] manual: @var{errno} should be @code{errno}
310 [24066] soft-fp: Inconsistent _FP_W_TYPE_SIZE check
311 [24088] libc: VSCR field is not being correctly read in ucontext_t on
312 ppc64le
313 [24096] time: Specifying '_' or '-' flag for "%EY" does not produce the
314 expected result
315 [24097] string: Can't use 64-bit register for size_t in assembly codes for
316 x32 (CVE-2019-6488)
317 [24110] hurd: SS_DISABLE never set in stack_t value returned by
318 sigaltstack
319 [24112] network: Do not send DNS queries for non-host names (where all
320 answers will be rejected)
321 [24130] libc: alpha __remqu corrupts $f3 register
322
323 \f
324 Version 2.28
325
326 Major new features:
327
328 * The localization data for ISO 14651 is updated to match the 2016
329 Edition 4 release of the standard, this matches data provided by
330 Unicode 9.0.0. This update introduces significant improvements to the
331 collation of Unicode characters. This release deviates slightly from
332 the standard in that the collation element ordering for lowercase and
333 uppercase LATIN script characters is adjusted to ensure that regular
334 expressions with ranges like [a-z] and [A-Z] don't interleave e.g. A
335 is not matched by [a-z]. With the update many locales have been
336 updated to take advantage of the new collation information. The new
337 collation information has increased the size of the compiled locale
338 archive or binary locales.
339
340 * The GNU C Library can now be compiled with support for Intel CET, AKA
341 Intel Control-flow Enforcement Technology. When the library is built
342 with --enable-cet, the resulting glibc is protected with indirect
343 branch tracking (IBT) and shadow stack (SHSTK). CET-enabled glibc is
344 compatible with all existing executables and shared libraries. This
345 feature is currently supported on i386, x86_64 and x32 with GCC 8 and
346 binutils 2.29 or later. Note that CET-enabled glibc requires CPUs
347 capable of multi-byte NOPs, like x86-64 processors as well as Intel
348 Pentium Pro or newer. NOTE: --enable-cet has been tested for i686,
349 x86_64 and x32 on non-CET processors. --enable-cet has been tested
350 for x86_64 and x32 on CET SDVs, but Intel CET support hasn't been
351 validated for i686.
352
353 * The GNU C Library now has correct support for ABSOLUTE symbols
354 (SHN_ABS-relative symbols). Previously such ABSOLUTE symbols were
355 relocated incorrectly or in some cases discarded. The GNU linker can
356 make use of the newer semantics, but it must communicate it to the
357 dynamic loader by setting the ELF file's identification (EI_ABIVERSION
358 field) to indicate such support is required.
359
360 * Unicode 11.0.0 Support: Character encoding, character type info, and
361 transliteration tables are all updated to Unicode 11.0.0, using
362 generator scripts contributed by Mike FABIAN (Red Hat).
363
364 * <math.h> functions that round their results to a narrower type are added
365 from TS 18661-1:2014 and TS 18661-3:2015:
366
367 - fadd, faddl, daddl and corresponding fMaddfN, fMaddfNx, fMxaddfN and
368 fMxaddfNx functions.
369
370 - fsub, fsubl, dsubl and corresponding fMsubfN, fMsubfNx, fMxsubfN and
371 fMxsubfNx functions.
372
373 - fmul, fmull, dmull and corresponding fMmulfN, fMmulfNx, fMxmulfN and
374 fMxmulfNx functions.
375
376 - fdiv, fdivl, ddivl and corresponding fMdivfN, fMdivfNx, fMxdivfN and
377 fMxdivfNx functions.
378
379 * Two grammatical forms of month names are now supported for the following
380 languages: Armenian, Asturian, Catalan, Czech, Kashubian, Occitan, Ossetian,
381 Scottish Gaelic, Upper Sorbian, and Walloon. The following languages now
382 support two grammatical forms in abbreviated month names: Catalan, Greek,
383 and Kashubian.
384
385 * Newly added locales: Lower Sorbian (dsb_DE) and Yakut (sah_RU) also
386 include the support for two grammatical forms of month names.
387
388 * Building and running on GNU/Hurd systems now works without out-of-tree
389 patches.
390
391 * The renameat2 function has been added, a variant of the renameat function
392 which has a flags argument. If the flags are zero, the renameat2 function
393 acts like renameat. If the flag is not zero and there is no kernel
394 support for renameat2, the function will fail with an errno value of
395 EINVAL. This is different from the existing gnulib function renameatu,
396 which performs a plain rename operation in case of a RENAME_NOREPLACE
397 flags and a non-existing destination (and therefore has a race condition
398 that can clobber the destination inadvertently).
399
400 * The statx function has been added, a variant of the fstatat64
401 function with an additional flags argument. If there is no direct
402 kernel support for statx, glibc provides basic stat support based on
403 the fstatat64 function.
404
405 * IDN domain names in getaddrinfo and getnameinfo now use the system libidn2
406 library if installed. libidn2 version 2.0.5 or later is recommended. If
407 libidn2 is not available, internationalized domain names are not encoded
408 or decoded even if the AI_IDN or NI_IDN flags are passed to getaddrinfo or
409 getnameinfo. (getaddrinfo calls with non-ASCII names and AI_IDN will fail
410 with an encoding error.) Flags which used to change the IDN encoding and
411 decoding behavior (AI_IDN_ALLOW_UNASSIGNED, AI_IDN_USE_STD3_ASCII_RULES,
412 NI_IDN_ALLOW_UNASSIGNED, NI_IDN_USE_STD3_ASCII_RULES) have been
413 deprecated. They no longer have any effect.
414
415 * Parsing of dynamic string tokens in DT_RPATH, DT_RUNPATH, DT_NEEDED,
416 DT_AUXILIARY, and DT_FILTER has been expanded to support the full
417 range of ELF gABI expressions including such constructs as
418 '$ORIGIN$ORIGIN' (if valid). For SUID/GUID applications the rules
419 have been further restricted, and where in the past a dynamic string
420 token sequence may have been interpreted as a literal string it will
421 now cause a load failure. These load failures were always considered
422 unspecified behaviour from the perspective of the dynamic loader, and
423 for safety are now load errors e.g. /foo/${ORIGIN}.so in DT_NEEDED
424 results in a load failure now.
425
426 * Support for ISO C threads (ISO/IEC 9899:2011) has been added. The
427 implementation includes all the standard functions provided by
428 <threads.h>:
429
430 - thrd_current, thrd_equal, thrd_sleep, thrd_yield, thrd_create,
431 thrd_detach, thrd_exit, and thrd_join for thread management.
432
433 - mtx_init, mtx_lock, mtx_timedlock, mtx_trylock, mtx_unlock, and
434 mtx_destroy for mutual exclusion.
435
436 - call_once for function call synchronization.
437
438 - cnd_broadcast, cnd_destroy, cnd_init, cnd_signal, cnd_timedwait, and
439 cnd_wait for conditional variables.
440
441 - tss_create, tss_delete, tss_get, and tss_set for thread-local storage.
442
443 Application developers must link against libpthread to use ISO C threads.
444
445 Deprecated and removed features, and other changes affecting compatibility:
446
447 * The nonstandard header files <libio.h> and <_G_config.h> are no longer
448 installed. Software that was using either header should be updated to
449 use standard <stdio.h> interfaces instead.
450
451 * The stdio functions 'getc' and 'putc' are no longer defined as macros.
452 This was never required by the C standard, and the macros just expanded
453 to call alternative names for the same functions. If you hoped getc and
454 putc would provide performance improvements over fgetc and fputc, instead
455 investigate using (f)getc_unlocked and (f)putc_unlocked, and, if
456 necessary, flockfile and funlockfile.
457
458 * All stdio functions now treat end-of-file as a sticky condition. If you
459 read from a file until EOF, and then the file is enlarged by another
460 process, you must call clearerr or another function with the same effect
461 (e.g. fseek, rewind) before you can read the additional data. This
462 corrects a longstanding C99 conformance bug. It is most likely to affect
463 programs that use stdio to read interactive input from a terminal.
464 (Bug #1190.)
465
466 * The macros 'major', 'minor', and 'makedev' are now only available from
467 the header <sys/sysmacros.h>; not from <sys/types.h> or various other
468 headers that happen to include <sys/types.h>. These macros are rarely
469 used, not part of POSIX nor XSI, and their names frequently collide with
470 user code; see https://sourceware.org/bugzilla/show_bug.cgi?id=19239 for
471 further explanation.
472
473 <sys/sysmacros.h> is a GNU extension. Portable programs that require
474 these macros should first include <sys/types.h>, and then include
475 <sys/sysmacros.h> if __GNU_LIBRARY__ is defined.
476
477 * The tilegx*-*-linux-gnu configurations are no longer supported.
478
479 * The obsolete function ustat is no longer available to newly linked
480 binaries; the headers <ustat.h> and <sys/ustat.h> have been removed. This
481 function has been deprecated in favor of fstatfs and statfs.
482
483 * The obsolete function nfsservctl is no longer available to newly linked
484 binaries. This function was specific to systems using the Linux kernel
485 and could not usefully be used with the GNU C Library on systems with
486 version 3.1 or later of the Linux kernel.
487
488 * The obsolete function name llseek is no longer available to newly linked
489 binaries. This function was specific to systems using the Linux kernel
490 and was not declared in a header. Programs should use the lseek64 name
491 for this function instead.
492
493 * The AI_IDN_ALLOW_UNASSIGNED and NI_IDN_ALLOW_UNASSIGNED flags for the
494 getaddrinfo and getnameinfo functions have been deprecated. The behavior
495 previously selected by them is now always enabled.
496
497 * The AI_IDN_USE_STD3_ASCII_RULES and NI_IDN_USE_STD3_ASCII_RULES flags for
498 the getaddrinfo and getnameinfo functions have been deprecated. The STD3
499 restriction (rejecting '_' in host names, among other things) has been
500 removed, for increased compatibility with non-IDN name resolution.
501
502 * The fcntl function now have a Long File Support variant named fcntl64. It
503 is added to fix some Linux Open File Description (OFD) locks usage on non
504 LFS mode. As for others *64 functions, fcntl64 semantics are analogous with
505 fcntl and LFS support is handled transparently. Also for Linux, the OFD
506 locks act as a cancellation entrypoint.
507
508 * The obsolete functions encrypt, encrypt_r, setkey, setkey_r, cbc_crypt,
509 ecb_crypt, and des_setparity are no longer available to newly linked
510 binaries, and the headers <rpc/des_crypt.h> and <rpc/rpc_des.h> are no
511 longer installed. These functions encrypted and decrypted data with the
512 DES block cipher, which is no longer considered secure. Software that
513 still uses these functions should switch to a modern cryptography library,
514 such as libgcrypt.
515
516 * Reflecting the removal of the encrypt and setkey functions above, the
517 macro _XOPEN_CRYPT is no longer defined. As a consequence, the crypt
518 function is no longer declared unless _DEFAULT_SOURCE or _GNU_SOURCE is
519 enabled.
520
521 * The obsolete function fcrypt is no longer available to newly linked
522 binaries. It was just another name for the standard function crypt,
523 and it has not appeared in any header file in many years.
524
525 * We have tentative plans to hand off maintenance of the passphrase-hashing
526 library, libcrypt, to a separate development project that will, we hope,
527 keep up better with new passphrase-hashing algorithms. We will continue
528 to declare 'crypt' in <unistd.h>, and programs that use 'crypt' or
529 'crypt_r' should not need to change at all; however, distributions will
530 need to install <crypt.h> and libcrypt from a separate project.
531
532 In this release, if the configure option --disable-crypt is used, glibc
533 will not install <crypt.h> or libcrypt, making room for the separate
534 project's versions of these files. The plan is to make this the default
535 behavior in a future release.
536
537 Changes to build and runtime requirements:
538
539 GNU make 4.0 or later is now required to build glibc.
540
541 Security related changes:
542
543 CVE-2016-6261, CVE-2016-6263, CVE-2017-14062: Various vulnerabilities have
544 been fixed by removing the glibc-internal IDNA implementation and using
545 the system-provided libidn2 library instead. Originally reported by Hanno
546 Böck and Christian Weisgerber.
547
548 CVE-2017-18269: An SSE2-based memmove implementation for the i386
549 architecture could corrupt memory. Reported by Max Horn.
550
551 CVE-2018-11236: Very long pathname arguments to realpath function could
552 result in an integer overflow and buffer overflow. Reported by Alexey
553 Izbyshev.
554
555 CVE-2018-11237: The mempcpy implementation for the Intel Xeon Phi
556 architecture could write beyond the target buffer, resulting in a buffer
557 overflow. Reported by Andreas Schwab.
558
559 The following bugs are resolved with this release:
560
561 [1190] stdio: fgetc()/fread() behaviour is not POSIX compliant
562 [6889] manual: 'PWD' mentioned but not specified
563 [13575] libc: SSIZE_MAX defined as LONG_MAX is inconsistent with ssize_t,
564 when __WORDSIZE != 64
565 [13762] regex: re_search etc. should return -2 on memory exhaustion
566 [13888] build: /tmp usage during testing
567 [13932] math: dbl-64 pow unexpectedly slow for some inputs
568 [14092] nptl: Support C11 threads
569 [14095] localedata: Review / update collation data from Unicode / ISO
570 14651
571 [14508] libc: -Wformat warnings
572 [14553] libc: Namespace pollution loff_t in sys/types.h
573 [14890] libc: Make NT_PRFPREG canonical.
574 [15105] libc: Extra PLT references with -Os
575 [15512] libc: __bswap_constant_16 not compiled when -Werror -Wsign-
576 conversion is given
577 [16335] manual: Feature test macro documentation incomplete and out of
578 date
579 [16552] libc: Unify umount implementations in terms of umount2
580 [17082] libc: htons et al.: statement-expressions prevent use on global
581 scope with -O1 and higher
582 [17343] libc: Signed integer overflow in /stdlib/random_r.c
583 [17438] localedata: pt_BR: wrong d_fmt delimiter
584 [17662] libc: please implement binding for the new renameat2 syscall
585 [17721] libc: __restrict defined as /* Ignore */ even in c11
586 [17979] libc: inconsistency between uchar.h and stdint.h
587 [18018] dynamic-link: Additional $ORIGIN handling issues (CVE-2011-0536)
588 [18023] libc: extend_alloca is broken (questionable pointer comparison,
589 horrible machine code)
590 [18124] libc: hppa: setcontext erroneously returns -1 as exit code for
591 last constant.
592 [18471] libc: llseek should be a compat symbol
593 [18473] soft-fp: [powerpc-nofpu] __sqrtsf2, __sqrtdf2 should be compat
594 symbols
595 [18991] nss: nss_files skips large entry in database
596 [19239] libc: Including stdlib.h ends up with macros major and minor being
597 defined
598 [19463] libc: linknamespace failures when compiled with -Os
599 [19485] localedata: csb_PL: Update month translations + add yesstr/nostr
600 [19527] locale: Normalized charset name not recognized by setlocale
601 [19667] string: Missing Sanity Check for malloc calls in file 'testcopy.c'
602 [19668] libc: Missing Sanity Check for malloc() in file 'tst-setcontext-
603 fpscr.c'
604 [19728] network: out of bounds stack read in libidn function
605 idna_to_ascii_4i (CVE-2016-6261)
606 [19729] network: out of bounds heap read on invalid utf-8 inputs in
607 stringprep_utf8_nfkc_normalize (CVE-2016-6263)
608 [19818] dynamic-link: Absolute (SHN_ABS) symbols incorrectly relocated by
609 the base address
610 [20079] libc: Add SHT_X86_64_UNWIND to elf.h
611 [20251] libc: 32bit programs pass garbage in struct flock for OFD locks
612 [20419] dynamic-link: files with large allocated notes crash in
613 open_verify
614 [20530] libc: bswap_16 should use __builtin_bswap16() when available
615 [20890] dynamic-link: ldconfig: fsync the files before atomic rename
616 [20980] manual: CFLAGS environment variable replaces vital options
617 [21163] regex: Assertion failure in pop_fail_stack when executing a
618 malformed regexp (CVE-2015-8985)
619 [21234] manual: use of CFLAGS makes glibc detect no optimization
620 [21269] dynamic-link: i386 sigaction sa_restorer handling is wrong
621 [21313] build: Compile Error GCC 5.4.0 MIPS with -0S
622 [21314] build: Compile Error GCC 5.2.0 MIPS with -0s
623 [21508] locale: intl/tst-gettext failure with latest msgfmt
624 [21547] localedata: Tibetan script collation broken (Dzongkha and Tibetan)
625 [21812] network: getifaddrs() returns entries with ifa_name == NULL
626 [21895] libc: ppc64 setjmp/longjmp not fully interoperable with static
627 dlopen
628 [21942] dynamic-link: _dl_dst_substitute incorrectly handles $ORIGIN: with
629 AT_SECURE=1
630 [22241] localedata: New locale: Yakut (Sakha) locale for Russia (sah_RU)
631 [22247] network: Integer overflow in the decode_digit function in
632 puny_decode.c in libidn (CVE-2017-14062)
633 [22342] nscd: NSCD not properly caching netgroup
634 [22391] nptl: Signal function clear NPTL internal symbols inconsistently
635 [22550] localedata: es_ES locale (and other es_* locales): collation
636 should treat ñ as a primary different character, sync the collation
637 for Spanish with CLDR
638 [22638] dynamic-link: sparc: static binaries are broken if glibc is built
639 by gcc configured with --enable-default-pie
640 [22639] time: year 2039 bug for localtime etc. on 64-bit platforms
641 [22644] string: memmove-sse2-unaligned on 32bit x86 produces garbage when
642 crossing 2GB threshold (CVE-2017-18269)
643 [22646] localedata: redundant data (LC_TIME) for es_CL, es_CU, es_EC and
644 es_BO
645 [22735] time: Misleading typo in time.h source comment regarding
646 CLOCKS_PER_SECOND
647 [22753] libc: preadv2/pwritev2 fallback code should handle offset=-1
648 [22761] libc: No trailing `%n' conversion specifier in FMT passed from
649 `__assert_perror_fail ()' to `__assert_fail_base ()'
650 [22766] libc: all glibc internal dlopen should use RTLD_NOW for robust
651 dlopen failures
652 [22786] libc: Stack buffer overflow in realpath() if input size is close
653 to SSIZE_MAX (CVE-2018-11236)
654 [22787] dynamic-link: _dl_check_caller returns false when libc is linked
655 through an absolute DT_NEEDED path
656 [22792] build: tcb-offsets.h dependency dropped
657 [22797] libc: pkey_get() uses non-reserved name of argument
658 [22807] libc: PTRACE_* constants missing for powerpc
659 [22818] glob: posix/tst-glob_lstat_compat failure on alpha
660 [22827] dynamic-link: RISC-V ELF64 parser mis-reads flag in ldconfig
661 [22830] malloc: malloc_stats doesn't restore cancellation state on stderr
662 [22848] localedata: ca_ES: update date definitions from CLDR
663 [22862] build: _DEFAULT_SOURCE is defined even when _ISOC11_SOURCE is
664 [22884] math: RISCV fmax/fmin handle signalling NANs incorrectly
665 [22896] localedata: Update locale data for an_ES
666 [22902] math: float128 test failures with GCC 8
667 [22918] libc: multiple common of `__nss_shadow_database'
668 [22919] libc: sparc32: backtrace yields infinite backtrace with
669 makecontext
670 [22926] libc: FTBFS on powerpcspe
671 [22932] localedata: lt_LT: Update of abbreviated month names from CLDR
672 required
673 [22937] localedata: Greek (el_GR, el_CY) locales actually need ab_alt_mon
674 [22947] libc: FAIL: misc/tst-preadvwritev2
675 [22963] localedata: cs_CZ: Add alternative month names
676 [22987] math: [powerpc/sparc] fdim inlines errno, exceptions handling
677 [22996] localedata: change LC_PAPER to en_US in es_BO locale
678 [22998] dynamic-link: execstack tests are disabled when SELinux is
679 disabled
680 [23005] network: Crash in __res_context_send after memory allocation
681 failure
682 [23007] math: strtod cannot handle -nan
683 [23024] nss: getlogin_r is performing NSS lookups when loginid isn't set
684 [23036] regex: regex equivalence class regression
685 [23037] libc: initialize msg_flags to zero for sendmmsg() calls
686 [23069] libc: sigaction broken on riscv64-linux-gnu
687 [23094] localedata: hr_HR: wrong thousands_sep and mon_thousands_sep
688 [23102] dynamic-link: Incorrect parsing of multiple consecutive $variable
689 patterns in runpath entries (e.g. $ORIGIN$ORIGIN)
690 [23137] nptl: s390: pthread_join sometimes block indefinitely (on 31bit
691 and libc build with -Os)
692 [23140] localedata: More languages need two forms of month names
693 [23145] libc: _init/_fini aren't marked as hidden
694 [23152] localedata: gd_GB: Fix typo in "May" (abbreviated)
695 [23171] math: C++ iseqsig for long double converts arguments to double
696 [23178] nscd: sudo will fail when it is run in concurrent with commands
697 that changes /etc/passwd
698 [23196] string: __mempcpy_avx512_no_vzeroupper mishandles large copies
699 (CVE-2018-11237)
700 [23206] dynamic-link: static-pie + dlopen breaks debugger interaction
701 [23208] localedata: New locale - Lower Sorbian (dsb)
702 [23233] regex: Memory leak in build_charclass_op function in file
703 posix/regcomp.c
704 [23236] stdio: Harden function pointers in _IO_str_fields
705 [23250] nptl: Offset of __private_ss differs from GCC
706 [23253] math: tgamma test suite failures on i686 with -march=x86-64
707 -mtune=generic -mfpmath=sse
708 [23259] dynamic-link: Unsubstituted ${ORIGIN} remains in DT_NEEDED for
709 AT_SECURE
710 [23264] libc: posix_spawnp wrongly executes ENOEXEC in non compat mode
711 [23266] nis: stringop-truncation warning with new gcc8.1 in nisplus-
712 parser.c
713 [23272] math: fma(INFINITY,INFIITY,0.0) should be INFINITY
714 [23277] math: nan function should not have const attribute
715 [23279] math: scanf and strtod wrong for some hex floating-point
716 [23280] math: wscanf rounds wrong; wcstod is ok for negative numbers and
717 directed rounding
718 [23290] localedata: IBM273 is not equivalent to ISO-8859-1
719 [23303] build: undefined reference to symbol
720 '__parse_hwcap_and_convert_at_platform@@GLIBC_2.23'
721 [23307] dynamic-link: Absolute symbols whose value is zero ignored in
722 lookup
723 [23313] stdio: libio vtables validation and standard file object
724 interposition
725 [23329] libc: The __libc_freeres infrastructure is not properly run across
726 DSO boundaries.
727 [23349] libc: Various glibc headers no longer compatible with
728 <linux/time.h>
729 [23351] malloc: Remove unused code related to heap dumps and malloc
730 checking
731 [23363] stdio: stdio-common/tst-printf.c has non-free license
732 [23396] regex: Regex equivalence regression in single-byte locales
733 [23422] localedata: oc_FR: More updates of locale data
734 [23442] build: New warning with GCC 8
735 [23448] libc: Out of bounds access in IBM-1390 converter
736 [23456] libc: Wrong index_cpu_LZCNT
737 [23458] build: tst-get-cpu-features-static isn't added to tests
738 [23459] libc: COMMON_CPUID_INDEX_80000001 isn't populated for Intel
739 processors
740 [23467] dynamic-link: x86/CET: A property note parser bug
741
742 \f
743 Version 2.27
744
745 Major new features:
746
747 * The GNU C Library can now be compiled with support for building static
748 PIE executables (See --enable-static-pie in INSTALL). These static PIE
749 executables are like static executables but can be loaded at any address
750 and provide additional security hardening benefits at the cost of some
751 memory and performance. When the library is built with --enable-static-pie
752 the resulting libc.a is usable with GCC 8 and above to create static PIE
753 executables using the GCC option '-static-pie'. This feature is currently
754 supported on i386, x86_64 and x32 with binutils 2.29 or later, and on
755 aarch64 with binutils 2.30 or later.
756
757 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf,
758 sinf, sincosf and tan with FMA, contributed by Arjan van de Ven and
759 H.J. Lu from Intel.
760
761 * Optimized x86-64 trunc and truncf for processors with SSE4.1.
762
763 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf.
764
765 * In order to support faster and safer process termination the malloc API
766 family of functions will no longer print a failure address and stack
767 backtrace after detecting heap corruption. The goal is to minimize the
768 amount of work done after corruption is detected and to avoid potential
769 security issues in continued process execution. Reducing shutdown time
770 leads to lower overall process restart latency, so there is benefit both
771 from a security and performance perspective.
772
773 * The abort function terminates the process immediately, without flushing
774 stdio streams. Previous glibc versions used to flush streams, resulting
775 in deadlocks and further data corruption. This change also affects
776 process aborts as the result of assertion failures.
777
778 * On platforms where long double has the IEEE binary128 format (aarch64,
779 alpha, mips64, riscv, s390 and sparc), the math library now implements
780 _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015.
781 These are the same interfaces added in version 2.26 for some platforms where
782 this format is supported but is not the format of long double.
783
784 * On platforms with support for _Float64x (aarch64, alpha, i386, ia64,
785 mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now
786 implements interfaces for that type, as defined by ISO/IEC TS
787 18661-3:2015. These are corresponding interfaces to those supported for
788 _Float128.
789
790 * The math library now implements interfaces for the _Float32, _Float64 and
791 _Float32x types, as defined by ISO/IEC TS 18661-3:2015. These are
792 corresponding interfaces to those supported for _Float128.
793
794 * glibc now implements the memfd_create and mlock2 functions on Linux.
795
796 * Support for memory protection keys was added. The <sys/mman.h> header now
797 declares the functions pkey_alloc, pkey_free, pkey_mprotect, pkey_set,
798 pkey_get.
799
800 * The copy_file_range function was added.
801
802 * Optimized memcpy, mempcpy, memmove, and memset for sparc M7.
803
804 * The ldconfig utility now processes `include' directives using the C/POSIX
805 collation ordering. Previous glibc versions used locale-specific
806 ordering, the change might break systems that relied on that.
807
808 * Support for two grammatical forms of month names has been added.
809 In a call to strftime, the "%B" and "%b" format specifiers will now
810 produce the grammatical form required when the month is used as part
811 of a complete date. New "%OB" and "%Ob" specifiers produce the form
812 required when the month is named by itself. For instance, in Greek
813 and in many Slavic and Baltic languages, "%B" will produce the month
814 in genitive case, and "%OB" will produce the month in nominative case.
815
816 In a call to strptime, "%B", "%b", "%h", "%OB", "%Ob", and "%Oh"
817 are all valid and will all accept any known form of month
818 name---standalone or complete, abbreviated or full. In a call to
819 nl_langinfo, the query constants MON_1..12 and ABMON_1..12 return
820 the strings used by "%B" and "%b", respectively. New query
821 constants ALTMON_1..12 and _NL_ABALTMON_1..12 return the strings
822 used by "%OB" and "%Ob", respectively.
823
824 In a locale definition file, use "alt_mon" and "ab_alt_mon" to
825 define the strings for %OB and %Ob, respectively; these have the
826 same syntax as "mon" and "abmon". These arrays are optional; if they
827 are not provided then they have the same content as "mon" and "abmon",
828 respectively.
829
830 These features are provided for locales which define "alt_mon" and/or
831 "ab_alt_mon" in their locale source data. This release includes such
832 alternative month name data for the following languages: Belarusian,
833 Croatian, Greek, Lithuanian, Polish, Russian, and Ukrainian.
834
835 This feature is currently a GNU extension, but it is expected to
836 be added to the next revision of POSIX, and it is also already
837 available on some BSD-derived operating systems.
838
839 This feature will cause existing statically compiled applications
840 to fail to load locales and fall back to the builtin C/POSIX locales.
841 See notes below for other changes affecting compatibility.
842
843 * Support for the RISC-V ISA running on Linux has been added. This port
844 requires at least binutils-2.30, gcc-7.3.0, and linux-4.15; and is supported
845 for the following ISA and ABI pairs:
846
847 - rv64imac lp64
848 - rv64imafdc lp64
849 - rv64imafdc lp64d
850
851 Deprecated and removed features, and other changes affecting compatibility:
852
853 * Statically compiled applications attempting to load locales compiled for the
854 GNU C Library version 2.27 will fail and fall back to the builtin C/POSIX
855 locale. The reason for this is that the addition of the new "%OB" and "%Ob",
856 support for two grammatical forms of the month names, also extends the locale
857 data binary format. Static applications needing locale support must be
858 recompiled to match the runtime and data they are deployed with. In some
859 distributions there is an upgrade window where dynamically linked applications
860 may use a new library but the old locale data and also fall back to the
861 builtin C/POSIX locales; restarting the application process is sufficient to
862 fix this.
863
864 * Support for statically linked applications which call dlopen is deprecated
865 and will be removed in a future version of glibc. Applications which call
866 dlopen need to be linked dynamically instead.
867
868 * Support for old programs which use internal stdio data structures and
869 functions is deprecated. This includes programs which use the C++ streams
870 provided by libstdc++ in GCC 2.95. Programs which use the internal
871 symbols _IO_adjust_wcolumn, _IO_default_doallocate, _IO_default_finish,
872 _IO_default_pbackfail, _IO_default_uflow, _IO_default_xsgetn,
873 _IO_default_xsputn, _IO_doallocbuf, _IO_do_write, _IO_file_attach,
874 _IO_file_close, _IO_file_close_it, _IO_file_doallocate, _IO_file_fopen,
875 _IO_file_init, _IO_file_jumps, _IO_fileno, _IO_file_open,
876 _IO_file_overflow, _IO_file_read, _IO_file_seek, _IO_file_seekoff,
877 _IO_file_setbuf, _IO_file_stat, _IO_file_sync, _IO_file_underflow,
878 _IO_file_write, _IO_file_xsputn, _IO_flockfile, _IO_flush_all,
879 _IO_flush_all_linebuffered, _IO_free_backup_area, _IO_free_wbackup_area,
880 _IO_init, _IO_init_marker, _IO_init_wmarker, _IO_iter_begin, _IO_iter_end,
881 _IO_iter_file, _IO_iter_next, _IO_least_wmarker, _IO_link_in,
882 _IO_list_all, _IO_list_lock, _IO_list_resetlock, _IO_list_unlock,
883 _IO_marker_delta, _IO_marker_difference, _IO_remove_marker, _IO_seekmark,
884 _IO_seekwmark, _IO_str_init_readonly, _IO_str_init_static,
885 _IO_str_overflow, _IO_str_pbackfail, _IO_str_seekoff, _IO_str_underflow,
886 _IO_switch_to_main_wget_area, _IO_switch_to_wget_mode,
887 _IO_unsave_wmarkers, _IO_wdefault_doallocate, _IO_wdefault_finish,
888 _IO_wdefault_pbackfail, _IO_wdefault_setbuf, _IO_wdefault_uflow,
889 _IO_wdefault_xsgetn, _IO_wdefault_xsputn, _IO_wdoallocbuf, _IO_wdo_write,
890 _IO_wfile_jumps, _IO_wfile_overflow, _IO_wfile_sync, _IO_wfile_underflow,
891 _IO_wfile_xsputn, _IO_wmarker_delta, or _IO_wsetb may stop working with a
892 future version of glibc. Unlike other symbol removals, these old
893 applications will not be supported using compatibility symbols.
894
895 * On GNU/Linux, the obsolete Linux constant PTRACE_SEIZE_DEVEL is no longer
896 defined by <sys/ptrace.h>.
897
898 * libm no longer supports SVID error handling (calling a user-provided
899 matherr function on error) or the _LIB_VERSION variable to control error
900 handling. (SVID error handling and the _LIB_VERSION variable still work
901 for binaries linked against older versions of the GNU C Library.) The
902 libieee.a library is no longer provided. math.h no longer defines struct
903 exception, or the macros X_TLOSS, DOMAIN, SING, OVERFLOW, UNDERFLOW,
904 TLOSS, PLOSS and HUGE.
905
906 * The libm functions pow10, pow10f and pow10l are no longer supported for
907 new programs. Programs should use the standard names exp10, exp10f and
908 exp10l for these functions instead.
909
910 * The mcontext_t type is no longer the same as struct sigcontext. On
911 platforms where it was previously the same, this changes the C++ name
912 mangling for interfaces involving this type.
913
914 * The add-ons mechanism for building additional packages at the same time as
915 glibc has been removed. The --enable-add-ons configure option is now
916 ignored.
917
918 * The --without-fp configure option is now ignored. Whether hardware
919 floating-point instructions are used is now configured based on whether
920 the compiler used at configure time (without any options implied by a
921 --with-cpu= configure option) uses such instructions.
922
923 * The res_hnok, res_dnok, res_mailok and res_ownok functions now check that
924 the specified string can be parsed as a domain name.
925
926 * In the malloc_info output, the <heap> element may contain another <aspace>
927 element, "subheaps", which contains the number of sub-heaps.
928
929 * The libresolv function p_secstodate is no longer supported for new
930 programs.
931
932 * The tilepro-*-linux-gnu configuration is no longer supported.
933
934 * The nonstandard header files <libio.h> and <_G_config.h> are deprecated
935 and will be removed in a future release. Software that is still using
936 either header should be updated to use standard <stdio.h> interfaces
937 instead.
938
939 libio.h was originally the header for a set of supported GNU extensions,
940 but they have not been maintained as such in many years, they are now
941 standing in the way of improvements to stdio, and we don't think there are
942 any remaining external users. _G_config.h was never intended for public
943 use, but predates the bits convention.
944
945 Changes to build and runtime requirements:
946
947 * bison version 2.7 or later is required to generate code in the 'intl'
948 subdirectory.
949
950 Security related changes:
951
952 CVE-2009-5064: The ldd script would sometimes run the program under
953 examination directly, without preventing code execution through the
954 dynamic linker. (The glibc project disputes that this is a security
955 vulnerability; only trusted binaries must be examined using the ldd
956 script.)
957
958 CVE-2017-15670: The glob function, when invoked with GLOB_TILDE,
959 suffered from a one-byte overflow during ~ operator processing (either
960 on the stack or the heap, depending on the length of the user name).
961 Reported by Tim RĂŒhsen.
962
963 CVE-2017-15671: The glob function, when invoked with GLOB_TILDE,
964 would sometimes fail to free memory allocated during ~ operator
965 processing, leading to a memory leak and, potentially, to a denial
966 of service.
967
968 CVE-2017-15804: The glob function, when invoked with GLOB_TILDE and
969 without GLOB_NOESCAPE, could write past the end of a buffer while
970 unescaping user names. Reported by Tim RĂŒhsen.
971
972 CVE-2017-17426: The malloc function, when called with an object size near
973 the value SIZE_MAX, would return a pointer to a buffer which is too small,
974 instead of NULL. This was a regression introduced with the new malloc
975 thread cache in glibc 2.26. Reported by Iain Buclaw.
976
977 CVE-2017-1000408: Incorrect array size computation in _dl_init_paths leads
978 to the allocation of too much memory. (This is not a security bug per se,
979 it is mentioned here only because of the CVE assignment.) Reported by
980 Qualys.
981
982 CVE-2017-1000409: Buffer overflow in _dl_init_paths due to miscomputation
983 of the number of search path components. (This is not a security
984 vulnerability per se because no trust boundary is crossed if the fix for
985 CVE-2017-1000366 has been applied, but it is mentioned here only because
986 of the CVE assignment.) Reported by Qualys.
987
988 CVE-2017-16997: Incorrect handling of RPATH or RUNPATH containing $ORIGIN
989 for AT_SECURE or SUID binaries could be used to load libraries from the
990 current directory.
991
992 CVE-2018-1000001: Buffer underflow in realpath function when getcwd function
993 succeeds without returning an absolute path due to unexpected behaviour
994 of the Linux kernel getcwd syscall. Reported by halfdog.
995
996 CVE-2018-6485: The posix_memalign and memalign functions, when called with
997 an object size near the value of SIZE_MAX, would return a pointer to a
998 buffer which is too small, instead of NULL. Reported by Jakub Wilk.
999
1000 CVE-2018-6551: The malloc function, when called with an object size near
1001 the value of SIZE_MAX, would return a pointer to a buffer which is too
1002 small, instead of NULL.
1003
1004 The following bugs are resolved with this release:
1005
1006 [866] glob: glob should match dangling symlinks
1007 [1062] glob: sysdeps/generic/glob.c merge from gnulib (part 3 of 3)
1008 [2522] localedata: ca_ES@valencia: new Valencian (meridional Catalan)
1009 locale
1010 [5997] math: Very slow execution of sinf function
1011 [10580] localedata: hr_HR: updated locale
1012 [10871] locale: 'mon' array should contain both nominative and genitive
1013 cases
1014 [12349] localedata: eu_ES: incorrect thousands separator
1015 [13605] localedata: shn_MM: new Shan locale
1016 [13805] localedata: ru_RU: currency should use ',' as radix point
1017 [13953] localedata: km_KH: locale update
1018 [13994] localedata: mjw_IN: new locale
1019 [14121] build: make writes .mo files in po directory
1020 [14333] libc: Fix the race between atexit() and exit()
1021 [14681] dynamic-link: _dl_get_origin leaks memory via executable link map.
1022 [14925] localedata: bn_*: LC_IDENTIFICATION.language key should be
1023 "Bangla"
1024 [15260] localedata: LC_MESSAGES.{yes,no}{str,expr}: various errors
1025 [15261] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of full-
1026 width Latin characters
1027 [15332] localedata: es_CU: locale update
1028 [15436] stdio: Don't close or flush stdio streams on abort
1029 [15537] localedata: lv_LV: invalid collation for Latvian diacritical
1030 letters
1031 [16148] localedata: ca_ES: incorrect thousands separator
1032 [16750] dynamic-link: ldd should not try to execute the binaries
1033 (CVE-2009-5064)
1034 [16777] localedata: pl_PL: incorrect thousands separator in locale
1035 [16905] localedata: hanzi: new collation
1036 [17563] localedata: cmn_TW: add hanzi collation
1037 [17750] localedata: wrong collation order of diacritics in most locales
1038 [17804] libc: scandirat fails with ENOMEM because it checks for errno even
1039 if malloc succeeded
1040 [17956] build: Build fails on missing definitions from header file
1041 nss/nss.h when Mozilla NSS is used for cryptography
1042 [18203] libc: realpath() does not handle unreachable paths correctly
1043 [18572] dynamic-link: [arm] Lazy TLSDESC relocation has data race
1044 [18812] localedata: kab_DZ: new Kabyle Algeria locale
1045 [18822] libc: Internal functions are called via PLT
1046 [18858] string: _HAVE_STRING_ARCH_xxx aren't defined for i386 nor x86_64
1047 [19170] libc: __gmon_start__ defined in hppa in crtn.S
1048 [19574] libc: glibc should support building static PIE binaries
1049 [19852] localedata: charmaps/UTF-8: incorrect wcwidth for U+3099 and
1050 U+309A
1051 [19971] glob: glob: Do not skip entries with zero d_ino values
1052 [19982] localedata: fr.po: spelling mistake for error code EXDEV
1053 [20008] localedata: km_KH: convert to translit_neutral
1054 [20009] localedata: tr_TR: convert LC_CTYPE to i18n
1055 [20142] math: [x86_64] Add SSE4.1 trunc, truncf
1056 [20204] dynamic-link: _dl_open_hook and _dlfcn_hook hardening
1057 [20482] localedata: de_CH: abbreviated weekdays should be two letters
1058 [20498] localedata: miq_NI: new MĂ­skitu / Miskito (miq) language locale
1059 for Nicaragua
1060 [20532] nss: getaddrinfo uses errno and h_errno without guaranteeing
1061 they're set, wrong errors returned by gaih_inet when lookup functions
1062 are not found.
1063 [20756] localedata: [PATCH] Use Unicode wise thousands separator
1064 [20826] network: posix/tst-getaddrinfo5 fails on hosts without network
1065 access
1066 [20952] localedata: yuw_PG: new locale
1067 [21084] localedata: charmaps/IBM858: new codepage
1068 [21161] manual: [PATCH] fix typo in manual/arith.texi on strtoul prototype
1069 [21242] libc: assert gives pedantic warning in old gcc versions
1070 [21265] dynamic-link: _dl_runtime_resolve isn't compatible with Intel C++
1071 __regcall calling convention
1072 [21309] math: signed integer overflow in sysdeps/ieee754/dbl-64/e_pow.c
1073 [21326] libc: C99 functions are not declared for C++11 and later if
1074 _GNU_SOURCE is not predefined by g++
1075 [21457] libc: sys/ucontext.h namespace
1076 [21530] libc: tmpfile() should be implemented using O_TMPFILE
1077 [21660] math: GCC fails to compile a formula with tgmath.h
1078 [21672] nptl: sys-libs/glibc on ia64 crashes on thread exit: signal
1079 SIGSEGV, Segmentation fault: pthread_create.c:432: __madvise
1080 (pd->stackblock, freesize - PTHREAD_STACK_MIN, MADV_DONTNEED);
1081 [21684] math: tgmath.h handling of complex integers
1082 [21685] math: tgmath.h handling of bit-fields
1083 [21686] math: tgmath.h handling of __int128
1084 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
1085 locale
1086 [21745] libc: [powerpc64le] Extra PLT reference with --enable-stack-
1087 protector=all
1088 [21750] localedata: column width of characters incompatible with classical
1089 wcwidth
1090 [21754] malloc: malloc: Perform as little work as possible after heap
1091 consistency check failures
1092 [21780] libc: hppa: p{read,write}v2 does not set ENOSUP on invalid flag
1093 [21790] libc: Missing __memset_zero_constant_len_parameter in libc.so
1094 [21791] string: Unused XXX_chk_XXX functions in libc.a
1095 [21815] dynamic-link: FAIL: elf/tst-prelink-cmp with GCC is defaulted to
1096 PIE
1097 [21836] localedata: Removed redundant data (LC_MONETARY) in various Indian
1098 locales
1099 [21845] localedata: Added new Locale bho_NP
1100 [21853] localedata: Fix abday Which looks same as day in zh_SG
1101 [21854] localedata: Added New Locale en_SC
1102 [21864] libc: xmalloc.o is compiled with -DMODULE_NAME=libc
1103 [21871] dynamic-link: _dl_runtime_resolve_avx_opt is slower than
1104 _dl_runtime_resolve_avx_slow
1105 [21885] network: getaddrinfo: gethosts does not release resolver context
1106 on memory allocation failure
1107 [21899] libc: XPG4.2 sigaction namespace
1108 [21908] dynamic-link: dynamic linker broke on ia64 (mmap2 consolidation is
1109 the suspect)
1110 [21913] libc: static binaries SIGSEGV in __brk when host's gcc is pie-by-
1111 default (i386)
1112 [21915] nss: nss_files can return with NSS_STATUS_SUCCESS and a clobbered
1113 errno value, causing getaddrinfo to fail
1114 [21920] localedata: Fix p_cs_precedes/n_cs_precedes for mt_MT
1115 [21922] network: getaddrinfo with AF_INET/AF_INET6 returns EAI_NONAME
1116 instead of EAI_NODATA
1117 [21928] libc: sys/ptrace.h: remove obsolete temporary development Linux
1118 constant PTRACE_SEIZE_DEVEL
1119 [21930] math: C-only gcc builtins used in <math.h> isinf
1120 [21932] network: Unpaired __resolv_context_get in generic get*_r
1121 implementation
1122 [21941] math: powerpc: Wrong register constraint for xssqrtqp in sqrtf128
1123 [21944] libc: sigval namespace
1124 [21951] localedata: Update hanzi collation by stroke
1125 [21955] math: Wrong alignment of L(SP_RANGE)/L(SP_INF_0) in
1126 sysdeps/x86_64/fpu/e_expf.S
1127 [21956] libc: Stack allocation in MIPS syscall impl (ubounded stack
1128 allocation in syscall loops)
1129 [21959] localedata: Fix Country name for xh_ZA
1130 [21960] localedata: Fix abmon for bem_ZM
1131 [21966] math: AVX2 mathvec functions use FMA without checking
1132 [21967] math: When 512-bit AVX2 wrapper functions in mathvec are used?
1133 [21971] localedata: Added New Locale for mfe_MU
1134 [21972] libc: assert macro requires operator== (int) for its argument type
1135 [21973] math: [sparc] libm missing sqrtl compat symbol
1136 [21974] libc: Remove __bb_init_func and __bb_exit_func
1137 [21982] string: stratcliff.c: error: assuming signed overflow does not
1138 occur with -O3
1139 [21986] stdio: __guess_grouping is called incorrectly
1140 [21987] math: [sparc32] wrong bits/long-double.h installed
1141 [22019] localedata: Wrong placement of monetary symbol in el_GR (negative
1142 amounts)
1143 [22022] localedata: Missing country_name for mni_IN
1144 [22023] localedata: Removed redundant data (LC_TIME and LC_MESSAGES) for
1145 niu_NZ
1146 [22025] locale: iconv: Inconsistency between pointer mangling and NULL
1147 checks
1148 [22026] locale: iconv_open: heap overflow on gconv_init failure
1149 [22028] math: bits/math-finite.h _MSUF_ expansion namespace
1150 [22035] math: [m68k] bits/math-inline.h macro namespace
1151 [22038] localedata: Fix abbreviated weeks and months for Somali
1152 [22044] localedata: Remove redundant data for Limburgish Language
1153 [22050] malloc: Linking with -lmcheck does not hook
1154 __malloc_initialize_hook correctly
1155 [22051] libc: zero terminator in the middle of glibc's .eh_frame
1156 [22052] malloc: malloc failed to compile with GCC 7 and -O3
1157 [22070] localedata: charmaps/UTF-8: wcwidth for
1158 Prepended_Concatenation_Mark codepoints set to 0 (should be 1)
1159 [22074] localedata: charmaps/UTF-8: wcwidth for U+1160-U+11FF (Hangul
1160 Jungseong and Jongseong) should be 0
1161 [22078] nss: nss_files performance issue in multi mode
1162 [22082] math: bits/math-finite.h exp10 condition
1163 [22086] libc: pcprofiledump incorrect cross-endian condition
1164 [22093] dynamic-link: ld.so no longer searches in .../x86_64
1165 [22095] network: Name server address allocation memory leak in resolv.conf
1166 parsing after OOM
1167 [22096] network: __resolv_conf_attach can incorrectly free passed conf
1168 object
1169 [22100] localedata: om_KE: LC_TIME: copy redundant data from om_ET
1170 [22101] dynamic-link: Dynamic loader must ignore "debug" shared objects
1171 e.g. ET_GNU_DEBUG_*
1172 [22111] malloc: malloc: per thread cache is not returned when thread exits
1173 [22112] localedata: Fix LC_TELEPHONE/LC_NAME for az_AZ
1174 [22134] libc: [linux] implement fexecve with execveat
1175 [22142] libc: [powerpc] printf oupts a wrong value of DBL_MAX on ppc64 and
1176 ppc64le
1177 [22145] libc: ttyname() gives up too early in the face of namespaces
1178 [22146] math: C++ build issue with float128 on x86_64
1179 [22153] nptl: nptl: save error code before process termination
1180 [22156] libc: [hppa,ia64,microblaze] Executable stack default
1181 [22159] malloc: malloc: MALLOC_CHECK_ broken with --enable-tunables=no
1182 [22161] nscd: nscd cache prune for netgroups hangs after timeout bump
1183 [22165] libc: [hppa] Text relocations in libc.so
1184 [22180] libc: destructor registered via __cxa_atexit is called twice
1185 [22183] glob: commit 5554304f0ddd ("posix: Allow glob to match dangling
1186 symlinks") cause "make" segfaults
1187 [22189] math: [powerpc] math_private.h definitions of math_opt_barrier and
1188 math_force_eval
1189 [22207] libc: FAIL: stdlib/test-atexit-race
1190 [22225] math: nearbyint arithmetic moved before feholdexcept
1191 [22229] math: [sparc32] missing copysignl, fabsl, fmal compat symbols
1192 [22235] math: iscanonical in C++ and float128
1193 [22243] math: log2(0) and log10(0) are wrong in downward rounding without
1194 the svid compat wrapper
1195 [22244] math: ynf and yn are wrong without the svid compat wrapper
1196 [22273] libc: Improper assert in Linux posix_spawn implementation
1197 [22284] libc: -pg -pie doesn't work
1198 [22292] locale: localedef exits with error 4 when it should be error 1
1199 [22294] locale: Allow "" for int_currency_symbol definition in locales.
1200 [22295] locale: Don't warn on non-symbolic characters in locale sources in
1201 --verbose.
1202 [22296] math: glibc 2.26: signbit build issue with Gcc 5.5.0 on x86_64
1203 [22298] nptl: x32: lockups on recursive pthread_mutex_lock after upgrade
1204 to 2.26
1205 [22299] dynamic-link: Problem with $PLATFORM on x86_64 platform
1206 [22320] glob: Buffer overflow in glob with GLOB_TILDE (CVE-2017-15670)
1207 [22321] libc: sysconf(_SC_IOV_MAX) returns -1 on Linux
1208 [22322] libc: [mips64] wrong bits/long-double.h installed
1209 [22325] glob: Memory leak in glob with GLOB_TILDE (CVE-2017-15671)
1210 [22332] glob: Buffer overflow in glob with GLOB_TILDE in unescaping
1211 (CVE-2017-15804)
1212 [22336] localedata: cs_CZ LC_COLLATE does not use i18n
1213 [22343] malloc: Integer overflow in posix_memalign
1214 [22347] libc: getrandom() returns the number of bytes that were copied to
1215 the buffer even though the comments say "Return 0 on success and -1 on
1216 failure."
1217 [22353] string: sysdeps/i386/i586/strcpy.S isn't maintainable
1218 [22362] libc: Installed crt1.o, crti,.o and crtn.o files are used with
1219 -m32
1220 [22370] dynamic-link: Incorrect note padding check
1221 [22375] libc: malloc returns pointer from tcache_get when should return
1222 NULL (CVE-2017-17426)
1223 [22377] math: iseqsig, float128 and C++
1224 [22382] localedata: Error in tpi_PG locale
1225 [22387] localedata: Replace unicode sequences <Uxxxx> for characters
1226 inside the ASCII printable range
1227 [22402] math: [powerpc64le] __MATH_TG does not support _Float128 for
1228 -mlong-double-64
1229 [22403] localedata: Slash needs escaping in some locales
1230 [22408] malloc: malloc_info access heaps without arena lock, ignores heaps
1231 [22409] network: res_hnok does not accept some host names used on the
1232 Internet
1233 [22412] network: res_dnok, res_hnok should perform syntax checks
1234 [22413] network: ns_name_pton ignores syntactically invalid trailing
1235 backslash
1236 [22415] stdio: setvbuf can lead to invalid free/segfault
1237 [22432] build: Non-deterministic build
1238 [22439] malloc: malloc_info should compute summary statistics for all sub-
1239 heaps in an arena
1240 [22442] network: if_nametoindex could report index for the wrong
1241 networking interface
1242 [22446] build: aliasing violation calling readlink in handle_request
1243 [22447] build: unsafe call to strlen with a non-string in getlogin_r.c
1244 [22457] libc: Generic preadv/pwritev incorrectly calls __posix_memalign
1245 [22459] libc: FAIL: elf/check-localplt with __stack_chk_fail related to
1246 __nscd_hash/__nss_hash
1247 [22463] network: p_secstodate overflow handling
1248 [22469] localedata: pl_PL LC_COLLATE does not use i18n
1249 [22478] libc: sigwait can fail with EINTR
1250 [22505] libc: ldconfig processes include directive in locale-specific
1251 order
1252 [22515] localedata: hsb_DE LC_COLLATE does not use copy "iso14651_t1"
1253 [22517] localedata: et_EE LC_COLLATE does not use copy "iso14651_t1"
1254 [22519] localedata: is_IS LC_COLLATE does not use copy "iso14651_t1"
1255 [22524] localedata: lt_LT LC_COLLATE does not use copy "iso14651_t1"
1256 [22527] localedata: tr_TR LC_COLLATE does not use copy "iso14651_t1"
1257 [22534] localedata: Collation rules for Serbian and Bosnian should be the
1258 same as for Croatian
1259 [22561] math: [DR#471] cacosh (0 + iNaN) should return NaN +/- i pi/2
1260 [22568] math: [DR#471] ctanh (0 + iNaN), ctanh (0 + i Inf)
1261 [22577] libc: missing newline after "cannot allocate TLS data structures
1262 for initial thread"
1263 [22588] manual: manual/conf.texi: missing underscore in front of
1264 SC_SSIZE_MAX
1265 [22593] math: nextafter and nexttoward are declared with const attribute
1266 [22596] manual: manual: finite(nan) wrongly described as returning nonzero
1267 [22603] string: ia64 memchr overflows internal pointer check
1268 [22605] libc: SH clone does not set the exit code correctly
1269 [22606] dynamic-link: Incorrect array size computation in _dl_init_paths
1270 (CVE-2017-1000408)
1271 [22607] dynamic-link: Buffer Overflow in _dl_init_paths (CVE-2017-1000409)
1272 [22611] malloc: malloc/tst-realloc wrongly assumes that errno must not be
1273 modified in case of success
1274 [22614] build: gcc: error: unrecognized command line option ‘-no-pie’
1275 [22615] manual: manual: ambiguous wording about errno value in case of
1276 success
1277 [22624] libc: MIPS setjmp() saves incorrect 'o0' register in --enable-
1278 stack-protector=all
1279 [22625] dynamic-link: RPATH $ORIGIN replaced by PWD for AT_SECURE/SUID
1280 binaries or if /proc is not mounted (CVE-2017-16997)
1281 [22627] dynamic-link: $ORIGIN in $LD_LIBRARY_PATH is substituted twice
1282 [22630] build: $(no-pie-ldflag) is no longer effective
1283 [22631] math: [m68k] Bad const attributes in bits/mathinline.h
1284 [22635] nptl: pthread_self returns NULL before libpthread is loaded
1285 [22636] nptl: PTHREAD_STACK_MIN is too small on x86-64
1286 [22637] nptl: guard size is subtracted from thread stack size instead of
1287 adding it on top
1288 [22648] libc: getrlimit/setrlimit with RLIM_INFINITY broken on alpha
1289 [22657] localedata: hu_HU: Avoid double space in date
1290 [22660] math: fmax, fmin sNaN handling on alpha
1291 [22664] libc: New warning of GCC8
1292 [22665] math: alpha: ceil and floor raise inexact exceptions
1293 [22666] math: alpha: trunc raise inexact exceptions
1294 [22667] libc: makecontext lacks stack alignment on i386
1295 [22678] libc: prlimit fails for RLIM_INFINITY values on 32-bit machines
1296 [22679] libc: getcwd(3) can succeed without returning an absolute path
1297 (CVE-2018-1000001)
1298 [22685] libc: PowerPC: Static AT_SECURE binaries segfault with lock-
1299 elision and tunables
1300 [22687] math: [powerpc-nofpu] complex long double functions spurious
1301 "invalid" exception
1302 [22688] math: [powerpc-nofpu] remainderl wrong sign of zero result
1303 [22690] math: [ldbl-128ibm] lrintl, lroundl missing "invalid" exceptions
1304 [22691] math: [powerpc-nofpu] fmaxmagl, fminmagl spurious "invalid"
1305 exception
1306 [22693] math: [ldbl-128ibm] log1pl (-qNaN) spurious "invalid" exception
1307 [22697] math: [powerpc] llround spurious "inexact" exceptions on 32-bit
1308 power4
1309 [22701] nis: Incomplete removal of libnsl
1310 [22702] math: [powerpc-nofpu] nearbyintl traps with trapping "inexact"
1311 [22707] libc: Missing defines in elf.h for DF_1_STUB and DF_1_PIE.
1312 [22715] dynamic-link: FAIL: elf/tst-audit10
1313 [22719] libc: Backtrace tests fail on hppa
1314 [22742] libc: [aarch64] mcontext_t __reserved field got renamed
1315 [22743] nptl: __pthread_register_cancel corrupts stack after f81ddabffd
1316 [22765] crypt: (struct crypt_data *data)->initialized is not set to zero
1317 before the first call to crypt_r () in crypt/badsalttest.c
1318
1319 \f
1320 Version 2.26
1321
1322 Major new features:
1323
1324 * A per-thread cache has been added to malloc. Access to the cache requires
1325 no locks and therefore significantly accelerates the fast path to allocate
1326 and free small amounts of memory. Refilling an empty cache requires locking
1327 the underlying arena. Performance measurements show significant gains in a
1328 wide variety of user workloads. Workloads were captured using a special
1329 instrumented malloc and analyzed with a malloc simulator. Contributed by
1330 DJ Delorie with the help of Florian Weimer, and Carlos O'Donell.
1331
1332 * Unicode 10.0.0 Support: Character encoding, character type info, and
1333 transliteration tables are all updated to Unicode 10.0.0, using
1334 generator scripts contributed by Mike FABIAN (Red Hat).
1335 These updates cause user visible changes, especially the changes in
1336 wcwidth for many emoji characters cause problems when emoji sequences
1337 are rendered with pango, see for example:
1338 https://bugzilla.gnome.org/show_bug.cgi?id=780669#c5
1339
1340 * Collation of Hungarian has been overhauled and is now consistent with "The
1341 Rules of Hungarian Orthography, 12th edition" (Bug 18934). Contributed by
1342 Egmont Koblinger.
1343
1344 * Improvements to the DNS stub resolver, contributed by Florian Weimer:
1345
1346 - The GNU C Library will now detect when /etc/resolv.conf has been
1347 modified and reload the changed configuration. The new resolver option
1348 “no-reload” (RES_NORELOAD) disables this behavior.
1349
1350 - The GNU C Library now supports an arbitrary number of search domains
1351 (configured using the “search” directive in /etc/resolv.conf);
1352 previously, there was a hard limit of six domains. For backward
1353 compatibility, applications that directly modify the ‘_res’ global
1354 object are still limited to six search domains.
1355
1356 - When the “rotate” (RES_ROTATE) resolver option is active, the GNU C
1357 Library will now randomly pick a name server from the configuration as a
1358 starting point. (Previously, the second name server was always used.)
1359
1360 * The tunables feature is now enabled by default. This allows users to tweak
1361 behavior of the GNU C Library using the GLIBC_TUNABLES environment variable.
1362
1363 * New function reallocarray, which resizes an allocated block (like realloc)
1364 to the product of two sizes, with a guaranteed clean failure upon integer
1365 overflow in the multiplication. Originally from OpenBSD, contributed by
1366 Dennis Wölfing and RĂŒdiger Sonderfeld.
1367
1368 * New wrappers for the Linux-specific system calls preadv2 and pwritev2.
1369 These are extended versions of preadv and pwritev, respectively, taking an
1370 additional flags argument. The set of supported flags depends on the
1371 running kernel; full support currently requires kernel 4.7 or later.
1372
1373 * posix_spawnattr_setflags now supports the flag POSIX_SPAWN_SETSID, to
1374 create a new session ID for the spawned process. This feature is
1375 scheduled to be added to the next major revision of POSIX; for the time
1376 being, it is available under _GNU_SOURCE.
1377
1378 * errno.h is now safe to use from C-preprocessed assembly language on all
1379 supported operating systems. In this context, it will only define the
1380 Exxxx constants, as preprocessor macros expanding to integer literals.
1381
1382 * On ia64, powerpc64le, x86-32, and x86-64, the math library now implements
1383 128-bit floating point as defined by ISO/IEC/IEEE 60559:2011 (IEEE
1384 754-2008) and ISO/IEC TS 18661-3:2015. Contributed by Paul E. Murphy,
1385 Gabriel F. T. Gomes, Tulio Magno Quites Machado Filho, and Joseph Myers.
1386
1387 To compile programs that use this feature, the compiler must support
1388 128-bit floating point with the type name _Float128 (as defined by TS
1389 18661-3) or __float128 (the nonstandard name used by GCC for C++, and for
1390 C prior to version 7). _GNU_SOURCE or __STDC_WANT_IEC_60559_TYPES_EXT__
1391 must be defined to make the new interfaces visible.
1392
1393 The new functions and macros correspond to those present for other
1394 floating-point types (except for a few obsolescent interfaces not
1395 supported for the new type), with F128 or f128 suffixes; for example,
1396 strtof128, HUGE_VAL_F128 and cosf128. Following TS 18661-3, there are no
1397 printf or scanf formats for the new type; the strfromf128 and strtof128
1398 interfaces should be used instead.
1399
1400 Deprecated and removed features, and other changes affecting compatibility:
1401
1402 * The synchronization that pthread_spin_unlock performs has been changed to
1403 now be equivalent to a C11 atomic store with release memory order to the
1404 spin lock's memory location. Previously, several (but not all)
1405 architectures used stronger synchronization (e.g., containing what is
1406 often called a full barrier). This change can improve performance, but
1407 may affect odd fringe uses of spin locks that depend on the previous
1408 behavior (e.g., using spin locks as atomic variables to try to implement
1409 Dekker's mutual exclusion algorithm).
1410
1411 * The port to Native Client running on ARMv7-A (--host=arm-nacl) has been
1412 removed.
1413
1414 * Sun RPC is deprecated. The rpcgen program, librpcsvc, and Sun RPC headers
1415 will only be built and installed when the GNU C Library is configured with
1416 --enable-obsolete-rpc. This allows alternative RPC implementations, such
1417 as TIRPC or rpcsvc-proto, to be used.
1418
1419 * The NIS(+) name service modules, libnss_nis, libnss_nisplus, and
1420 libnss_compat, are deprecated, and will not be built or installed by
1421 default.
1422
1423 The NIS(+) support library, libnsl, is also deprecated. By default, a
1424 compatibility shared library will be built and installed, but not headers
1425 or development libraries. Only a few NIS-related programs require this
1426 library. (In particular, the GNU C Library has never required programs
1427 that use 'gethostbyname' to be linked with libnsl.)
1428
1429 Replacement implementations based on TIRPC, which additionally support
1430 IPv6, are available from <https://github.com/thkukuk/>. The configure
1431 option --enable-obsolete-nsl will cause libnsl's headers, and the NIS(+)
1432 name service modules, to be built and installed.
1433
1434 * The DNS stub resolver no longer performs EDNS fallback. If EDNS or DNSSEC
1435 support is enabled, the configured recursive resolver must support EDNS.
1436 (Responding to EDNS-enabled queries with responses which are not
1437 EDNS-enabled is fine, but FORMERR responses are not.)
1438
1439 * res_mkquery and res_nmkquery no longer support the IQUERY opcode. DNS
1440 servers have not supported this opcode for a long time.
1441
1442 * The _res_opcodes variable has been removed from libresolv. It had been
1443 exported by accident.
1444
1445 * <string.h> no longer includes inline versions of any string functions,
1446 as this kind of optimization is better done by the compiler. The macros
1447 __USE_STRING_INLINES and __NO_STRING_INLINES no longer have any effect.
1448
1449 * The nonstandard header <xlocale.h> has been removed. Most programs should
1450 use <locale.h> instead. If you have a specific need for the definition of
1451 locale_t with no other declarations, please contact
1452 libc-alpha@sourceware.org and explain.
1453
1454 * The obsolete header <sys/ultrasound.h> has been removed.
1455
1456 * The obsolete signal constant SIGUNUSED is no longer defined by <signal.h>.
1457
1458 * The obsolete function cfree has been removed. Applications should use
1459 free instead.
1460
1461 * The stack_t type no longer has the name struct sigaltstack. This changes
1462 the C++ name mangling for interfaces involving this type.
1463
1464 * The ucontext_t type no longer has the name struct ucontext. This changes
1465 the C++ name mangling for interfaces involving this type.
1466
1467 * On M68k GNU/Linux and MIPS GNU/Linux, the fpregset_t type no longer has
1468 the name struct fpregset. On Nios II GNU/Linux, the mcontext_t type no
1469 longer has the name struct mcontext. On SPARC GNU/Linux, the struct
1470 mc_fq, struct rwindow, struct fpq and struct fq types are no longer
1471 defined in sys/ucontext.h, the mc_fpu_t type no longer has the name struct
1472 mc_fpu, the gwindows_t type no longer has the name struct gwindows and the
1473 fpregset_t type no longer has the name struct fpu. This changes the C++
1474 name mangling for interfaces involving those types.
1475
1476 * On S/390 GNU/Linux, the constants defined by <sys/ptrace.h> have been
1477 synced with the kernel:
1478
1479 - PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS and PTRACE_SETFPREGS
1480 are not supported on this architecture and have been removed.
1481
1482 - PTRACE_SINGLEBLOCK, PTRACE_SECCOMP_GET_FILTER, PTRACE_PEEKUSR_AREA,
1483 PTRACE_POKEUSR_AREA, PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE,
1484 PTRACE_DISABLE_TE and PTRACE_TE_ABORT_RAND have been added.
1485
1486 Programs that assume the GET/SETREGS ptrace requests are universally
1487 available will now fail to build, instead of malfunctioning at runtime.
1488
1489 Changes to build and runtime requirements:
1490
1491 * Linux kernel 3.2 or later is required at runtime, on all architectures
1492 supported by that kernel. (This is a change from version 2.25 only for
1493 x86-32 and x86-64.)
1494
1495 * GNU Binutils 2.25 or later is now required to build the GNU C Library.
1496
1497 * On most architectures, GCC 4.9 or later is required to build the GNU C
1498 Library. On powerpc64le, GCC 6.2 or later is required.
1499
1500 Older GCC versions and non-GNU compilers are still supported when
1501 compiling programs that use the GNU C Library. (We do not know exactly
1502 how old, and some GNU extensions to C may be _de facto_ required. If you
1503 are interested in helping us make this statement less vague, please
1504 contact libc-alpha@sourceware.org.)
1505
1506 Security related changes:
1507
1508 * The DNS stub resolver limits the advertised UDP buffer size to 1200 bytes,
1509 to avoid fragmentation-based spoofing attacks (CVE-2017-12132).
1510
1511 * LD_LIBRARY_PATH is now ignored in binaries running in privileged AT_SECURE
1512 mode to guard against local privilege escalation attacks (CVE-2017-1000366).
1513
1514 * Avoid printing a backtrace from the __stack_chk_fail function since it is
1515 called on a corrupt stack and a backtrace is unreliable on a corrupt stack
1516 (CVE-2010-3192).
1517
1518 * A use-after-free vulnerability in clntudp_call in the Sun RPC system has been
1519 fixed (CVE-2017-12133).
1520
1521 The following bugs are resolved with this release:
1522
1523 [984] network: Respond to changed resolv.conf in gethostbyname
1524 [5010] network: sunrpc service cleanup causes unwanted port mapper traffic
1525 [12068] localedata: sc_IT: misspelled yesexpr/day/abday/mon/abmon/date_fmt
1526 fields
1527 [12189] libc: __stack_chk_fail should not attempt a backtrace
1528 (CVE-2010-3192)
1529 [14096] time: Race condition on timezone/tst-timezone.out
1530 [14172] localedata: az_IR: new locale
1531 [14995] build: glibc fails to build if gold is the default linker, even if
1532 ld.bfd is available
1533 [15998] build: [powerpc] Set arch_minimum_kernel for powerpc LE
1534 [16637] network: inet_pton function is accepting IPv6 with bad format
1535 [16640] string: string/strtok.c: undefined behaviour inconsistent between
1536 x86 and other generic code
1537 [16875] localedata: ko_KR: fix lang_name
1538 [17225] localedata: ar_SY: localized month names for May and June are
1539 incorrect
1540 [17297] localedata: da_DK: wrong date_fmt string
1541 [18907] stdio: Incorrect order of __wur __THROW in <printf.h>
1542 [18934] localedata: hu_HU: collate: fix multiple bugs and add tests
1543 [18988] nptl: pthread wastes memory with mlockall(MCL_FUTURE)
1544 [19066] localedata: ar_SA abbreviated day and month names are in English
1545 [19569] network: resolv: Support an arbitrary number of search domains
1546 [19570] network: Implement random DNS server selection in the stub
1547 resolver
1548 [19838] locale: localedef fails on PA-RISC
1549 [19919] localedata: iso14651_t1_common: Correct the Malayalam sorting
1550 order of 0D36 and 0D37
1551 [19922] localedata: iso14651_t1_common: Define collation for Malayalam
1552 chillu characters
1553 [20098] libc: FAIL: debug/backtrace-tst on hppa
1554 [20257] network: sunrpc: clntudp_call does not enforce timeout when
1555 receiving data
1556 [20275] localedata: locale day/abday/mon/abmon should not have trailing
1557 whitespace
1558 [20313] localedata: Update locale data to Unicode 9.0
1559 [20424] manual: Document how to provide a malloc replacement
1560 [20496] localedata: agr_PE: new language locale AwajĂșn / Aguaruna (agr)
1561 for Peru
1562 [20686] locale: Add el_GR@euro to SUPPORTED.
1563 [20831] dynamic-link: _dl_map_segments does not test for __mprotect
1564 failures consistently
1565 [21015] dynamic-link: Document and fix --enable-bind-now
1566 [21016] nptl: pthread_cond support is broken on hppa
1567 [21029] libc: glibc-2.23 (and later) fails to compile with -fno-omit-
1568 frame-pointer on i386
1569 [21049] libc: segfault in longjmp_chk() due to clobbered processor
1570 register
1571 [21075] libc: unused assigment to %g4 in sparc/sparc{64,32}/clone.S
1572 [21088] libc: Build fails with --enable-static-nss
1573 [21094] math: cosf(1.57079697) has 3 ulp error on targets where the
1574 generic c code is used
1575 [21109] libc: Tunables broken on big-endian
1576 [21112] math: powf has large ulp errors with base close to 1 and exponent
1577 around 4000
1578 [21115] network: sunrpc: Use-after-free in error path in clntudp_call
1579 (CVE-2017-12133)
1580 [21120] malloc: glibc malloc is incompatible with GCC 7
1581 [21130] math: Incorrect return from y0l (-inf) and y1l (-inf) when linking
1582 with -lieee
1583 [21134] math: Exception (divide by zero) not set for y0/y1 (0.0) and y0/y1
1584 (-0.0) when linking with -lieee
1585 [21171] math: log10, log2 and lgamma return incorrect results
1586 [21179] libc: handle R_SPARC_DISP64 and R_SPARC_REGISTER relocs
1587 [21182] libc: __memchr_sse2: regression in glibc-2.25 on i686
1588 [21207] localedata: ce_RU: update weekdays from CLDR
1589 [21209] dynamic-link: LD_HWCAP_MASK read in setuid binaries
1590 [21217] localedata: Update months from CLDR-31
1591 [21232] libc: miss posix_fadvise64 on MIPS64 when static linking
1592 [21243] libc: support_delete_temp_file should issue warning for failed
1593 remove()
1594 [21244] libc: support resolv_test_start() socket fd close should be
1595 checked for errors.
1596 [21253] libc: localedef randomly segfaults when using -fstack-check due to
1597 new posix_spawn implementation
1598 [21258] dynamic-link: Branch predication in _dl_runtime_resolve_avx512_opt
1599 leads to lower CPU frequency
1600 [21259] libc: [alpha] termios.h missing IXANY for POSIX
1601 [21261] libc: [sparc64] bits/setjmp.h namespace
1602 [21267] network: [mips] bits/socket.h IOC* namespace
1603 [21268] libc: [alpha] termios.h NL2, NL3 namespace
1604 [21270] libc: mmap64 silently truncates large offset values
1605 [21275] libc: posix_spawn always crashes on ia64 now
1606 [21277] libc: [alpha] termios.h missing IUCLC for UNIX98 and older
1607 [21280] math: [powerpc] logbl for POWER7 return incorrect results
1608 [21289] libc: Incorrect declaration for 32-bit platforms with
1609 _FILE_OFFSET_BITS=64 causes build error
1610 [21295] network: GETAI(AF_UNSPEC) drops IPv6 addresses if nss module does
1611 not support gethostbyname4_r
1612 [21298] nptl: rwlock can deadlock on frequent reader/writer phase
1613 switching
1614 [21338] malloc: mallopt M_ARENA_MAX doesn't set the maximum number of
1615 arenas
1616 [21340] libc: Support POSIX_SPAWN_SETSID
1617 [21357] libc: unwind-dw2-fde deadlock when using AddressSanitizer
1618 [21359] network: ns_name_pack needs additional byte in destination buffer
1619 [21361] network: resolv: Reduce advertised EDNS0 buffer size to guard
1620 against fragmentation attacks (CVE-2017-12132)
1621 [21369] network: resolv: Remove EDNS fallback
1622 [21371] libc: Missing timespec definition when compiled with _XOPEN_SOURCE
1623 and _POSIX_C_SOURCE
1624 [21386] nptl: Assertion in fork for distinct parent PID is incorrect
1625 [21391] dynamic-link: x86: Set dl_platform and dl_hwcap from CPU features
1626 [21393] stdio: Missing dup3 error check in freopen, freopen64
1627 [21396] libc: Use AVX2 memcpy/memset on Skylake server
1628 [21399] localedata: Bad description for U00EC in
1629 localedata/charmaps/CP1254
1630 [21411] malloc: realloc documentation error
1631 [21426] network: sys/socket.h uio.h namespace
1632 [21428] libc: [aarch64] tst-backtrace5 testsuite failure
1633 [21445] libc: signal.h bsd_signal namespace
1634 [21455] network: Network headers stdint.h namespace
1635 [21474] network: resolv: res_init does not use RES_DFLRETRY (2) but 4 for
1636 retry value
1637 [21475] network: resolv: Overlong search path is truncated mid-label
1638 [21511] libc: sigstack namespace
1639 [21512] libc: clone() ends up calling exit_group() through _exit() wrapper
1640 [21514] libc: sysdeps/unix/sysv/linux/sys/syscall.h:31:27: fatal error:
1641 bits/syscall.h: No such file or directory
1642 [21517] libc: struct sigaltstack namespace
1643 [21528] dynamic-link: Duplicated minimal strtoul implementations in ld.so
1644 [21533] localedata: Update locale data to Unicode 10.0
1645 [21537] libc:
1646 ../sysdeps/unix/sysv/linux/s390/s390-32/__makecontext_ret.S:44: Error:
1647 junk at end of line, first unrecognized character is `@'
1648 [21538] libc: SIG_HOLD missing for XPG4
1649 [21539] libc: S390: Mismatch between kernel and glibc ptrace.h with
1650 request 12: PTRACE_SINGLEBLOCK vs PTRACE_GETREGS.
1651 [21542] libc: Use conservative default for sysconf (_SC_NPROCESSORS_ONLN)
1652 [21543] libc: sigevent namespace
1653 [21548] libc: [mips] get/set/make/swap context for MIPS O32 assume wrong
1654 size for general purpose registers in mcontext_t structure
1655 [21550] libc: sigwait namespace
1656 [21552] libc: XPG4 bsd_signal namespace
1657 [21554] libc: sigpause namespace
1658 [21560] libc: sys/wait.h signal.h namespace
1659 [21561] libc: waitid namespace
1660 [21573] nptl: GCC 7: /usr/bin/install: cannot remove
1661 '/usr/include/stdlib.h': Permission denied
1662 [21575] libc: sys/wait.h missing struct rusage definition
1663 [21584] libc: sigaltstack etc namespace
1664 [21597] libc: siginterrupt namespace
1665 [21607] math: hppa: FAIL: math/test-tgmath
1666 [21609] dynamic-link: Incomplete workaround for GCC __tls_get_addr ABI
1667 issue on x86-64
1668 [21622] libc: [tile] missing SA_* for POSIX.1:2008
1669 [21624] dynamic-link: ld.so: Unsafe alloca allows local attackers to alias
1670 stack and heap (CVE-2017-1000366)
1671 [21625] libc: wait3 namespace
1672 [21654] nss: Incorrect pointer alignment in NSS group merge result
1673 construction
1674 [21657] network: Parse interface zone id for node-local multicast
1675 [21662] string: memcmp-avx2-movbe.S lacks saturating subtraction for
1676 between_2_3
1677 [21666] libc: .symver is used on common symbol
1678 [21668] network: resolv: res_init cross-thread broadcast introduces race
1679 conditions
1680 [21687] math: tgmath.h totalorder, totalordermag return type
1681 [21694] locale: Current Glibc Locale Does Not Support Tok-Pisin and Fiji
1682 Hindi Locale
1683 [21696] libc: Incorrect assumption of of __cpu_mask in
1684 posix/sched_cpucount.c
1685 [21697] libc: sysdeps/posix/spawni.c: 2 * suspicious condition ?
1686 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
1687 locale
1688 [21707] math: ppc64le: Invalid IFUNC resolver from libgcc calls getauxval,
1689 leading to relocation crash
1690 [21709] libc: resolv_conf.c:552: update_from_conf: Assertion
1691 `resolv_conf_matches (resp, conf)' failed.
1692 [21710] localedata: Added Samoan language locale for Samoa
1693 [21711] localedata: Pashto yesstr/nostr locale are missing
1694 [21715] nptl: sysdeps/nptl/bits/pthreadtypes.h: typedef guard
1695 __have_pthread_attr_t can cause redefinition of typedef ‘pthread_attr_t’
1696 [21721] localedata: Incorrect Full Weekday names for ks_IN@devanagari
1697 [21723] localedata: yesstr/nostr missing for Chinese language locale
1698 [21724] localedata: yesstr and nostr are missing for Xhosa [LC_MESSAGES]
1699 locale
1700 [21727] localedata: yesstr and nostr are missing for Tsonga [LC_MESSAGES]
1701 locale
1702 [21728] localedata: New Locale for Tongan language
1703 [21729] localedata: incorrect LC_NAME fields for hi_IN
1704 [21733] localedata: yesstr and nostr are missing for zh_HK
1705 [21734] localedata: Missing yesstr and nostr are for kw_GB
1706 [21738] libc: misc/tst-preadvwritev2 and misc/tst-preadvwritev64v2 fail
1707 [21741] libc: Undefined __memmove_chk_XXX and __memset_chk_XXX in libc.a
1708 [21742] libc: _dl_num_cache_relocations is undefined in libc.a
1709 [21743] localedata: ks_IN@devanagari: abday strings mismatch the day
1710 strings
1711 [21744] libc: Tests failing on --enable-tunables --enable-stack-
1712 protector=all
1713 [21749] localedata: Wrong abbreviated day name (“abday”) for
1714 ar_JO/ar_LB/ar_SY
1715 [21756] localedata: missing yesstr, nostr for nds_DE and nds_NL
1716 [21757] localedata: missing yesstr, nostr for pap_AW and pap_CW
1717 [21759] localedata: missing yesstr and nostr for Tigrinya
1718 [21760] localedata: Fix LC_MESSAGES and LC_ADDRESS for anp_IN
1719 [21766] localedata: Wrong LC_MESSAGES for om_ET Locale
1720 [21767] localedata: Missing Bislama locales
1721 [21768] localedata: Missing yesstr and nostr for aa_ET
1722 [21770] localedata: Missing Field in li_NL
1723 [21778] nptl: Robust mutex may deadlock
1724 [21779] libc: MicroBlaze segfaults when loading libpthread
1725 [21783] localedata: Fix int_select international_call_prefixes
1726 [21784] localedata: Inconsistency in country_isbn
1727 [21788] localedata: Missing Country Postal Abbreviations
1728 [21794] localedata: Added-country_isbn-for-Italy
1729 [21795] localedata: Add/Fix country_isbn for France
1730 [21796] localedata: Added country_isbn for Republic of Korea
1731 [21797] localedata: Fix inconsistency in country_isbn and missing prefixes
1732 [21799] localedata: Added int_select international_call_prefixes
1733 [21801] localedata: Added int_select international_call_prefixes
1734 [21804] nptl: Double semicolon in thread-shared-types.h
1735 [21807] localedata: LC_ADDRESS fix for pap_CW
1736 [21808] localedata: Fix LC_ADDRESS for pap_AW
1737 [21821] localedata: Added country_name in mai_IN
1738 [21822] localedata: Fix LC_TIME for mai_IN
1739 [21823] localedata: missing yesstr, nostr for sa_IN
1740 [21825] localedata: Fix name_mrs for mag_IN
1741 [21828] localedata: 2.26 changelog should mention user visible changes
1742 with unicode 9.0
1743 [21835] localedata: Added Maithili language locale for Nepal
1744 [21838] localedata: Removed redundant data for the_NP
1745 [21839] localedata: Fix LC_MONETARY for ta_LK
1746 [21844] localedata: Fix Latin characters and Months Sequence.
1747 [21848] localedata: Fix mai_NP Title Name
1748
1749 \f
1750 Version 2.25
1751
1752 * The feature test macro __STDC_WANT_LIB_EXT2__, from ISO/IEC TR
1753 24731-2:2010, is supported to enable declarations of functions from that
1754 TR. Note that not all functions from that TR are supported by the GNU C
1755 Library.
1756
1757 * The feature test macro __STDC_WANT_IEC_60559_BFP_EXT__, from ISO/IEC TS
1758 18661-1:2014, is supported to enable declarations of functions and macros
1759 from that TS. Note that not all features from that TS are supported by
1760 the GNU C Library.
1761
1762 * The feature test macro __STDC_WANT_IEC_60559_FUNCS_EXT__, from ISO/IEC TS
1763 18661-4:2015, is supported to enable declarations of functions and macros
1764 from that TS. Note that most features from that TS are not supported by
1765 the GNU C Library.
1766
1767 * The nonstandard feature selection macros _REENTRANT and _THREAD_SAFE are
1768 now treated as compatibility synonyms for _POSIX_C_SOURCE=199506L.
1769 Since the GNU C Library defaults to a much newer revision of POSIX, this
1770 will only affect programs that specifically request an old conformance
1771 mode. For instance, a program compiled with -std=c89 -D_REENTRANT will
1772 see a change in the visible declarations, but a program compiled with
1773 just -D_REENTRANT, or -std=c99 -D_POSIX_C_SOURCE=200809L -D_REENTRANT,
1774 will not.
1775
1776 Some C libraries once required _REENTRANT and/or _THREAD_SAFE to be
1777 defined by all multithreaded code, but glibc has not required this for
1778 many years.
1779
1780 * The inclusion of <sys/sysmacros.h> by <sys/types.h> is deprecated. This
1781 means that in a future release, the macros “major”, “minor”, and “makedev”
1782 will only be available from <sys/sysmacros.h>.
1783
1784 These macros are not part of POSIX nor XSI, and their names frequently
1785 collide with user code; see for instance glibc bug 19239 and Red Hat bug
1786 130601. <stdlib.h> includes <sys/types.h> under _GNU_SOURCE, and C++ code
1787 presently cannot avoid being compiled under _GNU_SOURCE, exacerbating the
1788 problem.
1789
1790 * New <fenv.h> features from TS 18661-1:2014 are added to libm: the
1791 fesetexcept, fetestexceptflag, fegetmode and fesetmode functions, the
1792 femode_t type and the FE_DFL_MODE and FE_SNANS_ALWAYS_SIGNAL macros.
1793
1794 * Integer width macros from TS 18661-1:2014 are added to <limits.h>:
1795 CHAR_WIDTH, SCHAR_WIDTH, UCHAR_WIDTH, SHRT_WIDTH, USHRT_WIDTH, INT_WIDTH,
1796 UINT_WIDTH, LONG_WIDTH, ULONG_WIDTH, LLONG_WIDTH, ULLONG_WIDTH; and to
1797 <stdint.h>: INT8_WIDTH, UINT8_WIDTH, INT16_WIDTH, UINT16_WIDTH,
1798 INT32_WIDTH, UINT32_WIDTH, INT64_WIDTH, UINT64_WIDTH, INT_LEAST8_WIDTH,
1799 UINT_LEAST8_WIDTH, INT_LEAST16_WIDTH, UINT_LEAST16_WIDTH,
1800 INT_LEAST32_WIDTH, UINT_LEAST32_WIDTH, INT_LEAST64_WIDTH,
1801 UINT_LEAST64_WIDTH, INT_FAST8_WIDTH, UINT_FAST8_WIDTH, INT_FAST16_WIDTH,
1802 UINT_FAST16_WIDTH, INT_FAST32_WIDTH, UINT_FAST32_WIDTH, INT_FAST64_WIDTH,
1803 UINT_FAST64_WIDTH, INTPTR_WIDTH, UINTPTR_WIDTH, INTMAX_WIDTH,
1804 UINTMAX_WIDTH, PTRDIFF_WIDTH, SIG_ATOMIC_WIDTH, SIZE_WIDTH, WCHAR_WIDTH,
1805 WINT_WIDTH.
1806
1807 * New <math.h> features are added from TS 18661-1:2014:
1808
1809 - Signaling NaN macros: SNANF, SNAN, SNANL.
1810
1811 - Nearest integer functions: roundeven, roundevenf, roundevenl, fromfp,
1812 fromfpf, fromfpl, ufromfp, ufromfpf, ufromfpl, fromfpx, fromfpxf,
1813 fromfpxl, ufromfpx, ufromfpxf, ufromfpxl.
1814
1815 - llogb functions: the llogb, llogbf and llogbl functions, and the
1816 FP_LLOGB0 and FP_LLOGBNAN macros.
1817
1818 - Max-min magnitude functions: fmaxmag, fmaxmagf, fmaxmagl, fminmag,
1819 fminmagf, fminmagl.
1820
1821 - Comparison macros: iseqsig.
1822
1823 - Classification macros: iscanonical, issubnormal, iszero.
1824
1825 - Total order functions: totalorder, totalorderf, totalorderl,
1826 totalordermag, totalordermagf, totalordermagl.
1827
1828 - Canonicalize functions: canonicalize, canonicalizef, canonicalizel.
1829
1830 - NaN functions: getpayload, getpayloadf, getpayloadl, setpayload,
1831 setpayloadf, setpayloadl, setpayloadsig, setpayloadsigf, setpayloadsigl.
1832
1833 * The functions strfromd, strfromf, and strfroml, from ISO/IEC TS 18661-1:2014,
1834 are added to libc. They convert a floating-point number into string.
1835
1836 * Most of glibc can now be built with the stack smashing protector enabled.
1837 It is recommended to build glibc with --enable-stack-protector=strong.
1838 Implemented by Nick Alcock (Oracle).
1839
1840 * The function explicit_bzero, from OpenBSD, has been added to libc. It is
1841 intended to be used instead of memset() to erase sensitive data after use;
1842 the compiler will not optimize out calls to explicit_bzero even if they
1843 are "unnecessary" (in the sense that no _correct_ program can observe the
1844 effects of the memory clear).
1845
1846 * On ColdFire, MicroBlaze, Nios II and SH3, the float_t type is now defined
1847 to float instead of double. This does not affect the ABI of any libraries
1848 that are part of the GNU C Library, but may affect the ABI of other
1849 libraries that use this type in their interfaces.
1850
1851 * On x86_64, when compiling with -mfpmath=387 or -mfpmath=sse+387, the
1852 float_t and double_t types are now defined to long double instead of float
1853 and double. These options are not the default, and this does not affect
1854 the ABI of any libraries that are part of the GNU C Library, but it may
1855 affect the ABI of other libraries that use this type in their interfaces,
1856 if they are compiled or used with those options.
1857
1858 * The getentropy and getrandom functions, and the <sys/random.h> header file
1859 have been added.
1860
1861 * The buffer size for byte-oriented stdio streams is now limited to 8192
1862 bytes by default. Previously, on Linux, the default buffer size on most
1863 file systems was 4096 bytes (and thus remains unchanged), except on
1864 network file systems, where the buffer size was unpredictable and could be
1865 as large as several megabytes.
1866
1867 * The <sys/quota.h> header now includes the <linux/quota.h> header. Support
1868 for the Linux quota interface which predates kernel version 2.4.22 has
1869 been removed.
1870
1871 * The malloc_get_state and malloc_set_state functions have been removed.
1872 Already-existing binaries that dynamically link to these functions will
1873 get a hidden implementation in which malloc_get_state is a stub. As far
1874 as we know, these functions are used only by GNU Emacs and this change
1875 will not adversely affect already-built Emacs executables. Any undumped
1876 Emacs executables, which normally exist only during an Emacs build, should
1877 be rebuilt by re-running “./configure; make” in the Emacs build tree.
1878
1879 * The “ip6-dotint” and “no-ip6-dotint” resolver options, and the
1880 corresponding RES_NOIP6DOTINT flag from <resolv.h> have been removed.
1881 “no-ip6-dotint” had already been the default, and support for the
1882 “ip6-dotint” option was removed from the Internet in 2006.
1883
1884 * The "ip6-bytestring" resolver option and the corresponding RES_USEBSTRING
1885 flag from <resolv.h> have been removed. The option relied on a
1886 backwards-incompatible DNS extension which was never deployed on the
1887 Internet.
1888
1889 * The flags RES_AAONLY, RES_PRIMARY, RES_NOCHECKNAME, RES_KEEPTSIG,
1890 RES_BLAST defined in the <resolv.h> header file have been deprecated.
1891 They were already unimplemented.
1892
1893 * The "inet6" option in /etc/resolv.conf and the RES_USE_INET6 flag for
1894 _res.flags are deprecated. The flag was standardized in RFC 2133, but
1895 removed again from the IETF name lookup interface specification in RFC
1896 2553. Applications should use getaddrinfo instead.
1897
1898 * DNSSEC-related declarations and definitions have been removed from the
1899 <arpa/nameser.h> header file, and libresolv will no longer attempt to
1900 decode the data part of DNSSEC record types. Previous versions of glibc
1901 only implemented minimal support for the previous version of DNSSEC, which
1902 is incompatible with the currently deployed version.
1903
1904 * The resource record type classification macros ns_t_qt_p, ns_t_mrr_p,
1905 ns_t_rr_p, ns_t_udp_p, ns_t_xfr_p have been removed from the
1906 <arpa/nameser.h> header file because the distinction between RR types and
1907 meta-RR types is not officially standardized, subject to revision, and
1908 thus not suitable for encoding in a macro.
1909
1910 * The types res_sendhookact, res_send_qhook, re_send_rhook, and the qhook
1911 and rhook members of the res_state type in <resolv.h> have been removed.
1912 The glibc stub resolver did not support these hooks, but the header file
1913 did not reflect that.
1914
1915 * For multi-arch support it is recommended to use a GCC which has
1916 been built with support for GNU indirect functions. This ensures
1917 that correct debugging information is generated for functions
1918 selected by IFUNC resolvers. This support can either be enabled by
1919 configuring GCC with '--enable-gnu-indirect-function', or by
1920 enabling it by default by setting 'default_gnu_indirect_function'
1921 variable for a particular architecture in the GCC source file
1922 'gcc/config.gcc'.
1923
1924 * GDB pretty printers have been added for mutex and condition variable
1925 structures in POSIX Threads. When installed and loaded in gdb these pretty
1926 printers show various pthread variables in human-readable form when read
1927 using the 'print' or 'display' commands in gdb.
1928
1929 * Tunables feature added to allow tweaking of the runtime for an application
1930 program. This feature can be enabled with the '--enable-tunables' configure
1931 flag. The GNU C Library manual has details on usage and README.tunables has
1932 instructions on adding new tunables to the library.
1933
1934 * A new version of condition variables functions have been implemented in
1935 the NPTL implementation of POSIX Threads to provide stronger ordering
1936 guarantees.
1937
1938 * A new version of pthread_rwlock functions have been implemented to use a more
1939 scalable algorithm primarily through not using a critical section anymore to
1940 make state changes.
1941
1942 Security related changes:
1943
1944 * On ARM EABI (32-bit), generating a backtrace for execution contexts which
1945 have been created with makecontext could fail to terminate due to a
1946 missing .cantunwind annotation. This has been observed to lead to a hang
1947 (denial of service) in some Go applications compiled with gccgo. Reported
1948 by Andreas Schwab. (CVE-2016-6323)
1949
1950 * The DNS stub resolver functions would crash due to a NULL pointer
1951 dereference when processing a query with a valid DNS question type which
1952 was used internally in the implementation. The stub resolver now uses a
1953 question type which is outside the range of valid question type values.
1954 (CVE-2015-5180)
1955
1956 The following bugs are resolved with this release:
1957
1958 [4099] stdio: Overly agressive caching by stream i/o functions.
1959 [7065] build: Support building glibc with -fstack-protector or -fstack-
1960 protector-all
1961 [9842] localedata: en_CA: incorrect date format
1962 [13165] nptl: pthread_cond_wait() can consume a signal that was sent
1963 before it started waiting
1964 [14139] manual: Do not hardcode platform names in manual/libm-err-tab.pl
1965 [15765] nptl: sem_open is wrongly a cancellation point
1966 [16421] network: IN6_IS_ADDR_UNSPECIFIED can use undefined s6_addr32
1967 [16458] libc: endian.h and netinet/in.h byte order macros return values of
1968 wrong type
1969 [16628] dynamic-link: Segfault after a binary without pthread dlopen()s a
1970 library linked with pthread
1971 [16630] nptl: Use SYSENTER for pthread_cond_broadcast/signal() (i.e. fix
1972 "FIXME: Ingo" issue)
1973 [16907] libc: <argp.h> compiled with --std=cXX disables __attribute__
1974 [17252] libc: getrandom and getentropy syscall
1975 [17730] malloc: thread-local storage is sometimes improperly free()'d
1976 after being __libc_memalign()'d
1977 [18241] stdio: failed fseek on memstream does not set errno and can fail
1978 when it shouldnt
1979 [18243] nptl: sem_wait, sem_timedwait are cancellation points shm_open is
1980 not
1981 [18463] nptl: pthread_cond_broadcast issue when surrounded by
1982 PTHREAD_PRIO_INHERIT mutex on ARM
1983 [18784] network: res_query and related function crash for special record
1984 type queries (CVE-2015-5180)
1985 [19380] math: strtod does not raise "inexact"
1986 [19387] string: Integer overflow in memchr
1987 [19390] string: Integer overflow in strncat
1988 [19398] build: linknamespace tests fail with massively parallel build
1989 [19402] nptl: Deadlock with robust shared mutex and asynchronous
1990 termination
1991 [19469] malloc: M_PERTURB in test-skeleton.c invalidates malloc tests
1992 [19473] malloc: Turn malloc_get_state etc. in compatibility symbols
1993 [19514] libc: [PATCH] Fix spelling errors in spelling
1994 "implement"/"implementation" in several places
1995 [19582] network: Deprecate RES_USE_INET6
1996 [19673] manual: clog10 docs appear to be erroneous
1997 [19810] dynamic-link: dlopen with both RTLD_NOLOAD and RTLD_NODELETE
1998 causes a segmentation fault
1999 [19826] libc: invalid pointer returned from __tls_get_addr with static
2000 linking
2001 [20016] network: resolv: Remove hooks support from the API
2002 [20019] dynamic-link: NULL pointer dereference in libc.so.6 IFUNC due to
2003 uninitialized GOT
2004 [20033] math: [x86_64] vectorized math function don't call the __finite
2005 versions
2006 [20116] nptl: use after free in pthread_create
2007 [20181] stdio: open_memstream(): writes not at end of stream corrupt data
2008 [20292] dynamic-link: Comparison in elf/dl-open.c _dl_addr_inside_object
2009 is always true.
2010 [20311] nptl: please install proc_service.h
2011 [20366] libc: Compilation errors in installed headers in strict-compliance
2012 modes
2013 [20370] malloc: malloc: Arena free list management is still racy
2014 (incorrect fix in bug 19243)
2015 [20386] libc: assert (X = 0) does not result in GCC warning
2016 [20432] malloc: malloc: Minimize interface required for interposition
2017 [20435] libc: Missing unwind info in __startcontext causes infinite loop
2018 in _Unwind_Backtrace (CVE-2016-6323)
2019 [20444] hurd: recvmsg: PF_LOCAL sockets and msg_name lead to SIGLOST
2020 [20452] nptl: Addition of sysdep.o to libpthread.a breaks relinking
2021 libpthread.a
2022 [20455] math: [powerpc] fesetexceptflag fails to clear FE_INVALID
2023 [20459] localedata: et_EE: locale has wrong {p,n}_cs_precedes value
2024 [20477] network: resolv: incorrect double-checked locking related to
2025 _res_hconf
2026 [20478] libc: libc_ifunc macro and similar usages leads to false debug-
2027 information.
2028 [20495] math: x86_64 performance degradation due to AVX/SSE transition
2029 penalty
2030 [20497] localedata: lt_LT: LC_TIME d_fmt used is obsolete
2031 [20508] dynamic-link: _dl_runtime_resolve_avx/_dl_runtime_profile_avx512
2032 cause transition penalty
2033 [20517] math: sparcv9 missing fdiml compat symbol
2034 [20524] manual: strverscmp is inconsistent
2035 [20525] libc: <sys/quota.h> should be based on kernel headers
2036 [20539] math: GCC 7 -static -lm fails to link at -O3
2037 [20554] libc: ld: bss-plt forced due to /usr/lib/libc_nonshared.a(ppc-
2038 mcount.oS)
2039 [20558] string: POSIX bcopy/bzero decls do not implement Object Size
2040 Checking
2041 [20591] network: Remove obsolete DNSSEC support
2042 [20592] network: DNS resource record type classification macros in
2043 <arpa/nameser.h> are incorrect
2044 [20593] network: Update DNS RR type definitions
2045 [20611] network: getaddrinfo accepts invalid numeric scope IDs
2046 [20615] build: glibc build fails when using --with-cpu=power9 --enable-
2047 multi-arch
2048 [20629] network: libresolv: Remove support for bitstring labels
2049 (RES_USEBSTRING)
2050 [20647] libc: GLIBC quitting every program - glibc on Pentium-S leads to
2051 assertion: "maxidx >= 2"
2052 [20660] math: [arm] Use VSQRT
2053 [20662] libc: checking whether x86_64-pc-linux-gnu-gcc implicitly enables
2054 -fstack-protector no (32bit gcc 6.2.0 pie and ssp enable)
2055 [20689] libc: Test for FMA should also check for AVX.
2056 [20707] glob: gl_pathv entries not set to NULL with GLOB_DOOFFS
2057 [20715] math: iszero macro breaks existing code
2058 [20718] math: [powerpc] copysignl raises "invalid" for sNaN
2059 [20728] libc: powerpc: Missing TOC stub in clone
2060 [20729] build: glibc-2.24 fails to build for i486 with -Os
2061 [20750] build: Build fails with default PIE enabled gcc-6.x
2062 [20768] math: [alpha] sqrt fegetenv namespace
2063 [20785] libc: binutils 2.28 fails to assemble power6/memset.S file in
2064 glibc
2065 [20787] math: float_t is defined as float incorrectly on x86_64 even if
2066 __FLT_EVAL_METHOD__ is 2
2067 [20790] network: rpcgen buffer overrun in get_prog_declaration
2068 [20822] nptl: race condition in __lll_unlock_elision on powerpc
2069 [20829] libc: crypt snprintf namespace
2070 [20847] libc: tst-vfork3 failure
2071 [20855] math: Default bits/mathdef.h has inappropriate float_t
2072 [20859] math: [sh4] FP_ILOGB0 invalid
2073 [20864] localedata: iconv: cp936 missing single-byte euro sign (0x80,
2074 U+20AC), not same as GBK
2075 [20915] dynamic-link: global-dynamic TLS broken on aarch64
2076 [20916] math: pow handling of sNaN arguments
2077 [20918] build: Building with --enable-nss-crypt fails tst-linkall-static.
2078 [20919] math: Bad pow (qNaN, 0) result with -lieee
2079 [20940] math: hypot sNaN handling
2080 [20947] math: fmax, fmin sNaN handling
2081 [20956] libc: debug/tst-backtrace3-6 don't work with -O3 anymore
2082 [20964] network: sunrpc: Stack-based buffer overflow in getrpcport with
2083 RES_USE_INET6
2084 [20971] string: powerpc64/power7 memchr overflows internal pointer check
2085 [20973] nptl: robust mutexes: Lost wake-ups
2086 [20974] locale: bs_BA: yesexpr/noexpr regexes accept any character
2087 [20978] nis: strlen on null pointer in nss_nisplus
2088 [20985] nptl: robust mutexes: lowlevelrobustlock assembly on x86 blocks on
2089 wrong condition
2090 [21014] string: i686 memchr overflows internal pointer check
2091 [21019] libc: [mips] n32 lseek incorrect on overflow
2092 [21022] libc: [microblaze] __backtrace get_frame_size namespace
2093 [21026] libc: [MIPS] readahead syscall is broken on n64
2094 [21028] math: Fallback fesetexceptflag should always succeed
2095 [21045] libc: [powerpc-nofpu] swapcontext does not restore signal mask
2096 [21047] math: arm: fpu_control.h: _FPU_GETCW/_FPU_SETCW is rejected by
2097 clang
2098 [21053] libc: [SH] Namespace pollution from sys/ucontext.h
2099 [21061] librt: [microblaze] librt lost clock_* exports
2100 [21073] libc: tunables: insecure environment variables passed to
2101 subprocesses with AT_SECURE
2102 [21081] string: Missing vzeroupper in memset-vec-unaligned-erms.S
2103 \f
2104 Version 2.24
2105
2106 * The minimum Linux kernel version that this version of the GNU C Library
2107 can be used with is 3.2, except on i[4567]86 and x86_64, where Linux
2108 kernel version 2.6.32 or later suffices (on architectures that already
2109 required kernel versions more recent than 3.2, those requirements remain
2110 unchanged). Linux 3.2 or later kernel headers are required on all
2111 architectures.
2112
2113 * The pap_AN locale has been deleted. This has been deprecated for a long
2114 time. It has been replaced by pap_AW & pap_CW, both of which have long
2115 been included in previous releases.
2116
2117 * The readdir_r and readdir64_r functions have been deprecated. It is
2118 recommended to use readdir and readdir64 instead.
2119
2120 * The type “union wait” has been removed. It was deprecated in the early
2121 1990s and never part of POSIX. Application code should use the int type
2122 instead of “union wait”.
2123
2124 * A new NSS action is added to facilitate large distributed system
2125 administration. The action, MERGE, allows remote user stores like LDAP
2126 to be merged into local user stores like /etc/groups in order to provide
2127 easy to use, updated, and managed sets of merged credentials. The new
2128 action can be used by configuring it in /etc/nsswitch.conf:
2129 group: files [SUCCESS=merge] nis
2130 Implemented by Stephen Gallagher (Red Hat).
2131
2132 * The deprecated __malloc_initialize_hook variable has been removed from the
2133 API.
2134
2135 * The long unused localedef --old-style option has been removed. It hasn't
2136 done anything in over 16 years. Scripts using this option can safely
2137 drop it.
2138
2139 * nextupl, nextup, nextupf, nextdownl, nextdown and nextdownf are added to
2140 libm. They are defined by TS 18661 and IEEE754-2008. The nextup functions
2141 return the next representable value in the direction of positive infinity
2142 and the nextdown functions return the next representable value in the
2143 direction of negative infinity. These are currently enabled as GNU
2144 extensions.
2145
2146 Security related changes:
2147
2148 * An unnecessary stack copy in _nss_dns_getnetbyname_r was removed. It
2149 could result in a stack overflow when getnetbyname was called with an
2150 overly long name. (CVE-2016-3075)
2151
2152 * Previously, getaddrinfo copied large amounts of address data to the stack,
2153 even after the fix for CVE-2013-4458 has been applied, potentially
2154 resulting in a stack overflow. getaddrinfo now uses a heap allocation
2155 instead. Reported by Michael Petlan. (CVE-2016-3706)
2156
2157 * The glob function suffered from a stack-based buffer overflow when it was
2158 called with the GLOB_ALTDIRFUNC flag and encountered a long file name.
2159 Reported by Alexander Cherepanov. (CVE-2016-1234)
2160
2161 * The Sun RPC UDP client could exhaust all available stack space when
2162 flooded with crafted ICMP and UDP messages. Reported by Aldy Hernandez'
2163 alloca plugin for GCC. (CVE-2016-4429)
2164
2165 * The IPv6 name server management code in libresolv could result in a memory
2166 leak for each thread which is created, performs a failing naming lookup,
2167 and exits. Over time, this could result in a denial of service due to
2168 memory exhaustion. Reported by Matthias Schiffer. (CVE-2016-5417)
2169
2170 The following bugs are resolved with this release:
2171
2172 [1170] localedata: ne_NP: update Nepali locale definition file
2173 [3629] manual: stpcpy description in string.texi refers to MS-DOG instead
2174 of MS-DOS.
2175 [6527] malloc: [powerpc] Malloc alignment insufficient for PowerPC
2176 [6796] math: fdim() does not set errno on overflow
2177 [10354] libc: posix_spawn should use vfork() in more cases than presently
2178 [11213] localedata: localedata: add copyright disclaimer to locale files
2179 [12143] localedata: chr_US: new Cherokee locale
2180 [12450] localedata: sgs_LT: new locale
2181 [12676] localedata: ln_CD: new locale
2182 [13237] localedata: LC_ADDRESS.country_name: update all locales w/latest
2183 CLDR data
2184 [13304] math: fma, fmaf, fmal produce wrong results
2185 [14259] build: --localedir arg to configure is ignored
2186 [14499] nptl: Does posix_spawn invoke atfork handlers / use vfork?
2187 [14750] libc: Race condition in posix_spawn vfork usage vs signal handlers
2188 [14934] localedata: es_CL: wrong first weekday chilean locale
2189 [15262] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of
2190 romanisation
2191 [15263] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of 1/0
2192 and +/-
2193 [15264] localedata: LC_MESSAGES.yesstr/nostr: lacking in many locales
2194 [15368] nptl: raise() is not async-signal-safe
2195 [15479] math: ceil, floor, round and trunc raise inexact exception
2196 [15578] localedata: kk_KZ: various updates
2197 [16003] localedata: pap_AN: punt old locale
2198 [16137] localedata: iw_IL: punt old locale
2199 [16190] localedata: eo: new esperanto locale
2200 [16374] localedata: lv_LV: change currency symbol in LC_MONETARY to euro
2201 [16742] malloc: race condition: pthread_atfork() called before first
2202 malloc() results in unexpected locking behaviour/deadlocks
2203 [16975] localedata: LC_MESSAGES.yesexpr/noexpr: revisit capitalization in
2204 all locales
2205 [16983] localedata: postal_fmt does not allow %l and %n modifiers
2206 [17565] localedata: pt_PT: wrong (work-)week start
2207 [17899] math: [powerpc] floorl returns negative zero with FE_DOWNWARD
2208 [17950] build: Build fails with -msse
2209 [18205] localedata: be_BY*: wrong first_weekday and first_workday
2210 [18433] libc: posix_spawn does not return correctly upon failure to
2211 execute
2212 [18453] localedata: charmaps/IBM875: incorrect codes
2213 [18712] string: bits/string2.h incompatible with -O2 -Werror=packed
2214 -Wsystem-headers
2215 [18896] localedata: he_IL: improvements for currency
2216 [18911] localedata: ro_RO: Correcting week day name for "Tuesday" in
2217 Romanian locale data
2218 [18960] locale: s390: _nl_locale_subfreeres uses larl opcode on misaligned
2219 symbol
2220 [19056] libc: Deprecate readdir_r
2221 [19133] localedata: pt_*: days & months should be lowercase in Portuguese
2222 language
2223 [19198] localedata: nl_NL: small improvements for Dutch locales
2224 [19257] network: Per-thread memory leak in __res_vinit with IPv6
2225 nameservers (CVE-2016-5417)
2226 [19269] build: tst-audit4 and tst-audit10 failures with gcc-6 on non avx
2227 machine
2228 [19400] locale: Language missing in "iso-639.def", trivial fix in
2229 description
2230 [19431] malloc: Deadlock between fflush, getdelim, and fork
2231 [19505] libc: Incorrect file descriptor validity checks in
2232 posix_spawn_file_actions_add{open,close,dup2}
2233 [19509] dynamic-link: dlsym, dlvsym do not report errors through dlerror
2234 when using RTLD_NEXT
2235 [19512] locale: Stale `#ifndef HAVE_BUILTIN_EXPECT' in
2236 `intl/{gettextP,loadinfo}.h'
2237 [19534] libc: execle, execlp may use malloc
2238 [19568] localedata: *_CH: Swiss locales have inconsistent start of week
2239 [19573] network: res_nclose and __res_maybe_init disagree about name
2240 server initialization, breaking Hesiod
2241 [19575] localedata: Status of GB18030 tables
2242 [19581] localedata: sr_* date_fmt string contains additional newline
2243 [19583] string: SSSE3_Fast_Copy_Backward flag needs to be enabled for AMD
2244 Excavator core
2245 [19592] math: [ldbl-128ibm] ceill incorrect in non-default rounding modes
2246 [19593] math: [ldbl-128ibm] truncl incorrect in non-default rounding modes
2247 [19594] math: [ldbl-128ibm] roundl incorrect in non-default rounding modes
2248 [19595] math: [ldbl-128ibm] fmodl incorrect for results in subnormal
2249 double range
2250 [19602] math: [ldbl-128ibm] fmodl handling of equal arguments with low
2251 part zero incorrect
2252 [19603] math: [ldbl-128ibm] remainderl, remquol incorrect sign handling in
2253 equality tests
2254 [19610] dynamic-link: ldconfig -X removes stale symbolic links
2255 [19613] libc: s390x (64 bit) macro expansion WCOREDUMP and others
2256 [19633] locale: strfmon_l applies global locale to number formatting
2257 [19642] network: Memory leak in getnameinfo
2258 [19648] libc: test-skeleton.c: Do not set RLIMIT_DATA
2259 [19653] libc: Potential for NULL pointer dereference (CWE-476) in
2260 glibc-2.22
2261 [19654] math: [x86_64] Need testcase for BZ #19590 fix
2262 [19671] localedata: Missing Sanity Check for malloc() in 'tst-fmon.c' &
2263 'tst-numeric.c'
2264 [19674] math: [ldbl-128ibm] powl incorrect overflow handling
2265 [19677] math: [ldbl-128ibm] remainderl equality test incorrect for zero
2266 low part
2267 [19678] math: [ldbl-128ibm] nextafterl, nexttowardl incorrect sign of zero
2268 result
2269 [19679] dynamic-link: gcc-4.9.3 C++ exception handling broken due to
2270 unaligned stack
2271 [19726] locale: Converting UCS4LE to INTERNAL with iconv() does not update
2272 pointers and lengths in error-case.
2273 [19727] locale: Converting from/to UTF-xx with iconv() does not always
2274 report errors on UTF-16 surrogates values.
2275 [19755] nscd: nscd assertion failure in gc
2276 [19758] dynamic-link: Typo in EXTRA_LD_ENVVARS for x86-64
2277 [19759] libc: mempcpy shouldn't be inlined
2278 [19762] dynamic-link: HAS_CPU_FEATURE/HAS_ARCH_FEATURE are easy to misuse
2279 [19765] libc: s390 needs an optimized mempcpy
2280 [19779] glob: glob: buffer overflow with GLOB_ALTDIRFUNC due to incorrect
2281 NAME_MAX limit assumption (CVE-2016-1234)
2282 [19783] build: benchtests don't support --enable-hardcoded-path-in-tests
2283 [19787] network: Missing and incorrect truncation checks in getnameinfo
2284 [19790] math: [ldbl-128ibm] nearbyintl incorrect in non-default rounding
2285 modes
2286 [19791] network: Assertion failure in res_query.c with un-connectable name
2287 server addresses
2288 [19792] libc: MIPS: backtrace yields infinite backtrace with makecontext
2289 [19822] math: libm.so install clobbers old version
2290 [19825] network: resolv: send_vc can return uninitialized data in second
2291 response to getaddrinfo
2292 [19830] network: nss_dns: should check RDATA length against buffer length
2293 [19831] network: nss_dns: getaddrinfo returns uninitialized data when
2294 confronted with A/AAAA records of invalid size
2295 [19837] nss: nss_db: No retries for some long lines with a larger buffer
2296 [19848] math: powl(10,n) for n=-4,-5,-6,-7 is off by more than 1 ULP
2297 [19853] stdio: Printing IBM long double in decimal with high precision is
2298 sometimes incorrect
2299 [19860] build: x86_64: compile errors for tst-audit10 and tst-auditmod10b
2300 [19861] nptl: libpthread IFUNC resolver for fork can lead to crash
2301 [19862] network: resolv, nss_dns: Remove remaining logging of unexpected
2302 record types
2303 [19865] network: Assertion failure or memory leak in
2304 _nss_dns_getcanonname_r
2305 [19868] network: nss_dns: netent code does not skip over non-PTR records
2306 [19879] network: nss_dns: Stack overflow in getnetbyname implementation
2307 (CVE-2016-3075)
2308 [19881] string: Improve x86-64 memset
2309 [19907] string: Incorrect memcpy tests
2310 [19916] dynamic-link: S390: fprs/vrs are not saved/restored while
2311 resolving symbols
2312 [19925] libc: termios.h XCASE namespace
2313 [19928] string: memmove-vec-unaligned-erms.S is slow with large data size
2314 [19929] libc: limits.h NL_NMAX namespace
2315 [19931] stdio: Memory leak in vfprintf
2316 [19957] libc: clone(CLONE_VM) access invalid parent memory
2317 [19963] localedata: en_IL: New locale
2318 [19989] stdio: stdio.h cuserid namespace
2319 [19994] network: getaddrinfo does not restore RES_USE_INET6 flag in
2320 gethosts
2321 [19996] locale: langinfo.h nl_langinfo_l namespace
2322 [20005] stdio: fflush on a file opened with fmemopen resets position to 0
2323 [20010] network: getaddrinfo: Stack overflow in hostent translation
2324 (CVE-2016-3706)
2325 [20012] stdio: libio: fmemopen append mode failure
2326 [20014] stdio: stdio.h namespace for pre-threads POSIX
2327 [20017] network: resolv: Use gmtime_r instead of gmtime in p_secstodate
2328 [20023] libc: fcntl.h timespec namespace
2329 [20024] math: [x86_64] vectorized sincos trashes the stack
2330 [20031] network: nss_hesiod: Heap overflow in get_txt_records
2331 [20041] time: sys/time.h timespec namespace
2332 [20043] libc: unistd.h missing cuserid for UNIX98 and before
2333 [20044] libc: unistd.h missing pthread_atfork for UNIX98
2334 [20051] libc: ttyslot in wrong header under wrong conditions
2335 [20054] libc: gethostname not declared for XPG4
2336 [20055] libc: termios.h missing tcgetsid for XPG4
2337 [20072] dynamic-link: x86 init_cpu_features is called twice in static
2338 executable
2339 [20073] libc: sys/stat.h fchmod namespace
2340 [20074] libc: stdlib.h rand_r namespace
2341 [20076] libc: sys/stat.h missing S_IFSOCK, S_ISSOCK for XPG4
2342 [20094] libc: stdlib.h should not declare grantpt, ptsname, unlockpt for
2343 XPG3
2344 [20111] libc: struct sockaddr_storage cannot be aggregate-copied
2345 [20112] network: sunrpc: stack (frame) overflow in Sun RPC clntudp_call
2346 (CVE-2016-4429)
2347 [20115] string: Extra alignment in memset-vec-unaligned-erms.S
2348 [20119] libc: Wrong mask for processors level type from CPUID
2349 [20139] dynamic-link: Upper part of zmm is zeroed if Glibc is built with
2350 AS not supporting AVX512
2351 [20151] math: [ldbl-128/ldbl-128ibm] j0l, j1l, y0l, y1l return sNaN for
2352 sNaN argument
2353 [20153] math: [ldbl-128ibm] sqrtl (sNaN) returns sNaN
2354 [20156] math: [ldbl-128ibm] ceill, rintl etc. return sNaN for sNaN
2355 argument
2356 [20157] math: [powerpc] fabsl (sNaN) wrongly raises "invalid"
2357 [20160] math: [powerpc] ceil, rint etc. return sNaN for sNaN input
2358 [20178] libc: posix_spawn{p} should not call exit
2359 [20191] stdio: libio: vtables hardening
2360 [20195] string: FMA4 detection requires CPUID execution with register
2361 eax=0x80000001
2362 [20198] libc: quick_exit incorrectly destroys C++11 thread objects.
2363 [20205] math: [i386/x86_64] nextafterl incorrect incrementing negative
2364 subnormals
2365 [20212] math: acos (sNaN) returns sNaN
2366 [20213] math: asin (sNaN) returns sNaN
2367 [20214] network: Linux header sync with linux/in6.h and ipv6.h again.
2368 [20218] math: [i386] asinhl (sNaN) returns sNaN
2369 [20219] math: [i386] atanhl (sNaN) returns sNaN
2370 [20222] stdio: fopencookie: Mangle function pointers
2371 [20224] math: [i386] cbrtl (sNaN) returns sNaN
2372 [20225] math: ldexp, scalbn, scalbln return sNaN for sNaN input
2373 [20226] math: [i386/x86_64] expl, exp10l, expm1l return sNaN for sNaN
2374 input
2375 [20227] math: [i386/x86_64] logl (sNaN) returns sNaN
2376 [20228] math: [i386/x86_64] log10l (sNaN) returns sNaN
2377 [20229] math: [i386/x86_64] log1pl (sNaN) returns sNaN
2378 [20232] math: [ldbl-128] expm1l (sNaN) returns sNaN
2379 [20233] math: [ldbl-128ibm] expm1l (sNaN) returns sNaN
2380 [20234] math: [ldbl-128ibm] log1pl (sNaN) returns sNaN
2381 [20235] math: [i386/x86_64] log2l (sNaN) returns sNaN
2382 [20237] nss: nss_db: get*ent segfaults without preceding set*ent
2383 [20240] math: modf (sNaN) returns sNaN
2384 [20248] libc: debug/tst-longjump_chk2 calls printf from a signal handler
2385 [20250] math: frexp (sNaN) returns sNaN
2386 [20252] math: atan2 (sNaN, qNaN) fails to raise "invalid"
2387 [20255] math: [i386] fdim, fdimf return with excess range and precision /
2388 double rounding
2389 [20256] math: [i386/x86_64] fdiml returns sNaN for sNaN input
2390 [20260] string: ../sysdeps/x86/bits/string.h:1092:3: error: array
2391 subscript is below array bounds [-Werror=array-bounds]
2392 [20262] nis: _nss_nis_initgroups_dyn always returns NSS_STATUS_NOTFOUND
2393 [20263] nptl: robust mutex deadlocks if other thread requests timedlock
2394 (Only arm/linux)
2395 [20277] libc: $dp is not initialized correctly in sysdeps/hppa/start.S
2396 [20284] malloc: malloc: Corrupt arena avoidance causes unnecessary mmap
2397 fallbacks
2398 [20296] math: [i386/x86_64] scalbl returns sNaN for sNaN input, missing
2399 "invalid" exceptions
2400 [20314] nptl: make[4]: *** [/usr/include/stdlib.h] Error 1
2401 [20316] localedata: id_ID: Februari instead of Pebruari
2402 [20327] string: POWER8 strcasecmp returns incorrect result
2403 [20347] math: Failure: Test: j0_downward (0xap+0)
2404 [20348] libc: FAIL: misc/tst-preadvwritev64
2405 [20349] libc: 64-bit value is passed differently in p{readv,writev}{64}
2406 [20350] libc: There is no test for p{read,write}64
2407 [20357] math: Incorrect cos result for 1.5174239687223976
2408 [20384] build: Don't run libmvec-sincos-avx* tests on non avx machines
2409 \f
2410 Version 2.23
2411
2412 * Unicode 8.0.0 Support: Character encoding, character type info, and
2413 transliteration tables are all updated to Unicode 8.0.0, using new
2414 and/or improved generator scripts contributed by Mike FABIAN (Red Hat).
2415 These updates cause user visible changes, such as the fixes for bugs
2416 89, 16061, and 18568.
2417
2418 * sched_setaffinity, pthread_setaffinity_np no longer attempt to guess the
2419 kernel-internal CPU set size. This means that requests that change the
2420 CPU affinity which failed before (for example, an all-ones CPU mask) will
2421 now succeed. Applications that need to determine the effective CPU
2422 affinities need to call sched_getaffinity or pthread_getaffinity_np after
2423 setting it because the kernel can adjust it (and the previous size check
2424 would not detect this in the majority of cases).
2425
2426 * The fts.h header can now be used with -D_FILE_OFFSET_BITS=64. With LFS
2427 the following new symbols are used: fts64_children, fts64_close,
2428 fts64_open, fts64_read and fts64_set.
2429
2430 * getaddrinfo now detects certain invalid responses on an internal netlink
2431 socket. If such responses are received, an affected process will
2432 terminate with an error message of "Unexpected error <number> on netlink
2433 descriptor <number>" or "Unexpected netlink response of size <number> on
2434 descriptor <number>". The most likely cause for these errors is a
2435 multi-threaded application which erroneously closes and reuses the netlink
2436 file descriptor while it is used by getaddrinfo.
2437
2438 * A defect in the malloc implementation, present since glibc 2.15 (2012) or
2439 glibc 2.10 via --enable-experimental-malloc (2009), could result in the
2440 unnecessary serialization of memory allocation requests across threads.
2441 The defect is now corrected. Users should see a substantial increase in
2442 the concurent throughput of allocation requests for applications which
2443 trigger this bug. Affected applications typically create and
2444 destroy threads frequently. (Bug 19048 was reported and analyzed by
2445 Ericsson.)
2446
2447 * There is now a --disable-timezone-tools configure option for disabling the
2448 building and installing of the timezone related utilities (zic, zdump, and
2449 tzselect). This is useful for people who build the timezone data and code
2450 independent of the GNU C Library.
2451
2452 * The obsolete header <regexp.h> has been removed. Programs that require
2453 this header must be updated to use <regex.h> instead.
2454
2455 * The obsolete functions bdflush, create_module, get_kernel_syms,
2456 query_module and uselib are no longer available to newly linked binaries;
2457 the header <sys/kdaemon.h> has been removed. These functions and header
2458 were specific to systems using the Linux kernel and could not usefully be
2459 used with the GNU C Library on systems with version 2.6 or later of the
2460 Linux kernel.
2461
2462 * Optimized string, wcsmbs and memory functions for IBM z13.
2463 Implemented by Stefan Liebler.
2464
2465 * Newly linked programs that define a variable called signgam will no longer
2466 have it set by the lgamma, lgammaf and lgammal functions. Programs that
2467 require signgam to be set by those functions must ensure that they use the
2468 variable provided by the GNU C Library and declared in <math.h>, without
2469 defining their own copy.
2470
2471 * The minimum GCC version that can be used to build this version of the GNU
2472 C Library is GCC 4.7. Older GCC versions, and non-GNU compilers, can
2473 still be used to compile programs using the GNU C Library.
2474
2475 Security related changes:
2476
2477 * An out-of-bounds value in a broken-out struct tm argument to strftime no
2478 longer causes a crash. Reported by Adam Nielsen. (CVE-2015-8776)
2479
2480 * The LD_POINTER_GUARD environment variable can no longer be used to disable
2481 the pointer guard feature. It is always enabled. Previously,
2482 LD_POINTER_GUARD could be used to disable security hardening in binaries
2483 running in privileged AT_SECURE mode. Reported by Hector Marco-Gisbert.
2484 (CVE-2015-8777)
2485
2486 * An integer overflow in hcreate and hcreate_r could lead to an
2487 out-of-bounds memory access. Reported by Szabolcs Nagy. (CVE-2015-8778)
2488
2489 * The catopen function no longer has unbounded stack usage. Reported by
2490 Max. (CVE-2015-8779)
2491
2492 * The nan, nanf and nanl functions no longer have unbounded stack usage
2493 depending on the length of the string passed as an argument to the
2494 functions. Reported by Joseph Myers. (CVE-2014-9761)
2495
2496 * A stack-based buffer overflow was found in libresolv when invoked from
2497 libnss_dns, allowing specially crafted DNS responses to seize control
2498 of execution flow in the DNS client. The buffer overflow occurs in
2499 the functions send_dg (send datagram) and send_vc (send TCP) for the
2500 NSS module libnss_dns.so.2 when calling getaddrinfo with AF_UNSPEC
2501 family. The use of AF_UNSPEC triggers the low-level resolver code to
2502 send out two parallel queries for A and AAAA. A mismanagement of the
2503 buffers used for those queries could result in the response of a query
2504 writing beyond the alloca allocated buffer created by
2505 _nss_dns_gethostbyname4_r. Buffer management is simplified to remove
2506 the overflow. Thanks to the Google Security Team and Red Hat for
2507 reporting the security impact of this issue, and Robert Holiday of
2508 Ciena for reporting the related bug 18665. (CVE-2015-7547)
2509
2510 The following bugs are resolved with this release:
2511
2512 [89] localedata: Locales nb_NO and nn_NO should transliterate ÊÞÄ
2513 [887] math: Math library function "logb" and "nextafter" inconsistent
2514 [2542] math: Incorrect return from float gamma (-0X1.FA471547C2FE5P+1)
2515 [2543] math: Incorrect return from float gamma (-0X1.9260DCP+1)
2516 [2558] math: Incorrect return from double gamma (-0X1.FA471547C2FE5P+1)
2517 [2898] libc: [improve] warning: the use of `mktemp' is dangerous, better
2518 use `mkstemp'
2519 [4404] localedata: German translation of "Alarm clock" is misleading
2520 [6799] math: nextafter() and nexttoward() doen't set errno on
2521 overflow/underflow errors
2522 [6803] math: scalb(), scalbln(), scalbn() do not set errno on
2523 overflow/underflow
2524 [10432] nis: _nss_nis_setnetgrent assertion failure
2525 [11460] libc: fts has no LFS support
2526 [12926] network: getaddrinfo()/make_request() may spin forever
2527 [13065] nptl: Race condition in pthread barriers
2528 [13690] nptl: pthread_mutex_unlock potentially cause invalid access
2529 [14341] dynamic-link: Dynamic linker crash when DT_JMPREL and DT_REL{,A}
2530 are not contiguous
2531 [14551] math: [ldbl-128ibm] strtold overflow handling for IBM long double
2532 [14912] libc: Rename non-installed bits/*.h headers
2533 [15002] libc: Avoid undefined behavior in posix_fallocate overflow check
2534 [15367] math: Let gcc use __builtin_isinf
2535 [15384] math: One constant fewer in ieee754/dbl-64/wordsize-64/s_finite.c
2536 [15421] math: lgamma wrongly sets signgam for ISO C
2537 [15470] math: [arm] On ARM llrintl() and llroundl() do not raise
2538 FE_INVALID with argument out of range
2539 [15491] math: [i386/x86_64] x86 nearbyint implementations wrongly clear
2540 all exceptions
2541 [15786] dynamic-link: ifunc resolver functions can smash function
2542 arguments
2543 [15918] math: Unnecessary check for equality in hypotf()
2544 [16061] localedata: Review / update transliteration data
2545 [16068] math: [i386/x86_64] x86 and x86_64 fesetenv exclude state they
2546 should include
2547 [16141] time: strptime %z offset restriction
2548 [16171] math: drem should be alias of remainder
2549 [16296] math: fegetround is pure?
2550 [16347] math: [ldbl-128ibm] ldbl-128/e_lgammal_r.c may not be suitable.
2551 [16364] libc: sleep may leave SIGCHLD blocked on sync cancellation on
2552 GNU/Linux
2553 [16399] math: [mips] lrint / llrint / lround / llround missing exceptions
2554 [16415] math: Clean up ldbl-128 / ldbl-128ibm expm1l for large positive
2555 arguments
2556 [16422] math: [powerpc] math-float, math-double failing llrint tests with
2557 "Exception "Inexact" set" on ppc32
2558 [16495] localedata: nl_NL: date_fmt: shuffle year/month around
2559 [16517] math: Missing underflow exception from tanf/tan/tanl
2560 [16519] math: Missing underflow exception from sinhf
2561 [16520] math: Missing underflow exception from tanhf
2562 [16521] math: Missing underflow exception from exp2
2563 [16620] math: [ldbl-128ibm] exp10l spurious overflows / bad directed
2564 rounding results
2565 [16734] stdio: fopen calls mmap to allocate its buffer
2566 [16961] math: nan function incorrect handling of bad sequences
2567 [16962] math: nan function unbounded stack allocation (CVE-2014-9761)
2568 [16973] localedata: Fix lang_lib/lang_term as per ISO 639-2
2569 [16985] locale: localedef: confusing error message when opening output
2570 fails
2571 [17118] math: ctanh(INFINITY + 2 * I) returns incorrect value
2572 [17197] locale: Redundant shift character in iconv conversion output at
2573 block boundary
2574 [17243] libc: trunk/posix/execl.c:53: va_args problem ?
2575 [17244] libc: trunk/sysdeps/unix/sysv/linux/semctl.c:116: va_args muxup ?
2576 [17250] dynamic-link: static linking breaks nss loading
2577 (getaddrinfo/getpwnam/etc...)
2578 [17404] libc: atomic_exchange_rel lacking a barrier on MIPS16, GCC before
2579 4.7?
2580 [17441] math: isnan() should use __builtin_isnan() in GCC
2581 [17514] nptl: Assert failure unlocking ERRORCHECK mutex after timedlock
2582 (related to lock elision)
2583 [17787] manual: Exponent on page 324 of the PDF ends prematurely
2584 [17886] time: strptime should be able to parse "Z" as a timezone with %z
2585 [17887] time: strptime should be able to parse "+01:00" style timezones
2586 [17905] libc: catopen() Multiple unbounded stack allocations
2587 (CVE-2015-8779)
2588 [18084] libc: backtrace (..., 0) dumps core on x86
2589 [18086] libc: nice() sets errno to 0 on success
2590 [18240] libc: hcreate, hcreate_r should fail with ENOMEM if element count
2591 is too large (CVE-2015-8778)
2592 [18251] dynamic-link: SONAME missing when audit modules provides path
2593 [18265] libc: add attributes for wchar string and memory functions
2594 [18370] math: csqrt missing underflows
2595 [18421] libc: [hppa] read-only segment has dynamic relocations
2596 [18472] libc: Obsolete syscall wrappers should be compat symbols
2597 [18480] libc: hppa glibc miscompilation in sched_setaffinity()
2598 [18491] localedata: Update tr_TR LC_CTYPE as part of Unicode updates
2599 [18525] localedata: Remove locale timezone information
2600 [18560] libc: [powerpc] spurious bits/ipc.h definitions
2601 [18568] localedata: Update locale data to Unicode 8.0
2602 [18589] locale: sort-test.sh fails at random
2603 [18595] math: ctan, ctanh missing underflows
2604 [18604] libc: assert macro-expands its argument
2605 [18610] math: S390: fetestexcept() reports any exception if DXC-code
2606 contains a vector instruction exception.
2607 [18611] math: j1, jn missing errno setting on underflow
2608 [18618] localedata: sync Chechen locale definitions with other *_RU
2609 locales
2610 [18647] math: powf(-0x1.000002p0, 0x1p30) returns 0 instead of +inf
2611 [18661] libc: Some x86-64 assembly codes don't align stack to 16 bytes
2612 [18665] network: In send_dg, the recvfrom function is NOT always using the
2613 buffer size of a newly created buffer (CVE-2015-7547)
2614 [18674] libc: [i386] trunk/sysdeps/i386/tst-auditmod3b.c:84: possible
2615 missing break ?
2616 [18675] libc: fpathconf(_PC_NAME_MAX) fails against large filesystems for
2617 32bit processes
2618 [18681] libc: regexp.h is obsolete and buggy, and should be desupported
2619 [18699] math: tilegx cproj() for various complex infinities does not yield
2620 infinity
2621 [18724] libc: Harden put*ent functions against data injection
2622 [18743] nptl: PowerPC: findutils testcase fails with --enable-lock-elision
2623 [18755] build: build errors with -DNDEBUG
2624 [18757] stdio: fmemopen fails to set errno on failure
2625 [18778] dynamic-link: ld.so crashes if failed dlopen causes libpthread to
2626 be forced unloaded
2627 [18781] libc: openat64 lacks O_LARGEFILE
2628 [18787] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/atomic.h:71:6:
2629 error: can’t find a register in class ‘R1_REGS’ while reloading ‘asm’
2630 [18789] math: [ldbl-128ibm] sinhl inaccurate near 0
2631 [18790] math: [ldbl-128ibm] tanhl inaccurate
2632 [18795] libc: stpncpy fortification misses buffer lengths that are
2633 statically too large
2634 [18796] build: build fails for --disable-mathvec
2635 [18803] math: hypot missing underflows
2636 [18820] stdio: fmemopen may leak memory on failure
2637 [18823] math: csqrt spurious underflows
2638 [18824] math: fma spurious underflows
2639 [18825] math: pow missing underflows
2640 [18857] math: [ldbl-128ibm] nearbyintl wrongly uses signaling comparisons
2641 [18868] nptl: pthread_barrier_init typo has in-theory-undefined behavior
2642 [18870] build: sem_open.c fails to compile with missing symbol
2643 FUTEX_SHARED
2644 [18872] stdio: Fix memory leak in printf_positional
2645 [18873] libc: posix_fallocate overflow check ineffective
2646 [18875] math: Excess precision leads incorrect libm
2647 [18877] libc: arm: mmap offset regression
2648 [18887] libc: memory corruption when using getmntent on blank lines
2649 [18918] localedata: hu_HU: change time to HH:MM:SS format
2650 [18921] libc: Regression: extraneous stat() and fstat() performed by
2651 opendir()
2652 [18928] dynamic-link: LD_POINTER_GUARD is not ignored for privileged
2653 binaries (CVE-2015-8777)
2654 [18951] math: tgamma missing underflows
2655 [18952] math: [ldbl-128/ldbl-128ibm] lgammal spurious "invalid", incorrect
2656 signgam
2657 [18953] localedata: lt_LT: change currency symbol to the euro
2658 [18956] math: powf inaccuracy
2659 [18961] math: [i386] exp missing underflows
2660 [18966] math: [i386] exp10 missing underflows
2661 [18967] math: math.h XSI POSIX namespace (gamma, isnan, scalb)
2662 [18969] build: multiple string test failures due to missing locale
2663 dependencies
2664 [18970] libc: Reference of pthread_setcancelstate in libc.a
2665 [18977] math: float / long double Bessel functions not in XSI POSIX
2666 [18980] math: i386 libm functions return with excess range and precision
2667 [18981] math: i386 scalb*, ldexp return with excess range and precision
2668 [18982] stdio: va_list and vprintf
2669 [18985] time: Passing out of range data to strftime() causes a segfault
2670 (CVE-2015-8776)
2671 [19003] math: [x86_64] fma4 version of pow inappropriate contraction
2672 [19007] libc: FAIL: elf/check-localplt with -z now and binutils 2.26
2673 [19012] locale: iconv_open leaks memory on error path
2674 [19016] math: clog, clog10 inaccuracy
2675 [19018] nptl: Mangle function pointers in tls_dtor_list
2676 [19032] math: [i386] acosh (-qNaN) spurious "invalid" exception
2677 [19046] math: ldbl-128 / ldbl-128ibm lgamma bad overflow handling
2678 [19048] malloc: malloc: arena free list can become cyclic, increasing
2679 contention
2680 [19049] math: [powerpc] erfc incorrect zero sign
2681 [19050] math: [powerpc] log* incorrect zero sign
2682 [19058] math: [x86_64] Link fail with -fopenmp and -flto
2683 [19059] math: nexttoward overflow incorrect in non-default rounding modes
2684 [19071] math: ldbl-96 lroundl incorrect just below powers of 2
2685 [19074] network: Data race in _res_hconf_reorder_addrs
2686 [19076] math: [ldbl-128ibm] log1pl (-1) wrong sign of infinity
2687 [19077] math: [ldbl-128ibm] logl (1) incorrect sign of zero result
2688 [19078] math: [ldbl-128ibm] expl overflow incorrect in non-default
2689 rounding modes
2690 [19079] math: dbl-64/wordsize-64 lround based on llround incorrect for
2691 ILP32
2692 [19085] math: ldbl-128 lrintl, lroundl missing exceptions for 32-bit long
2693 [19086] manual: posix_fallocate64 documented argument order is wrong.
2694 [19088] math: lround, llround missing exceptions close to overflow
2695 threshold
2696 [19094] math: lrint, llrint missing exceptions close to overflow threshold
2697 [19095] math: dbl-64 lrint incorrect for 64-bit long
2698 [19122] dynamic-link: Unnecessary PLT relocations in librtld.os
2699 [19124] dynamic-link: ld.so failed to build with older assmebler
2700 [19125] math: [powerpc32] llroundf, llround incorrect exceptions
2701 [19129] dynamic-link: [arm] Concurrent lazy TLSDESC resolution can crash
2702 [19134] math: [powerpc32] lround, lroundf spurious exceptions
2703 [19137] libc: i386/epoll_pwait.S doesn't support cancellation
2704 [19143] nptl: Remove CPU set size checking from sched_setaffinity,
2705 pthread_setaffinity_np
2706 [19156] math: [ldbl-128] j0l spurious underflows
2707 [19164] nptl: tst-getcpu fails with many possible CPUs
2708 [19168] math: math/test-ildoubl and math/test-ldouble failure
2709 [19174] nptl: PowerPC: TLE enabled pthread mutex performs poorly.
2710 [19178] dynamic-link: ELF_RTYPE_CLASS_EXTERN_PROTECTED_DATA confuses
2711 prelink
2712 [19181] math: [i386/x86_64] fesetenv (FE_DFL_ENV), fesetenv
2713 (FE_NOMASK_ENV) do not clear SSE exceptions
2714 [19182] malloc: malloc deadlock between ptmalloc_lock_all and
2715 _int_new_arena/reused_arena
2716 [19189] math: [ldbl-128] log1pl (-qNaN) spurious "invalid" exception
2717 [19201] math: dbl-64 remainder incorrect sign of zero result
2718 [19205] math: bits/math-finite.h conditions do not match math.h and
2719 bits/mathcalls.h
2720 [19209] math: bits/math-finite.h wrongly maps ldexp to scalbn
2721 [19211] math: lgamma functions do not set signgam for -ffinite-math-only
2722 for C99-based standards
2723 [19212] libc: features.h not -Wundef clean
2724 [19213] math: [i386/x86_64] log* (1) incorrect zero sign for -ffinite-
2725 math-only
2726 [19214] libc: Family and model identification for AMD CPU's are incorrect.
2727 [19219] libc: GLIBC build fails for ia64 with missing __nearbyintl
2728 [19228] math: [powerpc] nearbyint wrongly clears "inexact", leaves traps
2729 disabled
2730 [19235] math: [powerpc64] lround, lroundf, llround, llroundf spurious
2731 "inexact" exceptions
2732 [19238] math: [powerpc] round, roundf spurious "inexact" for integer
2733 arguments
2734 [19242] libc: strtol incorrect in Turkish locales
2735 [19243] malloc: reused_arena can pick an arena on the free list, leading
2736 to an assertion failure and reference count corruption
2737 [19253] time: tzset() ineffective when temporary TZ did not include DST
2738 rules
2739 [19266] math: strtod ("NAN(I)") incorrect in Turkish locales
2740 [19270] math: [hppa] Shared libm missing __isnanl
2741 [19285] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/mman.h: missing
2742 MAP_HUGETLB and MAP_STACK defines
2743 [19313] nptl: Wrong __cpu_mask for x32
2744 [19347] libc: grantpt: try to force a specific gid even without pt_chown
2745 [19349] math: [ldbl-128ibm] tanhl inaccurate for small arguments
2746 [19350] math: [ldbl-128ibm] sinhl spurious overflows
2747 [19351] math: [ldbl-128ibm] logl inaccurate near 1
2748 [19363] time: x32: times() return value wrongly truncates/sign extends
2749 from 32bit
2750 [19367] dynamic-link: Improve branch prediction on Silvermont
2751 [19369] network: Default domain name not reset by res_ninit when "search"
2752 / "domain" entry is removed from resolv.conf
2753 [19375] math: powerpc: incorrect results for POWER7 logb with negative
2754 subnormals
2755 [19385] localedata: bg_BG: time separator should be colon, not comma
2756 [19408] libc: linux personality syscall wrapper may erroneously return an
2757 error on 32-bit architectures
2758 [19415] libc: dladdr returns wrong names on hppa
2759 [19432] libc: iconv rejects redundant escape sequences in IBM900, IBM903,
2760 IBM905, IBM907, and IBM909
2761 [19439] math: Unix98 isinf and isnan functions conflict with C++11
2762 [19443] build: build failures with -DDEBUG
2763 [19451] build: Make check fails on test-double-vlen2
2764 [19462] libc: Glibc failed to build with -Os
2765 [19465] math: Wrong code with -Os
2766 [19466] time: time/tst-mktime2.c is compiled into an infinite loop with
2767 -Os
2768 [19467] string: Fast_Unaligned_Load needs to be enabled for Excavator core
2769 CPU's.
2770 [19475] libc: Glibc 2.22 doesn't build on sparc [PATCH]
2771 [19486] math: S390: Math tests fail with "Exception Inexact set".
2772 [19529] libc: [ARM]: FAIL: stdlib/tst-makecontext
2773 [19550] libc: [mips] mmap negative offset handling inconsistent with other
2774 architectures
2775 [19590] math: Fail to build shared objects that use libmvec.so functions.
2776 \f
2777 Version 2.22
2778
2779 * The following bugs are resolved with this release:
2780
2781 438, 2981, 4719, 6544, 6792, 11216, 12836, 13028, 13064, 13151, 13152,
2782 14094, 14113, 14292, 14841, 14906, 14958, 15319, 15467, 15790, 15969,
2783 16159, 16339, 16350, 16351, 16352, 16353, 16361, 16512, 16526, 16538,
2784 16559, 16560, 16704, 16783, 16850, 17053, 17090, 17195, 17269, 17293,
2785 17322, 17403, 17475, 17523, 17542, 17569, 17581, 17588, 17596, 17620,
2786 17621, 17628, 17631, 17692, 17711, 17715, 17776, 17779, 17792, 17833,
2787 17836, 17841, 17912, 17916, 17930, 17932, 17944, 17949, 17964, 17965,
2788 17967, 17969, 17977, 17978, 17987, 17991, 17996, 17998, 17999, 18007,
2789 18019, 18020, 18029, 18030, 18032, 18034, 18036, 18038, 18039, 18042,
2790 18043, 18046, 18047, 18049, 18068, 18078, 18080, 18093, 18100, 18104,
2791 18110, 18111, 18116, 18125, 18128, 18134, 18138, 18185, 18196, 18197,
2792 18206, 18210, 18211, 18217, 18219, 18220, 18221, 18234, 18244, 18245,
2793 18247, 18287, 18319, 18324, 18333, 18346, 18371, 18383, 18397, 18400,
2794 18409, 18410, 18412, 18418, 18422, 18434, 18444, 18457, 18468, 18469,
2795 18470, 18479, 18483, 18495, 18496, 18497, 18498, 18502, 18507, 18508,
2796 18512, 18513, 18519, 18520, 18522, 18527, 18528, 18529, 18530, 18532,
2797 18533, 18534, 18536, 18539, 18540, 18542, 18544, 18545, 18546, 18547,
2798 18549, 18553, 18557, 18558, 18569, 18583, 18585, 18586, 18592, 18593,
2799 18594, 18602, 18612, 18613, 18619, 18633, 18635, 18641, 18643, 18648,
2800 18657, 18676, 18694, 18696, 18887.
2801
2802 * Cache information can be queried via sysconf() function on s390 e.g. with
2803 _SC_LEVEL1_ICACHE_SIZE as argument.
2804
2805 * A buffer overflow in gethostbyname_r and related functions performing DNS
2806 requests has been fixed. If the NSS functions were called with a
2807 misaligned buffer, the buffer length change due to pointer alignment was
2808 not taken into account. This could result in application crashes or,
2809 potentially arbitrary code execution, using crafted, but syntactically
2810 valid DNS responses. (CVE-2015-1781)
2811
2812 * The time zone file parser has been made more robust against crafted time
2813 zone files, avoiding heap buffer overflows related to the processing of
2814 the tzh_ttisstdcnt and tzh_ttisgmtcnt fields, and a stack overflow due to
2815 large time zone data files. Overly long time zone specifiers in the TZ
2816 variable no longer result in stack overflows and crashes.
2817
2818 * A powerpc and powerpc64 optimization for TLS, similar to TLS descriptors
2819 for LD and GD on x86 and x86-64, has been implemented. You will need
2820 binutils-2.24 or later to enable this optimization.
2821
2822 * Character encoding and ctype tables were updated to Unicode 7.0.0, using
2823 new generator scripts contributed by Pravin Satpute and Mike FABIAN (Red
2824 Hat). These updates cause user visible changes, such as the fix for bug
2825 17998.
2826
2827 * CVE-2014-8121 The NSS backends shared internal state between the getXXent
2828 and getXXbyYY NSS calls for the same database, causing a denial-of-service
2829 condition in some applications.
2830
2831 * Added vector math library named libmvec with the following vectorized x86_64
2832 implementations: cos, cosf, sin, sinf, sincos, sincosf, log, logf, exp, expf,
2833 pow, powf.
2834 The library can be disabled with --disable-mathvec. Use of the functions is
2835 enabled with -fopenmp -ffast-math starting from -O1 for GCC version >= 4.9.0.
2836 Shared library libmvec.so is linked in as needed when using -lm (no need to
2837 specify -lmvec explicitly for not static builds).
2838 Visit <https://sourceware.org/glibc/wiki/libmvec> for detailed information.
2839
2840 * A new fmemopen implementation has been added with the goal of POSIX
2841 compliance. The new implementation fixes the following long-standing
2842 issues: BZ#6544, BZ#11216, BZ#12836, BZ#13151, BZ#13152, and BZ#14292. The
2843 old implementation is still present for use be by existing binaries.
2844
2845 * The 32-bit sparc sigaction ABI was inadvertently broken in the 2.20 and 2.21
2846 releases. It has been fixed to match 2.19 and older, but binaries built
2847 against 2.20 and 2.21 might need to be recompiled. See BZ#18694.
2848
2849 * Port to Native Client running on ARMv7-A (--host=arm-nacl).
2850 Contributed by Roland McGrath (Google).
2851
2852 * The header <regexp.h> is deprecated, and will be removed in a future
2853 release. Use of this header will trigger a deprecation warning.
2854 Application developers should update their code to use <regex.h> instead.
2855
2856 This header was formerly part of SUS, but was deprecated in 1994 and
2857 removed from the standard in 2001. Also, the glibc implementation
2858 leaks memory. See BZ#18681 for more details.
2859 \f
2860 Version 2.21
2861
2862 * The following bugs are resolved with this release:
2863
2864 6652, 10672, 12674, 12847, 12926, 13862, 14132, 14138, 14171, 14498,
2865 15215, 15378, 15884, 16009, 16418, 16191, 16469, 16576, 16617, 16618,
2866 16619, 16657, 16740, 16857, 17192, 17266, 17273, 17344, 17363, 17370,
2867 17371, 17411, 17460, 17475, 17485, 17501, 17506, 17508, 17522, 17555,
2868 17570, 17571, 17572, 17573, 17574, 17582, 17583, 17584, 17585, 17589,
2869 17594, 17601, 17608, 17616, 17625, 17630, 17633, 17634, 17635, 17647,
2870 17653, 17657, 17658, 17664, 17665, 17668, 17682, 17702, 17717, 17719,
2871 17722, 17723, 17724, 17725, 17732, 17733, 17744, 17745, 17746, 17747,
2872 17748, 17775, 17777, 17780, 17781, 17782, 17791, 17793, 17796, 17797,
2873 17801, 17803, 17806, 17834, 17844, 17848, 17868, 17869, 17870, 17885,
2874 17892.
2875
2876 * CVE-2015-1472 CVE-2015-1473 Under certain conditions wscanf can allocate
2877 too little memory for the to-be-scanned arguments and overflow the
2878 allocated buffer. The implementation now correctly computes the required
2879 buffer size when using malloc, and switches to malloc from alloca as
2880 intended.
2881
2882 * A new semaphore algorithm has been implemented in generic C code for all
2883 machines. Previous custom assembly implementations of semaphore were
2884 difficult to reason about or ensure that they were safe. The new version
2885 of semaphore supports machines with 64-bit or 32-bit atomic operations.
2886 The new semaphore algorithm is used by sem_init, sem_open, sem_post,
2887 sem_wait, sem_timedwait, sem_trywait, and sem_getvalue.
2888
2889 * Port to Altera Nios II has been contributed by Mentor Graphics.
2890
2891 * Optimized strcpy, stpcpy, strncpy, stpncpy, strcmp, and strncmp
2892 implementations for powerpc64/powerpc64le.
2893 Implemented by Adhemerval Zanella (IBM).
2894
2895 * Added support for TSX lock elision of pthread mutexes on powerpc32, powerpc64
2896 and powerpc64le. This may improve lock scaling of existing programs on
2897 HTM capable systems. The lock elision code is only enabled with
2898 --enable-lock-elision=yes. Also, the TSX lock elision implementation for
2899 powerpc will issue a transaction abort on every syscall to avoid side
2900 effects being visible outside transactions.
2901
2902 * Optimized strcpy, stpcpy, strchrnul and strrchr implementations for
2903 AArch64. Contributed by ARM Ltd.
2904
2905 * i386 memcpy functions optimized with SSE2 unaligned load/store.
2906
2907 * CVE-2104-7817 The wordexp function could ignore the WRDE_NOCMD flag
2908 under certain input conditions resulting in the execution of a shell for
2909 command substitution when the application did not request it. The
2910 implementation now checks WRDE_NOCMD immediately before executing the
2911 shell and returns the error WRDE_CMDSUB as expected.
2912
2913 * CVE-2012-3406 printf-style functions could run into a stack overflow when
2914 processing format strings with a large number of format specifiers.
2915
2916 * CVE-2014-9402 The nss_dns implementation of getnetbyname could run into an
2917 infinite loop if the DNS response contained a PTR record of an unexpected
2918 format.
2919
2920 * The minimum GCC version that can be used to build this version of the GNU
2921 C Library is GCC 4.6. Older GCC versions, and non-GNU compilers, can
2922 still be used to compile programs using the GNU C Library.
2923
2924 * The GNU C Library is now built with -Werror by default. This can be
2925 disabled by configuring with --disable-werror.
2926
2927 * New locales: tu_IN, bh_IN, raj_IN, ce_RU.
2928
2929 * The obsolete sigvec function has been removed. This was the original
2930 4.2BSD interface that inspired the POSIX.1 sigaction interface, which
2931 programs have been using instead for about 25 years. Of course, ABI
2932 compatibility for old binaries using sigvec remains intact.
2933
2934 * Merged gettext 0.19.3 into the intl subdirectory. This fixes building
2935 with newer versions of bison.
2936
2937 * Support for MIPS o32 FPXX, FP64A and FP64 ABI Extensions.
2938 The original MIPS o32 hard-float ABI requires an FPU where double-precision
2939 registers overlay two consecutive single-precision registers. MIPS32R2
2940 introduced a new FPU mode (FR=1) where double-precision registers extend the
2941 corresponding single-precision registers which is incompatible with the
2942 o32 hard-float ABI. The MIPS SIMD ASE and the MIPSR6 architecture both
2943 require the use of FR=1 making a transition necessary. New o32 ABI
2944 extensions enable users to migrate over time from the original o32 ABI
2945 through to the updated o32 FP64 ABI. To achieve this the dynamic linker now
2946 tracks the ABI of any loaded object and verifies that new objects are
2947 compatible. Mode transitions will also be requested as required and
2948 unsupportable objects will be rejected. The ABI checks include both soft and
2949 hard float ABIs for o32, n32 and n64.
2950
2951 GCC 5 with GNU binutils 2.25 onwards:
2952 It is strongly recommended that all o32 system libraries are built using the
2953 new o32 FPXX ABI (-mfpxx) to facilitate the transition as this is compatible
2954 with the original and all new o32 ABI extensions. Configure a MIPS GCC
2955 compiler using --with-fp-32=xx to set this by default.
2956 \f
2957 Version 2.20
2958
2959 * The following bugs are resolved with this release:
2960
2961 6804, 9894, 11505, 12994, 13347, 13651, 14308, 14770, 15119, 15132, 15347,
2962 15514, 15698, 15804, 15894, 15946, 16002, 16064, 16095, 16194, 16198,
2963 16275, 16284, 16287, 16315, 16348, 16349, 16354, 16357, 16362, 16447,
2964 16516, 16532, 16539, 16545, 16561, 16562, 16564, 16574, 16599, 16600,
2965 16609, 16610, 16611, 16613, 16619, 16623, 16629, 16632, 16634, 16639,
2966 16642, 16648, 16649, 16670, 16674, 16677, 16680, 16681, 16683, 16689,
2967 16695, 16701, 16706, 16707, 16712, 16713, 16714, 16724, 16731, 16739,
2968 16740, 16743, 16754, 16758, 16759, 16760, 16770, 16786, 16789, 16791,
2969 16796, 16799, 16800, 16815, 16823, 16824, 16831, 16838, 16839, 16849,
2970 16854, 16876, 16877, 16878, 16882, 16885, 16888, 16890, 16892, 16912,
2971 16915, 16916, 16917, 16918, 16922, 16927, 16928, 16932, 16943, 16958,
2972 16965, 16966, 16967, 16977, 16978, 16984, 16990, 16996, 17009, 17022,
2973 17031, 17042, 17048, 17050, 17058, 17061, 17062, 17069, 17075, 17078,
2974 17079, 17084, 17086, 17088, 17092, 17097, 17125, 17135, 17137, 17150,
2975 17153, 17187, 17213, 17259, 17261, 17262, 17263, 17319, 17325, 17354.
2976
2977 * Reverted change of ABI data structures for s390 and s390x:
2978 On s390 and s390x the size of struct ucontext and jmp_buf was increased in
2979 2.19. This change is reverted in 2.20. The introduced 2.19 symbol versions
2980 of getcontext, setjmp, _setjmp, __sigsetjmp, longjmp, _longjmp, siglongjmp
2981 are preserved pointing straight to the same implementation as the old ones.
2982 Given that, new callers will simply provide a too-big buffer to these
2983 functions. Any applications/libraries out there that embed jmp_buf or
2984 ucontext_t in an ABI-relevant data structure that have already been rebuilt
2985 against 2.19 headers will have to rebuilt again. This is necessary in any
2986 case to revert the breakage in their ABI caused by the glibc change.
2987
2988 * Support for file description locks is added to systems running the
2989 Linux kernel. The standard file locking interfaces are extended to
2990 operate on file descriptions, not file descriptors, via the use of
2991 F_OFD_GETLK, F_OFD_SETLK, and F_OFD_SETLKW. File description locks
2992 are associated with an open file instead of a process.
2993
2994 * Optimized strchr implementation for AArch64. Contributed by ARM Ltd.
2995
2996 * The minimum Linux kernel version that this version of the GNU C Library
2997 can be used with is 2.6.32.
2998
2999 * Running the testsuite no longer terminates as soon as a test fails.
3000 Instead, a file tests.sum (xtests.sum from "make xcheck") is generated,
3001 with PASS or FAIL lines for individual tests. A summary of the results is
3002 printed, including a list of failing lists, and "make check" exits with
3003 error status if there were any unexpected failures. "make check
3004 stop-on-test-failure=y" may be used to keep the old behavior.
3005
3006 * The am33 port, which had not worked for several years, has been removed
3007 from ports.
3008
3009 * The _BSD_SOURCE and _SVID_SOURCE feature test macros are no longer
3010 supported; they now act the same as _DEFAULT_SOURCE (but generate a
3011 warning). Except for cases where _BSD_SOURCE enabled BSD interfaces that
3012 conflicted with POSIX (support for which was removed in 2.19), the
3013 interfaces those macros enabled remain available when compiling with
3014 _GNU_SOURCE defined, with _DEFAULT_SOURCE defined, or without any feature
3015 test macros defined.
3016
3017 * Optimized strcmp implementation for ARMv7. Contributed by ARM Ltd.
3018
3019 * Added support for TX lock elision of pthread mutexes on s390 and s390x.
3020 This may improve lock scaling of existing programs on TX capable systems.
3021 The lock elision code is only built with --enable-lock-elision=yes and
3022 then requires a GCC version supporting the TX builtins. With lock elision
3023 default mutexes are elided via __builtin_tbegin, if the cpu supports
3024 transactions. By default lock elision is not enabled and the elision code
3025 is not built.
3026
3027 * CVE-2014-4043 The posix_spawn_file_actions_addopen implementation did not
3028 copy the path argument. This allowed programs to cause posix_spawn to
3029 deference a dangling pointer, or use an unexpected pathname argument if
3030 the string was modified after the posix_spawn_file_actions_addopen
3031 invocation.
3032
3033 * All supported architectures now use the main glibc sysdeps directory
3034 instead of some being in a separate "ports" directory (which was
3035 distributed separately before glibc 2.17).
3036
3037 * The NPTL implementation of POSIX pthreads is no longer an "add-on".
3038 On configurations that support it (all Linux configurations), it's now
3039 used regardless of the --enable-add-ons switch to configure. It is no
3040 longer possible to build such configurations without pthreads support.
3041
3042 * Locale names, including those obtained from environment variables (LANG
3043 and the LC_* variables), are more tightly checked for proper syntax.
3044 setlocale will now fail (with EINVAL) for locale names that are overly
3045 long, contain slashes without starting with a slash, or contain ".." path
3046 components. (CVE-2014-0475) Previously, some valid locale names were
3047 silently replaced with the "C" locale when running in AT_SECURE mode
3048 (e.g., in a SUID program). This is no longer necessary because of the
3049 additional checks.
3050
3051 * On x86-64, the dynamic linker's lazy-binding support is now compatible
3052 with application code using Intel MPX instructions. (With all previous
3053 versions, the MPX register state could be clobbered when making calls
3054 into or out of a shared library.) Note that while the new dynamic
3055 linker is compatible with all known x86 hardware whether or not it
3056 supports Intel MPX, some x86 instruction-set emulators might fail to
3057 handle the new instruction encodings. This is known to affect Valgrind
3058 versions up through 3.9 (but will be fixed in the forthcoming 3.10
3059 release), and might affect other tools that do instruction emulation.
3060
3061 * Support for loadable gconv transliteration modules has been removed.
3062 The support for transliteration modules has been non-functional for
3063 over a decade, and the removal is prompted by security defects. The
3064 normal gconv conversion modules are still supported. Transliteration
3065 with //TRANSLIT is still possible, and the //IGNORE specifier
3066 continues to be supported. (CVE-2014-5119)
3067
3068 * Decoding a crafted input sequence in the character sets IBM933, IBM935,
3069 IBM937, IBM939, IBM1364 could result in an out-of-bounds array read,
3070 resulting a denial-of-service security vulnerability in applications which
3071 use functions related to iconv. (CVE-2014-6040)
3072 \f
3073 Version 2.19
3074
3075 * The following bugs are resolved with this release:
3076
3077 156, 387, 431, 762, 832, 926, 2801, 4772, 6786, 6787, 6807, 6810, 6981,
3078 7003, 9721, 9954, 10253, 10278, 11087, 11157, 11214, 12100, 12486, 12751,
3079 12986, 13028, 13982, 13985, 14029, 14032, 14120, 14143, 14155, 14286,
3080 14547, 14699, 14752, 14782, 14876, 14910, 15004, 15048, 15073, 15089,
3081 15128, 15218, 15268, 15277, 15308, 15362, 15374, 15400, 15425, 15427,
3082 15483, 15522, 15531, 15532, 15593, 15601, 15608, 15609, 15610, 15632,
3083 15640, 15670, 15672, 15680, 15681, 15723, 15734, 15735, 15736, 15748,
3084 15749, 15754, 15760, 15763, 15764, 15797, 15799, 15825, 15843, 15844,
3085 15846, 15847, 15849, 15850, 15855, 15856, 15857, 15859, 15867, 15886,
3086 15887, 15890, 15892, 15893, 15895, 15897, 15901, 15905, 15909, 15915,
3087 15917, 15919, 15921, 15923, 15939, 15941, 15948, 15963, 15966, 15968,
3088 15985, 15988, 15997, 16032, 16034, 16036, 16037, 16038, 16041, 16046,
3089 16055, 16071, 16072, 16074, 16077, 16078, 16103, 16112, 16143, 16144,
3090 16146, 16150, 16151, 16153, 16167, 16169, 16172, 16195, 16214, 16245,
3091 16271, 16274, 16283, 16289, 16293, 16314, 16316, 16330, 16337, 16338,
3092 16356, 16365, 16366, 16369, 16372, 16375, 16379, 16381, 16384, 16385,
3093 16386, 16387, 16390, 16394, 16398, 16400, 16407, 16408, 16414, 16430,
3094 16431, 16453, 16474, 16506, 16510, 16529
3095
3096 * Slovenian translations for glibc messages have been contributed by the
3097 Translation Project's Slovenian team of translators.
3098
3099 * The public headers no longer use __unused nor __block. This change is to
3100 support compiling programs that are derived from BSD sources and use
3101 __unused internally, and to support compiling with Clang's -fblock
3102 extension which uses __block.
3103
3104 * CVE-2012-4412 The strcoll implementation caches indices and rules for
3105 large collation sequences to optimize multiple passes. This cache
3106 computation may overflow for large collation sequences and may cause a
3107 stack or buffer overflow. This is now fixed to use a slower algorithm
3108 which does not use a cache if there is an integer overflow.
3109
3110 * CVE-2012-4424 The strcoll implementation uses malloc to cache indices and
3111 rules for large collation sequences to optimize multiple passes and falls
3112 back to alloca if malloc fails, resulting in a possible stack overflow.
3113 The implementation now falls back to an uncached collation sequence lookup
3114 if malloc fails.
3115
3116 * CVE-2013-4788 The pointer guard used for pointer mangling was not
3117 initialized for static applications resulting in the security feature
3118 being disabled. The pointer guard is now correctly initialized to a
3119 random value for static applications. Existing static applications need
3120 to be recompiled to take advantage of the fix (bug 15754).
3121
3122 * CVE-2013-4237 The readdir_r function could write more than NAME_MAX bytes
3123 to the d_name member of struct dirent, or omit the terminating NUL
3124 character. (Bugzilla #14699).
3125
3126 * CVE-2013-4332 The pvalloc, valloc, memalign, posix_memalign and
3127 aligned_alloc functions could allocate too few bytes or corrupt the
3128 heap when passed very large allocation size values (Bugzilla #15855,
3129 #15856, #15857).
3130
3131 * CVE-2013-4458 Stack overflow in getaddrinfo with large number of results
3132 for AF_INET6 has been fixed (Bugzilla #16072).
3133
3134 * New locales: ak_GH, anp_IN, ar_SS, cmn_TW, hak_TW, lzh_TW, nan_TW, pap_AW,
3135 pap_CW, quz_PE, the_NP.
3136
3137 * Substantially revised locales: gd_GB, ht_HT
3138
3139 * The LC_ADDRESS field was updated to support country_car for almost all
3140 supported locales.
3141
3142 * ISO 1427 definitions were updated.
3143
3144 * ISO 3166 definitions were updated.
3145
3146 * The localedef utility now supports --big-endian and --little-endian
3147 command-line options to generate locales for a different system from that
3148 for which the C library was built.
3149
3150 * Binary locale files now only depend on the endianness of the system for
3151 which they are generated and not on other properties of that system. As a
3152 consequence, binary files generated with new localedef may be incompatible
3153 with old versions of the GNU C Library, and binary files generated with
3154 old localedef may be incompatible with this version of the GNU C Library,
3155 in the following circumstances:
3156
3157 + Locale files may be incompatible on m68k systems.
3158
3159 + Locale archive files (but not separate files for individual locales) may
3160 be incompatible on systems where plain "char" is signed.
3161
3162 * The configure option --disable-versioning has been removed. Builds with
3163 --disable-versioning had not worked for several years.
3164
3165 * ISO 639 definitions were updated for Chiga (cgg) and Chinese (gan, hak, czh,
3166 cjy, lzh, cmn, mnp, cdo, czo, cpx, wuu, hsn, yue).
3167
3168 * SystemTap probes for malloc have been introduced.
3169
3170 * SystemTap probes for slow multiple precision fallback paths of
3171 transcendental functions have been introduced.
3172
3173 * Support for powerpc64le has been added.
3174
3175 * The soft-float powerpc port now supports e500 processors.
3176
3177 * Support for STT_GNU_IFUNC symbols added for ppc32/power4+ and ppc64.
3178
3179 * A new feature test macro _DEFAULT_SOURCE is available to enable the same
3180 set of header declarations that are enabled by default, even when other
3181 feature test macros or compiler options such as -std=c99 would otherwise
3182 disable some of those declarations.
3183
3184 * The _BSD_SOURCE feature test macro no longer enables BSD interfaces that
3185 conflict with POSIX. The libbsd-compat library (which was a dummy library
3186 that did nothing) has also been removed.
3187
3188 * Preliminary documentation about Multi-Thread, Async-Signal and
3189 Async-Cancel Safety has been added.
3190
3191 * Change of ABI data structures for s390 and s390x:
3192 On s390 and s390x the size of struct ucontext and jmp_buf was increased to
3193 allow for future hardware extensions. All library functions that accept or
3194 return these structures were versioned in order to provide backwards
3195 compatibility with existing code. However, not all cases can be handled
3196 with interface versioning. If an application embeds either structure into
3197 another structure and that structure is passed to another compilation unit
3198 which expects the newer larger-sized object then any access to the new
3199 fields will result in undefined behaviour. Similarly any access to fields
3200 that were moved by the enlarging of the structures will cause undefined
3201 behaviour. To fix the undefined behaviour all objects that are part of the
3202 final in-memory image must agree on the size of structures, and this may
3203 require recompilation.
3204 \f
3205 Version 2.18
3206
3207 * The following bugs are resolved with this release:
3208
3209 2546, 2560, 5159, 6809, 7006, 10060, 10062, 10283, 10357, 10686, 11120,
3210 11561, 12310, 12387, 12492, 12515, 12723, 13550, 13889, 13951, 13988,
3211 14142, 14176, 14200, 14256, 14280, 14293, 14317, 14327, 14478, 14496,
3212 14582, 14686, 14812, 14888, 14894, 14907, 14908, 14909, 14920, 14941,
3213 14952, 14964, 14981, 14982, 14985, 14991, 14994, 14996, 15000, 15003,
3214 15006, 15007, 15014, 15020, 15022, 15023, 15036, 15054, 15055, 15062,
3215 15078, 15084, 15085, 15086, 15100, 15160, 15214, 15221, 15232, 15234,
3216 15283, 15285, 15287, 15304, 15305, 15307, 15309, 15327, 15330, 15335,
3217 15336, 15337, 15339, 15342, 15346, 15359, 15361, 15366, 15380, 15381,
3218 15394, 15395, 15405, 15406, 15409, 15416, 15418, 15419, 15423, 15424,
3219 15426, 15429, 15431, 15432, 15441, 15442, 15448, 15465, 15480, 15485,
3220 15488, 15490, 15492, 15493, 15497, 15506, 15529, 15536, 15553, 15577,
3221 15583, 15618, 15627, 15631, 15654, 15655, 15666, 15667, 15674, 15711,
3222 15755, 15759.
3223
3224 * CVE-2013-2207 Incorrectly granting access to another user's pseudo-terminal
3225 has been fixed by disabling the use of pt_chown (Bugzilla #15755).
3226 Distributions can re-enable building and using pt_chown via the new configure
3227 option `--enable-pt_chown'. Enabling the use of pt_chown carries with it
3228 considerable security risks and should only be used if the distribution
3229 understands and accepts the risks.
3230
3231 * CVE-2013-0242 Buffer overrun in regexp matcher has been fixed (Bugzilla
3232 #15078).
3233
3234 * CVE-2013-1914 Stack overflow in getaddrinfo with many results has been
3235 fixed (Bugzilla #15330).
3236
3237 * Add support for calling C++11 thread_local object destructors on thread
3238 and program exit. This needs compiler support for offloading C++11
3239 destructor calls to glibc.
3240
3241 * Improved worst case performance of libm functions with double inputs and
3242 output.
3243
3244 * Support for priority inherited mutexes in pthread condition variables on
3245 non-x86 architectures.
3246
3247 * Port to Xilinx MicroBlaze contributed by David Holsgrove.
3248
3249 * Optimized string functions for AArch64. Implemented by Marcus Shawcroft.
3250
3251 * Optimized string functions for ARM. Implemented by Will Newton and
3252 Richard Henderson.
3253
3254 * Optimized libm functions for SPARC. Implemented by David S. Miller.
3255
3256 * Support for building more of ARM glibc as Thumb-2 code. Implemented by
3257 Richard Henderson.
3258
3259 * Support for building most of MIPS glibc as MIPS16 code. Implemented by
3260 Chung-Lin Tang, Maciej W. Rozycki and Maxim Kuvyrkov.
3261
3262 * Added a benchmark framework to track performance of functions in glibc.
3263
3264 * New <math.h> macro named issignaling to check for a signaling NaN (sNaN).
3265 It is based on draft TS 18661 and currently enabled as a GNU extension.
3266
3267 * On Linux, the clock function now uses the clock_gettime system call
3268 for improved precision, rather than old times system call.
3269
3270 * Added support for version-3 tz format files. This is needed when using
3271 the tz database release 2013e or later, and affects a few unusual cases --
3272 currently only TZ='America/Godthab' for time stamps after 2037.
3273
3274 * Added new API functions pthread_getattr_default_np and
3275 pthread_setattr_default_np to get and set the default pthread
3276 attributes of a process.
3277
3278 * Added support for TSX lock elision for pthread mutexes on i386 and x86-64.
3279 This may improve lock scaling of existing programs on TSX capable systems.
3280 When the --enable-lock-elision=yes parameter is specified at configure
3281 time lock elision will be enabled by default for all PTHREAD_MUTEX_DEFAULT
3282 mutexes.
3283
3284 * Support added for AT_HWCAP2 (to coincide with Linux kernel 3.10
3285 availability). Implemented by Ryan S. Arnold.
3286
3287 * Support added for POWER8 platform. Implemented by Ryan S. Arnold.
3288 \f
3289 Version 2.17
3290
3291 * The following bugs are resolved with this release:
3292
3293 1349, 3439, 3479, 3665, 5044, 5246, 5298, 5400, 6530, 6677, 6778, 6808,
3294 9685, 9914, 10014, 10038, 10114, 10191, 10631, 10873, 11438, 11607, 11638,
3295 11741, 12140, 13013, 13412, 13542, 13601, 13603, 13604, 13629, 13679,
3296 13696, 13698, 13717, 13741, 13759, 13761, 13763, 13881, 13939, 13950,
3297 13952, 13966, 14042, 14047, 14090, 14150, 14151, 14152, 14154, 14157,
3298 14166, 14173, 14195, 14197, 14237, 14246, 14251, 14252, 14283, 14298,
3299 14303, 14307, 14328, 14331, 14336, 14337, 14347, 14349, 14368, 14376,
3300 14417, 14447, 14459, 14476, 14477, 14501, 14505, 14510, 14516, 14518,
3301 14519, 14530, 14532, 14538, 14543, 14544, 14545, 14557, 14562, 14568,
3302 14576, 14579, 14583, 14587, 14595, 14602, 14610, 14621, 14638, 14645,
3303 14648, 14652, 14660, 14661, 14669, 14672, 14683, 14694, 14716, 14719,
3304 14743, 14767, 14783, 14784, 14785, 14793, 14796, 14797, 14801, 14803,
3305 14805, 14807, 14811, 14815, 14821, 14822, 14824, 14828, 14831, 14833,
3306 14835, 14838, 14856, 14863, 14865, 14866, 14868, 14869, 14871, 14872,
3307 14879, 14889, 14893, 14898, 14914.
3308
3309 * Optimization of memcpy for MIPS.
3310
3311 * CVE-2011-4609 svc_run() produces high cpu usage when accept fails with
3312 EMFILE has been fixed (Bugzilla #14889).
3313
3314 * The add-on ports collection is now distributed in the "ports" subdirectory
3315 of the main GNU C Library distribution, rather than separately.
3316
3317 * Port to ARM AArch64 contributed by Linaro.
3318
3319 * Support for STT_GNU_IFUNC symbols added for s390 and s390x.
3320 Optimized versions of memcpy, memset, and memcmp added for System z10 and
3321 zEnterprise z196.
3322 Implemented by Andreas Krebbel.
3323
3324 * The new function secure_getenv allows secure access to the environment,
3325 returning NULL if running in a SUID/SGID process. This function replaces
3326 the internal function __secure_getenv.
3327
3328 * SystemTap static probes have been added into the dynamic linker.
3329 Implemented by Gary Benson.
3330
3331 * Optimizations of string functions strstr, strcasestr and memmem.
3332 Implemented by Maxim Kuvyrkov.
3333
3334 * The minimum Linux kernel version that this version of the GNU C Library
3335 can be used with is 2.6.16.
3336
3337 * Optimizations of string functions memchr, wcschr, wcscpy, and wcsrchr for
3338 powerpc POWER7. Implemented by Will Schmidt.
3339
3340 * New configure option --disable-nscd builds the C library such that it
3341 never attempts to contact the Name Service Caching Daemon (nscd).
3342 New configure option --disable-build-nscd avoids building nscd itself;
3343 this is the default if --disable-nscd is used.
3344
3345 * Improved support for cross-compilation, including cross-testing and
3346 bootstrap builds without a previously built glibc.
3347
3348 * Several testsuite tests are now able to test multiple IFUNC variants of an
3349 interface, rather than just testing the one that would be chooen by
3350 default.
3351
3352 * New configure options --with-bugurl and --with-pkgversion, for
3353 distributors to use to embed their bug-reporting and package version
3354 information in --help and --version output.
3355
3356 * The ttyname and ttyname_r functions on Linux now fall back to searching for
3357 the tty file descriptor in /dev/pts or /dev if /proc is not available. This
3358 allows creation of chroots without the procfs mounted on /proc.
3359
3360 * The `crypt' function now fails if passed salt bytes that violate the
3361 specification for those values. On Linux, the `crypt' function will
3362 consult /proc/sys/crypto/fips_enabled to determine if "FIPS mode" is
3363 enabled, and fail on encrypted strings using the MD5 or DES algorithm
3364 when the mode is enabled.
3365
3366 * The `clock_*' suite of functions (declared in <time.h>) is now available
3367 directly in the main C library. Previously it was necessary to link with
3368 -lrt to use these functions. This change has the effect that a
3369 single-threaded program that uses a function such as `clock_gettime' (and
3370 is not linked with -lrt) will no longer implicitly load the pthreads
3371 library at runtime and so will not suffer the overheads associated with
3372 multi-thread support in other code such as the C++ runtime library.
3373
3374 * New locales: ayc_PE, doi_IN, ia_FR, mni_IN, nhn_MX, niu_NU, niu_NZ,
3375 sat_IN, and szl_PL.
3376 \f
3377 Version 2.16
3378
3379 * The following bugs are resolved with this release:
3380
3381 174, 208, 350, 369, 411, 706, 766, 2074, 2541, 2547, 2548, 2550, 2551,
3382 2552, 2553, 2554, 2562, 2563, 2565, 2566, 2570, 2576, 2636, 2678, 3335,
3383 3440, 3748, 3768, 3866, 3868, 3906, 3976, 3992, 4026, 4108, 4596, 4822,
3384 5077, 5461, 5805, 5993, 6471, 6486, 6578, 6649, 6730, 6770, 6794, 6884,
3385 6890, 6894, 6895, 6907, 6911, 6959, 7064, 9739, 9902, 10110, 10135, 10140,
3386 10153, 10210, 10254, 10346, 10375, 10545, 10716, 10846, 10882, 11174,
3387 11261, 11322, 11365, 11451, 11494, 11521, 11677, 11837, 11959, 12047,
3388 12097, 12193, 12194, 12297, 12298, 12301, 12340, 12354, 12416, 12495,
3389 13058, 13223, 13361, 13525, 13526, 13527, 13528, 13529, 13530, 13531,
3390 13532, 13533, 13547, 13551, 13552, 13553, 13555, 13556, 13559, 13563,
3391 13566, 13576, 13579, 13583, 13592, 13594, 13613, 13618, 13637, 13656,
3392 13658, 13673, 13691, 13695, 13704, 13705, 13706, 13718, 13726, 13738,
3393 13739, 13743, 13750, 13758, 13760, 13761, 13775, 13786, 13787, 13792,
3394 13806, 13824, 13840, 13841, 13844, 13846, 13848, 13851, 13852, 13854,
3395 13871, 13872, 13873, 13879, 13882, 13883, 13884, 13885, 13886, 13892,
3396 13895, 13908, 13910, 13911, 13912, 13913, 13914, 13915, 13916, 13917,
3397 13918, 13919, 13920, 13921, 13922, 13923, 13924, 13926, 13927, 13928,
3398 13938, 13941, 13942, 13954, 13955, 13956, 13963, 13967, 13968, 13970,
3399 13973, 13979, 13983, 13986, 13996, 14012, 14027, 14033, 14034, 14036,
3400 14040, 14043, 14044, 14048, 14049, 14050, 14053, 14055, 14059, 14064,
3401 14075, 14080, 14083, 14103, 14104, 14109, 14112, 14117, 14122, 14123,
3402 14134, 14153, 14183, 14188, 14199, 14210, 14218, 14229, 14241, 14273,
3403 14277, 14278.
3404
3405 * Support for the x32 ABI on x86-64 added. The x32 target is selected by
3406 configuring glibc with:
3407 BUILD_CC='gcc' CC='gcc -mx32' CXX='g++ -mx32'
3408 Visit <http://sites.google.com/site/x32abi/> for more x32 ABI info.
3409 Implemented by H.J. Lu.
3410
3411 * ISO C11 support:
3412
3413 + define static_assert
3414
3415 + do not declare gets
3416
3417 + declare at_quick_exit and quick_exit also for ISO C11
3418
3419 + aligned_alloc. NB: The code is deliberately allows the size parameter
3420 to not be a multiple of the alignment. This is a moronic requirement
3421 in the standard but it is only a requirement on the caller, not the
3422 implementation.
3423
3424 + timespec_get added
3425
3426 + uchar.h support added
3427
3428 + CMPLX, CMPLXF, CMPLXL added
3429
3430 Implemented by Ulrich Drepper.
3431
3432 * Support for the IA-64 has been moved to ports.
3433
3434 * Remove support for anything but ELF binary format
3435
3436 * Checking versions of poll, ppoll added.
3437 Implemented by Ulrich Drepper.
3438
3439 * More generic and 64-bit performance optimizations to math functions.
3440 Implemented by Ulrich Drepper.
3441
3442 * New configure option --enable-obsolete-rpc makes the deprecated RPC
3443 headers and functions available at compile time as they were before
3444 version 2.14. This option will be removed at some time in the future
3445 after the TI-RPC library becomes fully sufficient for the needs of
3446 existing applications.
3447
3448 * Compatibility code for Linux kernel versions before 2.4 has been removed.
3449 Note that glibc is not expected to work with any Linux kernel version
3450 before 2.6.
3451
3452 * New header <sys/auxv.h> and function getauxval allowing easy access to
3453 the AT_* key-value pairs passed from the Linux kernel. The header also
3454 defines the HWCAP_* bits associated with the AT_HWCAP key.
3455
3456 * New locales: mag_IN
3457
3458 * New configure option --enable-systemtap builds SystemTap static probes
3459 into libc for setjmp and longjmp and into libpthread for various operations.
3460 So far the setjmp/longjmp probes and some of the libpthread probes are
3461 provided only for i*86 and x86_64.
3462 Implemented by Roland McGrath and Rayson Ho.
3463
3464 * Optimized expf for x86-32 and x86-64. Implemented by Liubov Dmitrieva.
3465
3466 * More optimized functions for PowerPC. Implemented by Adhemerval Zanella
3467 and Will Schmidt.
3468
3469 * More optimized functions for SPARC. Implemented by David S. Miller.
3470
3471 * Improved support for cross-compilation, especially bootstrap builds
3472 without a previously built glibc.
3473
3474 * Ports for the TILE-Gx and TILEPro families of processors. Contributed by
3475 Chris Metcalf from Tilera.
3476
3477 * Support for the old ARM ABI has been removed from ports. Only the EABI is
3478 now supported for ARM processors.
3479
3480 * The hard-float variant of the ARM EABI now uses /lib/ld-linux-armhf.so.3
3481 as the name of the dynamic linker, to distinguish it from the
3482 /lib/ld-linux.so.3 dynamic linker used for the base EABI.
3483
3484 * Support for CRIS has been removed from ports.
3485
3486 * A new class of installed header has been documented for low-level
3487 platform-specific functionality. PowerPC added the first instance with a
3488 function to provide time base register access. Contributed by Tulio
3489 Magno Quites Machado Filho.
3490
3491 * ix86 configurations now install header files that are consistent with
3492 what x86-64 configurations install. These same header files can be used
3493 for -m32, -m64, or -mx32 builds.
3494 Contributed by H.J. Lu.
3495
3496 * Math library bug fixes. A thorough audit of all open math library bugs was
3497 conducted by Joseph Myers. Significant progress was made on many math
3498 library bugs resulting in more accurate exceptions and function results.
3499 Many thanks to all those that contributed including Andreas Jaeger for his
3500 patch review and work on the x87 trigonometric instruction issues.
3501
3502 * Timezone data is no longer installed. Timezone-related binaries and scripts
3503 will continue to be installed. Users should obtain their timezone data from
3504 their distribution provider or from the tzdata package at
3505 <ftp://munnari.oz.au/pub/>.
3506 \f
3507 Version 2.15
3508
3509 * The following bugs are resolved with this release:
3510
3511 6779, 6783, 9696, 10103, 10709, 11589, 11929, 12403, 12786, 12840, 12847,
3512 12868, 12852, 12874, 12885, 12892, 12906, 12907, 12922, 12935, 12962,
3513 13007, 13021, 13061, 13062, 13067, 13068, 13085, 13088, 13090, 13092,
3514 13096, 13114, 13118, 13123, 13134, 13138, 13147, 13150, 13166, 13179,
3515 13185, 13189, 13192, 13268, 13276, 13282, 13291, 13305, 13328, 13335,
3516 13337, 13344, 13358, 13367, 13413, 13416, 13423, 13439, 13446, 13472,
3517 13484, 13506, 13515, 13523, 13524, 13538, 13540
3518
3519 * New program pldd to list loaded object of a process
3520 Implemented by Ulrich Drepper.
3521
3522 * Add nss_db support back to glibc. No more dependency on Berkeley db
3523 and support for initgroups lookups.
3524 Implemented by Ulrich Drepper.
3525
3526 * Optimized strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-32.
3527 Contributed by HJ Lu.
3528
3529 * Improved strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-64.
3530 Contributed by HJ Lu.
3531
3532 * Optimized strcat, strncat on x86-64 and optimized wcscmp, wcslen, strnlen
3533 on x86-32 and x86-64.
3534 Contributed by Liubov Dmitrieva.
3535
3536 * Optimized strchr and strrchr for SSE on x86-32.
3537 Contributed by Liubov Dmitrieva.
3538
3539 * Optimized memchr, memrchr, rawmemchr, memcmp, wmemcmp, wcschr, wcscpy
3540 for x86-64 and x86-32.
3541 Contributed by Liubov Dmitrieva.
3542
3543 * New interfaces: scandirat, scandirat64
3544 Implemented by Ulrich Drepper.
3545
3546 * Checking versions of FD_SET, FD_CLR, and FD_ISSET added.
3547 Implemented by Ulrich Drepper.
3548
3549 * nscd now also caches the netgroup database.
3550 Implemented by Ulrich Drepper.
3551
3552 * Integrate libm with gcc's -ffinite-math-only option.
3553 Implemented by Ulrich Drepper.
3554
3555 * Lots of generic, 64-bit, and x86-64-specific performance optimizations
3556 to math functions. Implemented by Ulrich Drepper.
3557
3558 * Optimized strcasecmp and strncasecmp for AVX on x86-64.
3559 Implemented by Ulrich Drepper.
3560
3561 * New Linux interfaces: process_vm_readv, process_vm_writev
3562
3563 * Optimized strcasecmp and strncasecmp for SSSE3 and SSE4.2 on x86-32.
3564 Implemented by Ulrich Drepper.
3565
3566 * Optimized nearbyint and strcasecmp for PPC.
3567 Implemented by Adhemerval Zanella.
3568
3569 * New locales: bho_IN, unm_US, es_CU, ta_LK
3570 \f
3571 Version 2.14
3572
3573 * The following bugs are resolved with this release:
3574
3575 386, 6420, 7101, 9730, 9732, 9809, 10138, 10149, 10157, 11099, 11257,
3576 11258, 11487, 11532, 11558, 11578, 11634, 11653, 11668, 11697, 11724,
3577 11781, 11799, 11820, 11837, 11857, 11884, 11892, 11895, 11901, 11945,
3578 11947, 11952, 11987, 12052, 12083, 12158, 12178, 12200, 12346, 12350,
3579 12393, 12420, 12432, 12445, 12449, 12453, 12454, 12460, 12469, 12489,
3580 12509, 12510, 12511, 12518, 12527, 12541, 12545, 12551, 12582, 12583,
3581 12587, 12597, 12601, 12611, 12625, 12626, 12631, 12650, 12653, 12655,
3582 12660, 12671, 12681, 12684, 12685, 12711, 12713, 12714, 12717, 12723,
3583 12724, 12734, 12738, 12746, 12766, 12775, 12777, 12782, 12788, 12792,
3584 12795, 12811, 12813, 12814, 12841
3585
3586 * The RPC implementation in libc is obsoleted. Old programs keep working
3587 but new programs cannot be linked with the routines in libc anymore.
3588 Programs in need of RPC functionality must be linked against TI-RPC.
3589 The TI-RPC implementation is IPv6 enabled and there are other benefits.
3590
3591 Visible changes of this change include (obviously) the inability to link
3592 programs using RPC functions without referencing the TI-RPC library and the
3593 removal of the RPC headers from the glibc headers.
3594 Implemented by Ulrich Drepper.
3595
3596 * New Linux interfaces: clock_adjtime, name_to_handle_at, open_by_handle_at,
3597 syncfs, setns, sendmmsg
3598
3599 * New locales: os_RU, bem_ZA, en_ZA, ff_SN, sw_KE, sw_TZ, lb_LU, wae_CH,
3600 yue_HK, lij_IT, mhr_RU
3601
3602 * New encodings: CP770, CP771, CP772, CP773, CP774
3603
3604 * New program sotruss to trace calls through PLTs
3605 Implemented by Ulrich Drepper.
3606
3607 * The malloc hook implementation is marked deprecated and will be removed
3608 from the default implementation in the next version. The design never
3609 worked ever since the introduction of threads. Even programs which do
3610 not create threads themselves can use multiple threads created internally.
3611 \f
3612 Version 2.13
3613
3614 * The following bugs are resolved with this release:
3615
3616 3268, 6812, 7066, 10085, 10484, 10563, 10851, 11149, 11155, 11577, 11611,
3617 11640, 11655, 11701, 11840, 11856, 11883, 11903, 11904, 11968, 11979,
3618 12005, 12037, 12067, 12077, 12078, 12092, 12093, 12107, 12108, 12113,
3619 12140, 12159, 12167, 12191, 12194, 12201, 12204, 12205, 12207, 12348,
3620 12378, 12394, 12397
3621
3622 * New Linux interfaces: prlimit, prlimit64, fanotify_init, fanotify_mark
3623
3624 * POWER7 optimizations: memset, memcmp, strncmp
3625
3626 * New optimized string functions for x86-64: strnlen (SSE2),
3627 strcasecmp (SSE2, SSSE3, SSE4.2), strncasecmp (SSE2, SSSE3, SSE4.2)
3628 Implemented by Ulrich Drepper.
3629 \f
3630 Version 2.12
3631
3632 * The following bugs are resolved with this release:
3633
3634 3662, 4457, 5553, 10162, 10401, 10414, 10554, 10824, 10864, 10915,
3635 10918, 10936, 10939, 10958, 10968, 10969, 10972, 10992, 11000, 11001,
3636 11007, 11010, 11027, 11039, 11040, 11041, 11043, 11046, 11056, 11070,
3637 11093, 11115, 11125, 11126, 11127, 11134, 11141, 11149, 11183, 11184,
3638 11185, 11186, 11187, 11188, 11189, 11190, 11191, 11192, 11193, 11194,
3639 11200, 11230, 11235, 11242, 11254, 11258, 11271, 11272, 11276, 11279,
3640 11287, 11292, 11319, 11332, 11333, 11387, 11389, 11390, 11394, 11397,
3641 11410, 11438, 11449, 11470, 11471, 11520, 11537, 11538, 11571
3642
3643 * New interfaces: pthread_getname_np, pthread_setname_np
3644
3645 * New Linux interface: recvmmsg
3646
3647 * STT_GNU_IFUNC implemented for Sparc by David Miller.
3648
3649 * The dynamic linker now recognizes supported ABI versions from the
3650 EI_ABIVERSION field in the ELF header.
3651 Implemented by Ulrich Drepper.
3652
3653 * New NIS mode selector ADJUNCT_AS_SHADOW. The passwd.adjunct.byname table
3654 will not be used to fill in password fields in the passwd.byname replies.
3655 Instead it is used to synthesize the shadow.byname table, should it be
3656 missing. This is a useful mode in some installations involving Solaris.
3657 Implemented by Ulrich Drepper.
3658
3659 * New locales: kok_IN, sq_MK, cv_RU
3660 \f
3661 Version 2.11
3662
3663 * The following bugs are resolved with this release:
3664
3665 3270, 7094, 9924, 9986, 10011, 10085, 10107, 10159, 10162, 10166, 10183,
3666 10195, 10196, 10200, 10203, 10207, 10211, 10213, 10217, 10221, 10229,
3667 10262, 10286, 10312, 10315, 10319, 10349, 10360, 10391, 10402, 10416,
3668 10418, 10422, 10425, 10446, 10448, 10452, 10540, 10553, 10560, 10564,
3669 10609, 10643, 10692, 10713, 10717, 10730, 10731, 10742, 10780, 10784,
3670 10789, 10817, 10823, 10840, 10847
3671
3672 * New interfaces: execvpe, pthread_sigqueue, mkstemps, mkstemps64, mkostemps,
3673 mkostemps64
3674 Implemented by Ulrich Drepper.
3675
3676 * Checking version of longjmp added that fails if an uninitialized stack
3677 frame would be created. Implemented by Ulrich Drepper.
3678
3679 * STT_GNU_IFUNC is now supported in static executables.
3680 Implemented by H.J. Lu.
3681
3682 * STT_GNU_IFUNC implemented for PPC by Alan Modra.
3683
3684 * New optimized string functions for x86-64: strstr, strcasestr, memcmp,
3685 strcspn, strpbrk, strspn, strcpy, stpcpy, strncpy, strcmp (SSE2, SSE4.2),
3686 strncmp (SSE2, SSE4.2), strchr (SSE4.2), strrchr (SSE4.2).
3687 Contributed by H.J. Lu.
3688
3689 strlen, rawmemchr, strcmp (SSSE3), strncmp (SSSE3).
3690 Implemented by Ulrich Drepper.
3691
3692 * New optimized string functions for x86: strlen, strcspn, strspn, strpbrk,
3693 strstr, strcasestr.
3694 Contributed by H.J. Lu.
3695
3696 * Support for fma instruction in AVX on x86-64.
3697 Implemented by H.J. Lu and Ulrich Drepper.
3698
3699 * AVX support in x86-64 auditing support in ld.so.
3700 Implemented by H.J. Lu.
3701
3702 * STB_GNU_UNIQUE support added.
3703 Implemented by Ulrich Drepper.
3704
3705 * Implement second fallback mode for optimized DNS lookup for even more
3706 broken environments. If two requests from the same port are not
3707 handled correctly close the socket and open a new one before sending
3708 the second request. The 'single-request-reopen' option in /etc/resolv.conf
3709 can be used to select this mode right away, instead of rediscovering the
3710 necessity is every process again.
3711 Implemented by Ulrich Drepper.
3712
3713 * New resolver flag RES_USE_DNSSEC to enable use of verified lookup.
3714 Implemented by Adam Tkac.
3715
3716 * Optimized iconv conversions for S390x.
3717 Implemented by Andreas Krebbel.
3718
3719 * Using condvars with PI mutexes is now more efficient due to kernel
3720 support for requeueing to PI futexes. NPTL support added for x86-64.
3721 Implemented by Ulrich Drepper.
3722
3723 * New locales: ps_AF, my_MM
3724 \f
3725 Version 2.10
3726
3727 * The following bugs are resolved with this release:
3728
3729 697, 5381, 5807, 6411, 6545, 7040, 7058, 7067, 7080, 7083, 7095, 7098,
3730 9677, 9694, 9697, 9701, 9704, 9705, 9706, 9720, 9726, 9730, 9733, 9736,
3731 9741, 9750, 9753, 9759, 9781, 9791, 9793, 9823, 9833, 9844, 9880, 9881,
3732 9891, 9893, 9895, 9913, 9916, 9920, 9948, 9955, 9957, 9963, 9985, 10052,
3733 10069, 10086, 10087, 10090, 10092, 10093, 10100, 10118, 10128
3734
3735 * New interfaces: preadv, preadv64, pwritev, pwritev64, malloc_info
3736 Implemented by Ulrich Drepper.
3737
3738 * New Linux interfaces: accept4, fallocate, fallocate64.
3739 Implemented by Ulrich Drepper.
3740
3741 * Correct declarations of string function when used in C++ code. This
3742 could lead to compile errors for invalid C++ code.
3743
3744 * XPG7/POSIX 2008 compilation environment. Many GNU-specific functions are
3745 now in POSIX.
3746
3747 * New POSIX 2008 interface: psiginfo
3748 Implemented by Ulrich Drepper.
3749
3750 * New ISO C++1x interfaces: quick_exit, at_quick_exit
3751 Implemented by Ulrich Drepper.
3752
3753 * Support for selecting between multiple function definitions at runtime
3754 using STT_GNU_IFUNC symbols. Implemented by Ulrich Drepper.
3755
3756 * The libcrypt library can now use the hash function implementations in
3757 NSS. Implemented by Ulrich Drepper.
3758
3759 * The malloc implementation can be compiled to be less memory efficient
3760 but higher performing in multi-threaded programs.
3761 Implemented by Ulrich Drepper.
3762
3763 * New locales: nan_TW@latin, ks_IN
3764
3765 * Faster strlen, strchr, strchrnul, strrchr, memchr, and rawmemchr for x86-64.
3766 Implemented by Ulrich Drepper.
3767
3768 * Extended printf hook support. It is possible to use user-defined types
3769 and extend existing format specifiers.
3770 Implemented by Ulrich Drepper.
3771
3772 * Handling for group shadow files has been added.
3773 Implemented by Ulrich Drepper.
3774
3775 * DNS IPv4-IPv6 parallel lookup now deals better with broken DNS
3776 servers (the case, e.g., for some people using the built-in DNS
3777 server in ADSL modems/routers). There is a once-per-process timeout
3778 in case of a broken server. To avoid it, users can run nscd or put
3779 'options single-request' in /etc/resolv.conf.
3780 Implemented by Ulrich Drepper.
3781 \f
3782 Version 2.9
3783
3784 * The following bugs are resolved with this release:
3785
3786 3406, 5209, 5210, 5381, 5794, 5814, 5911, 6428, 6438, 6442, 6447, 6461,
3787 6472, 6505, 6544, 6589, 6612, 6634, 6653, 6654, 6657, 6698, 6712, 6713,
3788 6719, 6723, 6724, 6740, 6763, 6771, 6790, 6791, 6817, 6824, 6839, 6843,
3789 6845, 6856, 6857, 6860, 6867, 6875, 6919, 6920, 6942, 6947, 6955, 6966,
3790 6968, 6974, 6980, 6995, 7008, 7009, 7029
3791
3792 * Unified lookup for getaddrinfo: IPv4 and IPv6 addresses are now looked
3793 up at the same time. Implemented by Ulrich Drepper.
3794
3795 * TLS descriptors for LD and GD on x86 and x86-64.
3796 Implemented by Alexandre Oliva.
3797
3798 * getaddrinfo now handles DCCP and UDPlite.
3799 Implemented by Ulrich Drepper.
3800
3801 * New fixed-size conversion macros: htobe16, htole16, be16toh, le16toh,
3802 htobe32, htole32, be32toh, le32toh, htobe64, htole64, be64toh, le64toh.
3803 Implemented by Ulrich Drepper.
3804
3805 * New implementation of memmem, strstr, and strcasestr which is O(n).
3806 Implemented by Eric Blake.
3807
3808 * New Linux interfaces: inotify_init1, dup3, epoll_create1, pipe2
3809
3810 * Implement "e" option for popen to open file descriptor with the
3811 close-on-exec flag set. Implemented by Ulrich Drepper.
3812
3813 * Implement "b" mode for fmemopen. In this mode writes writes don't
3814 implicitly add a NUL byte and seeks from the end of the buffer really
3815 use the buffer end, not the string length as the basis.
3816 Implemented by Ulrich Drepper.
3817
3818 * Many functions, exported and internal, now atomically set the close-on-exec
3819 flag when run on a sufficiently new kernel. Implemented by Ulrich Drepper.
3820
3821 * Sorting rules for some Indian languages (Kannada, Gurumukhi, Telugu,
3822 Sinhala)
3823 Implemented by Pravin Satpute.
3824
3825 * New locales: sd_IN, sd_IN@devanagari, ks_IN@devanagari, ht_HT, en_AG, nl_AW.
3826 \f
3827 Version 2.8
3828
3829 * The following bugs are resolved with this release:
3830
3831 2549, 3112, 4314, 4407, 4997, 5012, 5040, 5112, 5204, 5208, 5209, 5220,
3832 5222, 5224, 5225, 5237, 5238, 5240, 5245, 5277, 5346, 5375, 5378, 5382,
3833 5424, 5427, 5428, 5435, 5436, 5439, 5441, 5442, 5443, 5451, 5452, 5454,
3834 5455, 5463, 5464, 5465, 5475, 5477, 5478, 5520, 5541, 5545, 5553, 5600,
3835 5602, 5607, 5614, 5627, 5628, 5736, 5737, 5741, 5753, 5760, 5762, 5768,
3836 5774, 5777, 5778, 5779, 5786, 5790, 5818, 5854, 5857, 5903, 5939, 5979,
3837 5983, 5995, 5998, 6004, 6007, 6020, 6021, 6024, 6040, 6042
3838
3839 * New locales: bo_CN, bo_IN, shs_CA.
3840
3841 * New encoding: HP-ROMAN9, HP-GREEK8, HP-THAI8, HP-TURKISH8.
3842
3843 * Sorting rules for some Indian languages (Devanagari and Gujarati).
3844 Implemented by Pravin Satpute.
3845
3846 * IPV6 addresses in /etc/resolv.conf can now have a scope ID
3847
3848 * nscd caches now all timeouts for DNS entries
3849 Implemented by Ulrich Drepper.
3850
3851 * nscd is more efficient and wakes up less often.
3852 Implemented by Ulrich Drepper.
3853
3854 * More checking functions: asprintf, dprintf, obstack_printf, vasprintf,
3855 vdprintf, and obstack_vprintf.
3856 Implemented by Jakub Jelinek.
3857
3858 * Faster memset for x86-64.
3859 Implemented by Harsha Jagasia and H.J. Lu.
3860
3861 * Faster memcpy on x86.
3862 Implemented by Ulrich Drepper.
3863
3864 * ARG_MAX is not anymore constant on Linux. Use sysconf(_SC_ARG_MAX).
3865 Implemented by Ulrich Drepper.
3866
3867 * Faster sqrt and sqrtf implemention for some PPC variants.
3868 Implemented by Steven Munroe.
3869 \f
3870 Version 2.7
3871
3872 * The following bugs are resolved with this release:
3873
3874 4125, 4126, 4178, 4359, 4407, 4512, 4514, 4525, 4554, 4556, 4557, 4566,
3875 4582, 4586, 4588, 4599, 4610, 4647, 4702, 4705, 4726, 4745, 4772, 4773,
3876 4775, 4776, 4792, 4813, 4814, 4816, 4833, 4858, 4860, 4896, 4905, 4925,
3877 4936, 4937, 4938, 4941, 4946, 4963, 4972, 5010, 5028, 5043, 5058, 5063,
3878 5071, 5103, 5104, 5112, 5113, 5184, 5186
3879
3880 * More checking functions: fread, fread_unlocked, open*, mq_open.
3881 Implemented by Jakub Jelinek and Ulrich Drepper.
3882
3883 * Extend fortification to C++. Implemented by Jakub Jelinek.
3884
3885 * Implement 'm' modifier for scanf. Add stricter C99/SUS compliance
3886 by not recognizing 'a' as a modifier when those specs are requested.
3887 Implemented by Jakub Jelinek.
3888
3889 * PPC optimizations to math and string functions.
3890 Implemented by Steven Munroe.
3891
3892 * New interfaces: mkostemp, mkostemp64. Like mkstemp* but allow additional
3893 options to be passed. Implemented by Ulrich Drepper.
3894
3895 * More CPU set manipulation functions. Implemented by Ulrich Drepper.
3896
3897 * New Linux interfaces: signalfd, eventfd, eventfd_read, and eventfd_write.
3898 Implemented by Ulrich Drepper.
3899
3900 * Handle private futexes in the NPTL implementation.
3901 Implemented by Jakub Jelinek and Ulrich Drepper.
3902
3903 * Add support for O_CLOEXEC. Implement in Hurd. Use throughout libc.
3904 Implemented by Roland McGrath and Ulrich Drepper.
3905
3906 * Linux/x86-64 vDSO support. Implemented by Ulrich Drepper.
3907
3908 * SHA-256 and SHA-512 based password encryption.
3909 Implemented by Ulrich Drepper.
3910
3911 * New locales: ber_DZ, ber_MA, en_NG, fil_PH, fur_IT, fy_DE, ha_NG, ig_NG,
3912 ik_CA, iu_CA, li_BE, li_NL, nds_DE, nds_NL, pap_AN, sc_IT, tk_TM, ug_CN,
3913 yo_NG.
3914
3915 + New iconv modules: MAC-CENTRALEUROPE, ISO-8859-9E, KOI8-RU.
3916 Implemented by Ulrich Drepper.
3917 \f
3918 Version 2.6
3919
3920 * The following bugs are resolved with this release:
3921
3922 3156, 3213, 3285, 3291, 3306, 3313, 3320, 3322, 3325, 3326, 3334, 3348,
3923 3352, 3363, 3369, 3426, 3427, 3429, 3451, 3458, 3465, 3480, 3483, 3493,
3924 3514, 3515, 3559, 3632, 3664, 3673, 3674, 3745, 3747, 3818, 3840, 3842,
3925 3851, 3855, 3884, 3885, 3902, 3905, 3919, 3944, 3954, 3955, 3957, 3991,
3926 3995, 3996, 4024, 4040, 4069, 4070, 4074, 4076, 4096, 4101, 4102, 4114,
3927 4130, 4131, 4181, 4276, 4306, 4342, 4344, 4349, 4364, 4368, 4381, 4392,
3928 4403, 4405, 4406, 4411, 4438, 4439, 4455, 4456, 4465, 4512, 4514, 4586,
3929 4702, 4858
3930
3931 * New Linux interfaces: epoll_pwait, sched_getcpu.
3932
3933 * New generic interfaces: strerror_l.
3934
3935 * nscd can now cache the services database. Implemented by Ulrich Drepper.
3936 \f
3937 Version 2.5
3938
3939 * The following bugs are resolved with this release:
3940
3941 39, 68, 192, 352, 388, 531, 935, 1006, 1201, 1203, 1386, 1782, 1783, 1784,
3942 1785, 1786, 1787, 1951, 1974, 1996, 2072, 2098, 2125, 2134, 2158, 2182,
3943 2349, 2376, 2386, 2389, 2415, 2418, 2420, 2423, 2450, 2451, 2466, 2472,
3944 2473, 2477, 2498, 2499, 2501, 2502, 2505, 2507, 2508, 2509, 2517, 2526,
3945 2569, 2571, 2592, 2611, 2625, 2632, 2644, 2662, 2680, 2683, 2684, 2693,
3946 2695, 2703, 2734, 2751, 2764, 2766, 2775, 2792, 2821, 2823, 2832, 2841,
3947 2843, 2883, 2892, 2908, 2914, 2926, 2961, 2978, 2980, 2987, 2997, 2998,
3948 3013, 3018, 3034, 3040, 3044, 3123, 3124, 3137, 3138, 3143, 3155, 3189,
3949 3225, 3251, 3252, 3253, 3273
3950
3951 * For Linux, the sorting of addresses returned by getaddrinfo now also
3952 handles rules 3, 4, and 7 from RFC 3484. I.e., all rules are handled.
3953 Implemented by Ulrich Drepper.
3954
3955 * Allow system admin to configure getaddrinfo with the /etc/gai.conf file.
3956 Implemented by Ulrich Drepper.
3957
3958 * New Linux interfaces: splice, tee, sync_file_range, vmsplice.
3959
3960 * New iconv module for MIK. Contributed by Alexander Shopov.
3961
3962 * For sites with broken group and/or passwd database, the auto-propagate
3963 option of nscd can prevent creating ID lookup entries from the results
3964 of a name lookup and vice versa. This usually is no problem but some
3965 site might have problems with the default behavior.
3966 Implemented by Ulrich Drepper.
3967
3968 * Iterating over entire database in NIS can be slow. With the
3969 SETENT_BATCH_READ option in /etc/default/nss a system admin can decide
3970 to trade time for memory. The entire database will be read at once.
3971 Implemented by Ulrich Drepper.
3972
3973 * The interfaces introduced in RFC 3542 have been implemented by
3974 Ulrich Drepper.
3975
3976 * Support for the new ELF hash table format was added by Ulrich Drepper.
3977
3978 * Support for priority inheritance mutexes added by Jakub Jelinek and
3979 Ulrich Drepper.
3980
3981 * Support for priority protected mutexes added by Jakub Jelinek.
3982
3983 * New locales: nr_ZA, pa_PK, ca_AD, ca_FR, ca_IT, el_CY, tr_CY, as_IN,
3984 or_IN, csb_PL, fy_NL, sr_ME.
3985 \f
3986 Version 2.4
3987
3988 * More overflow detection functions.
3989
3990 * New iconv converters for IBM1025, IBM1122, IBM1137, IBM1153,
3991 IBM1154, IBM1155, IBM1156, IBM1157, and IBM1158 contributed by Jiro Sekiba.
3992
3993 More converters for IBM803, IBM901, IBM902, IBM921, IBM1008, IBM1097,
3994 IBM1112, IBM1123, IBM1130, IBM1140, IBM1141, IBM1142, IBM1143, IBM1144,
3995 IBM1145, IBM1146, IBM1147, IBM1148, IBM1149, IBM1166, IBM1167, IBM4517,
3996 IBM4899, IBM4909, IBM4971, IBM5347, IBM9030, IBM9066, IBM9448, IBM12712,
3997 IBM16804, IBM1364, IBM1371, IBM1388, IBM1390, and IBM1399 contributed
3998 by Masahide Washizawa.
3999
4000 * It is now possible to install audit libraries for the dynamic linker using
4001 LD_AUDIT. Implemented by Ulrich Drepper.
4002
4003 * The LinuxThreads add-on, providing pthreads on Linux 2.4 kernels, is no
4004 longer supported. The new NPTL implementation requires Linux 2.6 kernels.
4005 For a libc and libpthread that works well on Linux 2.4 kernels, we
4006 recommend using the stable 2.3 branch.
4007
4008 * The new function open_wmemstream acts like open_memstream,
4009 but uses a wchar_t wide character buffer.
4010
4011 * The new function ppoll is an improved version of poll, similar to pselect.
4012
4013 * New interfaces: fdopendir, openat, fstatat, fchmodat, fchownat,
4014 futimesat, faccessat, mkdirat, mkfifoat, mknodat,
4015 renameat, unlinkat, linkat, symlinkat, readlinkat.
4016
4017 * New Linux kernel system calls: unshare,
4018 inotify_init, inotify_add_watch, inotify_rm_watch.
4019
4020 * The euidaccess function is now also known by the alias eaccess,
4021 for compatibility with some other systems.
4022
4023 * Timezone data updated to 2006b version.
4024 \f
4025 Version 2.3.6
4026
4027 * The following bugs are resolved with this release:
4028
4029 38, 253, 549, 622, 653, 721, 758, 851, 877, 915, 934, 955, 961,
4030 1016, 1037, 1076, 1079, 1080, 1081, 1082, 1083, 1084, 1085, 1086,
4031 1087, 1088, 1090, 1091, 1092, 1093, 1094, 1095, 1096, 1097, 1098,
4032 1099, 1100, 1101, 1102, 1103, 1104, 1105, 1106, 1107, 1108, 1109,
4033 1110, 1111, 1112, 1113, 1125, 1137, 1138, 1249, 1250, 1251, 1252,
4034 1253, 1254, 1350, 1358, 1394, 1438, 1498, 1534
4035
4036 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
4037
4038 * As of this release, GCC 4 can be used to compile the C Library.
4039
4040 * Timezone data updated to 2005m version.
4041 \f
4042 Version 2.3.5
4043
4044 * The following bugs are resolved with this release:
4045
4046 284, 592, 602, 626, 633, 640, 650, 661, 671, 681, 693, 700, 710, 719,
4047 722, 723, 725, 726, 727, 728, 729, 730, 731, 732, 733, 734, 735, 736,
4048 737, 738, 739, 740, 741, 742, 743, 744, 745, 765, 767, 768, 769, 776,
4049 777, 787, 821, 822, 823, 825
4050
4051 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
4052 \f
4053 Version 2.3.4
4054
4055 * Support for RFC 3678. Real implementations exist only for Linux so far.
4056 Implemented by Ulrich Drepper.
4057
4058 * nscd can now cache entries persistently. Expiring entries are reloaded.
4059 For speedups the cache can be shared in memory with client processes.
4060 Implemented by Ulrich Drepper.
4061
4062 * nscd can now perform SELinux checks.
4063 Implemented by Matthew Rickard <mjricka@epoch.ncsc.mil>.
4064
4065 * getaddrinfo queries are now cached. Canonical name lookup is performed
4066 efficiently.
4067 Implemented by Ulrich Drepper.
4068
4069 * The nothrow function attribute is used when headers are used by gcc when
4070 compiling C code. This can avoid significant amounts of exception
4071 handling data.
4072
4073 * The malloc functions perform more error checking and are stricter when
4074 it comes to reacting on errors. The default action is to terminate
4075 the process after showing an error message. Implemented by Ulrich Drepper.
4076
4077 * Reverse lookups of IPv6 addresses does not use bit string or .ip6.int
4078 lookups anymore unless explicitly requested. Implemented by Ulrich Drepper.
4079
4080 * Namespaces in ld.so are implemented. DSOs can be loaded in separate
4081 namespaces using the new function dlmopen(). This feature is of course,
4082 like most other dynamic loading functionality, not available in statically
4083 linked applications. Implemented by Ulrich Drepper.
4084
4085 * Low-overhead boundary checking variants of string and some stdio functions
4086 were added. These are to be used in conjunction with a gcc patch by
4087 Jakub Jelinek which adds calls to these functions if possible.
4088 Implemented by Jakub Jelinek and Ulrich Drepper.
4089
4090 * Old code for several operating systems and machine architectures that
4091 have not been in working condition in a long time have been removed from
4092 the main source tree maintained by the GNU C Library's maintainers.
4093 These files are now reside in the separate `ports' source module
4094 that is usable as an add-on when building the library.
4095 \f
4096 Version 2.3.3
4097
4098 * New functions `dladdr1' and `dlinfo' in <dlfcn.h> provide more ways to
4099 interrogate the dynamic linker, compatible with the Solaris interface.
4100
4101 * ELF thread-local storage support (TLS) now works on PowerPC and PowerPC64;
4102 implemented by Paul Mackerras, Steven Munroe, and Roland McGrath.
4103
4104 * getifaddrs now uses the netlink interface on Linux to get its information.
4105 Implemented by Thorsten Kukuk.
4106
4107 * getaddrinfo now implements AI_V4MAPPED, AI_ALL, and AI_ADDRCONF.
4108 Implemented by Ulrich Drepper.
4109
4110 * support for non-executable stacks on x86 has been added. Changes mostly
4111 by Roland McGrath.
4112
4113 * regex is now much faster for multibyte locales. Changes by Jakub Jelinek
4114 and Ulrich Drepper.
4115
4116 * getaddrinfo now performs destination address selection according to
4117 RFC 3484.
4118 \f
4119 Version 2.3.2
4120
4121 * Thread-safe interfaces for many functions that access locale data
4122 were added in version 2.3, but these features were omitted from NEWS.
4123 Many functions have variants with an `_l' suffix that take a `locale_t'
4124 object as a parameter rather than consulting the current locale.
4125 The new functions `newlocale', `duplocale', and `freelocale' in <locale.h>
4126 create and maintain `locale_t' objects. Additionally, the new function
4127 `uselocale' sets "the current locale" (as used by functions not so
4128 parameterized) set for an individual thread. These features were added
4129 in version 2.3, implemented by Ulrich Drepper and Roland McGrath.
4130
4131 * The functions getresuid, getresgid, setresuid, and setresgid, which
4132 have long been available on Linux, are now declared in <unistd.h>
4133 and are now also available on the Hurd.
4134
4135 * ELF thread-local storage support (TLS) now works on x86-64.
4136
4137 * The new dynamic string token $LIB is expanded in shared library names.
4138 This normally expands to lib, but on some 64-bit platforms to lib64 instead.
4139
4140 * Aldy Hernandez contributed complete software floating point support for
4141 PowerPC machines with no FPU.
4142
4143 * fexecve is implemented on Linux.
4144
4145 * The `btowc' function should work at least twice as fast due to
4146 specialized callbacks in the iconv modules. Implemented by Bruno Haible.
4147
4148 * With approriate thread add-ons cancelable functions are now implemented
4149 in libc.so as well. No need to call the function in libpthread. This
4150 change allowed to finally disable the incorrect and expensive handling
4151 of weak definition in ld.so.
4152
4153 * Yet more PLT entries in libc.so have been removed. We finally arrived
4154 at the bare minimum. Startup times improved appropriately.
4155
4156 * Support for the new Linux/x86 system call interface was added. The
4157 AT_SYSINFO auxiliary vector entry is recognized and handled.
4158 \f
4159 Version 2.3
4160
4161 * Masahide Washizawa contributed iconv modules for IBM1163 and IBM1164
4162 charsets.
4163
4164 * iconv (the program and the interface) now accepts empty names (excluding
4165 options like //TRANSLIT) to mean "use charset of current locale".
4166
4167 * localedef can now transliterate characters in strings which are not in
4168 the provided charmap. The information from the input locale is used.
4169
4170 * Prelinking support was added for ELF targets. This requires additional
4171 tools and recent versions of the GNU binutils. Contributed by Jakub Jelinek.
4172
4173 * Read-only stdio streams now use mmap to speed up operation by eliminating
4174 copying and buffer underflows. To use add 'm' to the mode string of
4175 the fopen/fdopen/freopen call. Implemented by Ulrich Drepper.
4176
4177 * The malloc functions were completely rewritten by Wolfram Gloger based
4178 on Doug Lea's malloc-2.7.0.c.
4179
4180 * Isamu Hasegawa contributed a completely new and POSIX-conformant
4181 implementation of regex.
4182
4183 * Bruno Haible upgraded the iconv and locale implementation to support
4184 Unicode 3.2.
4185
4186 * Contents of the LC_* and LANG environment variables in the CEN style are
4187 not recognized anymore. It never was used. Change by Ulrich Drepper.
4188
4189 * The runtime (ld.so, libc, libpthread for Linux) now can handle the ELF
4190 thread-local storage (TLS) ABI on some platforms.
4191 Changes by Ulrich Drepper. SH support by Kaz Kojima.
4192
4193 * Bruno Haible contributed iconv converters for ISO-2022-JP-3, SHIFT JIS-X0213,
4194 EUC-JISX0213, and TSCII.
4195
4196 * New header <ifaddrs.h> with functions `getifaddrs' and `freeifaddrs':
4197 BSD-compatible interface for getting all network interface addresses.
4198 Implementation for IPv4 by Roland McGrath.
4199
4200 * Loading of locale data is faster due to the introduction of a locale
4201 archive. Implemented by Roland McGrath and Ulrich Drepper.
4202
4203 * Startup times are significantly reduced by not using exported functions
4204 inside the library itself. Changes by Jakub Jelinek, Roland McGrath,
4205 and Ulrich Drepper.
4206
4207 * Steven Munroe contributed a port to PowerPC64/Linux.
4208 \f
4209 Version 2.2.6
4210
4211 * The Hurd now uses the GNU libio implementation of stdio rather than the
4212 old GNU stdio implementation, and uses a new ABI (libc.so.0.3).
4213
4214 * The Hurd on x86 now has the `ioperm' function and <sys/io.h> header file
4215 with the same behavior as the Linux system call of the same name.
4216 \f
4217 Version 2.2.5
4218
4219 * Stephen Moshier implemented log2, log10, powl and cbrtl for the
4220 128-bit long double format.
4221
4222 * Masahide Washizawa contributed iconv modules for IBM1132, IBM1133, IBM1160,
4223 IBM1161, and IBM1162 charsets.
4224
4225 * Andreas Jaeger contributed a port to x86-64/Linux.
4226
4227 * Peter Bruin contributed a port to PowerPC/Hurd.
4228
4229 * libc functions using I/O streams now can handle wide-oriented streams
4230 as well.
4231
4232 * optimizations in the dynamic linker. Binaries created by recent binutils
4233 versions start up quicker due to reduced time spent on relocations.
4234
4235 * Support for use of gcc3 added by Jakub Jelinek and HJ Lu.
4236 \f
4237 Version 2.2.4
4238
4239 * Stephen Moshier implemented cosh, expm1, log1p, acos, sinh, tanh,
4240 asinh, atanh, j0 for the 128-bit long double format.
4241
4242 * Bruno Haible updated all the code handling Unicode in some form to
4243 support Unicode 3.1.
4244
4245 * Speed of regex for single-byte locales is back to previous levels.
4246 Patch by Isamu Hasegawa.
4247
4248 * Alpha, SPARC, and IA-64 now also using floating stacks.
4249
4250 * Startup time of internationalized applications greatly improved through
4251 iconv cache. Use iconvconfig to generate the cache file.
4252 Contributed by Ulrich Drepper.
4253
4254 * The IA-64 specific part of ld.so was rewritten to eliminate some pretty
4255 severe performance problems. Patch by David Mosberger.
4256
4257 * The Hurd port got a lot more functionality like AIO, various stdio
4258 extensions, etc. Mainly done by Roland McGrath.
4259
4260 * mtrace can now lookup symbols in shared libraries.
4261 \f
4262 Version 2.2.3
4263
4264 * Intel's IA-64 math library is largely integrated. It provides fast and
4265 accurate implementatations for most basic and standard math functions
4266 in float, double, and long double format.
4267
4268 * Stephen Moshier implemented j0, j1, jn, y0, y1, yn, lgamma, erf, erfc,
4269 and asin for the 96-bit long double format and asin, log, tan for the
4270 128-bit long double format.
4271
4272 * The beginning of a last-bit accurate math library by IBM Haifa were added.
4273 The basic double functions exist today. Contributed by Abraham Ziv
4274 <ziv@il.ibm.com>, Moshe Olshansky <olshansk@il.ibm.com>, Ealan Henis
4275 <ealan@il.ibm.com>, and Anna Reitman <reitman@il.ibm.com>.
4276
4277 * An asynchronous name lookup library was added. The interface is designed
4278 after POSIX AIO. The proposal was circulated beforehand to get comments.
4279 No negative ones came in. Implemented by Ulrich Drepper.
4280
4281 * Port to S390/64bit contributed by Martin Schwidefsky
4282 <schwidefsky@de.ibm.com>.
4283
4284 * David Mosberger <davidm@hpl.hp.com> implemented the setcontext family
4285 of functions for Linux/IA-64.
4286
4287 * The RPC code is now thread safe. Threads can now use the same service
4288 of different services at the same time. Patch by Eric Norum
4289 <eric.norum@usask.ca> with some help by Ulrich Drepper.
4290
4291 * Martin Schwidefsky <schwidefsky@de.ibm.com> implemented the setcontext
4292 family of functions for Linux/S390.
4293
4294 * Ulrich Drepper <drepper@redhat.com> implemented the setcontext family
4295 of functions for Linux/x86.
4296
4297 * Port to Linux/CRIS contributed by Axis Communications.
4298 \f
4299 Version 2.2.2
4300
4301 * Lots of headers were cleaned up. Using the tool in the conform/ subdir
4302 we can now check for namespace violations and missing declarations. The
4303 result is that almost all headers are now Unix-compliant (as defined in
4304 the upcoming XPG6). The negative side is that some programs might need
4305 corrections, too, if they depend on the incorrect form of the headers in
4306 previous versions which defined too many symbols and included too many
4307 other headers.
4308
4309 * regex now handles multibyte character sets correctly.
4310 Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
4311
4312 * iconv (the program) does now conform to the upcoming XPG6 and handles
4313 charmaps. Instead of the charset names the path of charmaps can be
4314 provided and the conversion happens based on this data.
4315 Contributed by Ulrich Drepper.
4316
4317 * The locale program now provides detailed information about the installed
4318 locales. While
4319
4320 locale -a
4321
4322 only lists the names of the supported locales
4323
4324 locale -a --verbose
4325
4326 provides details such as country, language, and codeset name.
4327 Contributed by Ulrich Drepper.
4328 \f
4329 Version 2.2.1
4330
4331 * The gencat program now parses the input file according to the charset
4332 selected by the LC_CTYPE category. This is important for stateful
4333 character sets. To make generating catalogs easier there is a way
4334 to overwrite the charset selected by the locale: before the first
4335 message or $ quote line the catalog can contain a line like
4336
4337 $ codeset=ISO-8859-2
4338
4339 to select the charset (ISO-8859-2 in this case).
4340
4341 Implemented by Shinya Hanataka and Ulrich Drepper.
4342
4343 * New codeset conversion modules: IBM-922 (Estonia PC codeset),
4344 IBM-1124 (Ukraine PC codeset), IBM-1129 (Vietnamese PC codeset).
4345 Contributed by Masahide Washizawa <washi@jp.ibm.com>.
4346
4347 * Optimized string functions for Alpha ev6 and ev67 by Richard Henderson
4348 <rth@redhat.com> and Rick Gorton <rick.gorton@alpha-processor.com>.
4349
4350 * The LANGUAGE environment variable is now ignored unless the locale is
4351 changed from the default "C" locale.
4352
4353 * The usual bug fixes.
4354 \f
4355 Version 2.2
4356
4357 * Greg McGary added runtime support for bounds checking using gcc's
4358 new -fbounded-pointers option. ix86 target is complete. PowerPC
4359 is in progress.
4360
4361 * Thorsten Kukuk added secure mode support to nscd.
4362
4363 * The Berkeley db libraries have been removed.
4364
4365 Related, the nss_db module is now in a separate package since it
4366 obviously requires a database library being available.
4367
4368 * Wide character I/O streams implemented by Ulrich Drepper.
4369
4370 * Functions from the extended socket API added by Ulrich Drepper.
4371
4372 * Functions feenableexcept and fedisableexcept to control the
4373 behaviour of individual exceptions have been added by Andreas Jaeger.
4374
4375 * ldconfig program added by Andreas Jaeger and Jakub Jelinek.
4376
4377 * The resolver code has been updated from bind 8.2.3-T5B which supports
4378 threads. The integration was done by Andreas Jaeger, Adam D. Bradley,
4379 and Mark Kettenis.
4380
4381 This change could in some situations effect backward compatibility. Since
4382 now `_res' is a thread-local instead of a global variable, modifying it
4383 in one thread does not have any effect in other threads.
4384
4385 The resolver library was also extended to allow IPv6 as the transport
4386 protocol for the requests. This work was done by Stig Venaas.
4387
4388 * Compatibility code for K&R C compilers has been removed from the
4389 header files. A ISO C compiler is needed to use the library
4390 (conforming to either C89 or C99 standard).
4391
4392 * Complete rewrite of the localedef program to support multibyte character
4393 sets. Implement handling of ISO 14651 and ISO 14652. Rewrite strcoll,
4394 strxfrm, wcscoll, and wcsxfrm functions. Make isw*() functions work.
4395 Implemented by Ulrich Drepper.
4396
4397 Bruno Haible significantly improved the generation and use of the data
4398 structures for the wide character tables.
4399
4400 * Plural handling in gettext implemented by Ulrich Drepper.
4401
4402 * The utmp daemon has been removed.
4403
4404 * The port to MIPS-Linux has been finished by Andreas Jaeger.
4405
4406 * A port to Hitachi SH3 and SH4 has been contributed by Kazumoto Kojima
4407 and Yutaka Niibe.
4408
4409 * POSIX clocks and timers implemented by Kaz Kylheku and Ulrich Drepper.
4410
4411 * POSIX barriers implemented by Kaz Kylheku.
4412
4413 * POSIX spawn function family implemented by Ulrich Drepper.
4414
4415 * POSIX spinlocks are now available.
4416
4417 * Timed wait functions for mutex, rwlock, and semaphores are implemented.
4418
4419 * the configure option --enable-kernel=X.Y.Z allows to strip out
4420 compatibility for kernel versions before X.Y.Z. This is currently only
4421 implemented for Linux.
4422
4423 * the sockaddr_in6 structure changed. The IPv6 working group added a new
4424 field sin6_scope_id. This means that all programs using IPv6 should be
4425 recompiled. Don't expect binary compatibility with previous glibc
4426 versions.
4427
4428 * various conversion modules for IBM character sets contributed by
4429 Masahide Washizawa.
4430
4431 * IA-64 port by Jes Sorensen and HJ Lu.
4432 \f
4433 Version 2.1.3
4434
4435 * bug fixes
4436
4437 \f
4438 Version 2.1.2
4439
4440 * bug fixes
4441
4442 \f
4443 Version 2.1.1
4444
4445 * New ISO C 9x function _Exit, imaxabs, and imaxdiv are added.
4446
4447 * New xdr functions are added; some rpc functions are now 64bit clean.
4448
4449 * Fixed a number of bugs and memory leaks (especially in NIS+ code).
4450
4451 * Fixed known incompatibilities with glibc 2.0.
4452
4453 * New functions lock64, strchrnul, rawmemchr, getutmp and getutmpx.
4454
4455 * Optimized a number of functions (especially the ELF dynamic loader).
4456
4457 * Update timezone data files.
4458
4459 * lots of charmaps corrections
4460
4461 * some new locale definitions and charmaps
4462
4463 \f
4464 Version 2.1
4465
4466 * Richard Henderson corrected size of struct timeval on Linux/Alpha to
4467 conform to POSIX member type requirements. Symbol versions have been
4468 adjusted as needed within the library, and for direct use by applications,
4469 but there is potential for problems if third-party libraries use
4470 struct timeval as part of their interface. This does not present
4471 a problem for X and other "essential" system libraries.
4472
4473 * An additional locale model to support C++ Standard Library locale
4474 model and probably more was implemented by Ulrich Drepper.
4475
4476 * Eric Youngdale and Ulrich Drepper implemented versioning of objects on
4477 symbol level.
4478
4479 * Miles Bader provided the `argp' function family to support hierachical
4480 command line argument parsing, layered on top of getopt.
4481
4482 * strtod accepts new hexadecimal floating-point format from ISO C 9X.
4483
4484 * printf has two new specifiers %a and %A to print hexadecimal flaoting-point
4485 numbers.
4486
4487 * scanf recognizes the %a and %A format for scanning floating point numbers.
4488
4489 * the new headers <stdint.h> and <inttypes.h> from ISO C 9X provides
4490 information and interfaces for the available integer types.
4491
4492 * about 130 new math functions were added to implement the ISO C9x math
4493 library.
4494
4495 * the new header <complex.h> contains definitions of the complex math
4496 functions from ISO C 9X.
4497
4498 * the new header <tgmath.h> defines generic macros to use complex or
4499 real valued functions.
4500
4501 * Thorsten Kukuk provided an implementation for NIS+, securelevel 0, 1 and 2.
4502
4503 * Andreas Jaeger provided a test suite for the math library.
4504
4505 * Mark Kettenis implemented the utmpx interface and an utmp daemon.
4506
4507 * Ulrich Drepper added character set conversion functions (iconv).
4508
4509 * Optimized string functions have been added.
4510
4511 * The localedata addon is now part of glibc.
4512
4513 * An implementation of profiling shared libraries was added by Ulrich Drepper.
4514
4515 * Thorsten Kukuk and Ulrich Drepper provided an implementation for a caching
4516 daemon for NSS (nscd).
4517
4518 Missing a better place here are some numbers on improvements. Under
4519 Linux 2.1.125 un-tar-ing the kernel sources takes
4520
4521 user system wall
4522
4523 using local files 12.19s 6.88s 22.91s
4524
4525 using NIS 13.92s 8.91s 26.34s
4526
4527 using NIS & nscd 10.37s 7.34s 25.30s
4528
4529 using NIS+ 27.57s 30.37s 640.46s
4530
4531 using NIS+ & nscd 10.25s 7.83s 26.51s
4532
4533 using NIS & old nscd [1] 13.83s 8.32s 29.60s
4534
4535 Keep in mind that non-namelookup related operations dominate above times.
4536 It was just a common complain that using NIS+ unpacking the kernel is
4537 horribly slow.
4538
4539 [1] The old nscd implementation is not available anymore since it was
4540 distributed with glibc up to version 2.0.98 and thus is now replaced.
4541
4542 * Tim Waugh provided an implementation of the POSIX.2 wordexp function family.
4543
4544 * Mark Kettenis provided a Hesiod NSS module.
4545
4546 * The ELF dynamic loader knows how to process dynamic string tokens ($ORIGIN
4547 and $PLATFORM) in RPATHs and similar strings (Ulrich Drepper).
4548
4549 * rcmd can now handle netgroups (Dick Streefland).
4550
4551 * A port to the ARM was contributed by Phil Blundell, Pat Beirne and Scott
4552 Bambrough.
4553
4554 * Support for the IPv6 protocol has been added to the socket API, as per the
4555 latest draft standards.
4556
4557 * Support for Linux 2.2 has been added.
4558
4559 * Interface changes relative to the latest 2.0.x release:
4560 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4561 addseverity NEW: Unix98
4562 alphasort64 NEW: LFS
4563 argp_err_exit_status NEW: argp, GNU ext
4564 argp_error NEW: argp, GNU ext
4565 argp_failure NEW: argp, GNU ext
4566 argp_help NEW: argp, GNU ext
4567 argp_parse NEW: argp, GNU ext
4568 argp_program_bug_address NEW: argp, GNU ext
4569 argp_program_version NEW: argp, GNU ext
4570 argp_program_version_hook NEW: argp, GNU ext
4571 argp_state_help NEW: argp, GNU ext
4572 argp_usage NEW: argp, GNU ext
4573 authdes_create NEW: Secure RPC
4574 authdes_getucred NEW: Secure RPC
4575 authdes_pk_create NEW: Secure RPC
4576 backtrace NEW: GNU ext.
4577 backtrace_symbols NEW: GNU ext.
4578 backtrace_symbols_fd NEW: GNU ext.
4579 cacos NEW: ISO C 9x
4580 cacosf NEW: ISO C 9x
4581 cacosh NEW: ISO C 9x
4582 cacoshf NEW: ISO C 9x
4583 cacoshl NEW: ISO C 9x
4584 cacosl NEW: ISO C 9x
4585 capget NEW: kernel
4586 capset NEW: kernel
4587 carg NEW: ISO C 9x
4588 cargf NEW: ISO C 9x
4589 cargl NEW: ISO C 9x
4590 casin NEW: ISO C 9x
4591 casinf NEW: ISO C 9x
4592 casinh NEW: ISO C 9x
4593 casinhf NEW: ISO C 9x
4594 casinhl NEW: ISO C 9x
4595 casinl NEW: ISO C 9x
4596 catan NEW: ISO C 9x
4597 catanf NEW: ISO C 9x
4598 catanh NEW: ISO C 9x
4599 catanhf NEW: ISO C 9x
4600 catanhl NEW: ISO C 9x
4601 catanl NEW: ISO C 9x
4602 cbc_crypt NEW: Secure RPC
4603 ccos NEW: ISO C 9x
4604 ccosf NEW: ISO C 9x
4605 ccosh NEW: ISO C 9x
4606 ccoshf NEW: ISO C 9x
4607 ccoshl NEW: ISO C 9x
4608 ccosl NEW: ISO C 9x
4609 cexp NEW: ISO C 9x
4610 cexpf NEW: ISO C 9x
4611 cexpl NEW: ISO C 9x
4612 cimag NEW: ISO C 9x
4613 cimagf NEW: ISO C 9x
4614 cimagl NEW: ISO C 9x
4615 clearerr_locked REMOVED
4616 clntunix_create NEW: sunrpc ext
4617 clog NEW: ISO C 9x
4618 clog10 NEW: ISO C 9x
4619 clog10f NEW: ISO C 9x
4620 clog10l NEW: ISO C 9x
4621 clogf NEW: ISO C 9x
4622 clogl NEW: ISO C 9x
4623 conj NEW: ISO C 9x
4624 conjf NEW: ISO C 9x
4625 conjl NEW: ISO C 9x
4626 cpow NEW: ISO C 9x
4627 cpowf NEW: ISO C 9x
4628 cpowl NEW: ISO C 9x
4629 cproj NEW: ISO C 9x
4630 cprojf NEW: ISO C 9x
4631 cprojl NEW: ISO C 9x
4632 creal NEW: ISO C 9x
4633 crealf NEW: ISO C 9x
4634 creall NEW: ISO C 9x
4635 creat64 NEW: LFS
4636 csin NEW: ISO C 9x
4637 csinf NEW: ISO C 9x
4638 csinh NEW: ISO C 9x
4639 csinhf NEW: ISO C 9x
4640 csinhl NEW: ISO C 9x
4641 csinl NEW: ISO C 9x
4642 csqrt NEW: ISO C 9x
4643 csqrtf NEW: ISO C 9x
4644 csqrtl NEW: ISO C 9x
4645 ctan NEW: ISO C 9x
4646 ctanf NEW: ISO C 9x
4647 ctanh NEW: ISO C 9x
4648 ctanhf NEW: ISO C 9x
4649 ctanhl NEW: ISO C 9x
4650 ctanl NEW: ISO C 9x
4651 des_setparity NEW: Secure RPC
4652 ecb_crypt NEW: Secure RPC
4653 endutxent NEW: Unix98
4654 exp10 NEW: ISO C 9x
4655 exp10f NEW: ISO C 9x
4656 exp10l NEW: ISO C 9x
4657 exp2 NEW: ISO C 9x
4658 exp2f NEW: ISO C 9x
4659 exp2l NEW: ISO C 9x
4660 fattach NEW: STREAMS
4661 fdetach NEW: STREAMS
4662 fdim NEW: ISO C 9x
4663 fdimf NEW: ISO C 9x
4664 fdiml NEW: ISO C 9x
4665 feclearexcept NEW: ISO C 9x
4666 fegetenv NEW: ISO C 9x
4667 fegetexceptflag NEW: ISO C 9x
4668 fegetround NEW: ISO C 9x
4669 feholdexcept NEW: ISO C 9x
4670 feof_locked REMOVED
4671 feraiseexcept NEW: ISO C 9x
4672 ferror_locked REMOVED
4673 fesetenv NEW: ISO C 9x
4674 fesetexceptflag NEW: ISO C 9x
4675 fesetround NEW: ISO C 9x
4676 fetestexcept NEW: ISO C 9x
4677 feupdateenv NEW: ISO C 9x
4678 fflush_locked REMOVED
4679 ffsl NEW: GNU ext.
4680 ffsll NEW: GNU ext.
4681 fgetpos64 NEW: LFS
4682 fgets_unlocked NEW: GNU ext.
4683 fileno_locked REMOVED
4684 fma NEW: ISO C 9x
4685 fmaf NEW: ISO C 9x
4686 fmal NEW: ISO C 9x
4687 fmax NEW: ISO C 9x
4688 fmaxf NEW: ISO C 9x
4689 fmaxl NEW: ISO C 9x
4690 fmin NEW: ISO C 9x
4691 fminf NEW: ISO C 9x
4692 fminl NEW: ISO C 9x
4693 fmtmsg NEW: Unix98
4694 fopen64 NEW: LFS
4695 fputc_locked REMOVED
4696 fputs_unlocked NEW: GNU ext.
4697 fread_unlocked NEW: GNU ext.
4698 freopen64 NEW: LFS
4699 fseeko NEW: Unix98
4700 fsetpos64 NEW: LFS
4701 fstatfs64 NEW: LFS
4702 fstatvfs NEW: Unix98
4703 fstatvfs64 NEW: LFS
4704 ftello NEW: Unix98
4705 ftello64 NEW: LFS
4706 ftruncate64 NEW: LFS
4707 ftw64 NEW: LFS
4708 fwrite_unlocked NEW: GNU ext.
4709 gai_strerror NEW: IPv6
4710 gamma_r REMOVED
4711 gammaf_r REMOVED
4712 gammal_r REMOVED
4713 getchar_locked REMOVED
4714 getdate NEW: Unix98
4715 getdate_err NEW: Unix98
4716 getdate_r NEW: GNU ext.
4717 getmsg NEW: STREAMS
4718 getnameinfo NEW: IPv6
4719 getnetname NEW: Secure RPC
4720 getpmsg NEW: STREAMS
4721 getpt NEW: Unix98 PTY
4722 getrlimit64 NEW: LFS
4723 getutxent NEW: Unix98
4724 getutxid NEW: Unix98
4725 getutxline NEW: Unix98
4726 glob64 NEW: GNU ext.
4727 globfree64 NEW: GNU ext.
4728 gnu_get_libc_release NEW: GNU ext.
4729 gnu_get_libc_version NEW: GNU ext.
4730 grantpt NEW: Unix98 PTY
4731 host2netname NEW: Secure RPC
4732 iconv NEW: iconv
4733 iconv_close NEW: iconv
4734 iconv_open NEW: iconv
4735 if_freenameindex NEW: IPv6
4736 if_indextoname NEW: IPv6
4737 if_nameindex NEW: IPv6
4738 if_nametoindex NEW: IPv6
4739 in6addr_any NEW: IPv6
4740 in6addr_loopback NEW: IPv6
4741 inet6_isipv4mapped NEW: IPv6
4742 isastream NEW: STREAMS
4743 iswblank NEW: GNU ext.
4744 key_decryptsession NEW: Secure RPC
4745 key_decryptsession_pk NEW: Secure RPC
4746 key_encryptsession NEW: Secure RPC
4747 key_encryptsession_pk NEW: Secure RPC
4748 key_gendes NEW: Secure RPC
4749 key_get_conv NEW: Secure RPC
4750 key_secretkey_is_set NEW: Secure RPC
4751 key_setnet NEW: Secure RPC
4752 key_setsecret NEW: Secure RPC
4753 llrint NEW: ISO C 9x
4754 llrintf NEW: ISO C 9x
4755 llrintl NEW: ISO C 9x
4756 llround NEW: ISO C 9x
4757 llroundf NEW: ISO C 9x
4758 llroundl NEW: ISO C 9x
4759 log2 NEW: ISO C 9x
4760 log2f NEW: ISO C 9x
4761 log2l NEW: ISO C 9x
4762 lrint NEW: ISO C 9x
4763 lrintf NEW: ISO C 9x
4764 lrintl NEW: ISO C 9x
4765 lround NEW: ISO C 9x
4766 lroundf NEW: ISO C 9x
4767 lroundl NEW: ISO C 9x
4768 lseek64 NEW: LFS
4769 makecontext NEW: Unix98
4770 mempcpy NEW: GNU ext.
4771 mmap64 NEW: LFS
4772 moncontrol REMOVED
4773 modify_ldt NEW: kernel
4774 nan NEW: ISO C 9x
4775 nanf NEW: ISO C 9x
4776 nanl NEW: ISO C 9x
4777 nearbyint NEW: ISO C 9x
4778 nearbyintf NEW: ISO C 9x
4779 nearbyintl NEW: ISO C 9x
4780 netname2host NEW: Secure RPC
4781 netname2user NEW: Secure RPC
4782 nexttoward NEW: ISO C 9x
4783 nexttowardf NEW: ISO C 9x
4784 nexttowardl NEW: ISO C 9x
4785 nftw NEW: Unix98
4786 nftw64 NEW: LFS
4787 open64 NEW: LFS
4788 passwd2des NEW: Secure RPC
4789 pow10 NEW: GNU ext.
4790 pow10f NEW: GNU ext.
4791 pow10l NEW: GNU ext.
4792 pread NEW: Unix98
4793 pread64 NEW: LFS
4794 printf_size NEW: GNU ext.
4795 printf_size_info NEW: GNU ext.
4796 profil_counter REMOVED
4797 pthread_mutexattr_getkind_np REPLACED
4798 pthread_mutexattr_setkind_np REPLACED
4799 ptsname NEW: Unix98 PTY
4800 ptsname_r NEW: Unix98 PTY
4801 putc_locked REMOVED
4802 putchar_locked REMOVED
4803 putgrent NEW: GNU ext.
4804 putmsg NEW: STREAMS
4805 putpmsg NEW: STREAMS
4806 pututxline NEW: Unix98
4807 pwrite NEW: Unix98
4808 pwrite64 NEW: LFS
4809 readdir64 NEW: LFS
4810 readdir64_r NEW: LFS
4811 remquo NEW: ISO C 9x
4812 remquof NEW: ISO C 9x
4813 remquol NEW: ISO C 9x
4814 round NEW: ISO C 9x
4815 roundf NEW: ISO C 9x
4816 roundl NEW: ISO C 9x
4817 rtime NEW: GNU ext.
4818 scalbln NEW: ISO C 9x
4819 scalblnf NEW: ISO C 9x
4820 scalblnl NEW: ISO C 9x
4821 scandir64 NEW: LFS
4822 sendfile NEW: kernel
4823 setcontext NEW: Unix98
4824 setrlimit64 NEW: LFS
4825 setutxent NEW: Unix98
4826 sighold NEW: Unix98
4827 sigignore NEW: Unix98
4828 sigqueue NEW: POSIX.1b
4829 sigrelse NEW: Unix98
4830 sigset NEW: POSIX.1b
4831 sigtimedwait NEW: POSIX.1b
4832 sigwaitinfo NEW: POSIX.1b
4833 sincos NEW: GNU ext.
4834 sincosf NEW: GNU ext.
4835 sincosl NEW: GNU ext.
4836 statfs64 NEW: LFS
4837 statvfs NEW: Unix98
4838 statvfs64 NEW: LFS
4839 strcasestr NEW: GNU ext.
4840 strtoimax NEW: ISO C 9x
4841 strtoumax NEW: ISO C 9x
4842 strverscmp NEW: GNU ext.
4843 svcauthdes_stats NEW: Secure RPC
4844 svcunix_create NEW: sunrpc etx
4845 svcunixfd_create NEW: sunrpc ext
4846 swapcontext NEW: Unix98
4847 tcgetsid NEW: Unix98 PTY
4848 tdestroy NEW: GNU ext.
4849 tgamma NEW: ISO C 9x
4850 tgammaf NEW: ISO C 9x
4851 tgammal NEW: ISO C 9x
4852 tmpfile64 NEW: LFS
4853 trunc NEW: ISO C 9x
4854 truncate64 NEW: LFS
4855 truncf NEW: ISO C 9x
4856 truncl NEW: ISO C 9x
4857 umount2 NEW: kernel
4858 unlockpt NEW: Unix98 PTY
4859 updwtmpx NEW: Unix98
4860 user2netname NEW: Secure RPC
4861 utmpxname NEW: Unix98
4862 versionsort NEW: GNU ext.
4863 versionsort64 NEW: GNU ext.
4864 waitid NEW: Unix98
4865 wcscasecmp NEW: GNU ext.
4866 wcsncasecmp NEW: GNU ext.
4867 wcsnlen NEW: GNU ext.
4868 wcstoimax NEW: ISO C 9x
4869 wcstoll NEW: ISO C 9x
4870 wcstoull NEW: ISO C 9x
4871 wcstoumax NEW: ISO C 9x
4872 wcswcs NEW: Unix98
4873 wordexp NEW: POSIX.2
4874 wordfree NEW: POSIX.2
4875 write_profiling REMOVED
4876 xdecrypt NEW: Secure RPC
4877 xdr_authdes_cred NEW: Secure RPC
4878 xdr_authdes_verf NEW: Secure RPC
4879 xdr_cryptkeyarg NEW: Secure RPC
4880 xdr_cryptkeyarg2 NEW: Secure RPC
4881 xdr_cryptkeyres NEW: Secure RPC
4882 xdr_getcredres NEW: Secure RPC
4883 xdr_key_netstarg NEW: Secure RPC
4884 xdr_key_netstres NEW: Secure RPC
4885 xdr_keybuf NEW: Secure RPC
4886 xdr_keystatus NEW: Secure RPC
4887 xdr_netnamestr NEW: Secure RPC
4888 xdr_sizeof NEW: Secure RPC
4889 xdr_unixcred NEW: sunrpc ext
4890 xencrypt NEW: Secure RPC
4891 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4892 \f
4893 Version 2.0.6
4894
4895 * more bug fixes
4896
4897 \f
4898 Version 2.0.5
4899
4900 * more bug fixes
4901
4902 * inet_ntoa is thread-safe
4903
4904 * updwtmp is moved from libutil to libc
4905
4906 * rewrite of cbrt function
4907
4908 * update of timezone data
4909 \f
4910 Version 2.0.4
4911
4912 * more bug fixes
4913 \f
4914 Version 2.0.3
4915
4916 * more bug fixes
4917 \f
4918 Version 2.0.2
4919
4920 * more bug fixes
4921
4922 * add atoll function
4923
4924 * fix complex problems in Berkeley DB code
4925
4926 * fix math functions
4927 \f
4928 Version 2.0.1
4929
4930 * fixed lots of header problems (especially Linux/GNU specific)
4931
4932 * dynamic loader preserves all registers
4933
4934 * Roland McGrath provided support for handling of auxiliary objects in
4935 the ELF dynamic loader.
4936
4937 * support for parallel builds is improved
4938 \f
4939 Version 2.0
4940
4941 * GNU extensions are no longer declared by default. To enable them you
4942 must define the macro `_GNU_SOURCE' in your program or compile with
4943 `-D_GNU_SOURCE'.
4944
4945 * The library has changed from using GNU ld symbol aliases to using weak
4946 symbols where available. The ELF object file format supports weak
4947 symbols; GNU ld also supports weak symbols in the a.out format. (There
4948 is also now support for other GNU ld extensions in ELF. Use the
4949 `--with-elf' option to configure to indicate you have ELF, and
4950 `--with-gnu-ld' if using GNU ld.) This change resulted in the deletion
4951 of many files which contained only symbol aliases, reducing the size of
4952 the source and the compiled library; many other files were renamed to
4953 less cryptic names previously occupied by the symbol alias files.
4954 There is a new header file <elf.h> for programs which operate on
4955 files in the ELF format.
4956
4957 * Converted to Autoconf version 2, so `configure' has more options.
4958 Run `configure --help' to see the details.
4959
4960 * The library can now be configured to build profiling, highly-optimized
4961 (but undebuggable), and/or shared libraries (ELF with GNU ld only). The
4962 `--enable-profile', `--enable-omitfp', and `--enable-shared' options to
4963 `configure' enable building these extra libraries. The shared library is
4964 built by default when using both ELF and GNU ld. When shared libraries
4965 are enabled, the new library `-ldl' is available for arbitrary run-time
4966 loading of shared objects; its interface is defined in <dlfcn.h>. The
4967 new header file <link.h> gives access to the internals of the run-time
4968 dynamic linker, `ld.so'. The shell script `ldd' is similar to the
4969 application of same name on other systems and it provides information
4970 about dynamically linked binaries.
4971
4972 * The C library now provides the run-time support code for profiling
4973 executables compiled with `-pg'. Programs can control the profiling code
4974 through the interface in <sys/gmon.h>. The `gmon.out' files written by
4975 the GNU C library can be read only by GNU `gprof' (from GNU binutils);
4976 the support for this file format was contributed by David Mosberger-Tang.
4977
4978 * The math code has been replaced with a math library based on fdlibm from
4979 Sun, and modified by JT Conklin and Ulrich Drepper with i387 support, by
4980 Ian Taylor with `float' functions and by Ulrich Drepper with `long double'
4981 functions. The math functions now reside in a separate library, so
4982 programs using them will need to use `-lm' their linking commands.
4983
4984 * John C. Bowman contributed optimized ix87 assembler inline functions.
4985
4986 * Ulrich Drepper has contributed support for an `/etc/nsswitch.conf'
4987 mechanism similar to that found in Solaris 2. This is now used for the
4988 group, passwd, hosts, networks, services, protocols, rpc, ethers,
4989 shadow, netgroup, publickey, and alias databases. The `nsswitch.conf'
4990 file controls what services are used for each individual database. This
4991 works by loading shared libraries with names specified in `nsswitch.conf',
4992 so service modules can be changed or added at any time without even
4993 relinking any program. Currently there are the file, db, and NIS based
4994 NSS services available.
4995
4996 * The new functions `strtoq' and `strtouq' parse integer values from
4997 strings, like `strtol' and `strtoul', but they return `long long int' and
4998 `unsigned long long int' values, respectively (64-bit quantities).
4999
5000 * The new functions `strtof' and `strtold' parse floating-point values from
5001 strings, like `strtod', but they return `float' and `long double' values,
5002 respectively (on some machines `double' and `long double' are the same).
5003
5004 * Ulrich Drepper has contributed new implementations of the floating-point
5005 printing and reading code used in the `printf' family of functions and
5006 `strtod', `strtof', and `strtold'. These new functions are perfectly
5007 accurate, and much faster than the old ones.
5008
5009 * The implementation of the POSIX locale model was completely rewritten by
5010 Ulrich Drepper. This includes the new programs `localedef' and `locale'
5011 to compile the POSIX locale definition.
5012
5013 * The former dummy implementations of the strcoll and strxfrm function are
5014 now replaced by fully functional code contributed by Ulrich Drepper. The
5015 collation information comes from the POSIX locale definitions.
5016
5017 * The new header <langinfo.h> defines an interface for accessing
5018 various locale-dependent data (using the locale chosen with `setlocale').
5019
5020 * Ulrich Drepper has contributed a new suite of functions for operation on
5021 wide-character and multibyte-character strings, in <wchar.h>;
5022 and classification and case conversion of wide characters, in <wctype.h>.
5023 These new functions are conforming to the ISO C, Amendement 1 specification.
5024
5025 * There is now a second implementation of the standard I/O library available.
5026 It comes from GNU libg++ as was written by Per Bothner, heavily modified
5027 by Hongjiu Lu and made thread safe by Ulrich Drepper.
5028
5029 * You can now use positional parameter specifications in format strings
5030 for the `printf' and `scanf' families of functions. For example,
5031 `printf ("Number %2$d, Mr %1$s\n", "Jones", 6);'' prints
5032 ``Number 6, Mr Jones''. This is mainly useful when providing different
5033 format strings for different languages, whose grammars may dictate
5034 different orderings of the values being printed. To support this
5035 feature, the interface for `register_printf_handler' has changed; see
5036 the header file <printf.h> for details.
5037
5038 * The `printf' and `scanf' families of functions now understand a new
5039 formatting flag for numeric conversions: the ' flag (e.g. %'d or %'f) says
5040 to group numbers as indicated by the locale; for `scanf' and friends, this
5041 says to accept as valid only a number with all the proper grouping
5042 separators in the right places. In the default "C" locale, numbers are
5043 not grouped; but locales for specific countries will define the usual
5044 conventions (i.e. separate thousands with `,' in the US locale).
5045
5046 * The pgrp functions have been regularized, slightly incompatibly but much
5047 less confusingly. The core functions are now `getpgid' and `setpgid',
5048 which take arguments for the PID to operate on; the POSIX.1 `getpgrp' (no
5049 argument) and BSD `setpgrp' (identical to `setpgid') functions are
5050 provided for compatibility. There is no longer an incompatible `getpgrp'
5051 with an argument declared under _BSD_SOURCE; no BSD code uses it.
5052
5053 * The new header file <fts.h> and suite of functions simplify programs that
5054 operate on directory trees. This code comes from 4.4 BSD.
5055
5056 * The resolver code has been updated from the BIND 4.9.5-P1 release.
5057 Parts of the code were heavily modified by Ulrich Drepper to fit in the
5058 NSS scheme used in glibc.
5059
5060 * There is a new malloc debugging hook `__memalign_hook'.
5061
5062 * There are new typedefs `ushort' for `unsigned short int' and `uint' for
5063 `unsigned int' in <sys/types.h>. These are for compatibility only and
5064 their use is discouraged.
5065
5066 * The `-lmcheck' library to enable standard malloc debugging hooks is now
5067 done differently, so that it works even without GNU ld.
5068
5069 * New function `euidaccess' checks allowed access to a file like `access',
5070 but using the effective IDs instead of the real IDs.
5071
5072 * The time zone data files have been updated for the latest and greatest
5073 local time conventions of the countries of the world.
5074
5075 * The new function `dirfd' extracts the file descriptor used by a DIR stream;
5076 see <dirent.h>.
5077
5078 * The new functions `ecvt', `fcvt', and `gcvt' provide an obsolete interface
5079 for formatting floating-point numbers. They are provided only for
5080 compatibility; new programs should use `sprintf' instead. There are
5081 also equivalent function for the `long double' floating-point type and
5082 all functions also exist in a reentrant form.
5083
5084 * The new auxiliary library `-lutil' from 4.4 BSD contains various
5085 functions for maintaining the login-record files (primarily of use to
5086 system programs such as `login'), and convenient functions for
5087 allocating and initializing a pseudo-terminal (pty) device.
5088
5089 * Ulrich Drepper has contributed new support for System V style
5090 shared memory and IPC on systems that support it.
5091
5092 * Ulrich Drepper has contributed several miscellaneous new functions found
5093 in System V: The `hsearch' family of functions provide an effective
5094 implementation of hash tables; `a64l' and `l64a' provide a very simple
5095 binary to ASCII mapping; `drand48' and friends provide a 48-bit random
5096 number generator.
5097
5098 * Ulrich Drepper has contributed new reentrant counterparts for the
5099 `random' and `hsearch' families of functions; `random_r', `hsearch_r', etc.
5100
5101 * Ulrich Drepper has contributed new, highly-optimized versions of several
5102 string functions for the i486/Pentium family of processors.
5103
5104 * Ulrich Drepper has updated the Linux-specific code, based largely
5105 on work done in Hongjiu Lu's version of GNU libc for Linux.
5106 The GNU library now supports Linux versions 2.0.10 and later,
5107 using the ELF object file format (i[3456]86-*-linux).
5108
5109 * Andreas Schwab has ported the C library to Linux/m68k (m68k-*-linux).
5110
5111 * David Mosberger-Tang and Richard Henderson have ported the C library
5112 to Linux/Alpha (alpha-*-linux). Richard Henderson contributed the
5113 dynamic linking support for ELF/Alpha.
5114
5115 * Richard Henderson contributed several Alpha optimized assembler function
5116 for arithmetic and string handling.
5117
5118 * Ulrich Drepper has contributed a new set of message catalog functions to
5119 support multiple languages using the <libintl.h> interface, for use with
5120 his new package GNU gettext. Translation volunteers have contributed
5121 catalogs of the library's messages in Spanish, German, and Korean.
5122
5123 * For compatibility with XPG4, Ulrich Drepper has contributed the `gencat'
5124 program and the `catgets' function for reading the catalog files it
5125 creates. (The <libintl.h> interface is preferred; we include the
5126 <nl_types.h> interface using `catgets' only for source compatibility with
5127 programs already written to use it.)
5128
5129 * New header file <values.h> gives SVID-compatible names for <limits.h>
5130 constants.
5131
5132 * Various new macros, declarations, and small header files for compatibility
5133 with 4.4 BSD.
5134
5135 * New function `group_member' is a convenient way to check if a process has
5136 a given effective group ID.
5137
5138 * When using GCC 2.7 and later, the socket functions are now declared in a
5139 special way so that passing an argument of type `struct sockaddr_in *',
5140 `struct sockaddr_ns *', or `struct sockaddr_un *' instead of the generic
5141 `struct sockaddr *' type, does not generate a type-clash warning.
5142
5143 * New function `error' declared in header file <error.h> is a convenient
5144 function for printing error messages and optionally exiting; this is the
5145 canonical function used in GNU programs. The new functions `err', `warn',
5146 and friends in header file <err.h> are the canonical 4.4 BSD interface for
5147 doing the same thing.
5148
5149 * The <glob.h> interface has several new flags from 4.4 BSD that extend the
5150 POSIX.2 `glob' function to do ~ and {...} expansion.
5151
5152 * New function `unsetenv' complements `setenv' for compatibility with 4.4 BSD.
5153 `clearenv' which is used in POSIX.9 is also available.
5154
5155 * New function `getsid' returns session ID number on systems that support it.
5156
5157 * We have incorporated the 4.4 BSD `db' library (version 1.85). New header
5158 files <db.h> and <mpool.h> provide a rich set of functions for several
5159 types of simple databases stored in memory and in files, and <ndbm.h> is
5160 an old `ndbm'-compatible interface using the `db' functions. Link with
5161 `-ldb' to get these functions.
5162
5163 * New macro `strdupa' copies a string like `strdup', but uses local stack
5164 space from `alloca' instead of dynamic heap space from `malloc'.
5165
5166 * New function `strnlen' is like `strlen' but searches only a given maximum
5167 number of characters for the null terminator. `stpncpy', `strndup' and
5168 `strndupa' are similar variants for the `stpcpy', `strdup' and `strdupa'
5169 function.
5170
5171 * New function `statfs' in header <sys/statfs.h>.
5172
5173 * The new <argz.h> and <envz.h> interfaces contributed by Miles Bader
5174 provide convenient functions for operating on blocks of null-terminated
5175 strings.
5176
5177 * A new suite of functions in <utmp.h> handle all the details of reading
5178 and writing the utmp file.
5179
5180 * An implementation of the NIS/YP(tm) based NSS service was contributed by
5181 Thorsten Kukuk.
5182
5183 * Paul Eggert and Ulrich Drepper modified the `strftime' function to be
5184 completely POSIX compliant and also implemented the extended functionality
5185 to handle alternate digit representation and alternate era date formats.
5186
5187 * Ulrich Drepper provided an implementation of the `strptime' function
5188 defined in XPG4.2 which transforms a string into a `struct tm' value.
5189
5190 * Paul Eggert provided the tzselect shell script as part of the timezone
5191 code. The shell script makes it easy to select the correct timezone
5192 specification.
5193
5194 * The implementation of the malloc family of functions is completely replaced
5195 by a new implementation by Doug Lea with many improvements by Wolfram Gloger.
5196 The implementation uses the mmap function (if available) and it is
5197 optimized for the use in multi threaded programs.
5198
5199 * Ulrich Drepper contributed a MD5 "encryption" for the crypt family of
5200 functions. This new functionality is usable by specifying a special
5201 salt string and it is compatible with implementation on *BSD systems.
5202
5203 * Lots of functions from the XPG4.2 standard were added by Ulrich Drepper:
5204 `getsubopt' to handle second level command line options, `bsd_signal'
5205 to access BSD style `signal' functionality, the obsolete `regexp' style
5206 expression matcher.
5207
5208 * the `lchown' function is available on system which support this
5209 functionality.
5210
5211 * The implementation of the shadow password handling function was contributed
5212 by Ulrich Drepper.
5213
5214 * David Mosberger-Tang changed the SunRPC implementation to be 64bit safe.
5215
5216 * POSIX.1g support was added. The <sys/select.h> header is available,
5217 `isfdtype' and `pselect' are implemented. Craig Metz contributed an
5218 implementation of `getaddrinfo'.
5219 \f
5220 Version 1.09
5221
5222 * For cross-compilation you should now set `BUILD_CC' instead of `HOST_CC'.
5223
5224 * New header file <fstab.h> and new functions `getfsspec', `getfsent' and
5225 friends, for parsing /etc/fstab. This code comes from 4.4 BSD.
5226
5227 * The new function `daemon' from 4.4 BSD is useful for server programs that
5228 want to put themselves in the background.
5229
5230 * Joel Sherrill has contributed support for several standalone boards that
5231 run without an operating system.
5232
5233 * `printf', `scanf' and friends now accept a `q' type modifier for long
5234 long int as well as `ll'. Formats using these might be `%qu' or `%lld'.
5235
5236 * All of the code taken from BSD (notably most of the math and networking
5237 routines) has been updated from the BSD 4.4-Lite release.
5238
5239 * The resolver code has been updated from the BIND-4.9.3-BETA9 release.
5240
5241 * The new functions `getdomainname' and `setdomainname' fetch or change the
5242 YP/NIS domain name. These are system calls which exist on systems which
5243 have YP (aka NIS).
5244
5245 * The time zone data files have been updated for the latest international
5246 conventions.
5247
5248 * The SunRPC programs `portmap' and `rpcinfo' are now installed in
5249 $(sbindir) (usually /usr/local/sbin) instead of $(bindir).
5250 \f
5251 Version 1.08
5252
5253 * The C library now includes support for Sun RPC, from Sun's free
5254 RPCSRC-4.0 distribution. The `portmap', `rpcinfo', and `rpcgen' programs
5255 are included. (There is still no support for YP.)
5256
5257 * Tom Quinn has contributed a port of the C library to SGI machines running
5258 Irix 4 (mips-sgi-irix4).
5259
5260 * The new `lockf' function is a simplified interface to the locking
5261 facilities of `fcntl', included for compatibility.
5262
5263 * New time functions `timegm', `timelocal', and `dysize' for compatibility.
5264
5265 * New header file <sys/timeb.h> and new function `ftime' for compatibility.
5266
5267 * New header files <poll.h> and <sys/poll.h> and new function `poll' for
5268 compatibility.
5269
5270 * The error message printed by `assert' for a failed assertion now includes
5271 the name of the program (if using GNU ld) and the name of the calling
5272 function (with versions of GCC that support this).
5273
5274 * The `psignal' function is now declared in <signal.h>, not <stdio.h>.
5275
5276 * The library now includes the <sys/mman.h> header file and memory
5277 management functions `mmap', `munmap', `mprotect', `msync', and
5278 `madvise', on systems that support those facilities.
5279
5280 * The interface for `mcheck' has changed slightly: the function called to
5281 abort the program when an allocation inconsistency is detected now takes
5282 an argument that indicates the type of failure. The new function
5283 `mprobe' lets you request a consistency check for a particular block at
5284 any time (checks are normally done only when you call `free' or `realloc'
5285 on a block).
5286
5287 * It is now possible to easily cross-compile the C library, building on one
5288 system a library to run on another machine and/or operating system. All
5289 you need to do is set the variable `HOST_CC' in `configparms' to the
5290 native compiler for programs to run on the machine you are building on (a
5291 few generator programs are used on Unix systems); set `CC' to the
5292 cross-compiler.
5293
5294 * The new function `fexecve' (only implemented on the GNU system) executes
5295 a program file given a file descriptor already open on the file.
5296 \f
5297 Version 1.07
5298
5299 * Brendan Kehoe has contributed most of a port to the DEC Alpha
5300 running OSF/1 (alpha-dec-osf1). He says it is 75% complete.
5301
5302 * You can set the variable `libprefix' in `configparms' to specify a prefix
5303 to be prepended to installed library files; this makes it easy to install
5304 the GNU C library to be linked as `-lgnuc' or whatever.
5305
5306 * The new `stpncpy' is a cross between `stpcpy' and `strncpy': It
5307 copies a limited number of characters from a string, and returns the
5308 address of the last character written.
5309
5310 * You no longer need to check for whether the installed `stddef.h' is
5311 compatible with the GNU C library. configure now checks for you.
5312
5313 * You can now define a per-stream `fileno' function to convert the
5314 stream's cookie into an integral file descriptor.
5315
5316 * ``malloc (0)'' no longer returns a null pointer. Instead, it
5317 allocates zero bytes of storage, and returns a unique pointer which
5318 you can pass to `realloc' or `free'. The behavior is undefined if
5319 you dereference this pointer.
5320
5321 * The C library now runs on Sony NEWS m68k machines running either
5322 NewsOS 3 or NewsOS 4.
5323
5324 * The new `syscall' function is a system-dependent primitive function
5325 for invoking system calls. It has the canonical behavior on Unix
5326 systems, including unreliable return values for some calls (such as
5327 `pipe', `fork' and `getppid').
5328
5329 * The error code `EWOULDBLOCK' is now obsolete; it is always defined
5330 to `EAGAIN', which is the preferred name. On systems whose kernels
5331 use two distinct codes, the C library now translates EWOULDBLOCK to
5332 EAGAIN in every system call function.
5333 \f
5334 Version 1.06
5335
5336 * The GNU C Library Reference Manual is now distributed with the library.
5337 `make dvi' will produce a DVI file of the printed manual.
5338 `make info' will produce Info files that you can read on line using C-h i
5339 in Emacs or the `info' program.
5340 Please send comments on the manual to bug-glibc-manual@gnu.org.
5341
5342 * The library now supports SVR4 on i386s (i386-unknown-sysv4).
5343
5344 * Brendan Kehoe has contributed a port to Sun SPARCs running Solaris 2.
5345
5346 * Jason Merrill has contributed a port to the Sequent Symmetry running
5347 Dynix version 3 (i386-sequent-dynix).
5348
5349 * The library has been ported to i386s running SCO 3.2.4 (also known as SCO
5350 ODT 2.0; i386-unknown-sco3.2.4) or SCO 3.2 (i386-unknown-sco3.2).
5351
5352 * New function `memory_warnings' lets you arrange to get warnings when
5353 malloc is running out of memory to allocate, like Emacs gives you.
5354
5355 * The C library now contains the relocating allocator used in Emacs 19 for
5356 its editing buffers. This allocator (ralloc) minimizes allocation
5357 overhead and fragmentation by moving allocated regions around whenever it
5358 needs to. You always refer to a ralloc'd region with a "handle" (a
5359 pointer to a pointer--an object of type `void **').
5360
5361 * There is a new `printf' format: `%m' gives you the string corresponding
5362 to the error code in `errno'.
5363
5364 * In `scanf' formats, you can now use `%as' or `%a[' to do the normal `%s'
5365 or `%[' conversion, but instead of filling in a fixed-sized buffer you
5366 pass, the `a' modifier says to fill in a `char **' you pass with a
5367 malloc'd string.
5368
5369 * The `fnmatch' function supports the new flag bits `FNM_LEADING_DIR' and
5370 `FNM_CASEFOLD'. `FNM_LEADING_DIR' lets a pattern like `foo*' match a
5371 name like `foo/bar'. `FNM_CASEFOLD' says to ignore case in matching.
5372
5373 * `mkstemp' is a traditional Unix function to atomically create and open a
5374 uniquely-named temporary file.
5375 \f
5376 Version 1.05
5377
5378 * The standard location for the file that says what the local timezone is
5379 has changed again. It is now `/usr/local/etc/localtime' (or more
5380 precisely, `${prefix}/etc/localtime') rather than `/etc/localtime'.
5381
5382 * The distribution no longer contains any files with names longer than 14
5383 characters.
5384
5385 * `struct ttyent' has two new flag bits: TTY_TRUSTED and TTY_CONSOLE.
5386 These are set by the new `trusted' and `console' keywords in `/etc/ttys'.
5387
5388 * New functions `ttyslot' and `syslog' from 4.4 BSD.
5389 \f
5390 Version 1.04
5391
5392 * The configuration process has changed quite a bit. The `configure'
5393 script is now used just like the configuration scripts for other GNU
5394 packages. The `sysdeps' directory hierarchy is much rearranged.
5395 The file `INSTALL' explains the new scheme in detail.
5396
5397 * The header files no longer need to be processed into ANSI C and
5398 traditional C versions. There is just one set of files to install, and
5399 it will work with ANSI or old C compilers (including `gcc -traditional').
5400
5401 * Brendan Kehoe and Ian Lance Taylor have ported the library to the
5402 MIPS DECStation running Ultrix 4.
5403
5404 * The Sun 4 startup code (crt0) can now properly load SunOS 4 shared libraries.
5405 Tom Quinn contributed the initial code. The GNU C library can NOT yet be
5406 made itself into a shared library.
5407
5408 * Yet further improved support for the i386, running 4.3 BSD-like systems
5409 (such as Mach 3 with the Unix single-server), or System V.
5410
5411 * New function `strncasecmp' to do case-insensitive string comparison
5412 with limited length.
5413
5414 * New function `strsep' is a reentrant alternative to `strtok'.
5415
5416 * New functions `scandir' and `alphasort' for searching directories.
5417
5418 * New function `setenv' is a better interface to `putenv'.
5419
5420 * Ian Lance Taylor has contributed an implementation of the SVID `ftw'
5421 function for traversing a directory tree.
5422
5423 * The GNU obstack package is now also part of the C library.
5424 The new function `open_obstack_stream' creates a stdio stream that
5425 writes onto an obstack; `obstack_printf' and `obstack_vprintf' do
5426 formatted output directly to an obstack.
5427
5428 * Miscellaneous new functions: reboot, nice, sigaltstack (4.4 BSD only),
5429 cfmakeraw, getusershell, getpass, swab, getttyent, seteuid, setegid.
5430
5431 * `FNM_FILE_NAME' is another name for `FNM_PATHNAME', used with `fnmatch'.
5432
5433 * The new functions `strfry' and `memfrob' do mysterious and wonderful
5434 things to your strings.
5435
5436 * There are some new test programs: test-fseek, testmb, and testrand.
5437
5438 * Some work has been done to begin porting the library to 4.4 BSD and Linux.
5439 These ports are not finished, but are a good starting place for really
5440 supporting those systems.
5441
5442 * `/etc/localtime' is now the standard location for the file that says what
5443 the local timezone is, rather than `/usr/local/lib/zoneinfo/localtime'.
5444 This follows the general principle that `/etc' is the place for all local
5445 configuration files.
5446
5447 * The C library header files now use `extern "C"' when used by the C++
5448 compiler, so the C library should now work with C++ code.
5449
5450 * The header file <bstring.h> is gone. <string.h> now declares bcopy,
5451 bcmp, bzero, and ffs. (Update: nowadays these functions are declared
5452 in <strings.h>.)
5453
5454 * Mike Haertel (of GNU e?grep and malloc fame) has written a new sorting
5455 function which uses the `merge sort' algorithm, and is said to be
5456 significantly faster than the old GNU `qsort' function. Merge sort is
5457 now the standard `qsort' function. The new algorithm can require a lot
5458 of temporary storage; so, the old sorting function is called when the
5459 required storage is not available.
5460
5461 * The C library now includes Michael Glad's Ultra Fast Crypt, which
5462 provides the Unix `crypt' function, plus some other entry points.
5463
5464 * The code and header files taken from 4.4 BSD have been updated with the
5465 latest files released from Berkeley.
5466 \f
5467 ----------------------------------------------------------------------
5468 Copyright information:
5469
5470 Copyright (C) 1992-2019 Free Software Foundation, Inc.
5471
5472 Permission is granted to anyone to make or distribute verbatim copies
5473 of this document as received, in any medium, provided that the
5474 copyright notice and this permission notice are preserved,
5475 thus giving the recipient permission to redistribute in turn.
5476
5477 Permission is granted to distribute modified versions
5478 of this document, or of portions of it,
5479 under the above conditions, provided also that they
5480 carry prominent notices stating who last changed them.
5481 \f
5482 Local variables:
5483 version-control: never
5484 fill-column: 76
5485 End: