Changes

   1 ==================== Changes in man-pages-5.01 ====================
   2
   3 Released: 2019-05-09, Munich
   4
   5
   6 Contributors
   7 ------------
   8
   9 The following people contributed patches/fixes or (noted in brackets
  10 in the changelog below) reports, notes, and ideas that have been
  11 incorporated in changes in this release:
  12
  13 Abhinav Upadhyay <er.abhinav.upadhyay@gmail.com>
  14 Andreas Korb <andreas.d.korb@gmail.com>
  15 Anisse Astier <anisse@astier.eu>
  16 Brice Goglin <Brice.Goglin@inria.fr>
  17 Carlos O'Donell <carlos@redhat.com>
  18 Dr. Jürgen Sauermann <mail@xn--jrgen-sauermann-zvb.de>
  19 Egmont Koblinger <egmont@gmail.com>
  20 Elias Benali <stackptr@users.sourceforge.net>
  21 Elliot Hughes <enh@google.com>
  22 Florian Weimer <fweimer@redhat.com>
  23 Hugues Evrard <hevrard@google.com>
  24 Jakub Nowak <jakub.jakub.nowak@gmail.com>
  25 Jakub Wilk <jwilk@jwilk.net>
  26 Keegan Saunders <keeganwsaunders@gmail.com>
  27 Lucas Werkmeister <mail@lucaswerkmeister.de>
  28 Marcus Huewe <suse-tux@gmx.de>
  29 Michael Kerrisk <mtk.manpages@gmail.com>
  30 Michael Witten <mfwitten@gmail.com>
  31 Seth Troisi <sethtroisi@google.com>
  32 Slavomir Kaslev <kaslevs@vmware.com>
  33 Vincent Lefevre <vincent@vinc17.net>
  34 Wladimir Mutel <muwlgr@gmail.com>
  35
  36 Apologies if I missed anyone!
  37
  38
  39 Newly documented interfaces in existing pages
  40 ---------------------------------------------
  41
  42 tsearch.3
  43     Florian Weimer  [Michael Kerrisk]
  44         Document the twalk_r() function added in glibc 2.30
  45
  46
  47 New and changed links
  48 ---------------------
  49
  50 twalk_r.3
  51     Michael Kerrisk
  52         New link to twalk(3) page
  53
  54
  55 Changes to individual pages
  56 ---------------------------
  57
  58 accept.2
  59     Michael Kerrisk
  60         Note that 'addrlen' is left unchanged in the event of an error
  61             See http://austingroupbugs.net/view.php?id=836.
  62
  63 bpf.2
  64     Michael Kerrisk
  65         Update kernel version info for JIT compiler
  66
  67 clone.2
  68     Michael Kerrisk  [Jakub Nowak]
  69         CLONE_CHILD_SETTID has effect before clone() returns *in the child*
  70             CLONE_CHILD_SETTID may not have had effect by the time clone()
  71             returns in the parent, which could be relevant if the
  72             CLONE_VM flag is employed. The relevant kernel code is in
  73             schedule_tail(), which is called in ret_from_fork()
  74             in the child.
  75
  76             See https://bugzilla.kernel.org/show_bug.cgi?id=203105
  77
  78 execve.2
  79 exec.3
  80     Michael Kerrisk  [Dr. Jürgen Sauermann]
  81         Consistently use the term 'pathname' (not 'path')
  82
  83 execve.2
  84     Michael Kerrisk
  85         Note that stack+environ size is also limited to 3/4 of _STK_LIM
  86             In fs/exec.c::prepare_arg_pages(), we have:
  87
  88                     limit = _STK_LIM / 4 * 3;
  89                     limit = min(limit, bprm->rlim_stack.rlim_cur / 4);
  90     Michael Kerrisk  [Dr. Jürgen Sauermann]
  91         SEE ALSO: refer to exec(3) (rather than execl(3))
  92
  93 pipe.2
  94     Michael Kerrisk
  95         Note that 'pipefd' is left unchanged in the event of an error
  96             See http://austingroupbugs.net/view.php?id=467.
  97
  98 sched_setaffinity.2
  99     Michael Kerrisk  [Brice Goglin]
 100         Correct details of return value of sched_getaffinity() syscall
 101             Things changed a little with:
 102
 103                 commit cd3d8031eb4311e516329aee03c79a08333141f1
 104                 Author: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
 105                 Date:   Fri Mar 12 16:15:36 2010 +0900
 106
 107 setfsgid.2
 108     Michael Kerrisk
 109         Rewrite for improved clarity and defer to setfsuid() for details
 110             Rewrite for improved clarity and defer to setfsuid(2) for the
 111             rationale of the fsGID rather than repeating the same details
 112             in this page.
 113
 114 setfsuid.2
 115     Michael Kerrisk
 116         Rewrite for improved clarity and to hint history more explicitly
 117             The current text reads somewhat clumsily. Rewrite it to introduce
 118             the eUID and fsUID in parallel, and more clearly hint at the the
 119             historical rationale for the fsUID, which is detailed lower in
 120             the page.
 121
 122 socketpair.2
 123     Michael Kerrisk
 124         Clarify that 'sv' is left unchanged in the event of an error
 125             See also http://austingroupbugs.net/view.php?id=483.
 126
 127 splice.2
 128     Slavomir Kaslev
 129         EAGAIN can occur when called on nonblocking file descriptors
 130
 131 syscalls.2
 132     Michael Kerrisk  [Andreas Korb]
 133         Remove crufty text about i386 syscall dispatch table
 134             The removed text long ago ceased to be accurate. Nowadays, the
 135             dispatch table is autogenerated when building the kernel (via
 136             the kernel makefile, arch/x86/entry/syscalls/Makefile).
 137
 138 tee.2
 139     Slavomir Kaslev
 140         EAGAIN can occur when called on nonblocking file descriptors
 141
 142 fopen.3
 143     Elliot Hughes
 144         Explain BSD vs glibc "a+" difference
 145             Where is the initial read position for an "a+" stream?
 146
 147             POSIX leaves this unspecified. Most BSD man pages are silent, and
 148             MacOS has the ambiguous "The stream is positioned at the end of
 149             the file", not differentiating between reads and writes other than
 150             to say that fseek(3) does not affect writes. glibc's documentation
 151             explicitly specifies that the initial read position is the
 152             beginning of the file.
 153
 154             My new wording is based on the BSD implementations, so you may
 155             prefer to replace the non-glibc section with "unspecified", or
 156             indeed remove all claims about the initial read position.
 157
 158 mallinfo.3
 159     Elliott Hughes
 160         Further discourage use of mallinfo()
 161             The BUGS section already explains why you need to be cautious
 162             about using mallinfo, but given the number of bug reports we see
 163             on Android, it seems not many people are reading that far. Call it
 164             out up front.
 165
 166 malloc_trim.3
 167     Carlos O'Donell
 168         Update trimming information
 169             Since glibc 2.8, commit 68631c8eb92, the malloc_trim function has
 170             iterated over all arenas and free'd back to the OS all page runs
 171             that were free.  This allows an application to call malloc_trim to
 172             consolidate fragmented chunks and free back any pages it can to
 173             potentially reduce RSS usage.
 174
 175 posix_memalign.3
 176     Elliot Hughes
 177         Some functions set errno
 178             True of bionic, glibc, and musl. (I didn't check elsewhere.)
 179
 180 resolver.3
 181     Michael Kerrisk  [Wladimir Mutel]
 182         Mention that some functions set 'h_errno'
 183
 184 stdarg.3
 185     Michael Kerrisk  [Vincent Lefevre]
 186         Remove the NOTES section describing the ancient varargs macros
 187             stdarg.h is now 30 years old, and gcc long ago (2004) ceased to
 188             implement <varargs.h>. There seems little value in keeping this
 189             text.
 190
 191             See https://bugzilla.kernel.org/show_bug.cgi?id=202907
 192     Michael Kerrisk  [Egmont Koblinger]
 193         Add a note that "..." in function signature means a variadic function
 194             Egmont suggested adding this, because the string "..." appears
 195             at several other points in the page, but just to indicate that
 196             some text is omitted from example code.
 197
 198 strerror.3
 199     Jakub Wilk
 200         Don't discuss buffer size for strerror_l()
 201             Unlike strerror_r(), strerror_l() doesn't take buffer length as an
 202             argument.
 203
 204 strtol.3
 205 strtoul.3
 206     Jakub Wilk
 207         SEE ALSO: add strtoimax(3), strtoumax(3)
 208
 209 sysconf.3
 210     Michael Kerrisk  [Hugues Evrard]
 211         Clearly note that _SC_PAGESIZE and _SC_PAGE_SIZE are synonyms
 212
 213 tsearch.3
 214     Florian Weimer
 215         Do not use const arguments in twalk() callback
 216             The const specifier is not part of the prototype (it only applies
 217             to the implementation), so showing it here confuses the reader.
 218     Michael Kerrisk
 219         SYNOPSIS: add missing definition of 'VISIT' type
 220     Michael Kerrisk
 221         Reformat twalk() and twalk_r() prototypes for easier readability
 222
 223 console_codes.4
 224     Jakub Wilk
 225         Document that \e[1;n] and \e[2;n] support 16 colors
 226             Source: setterm_command() in drivers/tty/vt/vt.c
 227
 228 elf.5
 229     Michael Kerrisk  [Keegan Saunders]
 230         A data segment does not have PF_X
 231
 232 proc.5
 233     Michael Witten  [Anisse Astier]
 234         Add missing Inode field to /proc/net/unix
 235
 236 hostname.7
 237     Florian Weimer
 238         HOSTALIASES/search path processing is DNS-specific
 239             Other NSS modules do not necessarily honor these settings.
 240
 241 inode.7
 242     Michael Kerrisk
 243         Note that timestamp fields measure time starting at the Epoch
 244     Michael Kerrisk
 245         Timestamp fields are structures that include a nanosecond component
 246     Michael Kerrisk
 247         Add references to execve(2) to describe set-UID/set-GID behaviors