Ready for 4.12

Removed trailing white space at end of lines

Changes, ldd.1, chown.2, epoll_wait.2, get_mempolicy.2, ioctl_getfsmap.2, madvise.2, membarrier.2, mount.2, pipe.2, readv.2, seccomp.2, sigaltstack.2, splice.2, tee.2, timer_create.2, vmsplice.2, acosh.3, asinh.3, atanh.3, bsd_signal.3, dl_iterate_phdr.3, dlsym.3, flockfile.3, fpathconf.3, ftw.3, getaddrinfo.3, getcontext.3, getgrent.3, if_nametoindex.3, malloc.3, nl_langinfo.3, posix_madvise.3, ptsname.3, rand.3, resolver.3, sigpause.3, sigwait.3, strtol.3, sysconf.3, ttyslot.3, unlocked_stdio.3, unlockpt.3, capabilities.7, cgroup_namespaces.7, inode.7, man.7, pipe.7, sigevent.7: tstamp

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

Changes: Tweaks

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getpid.2: srcfix: add FIXME

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

clone.2: Update BUGS to reflect fact that PID caching was removed in glibc 2.25

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getpid.2: More tidy-ups in discussion of removal of getpid() PID caching

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getpid.2: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getpid.2: Minor tweaks to Carlos's patch

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getpid.2: Note that PID caching is removed as of glibc 2.25

Since glibc 2.25 the PID cache is removed.

Rationale given in the release notes:
https://sourceware.org/glibc/wiki/Release/2.25#pid_cache_removal
~~~
3.2.3. Calls to getpid are no longer cached

The PID cache used by glibc has been removed. In certain scenarios
the cache was not 100% reliable and because of that it was deemed
safer to remove the cache than to potentially return a wrong
answer.

Applications performing getpid() calls in a loop will see the
worst case performance degradation as the library call will
perform a system call at each invocation. Such application uses
were known to exist at least in OpenSSL (fork()-based PRNG
invalidation), but supporting the performance of that specific
invalidation mechanism was not judged to have sufficient value
against immediate and long-term benefits of removing the cache.

Functional reasons exist for the PID cache removal including
problems with PID namespaces, interoperability with raw system
calls (BZ#17214, Chrome: Issue 800183004), and improvements to
spawn (BZ#19957). Performance is actually increased in
pthread_create() with the removal of the cache since the
implementation no longer needs to perform an invalidation step.

Applications performing getpid() in a loop that need to do some
level of fork()-based invalidation can instead use
pthread_atfork() to register handlers to handle the invalidation.
There is work-in-progress to make pthread_atfork() available to
applications that do not link against libpthread.so (Provide
pthread_atfork() without libpthread.so).

Other kinds of invalidation are not supported and the glibc
community will actively look at a kernel assisted mechanism for
state management across fork(), vfork(), clone() and other
interfaces which can benefit from such semantics. It is the same
type of solution required for crypto PRNG reset across such API
calls.
~~~

Signed-off-by: Carlos O'Donell <carlos@redhat.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getline.3: wfix

Signed-off-by: Long Wang <w@laoqinren.net>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

acosh.3, asinh.3, atanh.3: Remove C89 designation.

See https://bugzilla.kernel.org/show_bug.cgi?id=196319

[mtk: confirmed from review of draft of C90 standard]

Signed-off-by: Alex Henrie <alexhenrie24@gmail.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

capabilities.7: Fix reversed descriptions of CAP_MAC_OVERRIDE and CAP_MAC_ADMIN

Reported-by: David Lewis <davlewi3@cisco.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

bpf.2: tfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

seccomp.2: Minor tweaks to Mike's patch

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

seccomp(2): Expand SECCOMP_RET_KILL documentation

First clarify that the process cannot catch this SIGSYS signal.
While the text currently says that, it's easy (IMO) to read
ambiguously and that it's referring to default behavior (no
handler -> process exits).

Then add details regarding coredump behavior.  Before Linux 4.11,
there was no way to get coredumps from such crashes.  Now we can
at least get crashes from single threaded processes.

Signed-off-by: Mike Frysinger <vapier@chromium.org>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

execve.2: tfix

Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getaddrinfo.3: wsfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ld.so.8: Minor tweaks to Carlos's patch

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ld.so.8: Expand DT_RUNPATH details.

ld.so.8: Expand DT_RUNPATH details.

Every 3 years we get asked why DT_RUNPATH doesn't work like DT_RPATH.
The most recent question was here:
https://www.sourceware.org/ml/libc-help/2017-06/msg00013.html

We need to expand the description of DT_RUNPATH to cover this situation
and explain that the DT_RUNPATH entries apply only to the immediate
DT_NEEDED, not that of another, say dlopen'd child object.

Applies to master.

Signed-off-by: Carlos O'Donell <carlos@redhat.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getcontext.3: Exemplar structure should use 'ucontext_t'.

In glibc we just finished a round of purging 'struct ucontext'
which is not in the POSIX reserved namespace of *_t tags. This has
some consequences to applications using the non-standard struct
ucontext:
https://sourceware.org/glibc/wiki/Release/2.26#Removal_of_.27struct_ucontext.27
but it also fixes a namespace conformance issue which is always a
longterm pain for large portable programs.

It was noted by Peter Maydell
(https://sourceware.org/bugzilla/show_bug.cgi?id=21457) that the
linux man pages still had references to 'struct ucontext' but only
in the form of an exemplar structure.

The patch fixes the exemplar to match what is in glibc
and therefore it won't ever suggest users can use 'struct
ucontext'.

Signed-off-by: Carlos O'Donell <carlos@redhat.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

madvise.2: Remove bogus text re POSIX_MADV_NOREUSE

There is a POSIX_FADV_NOREUSE for posix_fadvise(),
but no POSIX_MADV_NOREUSE for any API in POSIX.

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

posix_madvise.3: Remove paragraph about POSIX_FADV_NOREUSE

POSIX_FADV_NOREUSE is documented for posix_fadvise, and a
corresponding POSIX_MADV_NOREUSE flag is not specified by POSIX.

Thanks to Marc Lehmann <debian-reportbug@plan9.de>
See https://bugs.debian.org/865699

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getrandom.2: tfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ttyslot.3: tfix

Signed-off-by: Jakub Wilk <jwilk@jwilk.net>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

tmpfs.5: ffix

Signed-off-by: Jakub Wilk <jwilk@jwilk.net>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

open.2: Note some further advantages of the *at() APIs

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

process-keyring.7: Minor fix: SEE ALSO: remove self-reference

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ttyname.3: tfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

request_key.2: tfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

malloc.3: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

malloc.3: grfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

malloc.3: Document the reallocarray() added in glibc 2.26

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

malloc.3: Minor clarification to CONFORMING TO

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

cfree.3: tfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

readv.2: glibc 2.26 adds library support for preadv2() and pwritev2()

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

readv.2: tfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

fpathconf.3: srcfix: Add self to copyright notice

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

fpathconf.3: Largely rewrite the description of _PC_CHOWN_RESTRICTED

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

fpathconf.3: Rewrite description of _PC_PIPE_BUF

The existing description was not accurate, and lacked details.

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

fpathconf.3: Add an errors section

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

fpathconf.3: srcfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

fpathconf.3: Minor wording fixes

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

fpathconf.3: Rework RETURN VALUE description to add more detail

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sysconf.3: Rework RETURN VALUE description to add more detail

Make the discussion clearer, and add a few details.
Also, fix the problem report from Pavel Tikhomirov
who noted that the man page falsely said that errno
is not changed on a successful return.

Addresses https://bugzilla.kernel.org/show_bug.cgi?id=195955

Reported-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sysconf.3: wfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sysconf.3: Add ERRORS section

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ioctl.2: SEE ALSO: add ioctl_getfsmap(2)

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ioctl_getfsmap.2: ERRORS: order alphabetically

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ioctl_getfsmap.2: add VERSIONS section

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ioctl_getfsmap.2: Fix ordering of sections

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ioctl_getfsmap.2: Minor fixes to Darrick Wong's page

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ioctl_getfsmap.2: Document the GETFSMAP ioctl

Document the new GETFSMAP ioctl that returns the physical layout of a
(disk-based) filesystem.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getgrent.3, sigwait.3: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getgrent.3: Note glibc 2.22 changes for feature test macro requirements

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

wait.2: Note glibc 2.26 changes to feature test macro requirements for waitid()

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sigevent.7: Mention signal.h header file

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sigpause.3: Note glibc 2.26 changes to feature test macro requirements

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sigpause.3: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

nl_langinfo.3: Note feature test macro requirements for nl_langinfo_l()

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sigwait.3: Note glibc 2.26 feature test macro changes

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

bsd_signal.3: Note feature test macro requirements changes for glibc 2.26

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

flockfile.3: Note glibc 2.24 feature test macro requirement changes

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

unlocked_stdio.3: Note glibc 2.24 feature test macro requirement changes

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

chmod.2: Note glibc 2.24 feature test macro requirements changes for fchmod()

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

chmod.2: Put fchmod() feature test macro requirements in a more readable format

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

inode.7: Note glibc 2.24 feature test macro changes for S_IFSOCK and S_ISSOCK()

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ptsname.3: Since glibc 2.26, ptsname_r() no longer gives EINVAL for buf==NULL

Reported-by: Arjun Shankar <arjun.is@lostca.se>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

grantpt.3, ptsname.3, unlockpt.3: Update feature test macro-requirements for glibc 2.24

Cowritten-by: Michael Kerrisk <mtk.manpages@gmail.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

rand.3: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

rand.3: Note glibc 2.24 feature test macro requirement changes for rand_r()

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ttyslot.3: Note feature test macro requirements changes in glibc 2.24

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ttyslot.3: Fix error in feature test macro requirements

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ttyslot.3: Clarify details of use of <stdlib.h> file

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ttyslot.3: wfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sigaction.2: Note feature test macro requirements for 'si_code' constants

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sigaction.2: wfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sigaction.2: Add a subheading for the description of 'si_code'

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

sigaction.2: TRAP_BRANCH and TRAP_HWBKPT are present only on IA64

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

resolver.3: Add basic notes on 'op' argument of res_nmkquery() and res_mkquery()

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

splice.2, tee.2, vmsplice.2: SEE ALSO: add pipe(7)

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

pipe.2: SEE ALSO: add tee(2) and vmsplice(2)

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

pipe.7: SEE ALSO: add tee(2) and vmsplice(2)

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

splice.2: ERRORS: add EINVAL for case where both descriptors refer to same pipe

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

splice.2: ERRORS: split EINVAL error cases

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

cgroup_namespaces.7: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

cgroup_namespaces.7: tfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

cgroup_namespaces.7: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

cgroup_namespaces.7: Add some further explanation of the example shell session

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

cgroup_namespaces.7: Fix a bug in shell session example

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

cgroup_namespaces.7: Some minor wording fixes and clarifications

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

ldconfig.8: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

zdump.8: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

zic.8: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

zic.8: wfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

zic.8: wfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

zdump.8: wfix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

nscd.8: ffix

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

nscd.8: tfix

Reported-by: Brad Bendily <bendily@gmail.com>
Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>

getaddrinfo.3: Fix _POSIX_C_SOURCE value for getaddrinfo()

The correct _POSIX_C_SOURCE value is 200112L, not 201112L in features.h.

Signed-off-by: Michael Kerrisk <mtk.manpages@gmail.com>