openbsd/src/lib/libc, branch OPENBSD_7_8A mirror of https://github.com/libressl/openbsd.git
https://git.lua4.win/openbsd/atom?h=OPENBSD_7_82025-08-04T14:11:37+00:00Sort NAME, RETURN VALUES, ERRORS, and STANDARDS in the same order as SYNOPSIS.2025-08-04T14:11:37+00:00schwarze2025-08-04T14:11:37+00:00urn:sha1:5a2842c24fa46be5dfa91f663185a741c07702be
Sort HISTORY chronologically.
No text change.
Implement the POSIX-2024 close-on-fork flag, but modified to be2025-08-04T04:59:31+00:00guenther2025-08-04T04:59:31+00:00urn:sha1:cc34b672c114e027e8e6bdb0c901bb4b2607ff3f
reset on exec as preserving it across exec is not necessary for its
original purpose and has security and usability concerns.
Many thanks to Ricardo Branco (rbranco (at) suse.de) who did an
independent implementation, caught that /dev/fd/* needed to be
handled, and provided a port of the illumos test suite. Thanks
to tb@ for assistance with that.
ok deraadt@
replace the flockfile backend with a per FILE recursive mutex.2025-08-04T01:44:33+00:00dlg2025-08-04T01:44:33+00:00urn:sha1:97f291299e43eb39dd2ecdc18098d938d467dac6
the flockfile implementation in thread/rthread_file.c used an
external lock, and associated it with the relevant FILE * as needed.
this isn't great for a lot of reasons, complexity being the big
one, but the straw that broke the camels back is that it uses a
single spinlock to coordinate all of this, which in turn generates
a lot of sched_yield syscalls.
this avoids all the code complexity and the spinlock by just embedding
a small __rctmx in every FILE.
tested by and ok tb@ jca@
ok claudio@
This thread-aware profiling support diff comes from Yuichiro NAITO.2025-07-16T16:22:58+00:00deraadt2025-07-16T16:22:58+00:00urn:sha1:5506a044dc35162b58584a83a4099b4268ceba2a
At runtime, profiling data is stored per-thread. Upon termination, the
per-thread profiling data pools are merged into a into a single record,
which is then written out (using the new kernel-assisted system). I think
the original gmon merging parts may have come from or been inspired by
someone at NetBSD.
This has been delayed because there were concerns about adding a TLS object
which would require other library cranks, so this rides all the major cranks
today.
ok kettenis tb claudio sthen etc etc etc
add libc specific __cmtx and __rcmtx locks for internal use.2025-07-12T23:59:44+00:00dlg2025-07-12T23:59:44+00:00urn:sha1:417b1213b262bbe6d34c708537dff4b062920bfa
__cmtx provides mutual exclusion using futex(2) and cas on archs
that support it, or _spinlocks on the rest. __rcmtx is a recursive
mutex built on top of __cmtx, so it inherits the use of futex and
cas/spinlock from __cmtx.
until now the options we had for locking between threads in libc
were spinlocks or pthread mutexes. spinlocks use sched_yield to
relax if they have to spin on a contended lock, which we are trying
to minimise the use of as much as possible. pthread_mutex is
relatively large in memory and offers a lot of complicated features
which are unecessary for most of libc. the non cas/futex version
of pthread_mutexes currently relies on __thrsleep and __thrwakeup,
which we also want to deprecate.
having a small futex based lock available everywhere will help us
move away from overuse of spinlocks, and deprecate __thrsleep and
__thrwakeup.
ok kettenis@
jca@ provided his eyes too.
make the argument to ether_ntoa const.2025-06-29T00:33:46+00:00dlg2025-06-29T00:33:46+00:00urn:sha1:a2df642900a546ab79508391ab506e0a6d69eab3
this drives me nuts when i want to print something out of what's already
const. casting it works, but feels gross.
ok guenther@ tb@ deraadt@
enh says this is already in bionic/glibc/musl
The mdoc(7) .Ft macro does not need quoting of its arguments, but about2025-06-13T18:34:00+00:00schwarze2025-06-13T18:34:00+00:00urn:sha1:d939c0f8976619330bcf368e12700a6ab1d14f99
10% of our manual pages using this macro employed useless quoting anyway.
Remove these quotes such that they do not incite fear, uncertainty,
and doubt in developers who happen to look at these pages.
jmc@ and tb@ agree with the direction.
<uvm/uvmexp.h> is clearly only for VM_MALLOC_CONF, but <sys/sysctl.h> takes2025-06-12T16:07:09+00:00deraadt2025-06-12T16:07:09+00:00urn:sha1:bef775ccefd97b9e2ee02eb252923ac7d27bfbdc
care of doing that include.
Revert 1.144 of lib/libc/stdlib/malloc.3. It was changed by accident2025-06-04T00:38:01+00:00yasuoka2025-06-04T00:38:01+00:00urn:sha1:0afcce73c60015d8a6f4c42857ca2338ec20ca31
by my previous commit.
Again. Make exit(), fclose(), fflush(), and freopen() comply with2025-06-03T14:15:53+00:00yasuoka2025-06-03T14:15:53+00:00urn:sha1:6ed3a1ddc5d30f8d2745857ce2b60d3f6756aa82
POSIX-2008 requirements for setting the underlying file position when
flushing read-mode streams, and make an fseek()-after-fflush() not
change the underlying file position. This commit fixes some minor
problems of the previous.
previous diff from guenther
Much testing, review, assistence form tb@
ok tb@ millert@ for the previous
ok asou