aboutsummaryrefslogtreecommitdiffstats
path: root/src/libunixonacid
Commit message (Collapse)AuthorAgeFilesLines
* Align ancillary buffers on struct cmsghdr and int boundariesLaurent Bercot2017-04-232-5/+18
|
* Fix unixmessage_putv() (the normal message wasn't taken into account!)Laurent Bercot2017-03-151-1/+1
|
* More superflous headers cleanup and tiny fixesLaurent Bercot2017-03-1349-85/+6
|
* Types change: big pass on libstddjb and libunixonacidLaurent Bercot2017-02-211-1/+0
| | | | libdatastruct still missing, library still not functional
* More types changesLaurent Bercot2017-02-1762-245/+269
| | | | | | | | | New disize Add size_t to the autotypes list Delete redundant and replace-libc files dev_t/ino_t pass Big size_t pass More things missing, still not operational yet
* add arc4random_addrandom sysdep; next batch of types changes.Laurent Bercot2017-02-152-2/+2
| | | | | The library isn't supposed to be functional yet: there are still a lot of type mismatches.
* Clean up and modernize librandom.Laurent Bercot2016-10-145-6/+6
| | | | | | | | | | | | | | | | | | | | | | Correct random number generation has historically been suprisingly painful to achieve. There was no standard, every system behaved in a subtly different way, and there were a few userland initiatives to get decent randomness, all incompatible of course. The situation is a bit better now, we're heading towards some standardization. The arc4random() series of functions is a good API, and available on a lot of systems - unfortunately not Linux, but on Linux the new getrandom() makes using /dev/random obsolete. So I removed the old crap in librandom, dropped EGD support, dropped dynamic backend selection, made a single API series (random_* instead of goodrandom_* and badrandom_*), added an arc4random backend and a getrandom backend, and defaulted to /dev/urandom backed up by SURF in the worst case. This should be much smaller and logical. However, it's a major API break, so the skarnet.org stack will be changed to adapt.
* ipc_timed_send workaround for the usual OpenBSD braindead headersLaurent Bercot2016-07-241-1/+4
|
* Actually call ipc_timed_recv ipc_timed_recvLaurent Bercot2016-07-201-1/+1
|
* Add ipc_timed_recvLaurent Bercot2016-07-202-0/+31
|
* Add ipc_timed_sendLaurent Bercot2016-07-201-0/+31
|
* Fix closecb support wherever a unixmessage may get dropped; add closecb ↵Laurent Bercot2016-04-226-15/+24
| | | | support to unixconnection
* Tentative re-commit of the closecb featureLaurent Bercot2016-02-024-10/+30
|
* Revert " Add support for advanced closing callbacks to ↵Laurent Bercot2016-02-024-30/+10
| | | | | | | | unixmessage_sender_flush." This reverts commit c9c7e241a7b68f67e0cc8a7d3874510dced532d6. (something is segfaulting, reverting until I find the fix)
* Add support for advanced closing callbacks to unixmessage_sender_flush.Laurent Bercot2016-01-184-10/+30
| | | | | (Typically, this will be used for refcounting when broadcasting copies of the same fd.)
* Add atomic_rm_rf and atomic_rm_rf_tmpLaurent Bercot2015-08-133-2/+43
|
* - add bitarray_countLaurent Bercot2015-05-161-0/+25
| | | | | - add openwritevnclose_at - rc for 2.3.5.0
* OpenBSD sucks, but everyone knew this already.v2.3.4.0Laurent Bercot2015-05-051-0/+3
|
* - add stat_atLaurent Bercot2015-05-051-0/+77
| | | | | | - make case_diff* work as a function is more cases - make sig_number case-insensitive - make child signal unblocking work in child_spawn*
* Applied Roman Kimov's fixesv2.3.1.3Laurent Bercot2015-03-131-1/+1
|
* - Bugfix: unixmessage_sender_init didn't initialize b->shortyLaurent Bercot2015-02-192-0/+3
| | | | - byte_zero the ancilbuf's padding to appease valgrind
* - Add siovec_truncLaurent Bercot2015-02-052-2/+4
| | | | | | | | - Add openwritevnclose - replace buffer_read/write with fd_readsv/writesv - add localtmn_from_tain_g - refactor cdb_make - rewrite buffer_getvall and buffer_putvall, with a bugfix
* Bugfixes:v2.2.1.0Laurent Bercot2015-01-231-1/+2
| | | | | | | - seek_set now uses SEEK_SET (those functions are useless now, they will go at the next major version update) - skaclient_server_init now correctly sanitizes unixmessage_timed_receive
* - Added unixconnectionLaurent Bercot2015-01-193-0/+26
| | | | | - Added sig_name & sig_number (suggestion from Olivier Brunel) - version bumped to 2.2.1.0, rc
* Without a segfault in unixmessage_sender_free, please.v2.2.0.0Laurent Bercot2015-01-101-5/+8
|
* Bugfix: unixmessage_sender_free didn't close all the pending fdsLaurent Bercot2015-01-103-12/+11
|
* Add unixmessage_unput, for transaction cancellationLaurent Bercot2015-01-091-0/+23
|
* Bugfix: unixmessage_put was recording the wrong number of fdsLaurent Bercot2015-01-092-3/+3
|
* - Bugfixes in unixmessage/skaclient (short writes / fd leakage / DoS)v2.1.0.0Laurent Bercot2015-01-0610-48/+130
| | | | | | - ABI change: unixmessage protocol header is now 6 bytes (was 8) - API change: skaclient_start(_async) now takes an "options" argument - version increase to 2.1.0.0
* Add unixmessage_drop and some stuff in skaclientLaurent Bercot2015-01-041-0/+13
|
* Remove redundant _XPG4_2 in unixmessage_sender_flushLaurent Bercot2015-01-031-4/+0
|
* ... without breaking everything, if possibleLaurent Bercot2015-01-031-1/+1
|
* Easier error reporting in unixmessage_handleLaurent Bercot2015-01-031-1/+2
|
* Really add it >.>Laurent Bercot2015-01-021-0/+13
|
* Add skaclient_syncify for easier async/sync handlingLaurent Bercot2015-01-023-15/+7
|
* Safe wrappers around sendmsg and recvmsg in unixmessage primitivesLaurent Bercot2014-12-312-5/+15
|
* So MSG_WAITALL can block even with MSG_DONTWAIT... wtfbsdseriously.Laurent Bercot2014-12-212-3/+13
| | | | | Only include MSG_WAITALL when it's tested nb, then. Also test yet another fix for the tryancilautoclose FreeBSD problem.
* So the non-portable MSG_DONTWAIT may alleviate the problem... mightLaurent Bercot2014-12-211-12/+11
| | | | as well use it on architectures that support it.
* Found the BSD unixmessage bug. recvmsg(..., MSG_WAITALL) blocksLaurent Bercot2014-12-201-6/+15
| | | | | | until buffer full or socket shutdown on BSD, *even if the socket is nonblocking*. So I added an "okwaitall" sysdeps, and won't set MSG_WAITALL on retarded systems.
* Lots of cosmetic fixes.Laurent Bercot2014-12-206-6/+6
| | | | sha256 documented.
* Fix fd management in new child_spawn1, make socket nb for skaclientLaurent Bercot2014-12-181-1/+1
|
* Separate socket from pipe in child_spawn. Less code in skaclient!Laurent Bercot2014-12-181-1/+1
|
* Fix timed_getlnmax: correctly update wLaurent Bercot2014-12-181-2/+4
|
* Fix: timed_buffer_flush -> buffer_timed_flushLaurent Bercot2014-12-151-1/+1
|
* unixmessage_receive: don't read more than 1 message at a timeLaurent Bercot2014-12-141-2/+6
|
* Fix test of sendmsg return value (compare against int...)Laurent Bercot2014-12-141-1/+1
|
* Fix initial value for the length of the kolbak_queue.Laurent Bercot2014-12-131-2/+1
| | | | | This was the reason for the s6-dnsname-filter segfault. It took me 2 days, valgrind and gdb to corner that one. >.>
* Correctly (hopefully) update the main cbuffer in unixmessage_receiver_fillLaurent Bercot2014-12-061-0/+1
|
* Ditch diuints in unixmessage_receiver_fill, revert recvmsg() interpretationLaurent Bercot2014-12-061-9/+3
|
* Fix: adjust unixmessage_receiver_fill() math after recvmsg(). </3Laurent Bercot2014-12-061-1/+1
|
generated by cgit v1.3.1 (git 2.54.0) at 2026-09-14 06:07:52 +0000