aboutsummaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* - Bugfixes in unixmessage/skaclient (short writes / fd leakage / DoS)v2.1.0.0Laurent Bercot2015-01-0612-73/+182
| | | | | | - 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 Bercot2015-01-043-0/+17
|
* More macros for asyncout in skaclient.hLaurent Bercot2015-01-041-0/+6
|
* 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
|
* Add unixmessage_receiver_isfullLaurent Bercot2015-01-031-0/+1
|
* Really add it >.>Laurent Bercot2015-01-021-0/+13
|
* Add skaclient_syncify for easier async/sync handlingLaurent Bercot2015-01-024-15/+9
|
* Remove the --enable-right-tz configure option.Laurent Bercot2015-01-026-44/+43
| | | | | | Make it a run-time autodetection instead. Update doc. Thanks to Paul Jarc for the suggestion.
* Safe wrappers around sendmsg and recvmsg in unixmessage primitivesLaurent Bercot2014-12-312-5/+15
|
* Add _GNU_SOURCE to trynbwaitall.c because glibc headersLaurent Bercot2014-12-311-0/+3
| | | | | don't define kill() with only _BSD_SOURCE. That's very disappointing from GNU.
* Renormalize buffer size, even if we lose 1 charv2.0.0.0Laurent Bercot2014-12-224-10/+10
|
* child_spawn1_internal fix for noposixspawnLaurent Bercot2014-12-211-1/+1
|
* OK, this one works, and the problem was actually on my side (failure toLaurent Bercot2014-12-213-19/+4
| | | | | | use CMSG_LEN where required in tryancilautoclose). Also, good news: the sys/param.h inclusion in FreeBSD isn't necessary anymore.
* next generation of the FreeBSD fix...Laurent Bercot2014-12-213-15/+5
|
* So MSG_WAITALL can block even with MSG_DONTWAIT... wtfbsdseriously.Laurent Bercot2014-12-213-16/+19
| | | | | Only include MSG_WAITALL when it's tested nb, then. Also test yet another fix for the tryancilautoclose FreeBSD problem.
* Stricter trynbwaitall to actually catch the problemsLaurent Bercot2014-12-211-3/+8
|
* So the non-portable MSG_DONTWAIT may alleviate the problem... mightLaurent Bercot2014-12-213-14/+30
| | | | as well use it on architectures that support it.
* Something is rotten in the state of FreeBSD... but we all knew itLaurent Bercot2014-12-201-1/+2
| | | | already. More tentative workarounds.
* FreeBSD now chokes on tryancilautoclose ; tentative fixLaurent Bercot2014-12-201-31/+33
|
* Found the BSD unixmessage bug. recvmsg(..., MSG_WAITALL) blocksLaurent Bercot2014-12-202-6/+132
| | | | | | 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 Bercot2014-12-20106-190/+188
| | | | sha256 documented.
* Cosmetic fixes in iobuffer (was looking for a bug, but it wasn't there)Laurent Bercot2014-12-204-18/+18
|
* Fix fd management in new child_spawn1, make socket nb for skaclientLaurent Bercot2014-12-183-15/+19
|
* fd leak fix in child_spawn1_internalLaurent Bercot2014-12-181-2/+2
|
* Separate socket from pipe in child_spawn. Less code in skaclient!Laurent Bercot2014-12-187-68/+89
|
* Fix timed_getlnmax: correctly update wLaurent Bercot2014-12-181-2/+4
|
* Remove intentional crashLaurent Bercot2014-12-171-2/+0
|
* Make build crash early on purpose (to test test suite...)Laurent Bercot2014-12-171-0/+2
|
* 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 Bercot2014-12-132-5/+4
| | | | | This was the reason for the s6-dnsname-filter segfault. It took me 2 days, valgrind and gdb to corner that one. >.>
* Add guards around EOVERFLOW defines. (Thanks, OpenBSD)Laurent Bercot2014-12-105-0/+10
|
* Fix child_spawn fd closing when n=1Laurent Bercot2014-12-101-1/+1
|
* Remove deprecated sig.c and useless select wrapperLaurent Bercot2014-12-093-47/+0
|
* 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
|
* Rewrite unixmessage_receiver without buffer, only cbufferLaurent Bercot2014-12-0615-126/+129
|
* It's possible to statically initialize unixmessage_receiver_0Laurent Bercot2014-12-054-30/+12
|
* Bugfix: skaclient_start and skaclient_startf need a flush before a receive!Laurent Bercot2014-12-052-0/+14
|
* Simplify child_spawn: remove syncpipe in the posix_spawn() caseLaurent Bercot2014-12-051-14/+9
|
* Add skagetln_nofillLaurent Bercot2014-12-053-10/+22
|
* Add tain_zeroLaurent Bercot2014-12-052-0/+6
|
* More adjustments to unixmessage and skaclientLaurent Bercot2014-12-054-12/+17
|
* Better interface for simple skaclient serversLaurent Bercot2014-12-047-8/+74
|
* initial commit with rc for skalibs-2.0.0.0Laurent Bercot2014-09-18853-0/+19363
generated by cgit v1.3.1 (git 2.54.0) at 2026-09-02 00:13:07 +0000