summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2020-01-27 Put correct /library.so back into MakefileLaurent Bercot
2020-01-27 Fix shared libs installation; prepare for 2.9.2.0Laurent Bercot
2019-10-26 Simplify endianness testLaurent Bercot
2019-10-26 Switch getrandom sysdep back to choose clLaurent Bercot
2019-10-25 Better bigkvLaurent Bercot
2019-10-25 Add bigkvLaurent Bercot
2019-10-25 Document that getrandom is now clrLaurent Bercot
2019-10-25 getrandom sysdep must be clr, THANKS GLIBCLaurent Bercot
2019-10-23 Also fix documentation for default pathLaurent Bercot
2019-10-23 Remove obsolete documentation for --enable-clockLaurent Bercot
2019-10-21 version: 2.9.1.0v2.9.1.0Laurent Bercot
2019-10-20 Real fix for undefined PATH_MAXLaurent Bercot
2019-10-20 Don't rely on the existence of PATH_MAXLaurent Bercot
2019-10-18 __GNU__ is actually preferred to __gnu_hurd__, it seemsLaurent Bercot
2019-10-17 nonposix.h: define _GNU_SOURCE for Hurd, remove __GLIBC__ testLaurent Bercot
__GLIBC__ and similar ftms are not defined in the compiler, but in features.h, which isn't POSIX so cannot be included, and standard headers can't be included either because nonposix.h defines ftms that must exist *prior to* std headers inclusion. The kernel ftms, however, are defined by the compiler, so we can use them.
2019-10-10 Fix path_canonicalize()'s returned length when absoluteLaurent Bercot
2019-10-09 Add path_canonicalize, prepare for 2.9.1.0Laurent Bercot
2019-10-09 Fix SKALIBS_SIZEOF* macro declarations in skalibs/sysdeps.hLaurent Bercot
2019-09-22 version: 2.9.0.0v2.9.0.0Laurent Bercot
2019-09-21 Update tainnow.lib mentions in docLaurent Bercot
2019-09-20 INSTALL now explains the easy cross-compilationLaurent Bercot
2019-09-20 THAT'S IT WE HAVE EASY CROSS-COMPILATION FOLKSLaurent Bercot
2019-09-20 We're down to ONE rogue sysdep, boys. ONE.Laurent Bercot
2019-09-20 Better endianness test, because BSDLaurent Bercot
2019-09-20 sysdeps redesign: first part: minimize clr testsLaurent Bercot
Remaining clr: emptyregex: can be safely guessed to no nullispointer: can't be safely guessed, but do we need the test? devurandom: can't be safely guessed malloc0: can more or less be safely guessed to no
2019-09-07 Update docLaurent Bercot
2019-09-06 Change tain_now_set_* API: also give an initial valueLaurent Bercot
2019-09-04 Also better when you actually pull the sysdepsLaurent Bercot
2019-09-04 With the correct sysdeps macros, pleaseLaurent Bercot
2019-09-03 Big wallclock/stopwatch refactor. It was long overdue.Laurent Bercot
* --enable-clock and --enable-monotonic are gone * tain_sysclock() has been renamed tain_wallclock_read() * tain_wallclock_read() reads from CLOCK_REALTIME (or gettimeofday()) * tain_clockmon[_init]() have been renamed to tain_stopwatch_[read|init]() and now accept a monotonic clock name as an extra argument * tain_now() points to the system (wall) clock by default * tain_now_set_[stopwatch|wallclock]() can be used to switch Now to make a pass on all skarnet.org programs and add a tain_now_set_stopwatch() call everywhere needed... >.>
2019-08-19 Add more reproducibility to the buildLaurent Bercot
2019-08-07 iopause_select: add protection against UB in FD_SETLaurent Bercot
2019-07-31 bugfix: env_dump() unterminated stringsLaurent Bercot
2019-07-23 Fix uint16.h and uint32.h dependenciesLaurent Bercot
2019-07-23 Typo fixLaurent Bercot
2019-07-22 Factor all the BSD complaints into one new headerLaurent Bercot
2019-07-22 Document changesLaurent Bercot
2019-07-22 Tentative magic to support out-of-path CC, especially in cross-buildsLaurent Bercot
2019-07-22 Rework skalibs/nonposix.h to avoid glibc squealing like a pigLaurent Bercot
Also move a conformance workaround from nonposix.h to posixishard.h This may break things on MacOS, needs testing.
2019-07-19 Simplify cross-build support some moreLaurent Bercot
2019-07-19 Accept configure invocation from any directoryLaurent Bercot
2019-07-19 Accept raw external $CC for cross-compilingLaurent Bercot
2019-07-19 bits-fmtscan refactor, part twoLaurent Bercot
2019-07-11 bugfix: memcmp doesn't guarantee not touching bytes after cmp fails.Laurent Bercot
2019-07-11 Better with the new files added >.>Laurent Bercot
2019-07-11 bits-fmtscan refactor, part oneLaurent Bercot
2019-06-21 Doc fixLaurent Bercot
2019-06-16 child_spawn*: make sure posix_spawn() doesn't clobber pipes with dup2+closeLaurent Bercot
2019-05-14 version: 2.8.1.0v2.8.1.0Laurent Bercot
2019-05-11 Add tryemptyregex/skalibs_regcomp, prepare for 2.8.1.0Laurent Bercot