--- cvsroot/Coro/Changes 2014/03/04 05:07:44 1.625 +++ cvsroot/Coro/Changes 2015/10/04 13:07:55 1.648 @@ -5,35 +5,73 @@ TODO: __GCC_HAVE_DWARF2_CFI_ASM TODO: swap_sv, maybe add scope_swap_sv, andallow to unswap by repeat? TODO: swap_sv not undone in pool +TODO: croak when async_pool tries to run canceled thread? -Program terminated with signal 11, Segmentation fault. -#0 coro_semaphore_adjust (av=0x0, adjust=0) at State.xs:2915 -2915 State.xs: No such file or directory. -(gdb) bt -#0 coro_semaphore_adjust (av=0x0, adjust=0) at State.xs:2915 -#1 0x00007f638f0c0703 in slf_destroy (coro=0x5957180) at State.xs:2605 -#2 coro_state_destroy (coro=0x5957180) at State.xs:1712 -#3 0x00007f638f0c150c in coro_state_destroy (coro=) at State.xs:1709 -#4 XS_Coro__destroy (cv=) at State.xs:3853 -#5 0x00000000004b572d in Perl_pp_entersub () at pp_hot.c:2881 -#6 0x00000000004adfe3 in Perl_runops_standard () at run.c:42 -#7 0x000000000044335f in S_run_body (oldscope=) at perl.c:2462 -#8 perl_run (my_perl=) at perl.c:2383 -#9 0x00007f638f0ba3a1 in cctx_run (arg=) at State.xs:1478 -#10 0x00007f638f0b74d5 in coro_init () at libcoro/coro.c:96 -(gdb) up -#1 0x00007f638f0c0703 in slf_destroy (coro=0x5957180) at State.xs:2605 -2605 in State.xs -(gdb) p coro->slf_frame -$1 = { - prepare = 0x7f638f0baf60 , - check = 0x7f638f0b6e70 , - data = 0x0, - destroy = 0x7f638f0bb820 - - -TODO: warnhook reports from wrong line number now, when arg is Math::GMP number. - - FATAL: $Coro::IDLE blocked itself will now use Carp::confess + - fix memory leak when Coro::Handle uses Coro::EV internally + (testcase by Sten Sten). + - update code to libev 4 API (internally, Coro still used the + version 3 API calls). + - remove unused hv_sig. + +6.47 Sat Jul 11 03:58:20 CEST 2015 + - swap_svs - NVs can be in the SV head beginning with 5.20. + +6.46 Tue Jun 30 14:36:30 CEST 2015 + - restore portability to perls with windows fork emulation + (patch by Petr Písař). + +6.45 Tue Jun 30 01:40:08 CEST 2015 + - restore portability to perl 5.16 and below. + +6.44 Tue Jun 30 00:41:54 CEST 2015 + - avoid segfaulting (or worse) when tracing a canceled thread. + - the Event module silently broke it's public hook API, causing + failures on perls compiled with -Duselongdouble. Requiring + the latest version of Event because I don't know which release + changed this (it's not mentioned in the ChangeLog). + - allow xs level enterleave hooks via CoroAPI. + +6.43 Thu Jun 4 15:38:14 CEST 2015 + - use stability canary. + - port to stableperl-5.22.0-1.001. + - update libecb. + +6.42 Wed Feb 11 20:29:52 CET 2015 + - Coro::SemaphoreSet->try did not actually work (analyzed by + SATO Kentaro). + - upgrade libecb to be C11 compliant. + +6.41 Sat Sep 6 22:08:46 CEST 2014 + - restore portability to perl 5.8.x. + - give new Coro's a valid GvHV(PL_hintgv) - this is slow and takes + up some memory, but fixes "use feature" and similar modules + when used inside a Coro. + - allow zero as argument to Coro::Channel to mean the same thing + as no argument. this works with older versions as well, + but wasn't legal until now. + - slightly better c header file detection. + +6.39 Mon Jun 2 00:00:08 CEST 2014 + - work around more incompatible changes in 5.20. + +6.38 Sun Jun 1 21:54:23 CEST 2014 + - check that perl slots actually have enough space to hold + interpreter variables. + - untested port to perl 5.19 (64 bit tmps indices) (reported + by Andreas König). + - croak when cancel is called without a thread context. + +6.37 Tue Mar 4 13:27:33 CET 2014 + - *sigh*, removed leftover debugging code from debugging a + perl bug, of all things. + +6.36 Tue Mar 4 07:11:59 CET 2014 + - semaphores would not clear the destroy hook when interrupted + by ->throw, causing segfaults or worse. + - ->throw on a thread waiting for a semaphore did not acquire + the semaphore, but also didn't wake up other waiters, + possibly causing a deadlock. + - "FATAL: $Coro::IDLE blocked itself" will now use Carp::confess to report a full stacktrace, which should help find out where the actual call is. - "atomically" destroy data in slf_destroy, because it is @@ -149,7 +187,7 @@ 6.01 Sun Jul 3 12:31:14 CEST 2011 - workarounds are good, but the test for whether pthreads are used - was not good. thisone should be better. + was not good. this one should be better. - check differently whether gcc generates cfi instructions itself. 6.0 Wed Jun 29 19:43:35 CEST 2011 @@ -157,7 +195,7 @@ destroyed and cleaned up automatically (e.g. async { schedule }). - implement a JIT compiler for part of the thread switch code, which gives a 50% speed improvement on threaded perls, and - about 4% on non-thraeded perls (so threaded perls now finally + about 4% on non-threaded perls (so threaded perls now finally reach about half the speed of non-threaded perls). - slightly modernise Coro::Intro, add section about rouse functions. - avoid DEFSV and ERRSV, giving another 10% improvement