ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/Coro/Changes
(Generate patch)

Comparing Coro/Changes (file contents):
Revision 1.554 by root, Thu Mar 31 04:24:19 2011 UTC vs.
Revision 1.626 by root, Tue Mar 4 06:13:23 2014 UTC

1Revision history for Perl extension Coro. 1Revision history for Perl extension Coro.
2 2
3TODO: should explore PerlIO::coroaio (perl leaks like hell). 3TODO: should explore PerlIO::coroaio (perl leaks like hell).
4TODO: channel->maxsize(newsize)? 4TODO: channel->maxsize(newsize)?
5TODO: http://www.microsoft.com/msj/archive/s2ce.aspx 5TODO: __GCC_HAVE_DWARF2_CFI_ASM
6TODO: swap_sv, maybe add scope_swap_sv, andallow to unswap by repeat?
7TODO: swap_sv not undone in pool
6 8
96.36 Tue Mar 4 07:11:59 CET 2014
10 - semaphores would not clear the destroy hook when interrupted
11 by ->throw, causing segfaults or worse.
12 - ->throw on a thread waiting for a semaphore did not acquire
13 the semaphore, but also didn't wake up other waiters,
14 possibly causing a deadlock.
15 - "FATAL: $Coro::IDLE blocked itself" will now use Carp::confess
16 to report a full stacktrace, which should help find out
17 where the actual call is.
18 - "atomically" destroy data in slf_destroy, because it is
19 the right thing to do, just in case.
20 - disable ecb.h libm dependency, to compile on ancient systems
21 or under adverse conditions.
22
236.33 Mon Nov 18 11:26:27 CET 2013
24 - do not crash when freeing padlists with holes (in 5.18).
25 - tentative SVt_BIND 5.19 port/fix.
26
276.32 Tue Nov 5 15:35:35 CET 2013
28 - use a new algorithm to derive padlists for perl 5.18. The old
29 one could lead to 0-pointer accesses inside perl (reported
30 by Darin McBride).
31
326.31 Thu May 9 07:39:48 CEST 2013
33 - Coro::AIO requests would crash if the thread was ready'd
34 while the request was ongoing.
35
366.29 Wed May 8 02:55:18 CEST 2013
37 - when an on_destroy handler destructs the coro currently being
38 destructed a perl scalar could be accessed after being freed,
39 likely causing a crash.
40
416.28 Wed Mar 6 06:58:02 CET 2013
42 - clean remnants of existing __DIE__ and __WARN__ handlers so
43 they lose their magic and will not cause segfaults later
44 (testcase by Andrey Sagulin).
45 - improved Coro::State documentation a bit.
46 - Coro::Debug::command now flushes the output.
47 - add hack detection code for x32 abi, because the braindead slugs
48 who designed that made it look exactly like x86_64 without
49 providing proper compile time symbols to test for it. as a result,
50 this detection cannot work reliably.
51 - valgrind stack registering was broken.
52 - do not rely on Time::HiRes anymore in Coro::Debug.
53
546.23 Fri Dec 7 23:36:37 CET 2012
55 - use experimental fiber implementation on native windows
56 perls.
57 - use sizeof (void *) as multiplication factor for stack sizes,
58 to accomodate the totally braindamaged microsoft 64 bit "os".
59 - changed verifier host from win2k-ap510-32 to win7-sp516-32/64.
60 activeperl 5.16 crashes when PerlIO_define_layer is called due
61 to some bug in the perl dll, strawberry perl at least passes
62 the testsuite.
63 - implement Coro::Handle->peeraddr/host/port, for slightly
64 improved compatibility with LWP.
65 - implement 5.17 compatibility by almost blindly applying a
66 good-looking patch by Father Chrysostomos.
67 - move stack management functions into libcoro 3.
68 - libcoro version 3 "released".
69 - support magic values as timed_io_once args.
70 - recommend AnyEvent 7+ or EV 4+, also require EV
71 version 4 or newer for Coro::EV.
72
736.10 Tue Oct 9 01:14:27 CEST 2012
74 - updated ecb.h, it had a typo that caused it to not compile on many
75 big endian systems (reported by many people).
76 - disable memory fences in ecb.h to improve portability.
77
786.09 Sat Oct 6 23:25:02 CEST 2012
79 - Coro::EV I/O watchers were not interruptible by exceptions
80 (Coro::State::throw) (testcase by sten).
81 - ->throw now puts threads into the ready queue, as this seems to
82 be expected by existing code, and code that doesn't cope with spurious
83 wakeups needs fixing anyway.
84 - use fd -1 in mmap.
85 - cast I32 to int in error message printf.
86 - warn about broken so-called "hardened" kernels.
87
886.08 Fri Apr 13 12:05:47 CEST 2012
89 - be more aggressive about exiting like perl does - formerly,
90 exiting from the non-main thread would not execute END blocks.
91
926.07 Fri Nov 11 21:21:48 CET 2011
93 - work around a bug in PerlIO (setting $SIG{__WARN__} to a PVCV).
94 - update ecb.h.
95
966.06 Mon Aug 8 23:59:48 CEST 2011
97 - cygwin unfortunately patches the stack at runtime, so we use the pthreads
98 backend, which is an order of magnitude slower. unfortunately, cygwins
99 pthread implementation isn't very complete either, so allocate the stack
100 twice just to be sure.
101 (note: cygwin also enables mymalloc, which is NOT THREADSAFE ON WINDOWS,
102 in its ithreaded perl - best recompile cygwin and use the 'w'indows
103 backend for much better performance. also disable ithreads for
104 even better performance...).
105
1066.05 Thu Aug 4 21:36:36 CEST 2011
107 - blush, condvar values would not be propagated from send to recv anymore
108 (reported by Chip Salzenberg).
109 - use exponential increase for the readline buffer length in
110 Coro::Handle. also reduce initial allocation to 1020 from 4096 bytes.
111
1126.04 Wed Aug 3 17:15:45 CEST 2011
113 - use even more efficient and more compatible condvars for
114 compatibility to AnyEvent 6.x :)
115 - more inconsequential ecb.h updates.
116
1176.03 Wed Aug 3 11:41:30 CEST 2011
118 - change how Coro patches AnyEvent condvars for compatibility to
119 AnyEvent 6.x.
120 - update ecb.h, to no longer include <pthread.h> in case WinNT.h
121 hasn't been included.
122
1236.02 Wed Jul 13 04:35:19 CEST 2011
124 - "improve portability to Gentoo" - gentoo manages to put perl variables
125 in memory areas that are farther than 2gb apart, which the jit couldn't
126 handle and barfed. now it's just a bit slower on gentoo and similar
127 systems.
128
1296.01 Sun Jul 3 12:31:14 CEST 2011
130 - workarounds are good, but the test for whether pthreads are used
131 was not good. thisone should be better.
132 - check differently whether gcc generates cfi instructions itself.
133
1346.0 Wed Jun 29 19:43:35 CEST 2011
135 - INCOMPATIBLE CHANGE: unreferenced coro objects will now be
136 destroyed and cleaned up automatically (e.g. async { schedule }).
137 - implement a JIT compiler for part of the thread switch code,
138 which gives a 50% speed improvement on threaded perls, and
139 about 4% on non-thraeded perls (so threaded perls now finally
140 reach about half the speed of non-threaded perls).
141 - slightly modernise Coro::Intro, add section about rouse functions.
142 - avoid DEFSV and ERRSV, giving another 10% improvement
143 in thread switching.
144 - Coro::State->is_destroyed is now called is_zombie.
145 - implement a Coro->safe_cancel method that might fail, but
146 cancels in a "safer" way if it succeeds.
147 - add preliminary support for DEBUGGING perls.
148 - get rid of two hash-accesses when initialising a new Coro - this
149 speeds up coro creation by almost a factor of two.
150 - croak when a coro that is being cancelled tries to block
151 (e.g. while executing a guard block), instead of crashing or
152 deadlocking.
153 - use a more robust and also faster method to identify Coro::State
154 objects - speeds up everything a bit.
155 - implement Coro->cancel in XS for a 20% speed improvement, and to
156 be able to implement mutual cancellation.
157 - speed up context switches by a percent or two by more efficiently
158 allocating context stack entries.
159 - implement Coro->join and Coro->on_destroy in XS for a speedup and
160 a reduction in memory use.
161 - cancelling a coro while it itself is cancelling another coro is
162 now supported and working, instead of triggering an assertion.
7 - be a bit more crash-resistant when calling (buggy) on_destroy 163 - be a bit more crash-resistant when calling (buggy) on_destroy
8 callbacks (bets effort). 164 callbacks (best effort).
165 - move on_destroy into the slf_frame, to allow extension slf
166 functions to have destructors.
167 - get rid if coro refcounting - simply crash in other interpreter
168 threads by nulling the pointers on clone.
169 - simplify warn/die hook handling when loading Coro - the convoluted
170 logic seems to be no longer neccessary.
171 - use libecb instead of our own home-grown gcc hacks.
172 - document alternatives to Coro::LWP. Please use them :)
173 - work around another mindless idiotic NEEDLESS bug in openbsd/mirbsds
174 sigaltstack. Really. wine suffers from it, erlang suffers from it,
175 and it's known since at least 2006.
9 176
105.372 Wed Feb 23 06:14:30 CET 2011 1775.372 Wed Feb 23 06:14:30 CET 2011
11 - apparently mingw doesn't provide a working gettimeofday, try to 178 - apparently mingw doesn't provide a working gettimeofday, try to
12 work around that by relying on Time::HiRes (indirectly brought to 179 work around that by relying on Time::HiRes (indirectly brought to
13 my attention by Max Maischein). 180 my attention by Max Maischein).
… …
331 - ->throw is now supported on Coro::State objects. 498 - ->throw is now supported on Coro::State objects.
332 - clean up cctx creation code a bit. 499 - clean up cctx creation code a bit.
333 - entersub is actually a UNOP, not a LOGOP (not a bugfix). 500 - entersub is actually a UNOP, not a LOGOP (not a bugfix).
334 501
3354.9 Sat Nov 8 17:45:27 CET 2008 5024.9 Sat Nov 8 17:45:27 CET 2008
336 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there). 503 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there).
337 - (libcoro) no longer leak threads in the experimental pthread backend, 504 - (libcoro) no longer leak threads in the experimental pthread backend,
338 also speed it up considerably. 505 also speed it up considerably.
339 - (libcoro) do not rely on makecontext passing void *'s unscathed. 506 - (libcoro) do not rely on makecontext passing void *'s unscathed.
340 - fix compiletime dependencies on libcoro in the Makefile. 507 - fix compiletime dependencies on libcoro in the Makefile.
341 - cctx_count wasn't always updated properly. 508 - cctx_count wasn't always updated properly.
… …
403 - remove debugging code related to MgPV_nolen_const, also try to 570 - remove debugging code related to MgPV_nolen_const, also try to
404 make it compile with perl 5.8.6 (yes, apple apparently loves 571 make it compile with perl 5.8.6 (yes, apple apparently loves
405 outdated software). Reported by John S. 572 outdated software). Reported by John S.
406 573
4074.744 Tue Jul 8 22:06:35 CEST 2008 5744.744 Tue Jul 8 22:06:35 CEST 2008
408 - correctly provide default DIE/WARN handlers as documented. 575 - correctly provide default DIE/WARN handlers as documented.
409 - also overwrite PL_vtbl_sigelem.svt_clear, even though current 576 - also overwrite PL_vtbl_sigelem.svt_clear, even though current
410 implementations inside perl work fine for us. 577 implementations inside perl work fine for us.
411 578
4124.743 Mon Jun 16 00:21:57 CEST 2008 5794.743 Mon Jun 16 00:21:57 CEST 2008
413 - when using Coro::EV without running EV::loop it could 580 - when using Coro::EV without running EV::loop it could
… …
435 - sprinkle "no warnings" freely over everything, also suppress 602 - sprinkle "no warnings" freely over everything, also suppress
436 warnings for some other modules. 603 warnings for some other modules.
437 - fix typo in WSAEWOULDBLOCK. 604 - fix typo in WSAEWOULDBLOCK.
438 605
4394.72 Sun May 25 05:14:36 CEST 2008 6064.72 Sun May 25 05:14:36 CEST 2008
440 - tweak META.yaml a bit, unfortunately, there is no documented way 607 - tweak META.yaml a bit, unfortunately, there is no documented way
441 to have optional dependencies with CPAN. doh :( 608 to have optional dependencies with CPAN. doh :(
442 - avoid running some tests on windows because they would fail due to 609 - avoid running some tests on windows because they would fail due to
443 perl bug (broken fork, broken pipes...). 610 perl bug (broken fork, broken pipes...).
444 - work around perl on windows bugs where perl returns undocumented 611 - work around perl on windows bugs where perl returns undocumented
445 error codes for sysread, syswrite etc. by taking advantage 612 error codes for sysread, syswrite etc. by taking advantage
… …
449 - use unix domain sockets in testsuite to work around 616 - use unix domain sockets in testsuite to work around
450 common perl implementation bugs on widows (they are emulated by 617 common perl implementation bugs on widows (they are emulated by
451 tcp sockets on windows. ugh.) 618 tcp sockets on windows. ugh.)
452 619
4534.71 Sat May 24 20:01:27 CEST 2008 6204.71 Sat May 24 20:01:27 CEST 2008
454 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()"). 621 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()").
455 - take advantage of async name resolution of AnyEvent::Util. 622 - take advantage of async name resolution of AnyEvent::Util.
456 - work around brutal inet_aton override in Coro::LWP. 623 - work around brutal inet_aton override in Coro::LWP.
457 - take advantage of the readyhook in Coro::EV, for smoother 624 - take advantage of the readyhook in Coro::EV, for smoother
458 scheduling. 625 scheduling.
459 626
4604.7 Sun May 11 00:32:19 CEST 2008 6274.7 Sun May 11 00:32:19 CEST 2008
461 - completely reworked the Coro manpage. 628 - completely reworked the Coro manpage.
462 - added Coro::AnyEvent, generic event loop integration. 629 - added Coro::AnyEvent, generic event loop integration.
463 - implement cancel, ready and kill commands in Coro::Debug. 630 - implement cancel, ready and kill commands in Coro::Debug.
464 - document find_coro in Coro::Debug. 631 - document find_coro in Coro::Debug.
465 - incompatible change: rename has_stack to has_cctx. 632 - incompatible change: rename has_stack to has_cctx.
466 - Coro::AIO and Coro::BDB no longer force event model detection, 633 - Coro::AIO and Coro::BDB no longer force event model detection,
467 use AnyEvent::AIO and AnyEvent::BDB. 634 use AnyEvent::AIO and AnyEvent::BDB.
… …
591 758
5924.11 Thu Oct 11 02:40:24 CEST 2007 7594.11 Thu Oct 11 02:40:24 CEST 2007
593 - port to threaded perls. 760 - port to threaded perls.
594 761
5954.1 Thu Oct 11 02:38:16 CEST 2007 7624.1 Thu Oct 11 02:38:16 CEST 2007
596 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now 763 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now
597 be local to each coro (see Coro::State). 764 be local to each coro (see Coro::State).
598 - incompatible change: for very deep reasons, cede and cede_notself 765 - incompatible change: for very deep reasons, cede and cede_notself
599 cannot return anything, so nothing will be returned. 766 cannot return anything, so nothing will be returned.
600 - possibly bring back 5.10 compatibility (untested). 767 - possibly bring back 5.10 compatibility (untested).
601 - work around stupid (and wrong) warning on 5.10 :(. 768 - work around stupid (and wrong) warning on 5.10 :(.
… …
673 but stateful semantics. 840 but stateful semantics.
674 - fixed a lot of typos in Coro.pm (patch submitted by David 841 - fixed a lot of typos in Coro.pm (patch submitted by David
675 Steinbrunner, which applied flawlessly). 842 Steinbrunner, which applied flawlessly).
676 843
6773.6 Sat Apr 14 17:13:31 CEST 2007 8443.6 Sat Apr 14 17:13:31 CEST 2007
678 - added some bugfixes to get eg/myhttpd working again. 845 - added some bugfixes to get eg/myhttpd working again.
679 - added Coro::Storable for often-cede'ing freeze/thaw. 846 - added Coro::Storable for often-cede'ing freeze/thaw.
680 - try to do a clean exit when a coroutine calls exit 847 - try to do a clean exit when a coroutine calls exit
681 (EXPERIMENTAL). 848 (EXPERIMENTAL).
682 - got rid of indirect call through _coro_init. 849 - got rid of indirect call through _coro_init.
683 - updated the partly antique examples in eg/ to 850 - updated the partly antique examples in eg/ to

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines