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

Comparing Coro/Changes (file contents):
Revision 1.565 by root, Fri May 13 19:39:26 2011 UTC vs.
Revision 1.597 by root, Mon Oct 8 23:13:36 2012 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: http://www.microsoft.com/msj/archive/s2ce.aspx
6TODO: __GCC_HAVE_DWARF2_CFI_ASM
6 7
7TODO: tutorial 86.10
9 - updated
10
116.09 Sat Oct 6 23:25:02 CEST 2012
12 - Coro::EV I/O watchers were not interruptible by exceptions
13 (Coro::State::throw) (testcase by sten).
14 - ->throw now puts threads into the ready queue, as this seems to
15 be expected by existing code, and code that doesn't cope with spurious
16 wakeups needs fixing anyway.
17 - use fd -1 in mmap.
18 - cast I32 to int in error message printf.
19 - warn about broken so-called "hardened" kernels.
20
216.08 Fri Apr 13 12:05:47 CEST 2012
22 - be more aggressive about exiting like perl does - formerly,
23 exiting from the non-main thread would not execute END blocks.
24
256.07 Fri Nov 11 21:21:48 CET 2011
26 - work around a bug in PerlIO (setting $SIG{__WARN__} to a PVCV).
27 - update ecb.h.
28
296.06 Mon Aug 8 23:59:48 CEST 2011
30 - cygwin unfortunately patches the stack at runtime, so we use the pthreads
31 backend, which is an order of magnitude slower. unfortunately, cygwins
32 pthread implementation isn't very complete either, so allocate the stack
33 twice just to be sure.
34 (note: cygwin also enables mymalloc, which is NOT THREADSAFE ON WINDOWS,
35 in its ithreaded perl - best recompile cygwin and use the 'w'indows
36 backend for much better performance. also disable ithreads for
37 even better performance...).
38
396.05 Thu Aug 4 21:36:36 CEST 2011
40 - blush, condvar values would not be propagated from send to recv anymore
41 (reported by Chip Salzenberg).
42 - use exponential increase for the readline buffer length in
43 Coro::Handle. also reduce initial allocation to 1020 from 4096 bytes.
44
456.04 Wed Aug 3 17:15:45 CEST 2011
46 - use even more efficient and more compatible condvars for
47 compatibility to AnyEvent 6.x :)
48 - more inconsequential ecb.h updates.
49
506.03 Wed Aug 3 11:41:30 CEST 2011
51 - change how Coro patches AnyEvent condvars for compatibility to
52 AnyEvent 6.x.
53 - update ecb.h, to no longer include <pthread.h> in case WinNT.h
54 hasn't been included.
55
566.02 Wed Jul 13 04:35:19 CEST 2011
57 - "improve portability to Gentoo" - gentoo manages to put perl variables
58 in memory areas that are farther than 2gb apart, which the jit couldn't
59 handle and barfed. now it's just a bit slower on gentoo and similar
60 systems.
61
626.01 Sun Jul 3 12:31:14 CEST 2011
63 - workarounds are good, but the test for whether pthreads are used
64 was not good. thisone should be better.
65 - check differently whether gcc generates cfi instructions itself.
66
676.0 Wed Jun 29 19:43:35 CEST 2011
8 - INCOMPATIBLE CHANGE: unreferenced coro objects will now be 68 - INCOMPATIBLE CHANGE: unreferenced coro objects will now be
9 destroyed and cleaned up automatically (e.g. async { schedule }). 69 destroyed and cleaned up automatically (e.g. async { schedule }).
70 - implement a JIT compiler for part of the thread switch code,
71 which gives a 50% speed improvement on threaded perls, and
72 about 4% on non-thraeded perls (so threaded perls now finally
73 reach about half the speed of non-threaded perls).
74 - slightly modernise Coro::Intro, add section about rouse functions.
75 - avoid DEFSV and ERRSV, giving another 10% improvement
76 in thread switching.
10 - Coro::State->is_destroyed is now called is_zombie. 77 - Coro::State->is_destroyed is now called is_zombie.
11 - implement a Coro->safe_cancel method that might fail, but 78 - implement a Coro->safe_cancel method that might fail, but
12 cancels in a "safer" way if it succeeds. 79 cancels in a "safer" way if it succeeds.
80 - add preliminary support for DEBUGGING perls.
81 - get rid of two hash-accesses when initialising a new Coro - this
82 speeds up coro creation by almost a factor of two.
13 - croak when a coro that is being cancelled tries to block 83 - croak when a coro that is being cancelled tries to block
14 (e.g. while executing a guard block), instead of crashing or 84 (e.g. while executing a guard block), instead of crashing or
15 deadlocking. 85 deadlocking.
16 - use a more robust and also faster method to identify Coro::State 86 - use a more robust and also faster method to identify Coro::State
17 objects - speeds up everything a bit. 87 objects - speeds up everything a bit.
… …
27 callbacks (best effort). 97 callbacks (best effort).
28 - move on_destroy into the slf_frame, to allow extension slf 98 - move on_destroy into the slf_frame, to allow extension slf
29 functions to have destructors. 99 functions to have destructors.
30 - get rid if coro refcounting - simply crash in other interpreter 100 - get rid if coro refcounting - simply crash in other interpreter
31 threads by nulling the pointers on clone. 101 threads by nulling the pointers on clone.
102 - simplify warn/die hook handling when loading Coro - the convoluted
103 logic seems to be no longer neccessary.
104 - use libecb instead of our own home-grown gcc hacks.
105 - document alternatives to Coro::LWP. Please use them :)
106 - work around another mindless idiotic NEEDLESS bug in openbsd/mirbsds
107 sigaltstack. Really. wine suffers from it, erlang suffers from it,
108 and it's known since at least 2006.
32 109
335.372 Wed Feb 23 06:14:30 CET 2011 1105.372 Wed Feb 23 06:14:30 CET 2011
34 - apparently mingw doesn't provide a working gettimeofday, try to 111 - apparently mingw doesn't provide a working gettimeofday, try to
35 work around that by relying on Time::HiRes (indirectly brought to 112 work around that by relying on Time::HiRes (indirectly brought to
36 my attention by Max Maischein). 113 my attention by Max Maischein).

Diff Legend

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