1 Revision history for indirect
3 0.33 2014-09-29 20:20 UTC
4 + Fix : [RT #99083] : Breaks eval in an END block in Win32 pseudo-forks.
5 Thanks Graham Knop for reporting.
6 + Fix : Segfaults during global destruction of a thread or a
9 0.32 2014-09-21 20:15 UTC
10 + Add : Support for the PERL_OP_PARENT optional feature introduced in
12 + Fix : [RT #92806] : Tests that use run_perl() fail on Android
13 Thanks Brian Fraser for the patch.
14 + Fix : indirect constructs will no longer segfault while inside
15 the empty package on perl 5.8.x. This fix may also prevent
16 some segfaults during global destruction.
18 0.31 2013-09-05 16:45 UTC
19 + Fix : [RT #88428] : no indirect in eval can trigger for direct calls
21 Thanks Graham Knop for reporting.
22 + Tst : Author tests are no longer bundled with this distribution.
23 They are only made available to authors in the git repository.
25 0.30 2013-05-16 15:55 UTC
26 + Fix : [RT #83806] : false positives with Devel::Declare
27 [RT #83839] : false positive using ? : syntax
28 Thanks Andrew Main for the patch.
29 However, please note that the reason this patch seems to fix
30 thinks has not been explained.
31 + Fix : [RT #84649] : incorrect RT link in metadata
32 Thanks Karen Etheridge for reporting.
34 0.29 2013-03-05 01:30 UTC
35 + Fix : [RT #83659] : false positives
36 Proper method calls in string-like environments (like
37 "@{[ $x->new ]}" will no longer be reported as indirect.
38 This was a regression in 0.28.
39 Thanks Andrew Main for reporting.
40 + Fix : Broken linkage on Windows with gcc 3.4, which appears in
41 particular when using ActivePerl's default compiler suite.
42 For those setups, the indirect shared library will now be
43 linked against the perl dll directly (instead of the import
46 0.28 2013-02-26 17:05 UTC
47 + Fix : [RT #83450] : newlines confuse indirect
48 Perl sometimes resets the line buffer between the object and
49 the method name (e.g. for "sort Class\n->method" outside of
50 eval), and this could cause direct method calls to be reported
52 Thanks Gianni Ceccarelli for reporting.
53 + Fix : Check functions are now replaced and restored in a thread-safe
54 manner, either by using the wrap_op_checker() function from perl
55 when it is available (starting from perl 5.16) or by taking the
56 OP_REFCNT mutex on older perls.
58 0.27 2013-01-30 19:00 UTC
59 + Fix : [RT #82562] : indirect/Devel::CallParser interaction
60 indirect has been taught to play nicely with Devel::CallParser.
61 Thanks Andrew Main for the patch.
62 + Tst : Author tests overhaul.
64 0.26 2011-10-23 14:25 UTC
65 + Add : "no indirect 'global'" enables the pragma for the whole program,
66 except for lexical scopes that "use indirect" explicitely.
67 + Chg : Passing both the 'fatal' and 'hook' options to unimport() is
68 now considered an error, and will result in an exception.
69 unimport() used to consider only the first passed option of
70 those two, and silently ignored the other.
71 + Tst : Test failures of t/41-threads-teardown.t and t/50-external.t on
72 Cygwin should have been addressed.
73 + Tst : Threads tests will not fail anymore if resources constraints
74 prevent the system from creating all the required threads.
76 0.25 2011-08-24 15:40 UTC
77 + Fix : RT #69291 is now also fixed for perl 5.8. The pragma will no
78 longer vivify the "indirect" entry in the hints hash %^H on
80 + Tst : Attempt to make t/50-external.t pass on Cygwin.
82 0.24 2011-07-17 23:15 UTC
83 + Fix : [RT #64521] : "no indirect" leaking into eval.
84 This is currently only fixed for perl 5.10 (perl 5.12 and
85 higher were never affected). It was caused by a very stupid
86 mistake of mine that was introduced in indirect version 0.23.
87 Thanks Michael G Schwern for reporting.
88 + Fix : [RT #69291] : indirect.pm breaks %^H.
89 This was caused by the same mistake as for the previous bug,
90 and as such it is also only fixed for perl 5.10 (and never
92 Thanks Andrew Main for reporting.
93 + Doc : C++ compilers are officially NOT supported.
95 0.23 2010-10-03 00:15 UTC
96 + Fix : Some indirect constructs could be incorrectly reported when
97 several modules were used in the same scope. This caused
98 t/30-scope.t to fail randomly.
99 + Tst : Threads tests are now only run on perl 5.13.4 and higher.
100 They could segfault randomly because of what seems to be an
101 internal bug of Perl, which has been addressed in 5.13.4.
102 There is also an environment variable that allows you to
103 forcefully run those tests, but it should be set only for
104 author testing and not for end users.
106 0.22 2010-08-16 16:00 UTC
107 + Add : Indirect constructs are now reported for code interpolated
108 in quote-like environments, like "${\( ... )}", "@{[ ... ]}",
109 s/pattern/ ... /e, qr/(?{ ... })/ or qr/(??{ ... })/.
110 + Add : You can now make the pragma lethal by passing anything matching
111 /^:?fatal$/i to import(), including "FATAL" and ":Fatal".
112 + Fix : [RT #60378] : segmentation fault on indirect_ck_method.
113 This caused constructs like "@{[ $obj->$meth ]}" to segfault
114 when $meth was a lexical.
115 Thanks Tokuhiro Matsuno for reporting.
117 0.21 2010-05-31 23:10 UTC
118 + Chg : perl 5.8.1 is now required (instead of 5.8.0).
119 + Fix : [RT #57699] : indirect fail with 64-bit int on 5.13.1.
120 It was actually a problem with thread destructors segfaulting
121 because they weren't called at the right time anymore.
122 Thanks Andrew Main for reporting.
123 + Tst : A few more regression tests about the scope leak bug.
125 0.20 2010-04-18 21:25 UTC
126 + Fix : [RT #50570] : "indirect" leaking into LWP.
127 Thanks Andrew Main for reporting.
128 More generally, the require propagation workaround on 5.8-5.10.0
129 has been overhauled, and other scope leaks should be fixed.
130 + Fix : Test failures with 5.12 on Windows where Strawberry Perl crashes
131 because the SystemRoot environment variable is missing.
132 + Fix : Work around Kwalitee test misfailures.
134 0.19 2009-08-28 18:40 UTC
135 + Add : The new constant I_FORKSAFE can be tested to know whether the
136 module will behave nicely when fork()ing. It's currently always
137 true except on Windows where you need perl 5.10.1 for it to be
139 + Fix : I_THREADSAFE and I_FORKSAFE ought to be true when
140 PERL_INDIRECT_PM_DISABLE is set.
141 + Fix : The pragma could still leak if you passed to the "hook" option
142 a reference to the same (named) subroutine from two different
143 require scopes. The fix also provides a better solution for
145 + Fix : Segfaults when indirect is loaded for the first time from
147 + Fix : Leaks of memory associated with the root interpreter.
148 + Opt : Less memory will be used for non-threaded perls version 5.10.0
149 and below, and for threaded perls from version 5.10.1.
151 0.18 2009-08-23 16:15 UTC
152 + Add : When set, the PERL_INDIRECT_PM_DISABLE environment variable
153 disables the pragma globally.
155 0.17 2009-07-16 12:10 UTC
156 + Fix : [RT #47902] : "no indirect" leaking again.
157 This actually turned out to be a bug in how the hook coderefs
158 were stored in the hints hash.
159 Thanks Andrew Main for reporting once again.
160 + Fix : t/80-regressions.t failing on Windows.
161 + Tst : Yet more cleanups.
163 0.16 2009-07-14 16:50 UTC
164 + Add : Indirect calls on blocks are now reported. For those, '{' is
165 passed to the hook as the object description.
166 + Add : The new indirect::msg() function publicizes the default
167 warning/exception message.
168 + Fix : [RT #47866] : Segfault with UTF-8 regexps.
169 Thanks Andrew Main for reporting.
172 0.15 2009-07-08 22:55 UTC
173 + Fix : Invalid constructs with the same method and package name were
175 + Fix : The error line number used to point to the end of the expression
176 instead of its beginning.
178 0.14 2009-06-04 21:55 UTC
179 + Fix : Prevent bogus invalid syntaxes caused by reallocated memory
180 chunks. Thanks Andrew Main for reporting with a reproducible
183 0.13 2009-05-24 18:50 UTC
184 + Add : The hook now receives the file name and the line where the
185 error happened in respectively $_[2] and $_[3].
186 + Fix : Pass mortalized copies of the method name and object to the
187 hook. This seems to fix some rare crashes.
188 + Fix : Work around a bug in perl 5.10.0 and lower. Thanks Andrew Main
189 for teaching me about this issue.
190 + Fix : Report the correct file in error messages (a regression from the
193 0.12 2009-05-03 14:30 UTC
194 + Add : You can specify the action to execute for each indirect
195 construct encountered with the new "hook => $coderef" unimport()
197 + Chg : A ptable is now used internally for the op => position mapping.
198 + Fix : The pragma is now built with thread safety features enabled
199 whenever possible (a notable exception is perl 5.8.x on Win32,
200 as something seems wrong with its context handling).
201 The new indirect::I_THREADSAFE() constant reflects this.
202 + Fix : A negation precedence nit in indirect_ck_entersub().
203 + Tst : "use/no indirect" while parsing an indirect construct.
204 + Tst : Thread safety.
206 0.11 2009-02-08 18:35 UTC
207 + Fix : Potential collisions by hashing pointers with a wrong format.
208 + Upd : Resources in META.yml.
210 0.10 2009-01-17 12:40 UTC
211 Re-release 0.09_01 as stable.
213 0.09_01 2008-12-08 17:55 UTC
214 + Fix : Invalid memory read with "${\(new Foo)}" constructs. The
215 corresponding test is turned back on.
216 + Tst : Refinements in t/30-scope.t
218 0.09 2008-12-05 20:35 UTC
219 + Add : Support for perl 5.8.
220 + Tst : Skip a test in t/10-good.t that randomly segfaults for (I guess)
221 systems stricter than linux in the way they manage their memory.
223 0.08 2008-10-22 14:45 UTC
224 + Fix : A rare edge case for package whose names are prefix of 'main'.
225 + Tst : Test $$ as variable and state variables.
227 0.07_03 2008-10-17 20:10 UTC
228 + Add : Support and tests for variables with spaces after the sigil.
229 + Upd : META.yml spec updated to 1.4.
231 0.07_02 2008-10-15 21:10 UTC
232 + Add : Support and tests for package variables.
233 + Tst : Coverage improved by removing dead code.
235 0.07_01 2008-10-15 16:00 UTC
236 + Fix : [RT #40055] : Not handling RV2SV => GV(SV) correctly, which
237 could cause 'no indirect; print' segfaults. Thanks Goro Fuji for
240 0.06 2008-10-11 16:45 UTC
242 + Tst : Test "no indirect 'anything'", "foo Class->bar", and indirect
243 uses of exec() and system().
245 0.05 2008-10-02 14:40 UTC
246 + Chg : Now the old check function is always called before storing an op
248 + Fix : Misc code and docs refinements.
250 0.04 2008-08-30 19:00 UTC
251 + Fix : Clean up the op->src hash when we're done with an entersub.
252 + Tst : No longer fork for testing. IPC::Cmd isn't required anymore.
254 0.03 2008-08-12 15:25 UTC
255 This release is kindly supported by Copenhagen Hotel Centrum WiFi.
256 + Fix : Tests used not to pass PERL5OPTS to their kids. This lead to
257 failures under CPAN. I think.
260 0.02 2008-08-11 15:55 UTC
261 + Fix : Some building failures with old gcc versions that didn't seem to
262 like the ((hint == 2) ? croak : warn)(msg) construct. I think.
263 + Rem : Unused cruft from a previous implementation.
264 + Tst : Fail more gracefully when we can't capture buffers or when the
265 child returned an error.
267 0.01 2008-08-10 20:40 UTC
268 First version, released on an unsuspecting world.