1 Revision history for indirect
3 0.36 2015-07-17 22:15 UTC
4 + Fix : [RT #104312] : fatal hides perl errors in modules
5 no indirect 'fatal' will no longer hide compilation errors
6 occurring before indirect constructs.
7 Thanks Lukas Mai for reporting.
9 0.35 2015-04-06 22:20 UTC
10 + Fix : The module could end being disabled in one thread if it was
11 first loaded in another thread and that thread was immediately
12 terminated. This is now fixed and should address test failures
13 of t//09-load-threads.t and t/42-threads-global.t.
15 0.34 2015-04-02 19:50 UTC
16 + Chg : The new environment variable to enable thread tests on older
17 perls is PERL_FORCE_TEST_THREADS. Note that this variable
18 should only be turned on by authors.
19 + Fix : [RT #100068] : add link to historical tchrist post
20 The link has been added to the documentation. Thanks Olivier
21 Mengué for reporting.
22 + Fix : Segfaults when the module is loaded by several threads (or
23 Windows emulated processes) ran in parallel.
24 + Fix : Update the Windows ActivePerl + gcc 3.4 workaround for
25 ExtUtils::MakeMaker 7.04. Thanks Christian Walde for reporting
26 and feedback on this issue.
27 + Fix : Be really compatible with the optional OP_PARENT feature.
28 + Tst : $ENV{$Config{ldlibpthname}} is now preserved on all platforms,
29 which will address failures of t/41-threads-teardown.t and
30 t/50-external.t with unusual compilers (like icc) that link all
31 their compiled objects to their own libraries.
33 0.33 2014-09-29 20:20 UTC
34 + Fix : [RT #99083] : Breaks eval in an END block in Win32 pseudo-forks.
35 Thanks Graham Knop for reporting.
36 + Fix : Segfaults during global destruction of a thread or a
39 0.32 2014-09-21 20:15 UTC
40 + Add : Support for the PERL_OP_PARENT optional feature introduced in
42 + Fix : [RT #92806] : Tests that use run_perl() fail on Android
43 Thanks Brian Fraser for the patch.
44 + Fix : indirect constructs will no longer segfault while inside
45 the empty package on perl 5.8.x. This fix may also prevent
46 some segfaults during global destruction.
48 0.31 2013-09-05 16:45 UTC
49 + Fix : [RT #88428] : no indirect in eval can trigger for direct calls
51 Thanks Graham Knop for reporting.
52 + Tst : Author tests are no longer bundled with this distribution.
53 They are only made available to authors in the git repository.
55 0.30 2013-05-16 15:55 UTC
56 + Fix : [RT #83806] : false positives with Devel::Declare
57 [RT #83839] : false positive using ? : syntax
58 Thanks Andrew Main for the patch.
59 However, please note that the reason this patch seems to fix
60 thinks has not been explained.
61 + Fix : [RT #84649] : incorrect RT link in metadata
62 Thanks Karen Etheridge for reporting.
64 0.29 2013-03-05 01:30 UTC
65 + Fix : [RT #83659] : false positives
66 Proper method calls in string-like environments (like
67 "@{[ $x->new ]}" will no longer be reported as indirect.
68 This was a regression in 0.28.
69 Thanks Andrew Main for reporting.
70 + Fix : Broken linkage on Windows with gcc 3.4, which appears in
71 particular when using ActivePerl's default compiler suite.
72 For those setups, the indirect shared library will now be
73 linked against the perl dll directly (instead of the import
76 0.28 2013-02-26 17:05 UTC
77 + Fix : [RT #83450] : newlines confuse indirect
78 Perl sometimes resets the line buffer between the object and
79 the method name (e.g. for "sort Class\n->method" outside of
80 eval), and this could cause direct method calls to be reported
82 Thanks Gianni Ceccarelli for reporting.
83 + Fix : Check functions are now replaced and restored in a thread-safe
84 manner, either by using the wrap_op_checker() function from perl
85 when it is available (starting from perl 5.16) or by taking the
86 OP_REFCNT mutex on older perls.
88 0.27 2013-01-30 19:00 UTC
89 + Fix : [RT #82562] : indirect/Devel::CallParser interaction
90 indirect has been taught to play nicely with Devel::CallParser.
91 Thanks Andrew Main for the patch.
92 + Tst : Author tests overhaul.
94 0.26 2011-10-23 14:25 UTC
95 + Add : "no indirect 'global'" enables the pragma for the whole program,
96 except for lexical scopes that "use indirect" explicitely.
97 + Chg : Passing both the 'fatal' and 'hook' options to unimport() is
98 now considered an error, and will result in an exception.
99 unimport() used to consider only the first passed option of
100 those two, and silently ignored the other.
101 + Tst : Test failures of t/41-threads-teardown.t and t/50-external.t on
102 Cygwin should have been addressed.
103 + Tst : Threads tests will not fail anymore if resources constraints
104 prevent the system from creating all the required threads.
106 0.25 2011-08-24 15:40 UTC
107 + Fix : RT #69291 is now also fixed for perl 5.8. The pragma will no
108 longer vivify the "indirect" entry in the hints hash %^H on
110 + Tst : Attempt to make t/50-external.t pass on Cygwin.
112 0.24 2011-07-17 23:15 UTC
113 + Fix : [RT #64521] : "no indirect" leaking into eval.
114 This is currently only fixed for perl 5.10 (perl 5.12 and
115 higher were never affected). It was caused by a very stupid
116 mistake of mine that was introduced in indirect version 0.23.
117 Thanks Michael G Schwern for reporting.
118 + Fix : [RT #69291] : indirect.pm breaks %^H.
119 This was caused by the same mistake as for the previous bug,
120 and as such it is also only fixed for perl 5.10 (and never
122 Thanks Andrew Main for reporting.
123 + Doc : C++ compilers are officially NOT supported.
125 0.23 2010-10-03 00:15 UTC
126 + Fix : Some indirect constructs could be incorrectly reported when
127 several modules were used in the same scope. This caused
128 t/30-scope.t to fail randomly.
129 + Tst : Threads tests are now only run on perl 5.13.4 and higher.
130 They could segfault randomly because of what seems to be an
131 internal bug of Perl, which has been addressed in 5.13.4.
132 There is also an environment variable that allows you to
133 forcefully run those tests, but it should be set only for
134 author testing and not for end users.
136 0.22 2010-08-16 16:00 UTC
137 + Add : Indirect constructs are now reported for code interpolated
138 in quote-like environments, like "${\( ... )}", "@{[ ... ]}",
139 s/pattern/ ... /e, qr/(?{ ... })/ or qr/(??{ ... })/.
140 + Add : You can now make the pragma lethal by passing anything matching
141 /^:?fatal$/i to import(), including "FATAL" and ":Fatal".
142 + Fix : [RT #60378] : segmentation fault on indirect_ck_method.
143 This caused constructs like "@{[ $obj->$meth ]}" to segfault
144 when $meth was a lexical.
145 Thanks Tokuhiro Matsuno for reporting.
147 0.21 2010-05-31 23:10 UTC
148 + Chg : perl 5.8.1 is now required (instead of 5.8.0).
149 + Fix : [RT #57699] : indirect fail with 64-bit int on 5.13.1.
150 It was actually a problem with thread destructors segfaulting
151 because they weren't called at the right time anymore.
152 Thanks Andrew Main for reporting.
153 + Tst : A few more regression tests about the scope leak bug.
155 0.20 2010-04-18 21:25 UTC
156 + Fix : [RT #50570] : "indirect" leaking into LWP.
157 Thanks Andrew Main for reporting.
158 More generally, the require propagation workaround on 5.8-5.10.0
159 has been overhauled, and other scope leaks should be fixed.
160 + Fix : Test failures with 5.12 on Windows where Strawberry Perl crashes
161 because the SystemRoot environment variable is missing.
162 + Fix : Work around Kwalitee test misfailures.
164 0.19 2009-08-28 18:40 UTC
165 + Add : The new constant I_FORKSAFE can be tested to know whether the
166 module will behave nicely when fork()ing. It's currently always
167 true except on Windows where you need perl 5.10.1 for it to be
169 + Fix : I_THREADSAFE and I_FORKSAFE ought to be true when
170 PERL_INDIRECT_PM_DISABLE is set.
171 + Fix : The pragma could still leak if you passed to the "hook" option
172 a reference to the same (named) subroutine from two different
173 require scopes. The fix also provides a better solution for
175 + Fix : Segfaults when indirect is loaded for the first time from
177 + Fix : Leaks of memory associated with the root interpreter.
178 + Opt : Less memory will be used for non-threaded perls version 5.10.0
179 and below, and for threaded perls from version 5.10.1.
181 0.18 2009-08-23 16:15 UTC
182 + Add : When set, the PERL_INDIRECT_PM_DISABLE environment variable
183 disables the pragma globally.
185 0.17 2009-07-16 12:10 UTC
186 + Fix : [RT #47902] : "no indirect" leaking again.
187 This actually turned out to be a bug in how the hook coderefs
188 were stored in the hints hash.
189 Thanks Andrew Main for reporting once again.
190 + Fix : t/80-regressions.t failing on Windows.
191 + Tst : Yet more cleanups.
193 0.16 2009-07-14 16:50 UTC
194 + Add : Indirect calls on blocks are now reported. For those, '{' is
195 passed to the hook as the object description.
196 + Add : The new indirect::msg() function publicizes the default
197 warning/exception message.
198 + Fix : [RT #47866] : Segfault with UTF-8 regexps.
199 Thanks Andrew Main for reporting.
202 0.15 2009-07-08 22:55 UTC
203 + Fix : Invalid constructs with the same method and package name were
205 + Fix : The error line number used to point to the end of the expression
206 instead of its beginning.
208 0.14 2009-06-04 21:55 UTC
209 + Fix : Prevent bogus invalid syntaxes caused by reallocated memory
210 chunks. Thanks Andrew Main for reporting with a reproducible
213 0.13 2009-05-24 18:50 UTC
214 + Add : The hook now receives the file name and the line where the
215 error happened in respectively $_[2] and $_[3].
216 + Fix : Pass mortalized copies of the method name and object to the
217 hook. This seems to fix some rare crashes.
218 + Fix : Work around a bug in perl 5.10.0 and lower. Thanks Andrew Main
219 for teaching me about this issue.
220 + Fix : Report the correct file in error messages (a regression from the
223 0.12 2009-05-03 14:30 UTC
224 + Add : You can specify the action to execute for each indirect
225 construct encountered with the new "hook => $coderef" unimport()
227 + Chg : A ptable is now used internally for the op => position mapping.
228 + Fix : The pragma is now built with thread safety features enabled
229 whenever possible (a notable exception is perl 5.8.x on Win32,
230 as something seems wrong with its context handling).
231 The new indirect::I_THREADSAFE() constant reflects this.
232 + Fix : A negation precedence nit in indirect_ck_entersub().
233 + Tst : "use/no indirect" while parsing an indirect construct.
234 + Tst : Thread safety.
236 0.11 2009-02-08 18:35 UTC
237 + Fix : Potential collisions by hashing pointers with a wrong format.
238 + Upd : Resources in META.yml.
240 0.10 2009-01-17 12:40 UTC
241 Re-release 0.09_01 as stable.
243 0.09_01 2008-12-08 17:55 UTC
244 + Fix : Invalid memory read with "${\(new Foo)}" constructs. The
245 corresponding test is turned back on.
246 + Tst : Refinements in t/30-scope.t
248 0.09 2008-12-05 20:35 UTC
249 + Add : Support for perl 5.8.
250 + Tst : Skip a test in t/10-good.t that randomly segfaults for (I guess)
251 systems stricter than linux in the way they manage their memory.
253 0.08 2008-10-22 14:45 UTC
254 + Fix : A rare edge case for package whose names are prefix of 'main'.
255 + Tst : Test $$ as variable and state variables.
257 0.07_03 2008-10-17 20:10 UTC
258 + Add : Support and tests for variables with spaces after the sigil.
259 + Upd : META.yml spec updated to 1.4.
261 0.07_02 2008-10-15 21:10 UTC
262 + Add : Support and tests for package variables.
263 + Tst : Coverage improved by removing dead code.
265 0.07_01 2008-10-15 16:00 UTC
266 + Fix : [RT #40055] : Not handling RV2SV => GV(SV) correctly, which
267 could cause 'no indirect; print' segfaults. Thanks Goro Fuji for
270 0.06 2008-10-11 16:45 UTC
272 + Tst : Test "no indirect 'anything'", "foo Class->bar", and indirect
273 uses of exec() and system().
275 0.05 2008-10-02 14:40 UTC
276 + Chg : Now the old check function is always called before storing an op
278 + Fix : Misc code and docs refinements.
280 0.04 2008-08-30 19:00 UTC
281 + Fix : Clean up the op->src hash when we're done with an entersub.
282 + Tst : No longer fork for testing. IPC::Cmd isn't required anymore.
284 0.03 2008-08-12 15:25 UTC
285 This release is kindly supported by Copenhagen Hotel Centrum WiFi.
286 + Fix : Tests used not to pass PERL5OPTS to their kids. This lead to
287 failures under CPAN. I think.
290 0.02 2008-08-11 15:55 UTC
291 + Fix : Some building failures with old gcc versions that didn't seem to
292 like the ((hint == 2) ? croak : warn)(msg) construct. I think.
293 + Rem : Unused cruft from a previous implementation.
294 + Tst : Fail more gracefully when we can't capture buffers or when the
295 child returned an error.
297 0.01 2008-08-10 20:40 UTC
298 First version, released on an unsuspecting world.