1 Revision history for indirect
3 0.37 2016-07-12 16:40 UTC
4 + Chg : A large chunk of boilerplate XS code, which is also used in
5 other XS modules, has been factored out of the main .xs file
6 to a collection of .h files in the xsh subdirectory.
7 + Fix : [RT #115392] : Intermittent segfaults with heredocs
8 Heredocs should now be handled correctly.
9 Thanks Graham Knop for reporting.
11 0.36 2015-07-17 22:15 UTC
12 + Fix : [RT #104312] : fatal hides perl errors in modules
13 no indirect 'fatal' will no longer hide compilation errors
14 occurring before indirect constructs.
15 Thanks Lukas Mai for reporting.
17 0.35 2015-04-06 22:20 UTC
18 + Fix : The module could end being disabled in one thread if it was
19 first loaded in another thread and that thread was immediately
20 terminated. This is now fixed and should address test failures
21 of t//09-load-threads.t and t/42-threads-global.t.
23 0.34 2015-04-02 19:50 UTC
24 + Chg : The new environment variable to enable thread tests on older
25 perls is PERL_FORCE_TEST_THREADS. Note that this variable
26 should only be turned on by authors.
27 + Fix : [RT #100068] : add link to historical tchrist post
28 The link has been added to the documentation. Thanks Olivier
29 Mengué for reporting.
30 + Fix : Segfaults when the module is loaded by several threads (or
31 Windows emulated processes) ran in parallel.
32 + Fix : Update the Windows ActivePerl + gcc 3.4 workaround for
33 ExtUtils::MakeMaker 7.04. Thanks Christian Walde for reporting
34 and feedback on this issue.
35 + Fix : Be really compatible with the optional OP_PARENT feature.
36 + Tst : $ENV{$Config{ldlibpthname}} is now preserved on all platforms,
37 which will address failures of t/41-threads-teardown.t and
38 t/50-external.t with unusual compilers (like icc) that link all
39 their compiled objects to their own libraries.
41 0.33 2014-09-29 20:20 UTC
42 + Fix : [RT #99083] : Breaks eval in an END block in Win32 pseudo-forks.
43 Thanks Graham Knop for reporting.
44 + Fix : Segfaults during global destruction of a thread or a
47 0.32 2014-09-21 20:15 UTC
48 + Add : Support for the PERL_OP_PARENT optional feature introduced in
50 + Fix : [RT #92806] : Tests that use run_perl() fail on Android
51 Thanks Brian Fraser for the patch.
52 + Fix : indirect constructs will no longer segfault while inside
53 the empty package on perl 5.8.x. This fix may also prevent
54 some segfaults during global destruction.
56 0.31 2013-09-05 16:45 UTC
57 + Fix : [RT #88428] : no indirect in eval can trigger for direct calls
59 Thanks Graham Knop for reporting.
60 + Tst : Author tests are no longer bundled with this distribution.
61 They are only made available to authors in the git repository.
63 0.30 2013-05-16 15:55 UTC
64 + Fix : [RT #83806] : false positives with Devel::Declare
65 [RT #83839] : false positive using ? : syntax
66 Thanks Andrew Main for the patch.
67 However, please note that the reason this patch seems to fix
68 thinks has not been explained.
69 + Fix : [RT #84649] : incorrect RT link in metadata
70 Thanks Karen Etheridge for reporting.
72 0.29 2013-03-05 01:30 UTC
73 + Fix : [RT #83659] : false positives
74 Proper method calls in string-like environments (like
75 "@{[ $x->new ]}" will no longer be reported as indirect.
76 This was a regression in 0.28.
77 Thanks Andrew Main for reporting.
78 + Fix : Broken linkage on Windows with gcc 3.4, which appears in
79 particular when using ActivePerl's default compiler suite.
80 For those setups, the indirect shared library will now be
81 linked against the perl dll directly (instead of the import
84 0.28 2013-02-26 17:05 UTC
85 + Fix : [RT #83450] : newlines confuse indirect
86 Perl sometimes resets the line buffer between the object and
87 the method name (e.g. for "sort Class\n->method" outside of
88 eval), and this could cause direct method calls to be reported
90 Thanks Gianni Ceccarelli for reporting.
91 + Fix : Check functions are now replaced and restored in a thread-safe
92 manner, either by using the wrap_op_checker() function from perl
93 when it is available (starting from perl 5.16) or by taking the
94 OP_REFCNT mutex on older perls.
96 0.27 2013-01-30 19:00 UTC
97 + Fix : [RT #82562] : indirect/Devel::CallParser interaction
98 indirect has been taught to play nicely with Devel::CallParser.
99 Thanks Andrew Main for the patch.
100 + Tst : Author tests overhaul.
102 0.26 2011-10-23 14:25 UTC
103 + Add : "no indirect 'global'" enables the pragma for the whole program,
104 except for lexical scopes that "use indirect" explicitely.
105 + Chg : Passing both the 'fatal' and 'hook' options to unimport() is
106 now considered an error, and will result in an exception.
107 unimport() used to consider only the first passed option of
108 those two, and silently ignored the other.
109 + Tst : Test failures of t/41-threads-teardown.t and t/50-external.t on
110 Cygwin should have been addressed.
111 + Tst : Threads tests will not fail anymore if resources constraints
112 prevent the system from creating all the required threads.
114 0.25 2011-08-24 15:40 UTC
115 + Fix : RT #69291 is now also fixed for perl 5.8. The pragma will no
116 longer vivify the "indirect" entry in the hints hash %^H on
118 + Tst : Attempt to make t/50-external.t pass on Cygwin.
120 0.24 2011-07-17 23:15 UTC
121 + Fix : [RT #64521] : "no indirect" leaking into eval.
122 This is currently only fixed for perl 5.10 (perl 5.12 and
123 higher were never affected). It was caused by a very stupid
124 mistake of mine that was introduced in indirect version 0.23.
125 Thanks Michael G Schwern for reporting.
126 + Fix : [RT #69291] : indirect.pm breaks %^H.
127 This was caused by the same mistake as for the previous bug,
128 and as such it is also only fixed for perl 5.10 (and never
130 Thanks Andrew Main for reporting.
131 + Doc : C++ compilers are officially NOT supported.
133 0.23 2010-10-03 00:15 UTC
134 + Fix : Some indirect constructs could be incorrectly reported when
135 several modules were used in the same scope. This caused
136 t/30-scope.t to fail randomly.
137 + Tst : Threads tests are now only run on perl 5.13.4 and higher.
138 They could segfault randomly because of what seems to be an
139 internal bug of Perl, which has been addressed in 5.13.4.
140 There is also an environment variable that allows you to
141 forcefully run those tests, but it should be set only for
142 author testing and not for end users.
144 0.22 2010-08-16 16:00 UTC
145 + Add : Indirect constructs are now reported for code interpolated
146 in quote-like environments, like "${\( ... )}", "@{[ ... ]}",
147 s/pattern/ ... /e, qr/(?{ ... })/ or qr/(??{ ... })/.
148 + Add : You can now make the pragma lethal by passing anything matching
149 /^:?fatal$/i to import(), including "FATAL" and ":Fatal".
150 + Fix : [RT #60378] : segmentation fault on indirect_ck_method.
151 This caused constructs like "@{[ $obj->$meth ]}" to segfault
152 when $meth was a lexical.
153 Thanks Tokuhiro Matsuno for reporting.
155 0.21 2010-05-31 23:10 UTC
156 + Chg : perl 5.8.1 is now required (instead of 5.8.0).
157 + Fix : [RT #57699] : indirect fail with 64-bit int on 5.13.1.
158 It was actually a problem with thread destructors segfaulting
159 because they weren't called at the right time anymore.
160 Thanks Andrew Main for reporting.
161 + Tst : A few more regression tests about the scope leak bug.
163 0.20 2010-04-18 21:25 UTC
164 + Fix : [RT #50570] : "indirect" leaking into LWP.
165 Thanks Andrew Main for reporting.
166 More generally, the require propagation workaround on 5.8-5.10.0
167 has been overhauled, and other scope leaks should be fixed.
168 + Fix : Test failures with 5.12 on Windows where Strawberry Perl crashes
169 because the SystemRoot environment variable is missing.
170 + Fix : Work around Kwalitee test misfailures.
172 0.19 2009-08-28 18:40 UTC
173 + Add : The new constant I_FORKSAFE can be tested to know whether the
174 module will behave nicely when fork()ing. It's currently always
175 true except on Windows where you need perl 5.10.1 for it to be
177 + Fix : I_THREADSAFE and I_FORKSAFE ought to be true when
178 PERL_INDIRECT_PM_DISABLE is set.
179 + Fix : The pragma could still leak if you passed to the "hook" option
180 a reference to the same (named) subroutine from two different
181 require scopes. The fix also provides a better solution for
183 + Fix : Segfaults when indirect is loaded for the first time from
185 + Fix : Leaks of memory associated with the root interpreter.
186 + Opt : Less memory will be used for non-threaded perls version 5.10.0
187 and below, and for threaded perls from version 5.10.1.
189 0.18 2009-08-23 16:15 UTC
190 + Add : When set, the PERL_INDIRECT_PM_DISABLE environment variable
191 disables the pragma globally.
193 0.17 2009-07-16 12:10 UTC
194 + Fix : [RT #47902] : "no indirect" leaking again.
195 This actually turned out to be a bug in how the hook coderefs
196 were stored in the hints hash.
197 Thanks Andrew Main for reporting once again.
198 + Fix : t/80-regressions.t failing on Windows.
199 + Tst : Yet more cleanups.
201 0.16 2009-07-14 16:50 UTC
202 + Add : Indirect calls on blocks are now reported. For those, '{' is
203 passed to the hook as the object description.
204 + Add : The new indirect::msg() function publicizes the default
205 warning/exception message.
206 + Fix : [RT #47866] : Segfault with UTF-8 regexps.
207 Thanks Andrew Main for reporting.
210 0.15 2009-07-08 22:55 UTC
211 + Fix : Invalid constructs with the same method and package name were
213 + Fix : The error line number used to point to the end of the expression
214 instead of its beginning.
216 0.14 2009-06-04 21:55 UTC
217 + Fix : Prevent bogus invalid syntaxes caused by reallocated memory
218 chunks. Thanks Andrew Main for reporting with a reproducible
221 0.13 2009-05-24 18:50 UTC
222 + Add : The hook now receives the file name and the line where the
223 error happened in respectively $_[2] and $_[3].
224 + Fix : Pass mortalized copies of the method name and object to the
225 hook. This seems to fix some rare crashes.
226 + Fix : Work around a bug in perl 5.10.0 and lower. Thanks Andrew Main
227 for teaching me about this issue.
228 + Fix : Report the correct file in error messages (a regression from the
231 0.12 2009-05-03 14:30 UTC
232 + Add : You can specify the action to execute for each indirect
233 construct encountered with the new "hook => $coderef" unimport()
235 + Chg : A ptable is now used internally for the op => position mapping.
236 + Fix : The pragma is now built with thread safety features enabled
237 whenever possible (a notable exception is perl 5.8.x on Win32,
238 as something seems wrong with its context handling).
239 The new indirect::I_THREADSAFE() constant reflects this.
240 + Fix : A negation precedence nit in indirect_ck_entersub().
241 + Tst : "use/no indirect" while parsing an indirect construct.
242 + Tst : Thread safety.
244 0.11 2009-02-08 18:35 UTC
245 + Fix : Potential collisions by hashing pointers with a wrong format.
246 + Upd : Resources in META.yml.
248 0.10 2009-01-17 12:40 UTC
249 Re-release 0.09_01 as stable.
251 0.09_01 2008-12-08 17:55 UTC
252 + Fix : Invalid memory read with "${\(new Foo)}" constructs. The
253 corresponding test is turned back on.
254 + Tst : Refinements in t/30-scope.t
256 0.09 2008-12-05 20:35 UTC
257 + Add : Support for perl 5.8.
258 + Tst : Skip a test in t/10-good.t that randomly segfaults for (I guess)
259 systems stricter than linux in the way they manage their memory.
261 0.08 2008-10-22 14:45 UTC
262 + Fix : A rare edge case for package whose names are prefix of 'main'.
263 + Tst : Test $$ as variable and state variables.
265 0.07_03 2008-10-17 20:10 UTC
266 + Add : Support and tests for variables with spaces after the sigil.
267 + Upd : META.yml spec updated to 1.4.
269 0.07_02 2008-10-15 21:10 UTC
270 + Add : Support and tests for package variables.
271 + Tst : Coverage improved by removing dead code.
273 0.07_01 2008-10-15 16:00 UTC
274 + Fix : [RT #40055] : Not handling RV2SV => GV(SV) correctly, which
275 could cause 'no indirect; print' segfaults. Thanks Goro Fuji for
278 0.06 2008-10-11 16:45 UTC
280 + Tst : Test "no indirect 'anything'", "foo Class->bar", and indirect
281 uses of exec() and system().
283 0.05 2008-10-02 14:40 UTC
284 + Chg : Now the old check function is always called before storing an op
286 + Fix : Misc code and docs refinements.
288 0.04 2008-08-30 19:00 UTC
289 + Fix : Clean up the op->src hash when we're done with an entersub.
290 + Tst : No longer fork for testing. IPC::Cmd isn't required anymore.
292 0.03 2008-08-12 15:25 UTC
293 This release is kindly supported by Copenhagen Hotel Centrum WiFi.
294 + Fix : Tests used not to pass PERL5OPTS to their kids. This lead to
295 failures under CPAN. I think.
298 0.02 2008-08-11 15:55 UTC
299 + Fix : Some building failures with old gcc versions that didn't seem to
300 like the ((hint == 2) ? croak : warn)(msg) construct. I think.
301 + Rem : Unused cruft from a previous implementation.
302 + Tst : Fail more gracefully when we can't capture buffers or when the
303 child returned an error.
305 0.01 2008-08-10 20:40 UTC
306 First version, released on an unsuspecting world.