This program built for x86_64-redhat-linux-gnu Usage: make [options] [target] ... Tested revisions: r161428 - crash r161383 - crash r161170 - OK -- Summary: [4.6 Regression] ICE: in tree_nrv, at tree-nrv.c:155 The STE||AR Group member brycelelbach commented Jul 10, 2012 [comment by vamatya] [Trac time Sun Aug 7 19:17:03 2011] svnversion : 5210 (HPX) gcc: 4.4.5-2 boost: 1_47_0 configure command: cmake -DBOOST_INCLUDE_DIR=$BOOST_ROOT/include When using -O0 instead, the code produces the expected result. weblink
In file included from /home/manderson/boost-trunk/boost/tuple/tuple.hpp:33, from /home/manderson/parallex/trunk/hpx/external/coroutine/boost/coroutine/coroutine.hpp:37, from /home/manderson/parallex/trunk/hpx/hpx/hpx_fwd.hpp:28, from /home/manderson/parallex/trunk/hpx/src/runtime/components/console_error_sink.cpp:7: /home/manderson/boost-trunk/boost/tuple/detail/tuple_basic.hpp: In function ‘const boost::tuples::null_type boost::tuples::detail::cnull()’: /home/manderson/boost-trunk/boost/tuple/detail/tuple_basic.hpp:52: internal compiler error: in tree_nrv, at tree-nrv.c:143 Please submit a full bug report, See http://bugzilla.redhat.com/bugzilla for instructions. Format For Printing -XML -Clone This Bug -Top of page Home | New | Browse | Search | [?] | Reports | Help | NewAccount | Log In Remember [x] | The expected output is this: a*b=(17,17,17,17) a*c=(119,119,119,119) b*c=(7,7,7,7) What happened?
Comment 3 by [email protected], Apr 9 2013 Processing Status: Started This bug is worked around by using the flag -fno-strict-aliasing with -O2, or by just using -O1 instead of -O2. See http://bugzilla.redhat.com/bugzilla for instructions. See
I encountered this crash when simply upgrading to the latest boost (1.47.0), which uses attribute optimize, so this is likely to affect other people. Personal Open source Business Explore Sign up Sign in Pricing Blog Support Search GitHub This repository Watch 79 Star 517 Fork 133 STEllAR-GROUP/hpx Code Issues 101 Pull requests 5 Projects Test Application Run: accumulator_client -r Segmentation fault (core dumped) p.s. See http://bugzilla.redhat.com/bugzilla for instructions.
No general (`VPATH' variable) search path. # of strings in strcache: 0 # of strcache buffers: 0 strcache size: total = 0 / max = 0 / min = 4096 / Comment 5 by [email protected], Apr 10 2013 Processing Can you try using -O2 -fno-strict-aliasing and measure the actual performance impact? In file included from /home/manderson/boost-trunk/boost/tuple/tuple.hpp:33, from /home/manderson/parallex/trunk/hpx/external/coroutine/boost/coroutine/coroutine.hpp:37, from /home/manderson/parallex/trunk/hpx/hpx/hpx_fwd.hpp:28, from /home/manderson/parallex/trunk/hpx/src/performance_counters/server/raw_counter.cpp:6: /home/manderson/boost-trunk/boost/tuple/detail/tuple_basic.hpp: In function ‘const boost::tuples::null_type boost::tuples::detail::cnull()’: /home/manderson/boost-trunk/boost/tuple/detail/tuple_basic.hpp:52: internal compiler error: in tree_nrv, at tree-nrv.c:143 Please submit a full bug report, https://github.com/STEllAR-GROUP/hpx/issues/20 The workaround I tried for this was to insert a "#pragma GCC optimize("O0")", which does indeed fix the problem (see the commented pragma in the cpp file).
Comment 1 Lorenzo Castelli 2010-12-19 10:18:34 UTC Created attachment 22819 [details] Code to reproduce the error Comment 2 Mat Hostetter 2011-09-20 12:31:56 UTC Created attachment 25327 [details] bug showing ICE with bugreport.cpp 1.4 KB View Download Comment 1 by [email protected], Apr 2 2013 Processing Cc: [email protected] [email protected] Owner: [email protected] Comment 2 by [email protected], Apr 8 2013 Processing Labels: -Pri-2 Pri-1 M-28 Status: This is on Release build. There is no flag to specifically disable it.
Binary file core.xxxxxx present. read this article Finally, I need the command line options that the application was run with. Comment 3 Ben Bucksch (:BenB) 2012-03-31 15:43:41 PDT I can't easily install another compiler, I didn't find newer official packages. In file included from /home/manderson/boost-trunk/boost/filesystem/v3/path_traits.hpp:23, from /home/manderson/boost-trunk/boost/filesystem/v3/path.hpp:25, from /home/manderson/boost-trunk/boost/filesystem/path.hpp:31, from /home/manderson/parallex/trunk/hpx/external/logging/boost/logging/format/destination/rolling_file.hpp:35, from /home/manderson/parallex/trunk/hpx/external/logging/boost/logging/writer/named_write.hpp:29, from /home/manderson/parallex/trunk/hpx/external/logging/boost/logging/format/named_write.hpp:24, from /home/manderson/parallex/trunk/hpx/hpx/util/logging.hpp:12, from /home/manderson/parallex/trunk/hpx/hpx/exception.hpp:16, from /home/manderson/parallex/trunk/hpx/hpx/runtime.hpp:11, from /home/manderson/parallex/trunk/hpx/src/state.cpp:10: /home/manderson/boost-trunk/boost/system/error_code.hpp: In member function ‘virtual boost::system::error_condition boost::system::error_category::default_error_condition(int) const’:
I will go with that then. http://wiiplay.net/internal-compiler/internal-compiler-error-cl-exe.html Tinderbox is using 4.5.x IIRC. Comment 4 by [email protected], Apr 10 2013 Processing The problem with that workaround is that the offending code lives in a header which is included in many source files. It can silently generate incorrect code, which took me longer to track down.
That looks like release with debug info. This is free software; see the source for copying conditions. Project Member Comment 8 by [email protected], May 8 2013 Processing Labels: -M-28 MovedFrom-28 M-29 Moving all non essential bugs to the next Milestone. http://wiiplay.net/internal-compiler/internal-compiler-error-gcc.html See http://bugzilla.redhat.com/bugzilla for instructions.
Username: Password: Have you forgotten your login information? See http://bugzilla.redhat.com/bugzilla for instructions. HPX_LOGLEVEL=4 The STE||AR Group member brycelelbach commented Jul 10, 2012 [comment by manderson] [Trac time Mon Aug 8 15:50:46 2011] compiler: gcc 4.4.4 Boost: top of trunk on 8 Aug 2011
The -fno-strict-aliasing switch exists to tell the compiler that exactly this sort of nonstandard type-punning is being done. make: ** [src/CMakeFiles/hpx.dir/hpx_init.cpp.o] Error 1 make: *** [src/CMakeFiles/hpx.dir/all] Error 2 make: *** [all] Error 2 Building in Debug mode: configure script: cmake -DBOOST_INCLUDE_DIR=$BOOST_ROOT/include -DBOOST_LIB_DIR=$BOOST_ROOT/lib -DCMAKE_PREFIX=/home/vamatya/packages/inst_/hpx ../../src/hpx11/ -DCMAKE_BUILD_TYPE=Debug Build Succesfull. See
Preprocessed source stored into /tmp/ccmYrTaS.out file, please attach this to your bugreport. See http://bugzilla.redhat.com/bugzilla for instructions. I have checked through the GCC release notes and bug tracker, and I've been unable to find any information on this cause of this. http://wiiplay.net/internal-compiler/internal-compiler-error-net.html make: invalid option -- 'a' make: invalid option -- '-' GNU Make 3.81 Copyright (C) 2006 Free Software Foundation, Inc.
© Copyright 2017 wiiplay.net. All rights reserved.