[Bug c++/70558] New: POD datatype array crash in initialization when using optimization
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70558 Bug ID: 70558 Summary: POD datatype array crash in initialization when using optimization Product: gcc Version: 5.2.1 Status: UNCONFIRMED Severity: major Priority: P3 Component: c++ Assignee: unassigned at gcc dot gnu.org Reporter: hlprasu at gmail dot com Target Milestone: --- Created attachment 38197 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=38197&action=edit Contains example code to reproduce this problem I have created a simple POD class named Double, which encapsulates 'double' type so that comparison operators can be handled in a numerically correct fashion. This is in the attached .tar.gz file. My problem is that if a create a fixed array with it and try to initialize all elements in a single statement then it crashes only when optimization is enabled. E.g.: Double v[3]; v[0] = v[1] = v[2] = 0.0; // <- crash here with -O or -O2 or -O3 Funnily enough, this does not crash if you do v[0] = 0.0; v[1] = 0.0; v[2] = 0.0; To reproduce this error, 1. Untar the attached file and go to the directory tar -xvzf double-pod-crash.tar.gz cd double-pod-crash/ 2. Run 'make' and execute the test file. make ./test # <- This segfaults at the above statement I saw the crash to be at the above statement by using gdb after recompiling with -g option. Please let me know a solution to this problem.
[Bug libstdc++/41949] std::endl documentation contains bad link
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=41949 Prasad H. L. changed: What|Removed |Added CC||hlprasu at gmail dot com --- Comment #4 from Prasad H. L. --- (In reply to Jonathan Wakely from comment #3) > Fixed for 4.5.0 The link seems to be broken again. Could you please fix it?
[Bug c++/44855] Static members not initialised in explicit template instances of library
--- Comment #6 from hlprasu at gmail dot com 2010-07-25 06:35 --- Subject: Re: Static members not initialised in explicit template instances of library Output of "gcc -v" is given below. I'm using Fedora 13 (with updates till July 24, 2010). The error still persists. === Using built-in specs. Target: i686-redhat-linux Configured with: ../configure --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-bugurl=http://bugzilla.redhat.com/bugzilla --enable-bootstrap --enable-shared --enable-threads=posix --enable-checking=release --with-system-zlib --enable-__cxa_atexit --disable-libunwind-exceptions --enable-gnu-unique-object --enable-languages=c,c++,objc,obj-c++,java,fortran,ada --enable-java-awt=gtk --disable-dssi --with-java-home=/usr/lib/jvm/java-1.5.0-gcj-1.5.0.0/jre --enable-libgcj-multifile --enable-java-maintainer-mode --with-ecj-jar=/usr/share/java/eclipse-ecj.jar --disable-libjava-multilib --with-ppl --with-cloog --with-tune=generic --with-arch=i686 --build=i686-redhat-linux Thread model: posix gcc version 4.4.4 20100630 (Red Hat 4.4.4-10) (GCC) === On 25 July 2010 07:43, pinskia at gcc dot gnu dot org wrote: > --- Comment #5 from pinskia at gcc dot gnu dot org  2010-07-25 02:13 > --- > Works for me on the trunk and in 4.3.2. > > Can you give the output of "gcc -v"? > -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44855
[Bug c++/44855] New: Static members not initialised in explicit template instances of library
Consider the following C++ source code files
=> testLib.h <===
#include
#include
#define DEBUG1(X) std::cout<
class CompileTimeFixedSizeData
{
protected:
X _data[Size]; //!< The storage of the data
static IndexType _size; //!< The size of the data.
public:
CompileTimeFixedSizeData()
{
DEBUG1("Template value = "<< Size << std::endl);
DEBUG1("Static variable value = "<< _size<< std::endl);
}
virtual ~CompileTimeFixedSizeData()
{}
};
//setting the value of _size, which is static
template
IndexType CompileTimeFixedSizeData::_size=Size;
/** Create an instance of CompileTimeFixedSizeData and associated functions.
*/
#define COMPILE_TIME_FIXED_SIZE_DATA_INSTANCE(TYPE,SIZE) \
template class CompileTimeFixedSizeData;
}
=> testLib.cpp <===
#include "testLib.h"
namespace Upagraha
{
COMPILE_TIME_FIXED_SIZE_DATA_INSTANCE(double,3)
}
=> main.cpp <===
#include "testLib.h"
namespace Upagraha {
//extern template class CompileTimeFixedSizeData;
// FIXME: Uncomment to see a change in behavior
}
int main()
{
Upagraha::CompileTimeFixedSizeData a;
}
=> makefile <===
#OPTIONS=-rdynamic -fno-implicit-templates -fPIC
OPTIONS=-rdynamic -fno-implicit-templates
# FIXME: Uncomment the top one to see change in behaviour
test:libtestLib.so testLib.h main.cpp
g++ $(OPTIONS) $(TOPTIONS) main.cpp -L"${PWD}" -ltestLib -o main
libtestLib.so: testLib.cpp testLib.h
g++ $(OPTIONS) $(TOPTIONS) -shared testLib.cpp -o libtestLib.so
clean:
rm main libtestLib.so *~
Put the above four files in same the directory and do the following
*) make
*) export LD_LIBRARY_PATH="$PWD"
*) ./main
You'll see the output as
=
Template value = 3
Static variable value = 0
=
Now, follow either of the two "FIXME"s and build the program and run to get
=
Template value = 3
Static variable value = 3
=
Why this difference in behaviour in initialisation of static variable
Upagraha::CompileTimeFixedSizeData::_size? Could somebody please
explain this?
--
Summary: Static members not initialised in explicit template
instances of library
Product: gcc
Version: 4.4.4
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: hlprasu at gmail dot com
GCC build triplet: i686-redhat-linux
GCC host triplet: i686-redhat-linux
GCC target triplet: i686-redhat-linux
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44855
[Bug c++/44855] Static members not initialised in explicit template instances of library
--- Comment #1 from hlprasu at gmail dot com 2010-07-07 07:22 --- Created an attachment (id=21120) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=21120&action=view) The test case attached To simplify the task of whomsoever who wants to see the behaviour, all the files mentioned in my previous post is packaged in a tar.bz2 file attached here. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44855
[Bug c++/44855] Static members not initialised in explicit template instances of library
--- Comment #3 from hlprasu at gmail dot com 2010-07-07 09:30 --- Subject: Re: Static members not initialised in explicit template instances of library But, as I know, no where is it mentioned to be necessary for static member initialisation. Why should the compiler allow for such a subtle error? Or is this documented somewhere? Besides this issue, I've not faced any other problem in using shared library by not using -fPIC. Could you please point out the necessary implications that I am missing here? On 7 July 2010 14:48, redi at gcc dot gnu dot org wrote: > > > --- Comment #2 from redi at gcc dot gnu dot org  2010-07-07 09:18 --- > You need to use -fPIC for a shared object, so the version without that is not > valid. > > -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44855
