Intel Software Network - <![CDATA[ Intel® C++ Compiler ]]> feed http://software.intel.com/en-us/forums/intel-c-compiler en-us Intel Compiler bug?
I have boiled down a complex "bug" into a very small program which fails to catch an exception when compiled by the Intel C++ compiler (v11), however it works fine on gcc (4.x).  Note that on the Intel compiler, it doesn't crash, it just sits there forever.  When I ran it through gdb I noticed it seemed to be going into an endless loop in the std::vector destructor (see the code below).

What triggers this bug seems to be the inclusion of the OpenMP parallel directive.  If I serialize the little loop by commenting out the #pragma omp parallel for, the exception is succesfully caught, even on the Intel compiler.  So it seems to have something to do with the Intel compiler's OpenMP implementation, and how it plays with exceptions.  It is a compiler bug, I think, since this code doesn't throw any exceptions inside the parallel region.  Please see below:


#include <string>
#include <vector>
#include <iostream>

std::string
bad_func()
{
throw std::exception();
}

int
main(int argc, char* argv[])
{
try {
std::vector<float> v0;
std::vector<float> v1;
bad_func();
#pragma omp parallel for
for(unsigned int q=0; q<8; ++q) {
std::vector<float> u,v;
}       
}
catch(std::exception& e) {
std::cerr << e.what() << std::endl;
}
}
]]> http://software.intel.com/en-us/forums/intel-c-compiler/topic/70144/ Wed, 25 Nov 2009 13:33:38 -0800 http://software.intel.com/en-us/forums/intel-c-compiler/topic/70144/ ISN General icc: command line error: invalid argument for option &#39;-std&#39;
configure:5649: checking whether icc supports -Wmissing-format-attribute
configure:5686: icc -o conftest -std=iso9899:1990 -W -Wall -Wchar-subscripts -Wshadow -Wpointer-arith -Wmissing-prototypes -Wwrite-strings -D_XOPEN_SOURCE=500 -D_BSD_SOURCE -O2   -Wmissing-format-attribute   conftest.c  >&5
icc: command line error: invalid argument for option '-std'
configure:5693: $? = 1
configure: failed program was:
| /* confdefs.h.  */
| #define PACKAGE_NAME ""
| #define PACKAGE_TARNAME ""
| #define PACKAGE_VERSION ""
| #define PACKAGE_STRING ""
| #define PACKAGE_BUGREPORT ""
| #define PACKAGE "kdelibs"
| #define VERSION "3.5.10"
| /* end confdefs.h.  */
|
| int
| main ()
| {
|  return 0;
|   ;
|   return 0;
| }
configure:5728: result: no
]]> http://software.intel.com/en-us/forums/intel-c-compiler/topic/70102/ Wed, 25 Nov 2009 12:25:21 -0800 http://software.intel.com/en-us/forums/intel-c-compiler/topic/70102/ ISN General Windows 7/Windows Server 2008 Users - Please update to 11.1.048 or newer The Intel C++ Compiler for Windows 11.1.048 (update 2 revised) version has just released. It contains the complete fix for a bug DPD200139120 - installation hangs on Windows 7.

You can download it from Intel Registration Center or from our eval center.

Thanks,
Jennifer


Title: Errata Reference Number DPD200139120  

Problem:  Applications built with the Intel® C++ and Visual Fortran Compilers Professional Edition, version 11.1, or Intel® Parallel Composer may hang or encounter a segmentation-fault when run on Microsoft* Windows 7* or Windows Server 2008* R2.  Also, a similar issue may occur during installation of these compiler versions on these operating system versions.

Version:

  • Intel® C++ & Visual Fortran Compilers for Windows* 11.1 initial release and update 1 release:
    • Intel® C++ & Visual Fortran Compiler for Windows* 11.1.035
    • Intel® C++ & Visual Fortran Compiler for Windows* 11.1.038
  • Intel® Parallel Studio
  • Intel® Parallel Composer initial release, and update 1 release

Previous versions of the Intel® Compilers (version 11.0 or earlier) do not have this issue.  Please refer to the corresponding Release Notes for additional information on minimum system requirements for which operating systems are supported by that product version.

For applications built with the Intel Compilers version 11.1, the problem is contained within the libirc.lib, libircmt.lib on Windows.

 

Resolution:    

For the Intel® Compiler Professional Editions 11.1, this issue is resolved in update release 11.1.048 available on Intel Registration Center.  

For Intel® Parallel Composer, this issue is resolved in the update 2 revised release, available on Intel Registration Center.

Intel is still validating these configurations and will update the Systems Requirements when validation is complete.

Please find additional information on the install issue here http://software.intel.com/en-us/articles/the-installation-or-uninstallation-hangs-at-start-up-dialog-during-reading-existing-licenses/ and here http://software.intel.com/en-us/articles/program-crashes-or-hangs-on-some-systems/

Intel® Compilers Support Team

*Other names and brands may be claimed as the property of others

]]> http://software.intel.com/en-us/forums/intel-c-compiler/topic/68939/ Wed, 25 Nov 2009 12:16:09 -0800 http://software.intel.com/en-us/forums/intel-c-compiler/topic/68939/ ISN General Meet Intel® MKL team at SC ’09 in Portland Nov 16-19, 2009 http://software.intel.com/en-us/forums/intel-math-kernel-library/topic/69328/

Jennifer ]]> http://software.intel.com/en-us/forums/intel-c-compiler/topic/69422/ Wed, 25 Nov 2009 12:15:09 -0800 http://software.intel.com/en-us/forums/intel-c-compiler/topic/69422/ ISN General Intel-Compiled Boost Libraries Has anyone in the community compiled the C++ Boost libraries using the Intel Compiler? It would be a great time saver if the binary files could be posted for others to use. There are MSVS binaries posted on other sites, we love our Intel Compiler ;)

]]> http://software.intel.com/en-us/forums/intel-c-compiler/topic/70146/ Wed, 25 Nov 2009 07:41:19 -0800 http://software.intel.com/en-us/forums/intel-c-compiler/topic/70146/ ISN General Intel on Linux...what directory?
By some bumbling miracle I've managed to install the Intel Compiler Suite Pro on our Linux box (Red Hat Enterprise). Since I know next to nothing about Linux I consider this a major victory! Anyway, I installed to the default directory...whatever that is, but I'm perusing the file system and I don't see anywhere the compiler was installed to. I see the directory I created that the download was unpacked into, but I'd like to know the location of the actual installation. Anyone? Thanks in advance.

-L
]]> http://software.intel.com/en-us/forums/intel-c-compiler/topic/70114/ Tue, 24 Nov 2009 13:32:15 -0800 http://software.intel.com/en-us/forums/intel-c-compiler/topic/70114/ ISN General icpc: command line warning #10156: ignoring option &#39;-W&#39;; no argument required Am i right that gcc have analog for icc's -W<things> ?

configure:5560: checking whether icpc supports -Wmissing-format-attribute
configure:5600: rm -rf SunWS_cache; icpc -o conftest -ansi -D_XOPEN_SOURCE=500 -D_BSD_SOURCE -Wcast-align -Wchar-subscripts -Wall -W -Wpointer-arith -O2   -Wmissing-format-attribute   conftest.cpp  1>&5
icpc: command line warning #10156: ignoring option '-W'; no argument required
icpc: command line warning #10156: ignoring option '-W'; no argument required
icpc: command line remark #10148: option '-W' not supported
icpc: command line warning #10156: ignoring option '-W'; no argument required
conftest.cpp(16): remark #111: statement is unreachable
return 0;
^

configure:5607: $? = 0
configure:5637: result: yes
]]> http://software.intel.com/en-us/forums/intel-c-compiler/topic/70103/ Tue, 24 Nov 2009 02:51:59 -0800 http://software.intel.com/en-us/forums/intel-c-compiler/topic/70103/ ISN General Missing stdint.h with C99 i want to build a tool (in Windows), but the compiler didn´t find the stdint.h, with the /Qstd:c99 option.
The error message: "catastrophic error: could not open source file "stdint.h"
#include <stdint.h>"
where is the Failure?
I use Intel C++ Windows 11.1.048
Can anybody help me?
Thx
Brian ]]> http://software.intel.com/en-us/forums/intel-c-compiler/topic/70079/ Mon, 23 Nov 2009 03:12:58 -0800 http://software.intel.com/en-us/forums/intel-c-compiler/topic/70079/ ISN General Error on partial template specialization
my full template looks like this:

template <class ELEMENT_TYPE, typename STORAGE_TYPE, ParallelMode::Mode PARALLEL_MODE>
void Matrix<ELEMENT_TYPE, STORAGE_TYPE, PARALLEL_MODE>::Avarage()

But when i want to use a partial specialization on the folowing line

template <class ELEMENT_TYPE, typename STORAGE_TYPE>
void Matrix<ELEMENT_TYPE, STORAGE_TYPE, ParallelMode::MPI>::Avarage()

I get the folowing error

error: template argument list must match the parameter list

Any suggestions?

]]> http://software.intel.com/en-us/forums/intel-c-compiler/topic/70035/ Sun, 22 Nov 2009 22:40:44 -0800 http://software.intel.com/en-us/forums/intel-c-compiler/topic/70035/ ISN General RTC failure on FD_ZERO in any program compiled with -debug -check-uninit on openSUSE 11.2 or Ubuntu 9.10
I've just installed Intel C++ compiler 11.1/059 on openSUSE 11.1 for x86_64.  It appears that any program contaning the FD_ZERO macro with abort with an RTC failure if compiled in -debug -check-uninit mode. For example, put the following into a file called /tmp/foo.cpp:

#include <stdlib.h>
#include <stdio.h>
#include <sys/select.h>
int main (int argc, char ** argv)
{
fd_set readfds;
FD_ZERO (&readfds);
return EXIT_SUCCESS;
}

Then compile it:
. /opt/intel/Compiler/11.1/059/bin/iccvars.sh  intel64
icpc -debug -check-uninit -c -o foo.o foo.cpp
icpc   -o foo foo.o.

Then run it and observe the crash:
$ /tmp/foo

Run-Time Check Failure: The variable '__d1' is being used without being initialized

Aborted


The '__d1' variable is used in the GLIBC's definition of the FD_ZERO macro in /usr/include/bits/select.h:
....
# if __WORDSIZE == 64
#  define __FD_ZERO_STOS "stosq"
#  define __FD_SET_BTS "btsq"
#  define __FD_CLR_BTR "btrq"
#  define __FD_ISSET_BT "btq"
# else
...
# endif

# define __FD_ZERO(fdsp) \
do {<><------><------><------><------><------><------><------><------>      \
int __d0, __d1;<---><------><------><------><------><------><------>      \
__asm__ __volatile__ ("cld; rep; " __FD_ZERO_STOS<-><------><------>      \
<------><------><------>  : "=c" (__d0), "=D" (__d1)<--><------><------>      \
<------><------><------>  : "a" (0), "0" (sizeof (fd_set)<-----><------>      \
<------><------><------><------><------>  / sizeof (__fd_mask)),<------>      \
<------><------><------>    "1" (&__FDS_BITS (fdsp)[0])><------><------>      \
<------><------><------>  : "memory");<><------><------><------><------>      \
} while (0)


Any suggestions?

$ rpm -q gcc glibc
gcc-4.4-4.2.x86_64
glibc-2.10.1-10.4.x86_64
]]> http://software.intel.com/en-us/forums/intel-c-compiler/topic/69937/ Sun, 22 Nov 2009 10:56:45 -0800 http://software.intel.com/en-us/forums/intel-c-compiler/topic/69937/ ISN General