Intel Software Network feed

Intel name mangling planning

Sun, 19 Jul 2009 20:32:24 -0700

The name mangling for types __m64, __m128, __m128d, __m128i differs between Intel and Gnu compilers.

For example, gcc mangles __m128 as "U8__vectorf", while icc makes it "6__m128"

C++ example:
--------------------------------
#include <emmintrin.h>
int TestFunction(__m128 x) {return 1;}
--------------------------------
Output for g++ -S:
_Z12TestFunctionU8__vectorf
--------------------------------
Output for icc -S:
_Z12TestFunction6__m128
--------------------------------

Anyone who can tell me is this change coordinated with the gcc developers?

Thanks,
Wise

Problem with License Manager

Sun, 19 Jul 2009 20:26:37 -0700

Hi!
I received following error message while installing Intel license manager:
"Setup was unable to start the Intel License Manager for FLEXlm. Make sure you have a valid license file"
This error appear when i use combined license file and even i use initial license files received from Intel.

Environment for Intel license manager installing is Windows 2008 Server Standard 32-bit.
Host name and host id parameters correspond with server name and mac address one of NICs.

Anyone have ideas?

vm_strings.h - compilation error.

Sun, 19 Jul 2009 19:50:33 -0700

I am trying to compile my program in Visual Studio 6. I have just included all the necessary directories and libraries of IPP. When i try to compile my program it gives following error:

Compiling...
MemorySink.cpp
c:\program files\microsoft visual studio\vc98\include\errno.h(76) : warning C4005: 'EINTR' : macro redefinition
d:\work\dx source filter\live555\live\groupsock\include\netcommon.h(41) : see previous definition of 'EINTR' c:\program files\microsoft visual studio\vc98\include\errno.h(83) : warning C4005: 'EAGAIN' : macro redefinition
d:\work\dx source filter\live555\live\groupsock\include\netcommon.h(40) : see previous definition of 'EAGAIN'

d:\work\intel_ipp\data\w_ipp-samples_p_6.0.0.117\ipp-samples\audio-video-codecs\core\vm\include\vm_strings.h(134) : error C2146: syntax error : missing ';' before identifier 'vm_string_findfirst'

d:\work\intel_ipp\data\w_ipp-samples_p_6.0.0.117\ipp-samples\audio-video-codecs\core\vm\include\vm_strings.h(134) : error C2501: 'intptr_t' : missing storage-class or type specifiers

d:\work\intel_ipp\data\w_ipp-samples_p_6.0.0.117\ipp-samples\audio-video-codecs\core\vm\include\vm_strings.h(134) : fatal error C1004: unexpected end of file found
Error executing cl.exe.

OpenRTSPStreamConsole.exe - 3 error(s), 2 warning(s)

Following is my include paths:

..\..\..\..\..\..\..\Dx Source Filter\Live555\live\liveMedia\include,
..\..\..\..\..\..\..\Dx Source Filter\Live555\live\UsageEnvironment\include,
..\..\..\..\..\..\..\Dx Source Filter\Live555\live\groupsock\include,
..\..\..\..\..\..\..\Dx Source Filter\Live555\live\BasicUsageEnvironment\include,
..\..\..\audio-video-codecs\codec\color_space_converter\include,
..\..\..\audio-video-codecs\core\umc\include,
..\..\..\audio-video-codecs\core\vm\include,
..\..\..\audio-video-codecs\pipeline\umc_pipeline\include\,
..\..\..\audio-video-codecs\core\vm_plus\include,
..\..\..\audio-video-codecs\codec\mpeg4_dec\include,
..\..\..\audio-video-codecs\codec\h264_dec\include,
..\..\..\audio-video-codecs\codec\jpeg\include

Please help.

IPP samples for Mac OS

Sun, 19 Jul 2009 19:46:21 -0700

Hi all,

I'm quite new in IPP and I'm working on Mac.
I have downloaded sample codes for Mac OS, but they are not working.
Are there any tutorials that can give a good start for IPP under Mac?

Another question I have:
Apply offers several frameworks for app development. Is there any preference of which one of them to use for IPP (for instance: Cocoa or Carbon)

Thanks.

libstdc++.so.5 is missing...openSuse 11.1 64-bit

Sun, 19 Jul 2009 18:30:32 -0700

Hi there,

I am trying to install the intel fortran compiler for linux version 11.1 on a suse 11.1 machine. This is my uname -a gives:
Linux linux-3pe2 2.6.27.7-9-default #1 SMP 2008-12-04 18:10:04 +0100 x86_64 x86_64 x86_64 GNU/Linux

The error the installation gives is:

Step no: 4 of 7 | Installation configuration - Missing Critical Pre-requisite
--------------------------------------------------------------------------------
The following required for installation commands are missing:
libstdc++.so.5 (library)

I did a whereis libstdc++.so.5 and this is what was given:
libstdc++.so: /usr/lib/libstdc++.so.6 /usr/lib64/libstdc++.so.5 /usr/lib64/libstdc++.so.6

I dont understand why the installation does not see where the files are. How to get the installation route to the right directory either x86 or x64?

Please help to solve this problem. Someone must have come across this.

Thanks
Mald

[Regression Intel Fortran 11.1] Bug report

Sun, 19 Jul 2009 17:35:05 -0700

The following part of our code does compile with ifort 11.1 (both MAC OS X and Linux) when compiling just with -c, but not with the -g -c flags. Also when compiling with -c only the produced module and object files seems to be incorrect, because our code crashed further on when calling the compiled module expressions.o.
The error when compiling with -g -c is:
ifort111: error #10105: /opt/ifort/intel/Compiler/11.1/038/bin/ia32/fortcom: core dumped
ifort111: error #10106: Fatal error in /opt/ifort/intel/Compiler/11.1/038/bin/ia32/fortcom, terminated by unknown signal(139)
compilation aborted for expressions.f90 (code 1)

Here is the code:
*****
module expressions

implicit none
private

type :: prt_t
private
integer :: type = 0
end type prt_t

type :: var_entry_t
private
integer, pointer :: ival => null ()
procedure(obs_binary_int), nopass, pointer :: obs2_int => null ()
type(var_entry_t), pointer :: next => null ()
end type var_entry_t

type :: var_list_t
private
type(var_entry_t), pointer :: first => null ()
end type var_list_t

abstract interface
function obs_binary_int (prt1, prt2) result (ival)
import
integer :: ival
type(prt_t), intent(in) :: prt1, prt2
end function obs_binary_int
end interface

contains

recursive function var_list_get_var_ptr (var_list) result (var)
type(var_entry_t), pointer :: var
type(var_list_t), intent(in), target :: var_list
var => var_list%first
end function var_list_get_var_ptr

end module expressions
*****

ifort 9 compiles, ifort 11 not

Sun, 19 Jul 2009 15:24:41 -0700

module_test.f90:
--------------------
module mod_test
implicit none
contains

real function add_sum (a,b,c)
real,intent(in) :: a,b,c
real :: nothi

add_sum = a + b + nothi(c)

end function add_sum

real function nothi (d)
real,intent(in) :: d

nothi = d

end function nothi

end module mod_test

main_test.f90:
-----------------
program test
use mod_test
implicit none

real :: a,b,c,res

a=1.
b=1.
c=1.

res = add_sum ( a, b, c )
write(*,*)res

end program test

********************************************
"ifort module_test.f90 main_test.f90" gives me:

/tmp/ifortvZIo3p.o: In function `mod_test_mp_add_sum_':
module_test.f90:(.text+0x21): undefined reference to `nothi_'

But if i compile with ifort version 9 produces an executable a.out.

Debugging NaN -- ISNAN() in debugger?

Sun, 19 Jul 2009 12:39:06 -0700

How can I create a breakpoint that tests for NaN?

I am using Visual Studio 2005 and Intel Fortran 10 to debug a legacy Fortran77 program in which elements in a large array (~100000 elements) inexplicably take on NaN values.

If I create a watchpoint condition such as ISNAN(x) to test whether the variable x is NaN, the message in the watch window is "Undefinded variable isnan."

Thanks.

Glenn

Very, Very new to Intel TBB

Sun, 19 Jul 2009 11:05:01 -0700

...so please be kind... I am a programmer with 25 years of experiance but clueless in here. I'm going to take the dive but what O.S. should I focus on? I have a Vista Ultimate machine, XP Pro machine, Windows Server 2003 and Linux. I would prefer to start on XP because the contract I am working on uses that environment. I may have to switch them to Linux 2.6 possibly down the road but that would cause a much longer development cycle and require my ultimate solution to involve changing over all systems in the field. Client has a multiple application app running on a quad core cpu system under XP at the moment. He is having throughput problems. Will applying TBB under XP kernel make any sense? Sorry if this has been addressed before and question is redundant, I havn't learned how to look around here yet but will do better soon. Thanks, Dennis in Camarillo CA.

GMA950 glShadeModel(GL_FLAT) bug

Sun, 19 Jul 2009 06:30:57 -0700

I've stumbled on a driver bug, where when using glShadeModel(GL_FLAT) the triangle color is not always taken from the last vertex as it should.

Here's how this looks with the Intel card:

... and the same thing with ATI and NVidia cards:

Executable (requires glut32.dll):

Source:

#include "stdafx.h"
#include <GL/glut.h>

void display(void)
{
	int colors[7] = { 0xff0000,0x00ff00,0x0000ff, 0xff00ff,0xff00ff,0xff00ff,0xff00ff };
	int indexes[18] = { 1,2,0,  2,3,0, 0,4,1, 4,5,1, 1,5,2, 5,6,2 };
	float verts[21] = { 0,-2,0, 0,0,0, -1,1,0, -1,-1,0, 1,-1,0, 1,1,0, 0,2,0 };

	glClear( GL_COLOR_BUFFER_BIT );

	glEnableClientState( GL_VERTEX_ARRAY );
	glEnableClientState( GL_COLOR_ARRAY );

	glColorPointer( 4, GL_UNSIGNED_BYTE, 0, colors );
	glVertexPointer( 3, GL_FLOAT, 0, verts );
	glDrawElements( GL_TRIANGLES, 18, GL_UNSIGNED_INT, indexes );

	glDisableClientState( GL_VERTEX_ARRAY );
	glDisableClientState( GL_COLOR_ARRAY );
	
	glFlush();
}

void init(void)
{
	glShadeModel( GL_FLAT );
	glMatrixMode( GL_PROJECTION );
	glLoadIdentity();
	glOrtho( -2, 2, -3, 3, -1, 1 );
}

int _tmain( int argc, char** argv )
{
	glutInit( &argc, argv );
	glutInitDisplayMode( GLUT_SINGLE | GLUT_RGB );
	glutInitWindowSize( 250, 250 );
	glutInitWindowPosition( 100, 100 );
	glutCreateWindow( "Fix me!!!" );
	init();
	glutDisplayFunc( display );
	glutMainLoop();
	return 0;
}

The same thing happens with normals and when using VBO.
Tested on an Asus eeepc 1000h, Windows XP, with drivers version 6.14.10.4906(the latest from Asus) and 6.14.10.4926
BTW, the VBOs on 4926 are very unstable. 4906 has no VBOs