HI,
could any one explain or describe how to force reboot a CE5300 board?
It gets stuck: on any valid CF9 I know and also using "reboot -f". Is there any other valid sequence to force reboot this SoC, but not included in SDK I have?
Kind regards
Marcin
We have identified what we consider to be a compiler bug. Verified with the latest version of the compiler (2013 SP1) just now.
A simple code sample to illustrate the issue can be found at this address:
https://gist.github.com/lemire/6642148
and I also include it below.
It suffices to compile and run the code as directed. A print-out will then show the result of the computation vs. the expected result. The function that triggers the issue has only 7 (simple) lines. It appears that the keyword "restrict" is needed for the bug to appear.
We checked the assembly generated by the compiler and it makes no sense to us.
Note that the same code sample was tested successfully with several compilers including gcc.
Context: we identified this bug while working on a fast integer compression library (https://github.com/lemire/FastPFor). After getting the library to pass all tests with clang, gcc, VS2012... the intel compiler gave us grief. The code sample is the simplest case we could come up with to trigger the bug. To get around the bug, we simply manually unrolled the loop. This is obviously not very desirable in general.
// compile with:
// icc -std=c99 -O2 iccbug.c -o iccbug
// then run iccbug
// Tested in Linux Ubuntu 12.10 (Intel Core i7)
#include <stdint.h>
#include <stdio.h>
// expect: out[0] = in[0] + in[1] + in[2] + in[3];
// out[1] = in[4] + in[5] + in[6] + in[7];
__attribute__((noinline))
void broken_with_O2(int * restrict in, int * out) {
for(int outer = 0; outer < 2; outer++) {
*out = *in++;
for (int inner = 1; inner < 4; inner++) {
*out += *in++;
}
++out;
}
}
int main() {
int in[8] = {1,1,1,1,1,1,1,1};
int out[2];
broken_with_O2(&in[0],&out[0]);
printf(" got = %d %dn",out[0], out[1]);
printf(" expected = %d %dn", 4,4);
}
I seem to be unable to get programs using std::random to work.I have been using icpc version 13.1.3
The follwong example (taken off the web) fails to compile with the command:icpc rand_test.cpp -std=c++11
The errors reported are:
rand_test.cpp(8): error: namespace "std" has no member "uniform_real_distribution"
std::uniform_real_distribution<> dis(1, 2);
^
rand_test.cpp(8): error: expected an expression
std::uniform_real_distribution<> dis(1, 2);
^
rand_test.cpp(10): error: identifier "dis" is undefined
std::cout << dis(gen) << '';
Here's rand_test.cpp :
#include <random>
#include <iostream>
int main()
{
std::random_device rd;
std::mt19937 gen(rd());
std::uniform_real_distribution<> dis(1, 2);
for (int n = 0; n < 10; ++n) {
std::cout << dis(gen) << '';
}
std::cout << '\n';
}
Does anybody have any suggestions ? Thanks,
- Amartya
When switching from the Intel C++ 13 compiler 14.0 I started seeing some stack corruption errors, after narrowing it down to a minimalist example I can not see that it is anything other than a new compiler bug. In the attached Visual Studio 2012 project the stack corruption only occurrs if you place a breakpoint in line 7 of TurtleParser.cpp and only with the Intel C++ 14.0 compiler. If you try the example with Intel Compiler C++ 13 or the Visual Studio compiler it works fine. Also, the stack corruption does not happen if line 14 of TurtleParser.cpp is commented out, even though this line of code is never called by any code.
Tested on Microsoft Visual Studio Professional 2012 Version 11.0.60610.01 Update 3.
Can others confirm?
| Attachment | Size |
|---|---|
Download turtle.zip | 30.78 KB |
Now, I agree that I am asking for trouble, but I find the following failure weird. The following code:
#include <iostream>
using namespace std;
template<typename ... Pack>
void weeble(Pack ... rest, double x) {
int y[] = {rest...};
for (int i = 0; i < sizeof(y)/sizeof(*y); ++i) cout << y[i] << "";
cout << x << endl;
}
int main () {
weeble(123,456,789,3.1416);
}
When compiled with icpc -std=c++11 and the beta compiler, gives:
junk.cpp(12): error: no instance of function template "weeble" matches the argument list
argument types are: (int, int, int, double)
weeble(123,456,789,3.1416);
^
I cannot find anything forbidding it in C++11, and it is parsable (in theory).
If function parameter packs must be final, why did I not get an error message when the function was parsed?
Compiler XE 14 and boost 1.54
errors:
error #373: "boost::asio::detail::noncopyable::noncopyable(const boost::asio::detail::noncopyable &)" (declared at line 32 of "path removed\boost/asio/detail/noncopyable.hpp") is inaccessible path removed\boost\asio\basic_streambuf.hpp
error : class "boost::asio::basic_socket_iostream<boost::asio::ip::tcp, boost::asio::stream_socket_service<boost::asio::ip::tcp>, boost::posix_time::ptime, boost::asio::time_traits<boost::posix_time::ptime>, boost::asio::deadline_timer_service<boost::posix_time::ptime, boost::asio::time_traits<boost::posix_time::ptime>>>" has no member "read_some" path removed\boost\asio\impl\read.hpp
Please explain.
The following code fails to link when using icpc 14.0.0 and boost 1.54 or 1.53:
#include <boost/lexical_cast.hpp>
#include <string>
int main()
{
const float f(0.123f);
boost::lexical_cast<std::string>(f);
return 0;
}
Commmand line:
icpc -std=c++11 -I boost_1_54_0/ icpc_14_builtin_signbit_failure.cpp
Output:
/tmp/icpcZV9IvT.o: In function `main':
icpc_14_builtin_signbit_failure.cpp:(.text+0x65): undefined reference to `__builtin_signbit'
icpc_14_builtin_signbit_failure.cpp:(.text+0xa8): undefined reference to `__builtin_signbit'
It compiles fine if -std=c++11 is not specified. It also compiles fine with icpc 13.1.3 and -std=c++11 defined.
I have attached a file with the complete source for reproduction.
This is a showstopper for upgrading to the new compilers.
| Attachment | Size |
|---|---|
| Downloadicpc-14-builtin-signbit-failure.tbz2 | 1.23 MB |
Hi,
There seems to be an incompability between Intel Compiler 14.0.1 and the "fenv.h" header in the GNU Standard C++ Library provided with GCC 4.8.1.
Here is an example code :
[mboisson@r103-n2 tmp]$ cat test_fenv.cpp
#include <fenv.h> fexcept_t data;
[mboisson@r103-n2 tmp]$ echo $CPLUS_INCLUDE_PATH
[mboisson@r103-n2 tmp]$ icpc -c test_fenv.cpp
#### compiles fine without the gnu stdlibc++
[mboisson@r103-n2 tmp]$ export CPLUS_INCLUDE_PATH=/software6/compilers/gcc/4.8.1/include/c++/4.8.1/
[mboisson@r103-n2 tmp]$ icpc -c test_fenv.cpp
test_fenv.cpp(3): error: identifier "fexcept_t" is undefined fexcept_t data; ^ compilation aborted for test_fenv.cpp (code 2)
#### does not compile either when specifying std=c++11
[mboisson@r103-n2 tmp]$ icpc -std=c++11 -c test_fenv.cpp
In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(58): error: the global scope has no "fenv_t" using ::fenv_t; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(59): error: the global scope has no "fexcept_t" using ::fexcept_t; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(62): error: the global scope has no "feclearexcept" using ::feclearexcept; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(63): error: the global scope has no "fegetexceptflag" using ::fegetexceptflag; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(64): error: the global scope has no "feraiseexcept" using ::feraiseexcept; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(65): error: the global scope has no "fesetexceptflag" using ::fesetexceptflag; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(66): error: the global scope has no "fetestexcept" using ::fetestexcept; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(68): error: the global scope has no "fegetround" using ::fegetround; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(69): error: the global scope has no "fesetround" using ::fesetround; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(71): error: the global scope has no "fegetenv" using ::fegetenv; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(72): error: the global scope has no "feholdexcept" using ::feholdexcept; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(73): error: the global scope has no "fesetenv" using ::fesetenv; ^ In file included from test_fenv.cpp(1): /software6/compilers/gcc/4.8.1/include/c++/4.8.1/fenv.h(74): error: the global scope has no "feupdateenv" using ::feupdateenv; ^ test_fenv.cpp(3): error: identifier "fexcept_t" is undefined fexcept_t data; ^ compilation aborted for test_fenv.cpp (code 2)
I migrated my Leopard system to Mavericks today, and now I'm running into compile/link problems on my existing projects (built with XCode 3.2.5).
The errors include missing "cxxabi.h" (from <stdexcept>) during compile, and missing "libstdc++" during link.
I've successfully migrated from previous OS releases before with no problems, so I'm wondering if this is fixable?
Unfortunately, I don't have my C++ compiler installer anymore, and I can't find a download link for it, so I can't try just reinstalling the Intel compiler. Anyone have a link?
I may upgrade to the latest 2013 compiler though, if it still works with Xcode 3.x.
I want to know when Intel® C++ Studio XE for Linux for ubuntu 13.10 will be released? Non-commericial version.
thanx
-tony
Hi, first of all, sorry if i'm posting this in the wrong forum. I'm currently trying to install the Intel C++ Studio XE for Windows on my computer. Unfortunately after unpacking the offline installation file i am getting the following error message "ERROR: The installation is terminated because of unknown error. Suggestions: visit the Support Web Site for a solution to this problem.".
So here i am. Can you help me? thank you in advance for the help.
| Attachment | Size |
|---|---|
| Downloadwellintel.png | 20.54 KB |
Hi, first of all, sorry if i'm posting this in the wrong forum. I'm currently trying to install the Intel C++ Studio XE for Windows on my computer. Unfortunately after unpacking the offline installation file i am getting the following error message "ERROR: The installation is terminated because of unknown error. Suggestions: visit the Support Web Site for a solution to this problem.".
So here i am. Can you help me? thank you in advance for the help.
| Attachment | Size |
|---|---|
| Downloadwellintel.png | 20.54 KB |
Trying to download the OS X installer all morning. It always comes down corrupt. Have tried several browsers and also from my Windows machine. What's up?
Trying to download the OS X installer all morning. It always comes down corrupt. Have tried several browsers and also from my Windows machine. What's up?
I noticed there is not user-level intrinsic ‘_mm512_sincos_ps’ or ‘_mm512_mask_sincos_ps’ defined in zmmintrin.h.
However, I have just found out that Intel compiler is emitting ‘__svml_sincosf16’ and ’ __svml_sincosf16_mask’ when it autovectorises code and finds ‘cos’ and ‘sin’ operations on the same value.
I have been doing some tests and if I define my own ‘_mm512_sincos_ps’ at user-space level, Intel compiler recognises it and translates it into the appropriate ‘__svml_sincosf16’. However, the result of my code is incorrect, maybe because the parameters of my function are not the expected.
Could anyone please tell me why ‘_mm512_sincos_ps’ has not been defined in zmmintrin.h and what the expected parameters are so that I can define it appropriately ?
Thank you in advance.
Best regards.
(Using icc 14.0.1)
I’m trying to find some source code to implement EGD (Ethernet Global Data) communications in Ansi C. I’m trying to communicate a computer which runs Ansi C code with a PLC which communication protocol is EGD. Does anybody know/have any C source code that might help?
If anyone could help me, it would be fully appreciated!
Thanks!
Jesus
5 October 2013 - Intel released Intel(R) Composer XE 2013 SP1 for OS* X, Update 1 today, 25 October 2013. This compiler and all subsequent compilers will support Xcode 5 on Mavericks and Mountain Lion. Also, core Mavericks support is first added to this Composer XE 2013 SP1 Update 1 compiler.
WHERE? All product downloads are available to customers with current support. Log in to Intel Registration Center, https://registrationcenter.intel.com
CONFUSED about names and versions? Go here: http://software.intel.com/en-us/articles/intel-compiler-and-composer-upd...
QUESTIONS: Ask questions on this Forum.
Hello everyone,
We’ve been working on creating some sample applications using Intel® Cilk Plus technology. So far we have published 10 samples. You can find them at Intel® C++ Compiler Code Sample.
All the samples are published under the same book. You can easily navigate the samples from the left panel. Each sample page contains the abstract and code snippets to show you how we applied the Intel Cilk Plus. It also includes the associated performance data collected.
The complete source code including project files or makefiles is available from each page. Each sample also contains a “Readme.htm” that is almost identical to the sample page including build instructions.
Check it out and hope you will find it useful. Any comments or feedbacks are welcome!
Go parallel !
Jennifer
Hello everyone,
Today we are introducing a new product called Intel® C++ Compiler 14.0 for Android*. It can be hosted on Windows*, OS X*, or Linux*. You can find more info at http://software.intel.com/en-us/c-compiler-android.
The previous release – the limited-time, free version of the Intel® C++ Compiler 13.0 for Android* OS – is no longer available.
For emphasis, the version available today is 14.0. It is available in 3 separately downloadable packages, one each for Windows, OS X, and Linux. You can download evals or purchase the compiler at the above link as well.
Support is offered through this forum. If you have any issues, suggestions, comments, etc., please let us know.
Thank you !
Jennifer Jiang
Hi,
i installed the SP1 update 1 onto my XCode 5.0.2 and reworked the project settings of my existing projects.
i had added the command line option -fasm-block, because my code is using some assambler code.
however, during compilation, the compiler exists with an errcode 100, file/directory "/XCode/Developer/usr/bin/as" not found.
i just copied the file from my older xcode installation( XCode 4.4.2, Intel XE 12.1 Update 13 )
i also had to copy the folder /XCode/Developer/usr/libexec/as
now it compiles but i get the same threading errors like in 12.1 where the assambler read a file that is currently beeing modified by the compiler
ending up with undefined symbols, or unexpcted end of line in a ".s" file ( dont know the name currently)
i am sure that this is a problem of the wrong version of the as file.
So why is "as" missing?
Rene