Stupid error, crazy performance boost

[Dann Corbit]

All the compile errors complain that you have defined a backtrace() that returns void, it seems. It complains on this line over and over again:

Code: Select all

             (*_os) << backtrace() << std::endl;

Nothing illegal about that, as long as backtrace() returns something that can be printed (ie., not void).

Won't help for Windows since backtrace() is a Linux system call.
http://man7.org/linux/man-pages/man3/backtrace.3.html

There is similar functionality in Windows:
https://msdn.microsoft.com/en-us/librar ... 85%29.aspx
but the API is different, so some work would be required to do the same thing.

[xr_a_y]

ok thank you all.

If you read the code carefully, you will see that the #ifdef for Windows is ok (and it indeed works every days for me under visual studio 2017) but what you call "windows" is cygwin, and there the backtrace function I defined is empty and return void. It should at least return an empty string. Like this :

Code: Select all

#ifndef __MINGW32__

#include "dbg_win.h"

inline
std::string backtrace() {
    std::stringstream buff;
    std::vector<windbg::StackFrame> stack = windbg::stack_trace();
    buff << "Callstack: \n";
    for (unsigned int i = 0; i < stack.size(); ++i)	{
        buff << "0x" << std::hex << stack[i].address << ": " << stack[i].name << "(" << std::dec << stack[i].line << ") in " << stack[i].module << "\n";
    }
    return buff.str();
}

#else // __MINGW32__
inline
std::string backtrace() {
    ///@todo
    return "";
}
#endif // __MINGW32__

[Dann Corbit]

The first branch probably works for MSYS2/mingw but not for Cygwin.

[xr_a_y]

Anyway, why not using visual studio instead ? It is such a great IDE and debugger.

[Sesse]

Won't help for Windows since backtrace() is a Linux system call.
http://man7.org/linux/man-pages/man3/backtrace.3.html

It will certainly help with the template errors. You won't necessarily get backtraces, of course.

In any case, backtrace() without arguments doesn't exist on Linux either. So this is certainly some kind of wrapper with an overloaded name.

[xr_a_y]

Won't help for Windows since backtrace() is a Linux system call.
http://man7.org/linux/man-pages/man3/backtrace.3.html

It will certainly help with the template errors. You won't necessarily get backtraces, of course.

In any case, backtrace() without arguments doesn't exist on Linux either. So this is certainly some kind of wrapper with an overloaded name.

it is indeed, as backtrace from libc is taking arguments and returns size_t (https://www.gnu.org/software/libc/manua ... races.html)
mine is not and returns std::string.

Hopefully, backtrace() shall be needed only for debug purpose in case of a fatal error, so it is mostly useless for the end-user.

I'll probably push a new release to github tonight to fix this.

[Sven]

it also has lots of template substitutions that are not legal.
No way to compile it under Windows.

Can you elaborate a little about this template issue please ?

On the Github page you have linked to, I only see a README file, was that intended?

Sorry for bumping up my unanswered question, but where are you guys seeing Weini source code? I tried the Github link from Vivien's signature but there I can't find source code. Is it public at all (Dann tried to compile it so it must be available somehow)?

[Guenther]

it also has lots of template substitutions that are not legal.
No way to compile it under Windows.

Can you elaborate a little about this template issue please ?

On the Github page you have linked to, I only see a README file, was that intended?

Sorry for bumping up my unanswered question, but where are you guys seeing Weini source code? I tried the Github link from Vivien's signature but there I can't find source code. Is it public at all (Dann tried to compile it so it must be available somehow)?

Sven, the source is for download under 'releases'.

[Sven]

it also has lots of template substitutions that are not legal.
No way to compile it under Windows.

Can you elaborate a little about this template issue please ?

On the Github page you have linked to, I only see a README file, was that intended?

Sorry for bumping up my unanswered question, but where are you guys seeing Weini source code? I tried the Github link from Vivien's signature but there I can't find source code. Is it public at all (Dann tried to compile it so it must be available somehow)?

Sven, the source is for download under 'releases'.

Thanks, got it, didn't expect it there ...

[xr_a_y]

You probably should download the 0.0.16 again, I've just update it (it was a bad merge) without changing the tag name.