Code rebase

Address comments and rebase the code

In D12037#233027, @cedric wrote:

As long as expedite test pass and everyone comments has been addressed, I think this is good to go. The thread API is not used by EFL in the hot path, so that part is not the important bit. Maintenance and fitting into the host OS is more important.

I don't think that even if you run expedite 20 times with your CPU power saving disable you would anything.

In D12037#233015, @walac wrote:

In D12037#233003, @vtorri wrote:

have you tested with expedite and the fractal code ?

iirc, I couldn't compile the fracal app. The expedite I had to hijack some files to compile. I remember I didn't find any significant difference. Anyway, I have to rebase the code and will test it again.

In D12037#233003, @vtorri wrote:

have you tested with expedite and the fractal code ?

Fix typos in the comments

Fix minor compatibility corner cases with the posix counterpart.

In D12041#229807, @bu5hm4n wrote:

Patch does not apply ...

Fix patch to apply to HEAD

Add tests and fix bugs

Address raster's review comments

Fix undefined behavior in eina_thread_self

Use native windows threads in mingw

Address review comments

Fix struct layout compatibility

In D12022#229393, @vtorri wrote:

i've looked at https://docs.microsoft.com/en-us/windows/win32/fileio/naming-a-file#fully-qualified-vs-relative-paths and i indeed see it can begin with \ (but not /) :

"A file name is relative to the current directory if it does not begin with one of the following etc... "
"A single backslash, for example, "\directory" or "\file.txt". This is also referred to as an absolute path."

i don't know if you should also consider /

Update the patch based on reviewer feedback

Fix typo

Fix patch update

Fix negative path length calculation

In D11993#228976, @vtorri wrote:

@walac your prefer this patch which modifies 143 files, ONLY for eina, to my patch, which modifies 5 files for 4 libraries, with the same result ?? that's crazy

In D11993#228977, @vtorri wrote:

do you realize that, with this kind of changes, you are going to change almost all the files of the EFL ?

That patch seems a step towards in the right direction IMHO. Things get tricky when EFL try to use each other since, in some places, EAPI must be __declspec(dllimport) and in others __declspec(dllexport), which leads walking on thin ice by playing around with the preprocessor. I understand that each library should define its XXX_API macro. The static library support is a win, but not the main point of the patch.

In D11970#228390, @stefan_schmidt wrote:

Gettign of a tons of errors with this patch and cxx bindings enabled. One example (there are more):

[219/398] Compiling C++ object 'src/tests/eina_cxx/6b3491e@@eina_cxx_suite@exe/eina_cxx_suite.cc.o'
In file included from ../src/tests/eina_cxx/eina_cxx_suite.h:8,
                 from ../src/tests/eina_cxx/eina_cxx_suite.cc:20:
../src/tests/eina_cxx/../efl_check.h: In function ‘int _efl_test_option_disp(int, char**, const Efl_Test_Case*)’:
../src/tests/eina_cxx/../efl_check.h:137:23: warning: ISO C++ forbids converting a string constant to ‘char*’ [-Wwrite-strings]
  137 |              nav[0] = "valgrind";
      |                       ^~~~~~~~~~
../src/tests/eina_cxx/../efl_check.h:138:23: warning: ISO C++ forbids converting a string constant to ‘char*’ [-Wwrite-strings]
  138 |              nav[1] = "--trace-children=yes";

Make C++ compiler happy

@vtorri done

Replace malloc by alloca

Remove all references to errno in eina_test_file.c