Re: [Mingw-users] S_IRWXG, S_IRWXO undefined when building gcc 4.9.2 with mingwrt 3.20

[Earnie <ea...@us...>]

> From: David Gressett
> 
> I had previously successfully built gcc 4.9.2 ; I am now trying it again
after having
> upgraded to the mingwrt 3.21 and I am having problems.
> 
> I get through a large part of the build, but shortly after generating
13,000 lines
> in the log file, the xgcc compiler chokes on this file:
> 
> gcc-4.9.2/libgfortran/io/unix.c
> 
> at line 1133.
> 
> The offending line is
> 
> mode_mask = umask (S_IXUSR | S_IRWXG | S_IRWXO);
> 
> The compiler has no problems with S_IXUSR but cannot find definitions for
> S_IRWXG and S_IRWXO.
> 

Those bits don't exist in Windows.


> S_IXUSR is defined in  sys/stat.h and I would expect  theother two to be
defined
> there as well, but they are not.
> 
> A bit of searching finds both of the missing items are defined  in a
directory fairly
> deeply buried in the msys directory structure, but the definitions there
look very
> unixish and won't do for use in the MinGW sys/stat.h.
> 

Not for Windows consumption.

> A search for the missing definitions in the source code for the previous
version
> of mingwrt turns up nothing.
> 
> The offending line of code is protected inside an #ifdef HAVE_UMASK.
> 
> This leads me to believe that something in mingwrt 3.21 is causing the gcc
4.9.2
> build to be misconfigured and it is causing HAVE_UMASK to be defined  when
it
> should not be, so that the build is compiling code that should not be
compiled
> for Windows. I should file a ticket for this, but I would like to be able
to propose
> a patch when I do that.
> 

Find the pieces of config.log to tell you why.

> Does anybody have any idea  what could be causing this? The next step is
to
> compare the present and previous versions of mingwrt, but I don't want to
> waste time reinventing wheels that someone else already knows about.
> 

What are your configure arguments?

--
Earnie