[CMake] Re: Troubles with the Nmake generator...

Shishir Ramam sramam at gmail.com
Sun May 29 12:29:07 EDT 2005


This looks to be a problem with the parser used for dependency determinations.
One of my source files had a comment of the form 

/* this is  a comment to mimic trouble
   #include<...> and we should have it! */

Preliminary findings indicate that '#include' inside the comment caused the 
spurious dependency on directory '...'

I have fixed my source code to work around, but this is rather time consuming 
to find and localize. Hopefully it will be fixed in the parser soon. 

thanks,
-shishir


On 5/29/05, Shishir Ramam <sramam at gmail.com> wrote:
> Hi,
> 
> I have been attempting to use the nmake generator, and the build fails.
> From the preliminary investigations, it looks like the problem occurs in -
> cmMakeDepend::DependWalk(..)
> on the attempt to open the file using 'info->m_FullPath.c_str()'
> 
> The problem seems to be that the path in this case has a '...'
> at the end, which the OS obviously fails to open.
> 
> Moreover, it looks like this is a problem only for binary directories.
> It's *not* an in-place build.
> 
> I am starting to look a little further into this, but would appreciate
> any help that I can get in localizing the problem - or atleast where to look.
> 
> thanks,
> -shishir
> 
> ps: I am using Cmake 2.0.6 on WinXp. The same cmake setup works
> well with a VC2005 generator. I have not been successful in making it
> work for mingw/make. But that might be problems with my setup. Haven't
> done too much of digging around.
> 
> 
> --
> The man who doesn't read good books has no advantage over the man who
> can't read them. - Mark Twain
> 


-- 
The man who doesn't read good books has no advantage over the man who
can't read them. - Mark Twain


More information about the CMake mailing list