Diagnosing build errors

[Ralf Corsepius]

On Tue, 2008-04-15 at 14:59 +1200, Nigel Spon wrote:
> I am now building the toolchain and RTEMS 4.8 successfully (I think)  
> from Xcode under MacOS 10.5, which is nice because it gives me an IDE  
> and I don't have to use the command line anymore. However, one  
> component of the system has proved resistant to this treatment.
> 
> The gdb 6.6 build fails at the final 'make install' stage. The last  
> few lines of the log say:
> 
> make[3]: Nothing to be done for `install'.
> make[2]: Nothing to be done for `install'.
> /bin/sh /Users/frederickspon/Documents/RTEMSX_4.8/RTEMS_Xcode/../tools/ 
> gdb-6.6/sim/ppc/../../mkinstalldirs /Users/frederickspon/Documents/ 
> RTEMSX_4.8/RTEMS_Xcode/Obj/bin
> n=`echo run | sed 's,^,powerpc-rtems4.8-,;'`; \
> 	/usr/bin/install -c run /Users/frederickspon/Documents/RTEMSX_4.8/ 
> RTEMS_Xcode/Obj/bin/$n
> install: run: No such file or directory

>  I'm not sure where to look next to figure out what happened  
> here.
This breakdown very likely is a followup to something having gone wrong
before. Check your build log for why "run" doesn't seem to have been
built. 

>  Its rather annoying to have it get so close to completing and  
> then fail;
There are several major differences between gdb and binutils/gcc:

* gdb has a significantly smaller user-base (i.e. is significantly less
robust to build). It's configuration is quite bugged.
* gdb has many more dependencies on other packages.

I would guess your issues are a mixture of both, probably missing one or
more dependencies which causes gdb's configuration to trip into a bug.

>  I presume it was about to copy the results into the /bin  
> directory.
Almost. It was about to install "run" as
"$(bindir)/powerpc-rtems4.8-run"
i.e. to copy and rename it at the same time.

BTW: I'd recommend to start with gdb-6.8. It's configuration has seem
some significant improvements, appears much robust to build and isn't
technically much different from gdb-6.6.

Ralf