Re: toolchain's output depends on toolchain used to build the compiler?

[Shachar Shemesh]

Omer Zak wrote:
I suggested before to check the compiler and linker flags which are
actually used.

Now I decided to check how to actually make the checks.

According to man gcc:
-v (verbose mode) reports the final form of the include path.
--help -v reports extra information which can be helpful.

I found that for example:
gcc --help -v 2>&1 | less

when run under Linux (Debian Lenny), displays - among lots of other
information - the flags passed
to /usr/lib/gcc/i486-linux-gnu/4.3.2/collect2 (my gcc version is 4.3.2).

DISCLAIMER:
I did not check the above when invoked by makefiles with custom
compile-time flags.  Therefore, YMMV.
  
Look, I really appreciate everyone trying to help, but it gets a bit 
frustrating to repeat.

The files that are different are different after I manually (so any 
question about compiler flags created by the make system are irrelevant) 
feeding the two compiler an identical output produced by gcc -E (so any 
question about file system paths is irrelevant).

Again, thank you for trying to help, but lets try to move forward.

Shachar
                                                   --- Omer

On Tue, 2010-02-09 at 22:08 +0200, Shachar Shemesh wrote:
  
Oleg Goldshmidt wrote: 
    
2010/2/9 Shachar Shemesh <shac...@shemesh.biz>
        
        The newlib libraries built are compared, and are identical
        down to the last assembly instruction. The client libraries
        are compared. Some are identical, some are not.
        

Just to dot all the i's and cross all the t's: are the
cross-binutils identical? 

      
As listed in the table that started the thread, they are compiled as
part of the compiler bring up. In other words, yes.

Shachar
    

  


--
Shachar Shemesh
Lingnu Open Source Consulting Ltd.
http://www.lingnu.com

_______________________________________________
Linux-il mailing list
Linux-il@cs.huji.ac.il
http://mailman.cs.huji.ac.il/mailman/listinfo/linux-il