[3/3] RFC - optional patch to restore DW_AT_data_member_location

Jan Kratochvil jan.kratochvil@redhat.com
Wed Sep 12 13:25:00 GMT 2012


On Tue, 11 Sep 2012 22:56:11 +0200, Tom Tromey wrote:
> >> +	.string	"GNU Fortran 4.6.3 20120306 (Red Hat 4.6.3-2) -mtune=generic -march=x86-64 -g -fintrinsic-modules-path /usr/lib/gcc/x86_64-redhat-linux/4.6.3/finclude"
> 
> Jan> Probably not needing to be rebuilt but I think for FSF GDB it would
> Jan> be better to precompile such .S files using FSF GCC, which could be
> Jan> possibly (more easily) reproduced in the future.
> 
> I had to edit the .S file.  I would not want to try to reproduce this.
> I guess if I was going to do it again I would try to use my DWARF
> assembler instead.
> 
> I changed the string to
> 
> "GNU Fortran 4.6.3 plus hand hackery"
> 
> ... which is reasonably true.

I see I did not make clear why I like the reproducibility of the compilation.
This is because sometimes it is good to know what was hand-edited in the file
and I find a bit excessive to keep both the original and the hand-edited .S
files in the repository due to it.  This is why I sometimes
did #if 0-#else-#endif or #if 0-#endif-#if 1-#endif constructs in the .S
files (I see it now only in gdb.dwarf2/dw2-const.S).

I sure do not insist on anything like that, just FYI.


Thanks,
Jan



More information about the Gdb-patches mailing list