Re: [PATCH] dwarf2out: Emit DWARF 6 DW_AT_language_{name,version}

[Richard Biener via Gdb <gdb@sourceware.org>]

On Wed, 8 Jan 2025, Jakub Jelinek wrote:

> On Wed, Jan 08, 2025 at 09:14:59AM +0100, Eric Botcazou wrote:
> > > So, this patch is an alternative to the
> > > https://gcc.gnu.org/pipermail/gcc-patches/2024-November/669671.html
> > > patch, which had the major problem that it required changing all the
> > > DWARF consumers to be able to debug C17 or later or C++17 or later
> > > sources.
> > 
> > Do you plan to salvage the non-obsoleted parts of the above change?
> 
> No.  The switches on DW_AT_language value are a GCC internal thing,
> if we never generate say the post-DWARF5 DW_LANG_C_plus_plus_23,
> we don't need to handle it in the switch.
> 
> Though, Ada and Fortran could have a similar change to the C/C++
> one, i.e. also add DW_AT_language_{name,version} for newer Ada and Fortran
> versions, say DW_LNAME_Ada and then dunno whether 2005, 2012 and 2022
> or 2007, 2012 and 2023, in https://dwarfstd.org/languages-v6.html
> the versioning scheme for Ada (as well as Fortran) is YYYY.
> Similarly DW_LNAME_Fortran 2018 and 2023.
> 
> The reason I haven't done it myself is that the Ada FE doesn't tell
> Ada version at all - there is just "GNU Ada" and dwarf2out.cc right now
> implies it is Ada 95 for DWARF3 and Ada 83 otherwise.
> And in the Fortran case, while the FE provides a version, it only
> does so for Fortran 2003 and 2008 (and just "GNU Fortran" implies Fortran 90
> in dwarf2out).  So, no marking of Fortran 2018 or Fortran 2023.

Btw, with DWARF only we now have the opportunity to add langhooks that
directly create DWARF DIEs or attributes (for early debug).

Richard.