On Fri, 2014-06-20 at 23:15 +0200, Petr Machata wrote: > ... instead of inlining equivalent code.
I didn't know we had that already. I was just contemplating whether to add something like that. It also gets rid of the DW_TAG_mutable_type issue (and in a nicer way than I did). > +2014-06-20 Petr Machata <[email protected]> > + > + * alpha_retval.c (alpha_return_value_location): Call > + dwarf_peeled_die_type instead of inlining equivalent code. > + * arm_retval.c (arm_return_value_location): Likewise. > + * i386_retval.c (i386_return_value_location): Likewise. > + * ia64_retval.c (ia64_return_value_location): Likewise. > + * ppc64_retval.c (ppc64_return_value_location): Likewise. > + * ppc_retval.c (ppc_return_value_location): Likewise. > + * s390_retval.c (s390_return_value_location): Likewise. > + * sh_retval.c (sh_return_value_location): Likewise. > + * sparc_retval.c (sparc_return_value_location): Likewise. > + * tilegx_retval.c (tilegx_return_value_location): Likewise. > + * x86_64_retval.c (x86_64_return_value_location): Likewise. Looks good. I'll rebase and push my 'Remove non-existing DW_TAG_mutable_type' patch after this goes in. Thanks, Mark
