Home Original page

[PATCH v2] LoongArch: Set e_flags to 0x40 of ELF while using new relocations.

Fangrui Song maskray@google.com
Tue Sep 27 16:49:33 GMT 2022 
On 2022-09-25, Xi Ruoyao wrote:
>On Sun, 2022-09-25 at 16:28 +0800, WANG Xuerui wrote:
>> > +
>> > +         if (elf_elfheader (output_bfd)->e_ident[EI_CLASS] == ELFCLASS32)
>> >             interpreter = "/lib32/ld.so.1";
>> > -         else if (EF_LOONGARCH_IS_LP64 (flags))
>> > +         else if (elf_elfheader (output_bfd)->e_ident[EI_CLASS] == ELFCLASS64)
>> >             interpreter = "/lib64/ld.so.1";
>> >           else
>> >             interpreter = "/lib/ld.so.1";
>> > +
>> Wait what, are we still carrying the old-world defaults for the
>> interpreter path here? I know it's harmless because in fact it gets
>> overridden by gcc and glibc, but I'd suggest fixing it in a separate
>> patch later for consistency.
>
>It seems not easy to make a "correct" default...  /lib64/ld-linux-
>loongarch-lp64d.so.1 will be incorrect for a musl-based target.  Maybe
>we can assume Glibc is used because we are the *GNU* linker anyway...
>
>/* snip */

For all other architectures, the default dynamic linker is just not
useful. (ld.lld encodes no string at all.) Ideally this should be empty,
but if an empty interpreter string uses a behavior similar to --no-dynamic-linker,
use an arbitrary dummy string.

>> I'd bikeshed a little and suggest "EF_LOONGARCH_OBJABI_foo" because
>> it's actually "ELF object file ABI v1" and not "ELF ABI object v1".
>
>In the doc we say "the ABI version of an ELF object", IMO it's a little
>misleading (can trick people to believe it could be used to distinguish
>old and new world).  From Wikipedia:
>
>   In computer software, an application binary interface (ABI) is an
>   interface between two binary program modules.
>
>I'm not sure if the (non-dynamic) relocation types belongs to the concept
>of ABI: the linker treats relocatable object files as an input (like how
>"vim" treats a text file), instead of a program module.  But I can't come
>up with a better terminology for now...
>
>-- 
>Xi Ruoyao <xry111@xry111.site>
>School of Aerospace Science and Technology, Xidian University
More information about the Binutils mailing list