[PATCH] D33673: [Object/RelocVisitor] - Add support for R_X86_64_DTPOFF32 relocation.

[Rafael Avila de Espindola via llvm-commits]

George Rimar <grimar at accesssoftek.com> writes:

>>One interesting thing to note is that we don't actually want to relocate
>>these sections. What we want is to read the relocation, in a similar
>>way that lld does on the scanRelocs. For example, when reading
>>DW_AT_low_pc, we don't want the final value, we want what it points to
>>(section and offset). With elf_rela we can find that with no knowledge about
>>relocations and with elf_rel we only need to know how to read the
>>implicit addend.
>>
>>Cheers,
>>Rafael
>
> Yes, for .debug_ranges we do not have to perform relocations and just need
> to know the section index.
> Though in general it is not clear that it worth to optimize here (we still need to scan relocations).
>
> While we do relocations on DWARF parsers side, LLD code remains clear. Each additional optimization for .gdb_index
> will make it more specific. Generally observing MSVS profiling reports I can't say we can win much here.
> Sure relocations are generally expensive, but we may probably want to focus on something not so .gdb_index specific,
> as far I understand after last commits we are "slow" just because parsers are too generic. (though .gdb_index generation atm is much faster than it was few month ago).

Yes, this is a case that needs some benchmarking before making any other
decision.

Cheers,
Rafael