Re: [PATCH] Fix seg fault with --write PR gdb/20948

[Simon Marchi <simark at simark dot ca>]

On 2018-03-13 18:43, Jozef Lawrynowicz wrote:
> Hi Simon,
>
> I missed the example in comment #3, you are right there is a general 
> issue with
> BFD rather than this being GDB specific.
>
> It seems that the internal representation of SHSTRTAB 
> (elf_shstrtab(bfd)) does
> not get initialized or populated when opening an existing BFD. This is 
> not an
> issue when only reading, as this specific internal representation is 
> not
> required. It is also not an issue (or at least doesn't cause any 
> errors) when
> only writing as a new shstrtab is created. But when reading and 
> writing,
> elf_shstrtab needs to be populated using the existing elf data in the 
> loaded
> BFD and this doesn't happen at the moment.
>
> I'm working on a patch.
>
> I reckon when the BFD issue is fixed, we may not need any specific 
> patch for
> GDB.
>
> Thanks,
> Jozef

Ok, thanks for looking into it.

Simon