Re: [Xen-devel] [PATCH 3/4] x86/linker: add a reloc section to ELF binary

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 19.06.19 at 17:06, <roger.pau@xxxxxxxxxx> wrote:
> On Wed, Jun 19, 2019 at 06:57:05AM -0600, Jan Beulich wrote:
>> >>> On 19.06.19 at 13:02, <roger.pau@xxxxxxxxxx> wrote:
>> > If the hypervisor has been built with EFI support (ie: multiboot2).
>> > This allows to position the .reloc section correctly in the output
>> > binary, or else the linker might place .reloc before the .text
>> > section.
>> > 
>> > Note that the .reloc section is moved before .bss for two reasons: in
>> > order for the resulting binary to not contain any section with data
>> > after .bss, so that the file size can be smaller than the loaded
>> > memory size, and because the data it contains is read-only, so it
>> > belongs with the other sections containing read-only data.
>> 
>> While this may be fine for ELF, I'm afraid it would be calling for
>> subtle issues with xen.efi (i.e. the PE binary): There a .reloc
>> section is generally expected to come after "normal" data
>> sections.
> 
> OK, would you like me to leave the .reloc section at the previous
> position for EFI builds then?

Well, this part is a requirement, not a question of me liking you
to do so.

> Or do we prefer to leave .reloc orphaned in the ELF build?

Daniel might have an opinion here with his plans to actually
add relocations there in the non-linker-generated-PE build. I
don't have a strong opinion either way, as long as the
current method of building gets left as is (or even simplified).

Also a remark regarding the title - in my builds there already is
a .reloc section in the ELF images, so "add" doesn't really seem
correct to me. It sits right after .rodata, and I would it doesn't
get folded into there because - for some reason - .rodata is
actually marked writable.

Jan



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel