On Fri, Oct 13, 2017 at 02:25:43PM -0700, Anatol Pomozov wrote:
> Hi
>
> On Fri, Oct 13, 2017 at 12:25 PM, Eduardo Habkost <ehabk...@redhat.com> wrote:
> > On Thu, Oct 12, 2017 at 04:54:37PM -0700, Anatol Pomozov wrote:
> >> x86 is not the only architecture supported by multiboot.
> >> For example GRUB supports MIPS architecture as well.
> >>
> >> Signed-off-by: Anatol Pomozov <anatol.pomo...@gmail.com>
> >> ---
> >> hw/i386/multiboot.c | 2 +-
> >> 1 file changed, 1 insertion(+), 1 deletion(-)
> >>
> >> diff --git a/hw/i386/multiboot.c b/hw/i386/multiboot.c
> >> index c9254f313e..7dacd6d827 100644
> >> --- a/hw/i386/multiboot.c
> >> +++ b/hw/i386/multiboot.c
> >> @@ -173,7 +173,7 @@ int load_multiboot(FWCfgState *fw_cfg,
> >> }
> >>
> >> kernel_size = load_elf(kernel_filename, NULL, NULL, &elf_entry,
> >> - &elf_low, &elf_high, 0, I386_ELF_MACHINE,
> >> + &elf_low, &elf_high, 0, EM_NONE,
> >> 0, 0);
> >
> > I assume we still want PC to reject non-x86 ELF files.
>
> Does multiboot spec states this restriction? I've heard that there are
> attempts to implement multiboot at ARM [1] [2]. Also multiboot2 spec
> mentions MIPS as one of the target architectures.
>
> [1] https://github.com/jncronin/rpi-boot/blob/master/MULTIBOOT-ARM
> [2] https://wiki.linaro.org/AndrePrzywara/Multiboot
I don't believe the spec restricts that, but I don't see why it
would be useful to load an ELF file that doesn't match the target
architecture (e.g. loading non-x86 ELF files on a x86 machine
like PC).
>
> > Isn't it
> > better to add a elf_machine argument to load_multiboot() so each
> > load_multiboot() caller can specify what's the expected
> > architecture?
> >
> >
> >> if (kernel_size < 0) {
> >> fprintf(stderr, "Error while loading elf kernel\n");
> >> --
> >> 2.15.0.rc0.271.g36b669edcc-goog
> >>
> >
> > --
> > Eduardo
--
Eduardo
