qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [RFC v1 3/5] hw/riscv: Extend the kernel loading suppor


From: Alistair Francis
Subject: Re: [Qemu-devel] [RFC v1 3/5] hw/riscv: Extend the kernel loading support
Date: Wed, 19 Jun 2019 15:06:40 -0700

On Wed, Jun 19, 2019 at 2:01 PM Alistair Francis <address@hidden> wrote:
>
> On Wed, Jun 19, 2019 at 8:16 AM Bin Meng <address@hidden> wrote:
> >
> > On Wed, Jun 19, 2019 at 8:56 AM Alistair Francis
> > <address@hidden> wrote:
> > >
> > > Extend the RISC-V kernel loader to support uImage and Image files.
> > > A Linux kernel can now be booted with:
> > >
> > >     qemu-system-riscv64 -machine virt -bios fw_jump.elf -kernel Image
> > >
> > > Signed-off-by: Alistair Francis <address@hidden>
> > > ---
> > >  hw/riscv/boot.c | 19 ++++++++++++++-----
> > >  1 file changed, 14 insertions(+), 5 deletions(-)
> > >
> > > diff --git a/hw/riscv/boot.c b/hw/riscv/boot.c
> > > index 392ca0cb2e..7f68035a3f 100644
> > > --- a/hw/riscv/boot.c
> > > +++ b/hw/riscv/boot.c
> > > @@ -67,13 +67,22 @@ target_ulong riscv_load_kernel(MachineState *machine,
> > >      uint64_t kernel_entry, kernel_high;
> > >
> > >      if (load_elf(kernel_filename, NULL, kernel_translate, machine,
> > > -                 &kernel_entry, NULL, &kernel_high,
> > > -                 0, EM_RISCV, 1, 0) < 0) {
> > > -        error_report("could not load kernel '%s'", kernel_filename);
> > > -        exit(1);
> > > +                 &kernel_entry, NULL, &kernel_high, 0, EM_RISCV, 1, 0) > 
> > > 0) {
> > > +        return kernel_entry;
> > > +    }
> > > +
> > > +    if (load_uimage_as(kernel_filename, &kernel_entry, NULL, NULL,
> > > +                       kernel_translate, machine, NULL) > 0) {
> >
> > We should not set the 'kernel_translate' here for uImage.
> >
> > In fact, the whole kernel_translate() is not necessary.
>
> I have removed the kernel_translate() function. I tested loading
> uImage files though and they don't seem to work as the image is loaded
> at the wrong address.
>
> I have removed uImage loading support from this series. We can look at
> it in the future if we decide we want it.

Ah, my mistake, it looks like my uImage wasn't built correctly. I'll
keep this in.

Alistair

>
> Alistair
>
> >
> > > +        return kernel_entry;
> > > +    }
> > > +
> >
> > Regards,
> > Bin



reply via email to

[Prev in Thread] Current Thread [Next in Thread]