On Tue, 12 Nov 2013, Måns Rullgård wrote:
> Nicolas Pitre <nicolas.pi...@linaro.org> writes:
>
> > What about this patch which I think is currently your best option. Note
> > it would need to use the facilities from asm/opcodes.h to make it endian
> > agnostic.
> >
> > diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c
> > index 6a1b8a81b1..379cffe4ab 100644
> > --- a/arch/arm/kernel/setup.c
> > +++ b/arch/arm/kernel/setup.c
> > @@ -383,6 +383,34 @@ static void __init cpuid_init_hwcaps(void)
> > elf_hwcap |= HWCAP_IDIVT;
> > }
> >
> > + /*
> > + * Patch our division routines with the corresponding opcode
> > + * if the hardware supports it.
> > + */
> > + if (IS_ENABLED(CONFIG_THUMB2_KERNEL) && (elf_hwcap & HWCAP_IDIVT)) {
> > + extern char __aeabi_uidiv, __aeabi_idiv;
>
> It would be safer to declare these as arrays of unspecified size.
> Otherwise the compiler might do evil things with what to it looks like
> out of bounds indexing.
Right.
>
> There should also be some cache maintenance after this patching, or is
> that already happening for some other reason?
This is so early during boot that the MMU isn't even fully initialized
yet. The cache will be flushed.
Nicolas
