[PATCH RFC] arm64: eBPF JIT compiler

[Z Lim]

On Wed, Jul 2, 2014 at 9:57 PM, Z Lim <zlim.lnx at gmail.com> wrote:
> On Wed, Jul 2, 2014 at 2:28 PM, Alexei Starovoitov <ast at plumgrid.com> wrote:
>> On Tue, Jul 1, 2014 at 10:20 PM, Zi Shen Lim <zlim.lnx at gmail.com> wrote:
>> Do you really need 'jump by register' then? Regular 'bl' would be much faster.
>
> We'll need BLR to cover all cases. BL instruction can only address
> +/-128MB (28-bits).
>
> BTW, what is the range of "imm" for JMP|CALL? I'm guessing since it's
> s32, so +/-512MB?

Oops... I meant: is it +/-2GB?