Re: [PATCH 0/4] KVM: x86 emulator: Move em_grp from switch statement to decode table

[Avi Kivity <avi@redhat.com>]

On 12/07/2011 06:51 PM, Avi Kivity wrote:
> On 12/06/2011 11:04 AM, Takuya Yoshikawa wrote:
> > I removed em_grp1a() and em_grp9() but kept em_grp2() and em_grp45()
> > because they would produce a lot of trivial functions.
>
> We could convert the various emulate_2op_blah() macros to generate
> functions instead of statements, that would greatly reduce the boilerplate.
>
> > Though I think it's almost done about the conversions, I do not have
> > strong opinion about how to treat the remaining instructions: define
> > em_default() and put those into it may be one way.
>
> With em_insn() you have direct visibility as to what a decode table
> entry describes.  That doesn't work with em_default() or em_grpX().
>
>
> > Anyway, now, rather readable than before.
> >
>
> Yes.
>

Forgot to mention, I applied 2-4, thanks.

-- 
error compiling committee.c: too many arguments to function