On Feb 7, 11:21*pm, j...@cix.compulink.co.uk wrote:
> In article
> <91be674f-2d08-4443-b5a1-fae9e633e...@m40g2000yqh.googlegroups.com>,
>
> Reva...@gmail.com (Revan76) wrote:
> > I wondered why cpuid was serialization instruction like RSM and many
> > other.
>
> > Indeed, I don't understand why it requires previouses instruction to
> > be completed.
>
> Well, if there are other instructions that can change what CPUID
> returns, that would be a good reason. Even if there weren't when it was
> first defined, the possibility of adding them exists.
>
> --
> John Dallman * * * * * * * * * * * * * * * * * * * * *j...@cix.co.uk
> * * * * * "C++ - the FORTRAN of the early 21st century."
Okay thank you very much !