Re: More on the pentium workaround - the gotchas

Linus Torvalds (
Mon, 17 Nov 1997 09:41:43 -0800 (PST)

On Mon, 17 Nov 1997, Alan Cox wrote:
> > and exceptions to be delayed after movw %*, %%ss or popw %%ss, resulting
> > in more complexity and possibly dangerous situations.]
> write a page of movw to SS and then map it to all of your addres space
> and run it one day..

This shouldn't actually work. As far as I know, moving to SS only delays
interrupts until the next insutruction, _and_ it's not cumulative. The
effect being that the CPU will actually take interrupts and debug
exceptions every second instruction.

That's how I think it went. If somebody wants to test, feel free..