On Wed, 2012-07-11 at 12:22 -0400, Steven Rostedt wrote:
> On Tue, 2012-07-03 at 17:29 +0900, Masami Hiramatsu wrote:
>
>
> >
> > > + /* Restore flags */
> > > + pushq EFLAGS(%rsp)
> > > + popfq
> > > +
> > > + MCOUNT_RESTORE_FRAME
> >
> > Here, if MCOUNT_RESTORE_FRAME has skip too, I think you don't
> > need to restore flags before restoring other registers, like
> > below;
> >
> > MCOUNT_RESTORE_FRAME 8
> > popfq
> >
> > And also, this will prevent to modify flags before return by
> > addq in MCOUNT_RESTORE_FRAME.
>
> Ah, because the addq will modify flags :-/
>
> Grumble, I guess I should implement this, even though it will make it a
> little more complex. I thought it was better to restore flags
> explicitly, but that's not the case.
>
I know why I did this. Do you want kprobes to be able to modify flags?
If so, then I need to add, before the restore:
movq EFLAGS(%rsp), %rax
movq %rax, SS(%rsp)
-- Steve
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
