OSLib bug fixes (various)

[Daniel Ellis]

On Mon 15 Jan, Tony van der Hoff wrote:
> On Sun, 14 Jan 2001, at 18:43:29, Timothy Baldwin
> <csyteb at comp.leeds.ac.uk> wrote on the subject "OSLib bug fixes
> (various)":
> 
> >On Fri, 12 Jan 2001, Daniel Ellis wrote:
> >> On Fri 12 Jan, Tony van der Hoff wrote:
> >
> >> > >+ os_int_on() and os_int_off() added as these can be called from user
> >> > >   and 32 bit modes.
> >> > Really? With valid APCS - this was the reason they weren't included in
> >> > the first place.
> >
> >In USR26 and USR32 modes fhe I flag can't be written to, there it's state
> >won't be restore on function exit, also APCS32 does not restore the flags so
> >the I flag will remain as set by os_int_on() or os_int_off().
> >
> If I understand you correctly, you're saying that in 26-bit privileged
> modes, these are effectively no-ops, but they work in all other modes.
> Fair enough, that can be documented. 
> 
> I certainly have no objection to enabling these modes, if that's what
> people want...

If they call OS_IntOn, and OS_IntOff as they suggestedly do, then it won't
matter whether your in 26-bit mode or not (assuming the whole OS is running
in the same mode).  Their whole purpose is so that interrupts can be turned
on and off from user mode surely.  If you're in SVC mode you can switch
interrupts with a single instruction (or couple of instructions) - calling a
SWI would be somewhat heavyweight!

-- 
Daniel Ellis, Senior Software Engineer,
Pace Micro Technology plc,
645 Newmarket Road,                               tel: +44 (0) 1223 518619
Cambridge, CB5 8PB                                WWW: http://www.pace.co.uk/