On 03/20/2012 10:10 PM, Matthew Hicks wrote:
On Tue, Mar 20, 2012 at 3:04 PM, Stefan Kristiansson
<[email protected] <mailto:[email protected]>>
wrote:
On 03/20/2012 09:53 PM, Matthew Hicks wrote:
On Tue, Mar 20, 2012 at 2:48 PM, Stefan Kristiansson
<[email protected]
<mailto:[email protected]>> wrote:
That's a truth with modification, isn't it?
You proposed changing an implementation (or1200) not the ISA,
and most felt it wasn't worth it since it could as well (and should)
be handled in software.
I don't think there was any discussion about unacceptable ABI breaks
and changing ISAs.
Stefan
http://lists.openrisc.net/pipermail/openrisc/2011-December/000533.html where
I first propose changing the ISA as well as the implementation.
Ah, yes you did, I stand corrected :)
But still, changing the specification to define r0 to be hardwired to 0
or that it _has_ to be regarded as a nop is probably a too big of a change,
clarifying that writing to r0 should have no side effects
(except possibly writing a nonzero value to it) is another story.
Stefan
I am confused. So are you saying that the ISA say that only writing a
non-zero value to r0 be undefined?
No, I am saying that at the moment there is no definition of what happens if you
indeed write to r0,
only that you shouldn't and it isn't necessarily hardwired to zero.
For all we know it could self destruct and burst into flames.
With that kind of definition it's kind of hard to defend that software _should_
write 0 to r0 on start-up.
Stefan
_______________________________________________
OpenRISC mailing list
[email protected]
http://lists.openrisc.net/listinfo/openrisc
