On Sun, Dec 2, 2012 at 6:46 PM, Peter Gavin <[email protected]> wrote:
> On Sun, Dec 2, 2012 at 1:11 PM, Julius Baxter <[email protected]>
> wrote:
>>
>> 1. Accessing SPRs with insufficient privileges
>>
>> I reckon accessing SPRs which are only accessible in supervisor mode,
>> while in user mode, should basically do nothing - writing should
>> behave like a l.nop and reading should return zero, as if the SPR is
>> unimplemented.
>>
>>
>> http://opencores.org/or1k/Architecture_Specification#Accessing_SPRs_with_insufficient_privileges
>
>
> Yes, this makes sense.
>
> On a related note, sometimes I wonder if the CY and OV bits should be
> directly readable from userspace, so that user code can use hardware
> overflow detection without relying on operating system support.  The CY flag
> is indirectly readable with addc, but the OV flag is not.  They could be
> duplicated into a read-only SPR that's accessible from user mode.

Stefan has also mentioned previously that he'd like an easy way of
transferring the SR[FLAG] bit to a GPR more easily. Perhaps we can
come up with something to accommodate all of these flags.

>
>>
>> 2. Optional overflow/carry detection and (range) exception.
>
>
> Good idea.
>
>>
>> 3. Behaviour when AECR/AESR not present
>
>
> This makes sense.
>
>>
>> 4. "Signed" in name of l.add[ic] and l.sub
>
>
> Looks good.
>
> -Pete

Thanks,

Julius
_______________________________________________
OpenRISC mailing list
[email protected]
http://lists.openrisc.net/listinfo/openrisc

Reply via email to