On 2021/8/12 上午1:56, Richard Henderson wrote:
> On 8/11/21 4:57 AM, LIU Zhiwei wrote:
>> I  still don't know why the value written sign-extended.  If that's 
>> the the rule of final specification, I will try to obey it although 
>> our Linux will not depend on the high part.
>
> The text that I'm looking at is
>
> https://github.com/riscv/riscv-isa-manual/releases/download/Ratified-IMFDQC-and-Priv-v1.11/riscv-privileged-20190608.pdf 
>
>
> 3.1.6.2 Base ISA Control in mstatus Register
>
> In the fifth paragraph, the requirement for sign-extension is detailed.
>
Thanks. I have already seen this rule.

/"Whenever XLEN in any mode is set to a value less than the widest 
supported XLEN, all operations////must ignore source operand register bits above the configured XLEN, and 
must sign-extend results////to fill the entire widest supported XLEN in the destination register.//" /

I still don't know why the specification has this constraint. It just 
requires that fill hardware registers with defined sign-extension value.
But it doesn't give the real benefit of this constraint.

If the software doesn't use the high part, who cares the really value in 
high part? Do you know the benefit?  Thanks again.

Best Regards,
Zhiwei//

>
> r~