[PATCH v4 07/12] RISC-V: check operands for Zqinx

[Jan Beulich]

On 27.07.2026 05:15, Nelson Chu wrote:
> On Fri, Jul 24, 2026 at 6:04 PM Jan Beulich <jbeulich@suse.com> wrote:
>> ---
>> Does Zqinx actually make much sense in RV32? Function argument passing,
>> for example, ends up odd: a0..a3 aren't suitable to hold an argument.
>> Hence only a single argument can be passed in registers (a2..a5).
> 
> Jiawei,
> 
> You should know the best for these finx changes, do you think zqinx
> should be allowed under rv32?

May I put the question a little differently: Shouldn't there be at least
a semi-official spec for Zqinx, which would then address the question
above?

Jan