Re: [PATCH] RISC-V: Allow quadratic LMUL cost for unknown niter loops

[Bohan Lei]

>  - Experiment with the above condition and check if they work.  If so, let's 

Just to be sure, does this mean, to use the
`LOOP_VINFO_FULLY_WITH_LENGTH_P && ...` condition in place of the
`rvv_lmul_scaling_mode == ...` condition in the patch and see if it works for
the related loops?

Thanks,
Bohan