riscv: Implement Zbb based strlen and prefer it over the RVV based strlen implementation when Zbb is available

[Jeffrey Law]

On 6/11/2026 11:41 AM, Peter Bergner wrote:
> On 6/10/26 11:09 PM, Jeffrey Law wrote:
>> We can see the Zbb is just better all around.  There wasn't a single
>> case where RVV won.  It's pretty obvious that the vector version has a
>> higher fixed overhead, but I really expected vector to overcome that
>> overhead as the strings got longer.  As it stands the data says quite
>> clearly that we should be using Zbb on the K3 design and likely the K1
>> design (currently being tested).
>>
>> Given the K1/K3 designs are what folks can get their hands on, I'd
>> recommend we make Zbb preferred over RVV.  We'll likely have to adjust
>> that as newer designs come into the market, but the decision should be
>> data driven.  I'm going to run this on our Veyron V2 design and Peter is
>> going to run on the Ascalon design, but neither of those are generally
>> available and probably shouldn't drive decisions, those are mostly for
>> informational purposes and to give a sense of whether or not higher
>> targeted designs are likely to benefit from the RVV variant when those
>> higher performance designs hit the market.
> Yes, I'll have someone run the tests on our Ascalon design and report
> back.
Thanks.

>
> That said, this is looking good for a normal multiarch build, but for
> a --disable-multiarch build, we'll need some sysdeps/riscv/preconfigure*
> and Implies file changes, specifying the subdirectory search priority.
> I believe "as is" with this patch, if we build with --disable-multiarch,
> we'll never look in the zbb subdir, meaning zbb isn't in the subdirectory
> search list at all.
I'm not at all familiar with the Implies stuff.  I'll try some things 
with --disable-multiarch to see what I get.

>
> With your data, it would seem we'd want to search the new zbb subdir
> before the rvv subdir...at least for strlen, but I doubt we'd want
> that for all other functions, right?  So do we go with, search rvv
> first, followed by zbb and just take the perf hit for non-multiarch
> builds?  Your thoughts?
I think we'd want to evaluate the other functions.  str[n]cmp for 
example before we make this decision.

jeff