[PATCH] math: Add optimization barrier to ensure a1 + u.d is not reused [BZ #30664]

[Dave Anglin]

On 2025-02-26 4:37 p.m., Joseph Myers wrote:
> On Wed, 26 Feb 2025, Adhemerval Zanella Netto wrote:
> 
>> Last time I checked a integer based implementation (musl), it was faster
>> than glibc one even on recent hardware.  But I don't think this is pressing
>> issue since now most hardware does have fma instructions.
> 
> We have an implementation based on soft-fp (in sysdeps/ieee754/soft-fp/).  
> That does require a suitable sfp-machine.h, including the parts 
> integrating with hardware exceptions and rounding modes in the case of 
> architectures supporting those.

The is a hppa implementation of sfp-machine.h in libgcc.  I'll test.

-- 
John David Anglin  jda.parisc@gmail.com