Re: [PATCH v2] RAS/AMD/ATL: Remove bitwise_xor_bits

[Borislav Petkov <bp@alien8.de>]

On Mon, Nov 24, 2025 at 08:03:02PM +0800, Kuan-Wei Chiu wrote:
> On Mon, Nov 24, 2025 at 12:05:26PM +0100, Borislav Petkov wrote:
> > On Mon, Nov 24, 2025 at 04:57:51PM +0800, Kuan-Wei Chiu wrote:
> > > > Both LLVM/GCC support a __builtin_parity function which is functionally
> > > > equivalent to the custom bitwise_xor_bits() one. Let's simplify the code by
> > > > relying on the built-in. No functional changes.
> > > 
> > > IIRC in some cases,
> > 
> > Which are those cases?
> > 
> > Do you have a trigger scenario?
> > 
> I did a quick search, and I believe it was this kernel test robot
> report [1] that reminded me of this compiler behavior.
> 
> [1]: https://lore.kernel.org/oe-kbuild-all/202501312159.l6jNRaYy-lkp@intel.com/

Interesting, thanks for the pointer.

@Nik, just use hweight16() but pls do check what asm the compiler generates
before and after so that at least there's some palpable improvement or gcc is
smart enough to replace the unrolled XORing with something slick.

Also put in the commit message why we're not using the builtin parity thing
and quote the link above.

Thx.

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette