Hi Jiri, On Fri, Mar 07, 2025 at 07:57:48AM +0100, Jiri Slaby wrote: > On 06. 03. 25, 17:25, Kuan-Wei Chiu wrote: > > Several parts of the kernel contain redundant implementations of parity > > calculations for 16/32/64-bit values. Introduces generic > > parity16/32/64() helpers in bitops.h, providing a standardized > > and optimized implementation. > > > > Subsequent patches refactor various kernel components to replace > > open-coded parity calculations with the new helpers, reducing code > > duplication and improving maintainability. > > > > Co-developed-by: Yu-Chun Lin <eleanor...@gmail.com> > > Signed-off-by: Yu-Chun Lin <eleanor...@gmail.com> > > Signed-off-by: Kuan-Wei Chiu <visitor...@gmail.com> > > --- > > In v3, I use parityXX() instead of the parity() macro since the > > parity() macro may generate suboptimal code and requires special hacks > > to make GCC happy. If anyone still prefers a single parity() macro, > > please let me know. > > What is suboptimal and where exactly it matters? Have you actually measured > it? > In the previous thread, David and Yury had different opinions regarding the implementation details of the parity() macro. I am trying to find a solution that satisfies most people while keeping it as simple as possible.
I cannot point to any specific users who are particularly concerned about efficiency, so personally, I am not really concerned about the generated code either. However, I am not a fan of the #if gcc #else approach, and Yury also mentioned that he does not like the >> 16 >> 16 hack. At the same time, David pointed out that GCC might generate double-register math. Given these concerns, I leaned toward reverting to the parityXX() approach. If you still prefer using the parity() macro, we can revisit and discuss its implementation details further. > > Additionally, I changed parityXX() << y users to !!parityXX() << y > > because, unlike C++, C does not guarantee that true casts to int as 1. > > How comes? ANSI C99 exactly states: > === > true > which expands to the integer constant 1, > === > I gave a more detailed response in my reply to Peter. If we can confirm that casting bool to int will only result in 1 or 0, I will remove the !! hack in the next version. Regards, Kuan-Wei
