[RESEND] [PATCH 1/12] PPC64: First in the series of patches implementing POWER8 vector math.

GT tnggil@protonmail.com
Wed Mar 6 22:22:00 GMT 2019


Differences from prior submission of same patch:

1. Removed the powerpc specific bits/math-vector.h file. With that gone I believe the patches can go into glibc master, as opposed to a branch. There is now no API for compilers to possibly produce erroneously vectorized calls to the SIMD math functions. That makes these functions totally inert until GCC comes along and needs to test its new PPC64 autovectorization. Then, a PPC specific bits/math-vector.h will be added back into glibc.

I verified that libmvec does build and pass make check without the powerpc bits/math-vector.h. I also verified that GCC produces the same instructions in an executable, with and without the pragma omp simd directives. That compilation used a new installation of glibc including libmvec.

So, the new glibc including PPC64 libmvec should not break any existing code.

2. Fixed names of constants to include leading double-underscores as required by their namespace.

3. Removed configure check for VSX builtins in compiler. As glibc requires GCC >= 6.2, we are guaranteed the presence of these builtins.

4. Changed patch subject line to now include m/n sequence indicator.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-PPC64-First-in-the-series-of-patches-implementing-PO.patch
Type: text/x-patch
Size: 18218 bytes
Desc: not available
URL: <http://sourceware.org/pipermail/libc-alpha/attachments/20190306/f6a4b137/attachment.bin>


More information about the Libc-alpha mailing list