This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH] PPC64: First in the series of patches implementing POWER8 vector math.
- From: Joseph Myers <joseph at codesourcery dot com>
- To: GT <tnggil at protonmail dot com>
- Cc: "libc-alpha at sourceware dot org" <libc-alpha at sourceware dot org>
- Date: Fri, 22 Feb 2019 01:42:48 +0000
- Subject: Re: [PATCH] PPC64: First in the series of patches implementing POWER8 vector math.
- References: <h_EcM7gwhPlTHJdH4BQs5ZLDeYZK_TJGdZjALt1Rt8SYtUkYN9b2BnuiK1Nio5sl3qwLViw1R6GfXbU8ksDkZZtmGjpprzfY1ONa9cQgey0=@protonmail.com> <alpine.DEB.2.21.1902192243330.5857@digraph.polyomino.org.uk> <eTXHa4W6SVVmEo2Bvpl89isbLgGzgO9IXxXtufHLp8MqbBMSLbmrXBy2sfMjuMMs7zt0kGESrIOqja4TgRgP5VRlRTxISJfDbH0Guo4fE3w=@protonmail.com>
On Fri, 22 Feb 2019, GT wrote:
> And running autoconf at the glibc source-tree root directory should
> regenerate configure scripts in all subdirectories which have
> configure.ac in them, right?
No. It's "autoconf /path/to/configure.ac > /path/to/configure" to
regenerate a subdirectory configure fragment. See the Makefile rules.
> Turns out POWER7 is the earliest POWERn that implements v2.06 of the ISA.
> v2.06 is when type vector double was introduced. So the behavior is
> as expected.
I.e., any processor supporting this functionality has VSX. So do you need
the plain AltiVec versions at all, or just the VSX ones? (At least for
double; maybe if you do float functions later, AltiVec versions will be
relevant there?)
(I'd suggest that, instead of configure tests, it would be best just to
build the functions in question with appropriate -m options, and arrange
for the tests not to run on processors not supporting the required
functionality. If that won't work for some reason, presumably you can
give a clear explanation of that reason for the commit message.)
--
Joseph S. Myers
joseph@codesourcery.com