Re: [E-devel] Neon and 32 bit arm.

2020-11-12 Thread Simon Lees
On 11/13/20 1:13 AM, Carsten Haitzler wrote: > On Thu, 12 Nov 2020 22:50:50 +1030 Simon Lees said: > > armv6 is almost dead except for some very niche cases these days. armv7 - > there > was precisely one soc in the wild without neon (tegra1). so we are talking > very > niche uses which is

Re: [E-devel] Neon and 32 bit arm.

2020-11-12 Thread Ross Vandegrift
On Thu, Nov 12, 2020 at 10:50:50PM +1030, Simon Lees wrote: > armv6 and some armv7 chips don't support neon, our current detection in > meson causes openSUSE's 32bit arm builds to fail. Is this something we > care about enough upstream to make the meson files better or should we > just keep

Re: [E-devel] Neon and 32 bit arm.

2020-11-12 Thread Carsten Haitzler
On Thu, 12 Nov 2020 22:50:50 +1030 Simon Lees said: armv6 is almost dead except for some very niche cases these days. armv7 - there was precisely one soc in the wild without neon (tegra1). so we are talking very niche uses which is why it's not been on any radar. neon arm/intrinsics in evas are