-----Original Message----- From: linaro-toolchain-bounces@lists.linaro.org [mailto:linaro- toolchain-bounces@lists.linaro.org] On Behalf Of Mans Rullgard Sent: 20 March 2013 14:17 To: Renato Golin Cc: Kristof Beyls; Linaro Toolchain; Tim Northover Subject: Re: LLVM ARM NEON VMUL.f32
On 19 March 2013 21:56, Renato Golin renato.golin@linaro.org wrote:
Hi folks,
I found an issue while fixing a test using the wrong VMUL.f32, and
I'd like
to know what should be our choice on this topic that is slightly controversial.
Basically, LLVM chooses to lower single-precision FMUL to NEON's
VMUL.f32
instead of VFP's version because, on some cores (A8, A5 and Apple's
Swift),
the VFP variant is really slow.
This is all cool and dandy, but NEON is not IEEE 754 compliant, so
the
result is slightly different. So slightly that only one test, that
was
really pushing the boundaries (ie. going below FLT_MIN) did catch it.
There are two ways we can go here:
- Strict IEEE compatibility and *only* lower NEON's VMUL if unsafe-
math is
on. This will make generic single-prec. code slower but you can
always turn
unsafe-math on if you want more speed.
- Continue using NEON for f32 by default and put a note somewhere
that
people should turn this option (FeatureNEONForFP) off on A5/A8 if
they
*really* care about maximum IEEE compliance.
Apple already said that for Darwin, 2 is still the option of choice.
Do we
agree and ignore this issue? Or for GNU/EABI we want strict
conformance by
default?
GCC uses fmuls...
The NEON vmul.f32 takes two possibly unexpected shortcuts: it flushes denormals to zero, and it ignores the selected rounding mode. Both of these can result in incorrect operation of code assuming standard behaviour.
This was the reason GCC disabled vectorization for a lot of fp operations for neon when in strict IEEE754 conformance mode for the ARM port which is the default. And I suspect you want LLVM to as well if it already doesn't :)
http://gcc.gnu.org/PR43703 is the bug report for more - if sourceware is back up and services running.
regards Ramana
-- IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.