[Bug target/57989] New: gcc for ARM defines __ARM_FEATURE_SIMD32, but does provide SIMD32 (ARMv6) intrinsics

[Bug target/57989] New: gcc for ARM defines __ARM_FEATURE_SIMD32, but does provide SIMD32 (ARMv6) intrinsics

[maratek at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57989

            Bug ID: 57989
           Summary: gcc for ARM defines __ARM_FEATURE_SIMD32, but does
                    provide SIMD32 (ARMv6) intrinsics
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: maratek at gmail dot com

Created attachment 30556
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=30556&action=edit
gcc -version

ARM C language extensions specification
(http://infocenter.arm.com/help/topic/com.arm.doc.ihi0053a/IHI0053A_acle.pdf)
describes the __ARM_FEATURE_SIMD32 macro as follows: "__ARM_FEATURE_SIMD32 is
defined to 1 if the 32-bit SIMD instructions are supported and the intrinsics
defined in 9.5 are available. This also implies support for the GE global flags
which indicate byte-by-byte comparison results."

gcc 4.8 does define __ARM_FEATURE_SIMD32 when the target supports SIMD32
instructions (i.e. targeting ARMv6 or higher), but the corresponding intrinsics
are not available in gcc. As per above specification, the intrinsics should be
defined in file arm_acle.h, but gcc seems to lack this header.

[Bug target/57989] gcc for ARM defines __ARM_FEATURE_SIMD32, but does provide SIMD32 (ARMv6) intrinsics

[rearnsha at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57989

Richard Earnshaw <rearnsha at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2013-07-30
     Ever confirmed|0                           |1

--- Comment #1 from Richard Earnshaw <rearnsha at gcc dot gnu.org> ---
Implementing ACLE is work-in-progress.  It's unfortunate that this define was
added before the intrinsics were.