Skip site navigation (1)Skip section navigation (2) 
Date:      Sun, 1 May 2016 23:51:27 +0300
From:      Rozhuk Ivan <rozhuk.im@gmail.com>
To:        <freebsd-hackers@freebsd.org>, Rozhuk Ivan <Rozhuk.IM@gmail.com>
Subject:   some clang AVX instricts looks broken
Message-ID:  <20160501235127.451d8812@rimwks>
next in thread | raw e-mail | index | archive | help 

Hi!

I try port some SSE code to AVX and found that clang instricts _mm256_extract* broken:

_mm256_extract_epi8(__aymm0, 0) - BAD result
_mm256_extract_epi8(__aymm0, 0) & 0xff - OK
_mm_extract_epi8(_mm256_extractf128_si256(__aymm0, 0), 0) - OK

Same time: _mm256_extract_epi8(__aymm0, 0) build with GCC - OK.


CLANG:
static __inline int __attribute__((__always_inline__, __nodebug__))
_mm256_extract_epi8(__m256i __a, int const __imm)
{
  __v32qi __b = (__v32qi)__a;
  return __b[__imm & 31];
}


GCC:
extern __inline int __attribute__((__gnu_inline__, __always_inline__, __artificial__)) 
_mm256_extract_epi8 (__m256i const __X, int const __N) 
{ 
  __m128i __Y = _mm256_extractf128_si256 (__X, __N >> 4); 
  return _mm_extract_epi8 (__Y, __N % 16); 
}

Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20160501235127.451d8812>