mirror of
https://github.com/zebrajr/pytorch.git
synced 2025-12-07 12:21:27 +01:00
0dfdc2cbdb
Summary: If we use clang with sse4 support, we will have the function redefinition error between [1] and [2]. This patch try to add some checkings to fix this problem. I just turn on USE_NATIVE_ARCH with clang, then I hit the redefinition error. [1] caffe2/operators/quantized/int8_simd.h [2] third_party/gemmlowp/gemmlowp/fixedpoint/fixedpoint_sse.h Pull Request resolved: https://github.com/pytorch/pytorch/pull/13859 Differential Revision: D13095694 Pulled By: ezyang fbshipit-source-id: c65166e4d5a04bb54e2b82c52740af00116ccb0d
24 lines
722 B
C
24 lines
722 B
C
#pragma once
|
|
|
|
// We want to allow 128-bit wide SIMD if either NEON is available (as
|
|
// detected by GEMMLOWP_NEON), or whether SSE4.2 and Clang is
|
|
// available (in which case we will use the neon_sse.h library to
|
|
// share source between the two implementations). We use SSE4.2 to
|
|
// ensure we can use the full neon2sse library, and we use Clang as
|
|
// GCC has issues correctly compiling some parts of the neon2sse
|
|
// library.
|
|
|
|
// Otherwise, the INT8_NEON_SIMD variable will be undefined.
|
|
|
|
#include "gemmlowp/fixedpoint/fixedpoint.h"
|
|
#include "gemmlowp/public/gemmlowp.h"
|
|
|
|
#ifdef GEMMLOWP_NEON
|
|
#define INT8_NEON_SIMD
|
|
#endif
|
|
|
|
#if defined(__SSE4_2__) && defined(__clang__)
|
|
#include "NEON_2_SSE.h"
|
|
#define INT8_NEON_SIMD
|
|
#endif
|