[incubator-mxnet] branch master updated: add inline for __half2float_warp (#20152)

[zh...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

zhasheng pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git


The following commit(s) were added to refs/heads/master by this push:
     new 957733d  add inline for __half2float_warp (#20152)
957733d is described below

commit 957733d8b6ccdc8de6c54d9ac4e6c6ce9420e77b
Author: khaotik <kh...@users.noreply.github.com>
AuthorDate: Mon Apr 12 04:13:09 2021 +0800

    add inline for __half2float_warp (#20152)
---
 3rdparty/mshadow/mshadow/half.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/3rdparty/mshadow/mshadow/half.h b/3rdparty/mshadow/mshadow/half.h
index f2abab8..a43552d 100644
--- a/3rdparty/mshadow/mshadow/half.h
+++ b/3rdparty/mshadow/mshadow/half.h
@@ -43,7 +43,7 @@
   #include <cuda_fp16.h>
   #if defined(__CUDA_ARCH__)
     /*! \brief __half2float_warp */
-    __host__ __device__ float __half2float_warp(const volatile __half& h) { /* NOLINT(*) */
+    MSHADOW_XINLINE float __half2float_warp(const volatile __half& h) { /* NOLINT(*) */
       __half val;
 #if CUDA_VERSION >= 9000
       val = const_cast<__half&>(h);