Use memcpy to avoid putting one byte at a time when copying blocks

`memcpy` is likely to be faster than the naive method due to the possibility of use of vector instructions and copying more than a byte at a time. Signed-off-by: Damian Jarek <[email protected]>
RockinRoel · Oct 7, 2019 · 962f01e · 962f01e
1 parent 85ab7a3
commit 962f01e
Showing 1 changed file with 2 additions and 3 deletions.
diff --git a/include/boost/beast/zlib/detail/deflate_stream.ipp b/include/boost/beast/zlib/detail/deflate_stream.ipp
@@ -1298,9 +1298,8 @@ copy_block(
  put_short((std::uint16_t)len);
  put_short((std::uint16_t)~len);
  }
- // VFALCO Use memcpy?
- while (len--)
- put_byte(*buf++);
+ std::memcpy(&pending_buf_[pending_], buf, len);
+ pending_ += len;
 }
 
 //------------------------------------------------------------------------------