Refactor pair_set_epi16 for speedup

Use _mm_set1_epi32 instead of _mm_set_epi16, less instructions produced
by compiler. This patch also removes the duplicate define of the same
function.

Speed test results:
1. Unittest for each test cases in SSE2/AV1LbdInvTxfm2d shows 60%~80%
speedup (except those case with TX_TYPE include iidentity)
2. A brief speed test shows that with this CL, for speed1 encoder speeds up
~3% and decoder speeds up ~1.8%.
(Baseline is 18976fa5f1ce9777276d969190a1ccb51afbcefe)

Change-Id: I2b0e12973fda05a21d6b6eb0f0efe11df6edfb84

aom_dsp/x86/txfm_common_sse2.h

diff --git a/aom_dsp/x86/txfm_common_sse2.h b/aom_dsp/x86/txfm_common_sse2.h
index cd856ff..f946f13 100644
--- a/aom_dsp/x86/txfm_common_sse2.h
+++ b/aom_dsp/x86/txfm_common_sse2.h
@@ -16,9 +16,8 @@
 #include "aom/aom_integer.h"
 #include "aom_dsp/x86/synonyms.h"
 
-#define pair_set_epi16(a, b)                                            \
-  _mm_set_epi16((int16_t)(b), (int16_t)(a), (int16_t)(b), (int16_t)(a), \
-                (int16_t)(b), (int16_t)(a), (int16_t)(b), (int16_t)(a))
+#define pair_set_epi16(a, b) \
+  _mm_set1_epi32((int32_t)(((uint16_t)(a)) | (((uint32_t)(b)) << 16)))
 
 #define dual_set_epi16(a, b)                                            \
   _mm_set_epi16((int16_t)(b), (int16_t)(b), (int16_t)(b), (int16_t)(b), \

av1/common/x86/av1_txfm_sse2.h

diff --git a/av1/common/x86/av1_txfm_sse2.h b/av1/common/x86/av1_txfm_sse2.h
index ae3578f..1ab3977 100644
--- a/av1/common/x86/av1_txfm_sse2.h
+++ b/av1/common/x86/av1_txfm_sse2.h
@@ -17,16 +17,13 @@
 #include "./av1_rtcd.h"
 #include "aom/aom_integer.h"
 #include "aom_dsp/x86/transpose_sse2.h"
+#include "aom_dsp/x86/txfm_common_sse2.h"
 #include "av1/common/av1_txfm.h"
 
 #ifdef __cplusplus
 extern "C" {
 #endif
 
-#define pair_set_epi16(a, b)                                            \
-  _mm_set_epi16((int16_t)(b), (int16_t)(a), (int16_t)(b), (int16_t)(a), \
-                (int16_t)(b), (int16_t)(a), (int16_t)(b), (int16_t)(a))
-
 #define btf_16_sse2(w0, w1, in0, in1, out0, out1) \
   {                                               \
     __m128i t0 = _mm_unpacklo_epi16(in0, in1);    \