diff options
author | Linfeng Zhang <linfengz@google.com> | 2016-05-17 12:42:55 -0700 |
---|---|---|
committer | Linfeng Zhang <linfengz@google.com> | 2016-05-27 09:51:16 -0700 |
commit | af7fb17c09158e9039bfe6d72c591347b7c0ceb9 (patch) | |
tree | b4a5d179a8a7173ded8982b137a66be7db189819 /test | |
parent | bdfbd6325ef8582beb0cb382f7d0f12cc488ecee (diff) | |
download | libvpx-af7fb17c09158e9039bfe6d72c591347b7c0ceb9.tar libvpx-af7fb17c09158e9039bfe6d72c591347b7c0ceb9.tar.gz libvpx-af7fb17c09158e9039bfe6d72c591347b7c0ceb9.tar.bz2 libvpx-af7fb17c09158e9039bfe6d72c591347b7c0ceb9.zip |
Upgrade fwht4x4_mmx() to fwht4x4_sse2() for vp9 and vp10.
Function level timing test shows about 27% time saving on
a Xeon E5-2680 v2 desktop.
Rename vp9_dct_sse2.c to vp9_dct_intrin_sse2.c for vp9 and
rename dct_sse2.c to dct_intrin_sse2.c for vp10 to avoid
duplicate basenames.
Actually vp9_fwht4x4_mmx/sse2() and vp10_fwht4x4_mmx/sse2()
are identical. TODO: They should be unified later if there is
no intention to keep a duplicate.
Change-Id: I3e537b7bbd9ba417c606cd7c68c4dbbfa583f77d
Diffstat (limited to 'test')
-rw-r--r-- | test/fdct4x4_test.cc | 12 |
1 files changed, 2 insertions, 10 deletions
diff --git a/test/fdct4x4_test.cc b/test/fdct4x4_test.cc index 0c91aee21..5a58830d5 100644 --- a/test/fdct4x4_test.cc +++ b/test/fdct4x4_test.cc @@ -487,19 +487,11 @@ INSTANTIATE_TEST_CASE_P( make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_neon, 3, VPX_BITS_8))); #endif // HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE -#if CONFIG_USE_X86INC && HAVE_MMX && !CONFIG_VP9_HIGHBITDEPTH && \ - !CONFIG_EMULATE_HARDWARE -INSTANTIATE_TEST_CASE_P( - MMX, Trans4x4WHT, - ::testing::Values( - make_tuple(&vp9_fwht4x4_mmx, &vpx_iwht4x4_16_add_c, 0, VPX_BITS_8))); -#endif - -#if CONFIG_USE_X86INC && HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && \ - !CONFIG_EMULATE_HARDWARE +#if CONFIG_USE_X86INC && HAVE_SSE2 && !CONFIG_EMULATE_HARDWARE INSTANTIATE_TEST_CASE_P( SSE2, Trans4x4WHT, ::testing::Values( + make_tuple(&vp9_fwht4x4_sse2, &vpx_iwht4x4_16_add_c, 0, VPX_BITS_8), make_tuple(&vp9_fwht4x4_c, &vpx_iwht4x4_16_add_sse2, 0, VPX_BITS_8))); #endif |