[Libva] [PATCH libva 2/2] trace: Cleanup some HEVC prints

Some fields in HEVC buffers had extraneous timestamps interspersed with the traced data. Signed-off-by: Scott D Phillips --- va/va_trace.c | 108 +- 1 file changed, 62 insertions(+), 46 deletions(-) diff --git a/va/va_trace.c b/va

[Libva] [PATCH libva 1/2] trace: Add `va_TracePrint`, `va_TraceVPrint` static functions

These functions are similar to va_TraceMsg() but without emitting a timestamp. Signed-off-by: Scott D Phillips --- va/va_trace.c | 50 ++ 1 file changed, 34 insertions(+), 16 deletions(-) diff --git a/va/va_trace.c b/va/va_trace.c index f632176

[Libva] [PATCH intel-driver] Remove extraneous OUT_BATCH in gen8_gpe_state_base_address

Line accidentally left behind in: a82f0be Fix the 48-bit address issue for gpe_util functions on Signed-off-by: Scott D Phillips --- src/i965_gpe_utils.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/src/i965_gpe_utils.c b/src/i965_gpe_utils.c index 9f1bf92..91d1192 100644 --- a/src

[Libva] [PATCH intel-driver] dri: return error for unimplemented subsampling types

Previously YUV422 surfaces were allowed to be rendered but got rendered improperly. Signed-off-by: Scott D Phillips --- src/i965_output_dri.c | 4 1 file changed, 4 insertions(+) diff --git a/src/i965_output_dri.c b/src/i965_output_dri.c index d36fec5..f141a67 100644 --- a/src

[Libva] [PATCH intel-driver] i965_drv: fix cb_cr_height for YUV422 formats

YUV422 has full vertical chroma resolution, not half. Signed-off-by: Scott D Phillips --- src/i965_drv_video.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/i965_drv_video.c b/src/i965_drv_video.c index ad48f22..ec67848 100644 --- a/src/i965_drv_video.c +++ b/src

[Libva] [PATCH intel-driver v2] dri: return error for unimplemented surface formats

Previously packed YUV422 surface were allowed to be renderd but got rendered improperly. Signed-off-by: Scott D Phillips --- Changes since v1: - blacklist certain fourcc's instead of all 4:2:2 surfaces src/i965_output_dri.c | 3 +++ 1 file changed, 3 insertions(+) diff --git

Re: [Libva] [PATCH intel-driver] dri: return error for unimplemented subsampling types

On Thu, Jul 21, 2016 at 10:46:06PM -0700, Xiang, Haihao wrote: > > > Previously YUV422 surfaces were allowed to be rendered but got > > rendered improperly. > > YUV422 (planar) surface works for me, but packed surface is not > supported. Ah, you're right. I wasn't able to get it work initially w

[Libva] [PATCH intel-driver 2/2] test: read jpeg test data from /dev/urandom

Reading from urandom seems to be faster than using the standard PRNG. Signed-off-by: Scott D Phillips --- test/i965_jpeg_test_data.cpp | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/test/i965_jpeg_test_data.cpp b/test/i965_jpeg_test_data.cpp index 956f7cf..082d43f

[Libva] [PATCH intel-driver 1/2] test: use valarray for raw image comparison

std::valarray can fuse elementwise operations across arrays for more efficient comparison. Signed-off-by: Scott D Phillips --- test/i965_jpeg_encode_test.cpp | 29 + 1 file changed, 13 insertions(+), 16 deletions(-) diff --git a/test/i965_jpeg_encode_test.cpp b/test

[Libva] [PATCH intel-driver 0/2] Speed up jpeg encode tests

For me these two patches take the Big encode test from 44sec to ~7sec. Scott D Phillips (2): test: use valarray for raw image comparison test: read jpeg test data from /dev/urandom test/i965_jpeg_encode_test.cpp | 29 + test/i965_jpeg_test_data.cpp | 6

Re: [Libva] [PATCH intel-driver 0/2] Speed up jpeg encode tests

> > > For me these two patches take the Big encode test from 44sec to > > ~7sec. > > > > Scott D Phillips (2): > >   test: use valarray for raw image comparison > >   test: read jpeg test data from /dev/urandom > > > >  test/i965_jpeg_encode_test.cpp

[Libva] [PATCH v2] test: use valarray for raw image comparison

std::valarray can fuse elementwise operations across arrays for more efficient comparison. Signed-off-by: Scott D Phillips --- Changes since v1: - s/width * height/sizes/ - Added another array 'signs' which ensures that 0 and 255 never compare as separated by one. Also, the patch