Commit Graph

2485 Commits

Author SHA1 Message Date
Even Rouault 8fa405ee15 IDWT 5x3: fix bug in AVX2 implementation (#953, #957) 2017-06-30 00:03:05 +02:00
Even Rouault 6239ed7be4 INSTALL.md: add section discussing how to enable CPU specific optimizations 2017-06-26 13:13:26 +02:00
Even Rouault 533fa2fdee Merge pull request #957 from rouault/idwt_53_improvements
IDWT 5x3 single-pass lifting and SSE2/AVX2 implementation
2017-06-26 12:45:34 +02:00
Even Rouault 6026786069 Style fix 2017-06-21 13:20:35 +02:00
Even Rouault 93aca84731 Fix mingw related warnings 2017-06-21 12:54:40 +02:00
Even Rouault cdd3e83bae Fix clang warning about extraneous parentheses 2017-06-21 12:49:01 +02:00
Even Rouault 4fe7620d4a .travis.yml: add a configuration to test compilation of AVX2 (but disable tests since Travis doesn't have AVX2 compatible machines) 2017-06-21 12:41:56 +02:00
Even Rouault fd0dc535ad IDWT 5x3: generalize SSE2 version for AVX2
Thanks to our macros that abstract SSE use, the functions can use
AVX2 when available (at compile time)

This brings an extra 23% speed improvement on bench_dwt in 64bit builds
with AVX2 compared to SSE2.
2017-06-21 12:12:58 +02:00
Even Rouault f6e3475cc9 dwt.c: small cleanup 2017-06-21 01:07:56 +02:00
Even Rouault f06cfadef8 Enable __SSE__ / __SSE2__ with Visual Studio 2017-06-20 18:24:21 +02:00
Even Rouault fa55b52d19 Improve performance of inverse DWT 5x3 (#953)
* Use single-pass lifting inverse wavelet transform.
* For vertical pass, use SSE2 when available so as to process 8 columns
  in parallel. This is the most beneficial improvement, since the
  vertical pass involves a lot of cache trashing.

With the bench_dwt utility with default arguments (16383x16383 image),
time goes from 4.064 s to 1.212 s.
2017-06-20 18:01:34 +02:00
Even Rouault 919ed5f8b8 Add bench_dwt program (compiled only if BUILD_BENCH_DWT=ON) 2017-06-20 17:56:19 +02:00
Even Rouault 5c56933daf Merge pull request #955 from rouault/remove_opj_nosanitize
Remove OPJ_NOSANITIZE in opj_bio_read() and opj_bio_write() (#761)
2017-06-18 00:49:20 +02:00
Even Rouault 8df2521a60 Remove OPJ_NOSANITIZE in opj_bio_read() and opj_bio_write() (#761)
Commit 29313eb5 introduced those flags to avoid issues with
-fsanitize=unsigned-integer-overflow
However it is better just to rewrite the loop to avoid such condition
to occur.
2017-06-17 19:15:00 +02:00
Even Rouault 32b20b93e0 Fix astyle issue 2017-06-17 16:37:56 +02:00
Even Rouault 5f596cb283 Fix warning about unused arguments 2017-06-17 14:10:15 +02:00
Even Rouault cc07aec6c7 Fix warnings with recent GCC versions 2017-06-17 14:09:31 +02:00
Jeroen Ooms a0839cca24 only build both static and dynamic on non-windows 2017-06-16 13:58:25 +02:00
Jeroen Ooms 1329b3240a build both shared and static library 2017-06-16 13:27:19 +02:00
Antonin Descampe 36dd87cea8 Merge pull request #928 from RussellMcOrmond/master
Quiet mode for opj_decompress via -quiet long parameter.
2017-06-14 17:23:06 +02:00
Even Rouault 9cbc9903c3 Merge branch 't1_flag_optimizations' 2017-06-13 12:09:52 +02:00
Even Rouault 2609fb8077 Packet header writing: set empty packet header bit to 0 when appropriate (small optimization) 2017-06-12 18:38:11 +02:00
Even Rouault 73d1510d47 Encoder: fix packet writing of empty sub-bands (#891, #892)
There are situations where, given a tile size, at a resolution level,
there are sub-bands with x0==x1 or y0==y1, that consequently don't have any
valid codeblocks, but the other sub-bands may be non-empty.
Given that we recycle the memory from one tile to another one, those
ghost codeblocks might be non-0 and thus candidate for packet inclusion.
2017-06-12 18:37:50 +02:00
Even Rouault 81c5311758 T1: fix BYPASS/LAZY, TERMALL/RESTART and PTERM/ERTERM encoding modes. (#674)
There were a number of defects regarding when and how the termination of
passes had to done and the computation of their rate.
2017-06-09 10:49:03 +02:00
Even Rouault 9a9b06911e opj_t1_dec_sigpass_raw/opj_t1_dec_refpass_raw: harmonize style with mqc methods 2017-06-02 19:22:15 +02:00
Even Rouault 532243f1fd MQC/RAW decoder: use an artificial 0xFF 0xFF terminating marker.
This saves comparing the current pointer with the end of buffer pointer.
This results at least in tiny speed improvement for raw decoding, and
smaller code size for MQC as well.

This kills the remains of the raw.h/.c files that were only used for
decoding. Encoding using the mqc structure already.
2017-06-02 18:24:07 +02:00
Even Rouault 9b39fc4bcc Fix documentation of opj_t1_decode_cblks() 2017-06-02 18:23:49 +02:00
Even Rouault dde6cbabc0 Simplify VSC handling: instead of masking out bits when reading the 4th row.
Do not set them when updating flags of the 1st row
2017-06-02 18:23:38 +02:00
Even Rouault 3d9940a35b Force inlining of mqc decoding and pass steps through heavy use of macros, so as to get better register allocation 2017-06-02 18:23:20 +02:00
Even Rouault 7e8b502842 t1_generate_luts.c: fix compiler warnings 2017-06-02 18:22:59 +02:00
Even Rouault 2ba861c37c Optimize opj_t1_update_flags() 2017-06-02 18:22:42 +02:00
Even Rouault a0861855c1 T1: remove use of neghalf variable. It is useless since bpno is always > 0 2017-06-02 18:22:21 +02:00
Even Rouault 10410fe72e T1: avoid pointer indirection for mqc and raw members of opj_t1_t 2017-06-02 18:21:54 +02:00
Even Rouault a5003787ff T1: remove flags_stride variable from opj_t1_t 2017-06-02 18:21:39 +02:00
Even Rouault 0ec842e1f1 Inline opj_raw_decode() 2017-06-02 18:21:21 +02:00
Even Rouault aa7a8a4398 T1: loop unrolling in dec_sigpass_raw and dec_refpass_raw 2017-06-02 18:20:58 +02:00
Even Rouault 68557ff503 T1: Transpose coder optimizations to decoder, and cleanup code 2017-06-02 18:20:35 +02:00
Even Rouault 1957a498b6 Fix compiler warnings 2017-05-23 17:06:46 +02:00
Even Rouault 40c0f42def Factor index computation for lut_enc_ctxno_sc and lut_enc_spb 2017-05-23 17:06:46 +02:00
Even Rouault d6907b9304 Optimize a bit opj_t1_enc_clnpass() 2017-05-23 17:06:46 +02:00
Even Rouault c76a592131 T1: remove unused code in decoder 2017-05-23 17:06:46 +02:00
Even Rouault 4068363ff5 T1: fix VSC mode in encoder 2017-05-23 16:16:32 +02:00
Even Rouault cd12414c6b T1: use more compact flags to optimize cache usage in encoder passes. (#172)
Ported from Carl Hetherington work (actually through Matthieu Darbois's port
on top of OpenJPEG 2.1.0)

Can reduce total encoding time by 10-15%

WARNING: VSC mode is not implemented, and so is a temporary regression
that must be fixed.
2017-05-23 16:16:32 +02:00
Even Rouault 53d46fc733 Merge pull request #936 from rouault/master_warnings
CMake: add stronger warnings for openjp2 lib/bin by default, and error out on declaration-after-statement
2017-05-23 16:15:55 +02:00
Even Rouault a8ca7c51f3 CMake: add stronger warnings for openjp2 lib/bin by default, and error out on declaration-after-statement
And remove occurences of unused arguments in src/lib/openjp2
2017-05-23 15:47:57 +02:00
Even Rouault 6e97d877b1 Merge pull request #935 from rouault/add_compress_vsc_test
Tests: test opj_compress in VSC mode (related to #172)
2017-05-23 14:49:38 +02:00
Even Rouault 2d2c368b19 Tests: test opj_compress in VSC mode (related to #172) 2017-05-23 14:31:39 +02:00
Even Rouault 8728cfbc79 t1.c: fix compiler warnings 2017-05-23 13:54:28 +02:00
Russell McOrmond 5f1e380b51 Fixed formatting issues ASTYLE 2017-05-16 09:31:07 -04:00
Russell McOrmond 9d8e1ecdeb Quiet mode for opj_decompress via -quiet long parameter. 2017-05-15 14:39:54 -04:00