9baccb9860
[ot-tags] Speed up hb_ot_tags_from_complex_language()
...
Part of https://github.com/harfbuzz/harfbuzz/issues/3591
2. All the subtag_matches outside the switch match long strings (>= 6 or so).
As such, check the tag for such length before going into any of them.
benchmark-ot, before:
----------------------------------------------------------------------------------------------
Benchmark Time CPU Iterations
----------------------------------------------------------------------------------------------
BM_hb_ot_tags_from_script_and_language/COMMON zh_CN 172 ns 171 ns 4083155
BM_hb_ot_tags_from_script_and_language/COMMON en_US 120 ns 119 ns 5849947
BM_hb_ot_tags_from_script_and_language/LATIN en_US 113 ns 112 ns 5840326
BM_hb_ot_tags_from_script_and_language/COMMON none 4.66 ns 4.64 ns 151396224
BM_hb_ot_tags_from_script_and_language/LATIN none 4.66 ns 4.64 ns 149019593
After:
----------------------------------------------------------------------------------------------
Benchmark Time CPU Iterations
----------------------------------------------------------------------------------------------
BM_hb_ot_tags_from_script_and_language/COMMON zh_CN 112 ns 112 ns 6357763
BM_hb_ot_tags_from_script_and_language/COMMON en_US 60.5 ns 60.3 ns 11475091
BM_hb_ot_tags_from_script_and_language/LATIN en_US 54.9 ns 54.8 ns 12575690
BM_hb_ot_tags_from_script_and_language/COMMON none 4.61 ns 4.59 ns 152388450
BM_hb_ot_tags_from_script_and_language/LATIN none 4.66 ns 4.64 ns 151497600
2022-05-17 13:34:34 -06:00
26d906b88b
[perf] Add benchmark-ot
2022-05-17 13:12:17 -06:00
629fa8ee87
[perf/benchmark-font] Test Roboto as variable even though it's not
2022-05-16 17:49:36 -06:00
71a0cda869
[perf/benchmark-font] Only certain fonts are variable
...
Don't test every font as variable.
2022-05-16 17:49:36 -06:00
fb413f5202
[subset/cff] Don't use bitfields for hot bools
...
The struct has room because of alignment, and these bools are hot.
2022-05-16 17:38:18 -06:00
a4d98b63ea
[subset/cff1] Collect glyph-to-sid map to avoid an O(n^2) algorithm
...
Saves 13 for largest benchmark:
BM_subset/subset_glyphs/SourceHanSans-Regular_subset.otf/10000 -0.1313 -0.1308 75 65 75 65
BM_subset/subset_codepoints/SourceHanSans-Regular_subset.otf/4096 -0.1009 -0.1004 54 48 54 48
BM_subset/subset_codepoints/SourceHanSans-Regular_subset.otf/10000 -0.1067 -0.1066 70 62 69 62
2022-05-16 17:38:18 -06:00
b87f48e948
[cff1] get_sid() move bounds check into each implementation
2022-05-16 17:38:18 -06:00
e1e359b4da
[cff1] Tighten up range_list_t a bit
2022-05-16 16:36:28 -06:00
3fbac0942d
[cff1] Lazy-load & sort glyph names
...
Improves subset benchmarks by up to 70% for small CFF1 subset of
non-CID fonts!
BM_subset/subset_glyphs/SourceSansPro-Regular.otf/10 -0.7067 -0.7071 1 0 1 0
BM_subset/subset_glyphs/SourceSansPro-Regular.otf/64 -0.4817 -0.4824 1 0 1 0
BM_subset/subset_glyphs/SourceSansPro-Regular.otf/512 -0.1948 -0.1956 2 2 2 2
BM_subset/subset_glyphs/SourceSansPro-Regular.otf/2000 -0.0767 -0.0761 6 6 6 6
2022-05-16 16:36:28 -06:00
b58bfd9818
[font] Minor move of code to silence gcc-12 warning
...
See mailing list discussion.
2022-05-16 11:21:45 -06:00
602e0ca79d
[cff] Minor restructure of struct
...
Surprisingly this shows tiny benchmark improvement consistently.
2022-05-16 10:14:34 -06:00
acdab17ed3
[cff] Cosmetic in parsed_values_t
2022-05-13 14:14:36 -06:00
b46c7faa9c
[cff] Check buf_len, not buf
...
Ouch!
2022-05-13 14:02:54 -06:00
19a8db8545
[subset] fix potential integer overflow in gname_t::cmp.
2022-05-13 13:55:39 -06:00
2d2f66e1a3
[cff-common] In INDEX, return empty bytes if length is zero
...
Before it was possible to return non-null arrayZ.
2022-05-13 13:53:17 -06:00
a2f132f1fc
[cff] Check glyph-name's length, not arrayZ
...
As the latter can be non-null while still zero-length.
2022-05-13 13:49:39 -06:00
dc09053f19
fix build requirements for fedora/centos in buiding document
2022-05-13 13:10:11 -06:00
c657c4e1f8
[meta] fix type traits on gcc 4.9 #3526
...
Signed-off-by: Thomas Devoogdt <thomas.devoogdt@barco.com>
2022-05-13 11:26:12 -06:00
e4e053c8b3
[perf] fix typo in perf Makefile.
2022-05-13 11:25:09 -06:00
e61234c5f7
[vector] Add tests for move constructor/assignment
2022-05-12 13:20:10 -06:00
7fa580bc4f
[map] Fix map copy/move constructors to actually work
...
Ouch!
2022-05-12 13:05:32 -06:00
a09dd87ca3
[set] Fix set copy/move constructors to actually work
...
Ouch!
2022-05-12 12:58:07 -06:00
76fc27713f
[vector] Remove explicit std::move
...
Was confusing compilers. Let them figure it out themselves.
Makes NotoNastaliqu subsetting/1000 benchmark more than twice faster:
Benchmark Time CPU Time Old Time New CPU Old CPU New
------------------------------------------------------------------------------------------------------------------------------------------------------------
BM_subset/subset_glyphs/NotoNastaliqUrdu-Regular.ttf/1000 -0.5064 -0.5065 111 55 110 55
BM_subset/subset_codepoints/NotoNastaliqUrdu-Regular.ttf/1000 -0.5494 -0.5493 132 59 131 59
2022-05-12 12:14:07 -06:00
c81198b5bc
[set] Tweak move operators a bit
...
Should be equivalent.
2022-05-12 12:14:02 -06:00
8dc072d20d
Merge pull request #3579 from harfbuzz/subset-retain-buffer
...
Subset retain buffer
2022-05-11 16:45:40 -06:00
175319cd89
[gsubgpos] Clean up OT::ClassDefFormat2::intersected_class_glyphs 0 case
2022-05-11 13:47:17 -06:00
137af3612b
[gsubgpos] Simplify OT::ClassDefFormat2::intersected_class_glyphs()
2022-05-11 13:39:30 -06:00
3261e05bdb
[subset] Optimize ClassDef1::intersected_class_glyphs() for class0
2022-05-11 13:16:31 -06:00
c78d8ba60b
[subset] Allocate same size as source table for GSUB/GPOS/name
2022-05-11 13:05:41 -06:00
2e7f1ae48f
[subset] Use vector.allocated size instead of tracking buf_size
2022-05-11 12:52:27 -06:00
f08537963b
[cff-subset] Pre-alloc vector for operator decoding
2022-05-11 12:14:49 -06:00
7edd54f3dd
[perf/benchmark-subset] Minor cleanup
2022-05-11 12:14:49 -06:00
aeb50b8942
[subset] Retain buffer across table subset operations
2022-05-11 12:14:49 -06:00
bff78e6515
[cff] Convert interpretation environment to use constructor
2022-05-10 16:50:38 -06:00
de053e2efb
[cff] Convert subr_subset_param_t to use constructor
2022-05-10 16:17:20 -06:00
96140db485
[cff] Convert cff2_extents_param_t to use constructor
2022-05-10 16:17:20 -06:00
54544f2a57
[cff] Convert cff1_extents_param_t to use constructor
2022-05-10 16:17:20 -06:00
377befd0c7
[cff] Convert get_seac_param_t to use constructor
2022-05-10 16:17:20 -06:00
8fd70362fa
[cff] Use hb_ubytes_t() instead of Null(hb_ubytes_t)
2022-05-10 16:17:20 -06:00
9033c7f99d
[cff-common] Optimize INDEX::operator[]
...
Previous try showed slowdown in benchmarks, suprisingly.
Rewrite it keeping the function, hopefully allowing better optimization.
2022-05-10 14:58:53 -06:00
3aace2431b
Revert "[cff-common] Optimize INDEX::operator[]"
...
This reverts commit 9edb03ac7a
2022-05-10 14:54:04 -06:00
b31ef081db
Revert "[cff] Add an unlikely()"
...
This reverts commit 9ba9adb7ed
2022-05-10 14:52:40 -06:00
9ba9adb7ed
[cff] Add an unlikely()
2022-05-10 14:43:00 -06:00
9edb03ac7a
[cff-common] Optimize INDEX::operator[]
2022-05-10 14:43:00 -06:00
52d59bf150
[perf] Make subset benchmark data driven.
2022-05-10 14:30:08 -06:00
0a42410dc8
[cff2] Change extents/shape stack to be just a number
...
Do the blending immediately.
Fixes https://github.com/harfbuzz/harfbuzz/issues/3559
Benchmark on AdobeVFPrototype shows 35% speedup. Now we're faster
than FreeType:
Benchmark Time CPU Time Old Time New CPU Old CPU New
------------------------------------------------------------------------------------------------------------------------------------------------
BM_Font/glyph_extents/AdobeVFPrototype.otf/hb -0.3792 -0.3792 1584 983 1581 982
BM_Font/glyph_extents/AdobeVFPrototype.otf/ft +0.0228 +0.0224 1220 1248 1218 1245
BM_Font/glyph_extents/AdobeVFPrototype.otf/var/hb -0.3513 -0.3518 1616 1048 1613 1046
BM_Font/glyph_extents/AdobeVFPrototype.otf/var/ft +0.0172 +0.0169 1232 1254 1230 1251
2022-05-10 13:41:12 -06:00
5277a5772b
[perf] Add benchmarks for CFF subsetting.
2022-05-10 12:16:44 -06:00
8f9f0c494b
[subset] Enforce cmap12 group ordering constraints in collect_mapping.
...
Fixes fuzzer issue: https://oss-fuzz.com/testcase-detail/6365271012540416
2022-05-10 12:15:09 -06:00
c99ad0f015
Merge pull request #3572 from harfbuzz/cff-stack
...
Cff stack
2022-05-09 18:52:19 -06:00
1b14d2ff13
[cff] Fix arg-stack peek() impl
2022-05-09 18:15:31 -06:00
Behdad Esfahbod
Garret Rieger
jeremiazhao
Thomas Devoogdt