walkero
/
harfbuzz
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
12,948 Commits 16 Branches 166 Tags 114 MiB
Commit Graph

12948 Commits

Author SHA1 Message Date
Behdad Esfahbod 8bfeea4828 [subset] Compute set max using previous() 2022-05-05 11:13:57 -06:00
Behdad Esfahbod 00cb8c629d [subset] Don't go into glyf table if it's empty 2022-05-05 11:13:57 -06:00
Behdad Esfahbod 4fe69bc413 [subset] Use del_range in _remove_invalid_gids 2022-05-05 11:13:57 -06:00
Behdad Esfahbod 2a42edccbe [subset] Cosmetic; use set bulk array population instead of for loop 2022-05-05 10:35:54 -06:00
Garret Rieger bc5129d7fa [perf] use option_t in subset benchmark to select between glyphs and codepoint subset. 2022-05-05 10:01:49 -06:00
Behdad Esfahbod 43938ecdc2 [subset] Remove outdated comment 
I tried something like that. It was slower because of the allocations.
 2022-05-04 16:59:28 -06:00
Garret Rieger 6212856ce8 [perf] benchmark subsetting via glyphs. 2022-05-04 16:50:06 -06:00
Behdad Esfahbod 6829dd30ad
Merge pull request #3562 from harfbuzz/subset-cmap-no-qsort 
[subset] In cmap planning, remove a qsort()
 2022-05-04 16:49:45 -06:00
Behdad Esfahbod 50db78ba83 [subset] In cmap planning, remove a qsort() 2022-05-04 16:18:27 -06:00
Behdad Esfahbod 052812b6ba
Merge pull request #3561 from googlefonts/cmap_opt 
[subset] Further cmap subsetting speed optimizations
 2022-05-04 15:38:30 -06:00
Garret Rieger 7cb36e4222 [subset] Re-introduce size threshold in choosing unicode collection method. 
Threshold is needed since the unicodes set might be an inverted set.
 2022-05-04 21:22:26 +00:00
Garret Rieger 42c54eba83 [subset] Presize unicode to gid list to unicodes + glyphs size. 2022-05-04 20:21:43 +00:00
Garret Rieger 7c7c01d28c [subset] Remove switch to alternate unicode collection at large subset sizes. 
Benchmarks show that the first path is always faster even at large subset sizes:

BM_subset_codepoints/subset_roboto/10_median                              +0.0324         +0.0325             0             0             0             0
BM_subset_codepoints/subset_roboto/64_median                              +0.0253         +0.0255             0             1             0             1
BM_subset_codepoints/subset_roboto/512_median                             +0.0126         +0.0128             1             1             1             1
BM_subset_codepoints/subset_roboto/4000_median                            +0.0500         +0.0491             6             7             6             7
BM_subset_codepoints/subset_amiri/10_median                               +0.0338         +0.0332             1             1             1             1
BM_subset_codepoints/subset_amiri/64_median                               +0.0238         +0.0234             1             1             1             1
BM_subset_codepoints/subset_amiri/512_median                              +0.0066         +0.0063             8             8             8             8
BM_subset_codepoints/subset_amiri/4000_median                             -0.0011         -0.0012            13            13            13            13
BM_subset_codepoints/subset_noto_nastaliq_urdu/10_median                  +0.0226         +0.0226             0             0             0             0
BM_subset_codepoints/subset_noto_nastaliq_urdu/64_median                  +0.0047         +0.0044            20            20            20            20
BM_subset_codepoints/subset_noto_nastaliq_urdu/512_median                 +0.0022         +0.0021           165           166           165           166
BM_subset_codepoints/subset_noto_nastaliq_urdu/1000_median                -0.0021         -0.0023           166           166           166           165
BM_subset_codepoints/subset_noto_devangari/10_median                      +0.0054         +0.0054             0             0             0             0
BM_subset_codepoints/subset_noto_devangari/64_median                      +0.0024         +0.0019             0             0             0             0
BM_subset_codepoints/subset_noto_devangari/512_median                     +0.0089         +0.0090             5             5             5             5
BM_subset_codepoints/subset_noto_devangari/1000_median                    -0.0028         -0.0019             5             5             5             5
BM_subset_codepoints/subset_mplus1p/10_median                             +0.0001         +0.0002             0             0             0             0
BM_subset_codepoints/subset_mplus1p/64_median                             +0.0073         +0.0075             1             1             1             1
BM_subset_codepoints/subset_mplus1p/512_median                            +0.0034         +0.0034             1             1             1             1
BM_subset_codepoints/subset_mplus1p/4096_median                           -0.1248         -0.1248             7             6             7             6
BM_subset_codepoints/subset_mplus1p/10000_median                          -0.0885         -0.0885            13            12            13            12
BM_subset_codepoints/subset_notocjk/10_median                             +0.0031         +0.0032             2             2             2             2
BM_subset_codepoints/subset_notocjk/64_median                             -0.0010         -0.0010             2             2             2             2
BM_subset_codepoints/subset_notocjk/512_median                            -0.0023         -0.0023             9             9             9             9
BM_subset_codepoints/subset_notocjk/4096_median                           -0.1725         -0.1726            28            23            28            23
BM_subset_codepoints/subset_notocjk/32768_median                          -0.0277         -0.0287           140           137           140           136
BM_subset_codepoints/subset_notocjk/100000_median                         -0.0929         -0.0926           162           147           162           147
 2022-05-03 22:45:39 +00:00
Garret Rieger f0c04114bc [subset] Embed unicode to gid list vector in subset plan. 2022-05-03 22:02:59 +00:00
Behdad Esfahbod f67e6bf79c [perf/benchmark-font] Add benchmark for glyph_h_advance 2022-05-02 16:59:48 -06:00
Behdad Esfahbod 1c0a3d4d16 [perf/benchmark-font] Add a couple Noto fonts 2022-05-02 16:50:54 -06:00
Behdad Esfahbod 15fa8afb21 Add fast-path for big-endian 32-bit byteswap 
Speeds up cmap format-12 decoding by some 40% as measured by
the newly added test in perf/benchmark-font!
 2022-05-02 16:46:41 -06:00
Behdad Esfahbod 3fff2e9182 [perf/benchmark-font] Cosmetic 2022-05-02 16:42:10 -06:00
Behdad Esfahbod 307d2d8bb6 [cmap] Sprinkle some 'unlikely's 2022-05-02 16:30:22 -06:00
Garret Rieger 85ec5cbcef [subset] In _populate_unicodes_to_retain populate unicodes in order. 
Allows the set insert to take advantage of page lookup cache.
 2022-05-02 22:29:43 +00:00
Behdad Esfahbod 0d1f8dcaf3 [perf/benchmark-font] Actually make nominal_glyph bench work 2022-05-02 16:18:53 -06:00
Behdad Esfahbod 6cf69d10e7 [perf/benchmark-font] Add back testing of is_variable 2022-05-02 16:07:32 -06:00
Behdad Esfahbod 3aa2ff7988 [perf/benchmark-font] Fix build without freetype 2022-05-02 16:01:22 -06:00
Behdad Esfahbod 58a0988b57 [perf/benchmark-font] Benchmark get_nominal_glyph 2022-05-02 15:57:19 -06:00
Behdad Esfahbod 03f16fab58
Merge pull request #3560 from harfbuzz/perf-cleanup 
Perf cleanup
 2022-05-02 15:44:41 -06:00
Garret Rieger 088133d939 [subset] cache cp to new gid list in subset plan. 
This avoids having to recompute the ordered list multiple times during cmap generation.
 2022-05-02 21:29:16 +00:00
Behdad Esfahbod 6d29903e86 [perf/benchmark-font] Parametrize test 2022-05-02 14:03:15 -06:00
Behdad Esfahbod 636c90e81c [perf/perf] Rename to benchmark-font 2022-05-02 13:41:49 -06:00
Behdad Esfahbod 036d03d2e9 [perf/perf] Move all logic to perf-draw, for now 
To be renamed.
 2022-05-02 13:40:13 -06:00
Behdad Esfahbod 746c3c03c5 [perf/perf] Remove ttf-parser backend 2022-05-02 13:27:32 -06:00
Behdad Esfahbod 4aaa0af7d9 [perf/perf] Rely on hb-draw to measure ft performance 2022-05-02 13:06:27 -06:00
Behdad Esfahbod a4522df378
Merge pull request #3558 from harfbuzz/set-optimize 
[perf] hb_set_t optimizations and perf suite improvements
 2022-04-29 18:34:00 -06:00
Garret Rieger 6922a2561f [subset] Change serialize_rangeoffset_glyid back to using iterator. 2022-04-29 23:30:32 +00:00
Garret Rieger c66fd50c26 [subset] in cmap4 serialization save cp to gid iter to memory. 
Iterator accesses are slow and it's iterated multiple times.
 2022-04-29 23:18:53 +00:00
Garret Rieger 17b98563dc [subset] In cmap4 serialization reduce unnessecary calls into the iterator. 
Gives ~20% speedup for large subsets.
 2022-04-29 22:49:02 +00:00
Garret Rieger 5e241094bf [subset] In unicodes cache cleanup if set insert fails. 2022-04-29 22:45:16 +00:00
Behdad Esfahbod 217d38dfc7 Try to fix distcheck 2022-04-29 16:19:10 -06:00
Garret Rieger a424a92ce5 [subset] s/void */intptr_t. 2022-04-29 22:14:03 +00:00
Garret Rieger aad67f5629 [subset] cache results of collect_unicodes. 2022-04-29 22:05:34 +00:00
Behdad Esfahbod 35681b3edb [benchmark-shape] Break lines and shape separately 2022-04-29 16:02:55 -06:00
Behdad Esfahbod be1ac9c572 [benchmark-shape] Data-driven test sets 2022-04-29 15:55:19 -06:00
Behdad Esfahbod ae3efc6424 [perf] Spawn off benchmark-shape from perf runner 2022-04-29 15:37:11 -06:00
Behdad Esfahbod 5f43ce825a [benchmark-set] Split SetLookup into an ordered and random version 2022-04-29 13:39:15 -06:00
Behdad Esfahbod ae9c7b861b [benchmark-set] At least increase needle by one in lookup benchmark 2022-04-29 13:39:04 -06:00
Behdad Esfahbod 68a9b83d15 [benchmark-set] At least increase needle by one in lookup benchmark 2022-04-29 13:28:07 -06:00
Garret Rieger b4236b7de6 [subset] Optimize Cmap4 collect_unicodes. 
Use set add_range() instead of individual add() calls.
 2022-04-29 19:22:00 +00:00
Behdad Esfahbod 5866ec05f5 [benchmark-map] Remove rand() overhead from benchmark 2022-04-29 13:14:41 -06:00
Behdad Esfahbod 067225a86d [set] Optimize const page_for() using last_page_lookup caching 
Similar to previous commit.

This speeds up SetLookup benchmark by 50%, but that's because that
lookup always hits the same page...
 2022-04-29 13:04:36 -06:00
Behdad Esfahbod c283e41ce3 [set] Optimize non-const page_for() using last_page_lookup caching 
This speeds up SetOrderedInsert tests by 15 to 40 percent, and the
subset_mplus1p benchmarks by 9 to 27 percent.
 2022-04-29 12:45:48 -06:00
Behdad Esfahbod dd005911b9 [benchmark-set] Reduce lookup benchmark overhead 
Turnsout 90% was overhead...  Now lookup is in the 4ns ballpark.
 2022-04-29 12:23:53 -06:00