walkero
/
harfbuzz
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[simd] Comments

This commit is contained in:
Behdad Esfahbod 2019-12-08 18:59:17 -06:00
parent c799742ac1
commit 291d30b1ff
1 changed files with 14 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
src/hb-simd.hh
View File

@ -47,15 +47,21 @@
* be a binary search in the Coverage table, but that's where the * be a binary search in the Coverage table, but that's where the
* hb_set_digest_t speedup came from: hb_set_digest_t are narrow (3 or 4 * hb_set_digest_t speedup came from: hb_set_digest_t are narrow (3 or 4
* integers) structures that implement approximate matching, similar to Bloom * integers) structures that implement approximate matching, similar to Bloom
* Filters or Quotient Filters. These digests do all bitwise operations, so * Filters or Quotient Filters. These digests do all their work using bitwise
* they can be easily vectorized. Combined with a gather operation, or just * operations, so they can be easily vectorized. Combined with a gather
* multiple fetches in a row (which should parallelize) when gather is not * operation, or just multiple fetches in a row (which should parallelize) when
* available. This will allow us to * skip over 8 or 16 glyphs at a time. * gather is not available. This will allow us to skip over 8 or 16 glyphs at
* a time.
* *
* For fast fonts, like simple Latin fonts, like Roboto, the majority of time * For fast fonts, like simple Latin fonts, like Roboto, the majority of time
* is spent in binary searching in the Coverage table of kern and liga lookups. * is spent in binary searching in the Coverage table of kern and liga lookups.
* We can, again, use vector gather and comparison operations to implement a * We can, again, use vector gather and comparison operations to implement a
* 8ary or 16ary search instead of binary search. * 9ary or 17ary search instead of binary search, which will reduce search
* depth by 3x / 4x respectively. It's important to keep in mind that a
* 16-at-a-time 17ary search is /not/ in any way 17 times faster. Only 4 times
* faster at best since the number of search steps compared to binary search is
* log(17)/log(2) ~= 4. That should be taken into account while assessing
* various designs.
* *
* The rest of this files adds facilities to implement those, and possibly * The rest of this files adds facilities to implement those, and possibly
* more. * more.
@ -137,7 +143,9 @@
* example, my 2019 ThinkPad Yoga X1 does *not* support it. We should * example, my 2019 ThinkPad Yoga X1 does *not* support it. We should
* definitely explore that, but not initially. Also, it is possible that the * definitely explore that, but not initially. Also, it is possible that the
* extra memory load that puts will defeat the speedup we can gain from it. * extra memory load that puts will defeat the speedup we can gain from it.
* Must be implemented and measured carefully. * Also do note that for the search usecase, doubling the bitwidth from 256
* to 512, as discussed, only has hard max benefit cap of less than 30%
* speedup (log(17) / log(9)). Must be implemented and measured carefully.
*/ */
/* DESIGN /* DESIGN