[hvgl] Optimize extremum transforms some more
f31fa0a
[hvgl] Micro-optimize
e7a05b1
[hvgl] Minor use a variable
4ba0692
[hvgl] Speed up transform variations
727b647
[hvgl] Simplify master transforms a bit more
be30bf9
[hvgl] Move variables around
93ddf8f
[hvgl] Split apply_transforms into two loops
49fd66e
[hvgl] Micro-optimize transforms
19114e0
[hvgl] Simplify transforms
b7a9f15
[hvgl] Iterate all four iterators together
3a8d6d6
[hvgl] Micro-optimize
a1fba4e
[hvgl] Speed up simd path
8f505a7
[hvgl] Simplify sanitize more
c111f08
[hvgl] Simplify more sanitize
d4600d8
[hvgl] Simplify sanitize
41ee17e
[hvgl/simd] Only enable if little-endian
4f3c3ee
Revert "[hvgl] Implement AVX"
8e7eeb5
[hvgl] Implement AVX
e7ecc8e
[hvgl] Minor macro def
8bd2467
[hvgl] Micro-optimize
c0f8877
[hvgl] Micro-optimize
5013743
[hvgl] Rewrite a loop
1e8c858
[hvgl] Remove an unnecessary check
07a97ad
[config] Add HB_NO_SIMD
6da4fb1
[hvgl] Micro-optimization leftover
2ba20cf
[hvgl/simd] Remove unaligned SIMD codepath
76017d5
[hvgl/simd] Reduce alignment requirement
c5e3fb5
[hvgl] Implement Apple SIMD 4x4 matrix usage
587b167
[hvgl] Move code
6252d43
[hvgl] Minor micro-optimize
640d778