11 Commits

Include renames

Author	SHA1	Message	Date
Henrik Rydgard	2b6ed975e2	Improve fast sse matrix multiply	2014-12-06 00:28:29 +01:00
Sacha	d1cccf2c26	Fix some compiler warnings.	2014-07-17 00:48:47 +10:00
Henrik Rydgård	b532ab8f15	Oops	2014-07-15 18:45:26 +02:00
Henrik Rydgård	fb9f1c2d6a	Fix native tools build on mac etc	2014-07-15 17:24:31 +02:00
Unknown W. Brackets	9a6d1ce7e7	Define fast_matrix_mul_4x4_neon with dual mangling. ... Some compilers/platforms want or don't want the underscore prefix. Both should work fine.	2014-06-20 08:48:44 -07:00
Sacha	d809cf9e37	Remove some redundant armv7 defines. Also, a stray QFile include.	2014-06-11 15:14:20 +10:00
Henrik Rydgård	daf2595eb0	Introduce arch.h, which provides the MSVC compatible arch defines on other platforms.	2014-04-17 00:05:34 +02:00
Henrik Rydgård	56f3a9733d	Add guards to fast_matrix_sse.c	2014-04-06 15:04:37 +02:00
Henrik Rydgård	8c33c6aa09	Fix build for x86 android etc.	2014-03-22 15:17:14 +01:00
Henrik Rydgard	811bc13beb	Have fast_matrix_mul_4x4_sse accept unaligned matrices	2014-03-22 09:41:25 +01:00
Henrik Rydgard	9a49f564ce	fast_math: SIMD function library that can select to use NEON at runtime. ... Only contains one function so far: fast_matrix_mul_4x4. C, SSE, and NEON impls.	2014-03-22 00:31:10 +01:00