Closed mikeversteeg closed 1 year ago
Alternatively a new SimdYuva420pToBgraV2
would work for me as well, I would then skip the 444 intermediate step. Either makes me happy :).
I added Base implementation of function SimdYuva444pToBgraV2. Optimizations will be soon.
Super, thanks!
I'm trying to compile but get
Error LNK2019 unresolved external symbol "void __cdecl Simd::Sse41::AbsDifference(unsigned char const *,unsigned __int64,unsigned char const *,unsigned __int64,unsigned char *,unsigned __int64,unsigned __int64,unsigned __int64)" (?AbsDifference@Sse41@Simd@@YAXPEBE_K01PEAE111@Z) referenced in function SimdAbsDifference Simd
Error LNK2019 unresolved external symbol "void __cdecl Simd::Sse41::SvmSumLinear(float const *,float const *,float const *,unsigned __int64,unsigned __int64,float *)" (?SvmSumLinear@Sse41@Simd@@YAXPEBM00_K1PEAM@Z) referenced in function SimdSvmSumLinear Simd
Error LNK2019 unresolved external symbol "void __cdecl Simd::Sse41::AbsDifference(unsigned char const *,unsigned __int64,unsigned char const *,unsigned __int64,unsigned char *,unsigned __int64,unsigned __int64,unsigned __int64)" (?AbsDifference@Sse41@Simd@@YAXPEBE_K01PEAE111@Z) referenced in function "bool __cdecl Test::AbsDifferenceAutoTest(void)" (?AbsDifferenceAutoTest@Test@@YA_NXZ) Test
Error LNK2019 unresolved external symbol "void __cdecl Simd::Sse41::SvmSumLinear(float const *,float const *,float const *,unsigned __int64,unsigned __int64,float *)" (?SvmSumLinear@Sse41@Simd@@YAXPEBM00_K1PEAM@Z) referenced in function "bool __cdecl Test::SvmSumLinearAutoTest(void)" (?SvmSumLinearAutoTest@Test@@YA_NXZ) Test
Thank you for bug report. I fixed this bug.
Compiles fine now. Will let you know how it performs.
I'm interested in
SimdYuva444pToBgraV2
. I currently create a reduced size of an YUV420P by first converting to YUV444P, resize it and then useSimdYuv444pToBgraV2
to display. However if the original is YUVA420P, I currently don't see an efficient way to end up with reduced size BGRA.Thanks.