Re: [libav-devel] [PATCH 2/2] checkasm: x86: post commit review fixes

On Tue, Dec 22, 2015 at 10:59 PM, Janne Grunau wrote: > Check the full FPU tag word instead of only the upper half and simplify > the comparison. It previously only checked the lower half, not the upper. > Use upper-case function base name as macro name to instantiate

Re: [libav-devel] [PATCH 1/2] x86: zero extend the 32-bit length in int32_to_float_fmul_scalar implicitly

On Tue, Dec 22, 2015 at 10:59 PM, Janne Grunau wrote: > This reverts commit 5dfe4edad63971d669ae456b0bc40ef9364cca80. > --- > libavcodec/x86/fmtconvert.asm | 5 + > 1 file changed, 1 insertion(+), 4 deletions(-) Ok. ___

[libav-devel] av_rdft_init() / av_rdft_calc() for halfcomplex format DFT?

Hey list, I'd like to use libav for some Discrete Fourier Transform work. Specifically, I'd like to generate the following kind of transform in FFTW, but with libav (if possible). The documentation for libav's DFT functions are scant, so I figured I'd ask. If I were using FFTW, I would