[FFmpeg-devel] [PATCH 2/2] avutil/float_dsp: add ff_vector_dmul_{sse2, avx}

Henrik Gramner henrik at gramner.com
Fri Sep 14 17:57:13 EEST 2018

On Fri, Sep 14, 2018 at 4:51 PM, Henrik Gramner <henrik at gramner.com> wrote:
> I can't really think of any scenario where using a 32-bit register
> address operand with a 64-bit destination for LEA is not a mistake.

To clarify on this, using a 32-bit memory operand means the calculated
effective address will be 32-bit, not 64-bit, so if the result exceeds
0xFFFFFFFF it will be truncated and zero-extended (not even
sign-extended like most other x86 things!). It's essentially i big
trap because it really doesn't do what you'd expect it do to.

More information about the ffmpeg-devel mailing list