Changes between Version 3 and Version 4 of swscale


Ignore:
Timestamp:
Mar 31, 2016, 6:00:38 PM (4 years ago)
Author:
rbultje
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • swscale

    v3 v4  
    1616    - sometimes we overread/overwrite
    1717  * the unscaled paths are almost universally underoptimized (only mmx/mmx2 coverage, very little sse2 and virtually no avx2)
     18  * fast conversion simd often uses contexts to pass around information, which isn't very portable
    1819
    1920A good fix-up of swscale would have the following elements:
     
    3637  * these should ideally not be exposed in the common code (i.e. a grep for x86 in libswscale/*.c should be mostly empty)
    3738  * simd should be allowed to set constraints. E.g., over-reading and -writing of buffers should be enabled in some scenarios where that makes the assembly easier to write. This should be conveyed to the user so they can choose fast conversion by having padded buffers, or a slower conversion by not having padded buffers
    38   * fast conversion simd often uses contexts to pass around information, which isn't very portable
     39  * a "cleaner" SIMD API, see libavcodec/libavfilter/libswresample for examples
    3940- documentation
    4041  * API docs