hevc - Mixtion of Debug exe and release exe for encoder and decoder -
this hevc encoder , decoder have
** encoder , decoder in debug mode ** encoder , decoder in release mode
i know release mode optimize something.
can use debug encoder , release decoder?(i have not verified this)
how release encoder , debug decoder? (i try this, ok)
i guess both should ok, still want know why yes or why not in theoretical level.
not entirely sure asking if e.g. @ linux makefile, you'll see this:
debug: $(make) -c lib/tlibvideoio debug mm32=$(m32) $(make) -c lib/tlibcommon debug mm32=$(m32) $(make) -c lib/tlibdecoder debug mm32=$(m32) $(make) -c lib/tlibencoder debug mm32=$(m32) $(make) -c lib/tappcommon debug mm32=$(m32) $(make) -c app/tappdecoder debug mm32=$(m32) $(make) -c app/tappencoder debug mm32=$(m32) $(make) -c utils/annexbbytecount debug mm32=$(m32) $(make) -c utils/convert_ntombit_ycbcr debug mm32=$(m32) release: $(make) -c lib/tlibvideoio release mm32=$(m32) $(make) -c lib/tlibcommon release mm32=$(m32) $(make) -c lib/tlibdecoder release mm32=$(m32) $(make) -c lib/tlibencoder release mm32=$(m32) $(make) -c lib/tappcommon release mm32=$(m32) $(make) -c app/tappdecoder release mm32=$(m32) $(make) -c app/tappencoder release mm32=$(m32) $(make) -c utils/annexbbytecount release mm32=$(m32) $(make) -c utils/convert_ntombit_ycbcr release mm32=$(m32)
and if follow makefiles, end in makefile.base
contains following part:
# # debug cpp flags debug_cppflags = -g -d_debug # # release cpp release_cppflags = -o3 -ffloat-store -wuninitialized
so there have differences between debug- , release-mode. generated , reconstructed bitstreams identical regardless of use debug-binary or release-binary.
you fine mixing debug , release binaries.
hope helps...
Comments
Post a Comment