* memeq and memzero are not used much and any remotely sane implementation
* is fast enough. memcpy/memmove speed matters in multi-call mode, but
* the kernel image is decompressed in single-call mode, in which only
* memeq and memzero are not used much and any remotely sane implementation
* is fast enough. memcpy/memmove speed matters in multi-call mode, but
* the kernel image is decompressed in single-call mode, in which only
* (LZMA2 stores uncompressible chunks in uncompressed form). Thus, the
* functions below should just be kept small; it's probably not worth
* optimizing for speed.
* (LZMA2 stores uncompressible chunks in uncompressed form). Thus, the
* functions below should just be kept small; it's probably not worth
* optimizing for speed.