ffmpeg | branch: master | Andreas Rheinhardt <andreas.rheinha...@googlemail.com> | Sun Nov 4 05:48:40 2018 +0100| [6df9020f45eaff66ba2c2bac98cda9ddaacb03f3] | committer: Mark Thompson
cbs_mpeg2: Improve performance of writing slices Instead of using a combination of bitreader and -writer for copying data, one can byte-align the (obsolete and removed) bitreader to improve performance. One can even use memcpy in the normal case. This improved the time needed for writing the slicedata from 33618 to 2370 decicycles when tested on a video originating from a DVD (4194394 runs). Signed-off-by: Andreas Rheinhardt <andreas.rheinha...@googlemail.com> Signed-off-by: Mark Thompson <s...@jkqxz.net> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=6df9020f45eaff66ba2c2bac98cda9ddaacb03f3 --- libavcodec/cbs_mpeg2.c | 39 +++++++++++++++++++++++++++------------ 1 file changed, 27 insertions(+), 12 deletions(-) diff --git a/libavcodec/cbs_mpeg2.c b/libavcodec/cbs_mpeg2.c index 0df4234b12..8b8b266563 100644 --- a/libavcodec/cbs_mpeg2.c +++ b/libavcodec/cbs_mpeg2.c @@ -264,8 +264,6 @@ static int cbs_mpeg2_write_slice(CodedBitstreamContext *ctx, PutBitContext *pbc) { MPEG2RawSlice *slice = unit->content; - GetBitContext gbc; - size_t bits_left; int err; err = cbs_mpeg2_write_slice_header(ctx, pbc, &slice->header); @@ -273,21 +271,38 @@ static int cbs_mpeg2_write_slice(CodedBitstreamContext *ctx, return err; if (slice->data) { + size_t rest = slice->data_size - (slice->data_bit_start + 7) / 8; + uint8_t *pos = slice->data + slice->data_bit_start / 8; + + av_assert0(slice->data_bit_start >= 0 && + 8 * slice->data_size > slice->data_bit_start); + if (slice->data_size * 8 + 8 > put_bits_left(pbc)) return AVERROR(ENOSPC); - init_get_bits(&gbc, slice->data, slice->data_size * 8); - skip_bits_long(&gbc, slice->data_bit_start); - - while (get_bits_left(&gbc) > 15) - put_bits(pbc, 16, get_bits(&gbc, 16)); + // First copy the remaining bits of the first byte + if (slice->data_bit_start % 8) + put_bits(pbc, 8 - slice->data_bit_start % 8, + *pos++ & MAX_UINT_BITS(8 - slice->data_bit_start % 8)); + + if (put_bits_count(pbc) % 8 == 0) { + // If the writer is aligned at this point, + // memcpy can be used to improve performance. + // This is the normal case. + flush_put_bits(pbc); + memcpy(put_bits_ptr(pbc), pos, rest); + skip_put_bytes(pbc, rest); + } else { + // If not, we have to copy manually: + for (; rest > 3; rest -= 4, pos += 4) + put_bits32(pbc, AV_RB32(pos)); - bits_left = get_bits_left(&gbc); - put_bits(pbc, bits_left, get_bits(&gbc, bits_left)); + for (; rest; rest--, pos++) + put_bits(pbc, 8, *pos); - // Align with zeroes. - while (put_bits_count(pbc) % 8 != 0) - put_bits(pbc, 1, 0); + // Align with zeros + put_bits(pbc, 8 - put_bits_count(pbc) % 8, 0); + } } return 0; _______________________________________________ ffmpeg-cvslog mailing list ffmpeg-cvslog@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-cvslog
