Originally, ATS included a home-grown HPACK encoder and decoder for HTTP/2 header compression, featuring a custom Huffman coding implementation. As HTTP/2 adoption grew and later with HTTP/3 evolving to QPACK, the need for high-performance header decompression became critical. Our original Huffman decoder was simple but performed bit-by-bit processing, which proved to be a performance bottleneck under heavy traffic.
To address this, we replaced the ATS home-grown version with the open-source, MIT-licensed LiteSpeed HPACK library. LiteSpeed's implementation trades static memory (a fixed-size, compile-time-allocated lookup table of 64K entries) for drastically reduced computation. In perf top testing with production traffic, the CPU usage of the huffman_decode function dropped from about 1% to effectively unmeasurable levels, delivering a significant performance improvement.
This integration maintains full HPACK compliance for HTTP/2 and seamlessly supports header compression performance for HTTP/3 (QPACK), leveraging the same Huffman decoding acceleration.
The current implementation pulled into ATS is based upon what is currently the latest version, v2.3.5.
The code is kept as close to upstream as practical, with one deliberate behavioral divergence that any future re-sync must preserve:
lshpack_dec_huff_decode (the fast decoder) rejects Huffman padding of 8 or more bits, as required by RFC 7541 section 5.2. Upstream's fast decoder accepts such padding when it follows the final symbol near the end of the input; the 4-bit FSM decoder (lshpack_dec_huff_decode_full) has always rejected it. See the commented tail check in lshpack.cc and the decode_overlong_padding test in src/proxy/hdrs/unit_tests/test_Huffmancode.cc, which fails if the check is dropped.