5 Commits

Author	SHA1	Message	Date
Paul Crowley	b0d2062abe	Use a constant-time MAX function ... Bug: 146520538 Test: atest HadamardTest Change-Id: Ife1012c14d697141e6ee0c583dc32eaacdb72b73	2020-03-13 04:02:09 -07:00
Paul Crowley	0080bde5fa	Speed up encoding ... Bug: 63928581 Test: atest HadamardTest Change-Id: I1e37a9559892288f76e69fe81a746b77e2bf7495	2019-12-23 21:40:18 -08:00
Paul Crowley	53c005ff34	Shuffle the encoding to prevent systematic errors ... Bug: 63928581 Test: atest HadamardTest Change-Id: Ide8ead4bf5efa629c631df52249cbb322265cc8c	2019-12-23 12:16:31 -08:00
Paul Crowley	c675b182b3	Transpose the encoding matrix ... Stripe together the encodings from each of the 16 codewords, so that if a 512-byte DRAM line is knocked out, it affects 256 bits from each codeword rather than 4096 bits from a single encoded codeword. Rather than using std::bitset, we directly set and read bits in the std::vector<uint8_t>, because the striping means that copying it will now cost not4k in allocation but 64k. Decode directly to a word, without using list decoding. It seems we don't need list decoding for the error rates that matter here, and we never completed the implementation of it anyway. Declare and test only the full interface, now that it doesn't decompose quite so neatly. Bug: 63928581 Test: atest HadamardTest Change-Id: If022d3f4a8d6fccdf68119d4666f83ce5005bccb	2019-12-20 12:03:21 -08:00
Tianjie Xu	a0a12cfcf2	Add hadamard utilities to encode keys ... Add the utility functions to encode & decode 32 bytes keys. The hadamard encoding will expand a 2 bytes word to 2^15 bits. And thus the 32 byte key will expand to 64KiB in space. The encoded value is more robust and we will be able to recover the key even if there is some corruption. Test: unittest pass Change-Id: Iae8a28a8c7c1699f8641f9250f0eccde5c2ff138	2019-12-18 22:45:54 -08:00