325 Commits

Author	SHA1	Message	Date
Dmitry Vagner	1f39053d9a	detailed accounting	2022-10-18 11:23:43 -04:00
Dmitry Vagner	6a6fbec9f5	remove fp6 test	2022-10-18 01:19:24 -04:00
Dmitry Vagner	b9a1b4413d	move fp6mul and add more comments	2022-10-18 00:21:12 -04:00
Dmitry Vagner	cc9559d954	Merge branch 'main' of github.com:mir-protocol/plonky2 into Fp12	2022-10-18 00:08:03 -04:00
Dmitry Vagner	d475ab9368	fp6 passes randomized tests	2022-10-17 23:19:14 -04:00
Dmitry Vagner	b534b221a0	fix fp6 subtraction---fp12 tests running!	2022-10-17 15:58:56 -04:00
Dmitry Vagner	a0f7e6619d	fix fp6, better test function	2022-10-17 14:57:32 -04:00
Dmitry Vagner	3cff0928c4	better test API	2022-10-17 11:02:24 -04:00
Daniel Lubarov	f125786d9b	More directories for MPT logic	2022-10-16 22:55:09 -07:00
Daniel Lubarov	7f366cdace	Treat storage tries as sub-tries of the state trie ... I.e. have leaves in the state trie point to the root of a storage trie	2022-10-16 20:26:37 -07:00
Daniel Lubarov	a8e30b0ca0	Switch a few uses of current-context memory to kernel memory ... E.g. make sure the RLP segment is only used with the kernel context. Using current-context memory would also work, we just need to be consistent. For transaction parsing etc, the context should be 0 anyway, but explicitly referring to kernel memory feels more idiomatic to me.	2022-10-15 08:45:01 -07:00
Daniel Lubarov	ecce5be9e3	MPT format tweaks ... Don't need value lengths in memory. Branches with null values recognized as value_ptr = null.	2022-10-14 18:09:43 -07:00
Dmitry Vagner	adc8c33b0d	Fp12	2022-10-13 20:06:19 -04:00
Dmitry Vagner	fc2ea6281d	Fp6 mul test passes	2022-10-13 18:09:27 -04:00
Dmitry Vagner	d3e2b982ee	new op codes	2022-10-13 17:28:17 -04:00
Dmitry Vagner	e01c91bdf0	Merge branch 'main' of github.com:mir-protocol/plonky2 into Fp12	2022-10-13 17:08:44 -04:00
Dmitry Vagner	d1bad81985	stuff	2022-10-12 10:06:34 -04:00
Daniel Lubarov	299aabf860	Fix branch hashing bug	2022-10-11 15:56:23 -07:00
Daniel Lubarov	caf928b11e	MPT logic for inserts into extension nodes	2022-10-10 18:33:45 -07:00
Daniel Lubarov	50002df8e4	MPT insert into leaf, overlapping keys case	2022-10-10 10:42:02 -07:00
Daniel Lubarov	cad0473e1d	More MPT insert logic	2022-10-10 09:40:28 -07:00
Daniel Lubarov	33dba3a23d	Insertion optimization for leaf case	2022-10-09 20:18:16 -07:00
Daniel Lubarov	4a055b3a76	MPT insert logic, part 2	2022-10-09 11:32:01 -07:00
Daniel Lubarov	6bb1ad94e8	MPT insert logic, part 1	2022-10-08 22:27:57 -07:00
Daniel Lubarov	8ee7265863	Tweak MPT value storage	2022-10-08 13:51:52 -07:00
Daniel Lubarov	817156cd47	Begin MPT insert	2022-10-08 13:23:00 -07:00
Dmitry Vagner	f83504b16e	separate module + stack comments	2022-10-07 18:19:46 -07:00
Dmitry Vagner	7b1db48846	Fp6 mult purely on stack	2022-10-07 18:13:32 -07:00
Dmitry Vagner	00534286a6	Fp12 mult + Fp6 macros	2022-10-07 15:41:46 -07:00
Daniel Lubarov	e6b5e3656f	Some more uses of %increment, %decrement	2022-10-07 12:03:37 -07:00
Dmitry Vagner	d02c9bdda5	Fp6 mult	2022-10-07 09:29:38 -07:00
Daniel Lubarov	f8c104033c	Hash MPT extension nodes	2022-10-06 20:50:32 -07:00
Daniel Lubarov	ed2aac3af3	MPT fixes	2022-10-06 16:28:52 -07:00
Daniel Lubarov	0424fe680d	mload_packing	2022-10-06 14:27:36 -07:00
Daniel Lubarov	f2cb42bbe8	MPT logic to hash branch nodes	2022-10-05 20:46:49 -07:00
Daniel Lubarov	0c9847abf3	Rework MPT hashing to support inlining <32 byte children	2022-10-05 15:23:06 -07:00
Dmitry Vagner	9ebbc033bc	fix padlength issue	2022-10-05 09:45:38 -07:00
Daniel Lubarov	d8bf30150f	Merge pull request #756 from mir-protocol/rlp_fixes ... RLP related fixes	2022-10-04 15:34:38 -07:00
Daniel Lubarov	0ccb340e40	RLP related fixes	2022-10-04 15:18:31 -07:00
Dima V	dd6c5a0d1a	Merge pull request #640 from mir-protocol/ripeMD ... RipeMD	2022-10-03 15:49:04 -07:00
Dmitry Vagner	53014b732f	almost done	2022-10-03 15:30:17 -07:00
Nicholas Ward	d48f63142f	Merge pull request #754 from mir-protocol/sha2_inline_consts ... Inline some SHA2 constants	2022-10-03 15:20:34 -07:00
Dmitry Vagner	1475cddb3d	rearrange	2022-10-03 15:01:59 -07:00
Daniel Lubarov	c03773bab1	Inline some SHA2 constants ... I think `%mload_kernel_code_u32` is good when we need to do random access, but since the indices are constant here, let's just hardcode them like this. This reduces the assembled size of `compression.asm` from 1827 to 1454 bytes. I think there's still a lot more we could do to shrink it, though it's not that important.	2022-10-03 14:59:33 -07:00
Dmitry Vagner	54885fefa0	fix merge problem	2022-10-03 14:57:04 -07:00
Daniel Lubarov	66c28e953d	Unroll num_bytes ... Since it's in some code paths where speed is critical.	2022-10-03 14:42:11 -07:00
Dmitry Vagner	6e5fe43c7e	delete duplicates	2022-10-03 14:31:56 -07:00
Dmitry Vagner	d42250d677	merge	2022-10-03 14:25:33 -07:00
Nicholas Ward	99fb730aea	rearranging and cleanup	2022-10-03 14:07:21 -07:00
Dmitry Vagner	681b6e31e7	minor	2022-10-03 14:01:47 -07:00