Commit graph

531 commits

Author SHA1 Message Date
Roman Khimov
b5d3c37a01
Merge pull request #375 from nspcc-dev/pushbytes-short-read-fix-361
vm: fix not failing PUSHBYTES* on short read
2019-09-09 15:02:56 +03:00
Roman Khimov
c59b51e7ab
Merge pull request #376 from nspcc-dev/use-mr-tron-base58-fix-355
crypto: use mr-tron/base58 for base58
2019-09-09 15:02:24 +03:00
Roman Khimov
8db9d2a976 crypto: use mr-tron/base58 for base58
Use more fast and reliable implementation. Add some tests for our wrappers,
deduplicate code in PublicKey.Adress(). Fixes #355.
2019-09-09 13:12:55 +03:00
Roman Khimov
a2a8981979 vm: fix not failing PUSHBYTES* on short read
Add some tests also. Fixes #361.
2019-09-09 12:09:05 +03:00
Roman Khimov
450063de7d
Merge pull request #362 from nspcc-dev/add-missing-opcodes-into-the-list
Add missing opcodes into the list
2019-09-09 11:24:39 +03:00
Roman Khimov
7b0c2781d2 vm: expand instruction list with commands from current VM spec
Current NEO documentation lists them:
https://docs.neo.org/docs/en-us/tooldev/advanced/neo_vm.html

CALL_* instructions were left out because of conflict with golint (but they're
removed in NEO 3.0 anyway, so wasting time on them makes no sense).

Update autogenerated instruction_string.go accordingly.
2019-09-08 20:46:35 +03:00
Roman Khimov
726245c1ec
Merge pull request #360 from nspcc-dev/feature/struct2array
vm: implement Array <-> Struct conversion
2019-09-08 18:38:42 +03:00
Evgenii
bcc8234155 vm: implement Array <-> Struct conversion
Reference VM implementation supports convertation from
Struct to Array via NEWARRAY and vice versa.
https://github.com/neo-project/neo-vm/pull/91
2019-09-08 13:40:21 +03:00
Roman Khimov
8fac66d4af
Merge pull request #371 from nspcc-dev/fix-some-todo-items
Fix some easy TODO items, refs. #244.
2019-09-07 15:35:36 +03:00
Roman Khimov
8c448ab890 util: add some tests to make GetVarSize() fully covered 2019-09-07 10:28:00 +03:00
Roman Khimov
7e8dc9ad66 vm: uncomment TestForLoopBigIter
It works just fine.
2019-09-07 10:10:11 +03:00
Roman Khimov
1c7d203993 compiler: drop TODO comment, #294 was fixed a while ago 2019-09-07 10:09:42 +03:00
Roman Khimov
8169edf04c vm: expand PUSHBYTES instructions
Allows to reliably use String() on any of those. Update autogenerated
instruction_string.go accordingly.
2019-09-06 19:14:47 +03:00
Roman Khimov
dd357c9119
Merge pull request #357 from nspcc-dev/complete-vm-array-instructions
Add UNPACK, REVERSE, REMOVE, plus some tests for PACK. Fixes #195.
2019-09-06 19:13:03 +03:00
Roman Khimov
aefe572df3 vm: implement UNPACK, REVERSE, REMOVE opcodes
Also expand makeStackItem() to accept slices of int for testing
convenience. Fixes #195.
2019-09-06 19:00:34 +03:00
Roman Khimov
f362ffcf3c
Merge pull request #358 from nspcc-dev/drop-redundant-dev-code-part-8
_pkg.dev: drop vm. This is the last one in series of #315, #318, #322, #328, #330, #338 and #351, so if finally fixes #307.
2019-09-06 18:54:30 +03:00
Roman Khimov
d4d1c79546 _pkg.dev: drop vm
This is a nice VM, it has an interesting stack items model, better separation
of elements from stack implementation, simpler stack implementation and a bit
nicer opcode implementation mechanism. At the same time it lacks so many
features and is so differing from our current VM (that is closely tied to the
compiler) that it makes its code very hard to reuse for master's VM
improvement.

Also, some differences are non-obvious to judge in terms of better or
worse. Stack item model seems to be more extensible, but at the same time we
know very well what kind of stack items we have and this doesn't change a
lot. Slice-based stack is simple but it's hard to say which one would perform
better for real-world smart contracts (it has different tradeoffs in
operations complexity).

Based on that, I'm dropping it. Some ideas will be reused during VM
refactoring, but no more than that. Refs. #307.
2019-09-06 15:02:18 +03:00
Roman Khimov
5bd50c7bcd vm: add some tests for PACK instruction 2019-09-06 13:15:30 +03:00
Roman Khimov
0feaaabe61
Merge pull request #356 from nspcc-dev/add-vm-string-instructions
vm: implement CAT, SUBSTR, LEFT and RIGHT
2019-09-06 13:14:44 +03:00
Roman Khimov
b3ed3d24ee vm: implement CAT, SUBSTR, LEFT and RIGHT
String instruction set.
2019-09-06 12:41:27 +03:00
Roman Khimov
c899ec3038
Merge pull request #354 from nspcc-dev/some-vm-opcodes-and-fixes
Implements NZ, PICK, TUCK, XDROP, INVERT. Fixes XTUCK.
2019-09-05 19:55:50 +03:00
Roman Khimov
3f40334979 vm: add generic stub for all unimplemented instructions 2019-09-05 17:35:30 +03:00
Roman Khimov
59de72f446 vm: implement INVERT instruction 2019-09-05 17:35:30 +03:00
Roman Khimov
b494e094c0 vm: add XDROP implementation 2019-09-05 17:35:30 +03:00
Roman Khimov
974f5db6bc vm: add a TUCK implementation with tests
Also fix stack's InsertAt for the edge case of inserting into a position that
currently is empty (the first good TUCK test illustrates this).
2019-09-05 17:35:30 +03:00
Roman Khimov
eb224aeebe vm: add some checks for OVER and PICK, add OVER tests
These actually behaved correctly because of deferred sigsegv recover, but I
think it would be nice to print something more meaningful for users.
2019-09-05 17:35:30 +03:00
Roman Khimov
85831e3e92 vm: fix XTUCK implementation, add tests
The code that we have actually implements XTUCK and not TUCK. And it's a bit
broken, so fix it and add some tests. The most interesting one (that required
to touch stack code) is the one when we have 1 element on the stack and are
trying to tell XTUCK to push 2 elements deep.
2019-09-05 17:34:25 +03:00
Roman Khimov
5f0f9e1638 vm: implement PICK instruction 2019-09-05 15:18:04 +03:00
Roman Khimov
b34a1b6c25 vm: harden tests against VM failures
Add a check for failed VM state, some of tests using logic values (like NZ
test) can succeed even with VM in FAULT state otherwise.
2019-09-05 15:09:08 +03:00
Roman Khimov
fcf9c1213b vm: implement the NZ opcode 2019-09-05 15:05:44 +03:00
Roman Khimov
bcd81bcfb2
Merge pull request #352 from nspcc-dev/drop-bad-crypto-fix-245
Really simplifies our crypto library and fixes #245.
2019-09-05 13:18:17 +03:00
Roman Khimov
e537dc9ee4 keys: improve publick key checks with a check against P
ANSI X9.62 says that if x or y coordinate are greater than or equal to
curve.Params().P, the conversion should return an error (see ANSI X9.62:2005
Section A.5.8 Step b, which invokes Section A.5.5, which does the check and
rejects when x or y are too big.

See https://github.com/golang/go/issues/20482 for more details.
2019-09-05 12:34:12 +03:00
Roman Khimov
f12194f3b0 keys: deduplicate DecodeBytes/DecodeBinary for PrivateKey
They shared prefix logic for no good reason, don't do that.
2019-09-05 12:34:12 +03:00
Roman Khimov
2c3e92923f keys: simplify error handling for PublicKey() and associated
PublicKey() for PrivateKey now just can't fail and it makes no sense to return
an error from it. There is a lot of associated functionality for which this
also is true, so adjust it accordingly and simplify a lot of code.
2019-09-05 12:34:12 +03:00
Roman Khimov
60bc2e8053 keys: simplify PublicKey() for PrivateKey
Public key is just a point, so use the coordinates obtained previously to
initialize the PublicKey structure without jumping through the hoops of
encoding/decoding.
2019-09-05 12:34:12 +03:00
Roman Khimov
f0fbe9f6c9 crypto: drop home-grown elliptic crypto, use crypto/elliptic
As NEO uses P256 we can use standard crypto/elliptic library for almost
everything, the only exception being decompression of the Y coordinate. For
some reason the standard library only supports uncompressed format in its
Marshal()/Unmarshal() functions. elliptic.P256() is known to have
constant-time implementation, so it fixes #245 (and the decompression using
big.Int operates on public key, so nobody really cares about that part being
constant-time).

New decompress function is inspired by
https://stackoverflow.com/questions/46283760, even though the previous one
really did the same thing just in a little less obvious way.
2019-09-05 12:33:27 +03:00
Roman Khimov
0b884b92b3 crypto: use PrivateKey to generate a key pair
It makes no sense to provide an API for throw-away public keys, so obtain it
via a new real keypair generation where appropriate (and that's only needed
for testing).
2019-09-05 09:38:13 +03:00
Roman Khimov
1ffda460bd
Merge pull request #351 from nspcc-dev/drop-redundant-dev-code-part-7
Drop redundant dev code part 7, the next one in series of #315, #318, #322, #328, #330, #338.
2019-09-04 19:53:56 +03:00
fabwa
1af97a7a87
Update README.md 2019-09-04 18:19:22 +02:00
Roman Khimov
cea983acc6 _pkg.dev: drop peer/peermgr/syncmgr
It was a nice attempt, but there is not a lot that we can reuse for the
master branch. Refs. #307.
2019-09-04 15:30:04 +03:00
Roman Khimov
441f1d3bf5 _pkg.dev: drop chain package
Duplicated by blockchain code in core. Refs. #307.
2019-09-04 12:12:02 +03:00
Roman Khimov
e636537844
Merge pull request #340 from nspcc-dev/fix-golint-213
Make our build green again! Fixes #213.
2019-09-03 19:54:29 +03:00
Roman Khimov
2c2f91564f rpc: comment out getBlock and getRawTransaction
After the 8388f4a55e Golint quickly pointed that
these are unused, so comment 'em out for now.
2019-09-03 18:25:19 +03:00
Roman Khimov
6bfcc615b2 *: fix some misspellings
Found by goreport.
2019-09-03 18:20:05 +03:00
Roman Khimov
22bea5eccf core: remove ineffective assignment from test code
Spotted by goreport:
  Line 89: warning: ineffectual assignment to hash (ineffassign)
2019-09-03 18:17:44 +03:00
Roman Khimov
8388f4a55e rpc: unexport methods returning internal structures
Golint:
pkg/rpc/rpc.go:15:67: exported method GetBlock returns unexported type *rpc.response, which can be annoying to use
pkg/rpc/rpc.go:82:64: exported method GetRawTransaction returns unexported type *rpc.response, which can be annoying to use
pkg/rpc/rpc.go:97:52: exported method SendRawTransaction returns unexported type *rpc.response, which can be annoying to use

Refs. #213.
2019-09-03 18:16:48 +03:00
Roman Khimov
1fd5e5928c rpc: fix bad name spellings noted by golint
pkg/rpc/neoScanBalanceGetter.go:54:56: method parameter assetIdUint should be assetIDUint
pkg/rpc/neoScanBalanceGetter.go:62:3: var assetId should be assetID
pkg/rpc/server_test.go:27:5: var testRpcCases should be testRPCCases
pkg/rpc/txTypes.go:19:3: struct field assetId should be assetID
pkg/rpc/txTypes.go:39:35: interface method parameter assetId should be assetID
pkg/rpc/types.go:115:2: struct field TxId should be TxID

Refs. #213.
2019-09-03 18:16:48 +03:00
Roman Khimov
db9e37b3bb *: fix golint's omit type suggestions
pkg/core/transaction/attribute.go:67:14: should omit type uint8 from declaration of var urllen; it will be inferred from the right-hand side
pkg/crypto/keys/publickey.go:184:8: should omit type []byte from declaration of var b; it will be inferred from the right-hand side
pkg/network/payload/version_test.go:15:12: should omit type bool from declaration of var relay; it will be inferred from the right-hand side

Refs. #213.
2019-09-03 18:16:48 +03:00
Roman Khimov
a74518691c crypto: improve error messages for base58
Golint:
  error strings should not be capitalized or end with punctuation or a newline

Refs. #213.
2019-09-03 18:16:48 +03:00
Roman Khimov
89ed00ddfb core: simplify GetTransationResults() based on golint suggestion
Golint:
  pkg/core/blockchain.go:796:9: if block ends with a return statement, so drop
    this else and outdent its block (move short variable declaration to its own
    line if necessary)

Refs. #213.
2019-09-03 18:16:48 +03:00