564 Commits

Author	SHA1	Message	Date
chriseth	2f7c847ce4	Merge pull request #1279 from ethereum/semver-helper ... Export major/minor/patch helpers on SemVerVersion	2016-10-25 13:20:08 +02:00
Alex Beregszaszi	7a988555c3	Export major/minor/patch helpers on SemVerVersion	2016-10-25 12:13:41 +01:00
chriseth	e00a4b47c0	Merge pull request #1264 from ethereum/988 ... State variable under contract's name	2016-10-25 12:50:24 +02:00
Yoichi Hirai	59f6c18c2b	analysis: determine if a member access on a contract is an l-value	2016-10-24 19:31:57 +02:00
chriseth	f25aa0c68b	More checks for missing mobile type.	2016-10-24 16:45:25 +02:00
Yoichi Hirai	5245a3cf74	codegen: refactor common code	2016-10-24 16:34:23 +02:00
Yoichi Hirai	acba7b92e5	codegen: if a member access has been resolved as a variable, follow that ... This fixes at least the first example in #988	2016-10-24 16:34:23 +02:00
Yoichi Hirai	7cee39fc17	codegen: skip contract L for L.Foo where Foo is a type ... Fixes #1116	2016-10-24 15:30:26 +02:00
chriseth	2b00804d1d	Fix crash in throw.	2016-10-24 14:32:49 +02:00
chriseth	cb1fcaf6f6	Merge pull request #1240 from ethereum/1151 ... ast: super contract type does not contain native members	2016-10-24 11:45:54 +02:00
Yoichi Hirai	9d9380d30e	ast: add a null check	2016-10-21 13:01:45 +02:00
Yoichi Hirai	83e8fdecc5	ast: ContractDefinition::inheritableMembers contains enums as well as structs ... This fixes #1131	2016-10-21 13:01:45 +02:00
Alex Beregszaszi	7b0fb5d06b	Support variable references within modifiers for inline assembly	2016-10-20 23:45:13 +01:00
Yoichi Hirai	7887045a48	ast: simplifications suggested by @chriseth	2016-10-20 16:09:52 +02:00
Yoichi Hirai	ff9039b2a8	When a contract type is super, its members do not contain the functions of itself ... Fixes #1151	2016-10-20 16:09:52 +02:00
Alex Beregszaszi	0e33b63468	Allow warnings for inline assembly block	2016-10-20 13:25:28 +01:00
Alex Beregszaszi	c3330faf21	Issue warnings if stack is not balanced after inline assembly block	2016-10-20 13:25:28 +01:00
Alex Beregszaszi	0dd75ac100	Use warning function in TypeChecker	2016-10-20 13:25:28 +01:00
Alex Beregszaszi	7ac27c957e	Use >> (SAR) to denote constant shifts	2016-10-20 00:03:09 +01:00
Alex Beregszaszi	33250eef9e	Reject negative shifts within constants	2016-10-20 00:03:09 +01:00
RJ	48f2028911	Support shifting constants	2016-10-20 00:03:09 +01:00
Alex Beregszaszi	d18e56661d	Disallow unsupported RValues in inline assembly	2016-10-19 20:44:15 +01:00
Alex Beregszaszi	9616470f67	Disallow magic variables in inline assembly	2016-10-19 20:44:15 +01:00
Alex Beregszaszi	fdd1108c1c	Omit non-convertible bound functions	2016-10-19 10:41:04 +01:00
Alex Beregszaszi	003359a0b6	Ensure that bound functions cannot be defined without self type	2016-10-19 10:41:04 +01:00
Yoichi Hirai	d2470a3e39	super's size on stack is zero, because the expression compiler does not push an address. ... This is different from `this`, which is translated to `ADDRESS` instruction.	2016-10-18 14:17:49 +02:00
Yoichi Hirai	8e11bac8de	Check if a fixedBytes fits an integer type ... before looking up the size of the integer type. Fixes #1150.	2016-10-13 20:17:13 +02:00
Alex Beregszaszi	48ac970677	Merge pull request #1169 from ethereum/inline-assembly-tags ... Fix assignment after tags in inline assembly	2016-10-12 00:00:28 +01:00
chriseth	821fe6e916	Merge pull request #1181 from ethereum/formal_ignore_pragma ... formal: ignore pragmas during Why3 code generation	2016-10-11 16:29:32 +02:00
Alex Beregszaszi	d0791fb365	Fix assignment after tags in inline assembly	2016-10-11 12:28:49 +01:00
Yoichi Hirai	ab1f4632aa	Chack for non-version pragmas	2016-10-11 00:07:11 +02:00
Yoichi Hirai	092e5829d8	formal: ignore pragmas during Why3 code generation ... Fixes #1177	2016-10-11 00:01:29 +02:00
Yoichi Hirai	7ae68e1e47	Fix pragma keyword check ... Fixes #1192	2016-10-11 00:00:22 +02:00
chriseth	d3f410d8a8	Merge pull request #1189 from NicolaiSoeborg/develop ... Add HexLiteral to grammar, fixes #1186	2016-10-10 23:41:55 +02:00
Nicolai	e923a5e190	Fix: HexLiteral must be even number of nibbles ... solc even allow 0 nibbles: bytes a = hex"";	2016-10-10 23:08:44 +02:00
chriseth	0b1d449057	Merge pull request #1168 from ethereum/rename-dev-sha3 ... Rename dev::sha3 to dev::keccak256	2016-10-10 21:23:27 +02:00
Nicolai	f6195c0a46	Add HexLiteral to grammar, fixes #1186	2016-10-10 19:42:42 +02:00
Alex Beregszaszi	ae0477744c	Support address in inline assembly	2016-10-06 22:05:01 +01:00
Alex Beregszaszi	34df80c502	Add alias keccak256() for sha3()	2016-10-06 16:34:41 +02:00
Alex Beregszaszi	aefb6e5fcf	Rename dev::sha3 to dev::keccak256	2016-10-06 14:53:40 +01:00
Alex Beregszaszi	6afdee5958	Support both suicide/selfdestruct in inline assembly	2016-10-06 12:44:33 +01:00
chriseth	212e016091	Merge pull request #1104 from ethereum/fixmemcosts ... Fix memory resize costs during call	2016-09-17 12:30:41 +02:00
chriseth	5a45990458	Access output memory area so that we do not pay for resize during call.	2016-09-17 11:29:54 +02:00
chriseth	727c6fac85	Allow value transfer to library functions.	2016-09-17 11:29:27 +02:00
chriseth	031fd568cc	Fix crash for TypeName[k].	2016-09-15 18:16:34 +02:00
Yoichi Hirai	4337e70cca	Prepare for leaky exceptions ... Now toFormalType() reports errors by exceptions, they will be sometimes leaked to the wider context. This commits adds a catch.	2016-09-09 20:15:13 +02:00
Yoichi Hirai	ac7c6ae7d2	toFormalType reports errors by an exception ... This allows error reporting without passing `ASTNode` to `toFormalType()`	2016-09-09 20:15:08 +02:00
Yoichi Hirai	c861cf579d	Translate mapping types into Why3 arrays when keys are integers ... Even when the keys are signed the translation is supposed to work because Why3 arrays allow negative indices.	2016-09-09 19:11:15 +02:00
chriseth	2c4e9ece07	Merge pull request #1047 from pirapira/address-in-prelude ... formal verification: Add Address module in the WhyML prelude	2016-09-09 17:34:52 +02:00
chriseth	6a55405160	Fix problem with release version string.	2016-09-08 17:56:28 +02:00