HTTP 500 errors for submissions

[holiman]

I uploaded a big batch of signatures, but got quite a lot of http 500. These were the ones causing errors:

addDataPoint(uint256,bool,string)
addDSource(string,byte,uint256)
addOptionChain(uint256,string,uint256,uint256,bytes32,address,int[])
buildCity(string,uint[2],uint[2])
changeVotingRules(uint256,uint256,int)
CharlyLifeLog(string,int)
Congress(uint256,uint256,int,address)
demintTokens(address,uint8)
directionCount(int,int,int,int)
Etheropt(uint256,string,uint256,uint256,bytes32,address,int[])
get_address(string)
getBlockHeader(int)
getConfigUint(bytes32)
getDataPoint(uint256,uint256)
getFeeAmount(int)
getMaxLossAfterTrade(address,uint256,int,int)
getMaxLossAfterTrade(address,uint256,uint256,int,int)
getMoneyness(int,uint256,uint256)
getPrevHash(int)
getUint(bytes32,string)
hint(bytes32,string,bytes20)
hintURL(bytes32,string)
id(Token)
mint(address,string)
mint(address,uint256)
mintBadge(address,uint256)
mint(bytes32)
mintGreen(address,uint256)
mintGrey(address,uint256)
mintToken(address,uint256)
mint(uint256)
mint(uint256,string)
moneySumAtSettlement(address,uint256,int,uint256)
moneySumAtSettlement(address,uint256,uint256,int,uint256,uint256)
oraclize_setProof(byte)
orderMatchTest(uint256,uint256,int,uint256,uint256,address,address,uint256,int)
orderMatchTest(uint256,uint256,uint256,int,uint256,uint256,address,address,int)
orderMatch(uint256,uint256,int,uint256,uint256,address,uint8,bytes32,bytes32,int)
orderMatch(uint256,uint256,uint256,int,uint256,uint256,address,uint8,bytes32,bytes32,int)
personAdd(string,int,int,string)
personUpdateDOB(uint256,int)
personUpdateDOD(uint256,int)
setConfigUint(bytes32,uint256)
setNextID(uint256,int)
setPreviousID(uint256,int)
setProofType(byte)
setRowcol(uint256,uint[2])
setUint(bytes32,string,uint256)
_stringGas(string)
stringToBytes32(string)
unhint(bytes32)
mint(address,uint256)
mint(uint256)
hintURL(bytes32,string)
mintToken(address,uint256)

[holiman]

Added a few

[holiman]

With the additions above and already submitted, these are the results for all method calls over 10K blocks:

Original

Coverage stats: 
Total calls: 41872
Identified : 27384 (65.399312) %
Missed     : 14488 (34.600688) %

After addition
Coverage stats:

Total calls: 41872
Identified : 37011 (88.390810) %
Missed     : 4861 (11.609190) %

[holiman]

Actually, I think these are the only problematic ones left:

mint(address,uint256)
mint(uint256)
hintURL(bytes32,string)
mintToken(address,uint256)

My guess is an overactive regexp which sees mint(address, uint256) and does some replace into mint128...

[pipermerriam]

Seems likely that the regex may be part of the problem. Given my utter over-extended-ness and that this is generally still functional I'm going to leave this be for now. Thanks for submitting all those signatures.

[veox]

Can't submit viable_print(uint256,uint256,uint64) - also gives a HTTP 500 error.

It's added a similar signature viable_print(int256,uint256,uint256,uint64) (first arg is int256 instead of uint256) the first time I tried to do it, with ID 31599, and now makes an ID gap every time I try to submit the "correct" one.

---

EDIT:

The similarity to @holiman's signatures is that the function has an int in the name also, not just in the args.

---

OT: That's not an "actual" signature, BTW - in that I don't have code for it; I'm just "mining" for one with a 0x00000000 selector; reasoning mentioned in the very last paragraph of this article.

[pipermerriam]

@veox it looks like it's a db error trying to create a duplicate entry in one of the tables with a uniqueness constraint. Maybe a race condition with the same signature being submitted twice in quick succession?

[holiman]

There are at least 257 signatures that looks like they have been mangled by this, signatures like uintMax(uint256). I'm submitting them now. Looks like it's working