Analyzing data flow in Swift¶

You can use CodeQL to track the flow of data through a Swift program to places where the data is used.

Note

CodeQL analysis for Swift is currently in beta. During the beta, analysis of Swift code, and the accompanying documentation, will not be as comprehensive as for other languages.

About this article¶

This article describes how data flow analysis is implemented in the CodeQL libraries for Swift and includes examples to help you write your own data flow queries. The following sections describe how to use the libraries for local data flow, global data flow, and taint tracking. For a more general introduction to modeling data flow, see “About data flow analysis.”

Local data flow¶

Local data flow tracks the flow of data within a single function. Local data flow is easier, faster, and more precise than global data flow. Before looking at more complex tracking, you should always consider local tracking because it is sufficient for many queries.

Using local data flow¶

You can use the local data flow library by importing the DataFlow module. The library uses the class Node to represent any element through which data can flow. The Node class has a number of useful subclasses, such as ExprNode for expressions and ParameterNode for parameters. You can map between data flow nodes and expressions/control-flow nodes using the member predicates asExpr and getCfgNode:

class Node {
  /**
   * Gets the expression that corresponds to this node, if any.
   */
  Expr asExpr() { ... }

  /**
   * Gets the control flow node that corresponds to this data flow node.
   */
  ControlFlowNode getCfgNode() { ... }

  ...
}

You can use the predicates exprNode and parameterNode to map from expressions and parameters to their data-flow node:

/**
 * Gets a node corresponding to expression `e`.
 */
ExprNode exprNode(DataFlowExpr e) { result.asExpr() = e }

/**
 * Gets the node corresponding to the value of parameter `p` at function entry.
 */
ParameterNode parameterNode(DataFlowParameter p) { result.getParameter() = p }

There can be multiple data-flow nodes associated with a single expression node in the AST.

The predicate localFlowStep(Node nodeFrom, Node nodeTo) holds if there is an immediate data flow edge from the node nodeFrom to the node nodeTo. You can apply the predicate recursively, by using the + and * operators, or you can use the predefined recursive predicate localFlow.

For example, you can find flow from an expression source to an expression sink in zero or more local steps:

DataFlow::localFlow(DataFlow::exprNode(source), DataFlow::exprNode(sink))

Using local taint tracking¶

Local taint tracking extends local data flow to include flow steps where values are not preserved, such as string manipulation. For example:

temp = x
y = temp + ", " + temp

If x is a tainted string then y is also tainted.

The local taint tracking library is in the module TaintTracking. Like local data flow, a predicate localTaintStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo) holds if there is an immediate taint propagation edge from the node nodeFrom to the node nodeTo. You can apply the predicate recursively, by using the + and * operators, or you can use the predefined recursive predicate localTaint.

For example, you can find taint propagation from an expression source to an expression sink in zero or more local steps:

TaintTracking::localTaint(DataFlow::exprNode(source), DataFlow::exprNode(sink))

Examples of local data flow¶

This query finds the format argument passed into each call to String.init(format:_:):

import swift

from CallExpr call, Method method
where
  call.getStaticTarget() = method and
  method.hasQualifiedName("String", "init(format:_:)")
select call.getArgument(0).getExpr()

Unfortunately this will only give the expression in the argument, not the values which could be passed to it. So we use local data flow to find all expressions that flow into the argument:

import swift
import codeql.swift.dataflow.DataFlow

from CallExpr call, Method method, Expr sourceExpr, Expr sinkExpr
where
  call.getStaticTarget() = method and
  method.hasQualifiedName("String", "init(format:_:)") and
  sinkExpr = call.getArgument(0).getExpr() and
  DataFlow::localFlow(DataFlow::exprNode(sourceExpr), DataFlow::exprNode(sinkExpr))
select sourceExpr, sinkExpr

We can vary the source, for example, making the source the parameter of a function rather than an expression. The following query finds where a parameter is used for the format:

import swift
import codeql.swift.dataflow.DataFlow

from CallExpr call, Method method, ParamDecl sourceParam, Expr sinkExpr
where
  call.getStaticTarget() = method and
  method.hasQualifiedName("String", "init(format:_:)") and
  sinkExpr = call.getArgument(0).getExpr() and
  DataFlow::localFlow(DataFlow::parameterNode(sourceParam), DataFlow::exprNode(sinkExpr))
select sourceParam, sinkExpr

The following example finds calls to String.init(format:_:) where the format string is not a hard-coded string literal:

import swift
import codeql.swift.dataflow.DataFlow

from CallExpr call, Method method, DataFlow::Node sinkNode
where
  call.getStaticTarget() = method and
  method.hasQualifiedName("String", "init(format:_:)") and
  sinkNode.asExpr() = call.getArgument(0).getExpr() and
  not exists(StringLiteralExpr sourceLiteral |
    DataFlow::localFlow(DataFlow::exprNode(sourceLiteral), sinkNode)
  )
select call, "Format argument to " + method.getName() + " isn't hard-coded."

Global data flow¶

Global data flow tracks data flow throughout the entire program, and is therefore more powerful than local data flow. However, global data flow is less precise than local data flow, and the analysis typically requires significantly more time and memory to perform.

Note

You can model data flow paths in CodeQL by creating path queries. To view data flow paths generated by a path query in CodeQL for VS Code, you need to make sure that it has the correct metadata and select clause. For more information, see Creating path queries.

Using global data flow¶

You can use the global data flow library by implementing the module DataFlow::ConfigSig:

import codeql.swift.dataflow.DataFlow

module MyDataFlowConfiguration implements DataFlow::ConfigSig {
  predicate isSource(DataFlow::Node source) {
    ...
  }

  predicate isSink(DataFlow::Node sink) {
    ...
  }
}

module MyDataFlow = DataFlow::Global<MyDataFlowConfiguration>;

These predicates are defined in the configuration:

isSource - defines where data may flow from.
isSink - defines where data may flow to.
isBarrier - optionally, restricts the data flow.
isAdditionalFlowStep - optionally, adds additional flow steps.

The last line (module MyDataFlow = ...) instantiates the parameterized module for data flow analysis by passing the configuration to the parameterized module. Data flow analysis can then be performed using MyDataFlow::flow(DataFlow::Node source, DataFlow::Node sink):

from DataFlow::Node source, DataFlow::Node sink
where MyDataFlow::flow(source, sink)
select source, "Dataflow to $@.", sink, sink.toString()

Using global taint tracking¶

Global taint tracking is to global data flow what local taint tracking is to local data flow. That is, global taint tracking extends global data flow with additional non-value-preserving steps. The global taint tracking library uses the same configuration module as the global data flow library. You can perform taint flow analysis using TaintTracking::Global:

module MyTaintFlow = TaintTracking::Global<MyDataFlowConfiguration>;

from DataFlow::Node source, DataFlow::Node sink
where MyTaintFlow::flow(source, sink)
select source, "Taint flow to $@.", sink, sink.toString()

Predefined sources¶

The data flow library module codeql.swift.dataflow.FlowSources contains a number of predefined sources that you can use to write security queries to track data flow and taint flow.

The class RemoteFlowSource represents data flow from remote network inputs and from other applications.
The class LocalFlowSource represents data flow from local user input.
The class FlowSource includes both of the above.

Examples of global data flow¶

The following global taint-tracking query finds places where a string literal is used in a function call argument named “password”.

Since this is a taint-tracking query, the TaintTracking::Global module is used.
The isSource predicate defines sources as any StringLiteralExpr.
The isSink predicate defines sinks as arguments to a CallExpr called “password”.
The sources and sinks may need tuning to a particular use, for example, if passwords are represented by a type other than String or passed in arguments of a different name than “password”.

import swift
import codeql.swift.dataflow.DataFlow
import codeql.swift.dataflow.TaintTracking

module ConstantPasswordConfig implements DataFlow::ConfigSig {
  predicate isSource(DataFlow::Node node) { node.asExpr() instanceof StringLiteralExpr }

  predicate isSink(DataFlow::Node node) {
    // any argument called `password`
    exists(CallExpr call | call.getArgumentWithLabel("password").getExpr() = node.asExpr())
  }

module ConstantPasswordFlow = TaintTracking::Global<ConstantPasswordConfig>;

from DataFlow::Node sourceNode, DataFlow::Node sinkNode
where ConstantPasswordFlow::flow(sourceNode, sinkNode)
select sinkNode, "The value $@ is used as a constant password.", sourceNode, sourceNode.toString()

The following global taint-tracking query finds places where a value from a remote or local user input is used as an argument to the SQLite Connection.execute(_:) function.

Since this is a taint-tracking query, the TaintTracking::Global module is used.
The isSource predicate defines sources as a FlowSource (remote or local user input).
The isSink predicate defines sinks as the first argument in any call to Connection.execute(_:).

import swift
import codeql.swift.dataflow.DataFlow
import codeql.swift.dataflow.TaintTracking
import codeql.swift.dataflow.FlowSources

module SqlInjectionConfig implements DataFlow::ConfigSig {
  predicate isSource(DataFlow::Node node) { node instanceof FlowSource }

  predicate isSink(DataFlow::Node node) {
    exists(CallExpr call |
      call.getStaticTarget().(Method).hasQualifiedName("Connection", "execute(_:)") and
      call.getArgument(0).getExpr() = node.asExpr()
    )
  }
}

module SqlInjectionFlow = TaintTracking::Global<SqlInjectionConfig>;

from DataFlow::Node sourceNode, DataFlow::Node sinkNode
where SqlInjectionFlow::flow(sourceNode, sinkNode)
select sinkNode, "This query depends on a $@.", sourceNode, "user-provided value"