C++: Add alert provenance plumbing.

aschackmull · aschackmull · commit 3f986bda4f96 · 2024-02-07T11:37:45.000+01:00
diff --git a/cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowPrivate.qll b/cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowPrivate.qll
@@ -282,6 +282,10 @@ predicate lambdaCall(DataFlowCall call, LambdaCallKind kind, Node receiver) { no
 /** Extra data-flow steps needed for lambda flow analysis. */
 predicate additionalLambdaFlowStep(Node nodeFrom, Node nodeTo, boolean preservesValue) { none() }
 
+predicate knownSourceModel(Node source, string model) { none() }
+
+predicate knownSinkModel(Node sink, string model) { none() }
+
 /**
  * Holds if flow is allowed to pass from parameter `p` and back to itself as a
  * side-effect, resulting in a summary from `p` to itself.
diff --git a/cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowUtil.qll b/cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowUtil.qll
@@ -514,7 +514,7 @@ private module ThisFlow {
  */
 cached
 predicate localFlowStep(Node nodeFrom, Node nodeTo) {
-  simpleLocalFlowStep(nodeFrom, nodeTo)
+  simpleLocalFlowStep(nodeFrom, nodeTo, _)
   or
   // Field flow is not strictly a "step" but covers the whole function
   // transitively. There's no way to get a step-like relation out of the global
@@ -528,64 +528,67 @@ predicate localFlowStep(Node nodeFrom, Node nodeTo) {
  * This is the local flow predicate that's used as a building block in global
  * data flow. It may have less flow than the `localFlowStep` predicate.
  */
-predicate simpleLocalFlowStep(Node nodeFrom, Node nodeTo) {
-  // Expr -> Expr
-  exprToExprStep_nocfg(nodeFrom.asExpr(), nodeTo.asExpr())
-  or
-  // Assignment -> LValue post-update node
-  //
-  // This is used for assignments whose left-hand side is not a variable
-  // assignment or a storeStep but is still modeled by other means. It could be
-  // a call to `operator*` or `operator[]` where taint should flow to the
-  // post-update node of the qualifier.
-  exists(AssignExpr assign |
-    nodeFrom.asExpr() = assign and
-    nodeTo.(PostUpdateNode).getPreUpdateNode().asExpr() = assign.getLValue()
-  )
-  or
-  // Node -> FlowVar -> VariableAccess
-  exists(FlowVar var |
-    (
-      exprToVarStep(nodeFrom.asExpr(), var)
-      or
-      varSourceBaseCase(var, nodeFrom.asParameter())
-      or
-      varSourceBaseCase(var, nodeFrom.asUninitialized())
-      or
-      var.definedPartiallyAt(nodeFrom.asPartialDefinition())
-    ) and
-    varToNodeStep(var, nodeTo)
-  )
-  or
-  // Expr -> DefinitionByReferenceNode
-  exprToDefinitionByReferenceStep(nodeFrom.asExpr(), nodeTo.asDefiningArgument())
-  or
-  // `this` -> adjacent-`this`
-  ThisFlow::adjacentThisRefs(nodeFrom, nodeTo)
-  or
-  // post-update-`this` -> following-`this`-ref
-  ThisFlow::adjacentThisRefs(nodeFrom.(PostUpdateNode).getPreUpdateNode(), nodeTo)
-  or
-  // In `f(&x->a)`, this step provides the flow from post-`&` to post-`x->a`,
-  // from which there is field flow to `x` via reverse read.
-  exists(PartialDefinition def, Expr inner, Expr outer |
-    def.definesExpressions(inner, outer) and
-    inner = nodeTo.(InnerPartialDefinitionNode).getPreUpdateNode().asExpr() and
-    outer = nodeFrom.(PartialDefinitionNode).getPreUpdateNode().asExpr()
-  )
-  or
-  // Reverse flow: data that flows from the post-update node of a reference
-  // returned by a function call, back into the qualifier of that function.
-  // This allows data to flow 'in' through references returned by a modeled
-  // function such as `operator[]`.
-  exists(DataFlowFunction f, Call call, FunctionInput inModel, FunctionOutput outModel |
-    call.getTarget() = f and
-    inModel.isReturnValueDeref() and
-    outModel.isQualifierObject() and
-    f.hasDataFlow(inModel, outModel) and
-    nodeFrom.(PostUpdateNode).getPreUpdateNode().asExpr() = call and
-    nodeTo.asDefiningArgument() = call.getQualifier()
-  )
+predicate simpleLocalFlowStep(Node nodeFrom, Node nodeTo, string model) {
+  (
+    // Expr -> Expr
+    exprToExprStep_nocfg(nodeFrom.asExpr(), nodeTo.asExpr())
+    or
+    // Assignment -> LValue post-update node
+    //
+    // This is used for assignments whose left-hand side is not a variable
+    // assignment or a storeStep but is still modeled by other means. It could be
+    // a call to `operator*` or `operator[]` where taint should flow to the
+    // post-update node of the qualifier.
+    exists(AssignExpr assign |
+      nodeFrom.asExpr() = assign and
+      nodeTo.(PostUpdateNode).getPreUpdateNode().asExpr() = assign.getLValue()
+    )
+    or
+    // Node -> FlowVar -> VariableAccess
+    exists(FlowVar var |
+      (
+        exprToVarStep(nodeFrom.asExpr(), var)
+        or
+        varSourceBaseCase(var, nodeFrom.asParameter())
+        or
+        varSourceBaseCase(var, nodeFrom.asUninitialized())
+        or
+        var.definedPartiallyAt(nodeFrom.asPartialDefinition())
+      ) and
+      varToNodeStep(var, nodeTo)
+    )
+    or
+    // Expr -> DefinitionByReferenceNode
+    exprToDefinitionByReferenceStep(nodeFrom.asExpr(), nodeTo.asDefiningArgument())
+    or
+    // `this` -> adjacent-`this`
+    ThisFlow::adjacentThisRefs(nodeFrom, nodeTo)
+    or
+    // post-update-`this` -> following-`this`-ref
+    ThisFlow::adjacentThisRefs(nodeFrom.(PostUpdateNode).getPreUpdateNode(), nodeTo)
+    or
+    // In `f(&x->a)`, this step provides the flow from post-`&` to post-`x->a`,
+    // from which there is field flow to `x` via reverse read.
+    exists(PartialDefinition def, Expr inner, Expr outer |
+      def.definesExpressions(inner, outer) and
+      inner = nodeTo.(InnerPartialDefinitionNode).getPreUpdateNode().asExpr() and
+      outer = nodeFrom.(PartialDefinitionNode).getPreUpdateNode().asExpr()
+    )
+    or
+    // Reverse flow: data that flows from the post-update node of a reference
+    // returned by a function call, back into the qualifier of that function.
+    // This allows data to flow 'in' through references returned by a modeled
+    // function such as `operator[]`.
+    exists(DataFlowFunction f, Call call, FunctionInput inModel, FunctionOutput outModel |
+      call.getTarget() = f and
+      inModel.isReturnValueDeref() and
+      outModel.isQualifierObject() and
+      f.hasDataFlow(inModel, outModel) and
+      nodeFrom.(PostUpdateNode).getPreUpdateNode().asExpr() = call and
+      nodeTo.asDefiningArgument() = call.getQualifier()
+    )
+  ) and
+  model = ""
 }
 
 /**
diff --git a/cpp/ql/lib/semmle/code/cpp/dataflow/internal/TaintTrackingUtil.qll b/cpp/ql/lib/semmle/code/cpp/dataflow/internal/TaintTrackingUtil.qll
@@ -30,8 +30,8 @@ predicate localTaintStep(DataFlow::Node src, DataFlow::Node sink) {
  * Holds if the additional step from `src` to `sink` should be included in all
  * global taint flow configurations.
  */
-predicate defaultAdditionalTaintStep(DataFlow::Node src, DataFlow::Node sink) {
-  localAdditionalTaintStep(src, sink)
+predicate defaultAdditionalTaintStep(DataFlow::Node src, DataFlow::Node sink, string model) {
+  localAdditionalTaintStep(src, sink) and model = ""
 }
 
 /**
diff --git a/cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowPrivate.qll b/cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowPrivate.qll
@@ -1020,6 +1020,10 @@ predicate lambdaCall(DataFlowCall call, LambdaCallKind kind, Node receiver) { no
 /** Extra data-flow steps needed for lambda flow analysis. */
 predicate additionalLambdaFlowStep(Node nodeFrom, Node nodeTo, boolean preservesValue) { none() }
 
+predicate knownSourceModel(Node source, string model) { none() }
+
+predicate knownSinkModel(Node sink, string model) { none() }
+
 /**
  * Holds if flow is allowed to pass from parameter `p` and back to itself as a
  * side-effect, resulting in a summary from `p` to itself.
diff --git a/cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowUtil.qll b/cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowUtil.qll
@@ -1874,7 +1874,7 @@ private module Cached {
    * (intra-procedural) step.
    */
   cached
-  predicate localFlowStep(Node nodeFrom, Node nodeTo) { simpleLocalFlowStep(nodeFrom, nodeTo) }
+  predicate localFlowStep(Node nodeFrom, Node nodeTo) { simpleLocalFlowStep(nodeFrom, nodeTo, _) }
 
   private predicate indirectionOperandFlow(RawIndirectOperand nodeFrom, Node nodeTo) {
     nodeFrom != nodeTo and
@@ -1944,41 +1944,45 @@ private module Cached {
    * data flow. It may have less flow than the `localFlowStep` predicate.
    */
   cached
-  predicate simpleLocalFlowStep(Node nodeFrom, Node nodeTo) {
-    // Post update node -> Node flow
-    Ssa::postUpdateFlow(nodeFrom, nodeTo)
-    or
-    // Def-use/Use-use flow
-    Ssa::ssaFlow(nodeFrom, nodeTo)
-    or
-    // Operand -> Instruction flow
-    simpleInstructionLocalFlowStep(nodeFrom.asOperand(), nodeTo.asInstruction())
-    or
-    // Instruction -> Operand flow
-    exists(Instruction iFrom, Operand opTo |
-      iFrom = nodeFrom.asInstruction() and opTo = nodeTo.asOperand()
-    |
-      simpleOperandLocalFlowStep(iFrom, opTo) and
-      // Omit when the instruction node also represents the operand.
-      not iFrom = Ssa::getIRRepresentationOfOperand(opTo)
-    )
-    or
-    // Phi node -> Node flow
-    Ssa::fromPhiNode(nodeFrom, nodeTo)
-    or
-    // Indirect operand -> (indirect) instruction flow
-    indirectionOperandFlow(nodeFrom, nodeTo)
-    or
-    // Indirect instruction -> indirect operand flow
-    indirectionInstructionFlow(nodeFrom, nodeTo)
+  predicate simpleLocalFlowStep(Node nodeFrom, Node nodeTo, string model) {
+    (
+      // Post update node -> Node flow
+      Ssa::postUpdateFlow(nodeFrom, nodeTo)
+      or
+      // Def-use/Use-use flow
+      Ssa::ssaFlow(nodeFrom, nodeTo)
+      or
+      // Operand -> Instruction flow
+      simpleInstructionLocalFlowStep(nodeFrom.asOperand(), nodeTo.asInstruction())
+      or
+      // Instruction -> Operand flow
+      exists(Instruction iFrom, Operand opTo |
+        iFrom = nodeFrom.asInstruction() and opTo = nodeTo.asOperand()
+      |
+        simpleOperandLocalFlowStep(iFrom, opTo) and
+        // Omit when the instruction node also represents the operand.
+        not iFrom = Ssa::getIRRepresentationOfOperand(opTo)
+      )
+      or
+      // Phi node -> Node flow
+      Ssa::fromPhiNode(nodeFrom, nodeTo)
+      or
+      // Indirect operand -> (indirect) instruction flow
+      indirectionOperandFlow(nodeFrom, nodeTo)
+      or
+      // Indirect instruction -> indirect operand flow
+      indirectionInstructionFlow(nodeFrom, nodeTo)
+    ) and
+    model = ""
     or
     // Flow through modeled functions
-    modelFlow(nodeFrom, nodeTo)
+    modelFlow(nodeFrom, nodeTo, model)
     or
     // Reverse flow: data that flows from the definition node back into the indirection returned
     // by a function. This allows data to flow 'in' through references returned by a modeled
     // function such as `operator[]`.
-    reverseFlow(nodeFrom, nodeTo)
+    reverseFlow(nodeFrom, nodeTo) and
+    model = ""
   }
 
   private predicate simpleInstructionLocalFlowStep(Operand opFrom, Instruction iTo) {
@@ -1993,12 +1997,13 @@ private module Cached {
     opTo.getDef() = iFrom
   }
 
-  private predicate modelFlow(Node nodeFrom, Node nodeTo) {
+  private predicate modelFlow(Node nodeFrom, Node nodeTo, string model) {
     exists(
       CallInstruction call, DataFlowFunction func, FunctionInput modelIn, FunctionOutput modelOut
     |
       call.getStaticCallTarget() = func and
-      func.hasDataFlow(modelIn, modelOut)
+      func.hasDataFlow(modelIn, modelOut) and
+      model = "DataFlowFunction"
     |
       nodeFrom = callInput(call, modelIn) and
       nodeTo = callOutput(call, modelOut)
diff --git a/cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/TaintTrackingUtil.qll b/cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/TaintTrackingUtil.qll
@@ -14,7 +14,7 @@ private import SsaInternals as Ssa
 predicate localTaintStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo) {
   DataFlow::localFlowStep(nodeFrom, nodeTo)
   or
-  localAdditionalTaintStep(nodeFrom, nodeTo)
+  localAdditionalTaintStep(nodeFrom, nodeTo, _)
 }
 
 /**
@@ -23,20 +23,23 @@ predicate localTaintStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo) {
  * different objects.
  */
 cached
-predicate localAdditionalTaintStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo) {
-  operandToInstructionTaintStep(nodeFrom.asOperand(), nodeTo.asInstruction())
+predicate localAdditionalTaintStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo, string model) {
+  operandToInstructionTaintStep(nodeFrom.asOperand(), nodeTo.asInstruction()) and
+  model = ""
   or
-  modeledTaintStep(nodeFrom, nodeTo)
+  modeledTaintStep(nodeFrom, nodeTo, model)
   or
   // Flow from (the indirection of) an operand of a pointer arithmetic instruction to the
   // indirection of the pointer arithmetic instruction. This provides flow from `source`
   // in `x[source]` to the result of the associated load instruction.
   exists(PointerArithmeticInstruction pai, int indirectionIndex |
     nodeHasOperand(nodeFrom, pai.getAnOperand(), pragma[only_bind_into](indirectionIndex)) and
     hasInstructionAndIndex(nodeTo, pai, indirectionIndex + 1)
-  )
+  ) and
+  model = ""
   or
-  any(Ssa::Indirection ind).isAdditionalTaintStep(nodeFrom, nodeTo)
+  any(Ssa::Indirection ind).isAdditionalTaintStep(nodeFrom, nodeTo) and
+  model = ""
 }
 
 /**
@@ -113,8 +116,8 @@ predicate localExprTaint(Expr e1, Expr e2) {
  * Holds if the additional step from `src` to `sink` should be included in all
  * global taint flow configurations.
  */
-predicate defaultAdditionalTaintStep(DataFlow::Node src, DataFlow::Node sink) {
-  localAdditionalTaintStep(src, sink)
+predicate defaultAdditionalTaintStep(DataFlow::Node src, DataFlow::Node sink, string model) {
+  localAdditionalTaintStep(src, sink, model)
 }
 
 /**
@@ -134,7 +137,7 @@ predicate defaultTaintSanitizer(DataFlow::Node node) { none() }
  * Holds if taint can flow from `nodeIn` to `nodeOut` through a call to a
  * modeled function.
  */
-predicate modeledTaintStep(DataFlow::Node nodeIn, DataFlow::Node nodeOut) {
+predicate modeledTaintStep(DataFlow::Node nodeIn, DataFlow::Node nodeOut, string model) {
   // Normal taint steps
   exists(CallInstruction call, TaintFunction func, FunctionInput modelIn, FunctionOutput modelOut |
     call.getStaticCallTarget() = func and
@@ -143,7 +146,8 @@ predicate modeledTaintStep(DataFlow::Node nodeIn, DataFlow::Node nodeOut) {
     nodeIn = callInput(call, modelIn) and nodeOut = callOutput(call, modelOut)
     or
     exists(int d | nodeIn = callInput(call, modelIn, d) and nodeOut = callOutput(call, modelOut, d))
-  )
+  ) and
+  model = "TaintFunction"
   or
   // Taint flow from one argument to another and data flow from an argument to a
   // return value. This happens in functions like `strcat` and `memcpy`. We
@@ -160,7 +164,8 @@ predicate modeledTaintStep(DataFlow::Node nodeIn, DataFlow::Node nodeOut) {
     func.(TaintFunction).hasTaintFlow(modelIn, modelMidOut) and
     func.(DataFlowFunction).hasDataFlow(modelMidIn, modelOut) and
     modelMidOut.isParameterDeref(indexMid) and
-    modelMidIn.isParameter(indexMid)
+    modelMidIn.isParameter(indexMid) and
+    model = "TaintFunction"
   )
   or
   // Taint flow from a pointer argument to an output, when the model specifies flow from the deref
@@ -173,9 +178,11 @@ predicate modeledTaintStep(DataFlow::Node nodeIn, DataFlow::Node nodeOut) {
     indirectArgument.hasAddressOperandAndIndirectionIndex(nodeIn.asOperand(), _) and
     call.getStaticCallTarget() = func and
     (
-      func.(DataFlowFunction).hasDataFlow(modelIn, modelOut)
+      func.(DataFlowFunction).hasDataFlow(modelIn, modelOut) and
+      model = "DataFlowFunction"
       or
-      func.(TaintFunction).hasTaintFlow(modelIn, modelOut)
+      func.(TaintFunction).hasTaintFlow(modelIn, modelOut) and
+      model = "TaintFunction"
     ) and
     nodeOut = callOutput(call, modelOut)
   )
