Merge pull request #6825 from MathiasVP/use-shared-ssa-in-ir-dataflow

C++: Redesign IR dataflow using the shared SSA library

Author: geoffw0

config/identical-files.json

@@ -449,7 +449,8 @@
     "csharp/ql/lib/semmle/code/csharp/controlflow/internal/pressa/SsaImplCommon.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/basessa/SsaImplCommon.qll",
     "csharp/ql/lib/semmle/code/cil/internal/SsaImplCommon.qll",
-    "ruby/ql/lib/codeql/ruby/dataflow/internal/SsaImplCommon.qll"
+    "ruby/ql/lib/codeql/ruby/dataflow/internal/SsaImplCommon.qll",
+    "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/SsaImplCommon.qll"
   ],
   "CryptoAlgorithms Python/JS": [
     "javascript/ql/lib/semmle/javascript/security/CryptoAlgorithms.qll",

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowDispatch.qll

@@ -63,8 +63,10 @@ private module VirtualDispatch {
       |
         // Call argument
         exists(DataFlowCall call, int i |
-          other.(DataFlow::ParameterNode).isParameterOf(call.getStaticCallTarget(), i) and
-          src.(ArgumentNode).argumentOf(call, i)
+          other
+              .(DataFlow::ParameterNode)
+              .isParameterOf(pragma[only_bind_into](call).getStaticCallTarget(), i) and
+          src.(ArgumentNode).argumentOf(call, pragma[only_bind_into](pragma[only_bind_out](i)))
         ) and
         allowOtherFromArg = true and
         allowFromArg = true
@@ -128,6 +130,7 @@ private module VirtualDispatch {
    *
    * Used to fix a join ordering issue in flowsFrom.
    */
+  pragma[noinline]
   private predicate returnNodeWithKindAndEnclosingCallable(
     ReturnNode node, ReturnKind kind, DataFlowCallable callable
   ) {

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowPrivate.qll

@@ -112,11 +112,9 @@ class ReturnNode extends InstructionNode {
   Instruction primary;
 
   ReturnNode() {
-    exists(ReturnValueInstruction ret | instr = ret.getReturnValue() and primary = ret)
+    exists(ReturnValueInstruction ret | instr = ret and primary = ret)
     or
-    exists(ReturnIndirectionInstruction rii |
-      instr = rii.getSideEffectOperand().getAnyDef() and primary = rii
-    )
+    exists(ReturnIndirectionInstruction rii | instr = rii and primary = rii)
   }
 
   /** Gets the kind of this returned value. */
@@ -190,108 +188,16 @@ OutNode getAnOutNode(DataFlowCall call, ReturnKind kind) {
  */
 predicate jumpStep(Node n1, Node n2) { none() }
 
-private predicate fieldStoreStepNoChi(Node node1, FieldContent f, PostUpdateNode node2) {
-  exists(StoreInstruction store, Class c |
-    store = node2.asInstruction() and
-    store.getSourceValueOperand() = node1.asOperand() and
-    getWrittenField(store, f.getAField(), c) and
-    f.hasOffset(c, _, _)
-  )
-}
-
-private FieldAddressInstruction getFieldInstruction(Instruction instr) {
-  result = instr or
-  result = instr.(CopyValueInstruction).getUnary()
-}
-
-pragma[noinline]
-private predicate getWrittenField(Instruction instr, Field f, Class c) {
-  exists(FieldAddressInstruction fa |
-    fa =
-      getFieldInstruction([
-          instr.(StoreInstruction).getDestinationAddress(),
-          instr.(WriteSideEffectInstruction).getDestinationAddress()
-        ]) and
-    f = fa.getField() and
-    c = f.getDeclaringType()
-  )
-}
-
-private predicate fieldStoreStepChi(Node node1, FieldContent f, PostUpdateNode node2) {
-  exists(ChiPartialOperand operand, ChiInstruction chi |
-    chi.getPartialOperand() = operand and
-    node1.asOperand() = operand and
-    node2.asInstruction() = chi and
-    exists(Class c |
-      c = chi.getResultType() and
-      exists(int startBit, int endBit |
-        chi.getUpdatedInterval(startBit, endBit) and
-        f.hasOffset(c, startBit, endBit)
-      )
-      or
-      getWrittenField(operand.getDef(), f.getAField(), c) and
-      f.hasOffset(c, _, _)
-    )
-  )
-}
-
-private predicate arrayStoreStepChi(Node node1, ArrayContent a, PostUpdateNode node2) {
-  exists(a) and
-  exists(ChiPartialOperand operand, ChiInstruction chi, StoreInstruction store |
-    chi.getPartialOperand() = operand and
-    store = operand.getDef() and
-    node1.asOperand() = operand and
-    // This `ChiInstruction` will always have a non-conflated result because both `ArrayStoreNode`
-    // and `PointerStoreNode` require it in their characteristic predicates.
-    node2.asInstruction() = chi and
-    (
-      // `x[i] = taint()`
-      // This matches the characteristic predicate in `ArrayStoreNode`.
-      store.getDestinationAddress() instanceof PointerAddInstruction
-      or
-      // `*p = taint()`
-      // This matches the characteristic predicate in `PointerStoreNode`.
-      store.getDestinationAddress().(CopyValueInstruction).getUnary() instanceof LoadInstruction
-    )
-  )
-}
-
 /**
  * Holds if data can flow from `node1` to `node2` via an assignment to `f`.
  * Thus, `node2` references an object with a field `f` that contains the
  * value of `node1`.
  */
-predicate storeStep(Node node1, Content f, PostUpdateNode node2) {
-  fieldStoreStepNoChi(node1, f, node2) or
-  fieldStoreStepChi(node1, f, node2) or
-  arrayStoreStepChi(node1, f, node2) or
-  fieldStoreStepAfterArraySuppression(node1, f, node2)
-}
-
-// This predicate pushes the correct `FieldContent` onto the access path when the
-// `suppressArrayRead` predicate has popped off an `ArrayContent`.
-private predicate fieldStoreStepAfterArraySuppression(
-  Node node1, FieldContent f, PostUpdateNode node2
-) {
-  exists(WriteSideEffectInstruction write, ChiInstruction chi, Class c |
-    not chi.isResultConflated() and
-    node1.asInstruction() = chi and
-    node2.asInstruction() = chi and
-    chi.getPartial() = write and
-    getWrittenField(write, f.getAField(), c) and
-    f.hasOffset(c, _, _)
-  )
-}
-
-bindingset[result, i]
-private int unbindInt(int i) { i <= result and i >= result }
-
-pragma[noinline]
-private predicate getLoadedField(LoadInstruction load, Field f, Class c) {
-  exists(FieldAddressInstruction fa |
-    fa = load.getSourceAddress() and
-    f = fa.getField() and
-    c = f.getDeclaringType()
+predicate storeStep(StoreNodeInstr node1, FieldContent f, StoreNodeInstr node2) {
+  exists(FieldAddressInstruction fai |
+    node1.getInstruction() = fai and
+    node2.getInstruction() = fai.getObjectAddress() and
+    f.getField() = fai.getField()
   )
 }
 
@@ -300,122 +206,14 @@ private predicate getLoadedField(LoadInstruction load, Field f, Class c) {
  * Thus, `node1` references an object with a field `f` whose value ends up in
  * `node2`.
  */
-private predicate fieldReadStep(Node node1, FieldContent f, Node node2) {
-  exists(LoadOperand operand |
-    node2.asOperand() = operand and
-    node1.asInstruction() = operand.getAnyDef() and
-    exists(Class c |
-      c = operand.getAnyDef().getResultType() and
-      exists(int startBit, int endBit |
-        operand.getUsedInterval(unbindInt(startBit), unbindInt(endBit)) and
-        f.hasOffset(c, startBit, endBit)
-      )
-      or
-      getLoadedField(operand.getUse(), f.getAField(), c) and
-      f.hasOffset(c, _, _)
-    )
+predicate readStep(ReadNode node1, FieldContent f, ReadNode node2) {
+  exists(FieldAddressInstruction fai |
+    node1.getInstruction() = fai.getObjectAddress() and
+    node2.getInstruction() = fai and
+    f.getField() = fai.getField()
   )
 }
 
-/**
- * When a store step happens in a function that looks like an array write such as:
- * ```cpp
- * void f(int* pa) {
- *   pa = source();
- * }
- * ```
- * it can be a write to an array, but it can also happen that `f` is called as `f(&a.x)`. If that is
- * the case, the `ArrayContent` that was written by the call to `f` should be popped off the access
- * path, and a `FieldContent` containing `x` should be pushed instead.
- * So this case pops `ArrayContent` off the access path, and the `fieldStoreStepAfterArraySuppression`
- * predicate in `storeStep` ensures that we push the right `FieldContent` onto the access path.
- */
-predicate suppressArrayRead(Node node1, ArrayContent a, Node node2) {
-  exists(a) and
-  exists(WriteSideEffectInstruction write, ChiInstruction chi |
-    node1.asInstruction() = write and
-    node2.asInstruction() = chi and
-    chi.getPartial() = write and
-    getWrittenField(write, _, _)
-  )
-}
-
-private class ArrayToPointerConvertInstruction extends ConvertInstruction {
-  ArrayToPointerConvertInstruction() {
-    this.getUnary().getResultType() instanceof ArrayType and
-    this.getResultType() instanceof PointerType
-  }
-}
-
-private Instruction skipOneCopyValueInstructionRec(CopyValueInstruction copy) {
-  copy.getUnary() = result and not result instanceof CopyValueInstruction
-  or
-  result = skipOneCopyValueInstructionRec(copy.getUnary())
-}
-
-private Instruction skipCopyValueInstructions(Operand op) {
-  not result instanceof CopyValueInstruction and result = op.getDef()
-  or
-  result = skipOneCopyValueInstructionRec(op.getDef())
-}
-
-private predicate arrayReadStep(Node node1, ArrayContent a, Node node2) {
-  exists(a) and
-  // Explicit dereferences such as `*p` or `p[i]` where `p` is a pointer or array.
-  exists(LoadOperand operand, Instruction address |
-    operand.isDefinitionInexact() and
-    node1.asInstruction() = operand.getAnyDef() and
-    operand = node2.asOperand() and
-    address = skipCopyValueInstructions(operand.getAddressOperand()) and
-    (
-      address instanceof LoadInstruction or
-      address instanceof ArrayToPointerConvertInstruction or
-      address instanceof PointerOffsetInstruction
-    )
-  )
-}
-
-/**
- * In cases such as:
- * ```cpp
- * void f(int* pa) {
- *   *pa = source();
- * }
- * ...
- * int x;
- * f(&x);
- * use(x);
- * ```
- * the load on `x` in `use(x)` will exactly overlap with its definition (in this case the definition
- * is a `WriteSideEffect`). This predicate pops the `ArrayContent` (pushed by the store in `f`)
- * from the access path.
- */
-private predicate exactReadStep(Node node1, ArrayContent a, Node node2) {
-  exists(a) and
-  exists(WriteSideEffectInstruction write, ChiInstruction chi |
-    not chi.isResultConflated() and
-    chi.getPartial() = write and
-    node1.asInstruction() = write and
-    node2.asInstruction() = chi and
-    // To distinquish this case from the `arrayReadStep` case we require that the entire variable was
-    // overwritten by the `WriteSideEffectInstruction` (i.e., there is a load that reads the
-    // entire variable).
-    exists(LoadInstruction load | load.getSourceValue() = chi)
-  )
-}
-
-/**
- * Holds if data can flow from `node1` to `node2` via a read of `f`.
- * Thus, `node1` references an object with a field `f` whose value ends up in
- * `node2`.
- */
-predicate readStep(Node node1, Content f, Node node2) {
-  fieldReadStep(node1, f, node2) or
-  arrayReadStep(node1, f, node2) or
-  exactReadStep(node1, f, node2) or
-  suppressArrayRead(node1, f, node2)
-}
-
 /**
  * Holds if values stored inside content `c` are cleared at node `n`.
  */
@@ -447,7 +245,7 @@ private predicate suppressUnusedNode(Node n) { any() }
 // Java QL library compatibility wrappers
 //////////////////////////////////////////////////////////////////////////////
 /** A node that performs a type cast. */
-class CastNode extends InstructionNode {
+class CastNode extends Node {
   CastNode() { none() } // stub implementation
 }