I took most of the docs from the corresponding predicates in
JavaScript's `CodeDuplication.qll`. Where JavaScript had a corresponding
predicate but didn't have QLDoc, I added new QLDoc to both.
The `cpp/too-few-arguments` query produced alerts for ambiguous
databases where a function had multiple possible declarations, with some
declarations having the right number of parameters and some having too
many. With this change, the query errs on the side of caution in those
cases and does not produce an alert.
This fixes false positives on racket/racket.
The new `hasDefiniteNumberOfParameters` is exactly the negation of the
old `hasZeroParamDecl`.
In a couple of cases, we use `glval<unknown>` as the result type of an instruction because we can't come up with anything better. Two examples are the result of `VariableAddress[#ellipsis]`, and the address of the temp variable that holds the lvalue result of the conditional operator in `(a ? b : c) = y`. In both cases, we call `getTypeForGLValue(any(UnknownType t))`, but that would have multiple results because `result.hasType(any(UnknownType t), true)` also holds for `CppFunctionGLValueType`. I tightened the result type to ensure we get the right one.
The QL compiler is about to be changed to emit a warning when overriding a deprecated predicate. This PR marks the existing overrides of deprecated predicates as `deprecated` themselves, which avoids the warning.
The `Print.qll` models seem to preserve the `isWideCharDefault()` predicate for backwards compatibility, so we can't remove them and must continue overriding them.
The `XML.qll` override is necessary because both superclasses declare the `getName()` predicate. One is `deprecated`, and the other is `abstract`, so we have to have an override.
The virtual-dispatch code for globals was missing any relationship
between the union field access and the global variable, which meant it
propagated function-pointer flow between any two fields of a global
struct. This resulted in false positives from
`cpp/tainted-format-string` on projects using SDL, such as
WohlSoft/PGE-Project.
In addition to fixing that bug, this commit also brings the code up to
date with the new style of modeling flow through global variables:
`DataFlow::Node.asVariable()`.
This test demonstrates that IR data flow conflates unrelated fields of a
global struct-typed variable and that this bug is not present in the old
AST-based implementation of `semmle.code.cpp.security.TaintTracking`.
This cleans up the test results, which were confusing because functions
like `sink` had multiple locations.
There are some additional results now involving casts to `const char *`
because previously it varied whether `sink` used `const`, and now it
always does.
The previous changes made the optimizer choose a bad join order for the RHS of the antijoin in `addressOperandAllocationAndOffset`. Once again, `unique` to the rescue.
