When this query was run as an upgrade script, the optimizer picked a bad
join order, making the upgrade very slow on large databases. It picked a
bad join order because upgrade scripts are run with no stats.
Lack of support for the GCC vector extensions was causing a bunch of sanity failures in the syntax zoo. This PR adds minimal IR generation support for these types.
Added `VectorAggregateLiteral`, and factored most of `ArrayAggregateLiteral` out into the common base class `ArrayOrVectorAggregateLiteral`. I'd be happy to merge these all into `ArrayAggregateLiteral` if we don't care about the distinction.
Made a few tweaks to `TranslatedArrayExpr` to compute the element type by looking at the result type of the `ArrayExpr`, not the type of the base operand. Note that this means that for `T a[10]; a[i] = foo;`, the result of the `PointerAdd` for `a[i]` will now be `glvalue<T>`, not `T*`. This is actually more faithful to the source language, and has no semantic difference on the IR.
Added some missing `getInstructionElementSize()` overrides.
Added the new `BuiltIn` opcode, renamed the existing `BuiltInInstruction` to `BuiltInOperationInstruction`, and made any `BuiltInOperation` that we don't specifically handle translate to `BuiltIn`. `BuiltInOperationInstruction` now has a way to get the specific `BuiltInOperation`.
Added `getCanonicalQLClass()` overrides for `GNUVectorType` and `BuiltInOperation`.
Added a simple IR test for vector types.
This file is now identical in all languages. Unifying this file led to
the following changes:
- The documentation spelling fixes and example from the C++ version
were copied to the other versions and updated.
- The steps through `NonLocalJumpNode` from C# were abstracted into a
`globalAdditionalTaintStep` predicate that's empty for C++ and Java.
- The `defaultTaintBarrier` predicate from Java is now present but empty
on C++ and C#.
- The C++ `isAdditionalFlowStep` predicate on
`TaintTracking::Configuration` no longer includes `localFlowStep`.
That should avoid some unnecessary tuple copying.
This class was copy-pasted in all `DataFlowN.qll` files without using
the identical-files system to keep the copies in sync. The class is now
moved to the `DataFlowImplN.qll` files.
This also has the effect of preventing recursion through first data flow
library copy for C/C++. Such recursion has been deprecated for over a
year, and some forms of recursions are already ruled out by the library
implementation.
To compensate for the lack of field flow, the taint tracking library has
previously considered taint to flow from fields to their containing
structs and back again from the structs to any of their fields. This
leads to false flow between unrelated fields and is not needed now that
we have proper flow through fields.
There's now a `localFlowStep` predicate for use directly in queries and
other libraries and a `simpleLocalFlowStep` for use only by the global
data flow library. The former predicate is intended to include field
flow, but the latter may not.
This will let Java and C# (and possibly C++ IR) avoid getting two kinds
of field flow at the same time, both from SSA and from the global data
flow library. It should let C++ AST add some form of field flow to
`localFlowStep` without making it an input to the global data flow
library.
