The --import-debug-info CLI flag and the ImporterOptions::
AdditionalDebugInfoPaths field it populated were a side-channel that
loaded extra DWARF files outside the normal ELF dependency-resolution
path. It bypassed the symbol-aware machinery and is no longer needed.
Unfortunately, in DWARF, there's no widespread indicator on whether a
function is regular ARM or Thumb. Therefore, in this commit, instead of
blindly creating a new function, we first check if, in the model, we
already have a function at that address.
This means that, if we have a Thumb function in DWARF, but it doesn't
already exist in the model, we will import as regular ARM, which is
wrong.
Specifically, this can happen with `revng model import debug-info`.
We used to use dynamic_symbol_begin which does not use segments (uses
sections) and also tampers with `st_value` for ARM symbols.
This commit makes it reuse the more accurate logic implemented in
ELFImporter.
DW_AT_abstract_origin chains are walked recursively to find a DIE's
name, with no cycle check. Some binaries (e.g. debian-bookworm-mipsel
/usr/sbin/pwck) ship DWARF with a cycle in that chain, so getName ended
up in infinite recursion.
Make the walk iterative.
The macro is no longer necessary: the code now ensures we never
need to assign anything to an array, which would be ill-advised
nevertheless given that array-typed rvalues in C immediately
decay to pointers.
The macro is no longer necessary: the code now ensures we never
need to bitcast anything to an array, which would be ill-advised
nevertheless given that array-typed rvalues in C immediately
decay to pointers.
assign_array is necessary as a fallback for when the decompiled C
code may need assignments among array types. bit_cast_to_array is
necessary because when the target type of a bitcast is an array
the regular bit_cast macro is not enough.
Rework primitive-types.h so that the various bit_cast macros share
a single implementation based on `typeof`/`__typeof__`. As a
consequence both keywords are added to the model's reserved-name
set so that user identifiers cannot collide with them.
Teach NameBuilder and ModelCBuilder how to produce names and
references for opaque-array types: NameBuilder gains
opaqueTypeName(ByteSize), ModelCBuilder gains the matching
reference and definition tags. Both rely on the new
Configuration::OpaqueTypePrefix.
This commit introduces an analysis to import prototypes from a SQLite
database of well-known prototypes, typically built from debug info of
operating systems.
This analysis supersedes import-well-known-models.
Refactor the binary import and dependency resolution infrastructure to
support multiple platforms (Linux, Windows, macOS) via a
configuration-driven root system.
Major changes:
* Overhaul PDB and DWARF importers for platform-aware debug info
loading.
* Refactor LDDTree into a template-based architecture with
platform-specific implementations (ELF, PE/COFF).
* Mostly rewrite the PDB importer, which had significant limitations.
Rework the python mixin infrastructure for the model. Mixins are now
specified explicitly instead of using the `ast` module. The mixins are
loaded relative to the generated python file, this allows to have two
models: `revng.model` and `revng.project.model`. The latter is augmented
via mixins to have project-specific facilities such as `get_artifact`
and artifact accessors.
Overhaul the logic and method names involved in enabling and disabling
reference caching in `TupleTree<T>`. `TupleTreeReference<T, U>` now
lazily caches the target and will traverse the path only when needed.
Also expose and use these functions in the new pipeline, which should
provide some speedup when executing a `Schedule`.
Alessandro Di Federico