59ffacf61b
localCompare: mention fields as they are skipped
2025-10-13 18:33:10 +03:00
b38185c979
Model: rewrite TTG documentation
...
Big chunks of this file were outdated or otherwise awkward,
this brings it up-to-date with the current state of the TTG.
2025-10-13 18:33:10 +03:00
10ca432cb2
TTG: simplify reference field usage
...
This allows to avoid having to specify the `rootType` (stable within
the entire tree) manually each time a reference is introduced.
Thanks to that, a much nicer syntax of:
```yaml
- name: MyTypeField
reference-to: TypeDefinition
```
can replace the current one:
```yaml
- name: MyTypeField
reference:
pointeeType: TypeDefinition
rootType: Binary
```
2025-10-13 18:33:10 +03:00
480443f358
TTG: drop upcastable: configuration option
...
Instead, detect whether `UpcastablePointer` wrapping is needed or not
based on the type.
2025-10-13 18:33:10 +03:00
e9bb8c43cf
TTG: make sure emission order is stable
...
Debugging is a lot harder when you cannot just diff outputs of
the generator.
This commit ensures types are always emitted in the same order,
no matter the input schema ordering or the target language.
2025-10-13 18:33:10 +03:00
d1fdb4d649
TTG: drop an unused global
2025-10-13 18:33:10 +03:00
2315248938
TTG: remove a broken KOT-compatibility check
2025-10-13 18:33:10 +03:00
38921fb85e
TTG: drop an unnecessary TODO
2025-10-13 18:33:10 +03:00
d9ee804211
Do not use std::vector in model TTG docs
...
Currently, using `std::vector` in these breaks invalidation, so to
avoid potential confusion, we shouldn't use it in docs either.
Note that it's fine to use `std::vector` in non-invalidation-enabled
TTG instances.
2025-10-13 18:33:10 +03:00
0dba3e0c69
Yield: improve TTG-schema formatting
2025-10-13 18:33:10 +03:00
de6db68394
EFA: improve TTG-schema formatting
2025-10-13 18:33:10 +03:00
95845277a9
TTG: verify given schema
2025-10-13 18:33:10 +03:00
00ea93c99e
TTG: move metaschema.yml to share/revng/
2025-10-13 18:33:10 +03:00
04fdc2028a
migrations: do not duplicate the template
2025-10-13 18:33:10 +03:00
89282f31da
Minor improvements
2025-10-13 18:33:10 +03:00
6b976f4fda
Merge branch 'feature/clift-directory-structure'
2025-10-13 15:31:59 +02:00
22d6f70ee2
Clift directory structure overhaul
2025-10-13 15:08:20 +02:00
3c191122dc
Rename Clift to CliftDialect, CliftOps to Clift
2025-10-13 15:08:20 +02:00
c34aed5565
Merge branch 'feature/split-llvm-container'
2025-10-13 10:44:35 +02:00
256d2310b0
DwarfImporter: drop an unnecessary check
2025-10-13 10:26:41 +03:00
df681aa573
Ensure consumed errors are printed somewhere
2025-10-13 10:26:41 +03:00
25538385c8
Avoid using !! unless necessary
...
The only necessary place is when passing it into the `BOOST_TEST`
macro, which doesn't convert its argument to `bool` internally.
2025-10-13 10:26:41 +03:00
a8e326f4d7
UpcastablePointer: improve documentation
2025-10-13 10:26:41 +03:00
e56f543b1a
Split module.bc.zstd
2025-10-13 10:26:41 +03:00
fa24a51b05
Don't ignore TargetRegistry::lookupTarget output
2025-10-13 10:26:41 +03:00
9bf93d3a30
Importers: add TODOs on diagnostic candidates
2025-10-13 10:26:41 +03:00
16985c37ea
Add comments clarifying error-related behavior
2025-10-13 10:26:41 +03:00
ed53e759c9
Do not use -Y flag
...
Yaml model are now the _only_ way to store them, no point in still
specifying the old flag for it.
2025-10-13 10:26:41 +03:00
5717d835f8
Add missing llvm::Error checks
2025-10-13 10:26:37 +03:00
c11761ceb5
Pipes: drop unused invalidateFromDiff
2025-10-08 12:22:34 +03:00
68d62428d8
Lift: drop free loadBinary function
2025-10-08 12:22:34 +03:00
d3bae70ba4
TupleTreeDiff: drop an unused overload
2025-10-08 12:22:34 +03:00
0d9ea6db18
Inspector: drop a no-op llvm::consumeError
2025-10-08 12:22:34 +03:00
dadd836035
MetaAddress: assert validity
2025-10-08 12:22:34 +03:00
e6d30936f3
Avoid llvm::ExitOnError in non-tool binaries
2025-10-08 12:22:34 +03:00
745b461568
Tools: use llvm::ExitOnError
2025-10-08 12:22:34 +03:00
a88747cf99
Drop unused includes
2025-10-08 12:22:34 +03:00
d1fbf2eee6
Model: Drop ToolHelpers.h
2025-10-08 12:22:34 +03:00
b6cb5c2ca3
UpcastablePointer: drop llvm::Error overload
2025-10-08 12:22:34 +03:00
d51ade0243
check-commit: add work-in-progress checks
2025-10-08 12:22:34 +03:00
ca712608f1
check-commit: more explicit error message
2025-10-08 12:22:34 +03:00
77ea07c364
Minor improvements
2025-10-08 12:22:30 +03:00
6d1fbdca97
Implement infrastructure for PipeRuns
...
Add the needed infrastructures that allow ephemeral classes (`PipeRun`s)
to be wrapped as a regular pypeline `Pipe` class. These automatically
have their dependencies tracked and propagated to the pypeline.
2025-10-03 17:40:45 +02:00
7783c0484a
AVI: sort entry points by dominance
...
This commit a non-deterministic bug on ARM's switch-disjoint-ranges
test.
The problem was that we were setting as initial nodes of the monotone
framework all the nodes that are not reachable from the nodes classified
as initial nodes *so far*.
This approach is susceptible to the visit order.
Classifying certain nodes as entry nodes degrades the analysis results
since we associates a top value to them.
Electing initial nodes by traversing a list of nodes sorted by dominance
solves this problem.
2025-10-02 14:32:17 +02:00
5d6bddf845
Merge branch 'feature/model-agnostic-clift'
2025-09-30 16:25:17 +02:00
676fdfe0f5
Split common parts of Clift backend into new class
...
CEmitter contains utilities for emitting types, attributes and
declarations. These will be useful for generating headers from Clift.
2025-09-29 18:05:55 +03:00
7fab96f119
Add dumpMlirModule for debugging
2025-09-29 18:05:55 +03:00
110fb23d58
Improve Clift backend string literal test
2025-09-29 18:05:55 +03:00
18773ba22c
Add Clift declaration attribute support
2025-09-29 18:05:55 +03:00
cd48916c8d
Convert Clift backend tests to use clift-opt
2025-09-29 18:05:55 +03:00
Ivan Krysak