Adds flag for enabling monolith opcodes and re-implements OP_AND, OP_OR, and OP_XOR.
Co-authored-by: Joshua Yabut <yabut.joshua@gmail.com>
Co-authored-by: Marcos Mayorga <mm@mm-studios.com>
Co-authored-by: Daniel Connolly <daniel@dconnolly.com>
Details
Details
Diff Detail
Diff Detail
- Repository
- rABC Bitcoin ABC
- Branch
- arcpatch-D1090_2
- Lint
Lint Skipped - Unit
No Test Coverage - Build Status
Buildable 2084 Build 2315: Bitcoin ABC Buildbot (legacy) Build 2314: arc lint + arc unit
Event Timeline
There are a very large number of changes, so older changes are hidden. Show Older Changes
| test/functional/test_framework/script.py | ||
|---|---|---|
| 344 ↗ | (On Diff #3150) | this should not be here, will fix. still getting used to phabricator |
| src/Makefile.test.include | ||
|---|---|---|
| 64 ↗ | (On Diff #3151) | op_codes. |
| src/script/interpreter.cpp | ||
| 351 ↗ | (On Diff #3151) | You should have prepared this flag with negative testing. That makes it very clear what new behavior is added to the flag with each diff. It would also ensure that all your diffs do not need to depend on each other, which would ensure you move much faster. |
| 808 ↗ | (On Diff #3151) | This is terrible. i original asked for C++11 style loops (so not this) but it turns out that the index is used in multiple container, so that's not even a good idea. |
| src/script/script_error.cpp | ||
| 50 ↗ | (On Diff #3151) | If this is an error about input length, then an error code specifying so would be more appropriate. |
| src/test/op_code.cpp | ||
| 11 ↗ | (On Diff #3151) | nope |
| 19 ↗ | (On Diff #3151) | Other tests do have an infrastructure to generate json tests and then run them. If you are going to go that road, it might as well be useful and generate reusable test cases. |
| 48 ↗ | (On Diff #3151) | Use explicit names. |
| 70 ↗ | (On Diff #3151) | dito. |
| 158 ↗ | (On Diff #3151) | There is no check for empty elements. More generally, build a set of element and expected result when you run them. Run through various set of flags (mandatory, standard and nothing seems like good ideas) and check they fail. Do the same with the monolith flag in addition, check they give the expected result. Both way as binary ops are commutative. Make sure the large vectors are filled with various data combination at the start and end so you exercise cases where possible signed stack items would show up. |
| 265 ↗ | (On Diff #3151) | I'm not sure where this test is going. If we want to have a set of test cases, script_tests.json seems like the appropriate place to put them. Obviously, having a custom test would allow to grill thing more exhaustively, so that's a good thing, but this test doesn't do this. We should probably have both. |