René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Generate Lib/opcode.py from Python/bytecodes.c","articleBody":"This could also auto-generate Include/opcode.h, Include/internal/pycore_opcode.h and Python/opcode_targets.h, subsuming both Tools/build/generate_opcode_h.py and Python/makeopcodetargets.py -- although the simplest approach would probably be to just keep those tools and only ensure that they are re-run after opcode.py is updated.\r\n\r\n## Variables to generate\r\n\r\nThe auto-generatable contents of opcode.py is as follows:\r\n- `opmap` -- make up opcode numbers in a pass after reading all input; special-case `CACHE = 0`\r\n- `HAVE_ARGUMENT` -- done while making up opcode numbers (group argument-less ones in lower group)\r\n- `ENABLE_SPECIALIZATION = True`\r\n- `EXTENDED_ARG` -- set from opcode for `EXTENDED_ARG` inst, if there is one\r\n- `opname` -- invert `opmap`\r\n- pseudo opcode definitions, can add DSL syntax `pseudo(NAME) = { name1, name2, ... };`\r\n- `hasarg` -- can be derived from `instr_format` metadata\r\n- `hasconst` -- could hardcode to `LOAD_CONST`\r\n- `hasname` -- may have to check for occurrences of `co_names` in code?\r\n- `hasjrel` -- check for `JUMPBY` with arg that doesn't start with `INLINE_CACHE_ENTRIES_`\r\n- `hasjabs`-- no longer used, set to `[]` for backwards compatibility\r\n- `haslocal` -- opcode name contains '_FAST'\r\n- `hascompare` -- opcode name starts with `COMPARE_`\r\n- `hasfree` -- opcode name ends in `DEREF` or `CELL` or `CLOSURE`\r\n- `hasexc` -- pseudo opcode, name starts with `SETUP_`\r\n- `MIN_PSEUDO_OPCODE = 256`\r\n- `MAX_PSEUDO_OPCODE` -- derive from pseudo opcodes\r\n- `__all__` -- hardcode\r\n\r\nThe following are not public but imported by `dis.py` so they are still needed:\r\n- `_nb_ops` -- just hardcode?\r\n- `_specializations` -- derive from families (with some adjustments)\r\n- `_specialized_instructions` -- compute from `_specializations`\r\n- `_specialization_stats` -- only used by `test__opcode.py`, move into there?\r\n- `_cache_format` -- compute from cache effects? (how to make up names?)\r\n- `_inline_cache_entries` -- compute from `_cache_format`\r\n\r\nThe hardcoded stuff can go in prologue and epilogue sections that are updated manually.\r\n\r\nThis project (if we decide to do it) might be a good reason to refactor `generate_cases.py` into a library (the code for this shouldn't be shoved into the main file).\r\n\r\n## Benefits\r\n\r\nWe can wait on this project until we are sure we need at least one of the following benefits:\r\n\r\n- Once we are generating the numeric opcode values it will be easier to also generate numeric values for micro-opcodes.\r\n- Avoid having to keep multiple definitions in sync (e.g. families, cache formats).\r\n- Easier to understand. E.g. where are the numeric opcode values for specialized instructions defined? (Would require also subsuming generate_opcode_h.py.)\r\n\n\n\u003c!-- gh-linked-prs --\u003e\n### Linked PRs\n* gh-102685\n\u003c!-- /gh-linked-prs --\u003e\n","author":{"url":"https://github.com/gvanrossum","@type":"Person","name":"gvanrossum"},"datePublished":"2023-03-14T00:04:26.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":3},"url":"https://github.com/102674/cpython/issues/102674"}