René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Remove `eval_breaker` and `kwnames` local variables from interpreter definition.","articleBody":"Currently there are quite a few local variables in `PyEval_EvalDefault()`.\r\nEach of these is:\r\n* A candidate for register allocation, confusing the compiler.\r\n* Potentially adds extra state to the interpreter\r\n* Adds complexity to automatically generated interpreters and compilers\r\n\r\nThe local variables break down into three classes:\r\n* The state of the interpreter\r\n* Transient values\r\n* Values for gathering stats, debugging, etc.\r\n\r\nThe last category are not present in release builds, so we can ignore those.\r\nThe transient values, like `opcode` and `oparg` are an artifact of the dispatching mechanism and will be absent for compiled code, or different for a different interpreter.\r\n\r\nIt is the remaining values that matter.\r\nEach of these will consume a register in compiled code, and need to kept in sync where there is redundancy.\r\nThe current values are:\r\n* `next_instr` -- The VM instruction pointer\r\n* `frame` -- Pointer to the current frame\r\n* `stackpointer` -- Pointer to the top of stack\r\n* `tstate` -- Pointer to the current thread\r\n* `eval_break` -- Pointer to the eval-breaker\r\n* `kwnames` -- Pointer to the keyword names, if any for the next call\r\n\r\n`next_instr` and `frame` are fundamental to VM operation.\r\n\r\nHypothetically `stackpointer` can be eliminated by a compiler, and `tstate` can be eliminated by some clever stack arrangement, or using thread-local storage. However, these approaches are complex and may not be worthwhile.\r\n\r\n`eval_breaker` can easily be eliminated, as it is redundant. `eval_breaker == tstate-\u003einterp-\u003eeval_breaker`\r\n\r\n`kwnames` is logically transient, in that it is only valid between the `KW_NAMES` instruction and a `CALL`. However from the very local view of a single instruction it is effectively an interpreter-wide value.\r\nIt can removed by pushing the kwnames to the stack, but that is likely to cause a slowdown. We need to experiment.\r\n\r\n \r\n\r\n\r\n\n\n\u003c!-- gh-linked-prs --\u003e\n### Linked PRs\n* gh-104581\n* gh-107383\n\u003c!-- /gh-linked-prs --\u003e\n","author":{"url":"https://github.com/markshannon","@type":"Person","name":"markshannon"},"datePublished":"2023-05-17T13:35:42.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":1},"url":"https://github.com/104580/cpython/issues/104580"}