diff options
| author | mpage <mpage@meta.com> | 2024-10-09 08:18:25 -0700 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-10-09 15:18:25 +0000 |
| commit | f978fb4f8d6eac0585057e463bb1701dc04a9900 (patch) | |
| tree | 72dc60eb608119b7c6f9315788132b7bb1b8ff92 /Python/optimizer_cases.c.h | |
| parent | b9a8ca0a6aa9251cb798f34f0c9d2cc95107eec6 (diff) | |
| download | cpython-f978fb4f8d6eac0585057e463bb1701dc04a9900.tar.gz cpython-f978fb4f8d6eac0585057e463bb1701dc04a9900.zip | |
gh-115999: Refactor `LOAD_GLOBAL` specializations to avoid reloading {globals, builtins} keys (gh-124953)
Each of the `LOAD_GLOBAL` specializations is implemented roughly as:
1. Load keys version.
2. Load cached keys version.
3. Deopt if (1) and (2) don't match.
4. Load keys.
5. Load cached index into keys.
6. Load object from (4) at offset from (5).
This is not thread-safe in free-threaded builds; the keys object may be replaced
in between steps (3) and (4).
This change refactors the specializations to avoid reloading the keys object and
instead pass the keys object from guards to be consumed by downstream uops.
Diffstat (limited to 'Python/optimizer_cases.c.h')
| -rw-r--r-- | Python/optimizer_cases.c.h | 60 |
1 files changed, 51 insertions, 9 deletions
diff --git a/Python/optimizer_cases.c.h b/Python/optimizer_cases.c.h index ae532fd2776..6ec9e69d1db 100644 --- a/Python/optimizer_cases.c.h +++ b/Python/optimizer_cases.c.h @@ -844,30 +844,48 @@ break; } - case _GUARD_BUILTINS_VERSION: { + case _GUARD_GLOBALS_VERSION_PUSH_KEYS: { + _Py_UopsSymbol *globals_keys; + uint16_t version = (uint16_t)this_instr->operand; + globals_keys = sym_new_unknown(ctx); + (void)version; + stack_pointer[0] = globals_keys; + stack_pointer += 1; + assert(WITHIN_STACK_BOUNDS()); break; } - case _LOAD_GLOBAL_MODULE: { + case _GUARD_BUILTINS_VERSION_PUSH_KEYS: { + _Py_UopsSymbol *builtins_keys; + uint16_t version = (uint16_t)this_instr->operand; + builtins_keys = sym_new_unknown(ctx); + (void)version; + stack_pointer[0] = builtins_keys; + stack_pointer += 1; + assert(WITHIN_STACK_BOUNDS()); + break; + } + + case _LOAD_GLOBAL_MODULE_FROM_KEYS: { _Py_UopsSymbol *res; _Py_UopsSymbol *null = NULL; res = sym_new_not_null(ctx); null = sym_new_null(ctx); - stack_pointer[0] = res; - if (oparg & 1) stack_pointer[1] = null; - stack_pointer += 1 + (oparg & 1); + stack_pointer[-1] = res; + if (oparg & 1) stack_pointer[0] = null; + stack_pointer += (oparg & 1); assert(WITHIN_STACK_BOUNDS()); break; } - case _LOAD_GLOBAL_BUILTINS: { + case _LOAD_GLOBAL_BUILTINS_FROM_KEYS: { _Py_UopsSymbol *res; _Py_UopsSymbol *null = NULL; res = sym_new_not_null(ctx); null = sym_new_null(ctx); - stack_pointer[0] = res; - if (oparg & 1) stack_pointer[1] = null; - stack_pointer += 1 + (oparg & 1); + stack_pointer[-1] = res; + if (oparg & 1) stack_pointer[0] = null; + stack_pointer += (oparg & 1); assert(WITHIN_STACK_BOUNDS()); break; } @@ -2419,6 +2437,30 @@ break; } + case _LOAD_GLOBAL_MODULE: { + _Py_UopsSymbol *res; + _Py_UopsSymbol *null = NULL; + res = sym_new_not_null(ctx); + null = sym_new_null(ctx); + stack_pointer[0] = res; + if (oparg & 1) stack_pointer[1] = null; + stack_pointer += 1 + (oparg & 1); + assert(WITHIN_STACK_BOUNDS()); + break; + } + + case _LOAD_GLOBAL_BUILTINS: { + _Py_UopsSymbol *res; + _Py_UopsSymbol *null = NULL; + res = sym_new_not_null(ctx); + null = sym_new_null(ctx); + stack_pointer[0] = res; + if (oparg & 1) stack_pointer[1] = null; + stack_pointer += 1 + (oparg & 1); + assert(WITHIN_STACK_BOUNDS()); + break; + } + case _INTERNAL_INCREMENT_OPT_COUNTER: { stack_pointer += -1; assert(WITHIN_STACK_BOUNDS()); |