diff options
Diffstat (limited to 'Python/compile.c')
| -rw-r--r-- | Python/compile.c | 432 |
1 files changed, 68 insertions, 364 deletions
diff --git a/Python/compile.c b/Python/compile.c index 68af5d6..83cf455 100644 --- a/Python/compile.c +++ b/Python/compile.c @@ -70,9 +70,7 @@ && ((C)->u->u_ste->ste_type == ModuleBlock)) typedef _PyCompilerSrcLocation location; -typedef _PyCfgInstruction cfg_instr; -typedef _PyCfgBasicblock basicblock; -typedef _PyCfgBuilder cfg_builder; +typedef struct _PyCfgBuilder cfg_builder; #define LOCATION(LNO, END_LNO, COL, END_COL) \ ((const _PyCompilerSrcLocation){(LNO), (END_LNO), (COL), (END_COL)}) @@ -101,7 +99,7 @@ static jump_target_label NO_LABEL = {-1}; } #define USE_LABEL(C, LBL) \ - RETURN_IF_ERROR(instr_sequence_use_label(INSTR_SEQUENCE(C), (LBL).id)) + RETURN_IF_ERROR(_PyCompile_InstructionSequence_UseLabel(INSTR_SEQUENCE(C), (LBL).id)) /* fblockinfo tracks the current frame block. @@ -218,8 +216,9 @@ instr_sequence_new_label(instr_sequence *seq) return lbl; } -static int -instr_sequence_use_label(instr_sequence *seq, int lbl) { +int +_PyCompile_InstructionSequence_UseLabel(instr_sequence *seq, int lbl) +{ int old_size = seq->s_labelmap_size; RETURN_IF_ERROR( _PyCompile_EnsureArrayLargeEnough(lbl, @@ -238,8 +237,9 @@ instr_sequence_use_label(instr_sequence *seq, int lbl) { #define MAX_OPCODE 511 -static int -instr_sequence_addop(instr_sequence *seq, int opcode, int oparg, location loc) +int +_PyCompile_InstructionSequence_Addop(instr_sequence *seq, int opcode, int oparg, + location loc) { assert(0 <= opcode && opcode <= MAX_OPCODE); assert(IS_WITHIN_OPCODE_RANGE(opcode)); @@ -288,10 +288,12 @@ instr_sequence_fini(instr_sequence *seq) { seq->s_instrs = NULL; } -static int -instr_sequence_to_cfg(instr_sequence *seq, cfg_builder *g) { - memset(g, 0, sizeof(cfg_builder)); - RETURN_IF_ERROR(_PyCfgBuilder_Init(g)); +static cfg_builder* +instr_sequence_to_cfg(instr_sequence *seq) { + cfg_builder *g = _PyCfgBuilder_New(); + if (g == NULL) { + return NULL; + } /* There can be more than one label for the same offset. The * offset2lbl maping selects one of them which we use consistently. @@ -300,7 +302,7 @@ instr_sequence_to_cfg(instr_sequence *seq, cfg_builder *g) { int *offset2lbl = PyMem_Malloc(seq->s_used * sizeof(int)); if (offset2lbl == NULL) { PyErr_NoMemory(); - return ERROR; + goto error; } for (int i = 0; i < seq->s_used; i++) { offset2lbl[i] = -1; @@ -336,23 +338,17 @@ instr_sequence_to_cfg(instr_sequence *seq, cfg_builder *g) { goto error; } } - PyMem_Free(offset2lbl); - - int nblocks = 0; - for (basicblock *b = g->g_block_list; b != NULL; b = b->b_list) { - nblocks++; - } - if ((size_t)nblocks > SIZE_MAX / sizeof(basicblock *)) { - PyErr_NoMemory(); - return ERROR; + if (_PyCfgBuilder_CheckSize(g) < 0) { + goto error; } - return SUCCESS; + PyMem_Free(offset2lbl); + return g; error: + _PyCfgBuilder_Free(g); PyMem_Free(offset2lbl); - return ERROR; + return NULL; } - /* The following items change on entry and exit of code blocks. They must be saved and restored when returning to a block. */ @@ -920,7 +916,7 @@ codegen_addop_noarg(instr_sequence *seq, int opcode, location loc) { assert(!OPCODE_HAS_ARG(opcode)); assert(!IS_ASSEMBLER_OPCODE(opcode)); - return instr_sequence_addop(seq, opcode, 0, loc); + return _PyCompile_InstructionSequence_Addop(seq, opcode, 0, loc); } static Py_ssize_t @@ -1153,7 +1149,7 @@ codegen_addop_i(instr_sequence *seq, int opcode, Py_ssize_t oparg, location loc) int oparg_ = Py_SAFE_DOWNCAST(oparg, Py_ssize_t, int); assert(!IS_ASSEMBLER_OPCODE(opcode)); - return instr_sequence_addop(seq, opcode, oparg_, loc); + return _PyCompile_InstructionSequence_Addop(seq, opcode, oparg_, loc); } static int @@ -1163,7 +1159,7 @@ codegen_addop_j(instr_sequence *seq, location loc, assert(IS_LABEL(target)); assert(OPCODE_HAS_JUMP(opcode) || IS_BLOCK_PUSH_OPCODE(opcode)); assert(!IS_ASSEMBLER_OPCODE(opcode)); - return instr_sequence_addop(seq, opcode, target.id, loc); + return _PyCompile_InstructionSequence_Addop(seq, opcode, target.id, loc); } #define RETURN_IF_ERROR_IN_SCOPE(C, CALL) { \ @@ -7492,201 +7488,6 @@ _PyCompile_ConstCacheMergeOne(PyObject *const_cache, PyObject **obj) return SUCCESS; } - -static int * -build_cellfixedoffsets(_PyCompile_CodeUnitMetadata *umd) -{ - int nlocals = (int)PyDict_GET_SIZE(umd->u_varnames); - int ncellvars = (int)PyDict_GET_SIZE(umd->u_cellvars); - int nfreevars = (int)PyDict_GET_SIZE(umd->u_freevars); - - int noffsets = ncellvars + nfreevars; - int *fixed = PyMem_New(int, noffsets); - if (fixed == NULL) { - PyErr_NoMemory(); - return NULL; - } - for (int i = 0; i < noffsets; i++) { - fixed[i] = nlocals + i; - } - - PyObject *varname, *cellindex; - Py_ssize_t pos = 0; - while (PyDict_Next(umd->u_cellvars, &pos, &varname, &cellindex)) { - PyObject *varindex = PyDict_GetItem(umd->u_varnames, varname); - if (varindex != NULL) { - assert(PyLong_AS_LONG(cellindex) < INT_MAX); - assert(PyLong_AS_LONG(varindex) < INT_MAX); - int oldindex = (int)PyLong_AS_LONG(cellindex); - int argoffset = (int)PyLong_AS_LONG(varindex); - fixed[oldindex] = argoffset; - } - } - - return fixed; -} - -#define IS_GENERATOR(CF) \ - ((CF) & (CO_GENERATOR | CO_COROUTINE | CO_ASYNC_GENERATOR)) - -static int -insert_prefix_instructions(_PyCompile_CodeUnitMetadata *umd, basicblock *entryblock, - int *fixed, int nfreevars, int code_flags) -{ - assert(umd->u_firstlineno > 0); - - /* Add the generator prefix instructions. */ - if (IS_GENERATOR(code_flags)) { - /* Note that RETURN_GENERATOR + POP_TOP have a net stack effect - * of 0. This is because RETURN_GENERATOR pushes an element - * with _PyFrame_StackPush before switching stacks. - */ - cfg_instr make_gen = { - .i_opcode = RETURN_GENERATOR, - .i_oparg = 0, - .i_loc = LOCATION(umd->u_firstlineno, umd->u_firstlineno, -1, -1), - .i_target = NULL, - }; - RETURN_IF_ERROR(_PyBasicblock_InsertInstruction(entryblock, 0, &make_gen)); - cfg_instr pop_top = { - .i_opcode = POP_TOP, - .i_oparg = 0, - .i_loc = NO_LOCATION, - .i_target = NULL, - }; - RETURN_IF_ERROR(_PyBasicblock_InsertInstruction(entryblock, 1, &pop_top)); - } - - /* Set up cells for any variable that escapes, to be put in a closure. */ - const int ncellvars = (int)PyDict_GET_SIZE(umd->u_cellvars); - if (ncellvars) { - // umd->u_cellvars has the cells out of order so we sort them - // before adding the MAKE_CELL instructions. Note that we - // adjust for arg cells, which come first. - const int nvars = ncellvars + (int)PyDict_GET_SIZE(umd->u_varnames); - int *sorted = PyMem_RawCalloc(nvars, sizeof(int)); - if (sorted == NULL) { - PyErr_NoMemory(); - return ERROR; - } - for (int i = 0; i < ncellvars; i++) { - sorted[fixed[i]] = i + 1; - } - for (int i = 0, ncellsused = 0; ncellsused < ncellvars; i++) { - int oldindex = sorted[i] - 1; - if (oldindex == -1) { - continue; - } - cfg_instr make_cell = { - .i_opcode = MAKE_CELL, - // This will get fixed in offset_derefs(). - .i_oparg = oldindex, - .i_loc = NO_LOCATION, - .i_target = NULL, - }; - if (_PyBasicblock_InsertInstruction(entryblock, ncellsused, &make_cell) < 0) { - PyMem_RawFree(sorted); - return ERROR; - } - ncellsused += 1; - } - PyMem_RawFree(sorted); - } - - if (nfreevars) { - cfg_instr copy_frees = { - .i_opcode = COPY_FREE_VARS, - .i_oparg = nfreevars, - .i_loc = NO_LOCATION, - .i_target = NULL, - }; - RETURN_IF_ERROR(_PyBasicblock_InsertInstruction(entryblock, 0, ©_frees)); - } - - return SUCCESS; -} - -static int -fix_cell_offsets(_PyCompile_CodeUnitMetadata *umd, basicblock *entryblock, int *fixedmap) -{ - int nlocals = (int)PyDict_GET_SIZE(umd->u_varnames); - int ncellvars = (int)PyDict_GET_SIZE(umd->u_cellvars); - int nfreevars = (int)PyDict_GET_SIZE(umd->u_freevars); - int noffsets = ncellvars + nfreevars; - - // First deal with duplicates (arg cells). - int numdropped = 0; - for (int i = 0; i < noffsets ; i++) { - if (fixedmap[i] == i + nlocals) { - fixedmap[i] -= numdropped; - } - else { - // It was a duplicate (cell/arg). - numdropped += 1; - } - } - - // Then update offsets, either relative to locals or by cell2arg. - for (basicblock *b = entryblock; b != NULL; b = b->b_next) { - for (int i = 0; i < b->b_iused; i++) { - cfg_instr *inst = &b->b_instr[i]; - // This is called before extended args are generated. - assert(inst->i_opcode != EXTENDED_ARG); - int oldoffset = inst->i_oparg; - switch(inst->i_opcode) { - case MAKE_CELL: - case LOAD_CLOSURE: - case LOAD_DEREF: - case STORE_DEREF: - case DELETE_DEREF: - case LOAD_FROM_DICT_OR_DEREF: - assert(oldoffset >= 0); - assert(oldoffset < noffsets); - assert(fixedmap[oldoffset] >= 0); - inst->i_oparg = fixedmap[oldoffset]; - } - } - } - - return numdropped; -} - - -static int -prepare_localsplus(_PyCompile_CodeUnitMetadata *umd, cfg_builder *g, int code_flags) -{ - assert(PyDict_GET_SIZE(umd->u_varnames) < INT_MAX); - assert(PyDict_GET_SIZE(umd->u_cellvars) < INT_MAX); - assert(PyDict_GET_SIZE(umd->u_freevars) < INT_MAX); - int nlocals = (int)PyDict_GET_SIZE(umd->u_varnames); - int ncellvars = (int)PyDict_GET_SIZE(umd->u_cellvars); - int nfreevars = (int)PyDict_GET_SIZE(umd->u_freevars); - assert(INT_MAX - nlocals - ncellvars > 0); - assert(INT_MAX - nlocals - ncellvars - nfreevars > 0); - int nlocalsplus = nlocals + ncellvars + nfreevars; - int* cellfixedoffsets = build_cellfixedoffsets(umd); - if (cellfixedoffsets == NULL) { - return ERROR; - } - - - // This must be called before fix_cell_offsets(). - if (insert_prefix_instructions(umd, g->g_entryblock, cellfixedoffsets, nfreevars, code_flags)) { - PyMem_Free(cellfixedoffsets); - return ERROR; - } - - int numdropped = fix_cell_offsets(umd, g->g_entryblock, cellfixedoffsets); - PyMem_Free(cellfixedoffsets); // At this point we're done with it. - cellfixedoffsets = NULL; - if (numdropped < 0) { - return ERROR; - } - - nlocalsplus -= numdropped; - return nlocalsplus; -} - static int add_return_at_end(struct compiler *c, int addNone) { @@ -7700,12 +7501,11 @@ add_return_at_end(struct compiler *c, int addNone) return SUCCESS; } -static int cfg_to_instr_sequence(cfg_builder *g, instr_sequence *seq); - static PyCodeObject * optimize_and_assemble_code_unit(struct compiler_unit *u, PyObject *const_cache, int code_flags, PyObject *filename) { + cfg_builder *g = NULL; instr_sequence optimized_instrs; memset(&optimized_instrs, 0, sizeof(instr_sequence)); @@ -7714,51 +7514,28 @@ optimize_and_assemble_code_unit(struct compiler_unit *u, PyObject *const_cache, if (consts == NULL) { goto error; } - cfg_builder g; - if (instr_sequence_to_cfg(&u->u_instr_sequence, &g) < 0) { + g = instr_sequence_to_cfg(&u->u_instr_sequence); + if (g == NULL) { goto error; } - int nparams = (int)PyList_GET_SIZE(u->u_ste->ste_varnames); int nlocals = (int)PyDict_GET_SIZE(u->u_metadata.u_varnames); + int nparams = (int)PyList_GET_SIZE(u->u_ste->ste_varnames); assert(u->u_metadata.u_firstlineno); - if (_PyCfg_OptimizeCodeUnit(&g, consts, const_cache, nlocals, + + if (_PyCfg_OptimizeCodeUnit(g, consts, const_cache, nlocals, nparams, u->u_metadata.u_firstlineno) < 0) { goto error; } - int stackdepth = _PyCfg_Stackdepth(&g); - if (stackdepth < 0) { + int stackdepth; + int nlocalsplus; + if (_PyCfg_OptimizedCfgToInstructionSequence(g, &u->u_metadata, code_flags, + &stackdepth, &nlocalsplus, + &optimized_instrs) < 0) { goto error; } - /* prepare_localsplus adds instructions for generators that push - * and pop an item on the stack. This assertion makes sure there - * is space on the stack for that. - * It should always be true, because a generator must have at - * least one expression or call to INTRINSIC_STOPITERATION_ERROR, - * which requires stackspace. - */ - assert(!(IS_GENERATOR(code_flags) && stackdepth == 0)); - /** Assembly **/ - int nlocalsplus = prepare_localsplus(&u->u_metadata, &g, code_flags); - if (nlocalsplus < 0) { - goto error; - } - - _PyCfg_ConvertPseudoOps(g.g_entryblock); - - /* Order of basic blocks must have been determined by now */ - - if (_PyCfg_ResolveJumps(&g) < 0) { - goto error; - } - - /* Can't modify the bytecode after computing jump offsets. */ - - if (cfg_to_instr_sequence(&g, &optimized_instrs) < 0) { - goto error; - } co = _PyAssemble_MakeCodeObject(&u->u_metadata, const_cache, consts, stackdepth, &optimized_instrs, nlocalsplus, @@ -7767,7 +7544,7 @@ optimize_and_assemble_code_unit(struct compiler_unit *u, PyObject *const_cache, error: Py_XDECREF(consts); instr_sequence_fini(&optimized_instrs); - _PyCfgBuilder_Fini(&g); + _PyCfgBuilder_Free(g); return co; } @@ -7790,39 +7567,6 @@ optimize_and_assemble(struct compiler *c, int addNone) return optimize_and_assemble_code_unit(u, const_cache, code_flags, filename); } -static int -cfg_to_instr_sequence(cfg_builder *g, instr_sequence *seq) -{ - int lbl = 0; - for (basicblock *b = g->g_entryblock; b != NULL; b = b->b_next) { - b->b_label = (jump_target_label){lbl}; - lbl += b->b_iused; - } - for (basicblock *b = g->g_entryblock; b != NULL; b = b->b_next) { - RETURN_IF_ERROR(instr_sequence_use_label(seq, b->b_label.id)); - for (int i = 0; i < b->b_iused; i++) { - cfg_instr *instr = &b->b_instr[i]; - if (OPCODE_HAS_JUMP(instr->i_opcode)) { - instr->i_oparg = instr->i_target->b_label.id; - } - RETURN_IF_ERROR( - instr_sequence_addop(seq, instr->i_opcode, instr->i_oparg, instr->i_loc)); - - _PyCompile_ExceptHandlerInfo *hi = &seq->s_instrs[seq->s_used-1].i_except_handler_info; - if (instr->i_except != NULL) { - hi->h_label = instr->i_except->b_label.id; - hi->h_startdepth = instr->i_except->b_startdepth; - hi->h_preserve_lasti = instr->i_except->b_preserve_lasti; - } - else { - hi->h_label = -1; - } - } - } - return SUCCESS; -} - - /* Access to compiler optimizations for unit tests. * * _PyCompile_CodeGen takes and AST, applies code-gen and @@ -7872,7 +7616,7 @@ instructions_to_instr_sequence(PyObject *instructions, instr_sequence *seq) for (int i = 0; i < num_insts; i++) { if (is_target[i]) { - if (instr_sequence_use_label(seq, i) < 0) { + if (_PyCompile_InstructionSequence_UseLabel(seq, i) < 0) { goto error; } } @@ -7912,7 +7656,7 @@ instructions_to_instr_sequence(PyObject *instructions, instr_sequence *seq) if (PyErr_Occurred()) { goto error; } - if (instr_sequence_addop(seq, opcode, oparg, loc) < 0) { + if (_PyCompile_InstructionSequence_Addop(seq, opcode, oparg, loc) < 0) { goto error; } } @@ -7923,23 +7667,26 @@ error: return ERROR; } -static int -instructions_to_cfg(PyObject *instructions, cfg_builder *g) +static cfg_builder* +instructions_to_cfg(PyObject *instructions) { + cfg_builder *g = NULL; instr_sequence seq; memset(&seq, 0, sizeof(instr_sequence)); if (instructions_to_instr_sequence(instructions, &seq) < 0) { goto error; } - if (instr_sequence_to_cfg(&seq, g) < 0) { + g = instr_sequence_to_cfg(&seq); + if (g == NULL) { goto error; } instr_sequence_fini(&seq); - return SUCCESS; + return g; error: + _PyCfgBuilder_Free(g); instr_sequence_fini(&seq); - return ERROR; + return NULL; } static PyObject * @@ -7978,42 +7725,14 @@ error: static PyObject * cfg_to_instructions(cfg_builder *g) { - PyObject *instructions = PyList_New(0); - if (instructions == NULL) { + instr_sequence seq; + memset(&seq, 0, sizeof(seq)); + if (_PyCfg_ToInstructionSequence(g, &seq) < 0) { return NULL; } - int lbl = 0; - for (basicblock *b = g->g_entryblock; b != NULL; b = b->b_next) { - b->b_label = (jump_target_label){lbl}; - lbl += b->b_iused; - } - for (basicblock *b = g->g_entryblock; b != NULL; b = b->b_next) { - for (int i = 0; i < b->b_iused; i++) { - cfg_instr *instr = &b->b_instr[i]; - location loc = instr->i_loc; - int arg = HAS_TARGET(instr->i_opcode) ? - instr->i_target->b_label.id : instr->i_oparg; - - PyObject *inst_tuple = Py_BuildValue( - "(iiiiii)", instr->i_opcode, arg, - loc.lineno, loc.end_lineno, - loc.col_offset, loc.end_col_offset); - if (inst_tuple == NULL) { - goto error; - } - - if (PyList_Append(instructions, inst_tuple) != 0) { - Py_DECREF(inst_tuple); - goto error; - } - Py_DECREF(inst_tuple); - } - } - - return instructions; -error: - Py_DECREF(instructions); - return NULL; + PyObject *res = instr_sequence_to_instructions(&seq); + instr_sequence_fini(&seq); + return res; } // C implementation of inspect.cleandoc() @@ -8195,34 +7914,36 @@ finally: PyObject * _PyCompile_OptimizeCfg(PyObject *instructions, PyObject *consts, int nlocals) { + cfg_builder *g = NULL; PyObject *res = NULL; PyObject *const_cache = PyDict_New(); if (const_cache == NULL) { return NULL; } - cfg_builder g; - if (instructions_to_cfg(instructions, &g) < 0) { + g = instructions_to_cfg(instructions); + if (g == NULL) { goto error; } int nparams = 0, firstlineno = 1; - if (_PyCfg_OptimizeCodeUnit(&g, consts, const_cache, nlocals, + if (_PyCfg_OptimizeCodeUnit(g, consts, const_cache, nlocals, nparams, firstlineno) < 0) { goto error; } - res = cfg_to_instructions(&g); + res = cfg_to_instructions(g); error: Py_DECREF(const_cache); - _PyCfgBuilder_Fini(&g); + _PyCfgBuilder_Free(g); return res; } -int _PyCfg_JumpLabelsToTargets(basicblock *entryblock); +int _PyCfg_JumpLabelsToTargets(cfg_builder *g); PyCodeObject * _PyCompile_Assemble(_PyCompile_CodeUnitMetadata *umd, PyObject *filename, PyObject *instructions) { + cfg_builder *g = NULL; PyCodeObject *co = NULL; instr_sequence optimized_instrs; memset(&optimized_instrs, 0, sizeof(instr_sequence)); @@ -8232,37 +7953,20 @@ _PyCompile_Assemble(_PyCompile_CodeUnitMetadata *umd, PyObject *filename, return NULL; } - cfg_builder g; - if (instructions_to_cfg(instructions, &g) < 0) { + g = instructions_to_cfg(instructions); + if (g == NULL) { goto error; } - if (_PyCfg_JumpLabelsToTargets(g.g_entryblock) < 0) { - goto error; - } - - int stackdepth = _PyCfg_Stackdepth(&g); - if (stackdepth < 0) { + if (_PyCfg_JumpLabelsToTargets(g) < 0) { goto error; } int code_flags = 0; - int nlocalsplus = prepare_localsplus(umd, &g, code_flags); - if (nlocalsplus < 0) { - goto error; - } - - _PyCfg_ConvertPseudoOps(g.g_entryblock); - - /* Order of basic blocks must have been determined by now */ - - if (_PyCfg_ResolveJumps(&g) < 0) { - goto error; - } - - /* Can't modify the bytecode after computing jump offsets. */ - - if (cfg_to_instr_sequence(&g, &optimized_instrs) < 0) { + int stackdepth, nlocalsplus; + if (_PyCfg_OptimizedCfgToInstructionSequence(g, umd, code_flags, + &stackdepth, &nlocalsplus, + &optimized_instrs) < 0) { goto error; } @@ -8277,7 +7981,7 @@ _PyCompile_Assemble(_PyCompile_CodeUnitMetadata *umd, PyObject *filename, error: Py_DECREF(const_cache); - _PyCfgBuilder_Fini(&g); + _PyCfgBuilder_Free(g); instr_sequence_fini(&optimized_instrs); return co; } |