From: Thomas Walker Lynch Date: Thu, 8 May 2025 09:18:43 +0000 (-0700) Subject: partial progress implementing #assign directive X-Git-Url: https://git.reasoningtechnology.com/usr/lib/python2.7/encodings/iso8859_9.py?a=commitdiff_plain;h=5c657e20c0d1ef7706db5274acffbff03c2e545a;p=RT-gcc partial progress implementing #assign directive --- diff --git "a/document\360\237\226\211/source/cpp.org" "b/document\360\237\226\211/source/cpp.org" new file mode 100644 index 0000000..b38c7ba --- /dev/null +++ "b/document\360\237\226\211/source/cpp.org" @@ -0,0 +1,503 @@ +#+TITLE: C Preprocessor Overview +#+AUTHOR: Thomas Walker Lynch & Caelestis Index +#+DESCRIPTION: High-level architectural partitioning of cpp (GCC 12.x) +#+FILETAGS: cpp preprocessor architecture gcc +#+OPTIONS: toc:nil + +* Preprocessing Pipeline (Diagram) + +#+BEGIN_SRC text + C Preprocessor (cpp) + ===================== + ++----------------------+ +| Source Code | ++----------------------+ + | + v ++----------------------+ +| Lexical Analysis | <- Part of: Lexical Analysis +| (tokenize input) | ++----------------------+ + | + v ++----------------------+ +| Directive Engine | <- Part of: Directive Handling +| (#define, #if, etc.) | ++----------------------+ + | + v ++----------------------+ +| Conditional Logic | <- Part of: Conditional Compilation +| (#if/#ifdef/#else) | ++----------------------+ + | + v ++----------------------+ +| Macro Expansion | <- Part of: Macro Expansion +| (object/function) | ++----------------------+ + | + v ++----------------------+ +| Callback Hooks | <- Part of: Hook and Callback Interface +| (cpp_callbacks) | ++----------------------+ + | + v ++----------------------+ +| Output Tokens | <- Output stream to compiler frontend +| (to GCC parser) | ++----------------------+ +#+END_SRC + +Each block corresponds to a major processing stage in `cpp`. The functional groups defined earlier align to these blocks as indicated, though some (like state management and diagnostics) operate globally across the pipeline. + + + +* Major Functional Partitions of the C Preprocessor (cpp) + +This section outlines the primary architectural components of the C preprocessor as implemented in GCC 12.x. These functional partitions help frame how cpp processes input and how its internal modules interact. + +** 1. Lexical Analysis +- Tokenizes input into =cpp_token= streams. +- Decodes: + - UTF-8 characters + - Trigraphs (e.g., =??=) + - Digraphs (e.g., =<: = for =[=) +- Central structure: =cpp_lexer= +- Produces tokens for macro expansion and conditional evaluation. +** 2. Directive Handling +- Processes all =#= directives, including: + - =#define=, =#undef=, =#include=, =#line=, =#error=, =#pragma= + - Extended directives like =#assign=, =#call= if supported. +- Managed via =directive_table= and dispatch functions like =do_define=, =do_include=, etc. + +** 3. Conditional Compilation +- Handles constructs like: + - =#if=, =#ifdef=, =#ifndef=, =#elif=, =#else=, =#endif= +- Used to include or exclude code based on macro definitions and constant expressions. +- Driven by the =if_stack= in =cpp_reader=. +- Central to controlling variant builds, platform-specific code, or staged compilation. +** 4. File Inclusion and Search Paths +- Resolves =#include= and maintains include history. +- Handles: + - System vs user includes (<...> vs "..."). + - Include path resolution via =cpp_search_path=. + - File change tracking via =file_stack=. +** 5. Macro Expansion +- Handles object-like and function-like macros: + - =#define PI 3.14= + - =#define SQR(x) ((x)*(x))= +- Manages: + - Argument collection and expansion + - Token-pasting (=##=) and stringification (=#=) +- Involves =macro_table=, =collect_args=, and =expand_macro()= + +** 6. Diagnostics and Error Recovery +- Reports syntax errors, macro misuse, directive misuse. +- Uses: + - =cpp_error=, =cpp_warning=, =cpp_notice= + - Tracks macro nesting, input location, and file state for context. + +** 7. Hook and Callback Interface +- Interface: =cpp_callbacks= +- Allows frontend or plugin to observe: + - Macro definitions + - File changes + - Token output stream +- Enables debugging tools, IDEs, or language servers to integrate preprocessor awareness. + +** 8. State Management and Scoping +- Maintains global and file-level preprocessor state. +- Tracks: + - Nested conditional state via =if_stack= + - Macro table lifetimes and shadowing + - Include guards and =#pragma once= heuristics + + +* cpplib.h -- Application Interface Overview + +This section documents the **interface** and **in-memory model** of the C preprocessor (`libcpp`) from GCC 12.2.0. +It covers core data structures (tokens, macros, readers) and the primary functions for working with them. + +** Key Data Structures + +*** Token & Token Metadata +- `enum cpp_ttype` :: All possible token types (operators, names, literals, etc.) +- `struct cpp_token` :: Represents a token in the stream (with union-based payload) +- `enum cpp_token_fld_kind` :: Discriminates the active field in `cpp_token.val` +- `struct cpp_string` :: Raw string representation with length and pointer + +*** Macros & Identifiers +- `struct cpp_macro` :: Describes macro kind, parameter list, and token expansion +- `enum cpp_macro_kind` :: ISO-style, traditional-style, and assertion macros +- `struct cpp_identifier` :: Canonical and original spellings of a name +- `struct cpp_macro_arg` :: Argument number and spelling for macro arguments + +*** Symbol Table +- `struct cpp_hashnode` :: Hash table node for identifiers/macros +- `enum node_type` :: Distinguishes macro types (arg/user/builtin) +- `union _cpp_hashnode_value` :: Payload (macro, arg index, etc.) +- `enum cpp_builtin_type` :: Reserved built-ins like `__LINE__`, `__FILE__`, `_Pragma` + +*** Reader & Configuration +- `struct cpp_reader` :: Forward-declared. Central structure for preprocessing. +- `struct cpp_options` :: Stores all language mode flags, warning flags, and feature toggles. +- `struct cpp_callbacks` :: Client hook interface for diagnostic, macro, and file events. +- `struct cpp_dir` :: Represents an `#include` search directory. + +*** Numerics +- `struct cpp_num` :: Two-part 64-bit integer (high, low), overflow flags +- `cpp_classify_number` :: Categorizes radix/type (e.g., `0x`, `u`, `LL`) +- Defines :: `CPP_N_*` classify bits (INTEGER, FLOATING, WIDTH, RADIX, SUFFIX) + +*** Charset Handling +- `typedef cppchar_t` :: 32-bit safe character representation +- `struct cpp_decoded_char` :: Result of UTF-8 decoding step +- `struct cpp_char_column_policy` :: Visual column handling for diagnostics +- `class cpp_display_width_computation` :: Converts UTF-8 sequence to visual width + +*** Comment Tracking +- `struct cpp_comment`, `cpp_comment_table` :: Captures all parsed comments (if enabled) + +** Core Functions + +*** Lifecycle & Reader Setup +- `cpp_create_reader(enum c_lang, ...)` :: Allocates and initializes `cpp_reader` +- `cpp_finish`, `cpp_destroy` :: Finalize and free the reader +- `cpp_post_options` :: Commit option changes after parsing flags + +*** Preprocessing Input +- `cpp_read_main_file` :: Begin reading and preprocessing a source file +- `cpp_get_token()` :: Fetch next token from stream +- `cpp_peek_token()` :: Peek ahead without consuming +- `cpp_backup_tokens()` :: Push tokens back for re-parsing +- `cpp_retrofit_as_include()` :: Treat main file as if included + +*** Macro System +- `cpp_define()`, `cpp_define_unused()`, `cpp_define_lazily()` :: Define macros +- `cpp_macro_definition()` :: Dump macro body as string +- `cpp_compare_macros()` :: Deep compare two macros +- `cpp_undef()`, `cpp_undef_all()` :: Remove macro(s) +- `cpp_set_deferred_macro()`, `cpp_get_deferred_macro()` :: Lazy macro substitution + +*** Symbol Lookup +- `cpp_lookup()` :: Lookup or create an identifier hashnode +- `cpp_forall_identifiers()` :: Iterate over all identifiers + +*** String & Char Evaluation +- `cpp_interpret_charconst()` :: Parse a character constant (e.g. `'a'`) +- `cpp_interpret_string()` :: Parse string literal(s) into `cpp_string` +- `cpp_interpret_integer()` :: Parse numeric token into `cpp_num` + +*** Diagnostics +- `cpp_error()`, `cpp_warning()`, `cpp_pedwarning()` :: General messages +- `cpp_error_at()` :: Message with source location (rich_location optional) +- `cpp_errno()` / `cpp_errno_filename()` :: Errors based on `errno` +- `cpp_warning_with_line()` :: Fallback location-based warnings +- `cpp_get_callbacks()` / `cpp_set_callbacks()` :: Manage diagnostic hooks + +*** Extension Hooks & Pragma +- `cpp_register_pragma()` :: Register custom `#pragma` handler +- `cpp_get_callbacks()` :: Access to client-supplied hook table +- `cpp_define_formatted()` :: Macro with `printf`-style input +- `cpp_directive_only_process()` :: Run directive-only logic on a token stream + +*** Includes & File Management +- `cpp_set_include_chains()` :: Set system and user include paths +- `cpp_push_buffer()` :: Manually push a buffer for parsing +- `cpp_included()`, `cpp_included_before()` :: Has this file been included? +- `cpp_get_converted_source()` :: Read a file in input charset, return decoded buffer + +** Token Types (cpp_ttype) + +A full enumeration of all tokens in the preprocessor: +- Operators: `CPP_PLUS`, `CPP_MINUS`, `CPP_EQ_EQ`, etc. +- Punctuation: `CPP_OPEN_PAREN`, `CPP_HASH`, `CPP_SEMICOLON` +- Literals: `CPP_STRING`, `CPP_WCHAR`, `CPP_NUMBER` +- Special: `CPP_MACRO_ARG`, `CPP_PRAGMA`, `CPP_EOF` + +Each token has: +- Type (`enum cpp_ttype`) +- Flags (`PREV_WHITE`, `DIGRAPH`, `NO_EXPAND`, etc.) +- Source location +- Union payload (e.g., string, macro arg, hashnode) + +** Interface Concepts Beyond Code +*** Unicode Handling +- Input is normalized per `cpp_normalize_level` +- UTF-8 is expanded into 32-bit code points (`cppchar_t`) +- Display width of characters is estimated for diagnostics +- Bidi (bidirectional) controls are optionally scanned/warned + +*** Client Extension Hooks +- Most preprocessing operations (macro use, `#include`, comments, errors) are callback-hooked +- Used by GCC frontend to track macro use, implement diagnostics, and guide `#pragma` processing + +*** Dependency Generation +- `cpp_finish()` accepts an output stream for dependency info +- Options control whether main file is included, phony targets are added, etc. + +** Summary + +`cpplib.h` serves as both API contract and internal representation guide. +- It offers a high-fidelity view of source tokens for later compiler stages. +- The entire macro system, character encoding, and diagnostic lifecycle are managed through this interface. + + + + +* Callback Hooks (cpp_callbacks) + +The `cpp_callbacks` struct in `cpplib.h` allows external consumers (e.g., GCC frontend, IDE integrations, or plugins) to receive notifications during preprocessing. Each function pointer in this struct represents a hookable event. + +** Overview + +Hooks are triggered at specific stages: +- After macro definition or undefinition +- Before and after file inclusion +- When tokens are emitted +- Upon encountering diagnostics +- During comment scanning (if enabled) +- On encountering special directives (e.g., `#pragma`) + +** Hook Structure + +#+BEGIN_SRC c +struct cpp_callbacks { + void (*define)(cpp_reader *, source_location, const cpp_hashnode *); + void (*undef)(cpp_reader *, source_location, const cpp_hashnode *); + void (*include)(cpp_reader *, const char *filename, int angle_brackets); + void (*file_change)(cpp_reader *, const struct line_map *); + void (*line_change)(cpp_reader *, source_location, int to_file, int to_line); + void (*ident)(cpp_reader *, const cpp_string *); + void (*invalid_directive)(cpp_reader *); + void (*def_pragma)(cpp_reader *, const cpp_token *); + void (*cb_comment)(cpp_reader *, const cpp_token *); +}; +#+END_SRC + +Each callback receives either a pointer to the `cpp_reader`, the affected token or structure, and optional contextual data. + +--- + +** `define` + +*** Trigger +- Fired immediately after a macro is defined with `#define`. + +*** Parameters +- `cpp_reader *pfile`: global preprocessor state (read-write). +- `source_location loc`: location of the `#define`. +- `const cpp_hashnode *node`: the macro name and metadata (read-only in this context). + +*** Semantics +- The `cpp_hashnode` holds the macro's name and a pointer to its `cpp_macro` definition. +- Modifying the macro at this point is possible but discouraged. Use `cpp_undef()` + `cpp_define()` instead if redefinition is needed. + +*** Uses +- GCC uses this to update dependency tracking and debug tables. +- Tools may track macro definitions, emit logs, or enforce naming policies. + +--- + +** `undef` + +*** Trigger +- Fired after `#undef` removes a macro. + +*** Parameters +- Same as `define`. + +*** Semantics +- The node is marked `undefined`, but the symbol remains in the hash table. +- No mutation should occur—only inspection or logging. + +*** Uses +- Enables reversal tracking or macro scoping analysis. + +--- + +** `include` + +*** Trigger +- Fired just before a file is opened via `#include`. + +*** Parameters +- `cpp_reader *pfile` +- `const char *filename`: string from the include directive (not normalized). +- `int angle_brackets`: nonzero for `<...>`, zero for `"..."`. + +*** Semantics +- Purely informational; does not affect include search or suppression. +- The filename is unverified and not guaranteed to exist. + +*** Uses +- IDEs and build tools use this to build include graphs. +- LSPs use it to track file references and symbol origins. + +--- + +** `file_change` + +*** Trigger +- Called when the active input file changes (entry or exit of `#include`). + +*** Parameters +- `cpp_reader *pfile` +- `const struct line_map *map`: describes the current file's location and context. + +*** Semantics +- `line_map` gives full access to file/line/column mapping. +- This structure is read-only; mutating it will corrupt diagnostics and tokenization. + +*** Uses +- Debug info (DWARF line tables), logging, stack-based include tracking. + +--- + +** `line_change` + +*** Trigger +- Fired on `#line` directives or line-mapping transitions. + +*** Parameters +- `cpp_reader *pfile` +- `source_location loc`: location in input stream. +- `int to_file`: non-zero if a new file name is being used. +- `int to_line`: new logical line number. + +*** Semantics +- Use this to remap locations or re-synchronize overlays. +- These values are inputs to the line map; do not write back. + +*** Uses +- Used in DWARF debug info to support accurate line-based breakpoints. + +--- + +** `ident` + +*** Trigger +- Called when a `#ident` directive is parsed. + +*** Parameters +- `cpp_reader *pfile` +- `const cpp_string *text`: payload of the identifier message. + +*** Semantics +- Informational only. Common in legacy systems or codegen traces. + +*** Uses +- Collect module identity, versioning hints, or logmarks. + +--- + +** `invalid_directive` + +*** Trigger +- Fired when an unrecognized or malformed directive is encountered. + +*** Parameters +- `cpp_reader *pfile` + +*** Semantics +- Hook has no extra context; use `cpp_get_token()` to recover. +- Hook may trigger fallback behavior or custom directive logic. + +*** Uses +- Used in `-fpreprocessed` mode to suppress diagnostics. +- External tools can use this to extend the directive set. + +--- + +** `def_pragma` + +*** Trigger +- Fired when a `#pragma` directive is parsed. + +*** Parameters +- `cpp_reader *pfile` +- `const cpp_token *pragma`: token stream beginning with `CPP_PRAGMA`. + +*** Semantics +- Read-only access to token stream. +- Mutation possible via `cpp_push_buffer()` to inject expanded tokens. + +*** Uses +- GCC plugins hook this to implement custom `#pragma` behavior. +- Can trigger front-end features (like `#pragma GCC diagnostic`). + +--- + +** `cb_comment` + +*** Trigger +- Optional. Enabled if comment tracking is requested. + +*** Parameters +- `cpp_reader *pfile` +- `const cpp_token *comment`: holds text of comment. + +*** Semantics +- Only line/block comment content is captured, not semantics. +- Read-only token; do not mutate token payload. + +*** Uses +- Used by source-to-source translators and formatters. +- Some static analyzers inspect comments for hints or disables. + +--- + +** Summary + +The `cpp_callbacks` interface enables observational and limited transformational interaction with the preprocessor pipeline. + +- Most parameters are read-only or shallow copies. +- For transformations, prefer using `cpp_define()`, `cpp_push_buffer()`, or `cpp_backup_tokens()` externally. +- Internal structures like `cpp_reader`, `cpp_token`, and `cpp_macro` should not be mutated unless explicitly permitted. + + + +* Plugin-Like Integration in libcpp + +Unlike the main GCC compiler, which supports a formal plugin system (`gcc-plugin.h`), `libcpp` (the C preprocessor library) does *not* support plugins in the dynamic or runtime-loaded sense. There is no system for loading shared libraries, registering handlers via symbols, or extending preprocessor behavior through runtime modules. + +** Static Hook Interface via cpp_callbacks + +Instead, `libcpp` exposes a *statically defined interface* (`struct cpp_callbacks`) for embedding applications to receive notifications of preprocessor events. These include: + +- Macro definitions and undefinitions +- Source file entry/exit +- Comment and pragma parsing +- Token emission and buffer transitions + +An embedding client (such as GCC's C/C++ frontend, or a third-party tool using libcpp) may assign function pointers directly into this struct during reader setup. + +#+BEGIN_SRC c +cpp_reader *r = cpp_create_reader(...); +cpp_callbacks *cb = cpp_get_callbacks(r); +cb->macro_defined = my_macro_handler; +cb->file_change = my_file_tracker; +#+END_SRC + +This pattern is analogous to a *plugin interface*, but all logic is statically linked at compile time. + +** Mutability and Access Scope + +The callback interface is primarily **observational**—that is, hooks are expected to inspect events, not mutate the `cpp_reader` state directly. However, advanced users can, with care, reach into the data structures passed to them (e.g., `cpp_macro`, `cpp_hashnode`) and affect behavior, though this is neither documented nor officially supported. + +In summary: + +| Feature | GCC Frontend Plugin | libcpp Callback Interface | +|--------------------------+---------------------+----------------------------| +| Dynamically loadable | Yes | No | +| Runtime extension API | Yes (`gcc-plugin.h`) | No | +| Assign custom handlers | Yes | Yes (via `cpp_callbacks`) | +| Mutate core structures | With care | With care (not endorsed) | +| Stability across versions| Best-effort | Internal API, may break | + +** Recommendation + +Use `cpp_callbacks` as a read-only interface to monitor preprocessing behavior. If deeper mutation or instrumentation is required, consider modifying or forking `libcpp` itself. There is currently no officially supported way to extend it at runtime. diff --git "a/document\360\237\226\211/source/cpp_reader.org" "b/document\360\237\226\211/source/cpp_reader.org" new file mode 100644 index 0000000..bc87d15 --- /dev/null +++ "b/document\360\237\226\211/source/cpp_reader.org" @@ -0,0 +1,147 @@ +#+TITLE: cpp_reader: Preprocessor State and Interface Guide +#+AUTHOR: Caelestis Index +#+FILETAGS: cpp, GCC internals, preprocessor, architecture + +* Overview +The =cpp_reader= struct in GCC's =libcpp= encapsulates the complete state of a single C preprocessor session. It governs token input, macro expansion, directive parsing, include stack management, and source map resolution. It is the central state object passed through nearly all parts of the C preprocessor. + +* 1. State Data + +** 1.1 Buffer and Lexing State +- ~buffer~, ~overlaid_buffer~: Input buffer stack for file and macro streams. +- ~cur_token~, ~cur_run~, ~base_run~: Active token buffer and tokenrun tracking. +- ~keep_tokens~: Whether to preserve old tokens (e.g., for diagnostics). +- ~a_buff~, ~u_buff~, ~free_buffs~: Temporary memory allocation pools. + +** 1.2 Parsing and Directive State +- ~state~: General lexer state (includes ~in_directive~ flag). +- ~state.in_directive~: Boolean flag indicating whether the preprocessor is currently parsing a directive line. If ~true~, token behavior (e.g., whitespace and line continuation) may differ. +- ~directive~, ~directive_line~: Currently parsed directive and its location. +- ~directive_result~: Token synthesized by a directive (if any). + +** 1.3 Macro Context and Expansion +- ~context~, ~base_context~: Macro expansion call stack. +- ~top_most_macro_node~: Current top-level macro under expansion. +- ~about_to_expand_macro_p~: Indicates if a macro is about to expand. +- ~macro_buffer~, ~macro_buffer_len~: Buffers for rendering macro string forms. + +** 1.4 Include and File Lookup State +- ~quote_include~, ~bracket_include~, ~no_search_path~: Search paths. +- ~all_files~, ~main_file~: Linked list of all known input files. +- ~file_hash~, ~dir_hash~: Hashtables for file path caching. +- ~nonexistent_file_hash~: Optimizes negative lookup caching. +- ~seen_once_only~: Tracks ~#pragma once~ semantics. + +** 1.5 Character Set Conversion +- ~narrow_cset_desc~, ~utf8_cset_desc~, ~wide_cset_desc~, etc.: Converters for source to execution character encodings. + +** 1.6 Location Mapping and Source Positioning +- ~line_table~: GCC's =line_maps= structure for virtual location tracking. +- ~invocation_location~, ~main_loc~, ~forced_token_location~: Positional context for diagnostics, token creation. + +** 1.7 Miscellaneous Flags and Utilities +- ~quote_ignores_source_dir~: Include resolution behavior flag. +- ~counter~: Value of the ~__COUNTER__~ macro. +- ~out~: Output buffer for traditional preprocessing mode. +- ~savedstate~: Used for dependency tracking with precompiled headers. +- ~comments~: Optional comment capture buffer. + +* 2. Core Interface Functions +** 2.1 Token Retrieval +- ~cpp_get_token(pfile)~: Public interface for retrieving the next logical token. +- ~cpp_peek_token(pfile, N)~: Look ahead without consuming. +- ~cpp_get_token_1(pfile)~: Internal token fetch used during macro expansion. + +** 2.2 Macro Definition and Expansion +- ~_cpp_new_macro(pfile, cmk_macro, obstack_ptr)~: Allocate and initialize a new macro definition. +- ~_cpp_mark_macro_used(node)~: Mark a macro as having been used. +- ~replace_args(...)~: Expand and replace macro arguments (not used during directive handling). +- ~collect_args(...)~: Collects arguments for a function-like macro invocation. +- ~collect_single_argument(...)~: Parses one macro argument and handles token accumulation. +- ~cpp_arguments_ok(...)~: Checks argument count and matching for a macro invocation. +- ~set_arg_token(...)~: Sets or appends a token in an argument’s expansion list. + +** 2.3 Directive Handling Helpers +- ~_cpp_skip_rest_of_line(pfile)~: Skip trailing tokens after directive arguments. +- ~lex_macro_node(pfile)~: Specialized lexer for parsing macro names. + +** 2.4 File/Include Handling +- ~cpp_push_include(pfile, filename)~: Add a new include to the stack. +- ~cpp_find_include_file(...)~: Path search logic. + +** 2.5 Location Utilities +- ~cpp_token_location(token)~: Extracts a =location_t= from a token. +- ~linemap_add(...)~: Adds a mapping between logical and physical line/column. + +** 2.6 Miscellaneous +- ~cpp_warning_with_line(...)~, ~cpp_error_with_line(...)~: Emit diagnostics with location. +- ~cpp_lookup(pfile, name, length)~: Interns an identifier and returns a ~cpp_hashnode *~. +- ~NODE_NAME(node)~: Expands to the null-terminated name of a macro node. + +* 3. Usage Examples + +** 3.1 Defining a Macro from a Directive +#+BEGIN_SRC c +cpp_hashnode *node = lex_macro_node(pfile); +cpp_macro *macro = _cpp_new_macro(pfile, cmk_macro, _cpp_reserve_room(pfile, 0, sizeof(cpp_macro))); +macro->count = 1; +macro->exp.tokens[0] = make_number_token("42"); +node->type = NT_USER_MACRO; +node->value.macro = macro; +_cpp_mark_macro_used(node); +#+END_SRC + +** 3.2 Parsing a Directive With Two Arguments +#+BEGIN_SRC c +cpp_token *arg1 = cpp_get_token(pfile); +cpp_token *comma = cpp_get_token(pfile); +if (comma->type != CPP_COMMA) + cpp_error(pfile, CPP_DL_ERROR, "expected ',' after macro name"); +cpp_token *arg2 = cpp_get_token(pfile); +_cpp_skip_rest_of_line(pfile); +#+END_SRC + +** 3.3 Controlling Directive Context +#+BEGIN_SRC c +bool saved = pfile->state.in_directive; +pfile->state.in_directive = false; +assign_handler(pfile); +pfile->state.in_directive = saved; +#+END_SRC + +** 3.4 Tokenization and Location Debugging +#+BEGIN_SRC c +const cpp_token *tok = cpp_get_token(pfile); +location_t loc = tok->src_loc; +printf("token at line: %d\n", LOCATION_LINE(loc)); +#+END_SRC + +* 4. directive.cc extensions to the reader +- ~lex_macro_node(pfile)~: Returns a ~cpp_hashnode *~ for the next identifier, used for directives like ~#define~ or custom ones like ~#assign~. +- ~_cpp_skip_rest_of_line(pfile)~: Advances the token stream to the next physical line. +- ~cpp_error_with_line(...)~, ~cpp_warning_with_line(...)~: Used for directive diagnostics. +- ~cpp_lookup(pfile, name, length)~: Interns a name as a hashnode symbol. +- ~cpp_reader->directive_result~: Used to push a synthesized token result into the stream (e.g., for ~#include_next~). +- ~pfile->state.in_directive~: Must be manually toggled when directive code calls into macro infrastructure. +* 5. macro.cc extensions to the reader + +*** 4.2.1 collect_args(...) +Accumulates macro arguments for a function-like macro. Reads and segments the input stream into a series of ~macro_arg~ entries, tracking nesting of parentheses and token boundaries. + +*** 4.2.2 collect_single_argument(...) +Parses and collects one macro argument, terminating on a comma or closing paren. Used internally by ~collect_args~, but can be called separately for single-argument macro handling. + +*** 4.2.3 replace_args(...) +Performs full substitution of macro arguments into the macro body. Handles token pasting (~##~), stringification (~#~), and recursive macro expansion. + +*** 4.2.4 cpp_arguments_ok(...) +Checks whether the number of provided arguments matches the macro’s parameter list. Validates ~paramc~ and variadic status. + +*** 4.2.5 set_arg_token(...) +Helper to insert or append a token into a ~macro_arg~. Used when building argument streams in ~collect_single_argument~. + +These routines enable fine-grained control over macro behavior and can be selectively reused to simulate macro expansion at directive time (e.g., ~#assign~, ~#bind~, or macro templating extensions). +* 6. Conclusion +~cpp_reader~ is the heart of the preprocessor, acting as a unifying context for token streams, macro tables, buffer management, diagnostics, and parser state. Understanding and safely manipulating it is key to extending the preprocessor (e.g., adding new directives like ~#assign~) without destabilizing expansion or include logic. + +Use ~in_directive~, ~context~, and ~cur_token~ fields with care, and follow the established patterns in ~directives.cc~ and ~macro.cc~ to ensure consistent behavior across parse and expansion phases. diff --git "a/document\360\237\226\211/source/internal_h.org" "b/document\360\237\226\211/source/internal_h.org" new file mode 100644 index 0000000..97d10de --- /dev/null +++ "b/document\360\237\226\211/source/internal_h.org" @@ -0,0 +1,227 @@ +#+TITLE: internal.h - Documentation Reference (Emacs Org Format) +#+AUTHOR: Thomas Walker Lynch & Caelestis Index +#+DESCRIPTION: Reference breakdown of types, macros, and helper declarations in GCC's libcpp/internal.h +#+FILETAGS: cpp preprocessor gcc headers internal +#+OPTIONS: toc:nil + +* Overview +`internal.h` contains declarations for data structures, constants, and utility macros central to GCC's internal C preprocessor logic. It defines memory buffers, macro contexts, lexer state tracking, token kinds, character classes, and preprocessor infrastructure such as file buffers and include tracking. + +* Included Headers +- `symtab.h`: Symbol table definitions used internally. +- `cpplib.h`: Public CPP interfaces for tokens and readers. +- ``: (conditionally) iconv conversion API. + +* Core Data Structures +** `_cpp_buff` +A generic buffer with pointer markers. Used throughout macro processing and string/token accumulation. + +** `cpp_context` +Represents the current token expansion context. May hold ISO macro token runs or traditional literal input. Tracks virtual locations if macro tracking is enabled. + +** `macro_context` +Holds virtual locations and associated macro node. Used to support `-ftrack-macro-expansion`. + +** `cpp_reader` +Global object managing state for a preprocessing run, including lexer state, file buffers, context stack, macro table, callbacks, charset converters, and diagnostics. + +** `cpp_buffer` +Represents the input buffer of a file or command. Tracks physical and logical line positions, associated file, character set conversion, and line notes. + +** `lexer_state` +Bitfield flags tracking parsing state, expansion behavior, and preprocessor conditionals. + +** `tokenrun` +Represents a sequence of `cpp_token`s. Token runs are chained and form a circular buffer. + +** `spec_nodes` +Holds special pre-defined nodes like `defined`, `true`, `__VA_ARGS__`, etc. Used by conditional expressions and macro substitution. + +** `def_pragma_macro` +Stores push/pop state for macros affected by `#pragma push_macro` and `#pragma pop_macro`. + +* cpp_reader Structure +The `cpp_reader` structure is the central object representing the full state of a preprocessor session in GCC's `libcpp`. It is passed to nearly every function across the subsystem and serves as the orchestration hub for lexing, macro expansion, buffer management, file inclusion, diagnostics, encoding conversion, and callback integration. + +** Purpose +`cpp_reader` encapsulates: +- The lexical stream and its position. +- The active and historical context stack. +- Preprocessor directives and include file tracking. +- Memory and token buffer management. +- Charset encoding conversions. +- Diagnostics and frontend callbacks. + +This makes it the definitive state carrier for a preprocessor run. + +** Usage in the Preprocessor + +The `cpp_reader` structure is instantiated once at the beginning of a preprocessing session via a function like `cpp_create_reader`. It is then initialized with options, encoding settings, and source input before being passed into most libcpp functions. It acts as the persistent environment for all operations and carries forward lexical position, macro state, memory buffers, and file context. + +Typical usage involves: + +1. **Initialization**: + - Create with `cpp_create_reader`. + - Configure options via `cpp_get_options`. + - Setup include paths and callbacks. + - Load source with `cpp_read_main_file`. + +2. **Tokenization Loop**: + - Repeatedly call `cpp_get_token(pfile)` to read tokens. + - Tokens are drawn from buffers, macro expansions, or virtual sources. + - `pfile->context` may be manipulated during macro expansions. + +3. **Directive and Macro Handling**: + - Functions like `_cpp_handle_directive`, `_cpp_create_definition`, or `_cpp_push_token_context` all mutate or inspect `pfile` to reflect state changes during preprocessing. + +4. **Finalization**: + - Clean up with `cpp_finish`, `cpp_destroy`, or related resource freeing logic. + +**Example** (simplified and partial): +```c +cpp_reader *pfile = cpp_create_reader(CLK_GNUC89, NULL, linemap); +cpp_get_options(pfile)->lang = CLK_GNUC89; +cpp_get_callbacks(pfile)->diagnostic = my_diagnostic_callback; +cpp_read_main_file(pfile, "myheader.h"); + +const cpp_token *tok; +while ((tok = cpp_get_token(pfile))->type != CPP_EOF) { + // Process token +} + +cpp_finish(pfile); + + +** Member-by-Member Overview + +- `cpp_buffer *buffer` :: Current input buffer, holding the text being preprocessed. +- `cpp_buffer *overlaid_buffer` :: A temporary buffer overlayed for special cases (e.g. `#include` insertions). +- `struct lexer_state state` :: Tracks current directive state (e.g., inside `#define`), comment retention, and conditional skipping. +- `class line_maps *line_table` :: Manages source line and file mapping for diagnostics and `__LINE__`/`__FILE__`. +- `location_t directive_line` :: Source location of the last encountered directive. +- `_cpp_buff *a_buff` :: Aligned buffer for allocations requiring native alignment (e.g., tokens). +- `_cpp_buff *u_buff` :: Unaligned buffer for simpler memory allocations. +- `_cpp_buff *free_buffs` :: Chain of reusable `_cpp_buff` structures. +- `cpp_context base_context` :: The base (top-level) token context. +- `cpp_context *context` :: Pointer to the current context on the expansion stack. +- `const struct directive *directive` :: Active directive, if in one. +- `cpp_token directive_result` :: The token result of directive evaluation. +- `location_t invocation_location` :: Location of a macro's invocation, used for expansion diagnostics. +- `cpp_hashnode *top_most_macro_node` :: Node of the macro currently being expanded at top level. +- `bool about_to_expand_macro_p` :: True if a macro is queued for expansion. +- `cpp_dir *quote_include` :: `#include "..."` search path. +- `cpp_dir *bracket_include` :: `#include <...>` search path. +- `cpp_dir no_search_path` :: A dummy path that disables search. +- `_cpp_file *all_files` :: List of all known files encountered. +- `_cpp_file *main_file` :: The initial input source file. +- `htab *file_hash`, `htab *dir_hash` :: Hash tables for file and directory caching. +- `file_hash_entry_pool *file_hash_entries` :: Pool allocator for file hash entries. +- `htab *nonexistent_file_hash` :: Cache of known missing files (for fast rejection). +- `obstack nonexistent_file_ob` :: Memory store for missing file data. +- `bool quote_ignores_source_dir` :: Controls whether to skip the current file's directory when resolving `#include "..."`. +- `bool seen_once_only` :: True if any `#pragma once` or `#import` was used. +- `const cpp_hashnode *mi_cmacro`, `mi_ind_cmacro` :: Cached macro guards used for multiple-include optimization. +- `bool mi_valid` :: Whether the multiple-inclusion optimization is currently valid. +- `cpp_token *cur_token` :: The current token being read or expanded. +- `tokenrun base_run, *cur_run` :: Token run (buffer) chain for macro-expanded tokens. +- `unsigned int lookaheads` :: Number of lookahead tokens buffered. +- `unsigned int keep_tokens` :: Whether to retain tokens for re-use or reprocessing. +- `unsigned char *macro_buffer` :: Buffer holding macro definition text for diagnostics or display. +- `unsigned int macro_buffer_len` :: Length of `macro_buffer`. +- `cset_converter narrow_cset_desc` :: Converter from source charset to execution charset (e.g. UTF-8). +- `cset_converter utf8_cset_desc`, `char16_cset_desc`, `char32_cset_desc`, `wide_cset_desc` :: Charset converters for UTF and wide characters. +- `const unsigned char *date`, `*time` :: Cached date/time strings used for `__DATE__` and `__TIME__`. +- `time_t time_stamp` :: Internal timestamp used for `__TIMESTAMP__`. +- `int time_stamp_kind` :: Metadata on how timestamp was acquired. +- `cpp_token avoid_paste`, `endarg` :: Special tokens used for controlling macro pasting behavior and argument marking. +- `mkdeps *deps` :: Opaque pointer to dependency tracking system (used for `-M` options). +- `obstack hash_ob`, `buffer_ob` :: Obstack memory pools for hash nodes and buffers, respectively. +- `pragma_entry *pragmas` :: List of user-defined or built-in pragma handlers. +- `cpp_callbacks cb` :: Callback structure for emitting diagnostics or user-visible events. +- `ht *hash_table` :: Identifier hash table. +- `op *op_stack`, `*op_limit` :: Stack used for evaluating constant expressions (e.g., in `#if`). +- `cpp_options opts` :: Holds all preprocessor option settings (e.g. pedantic mode, line directives). +- `spec_nodes spec_nodes` :: Special identifiers (`__VA_ARGS__`, `defined`, etc.). +- `bool our_hashtable` :: Whether this instance owns the hash table memory. +- `out { base, limit, cur, first_line }` :: Traditional output buffer. +- `saved_cur`, `saved_rlimit`, `saved_line_base` :: Saved pointers for buffer overlays. +- `cpp_savedstate *savedstate` :: Saved state for precompiled header support. +- `unsigned int counter` :: Value of `__COUNTER__` macro. +- `cpp_comment_table comments` :: Stores comments if `save_comments` is enabled. +- `def_pragma_macro *pushed_macros` :: List of macros pushed via `#pragma push_macro`. +- `location_t forced_token_location` :: Override location used for the next emitted token. +- `location_t main_loc` :: Marker for the location of the main file’s first line. + +** Summary +`cpp_reader` is a highly stateful construct. It abstracts preprocessing into a cooperative sequence of stages: file loading, lexical analysis, macro handling, directive parsing, and token expansion. Each of these is enabled or modulated via member fields. The design permits reuse of storage buffers, incremental context stacking, and precise location tracking across deeply nested macro expansions and file inclusions. +* Enums and Constants +** `include_type` +Represents how a file was included (e.g., `#include`, `#import`, `-include`, etc.). Used to manage buffer overlays and inclusion depth. + +** `context_tokens_kind` +Distinguishes how tokens are held in a `cpp_context`: direct, indirect, or extended. + +** Alignment Helpers +- `DEFAULT_ALIGNMENT`: Derived from struct alignment. +- `CPP_ALIGN2`, `CPP_ALIGN`: Ensure proper memory alignment. + +** Character Class Macros +- `is_idchar`, `is_numchar`, `is_hspace`, `is_vspace`, etc.: Type-safe wrappers over libc ctype behavior with preprocessor-specific adjustments. + +* Buffers and Memory +- `_cpp_get_buff`, `_cpp_release_buff`, `_cpp_extend_buff`, `_cpp_aligned_alloc`, `_cpp_unaligned_alloc`: Allocate and manage working buffers used during expansion. + +* Token and Macro Helpers +- `_cpp_mark_macro_used`: Marks a macro as used for diagnostics. +- `CPP_OPTION`, `CPP_BUFFER`, `CPP_INCREMENT_LINE`: Common access macros for reader state and buffer internals. +- `SEEN_EOL()`: Helper to check if the last token was EOF. + +* Function Declarations by File +** From macro.cc +- `_cpp_create_definition`, `_cpp_new_macro`, `_cpp_notify_macro_use`, `_cpp_push_token_context`, etc.: Manage macro creation, expansion, and context. + +** From directives.cc +- `_cpp_define_builtin`, `_cpp_handle_directive`, `_cpp_do__Pragma`, etc.: Directive parsing and #pragma handlers. + +** From files.cc +- `_cpp_find_file`, `_cpp_stack_include`, `_cpp_pop_file_buffer`: File inclusion management and include guards. + +** From lex.cc +- `_cpp_lex_token`, `_cpp_temp_token`, `_cpp_equiv_tokens`: Token lexing and temporary token generation. + +** From expr.cc +- `_cpp_parse_expr`, `_cpp_expand_op_stack`: Expression parsing in `#if`/`#elif`. + +** From charset.cc +- `_cpp_valid_utf8`, `_cpp_convert_input`, `_cpp_destroy_iconv`: Character encoding conversion routines. + +** From init.cc +- `_cpp_restore_special_builtin`, `cpp_named_operator2name`: Initialization helpers for macro state. + +** From identifiers.cc +- `_cpp_init_hashtable`, `_cpp_destroy_hashtable`: Identifier table setup and teardown. + +* Encoding and Normalization +** `normalize_state` +Tracks normalization level and combining characters for UCN validation and identifier processing. + +** `cset_converter` +Holds state for iconv-based charset conversion. Used for input and output charset normalization. + +* Accessor Inline Functions +- `_cpp_in_system_header`, `_cpp_in_main_source_file`, `_cpp_defined_macro_p`: Context-sensitive accessors. +- `ustrcmp`, `ustrlen`, `uxstrdup`, `ufputs`, etc.: UTF-aware string handling routines. + +* Diagnostic Integration +** `encoding_rich_location` +Subclass of `rich_location` that forces encoding escape visibility for diagnostics. Constructed from `cpp_reader`. + +* Notes +- This file is not compiled standalone but included in many CPP components. +- It contains bridge-level API elements that link between token processing, buffer management, and frontend logic. +- Care must be taken when editing alignment or buffer routines as they affect all downstream expansion logic. + +* TODO +- Document how iconv fallback works when `HAVE_ICONV` is not defined. +- Clarify lifecycle of pushed macro contexts during nested `#pragma push_macro` chains. +- Integrate doc with `macro.cc` and `lex.cc` references for cross-module tracing. diff --git "a/document\360\237\226\211/source/libcpp_h.org" "b/document\360\237\226\211/source/libcpp_h.org" new file mode 100644 index 0000000..e69de29 diff --git "a/document\360\237\226\211/source/macro_cc.org" "b/document\360\237\226\211/source/macro_cc.org" new file mode 100644 index 0000000..53d8b65 --- /dev/null +++ "b/document\360\237\226\211/source/macro_cc.org" @@ -0,0 +1,83 @@ +#+TITLE: macro.cc - Documentation Reference +#+AUTHOR: Thomas Walker Lynch & Caelestis Index +#+DESCRIPTION: High-level architectural partitioning of cpp (GCC 12.x) +#+FILETAGS: cpp preprocessor architecture gcc +#+OPTIONS: toc:nil + +* Overview +This file implements the core logic for macro parsing, macro definition, expansion, and deferred/lazy evaluation within the C preprocessor (CPP) subsystem in GCC's `libcpp`. It complements the infrastructure declared in `libcpp.h` and utilizes various helpers from supporting headers. + +* Included Headers and Their Purpose +- `config.h`: Compiler configuration macros. +- `system.h`: GCC-wide portability and utility macros. +- `intl.h`: Localization support. +- `cpplib.h`: Core interface for the C preprocessor. +- `internal.h`: Internal-only structures and definitions for the preprocessor. +- `macros.h`: Macro parsing and storage structures. +- `trad.h`: Traditional mode logic. +- `mkdeps.h`: Dependency output handling. +- `diagnostic-core.h`: Diagnostic emission interfaces. +- `cpp-id-data.h`: Identifier information, e.g. for argument naming. + +* Major Data Structures Used + +- `cpp_reader` (from `cpplib.h`): The global preprocessor context. +- `cpp_hashnode` (from `cpplib.h`): Represents identifiers, including macro definitions. +- `cpp_macro` (from `macros.h`): Stores a single macro definition, either traditional or ISO. +- `macro_arg` (from `macros.h`): Represents a single argument to a function-like macro. +- `macro_context` (internal): Used for tracking extended macro expansion location information. +- `_cpp_buff` (from `internal.h`): Temporary token or string storage buffer. +- `cpp_token` (from `cpplib.h`): Represents a preprocessor token. +- `cpp_string` (from `cpplib.h`): String-like wrapper for character sequences. + +* Functional Groups (Grouped per libcpp.h Theme) + +*** Token Context Management +- `_cpp_push_token_context`: Pushes a direct token sequence as context. +- `push_ptoken_context`: Pushes indirect token sequence. +- `push_extended_tokens_context`: Pushes context with virtual locations. +- `_cpp_pop_context`: Pops current macro or token context. + +*** Argument Expansion and Memory +- `expand_arg`: Expands a macro argument by recursively evaluating tokens. +- `alloc_expanded_arg_mem`: Allocates buffer space for an argument. +- `ensure_expanded_arg_room`: Doubles expansion buffer when needed. +- `set_arg_token` (external): Inserts expanded tokens into an argument. + +*** Macro Definition and Redefinition +- `_cpp_create_definition`: Top-level interface to create and store macro definition. +- `create_iso_definition`: Parses macro arguments and expansion tokens. +- `_cpp_save_parameter`: Saves a named parameter for a function-like macro. +- `_cpp_unsave_parameters`: Restores hashnodes after failed macro parse. +- `warn_of_redefinition`: Determines if a redefinition should trigger a warning. +- `cpp_compare_macros`: Compares two macros for semantic equality. + +*** Macro Instantiation and Lazy Expansion +- `get_deferred_or_lazy_macro`: Retrieves or forces realization of a deferred or lazy macro. +- `cpp_get_deferred_macro`: Resolves a deferred macro. +- `cpp_define_lazily`: Marks a macro for delayed definition. +- `_cpp_notify_macro_use`: Central notification hook that tracks macro use. + +*** Macro Definition Representation +- `cpp_macro_definition`: Renders a macro definition as a string. +- `cpp_macro_definition(pfile, node, macro)`: Core form with macro pointer. + +*** Lexing Helpers and Traditional Compatibility +- `lex_expansion_token`: Lexes one token in a macro body. +- `check_trad_stringification`: Warns if argument appears stringified in traditional C. +- `_cpp_new_macro`: Allocates and initializes a `cpp_macro`. + +* Integration with Other Subsystems +- Works closely with: `lex.c`, `directives.cc`, and `internal.c`. +- Interfaces with `linemap` for virtual location computation. +- Supports both ISO and traditional C macro handling. + +* Notes +- Token pasting (`##`) is carefully constrained per ISO rules. +- Parameter and macro use is tracked for diagnostics and DWARF output. +- Extra tokens such as padding and stringification markers carry encoded flags. + +* TODO +- Document edge cases and non-ISO behaviors (e.g., bare ellipsis). +- Link to relevant `libcpp.h` macro flags and diagnostic utilities. +- Cross-reference context expansion rules with `cpp_get_token_1`. diff --git "a/document\360\237\226\211/source/macro_registration.org" "b/document\360\237\226\211/source/macro_registration.org" new file mode 100644 index 0000000..56bf752 --- /dev/null +++ "b/document\360\237\226\211/source/macro_registration.org" @@ -0,0 +1,150 @@ +#+TITLE: Macro Symbol Registration in GCC 12 libcpp +#+AUTHOR: Caelestis Index +#+DESCRIPTION: Full lifecycle of defining and registering a macro in GCC's C preprocessor +#+OPTIONS: toc:nil +#+FILETAGS: gcc libcpp macro cpp_hashnode + +* Overview +This document explains the full lifecycle for defining a macro in GCC 12.x's =libcpp= preprocessor. It traces the required steps from token parsing through symbol table registration, highlighting where and how macro definitions become visible to the preprocessor engine. + +* 1. Obtaining the Macro Name’s Hash Node + +In =libcpp=, all identifiers — including macro names — are interned in a symbol table as =cpp_hashnode= entries. When the lexer emits a =CPP_NAME= token, it automatically fills: + +#+BEGIN_SRC c +token->val.node.node // type: cpp_hashnode * +#+END_SRC + +If the macro name comes from parsed input (e.g. `#assign` or `#define`), this node is already in the symbol table — no need to call =cpp_lookup= again. + +If you're defining a macro from a raw string (not a parsed token), you *would* use: + +#+BEGIN_SRC c +cpp_lookup(pfile ,name ,len); +#+END_SRC + +Note: =cpp_lookup= both interns new identifiers and retrieves existing ones. + +* 2. Creating and Populating a cpp_macro Object + +GCC uses a =cpp_macro= struct to hold the macro’s definition: number of parameters, replacement tokens, flags, etc. + +Allocation is done with: + +#+BEGIN_SRC c +cpp_macro *macro = _cpp_new_macro( + pfile, + cmk_macro, + _cpp_reserve_room(pfile ,0 ,sizeof(cpp_macro)) +); +#+END_SRC + +After that, populate its fields: + +#+BEGIN_SRC c +macro->fun_like = 0; +macro->paramc = 0; +macro->variadic = 0; +macro->count = 1; +macro->used = 1; + +cpp_token *tok = ¯o->exp.tokens[0]; +tok->type = CPP_NUMBER; +tok->val.str.text = (const unsigned char *) "42"; +tok->val.str.len = 2; +tok->flags = 0; +#+END_SRC + +Note: These macros are obstack-allocated; you don't free them manually. + +* 3. Handling Redefinitions (Optional, but Expected) + +If the symbol already has a macro: + +#+BEGIN_SRC c +if( cpp_macro_p(node) ) + warn_of_redefinition(pfile ,node ,macro); +#+END_SRC + +GCC allows redefinition only if the new macro is *identical*. If not, it issues a pedantic warning and overwrites the old definition. + +To remove the previous macro: + +#+BEGIN_SRC c +_cpp_free_definition(node); +#+END_SRC + +This clears the macro without deallocating it (obstack). + +* 4. Installing the Macro in the Symbol Table + +The macro is made active by assigning it to the symbol table: + +#+BEGIN_SRC c +node->type = NT_USER_MACRO; +node->value.macro = macro; +#+END_SRC + +This effectively *registers* the macro for expansion. + +There is no separate "symbol table insertion" step — the hash node was already in the table. + +GCC may also set flags: + +- =NODE_WARN= → warn if redefining built-in +- =NODE_CONDITIONAL= → cleared when explicitly defined + +* 5. Finalization Steps + +Some final steps after macro insertion: + +- Mark it used (optional): + + #+BEGIN_SRC c + _cpp_mark_macro_used(node); + #+END_SRC + +- Emit a diagnostic: + + #+BEGIN_SRC c + cpp_warning(pfile ,CPP_W_NONE ,"Assigned macro %s as 42" ,NODE_NAME(node)); + #+END_SRC + +- Clear the =NODE_USED= flag to reset unused-macro warnings: + + #+BEGIN_SRC c + node->flags &= ~NODE_USED; + #+END_SRC + +* Summary of Required Steps + +Here is the complete, valid sequence to register a macro manually: + +#+BEGIN_SRC c +cpp_token *name_token = assign_name_argument(pfile); +cpp_hashnode *node = name_token->val.node.node; + +cpp_macro *macro = _cpp_new_macro(...); // allocate and populate + +// fill token replacement list... +macro->count = 1; +macro->exp.tokens[0] = ...; + +node->type = NT_USER_MACRO; +node->value.macro = macro; +_cpp_mark_macro_used(node); +#+END_SRC + +That is sufficient to define and register a macro. =NODE_NAME(node)= is useful for diagnostics, but not required for registration. + +* Notes + +- If you already have a =cpp_token= from parsing, the hash node is *already interned*. +- Macros must be registered by setting =node->type= and =node->value.macro=. +- Redefinitions are allowed only if semantically identical unless explicitly undefined. +- No extra insertion or lookup step is needed unless building from raw text. + +* References +- =directives.cc= → =do_define= and redefinition checks +- =macro.cc= → =create_iso_definition= and macro assembly +- =cpplib.h= → =cpp_hashnode=, =cpp_macro=, enum flags diff --git "a/script_gcc_min-12\360\237\226\211/directives.cc" "b/script_gcc_min-12\360\237\226\211/directives.cc" new file mode 100644 index 0000000..8ee29b3 --- /dev/null +++ "b/script_gcc_min-12\360\237\226\211/directives.cc" @@ -0,0 +1,3058 @@ +/* CPP Library. (Directive handling.) + Copyright (C) 1986-2022 Free Software Foundation, Inc. + Contributed by Per Bothner, 1994-95. + Based on CCCP program by Paul Rubin, June 1986 + Adapted to ANSI C, Richard Stallman, Jan 1987 + +This program is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 3, or (at your option) any +later version. + +This program is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; see the file COPYING3. If not see +. */ + +#include "config.h" +#include "system.h" +#include "cpplib.h" +#include "internal.h" +#include "mkdeps.h" +#include "obstack.h" + +/* Stack of conditionals currently in progress + (including both successful and failing conditionals). */ +struct if_stack +{ + struct if_stack *next; + location_t line; /* Line where condition started. */ + const cpp_hashnode *mi_cmacro;/* macro name for #ifndef around entire file */ + bool skip_elses; /* Can future #else / #elif be skipped? */ + bool was_skipping; /* If were skipping on entry. */ + int type; /* Most recent conditional for diagnostics. */ +}; + +/* Contains a registered pragma or pragma namespace. */ +typedef void (*pragma_cb) (cpp_reader *); +struct pragma_entry +{ + struct pragma_entry *next; + const cpp_hashnode *pragma; /* Name and length. */ + bool is_nspace; + bool is_internal; + bool is_deferred; + bool allow_expansion; + union { + pragma_cb handler; + struct pragma_entry *space; + unsigned int ident; + } u; +}; + +/* Values for the origin field of struct directive. KANDR directives + come from traditional (K&R) C. STDC89 directives come from the + 1989 C standard. STDC2X directives come from the C2X standard. EXTENSION + directives are extensions. */ +#define KANDR 0 +#define STDC89 1 +#define STDC2X 2 +#define EXTENSION 3 + +/* Values for the flags field of struct directive. COND indicates a + conditional; IF_COND an opening conditional. INCL means to treat + "..." and <...> as q-char and h-char sequences respectively. IN_I + means this directive should be handled even if -fpreprocessed is in + effect (these are the directives with callback hooks). + + EXPAND is set on directives that are always macro-expanded. + + ELIFDEF is set on directives that are only handled for standards with the + #elifdef / #elifndef feature. */ +#define COND (1 << 0) +#define IF_COND (1 << 1) +#define INCL (1 << 2) +#define IN_I (1 << 3) +#define EXPAND (1 << 4) +#define DEPRECATED (1 << 5) +#define ELIFDEF (1 << 6) + +/* Defines one #-directive, including how to handle it. */ +typedef void (*directive_handler) (cpp_reader *); +typedef struct directive directive; +struct directive +{ + directive_handler handler; /* Function to handle directive. */ + const uchar *name; /* Name of directive. */ + unsigned short length; /* Length of name. */ + unsigned char origin; /* Origin of directive. */ + unsigned char flags; /* Flags describing this directive. */ +}; + +/* Forward declarations. */ + +static void skip_rest_of_line (cpp_reader *); +static void check_eol (cpp_reader *, bool); +static void start_directive (cpp_reader *); +static void prepare_directive_trad (cpp_reader *); +static void end_directive (cpp_reader *, int); +static void directive_diagnostics (cpp_reader *, const directive *, int); +static void run_directive (cpp_reader *, int, const char *, size_t); +static char *glue_header_name (cpp_reader *); +static const char *parse_include (cpp_reader *, int *, const cpp_token ***, + location_t *); +static void push_conditional (cpp_reader *, int, int, const cpp_hashnode *); +static unsigned int read_flag (cpp_reader *, unsigned int); +static bool strtolinenum (const uchar *, size_t, linenum_type *, bool *); +static void do_diagnostic (cpp_reader *, enum cpp_diagnostic_level code, + enum cpp_warning_reason reason, int); +static cpp_hashnode *lex_macro_node (cpp_reader *, bool); +static int undefine_macros (cpp_reader *, cpp_hashnode *, void *); +static void do_include_common (cpp_reader *, enum include_type); +static struct pragma_entry *lookup_pragma_entry (struct pragma_entry *, + const cpp_hashnode *); +static int count_registered_pragmas (struct pragma_entry *); +static char ** save_registered_pragmas (struct pragma_entry *, char **); +static char ** restore_registered_pragmas (cpp_reader *, struct pragma_entry *, + char **); +static void do_pragma_once (cpp_reader *); +static void do_pragma_poison (cpp_reader *); +static void do_pragma_system_header (cpp_reader *); +static void do_pragma_dependency (cpp_reader *); +static void do_pragma_warning_or_error (cpp_reader *, bool error); +static void do_pragma_warning (cpp_reader *); +static void do_pragma_error (cpp_reader *); +static void do_linemarker (cpp_reader *); +static const cpp_token *get_token_no_padding (cpp_reader *); +static const cpp_token *get__Pragma_string (cpp_reader *); +static void destringize_and_run (cpp_reader *, const cpp_string *, + location_t); +static bool parse_answer (cpp_reader *, int, location_t, cpp_macro **); +static cpp_hashnode *parse_assertion (cpp_reader *, int, cpp_macro **); +static cpp_macro **find_answer (cpp_hashnode *, const cpp_macro *); +static void handle_assertion (cpp_reader *, const char *, int); +static void do_pragma_push_macro (cpp_reader *); +static void do_pragma_pop_macro (cpp_reader *); +static void cpp_pop_definition (cpp_reader *, struct def_pragma_macro *); + +/* This is the table of directive handlers. All extensions other than + #warning, #include_next, and #import are deprecated. The name is + where the extension appears to have come from. */ + +#define DIRECTIVE_TABLE \ + D(define, T_DEFINE = 0, KANDR, IN_I) \ + D(include, T_INCLUDE, KANDR, INCL | EXPAND) \ + D(endif, T_ENDIF, KANDR, COND) \ + D(ifdef, T_IFDEF, KANDR, COND | IF_COND) \ + D(if, T_IF, KANDR, COND | IF_COND | EXPAND) \ + D(else, T_ELSE, KANDR, COND) \ + D(ifndef, T_IFNDEF, KANDR, COND | IF_COND) \ + D(undef, T_UNDEF, KANDR, IN_I) \ + D(line, T_LINE, KANDR, EXPAND) \ + D(elif, T_ELIF, STDC89, COND | EXPAND) \ + D(elifdef, T_ELIFDEF, STDC2X, COND | ELIFDEF) \ + D(elifndef, T_ELIFNDEF, STDC2X, COND | ELIFDEF) \ + D(error, T_ERROR, STDC89, 0) \ + D(pragma, T_PRAGMA, STDC89, IN_I) \ + D(warning, T_WARNING, EXTENSION, 0) \ + D(include_next, T_INCLUDE_NEXT, EXTENSION, INCL | EXPAND) \ + D(ident, T_IDENT, EXTENSION, IN_I) \ + D(import, T_IMPORT, EXTENSION, INCL | EXPAND) /* ObjC */ \ + D(assert, T_ASSERT, EXTENSION, DEPRECATED) /* SVR4 */ \ + D(unassert, T_UNASSERT, EXTENSION, DEPRECATED) /* SVR4 */ \ + D(sccs, T_SCCS, EXTENSION, IN_I) /* SVR4? */ \ + D(assign, T_ASSIGN, EXTENSION, IN_I) + +/* #sccs is synonymous with #ident. */ +#define do_sccs do_ident + +/* Use the table to generate a series of prototypes, an enum for the + directive names, and an array of directive handlers. */ + +#define D(name, t, o, f) static void do_##name (cpp_reader *); +DIRECTIVE_TABLE +#undef D + +#define D(n, tag, o, f) tag, +enum +{ + DIRECTIVE_TABLE + N_DIRECTIVES +}; +#undef D + +#define D(name, t, origin, flags) \ +{ do_##name, (const uchar *) #name, \ + sizeof #name - 1, origin, flags }, +static const directive dtable[] = +{ +DIRECTIVE_TABLE +}; +#undef D + +/* A NULL-terminated array of directive names for use + when suggesting corrections for misspelled directives. */ +#define D(name, t, origin, flags) #name, +static const char * const directive_names[] = { +DIRECTIVE_TABLE + NULL +}; +#undef D + +#undef DIRECTIVE_TABLE + +/* Wrapper struct directive for linemarkers. + The origin is more or less true - the original K+R cpp + did use this notation in its preprocessed output. */ +static const directive linemarker_dir = +{ + do_linemarker, UC"#", 1, KANDR, IN_I +}; + +/* Skip any remaining tokens in a directive. */ +static void +skip_rest_of_line (cpp_reader *pfile) +{ + /* Discard all stacked contexts. */ + while (pfile->context->prev) + _cpp_pop_context (pfile); + + /* Sweep up all tokens remaining on the line. */ + if (! SEEN_EOL ()) + while (_cpp_lex_token (pfile)->type != CPP_EOF) + ; +} + +/* Helper function for check_oel. */ + +static void +check_eol_1 (cpp_reader *pfile, bool expand, enum cpp_warning_reason reason) +{ + if (! SEEN_EOL () && (expand + ? cpp_get_token (pfile) + : _cpp_lex_token (pfile))->type != CPP_EOF) + cpp_pedwarning (pfile, reason, "extra tokens at end of #%s directive", + pfile->directive->name); +} + +/* Variant of check_eol used for Wendif-labels warnings. */ + +static void +check_eol_endif_labels (cpp_reader *pfile) +{ + check_eol_1 (pfile, false, CPP_W_ENDIF_LABELS); +} + +/* Ensure there are no stray tokens at the end of a directive. If + EXPAND is true, tokens macro-expanding to nothing are allowed. */ + +static void +check_eol (cpp_reader *pfile, bool expand) +{ + check_eol_1 (pfile, expand, CPP_W_NONE); +} + +/* Ensure there are no stray tokens other than comments at the end of + a directive, and gather the comments. */ +static const cpp_token ** +check_eol_return_comments (cpp_reader *pfile) +{ + size_t c; + size_t capacity = 8; + const cpp_token **buf; + + buf = XNEWVEC (const cpp_token *, capacity); + c = 0; + if (! SEEN_EOL ()) + { + while (1) + { + const cpp_token *tok; + + tok = _cpp_lex_token (pfile); + if (tok->type == CPP_EOF) + break; + if (tok->type != CPP_COMMENT) + cpp_error (pfile, CPP_DL_PEDWARN, + "extra tokens at end of #%s directive", + pfile->directive->name); + else + { + if (c + 1 >= capacity) + { + capacity *= 2; + buf = XRESIZEVEC (const cpp_token *, buf, capacity); + } + buf[c] = tok; + ++c; + } + } + } + buf[c] = NULL; + return buf; +} + +/* Called when entering a directive, _Pragma or command-line directive. */ +static void +start_directive (cpp_reader *pfile) +{ + /* Setup in-directive state. */ + pfile->state.in_directive = 1; + pfile->state.save_comments = 0; + pfile->directive_result.type = CPP_PADDING; + + /* Some handlers need the position of the # for diagnostics. */ + pfile->directive_line = pfile->line_table->highest_line; +} + +/* Called when leaving a directive, _Pragma or command-line directive. */ +static void +end_directive (cpp_reader *pfile, int skip_line) +{ + if (CPP_OPTION (pfile, traditional)) + { + /* Revert change of prepare_directive_trad. */ + if (!pfile->state.in_deferred_pragma) + pfile->state.prevent_expansion--; + + if (pfile->directive != &dtable[T_DEFINE]) + _cpp_remove_overlay (pfile); + } + else if (pfile->state.in_deferred_pragma) + ; + /* We don't skip for an assembler #. */ + else if (skip_line) + { + skip_rest_of_line (pfile); + if (!pfile->keep_tokens) + { + pfile->cur_run = &pfile->base_run; + pfile->cur_token = pfile->base_run.base; + } + } + + /* Restore state. */ + pfile->state.save_comments = ! CPP_OPTION (pfile, discard_comments); + pfile->state.in_directive = 0; + pfile->state.in_expression = 0; + pfile->state.angled_headers = 0; + pfile->directive = 0; +} + +/* Prepare to handle the directive in pfile->directive. */ +static void +prepare_directive_trad (cpp_reader *pfile) +{ + if (pfile->directive != &dtable[T_DEFINE]) + { + bool no_expand = (pfile->directive + && ! (pfile->directive->flags & EXPAND)); + bool was_skipping = pfile->state.skipping; + + pfile->state.in_expression = (pfile->directive == &dtable[T_IF] + || pfile->directive == &dtable[T_ELIF]); + if (pfile->state.in_expression) + pfile->state.skipping = false; + + if (no_expand) + pfile->state.prevent_expansion++; + _cpp_scan_out_logical_line (pfile, NULL, false); + if (no_expand) + pfile->state.prevent_expansion--; + + pfile->state.skipping = was_skipping; + _cpp_overlay_buffer (pfile, pfile->out.base, + pfile->out.cur - pfile->out.base); + } + + /* Stop ISO C from expanding anything. */ + pfile->state.prevent_expansion++; +} + +/* Output diagnostics for a directive DIR. INDENTED is nonzero if + the '#' was indented. */ +static void +directive_diagnostics (cpp_reader *pfile, const directive *dir, int indented) +{ + /* Issue -pedantic or deprecated warnings for extensions. We let + -pedantic take precedence if both are applicable. */ + if (! pfile->state.skipping) + { + if (dir->origin == EXTENSION + && !(dir == &dtable[T_IMPORT] && CPP_OPTION (pfile, objc)) + && CPP_PEDANTIC (pfile)) + cpp_error (pfile, CPP_DL_PEDWARN, "#%s is a GCC extension", dir->name); + else if (((dir->flags & DEPRECATED) != 0 + || (dir == &dtable[T_IMPORT] && !CPP_OPTION (pfile, objc))) + && CPP_OPTION (pfile, cpp_warn_deprecated)) + cpp_warning (pfile, CPP_W_DEPRECATED, + "#%s is a deprecated GCC extension", dir->name); + } + + /* Traditionally, a directive is ignored unless its # is in + column 1. Therefore in code intended to work with K+R + compilers, directives added by C89 must have their # + indented, and directives present in traditional C must not. + This is true even of directives in skipped conditional + blocks. #elif cannot be used at all. */ + if (CPP_WTRADITIONAL (pfile)) + { + if (dir == &dtable[T_ELIF]) + cpp_warning (pfile, CPP_W_TRADITIONAL, + "suggest not using #elif in traditional C"); + else if (indented && dir->origin == KANDR) + cpp_warning (pfile, CPP_W_TRADITIONAL, + "traditional C ignores #%s with the # indented", + dir->name); + else if (!indented && dir->origin != KANDR) + cpp_warning (pfile, CPP_W_TRADITIONAL, + "suggest hiding #%s from traditional C with an indented #", + dir->name); + } +} + +/* Check if we have a known directive. INDENTED is true if the + '#' of the directive was indented. This function is in this file + to save unnecessarily exporting dtable etc. to lex.cc. Returns + nonzero if the line of tokens has been handled, zero if we should + continue processing the line. */ +int +_cpp_handle_directive (cpp_reader *pfile, bool indented) +{ + const directive *dir = 0; + const cpp_token *dname; + bool was_parsing_args = pfile->state.parsing_args; + bool was_discarding_output = pfile->state.discarding_output; + int skip = 1; + + if (was_discarding_output) + pfile->state.prevent_expansion = 0; + + if (was_parsing_args) + { + if (CPP_OPTION (pfile, cpp_pedantic)) + cpp_error (pfile, CPP_DL_PEDWARN, + "embedding a directive within macro arguments is not portable"); + pfile->state.parsing_args = 0; + pfile->state.prevent_expansion = 0; + } + start_directive (pfile); + dname = _cpp_lex_token (pfile); + + if (dname->type == CPP_NAME) + { + if (dname->val.node.node->is_directive) + { + dir = &dtable[dname->val.node.node->directive_index]; + if ((dir->flags & ELIFDEF) + && !CPP_OPTION (pfile, elifdef) + /* For -std=gnu* modes elifdef is supported with + a pedwarn if pedantic. */ + && CPP_OPTION (pfile, std)) + dir = 0; + } + } + /* We do not recognize the # followed by a number extension in + assembler code. */ + else if (dname->type == CPP_NUMBER && CPP_OPTION (pfile, lang) != CLK_ASM) + { + dir = &linemarker_dir; + if (CPP_PEDANTIC (pfile) && ! CPP_OPTION (pfile, preprocessed) + && ! pfile->state.skipping) + cpp_error (pfile, CPP_DL_PEDWARN, + "style of line directive is a GCC extension"); + } + + if (dir) + { + /* If we have a directive that is not an opening conditional, + invalidate any control macro. */ + if (! (dir->flags & IF_COND)) + pfile->mi_valid = false; + + /* Kluge alert. In order to be sure that code like this + + #define HASH # + HASH define foo bar + + does not cause '#define foo bar' to get executed when + compiled with -save-temps, we recognize directives in + -fpreprocessed mode only if the # is in column 1. macro.cc + puts a space in front of any '#' at the start of a macro. + + We exclude the -fdirectives-only case because macro expansion + has not been performed yet, and block comments can cause spaces + to precede the directive. */ + if (CPP_OPTION (pfile, preprocessed) + && !CPP_OPTION (pfile, directives_only) + && (indented || !(dir->flags & IN_I))) + { + skip = 0; + dir = 0; + } + else + { + /* In failed conditional groups, all non-conditional + directives are ignored. Before doing that, whether + skipping or not, we should lex angle-bracketed headers + correctly, and maybe output some diagnostics. */ + pfile->state.angled_headers = dir->flags & INCL; + pfile->state.directive_wants_padding = dir->flags & INCL; + if (! CPP_OPTION (pfile, preprocessed)) + directive_diagnostics (pfile, dir, indented); + if (pfile->state.skipping && !(dir->flags & COND)) + dir = 0; + } + } + else if (dname->type == CPP_EOF) + ; /* CPP_EOF is the "null directive". */ + else + { + /* An unknown directive. Don't complain about it in assembly + source: we don't know where the comments are, and # may + introduce assembler pseudo-ops. Don't complain about invalid + directives in skipped conditional groups (6.10 p4). */ + if (CPP_OPTION (pfile, lang) == CLK_ASM) + skip = 0; + else if (!pfile->state.skipping) + { + const char *unrecognized + = (const char *)cpp_token_as_text (pfile, dname); + const char *hint = NULL; + + /* Call back into gcc to get a spelling suggestion. Ideally + we'd just use best_match from gcc/spellcheck.h (and filter + out the uncommon directives), but that requires moving it + to a support library. */ + if (pfile->cb.get_suggestion) + hint = pfile->cb.get_suggestion (pfile, unrecognized, + directive_names); + + if (hint) + { + rich_location richloc (pfile->line_table, dname->src_loc); + source_range misspelled_token_range + = get_range_from_loc (pfile->line_table, dname->src_loc); + richloc.add_fixit_replace (misspelled_token_range, hint); + cpp_error_at (pfile, CPP_DL_ERROR, &richloc, + "invalid preprocessing directive #%s;" + " did you mean #%s?", + unrecognized, hint); + } + else + cpp_error (pfile, CPP_DL_ERROR, + "invalid preprocessing directive #%s", + unrecognized); + } + } + + pfile->directive = dir; + if (CPP_OPTION (pfile, traditional)) + prepare_directive_trad (pfile); + + if (dir) + pfile->directive->handler (pfile); + else if (skip == 0) + _cpp_backup_tokens (pfile, 1); + + end_directive (pfile, skip); + if (was_parsing_args && !pfile->state.in_deferred_pragma) + { + /* Restore state when within macro args. */ + pfile->state.parsing_args = 2; + pfile->state.prevent_expansion = 1; + } + if (was_discarding_output) + pfile->state.prevent_expansion = 1; + return skip; +} + +/* Directive handler wrapper used by the command line option + processor. BUF is \n terminated. */ +static void +run_directive (cpp_reader *pfile, int dir_no, const char *buf, size_t count) +{ + cpp_push_buffer (pfile, (const uchar *) buf, count, + /* from_stage3 */ true); + start_directive (pfile); + + /* This is a short-term fix to prevent a leading '#' being + interpreted as a directive. */ + _cpp_clean_line (pfile); + + pfile->directive = &dtable[dir_no]; + if (CPP_OPTION (pfile, traditional)) + prepare_directive_trad (pfile); + pfile->directive->handler (pfile); + end_directive (pfile, 1); + _cpp_pop_buffer (pfile); +} + +/* Checks for validity the macro name in #define, #undef, #ifdef and + #ifndef directives. IS_DEF_OR_UNDEF is true if this call is + processing a #define or #undefine directive, and false + otherwise. */ +static cpp_hashnode * +lex_macro_node (cpp_reader *pfile, bool is_def_or_undef) +{ + const cpp_token *token = _cpp_lex_token (pfile); + + /* The token immediately after #define must be an identifier. That + identifier may not be "defined", per C99 6.10.8p4. + In C++, it may not be any of the "named operators" either, + per C++98 [lex.digraph], [lex.key]. + Finally, the identifier may not have been poisoned. (In that case + the lexer has issued the error message for us.) */ + + if (token->type == CPP_NAME) + { + cpp_hashnode *node = token->val.node.node; + + if (is_def_or_undef + && node == pfile->spec_nodes.n_defined) + cpp_error (pfile, CPP_DL_ERROR, + "\"%s\" cannot be used as a macro name", + NODE_NAME (node)); + else if (! (node->flags & NODE_POISONED)) + return node; + } + else if (token->flags & NAMED_OP) + cpp_error (pfile, CPP_DL_ERROR, + "\"%s\" cannot be used as a macro name as it is an operator in C++", + NODE_NAME (token->val.node.node)); + else if (token->type == CPP_EOF) + cpp_error (pfile, CPP_DL_ERROR, "no macro name given in #%s directive", + pfile->directive->name); + else + cpp_error (pfile, CPP_DL_ERROR, "macro names must be identifiers"); + + return NULL; +} + +/* Process a #define directive. Most work is done in macro.cc. */ +static void +do_define (cpp_reader *pfile) +{ + cpp_hashnode *node = lex_macro_node (pfile, true); + + if (node) + { + /* If we have been requested to expand comments into macros, + then re-enable saving of comments. */ + pfile->state.save_comments = + ! CPP_OPTION (pfile, discard_comments_in_macro_exp); + + if (pfile->cb.before_define) + pfile->cb.before_define (pfile); + + if (_cpp_create_definition (pfile, node)) + if (pfile->cb.define) + pfile->cb.define (pfile, pfile->directive_line, node); + + node->flags &= ~NODE_USED; + } +} + +/* Handle #undef. Mark the identifier NT_VOID in the hash table. */ +static void +do_undef (cpp_reader *pfile) +{ + cpp_hashnode *node = lex_macro_node (pfile, true); + + if (node) + { + if (pfile->cb.before_define) + pfile->cb.before_define (pfile); + + if (pfile->cb.undef) + pfile->cb.undef (pfile, pfile->directive_line, node); + + /* 6.10.3.5 paragraph 2: [#undef] is ignored if the specified + identifier is not currently defined as a macro name. */ + if (cpp_macro_p (node)) + { + if (node->flags & NODE_WARN) + cpp_error (pfile, CPP_DL_WARNING, + "undefining \"%s\"", NODE_NAME (node)); + else if (cpp_builtin_macro_p (node) + && CPP_OPTION (pfile, warn_builtin_macro_redefined)) + cpp_warning_with_line (pfile, CPP_W_BUILTIN_MACRO_REDEFINED, + pfile->directive_line, 0, + "undefining \"%s\"", NODE_NAME (node)); + + if (node->value.macro + && CPP_OPTION (pfile, warn_unused_macros)) + _cpp_warn_if_unused_macro (pfile, node, NULL); + + _cpp_free_definition (node); + } + } + + check_eol (pfile, false); +} + +/* Undefine a single macro/assertion/whatever. */ + +static int +undefine_macros (cpp_reader *pfile ATTRIBUTE_UNUSED, cpp_hashnode *h, + void *data_p ATTRIBUTE_UNUSED) +{ + /* Body of _cpp_free_definition inlined here for speed. + Macros and assertions no longer have anything to free. */ + h->type = NT_VOID; + h->value.answers = NULL; + h->flags &= ~(NODE_POISONED|NODE_DISABLED|NODE_USED); + return 1; +} + +/* Undefine all macros and assertions. */ + +void +cpp_undef_all (cpp_reader *pfile) +{ + cpp_forall_identifiers (pfile, undefine_macros, NULL); +} + + +/* Helper routine used by parse_include. Reinterpret the current line + as an h-char-sequence (< ... >); we are looking at the first token + after the <. Returns a malloced filename. */ +static char * +glue_header_name (cpp_reader *pfile) +{ + const cpp_token *token; + char *buffer; + size_t len, total_len = 0, capacity = 1024; + + /* To avoid lexed tokens overwriting our glued name, we can only + allocate from the string pool once we've lexed everything. */ + buffer = XNEWVEC (char, capacity); + for (;;) + { + token = get_token_no_padding (pfile); + + if (token->type == CPP_GREATER) + break; + if (token->type == CPP_EOF) + { + cpp_error (pfile, CPP_DL_ERROR, "missing terminating > character"); + break; + } + + len = cpp_token_len (token) + 2; /* Leading space, terminating \0. */ + if (total_len + len > capacity) + { + capacity = (capacity + len) * 2; + buffer = XRESIZEVEC (char, buffer, capacity); + } + + if (token->flags & PREV_WHITE) + buffer[total_len++] = ' '; + + total_len = (cpp_spell_token (pfile, token, (uchar *) &buffer[total_len], + true) + - (uchar *) buffer); + } + + buffer[total_len] = '\0'; + return buffer; +} + +/* Returns the file name of #include, #include_next, #import and + #pragma dependency. The string is malloced and the caller should + free it. Returns NULL on error. LOCATION is the source location + of the file name. */ + +static const char * +parse_include (cpp_reader *pfile, int *pangle_brackets, + const cpp_token ***buf, location_t *location) +{ + char *fname; + const cpp_token *header; + + /* Allow macro expansion. */ + header = get_token_no_padding (pfile); + *location = header->src_loc; + if ((header->type == CPP_STRING && header->val.str.text[0] != 'R') + || header->type == CPP_HEADER_NAME) + { + fname = XNEWVEC (char, header->val.str.len - 1); + memcpy (fname, header->val.str.text + 1, header->val.str.len - 2); + fname[header->val.str.len - 2] = '\0'; + *pangle_brackets = header->type == CPP_HEADER_NAME; + } + else if (header->type == CPP_LESS) + { + fname = glue_header_name (pfile); + *pangle_brackets = 1; + } + else + { + const unsigned char *dir; + + if (pfile->directive == &dtable[T_PRAGMA]) + dir = UC"pragma dependency"; + else + dir = pfile->directive->name; + cpp_error (pfile, CPP_DL_ERROR, "#%s expects \"FILENAME\" or ", + dir); + + return NULL; + } + + if (pfile->directive == &dtable[T_PRAGMA]) + { + /* This pragma allows extra tokens after the file name. */ + } + else if (buf == NULL || CPP_OPTION (pfile, discard_comments)) + check_eol (pfile, true); + else + { + /* If we are not discarding comments, then gather them while + doing the eol check. */ + *buf = check_eol_return_comments (pfile); + } + + return fname; +} + +/* Handle #include, #include_next and #import. */ +static void +do_include_common (cpp_reader *pfile, enum include_type type) +{ + const char *fname; + int angle_brackets; + const cpp_token **buf = NULL; + location_t location; + + /* Re-enable saving of comments if requested, so that the include + callback can dump comments which follow #include. */ + pfile->state.save_comments = ! CPP_OPTION (pfile, discard_comments); + + /* Tell the lexer this is an include directive -- we want it to + increment the line number even if this is the last line of a file. */ + pfile->state.in_directive = 2; + + fname = parse_include (pfile, &angle_brackets, &buf, &location); + if (!fname) + goto done; + + if (!*fname) + { + cpp_error_with_line (pfile, CPP_DL_ERROR, location, 0, + "empty filename in #%s", + pfile->directive->name); + goto done; + } + + /* Prevent #include recursion. */ + if (pfile->line_table->depth >= CPP_OPTION (pfile, max_include_depth)) + cpp_error (pfile, + CPP_DL_ERROR, + "#include nested depth %u exceeds maximum of %u" + " (use -fmax-include-depth=DEPTH to increase the maximum)", + pfile->line_table->depth, + CPP_OPTION (pfile, max_include_depth)); + else + { + /* Get out of macro context, if we are. */ + skip_rest_of_line (pfile); + + if (pfile->cb.include) + pfile->cb.include (pfile, pfile->directive_line, + pfile->directive->name, fname, angle_brackets, + buf); + + _cpp_stack_include (pfile, fname, angle_brackets, type, location); + } + + done: + XDELETEVEC (fname); + if (buf) + XDELETEVEC (buf); +} + +static void +do_include (cpp_reader *pfile) +{ + do_include_common (pfile, IT_INCLUDE); +} + +static void +do_import (cpp_reader *pfile) +{ + do_include_common (pfile, IT_IMPORT); +} + +static void +do_include_next (cpp_reader *pfile) +{ + enum include_type type = IT_INCLUDE_NEXT; + + /* If this is the primary source file, warn and use the normal + search logic. */ + if (_cpp_in_main_source_file (pfile)) + { + cpp_error (pfile, CPP_DL_WARNING, + "#include_next in primary source file"); + type = IT_INCLUDE; + } + do_include_common (pfile, type); +} + +/* Subroutine of do_linemarker. Read possible flags after file name. + LAST is the last flag seen; 0 if this is the first flag. Return the + flag if it is valid, 0 at the end of the directive. Otherwise + complain. */ +static unsigned int +read_flag (cpp_reader *pfile, unsigned int last) +{ + const cpp_token *token = _cpp_lex_token (pfile); + + if (token->type == CPP_NUMBER && token->val.str.len == 1) + { + unsigned int flag = token->val.str.text[0] - '0'; + + if (flag > last && flag <= 4 + && (flag != 4 || last == 3) + && (flag != 2 || last == 0)) + return flag; + } + + if (token->type != CPP_EOF) + cpp_error (pfile, CPP_DL_ERROR, "invalid flag \"%s\" in line directive", + cpp_token_as_text (pfile, token)); + return 0; +} + +/* Subroutine of do_line and do_linemarker. Convert a number in STR, + of length LEN, to binary; store it in NUMP, and return false if the + number was well-formed, true if not. WRAPPED is set to true if the + number did not fit into 'linenum_type'. */ +static bool +strtolinenum (const uchar *str, size_t len, linenum_type *nump, bool *wrapped) +{ + linenum_type reg = 0; + + uchar c; + bool seen_digit_sep = false; + *wrapped = false; + while (len--) + { + c = *str++; + if (!seen_digit_sep && c == '\'' && len) + { + seen_digit_sep = true; + continue; + } + if (!ISDIGIT (c)) + return true; + seen_digit_sep = false; + if (reg > ((linenum_type) -1) / 10) + *wrapped = true; + reg *= 10; + if (reg > ((linenum_type) -1) - (c - '0')) + *wrapped = true; + reg += c - '0'; + } + *nump = reg; + return false; +} + +/* Interpret #line command. + Note that the filename string (if any) is a true string constant + (escapes are interpreted). */ +static void +do_line (cpp_reader *pfile) +{ + class line_maps *line_table = pfile->line_table; + const line_map_ordinary *map = LINEMAPS_LAST_ORDINARY_MAP (line_table); + + /* skip_rest_of_line() may cause line table to be realloc()ed so note down + sysp right now. */ + + unsigned char map_sysp = ORDINARY_MAP_IN_SYSTEM_HEADER_P (map); + const cpp_token *token; + const char *new_file = ORDINARY_MAP_FILE_NAME (map); + linenum_type new_lineno; + + /* C99 raised the minimum limit on #line numbers. */ + linenum_type cap = CPP_OPTION (pfile, c99) ? 2147483647 : 32767; + bool wrapped; + + /* #line commands expand macros. */ + token = cpp_get_token (pfile); + if (token->type != CPP_NUMBER + || strtolinenum (token->val.str.text, token->val.str.len, + &new_lineno, &wrapped)) + { + if (token->type == CPP_EOF) + cpp_error (pfile, CPP_DL_ERROR, "unexpected end of file after #line"); + else + cpp_error (pfile, CPP_DL_ERROR, + "\"%s\" after #line is not a positive integer", + cpp_token_as_text (pfile, token)); + return; + } + + if (CPP_PEDANTIC (pfile) && (new_lineno == 0 || new_lineno > cap || wrapped)) + cpp_error (pfile, CPP_DL_PEDWARN, "line number out of range"); + else if (wrapped) + cpp_error (pfile, CPP_DL_WARNING, "line number out of range"); + + token = cpp_get_token (pfile); + if (token->type == CPP_STRING) + { + cpp_string s = { 0, 0 }; + if (cpp_interpret_string_notranslate (pfile, &token->val.str, 1, + &s, CPP_STRING)) + new_file = (const char *)s.text; + check_eol (pfile, true); + } + else if (token->type != CPP_EOF) + { + cpp_error (pfile, CPP_DL_ERROR, "\"%s\" is not a valid filename", + cpp_token_as_text (pfile, token)); + return; + } + + skip_rest_of_line (pfile); + _cpp_do_file_change (pfile, LC_RENAME_VERBATIM, new_file, new_lineno, + map_sysp); + line_table->seen_line_directive = true; +} + +/* Interpret the # 44 "file" [flags] notation, which has slightly + different syntax and semantics from #line: Flags are allowed, + and we never complain about the line number being too big. */ +static void +do_linemarker (cpp_reader *pfile) +{ + class line_maps *line_table = pfile->line_table; + const line_map_ordinary *map = LINEMAPS_LAST_ORDINARY_MAP (line_table); + const cpp_token *token; + const char *new_file = ORDINARY_MAP_FILE_NAME (map); + linenum_type new_lineno; + unsigned int new_sysp = ORDINARY_MAP_IN_SYSTEM_HEADER_P (map); + enum lc_reason reason = LC_RENAME_VERBATIM; + int flag; + bool wrapped; + + /* Back up so we can get the number again. Putting this in + _cpp_handle_directive risks two calls to _cpp_backup_tokens in + some circumstances, which can segfault. */ + _cpp_backup_tokens (pfile, 1); + + /* #line commands expand macros. */ + token = cpp_get_token (pfile); + if (token->type != CPP_NUMBER + || strtolinenum (token->val.str.text, token->val.str.len, + &new_lineno, &wrapped)) + { + /* Unlike #line, there does not seem to be a way to get an EOF + here. So, it should be safe to always spell the token. */ + cpp_error (pfile, CPP_DL_ERROR, + "\"%s\" after # is not a positive integer", + cpp_token_as_text (pfile, token)); + return; + } + + token = cpp_get_token (pfile); + if (token->type == CPP_STRING) + { + cpp_string s = { 0, 0 }; + if (cpp_interpret_string_notranslate (pfile, &token->val.str, + 1, &s, CPP_STRING)) + new_file = (const char *)s.text; + + new_sysp = 0; + flag = read_flag (pfile, 0); + if (flag == 1) + { + reason = LC_ENTER; + /* Fake an include for cpp_included (). */ + _cpp_fake_include (pfile, new_file); + flag = read_flag (pfile, flag); + } + else if (flag == 2) + { + reason = LC_LEAVE; + flag = read_flag (pfile, flag); + } + if (flag == 3) + { + new_sysp = 1; + flag = read_flag (pfile, flag); + if (flag == 4) + new_sysp = 2; + } + pfile->buffer->sysp = new_sysp; + + check_eol (pfile, false); + } + else if (token->type != CPP_EOF) + { + cpp_error (pfile, CPP_DL_ERROR, "\"%s\" is not a valid filename", + cpp_token_as_text (pfile, token)); + return; + } + + skip_rest_of_line (pfile); + + if (reason == LC_LEAVE) + { + /* Reread map since cpp_get_token can invalidate it with a + reallocation. */ + map = LINEMAPS_LAST_ORDINARY_MAP (line_table); + const line_map_ordinary *from + = linemap_included_from_linemap (line_table, map); + + if (!from) + /* Not nested. */; + else if (!new_file[0]) + /* Leaving to "" means fill in the popped-to name. */ + new_file = ORDINARY_MAP_FILE_NAME (from); + else if (filename_cmp (ORDINARY_MAP_FILE_NAME (from), new_file) != 0) + /* It's the wrong name, Grommit! */ + from = NULL; + + if (!from) + { + cpp_warning (pfile, CPP_W_NONE, + "file \"%s\" linemarker ignored due to " + "incorrect nesting", new_file); + return; + } + } + + /* Compensate for the increment in linemap_add that occurs in + _cpp_do_file_change. We're currently at the start of the line + *following* the #line directive. A separate location_t for this + location makes no sense (until we do the LC_LEAVE), and + complicates LAST_SOURCE_LINE_LOCATION. */ + pfile->line_table->highest_location--; + + _cpp_do_file_change (pfile, reason, new_file, new_lineno, new_sysp); + line_table->seen_line_directive = true; +} + +/* Arrange the file_change callback. Changing to TO_FILE:TO_LINE for + REASON. SYSP is 1 for a system header, 2 for a system header that + needs to be extern "C" protected, and zero otherwise. */ +void +_cpp_do_file_change (cpp_reader *pfile, enum lc_reason reason, + const char *to_file, linenum_type to_line, + unsigned int sysp) +{ + linemap_assert (reason != LC_ENTER_MACRO); + + const line_map_ordinary *ord_map = NULL; + if (!to_line && reason == LC_RENAME_VERBATIM) + { + /* A linemarker moving to line zero. If we're on the second + line of the current map, and it also starts at zero, just + rewind -- we're probably reading the builtins of a + preprocessed source. */ + line_map_ordinary *last = LINEMAPS_LAST_ORDINARY_MAP (pfile->line_table); + if (!ORDINARY_MAP_STARTING_LINE_NUMBER (last) + && 0 == filename_cmp (to_file, ORDINARY_MAP_FILE_NAME (last)) + && SOURCE_LINE (last, pfile->line_table->highest_line) == 2) + { + ord_map = last; + pfile->line_table->highest_location + = pfile->line_table->highest_line = MAP_START_LOCATION (last); + } + } + + if (!ord_map) + if (const line_map *map = linemap_add (pfile->line_table, reason, sysp, + to_file, to_line)) + { + ord_map = linemap_check_ordinary (map); + linemap_line_start (pfile->line_table, + ORDINARY_MAP_STARTING_LINE_NUMBER (ord_map), + 127); + } + + if (pfile->cb.file_change) + pfile->cb.file_change (pfile, ord_map); +} + +/* Report a warning or error detected by the program we are + processing. Use the directive's tokens in the error message. */ +static void +do_diagnostic (cpp_reader *pfile, enum cpp_diagnostic_level code, + enum cpp_warning_reason reason, int print_dir) +{ + const unsigned char *dir_name; + unsigned char *line; + location_t src_loc = pfile->cur_token[-1].src_loc; + + if (print_dir) + dir_name = pfile->directive->name; + else + dir_name = NULL; + pfile->state.prevent_expansion++; + line = cpp_output_line_to_string (pfile, dir_name); + pfile->state.prevent_expansion--; + + if (code == CPP_DL_WARNING_SYSHDR && reason) + cpp_warning_with_line_syshdr (pfile, reason, src_loc, 0, "%s", line); + else if (code == CPP_DL_WARNING && reason) + cpp_warning_with_line (pfile, reason, src_loc, 0, "%s", line); + else + cpp_error_with_line (pfile, code, src_loc, 0, "%s", line); + free (line); +} + +static void +do_error (cpp_reader *pfile) +{ + do_diagnostic (pfile, CPP_DL_ERROR, CPP_W_NONE, 1); +} + +static void +do_warning (cpp_reader *pfile) +{ + /* We want #warning diagnostics to be emitted in system headers too. */ + do_diagnostic (pfile, CPP_DL_WARNING_SYSHDR, CPP_W_WARNING_DIRECTIVE, 1); +} + +/* Report program identification. */ +static void +do_ident (cpp_reader *pfile) +{ + const cpp_token *str = cpp_get_token (pfile); + + if (str->type != CPP_STRING) + cpp_error (pfile, CPP_DL_ERROR, "invalid #%s directive", + pfile->directive->name); + else if (pfile->cb.ident) + pfile->cb.ident (pfile, pfile->directive_line, &str->val.str); + + check_eol (pfile, false); +} + +/* Lookup a PRAGMA name in a singly-linked CHAIN. Returns the + matching entry, or NULL if none is found. The returned entry could + be the start of a namespace chain, or a pragma. */ +static struct pragma_entry * +lookup_pragma_entry (struct pragma_entry *chain, const cpp_hashnode *pragma) +{ + while (chain && chain->pragma != pragma) + chain = chain->next; + + return chain; +} + +/* Create and insert a blank pragma entry at the beginning of a + singly-linked CHAIN. */ +static struct pragma_entry * +new_pragma_entry (cpp_reader *pfile, struct pragma_entry **chain) +{ + struct pragma_entry *new_entry; + + new_entry = (struct pragma_entry *) + _cpp_aligned_alloc (pfile, sizeof (struct pragma_entry)); + + memset (new_entry, 0, sizeof (struct pragma_entry)); + new_entry->next = *chain; + + *chain = new_entry; + return new_entry; +} + +/* Register a pragma NAME in namespace SPACE. If SPACE is null, it + goes in the global namespace. */ +static struct pragma_entry * +register_pragma_1 (cpp_reader *pfile, const char *space, const char *name, + bool allow_name_expansion) +{ + struct pragma_entry **chain = &pfile->pragmas; + struct pragma_entry *entry; + const cpp_hashnode *node; + + if (space) + { + node = cpp_lookup (pfile, UC space, strlen (space)); + entry = lookup_pragma_entry (*chain, node); + if (!entry) + { + entry = new_pragma_entry (pfile, chain); + entry->pragma = node; + entry->is_nspace = true; + entry->allow_expansion = allow_name_expansion; + } + else if (!entry->is_nspace) + goto clash; + else if (entry->allow_expansion != allow_name_expansion) + { + cpp_error (pfile, CPP_DL_ICE, + "registering pragmas in namespace \"%s\" with mismatched " + "name expansion", space); + return NULL; + } + chain = &entry->u.space; + } + else if (allow_name_expansion) + { + cpp_error (pfile, CPP_DL_ICE, + "registering pragma \"%s\" with name expansion " + "and no namespace", name); + return NULL; + } + + /* Check for duplicates. */ + node = cpp_lookup (pfile, UC name, strlen (name)); + entry = lookup_pragma_entry (*chain, node); + if (entry == NULL) + { + entry = new_pragma_entry (pfile, chain); + entry->pragma = node; + return entry; + } + + if (entry->is_nspace) + clash: + cpp_error (pfile, CPP_DL_ICE, + "registering \"%s\" as both a pragma and a pragma namespace", + NODE_NAME (node)); + else if (space) + cpp_error (pfile, CPP_DL_ICE, "#pragma %s %s is already registered", + space, name); + else + cpp_error (pfile, CPP_DL_ICE, "#pragma %s is already registered", name); + + return NULL; +} + +/* Register a cpplib internal pragma SPACE NAME with HANDLER. */ +static void +register_pragma_internal (cpp_reader *pfile, const char *space, + const char *name, pragma_cb handler) +{ + struct pragma_entry *entry; + + entry = register_pragma_1 (pfile, space, name, false); + entry->is_internal = true; + entry->u.handler = handler; +} + +/* Register a pragma NAME in namespace SPACE. If SPACE is null, it + goes in the global namespace. HANDLER is the handler it will call, + which must be non-NULL. If ALLOW_EXPANSION is set, allow macro + expansion while parsing pragma NAME. This function is exported + from libcpp. */ +void +cpp_register_pragma (cpp_reader *pfile, const char *space, const char *name, + pragma_cb handler, bool allow_expansion) +{ + struct pragma_entry *entry; + + if (!handler) + { + cpp_error (pfile, CPP_DL_ICE, "registering pragma with NULL handler"); + return; + } + + entry = register_pragma_1 (pfile, space, name, false); + if (entry) + { + entry->allow_expansion = allow_expansion; + entry->u.handler = handler; + } +} + +/* Similarly, but create mark the pragma for deferred processing. + When found, a CPP_PRAGMA token will be insertted into the stream + with IDENT in the token->u.pragma slot. */ +void +cpp_register_deferred_pragma (cpp_reader *pfile, const char *space, + const char *name, unsigned int ident, + bool allow_expansion, bool allow_name_expansion) +{ + struct pragma_entry *entry; + + entry = register_pragma_1 (pfile, space, name, allow_name_expansion); + if (entry) + { + entry->is_deferred = true; + entry->allow_expansion = allow_expansion; + entry->u.ident = ident; + } +} + +/* Register the pragmas the preprocessor itself handles. */ +void +_cpp_init_internal_pragmas (cpp_reader *pfile) +{ + /* Pragmas in the global namespace. */ + register_pragma_internal (pfile, 0, "once", do_pragma_once); + register_pragma_internal (pfile, 0, "push_macro", do_pragma_push_macro); + register_pragma_internal (pfile, 0, "pop_macro", do_pragma_pop_macro); + + /* New GCC-specific pragmas should be put in the GCC namespace. */ + register_pragma_internal (pfile, "GCC", "poison", do_pragma_poison); + register_pragma_internal (pfile, "GCC", "system_header", + do_pragma_system_header); + register_pragma_internal (pfile, "GCC", "dependency", do_pragma_dependency); + register_pragma_internal (pfile, "GCC", "warning", do_pragma_warning); + register_pragma_internal (pfile, "GCC", "error", do_pragma_error); +} + +/* Return the number of registered pragmas in PE. */ + +static int +count_registered_pragmas (struct pragma_entry *pe) +{ + int ct = 0; + for (; pe != NULL; pe = pe->next) + { + if (pe->is_nspace) + ct += count_registered_pragmas (pe->u.space); + ct++; + } + return ct; +} + +/* Save into SD the names of the registered pragmas referenced by PE, + and return a pointer to the next free space in SD. */ + +static char ** +save_registered_pragmas (struct pragma_entry *pe, char **sd) +{ + for (; pe != NULL; pe = pe->next) + { + if (pe->is_nspace) + sd = save_registered_pragmas (pe->u.space, sd); + *sd++ = (char *) xmemdup (HT_STR (&pe->pragma->ident), + HT_LEN (&pe->pragma->ident), + HT_LEN (&pe->pragma->ident) + 1); + } + return sd; +} + +/* Return a newly-allocated array which saves the names of the + registered pragmas. */ + +char ** +_cpp_save_pragma_names (cpp_reader *pfile) +{ + int ct = count_registered_pragmas (pfile->pragmas); + char **result = XNEWVEC (char *, ct); + (void) save_registered_pragmas (pfile->pragmas, result); + return result; +} + +/* Restore from SD the names of the registered pragmas referenced by PE, + and return a pointer to the next unused name in SD. */ + +static char ** +restore_registered_pragmas (cpp_reader *pfile, struct pragma_entry *pe, + char **sd) +{ + for (; pe != NULL; pe = pe->next) + { + if (pe->is_nspace) + sd = restore_registered_pragmas (pfile, pe->u.space, sd); + pe->pragma = cpp_lookup (pfile, UC *sd, strlen (*sd)); + free (*sd); + sd++; + } + return sd; +} + +/* Restore the names of the registered pragmas from SAVED. */ + +void +_cpp_restore_pragma_names (cpp_reader *pfile, char **saved) +{ + (void) restore_registered_pragmas (pfile, pfile->pragmas, saved); + free (saved); +} + +/* Pragmata handling. We handle some, and pass the rest on to the + front end. C99 defines three pragmas and says that no macro + expansion is to be performed on them; whether or not macro + expansion happens for other pragmas is implementation defined. + This implementation allows for a mix of both, since GCC did not + traditionally macro expand its (few) pragmas, whereas OpenMP + specifies that macro expansion should happen. */ +static void +do_pragma (cpp_reader *pfile) +{ + const struct pragma_entry *p = NULL; + const cpp_token *token, *pragma_token; + location_t pragma_token_virt_loc = 0; + cpp_token ns_token; + unsigned int count = 1; + + pfile->state.prevent_expansion++; + + pragma_token = token = cpp_get_token_with_location (pfile, + &pragma_token_virt_loc); + ns_token = *token; + if (token->type == CPP_NAME) + { + p = lookup_pragma_entry (pfile->pragmas, token->val.node.node); + if (p && p->is_nspace) + { + bool allow_name_expansion = p->allow_expansion; + if (allow_name_expansion) + pfile->state.prevent_expansion--; + + token = cpp_get_token (pfile); + if (token->type == CPP_NAME) + p = lookup_pragma_entry (p->u.space, token->val.node.node); + else + p = NULL; + if (allow_name_expansion) + pfile->state.prevent_expansion++; + count = 2; + } + } + + if (p) + { + if (p->is_deferred) + { + pfile->directive_result.src_loc = pragma_token_virt_loc; + pfile->directive_result.type = CPP_PRAGMA; + pfile->directive_result.flags = pragma_token->flags; + pfile->directive_result.val.pragma = p->u.ident; + pfile->state.in_deferred_pragma = true; + pfile->state.pragma_allow_expansion = p->allow_expansion; + if (!p->allow_expansion) + pfile->state.prevent_expansion++; + } + else + { + /* Since the handler below doesn't get the line number, that + it might need for diagnostics, make sure it has the right + numbers in place. */ + if (pfile->cb.line_change) + (*pfile->cb.line_change) (pfile, pragma_token, false); + if (p->allow_expansion) + pfile->state.prevent_expansion--; + (*p->u.handler) (pfile); + if (p->allow_expansion) + pfile->state.prevent_expansion++; + } + } + else if (pfile->cb.def_pragma) + { + if (count == 1 || pfile->context->prev == NULL) + _cpp_backup_tokens (pfile, count); + else + { + /* Invalid name comes from macro expansion, _cpp_backup_tokens + won't allow backing 2 tokens. */ + /* ??? The token buffer is leaked. Perhaps if def_pragma hook + reads both tokens, we could perhaps free it, but if it doesn't, + we don't know the exact lifespan. */ + cpp_token *toks = XNEWVEC (cpp_token, 2); + toks[0] = ns_token; + toks[0].flags |= NO_EXPAND; + toks[1] = *token; + toks[1].flags |= NO_EXPAND; + _cpp_push_token_context (pfile, NULL, toks, 2); + } + pfile->cb.def_pragma (pfile, pfile->directive_line); + } + + pfile->state.prevent_expansion--; +} + +/* Handle #pragma once. */ +static void +do_pragma_once (cpp_reader *pfile) +{ + if (_cpp_in_main_source_file (pfile)) + cpp_error (pfile, CPP_DL_WARNING, "#pragma once in main file"); + + check_eol (pfile, false); + _cpp_mark_file_once_only (pfile, pfile->buffer->file); +} + +/* Handle #pragma push_macro(STRING). */ +static void +do_pragma_push_macro (cpp_reader *pfile) +{ + cpp_hashnode *node; + size_t defnlen; + const uchar *defn = NULL; + char *macroname, *dest; + const char *limit, *src; + const cpp_token *txt; + struct def_pragma_macro *c; + + txt = get__Pragma_string (pfile); + if (!txt) + { + location_t src_loc = pfile->cur_token[-1].src_loc; + cpp_error_with_line (pfile, CPP_DL_ERROR, src_loc, 0, + "invalid #pragma push_macro directive"); + check_eol (pfile, false); + skip_rest_of_line (pfile); + return; + } + dest = macroname = (char *) alloca (txt->val.str.len + 2); + src = (const char *) (txt->val.str.text + 1 + (txt->val.str.text[0] == 'L')); + limit = (const char *) (txt->val.str.text + txt->val.str.len - 1); + while (src < limit) + { + /* We know there is a character following the backslash. */ + if (*src == '\\' && (src[1] == '\\' || src[1] == '"')) + src++; + *dest++ = *src++; + } + *dest = 0; + check_eol (pfile, false); + skip_rest_of_line (pfile); + c = XNEW (struct def_pragma_macro); + memset (c, 0, sizeof (struct def_pragma_macro)); + c->name = XNEWVAR (char, strlen (macroname) + 1); + strcpy (c->name, macroname); + c->next = pfile->pushed_macros; + node = _cpp_lex_identifier (pfile, c->name); + if (node->type == NT_VOID) + c->is_undef = 1; + else if (node->type == NT_BUILTIN_MACRO) + c->is_builtin = 1; + else + { + defn = cpp_macro_definition (pfile, node); + defnlen = ustrlen (defn); + c->definition = XNEWVEC (uchar, defnlen + 2); + c->definition[defnlen] = '\n'; + c->definition[defnlen + 1] = 0; + c->line = node->value.macro->line; + c->syshdr = node->value.macro->syshdr; + c->used = node->value.macro->used; + memcpy (c->definition, defn, defnlen); + } + + pfile->pushed_macros = c; +} + +/* Handle #pragma pop_macro(STRING). */ +static void +do_pragma_pop_macro (cpp_reader *pfile) +{ + char *macroname, *dest; + const char *limit, *src; + const cpp_token *txt; + struct def_pragma_macro *l = NULL, *c = pfile->pushed_macros; + txt = get__Pragma_string (pfile); + if (!txt) + { + location_t src_loc = pfile->cur_token[-1].src_loc; + cpp_error_with_line (pfile, CPP_DL_ERROR, src_loc, 0, + "invalid #pragma pop_macro directive"); + check_eol (pfile, false); + skip_rest_of_line (pfile); + return; + } + dest = macroname = (char *) alloca (txt->val.str.len + 2); + src = (const char *) (txt->val.str.text + 1 + (txt->val.str.text[0] == 'L')); + limit = (const char *) (txt->val.str.text + txt->val.str.len - 1); + while (src < limit) + { + /* We know there is a character following the backslash. */ + if (*src == '\\' && (src[1] == '\\' || src[1] == '"')) + src++; + *dest++ = *src++; + } + *dest = 0; + check_eol (pfile, false); + skip_rest_of_line (pfile); + + while (c != NULL) + { + if (!strcmp (c->name, macroname)) + { + if (!l) + pfile->pushed_macros = c->next; + else + l->next = c->next; + cpp_pop_definition (pfile, c); + free (c->definition); + free (c->name); + free (c); + break; + } + l = c; + c = c->next; + } +} + +/* Handle #pragma GCC poison, to poison one or more identifiers so + that the lexer produces a hard error for each subsequent usage. */ +static void +do_pragma_poison (cpp_reader *pfile) +{ + const cpp_token *tok; + cpp_hashnode *hp; + + pfile->state.poisoned_ok = 1; + for (;;) + { + tok = _cpp_lex_token (pfile); + if (tok->type == CPP_EOF) + break; + if (tok->type != CPP_NAME) + { + cpp_error (pfile, CPP_DL_ERROR, + "invalid #pragma GCC poison directive"); + break; + } + + hp = tok->val.node.node; + if (hp->flags & NODE_POISONED) + continue; + + if (cpp_macro_p (hp)) + cpp_error (pfile, CPP_DL_WARNING, "poisoning existing macro \"%s\"", + NODE_NAME (hp)); + _cpp_free_definition (hp); + hp->flags |= NODE_POISONED | NODE_DIAGNOSTIC; + } + pfile->state.poisoned_ok = 0; +} + +/* Mark the current header as a system header. This will suppress + some categories of warnings (notably those from -pedantic). It is + intended for use in system libraries that cannot be implemented in + conforming C, but cannot be certain that their headers appear in a + system include directory. To prevent abuse, it is rejected in the + primary source file. */ +static void +do_pragma_system_header (cpp_reader *pfile) +{ + if (_cpp_in_main_source_file (pfile)) + cpp_error (pfile, CPP_DL_WARNING, + "#pragma system_header ignored outside include file"); + else + { + check_eol (pfile, false); + skip_rest_of_line (pfile); + cpp_make_system_header (pfile, 1, 0); + } +} + +/* Check the modified date of the current include file against a specified + file. Issue a diagnostic, if the specified file is newer. We use this to + determine if a fixed header should be refixed. */ +static void +do_pragma_dependency (cpp_reader *pfile) +{ + const char *fname; + int angle_brackets, ordering; + location_t location; + + fname = parse_include (pfile, &angle_brackets, NULL, &location); + if (!fname) + return; + + ordering = _cpp_compare_file_date (pfile, fname, angle_brackets); + if (ordering < 0) + cpp_error (pfile, CPP_DL_WARNING, "cannot find source file %s", fname); + else if (ordering > 0) + { + cpp_error (pfile, CPP_DL_WARNING, + "current file is older than %s", fname); + if (cpp_get_token (pfile)->type != CPP_EOF) + { + _cpp_backup_tokens (pfile, 1); + do_diagnostic (pfile, CPP_DL_WARNING, CPP_W_NONE, 0); + } + } + + free ((void *) fname); +} + +/* Issue a diagnostic with the message taken from the pragma. If + ERROR is true, the diagnostic is a warning, otherwise, it is an + error. */ +static void +do_pragma_warning_or_error (cpp_reader *pfile, bool error) +{ + const cpp_token *tok = _cpp_lex_token (pfile); + cpp_string str; + if (tok->type != CPP_STRING + || !cpp_interpret_string_notranslate (pfile, &tok->val.str, 1, &str, + CPP_STRING) + || str.len == 0) + { + cpp_error (pfile, CPP_DL_ERROR, "invalid \"#pragma GCC %s\" directive", + error ? "error" : "warning"); + return; + } + cpp_error (pfile, error ? CPP_DL_ERROR : CPP_DL_WARNING, + "%s", str.text); + free ((void *)str.text); +} + +/* Issue a warning diagnostic. */ +static void +do_pragma_warning (cpp_reader *pfile) +{ + do_pragma_warning_or_error (pfile, false); +} + +/* Issue an error diagnostic. */ +static void +do_pragma_error (cpp_reader *pfile) +{ + do_pragma_warning_or_error (pfile, true); +} + +/* Get a token but skip padding. */ +static const cpp_token * +get_token_no_padding (cpp_reader *pfile) +{ + for (;;) + { + const cpp_token *result = cpp_get_token (pfile); + if (result->type != CPP_PADDING) + return result; + } +} + +/* Check syntax is "(string-literal)". Returns the string on success, + or NULL on failure. */ +static const cpp_token * +get__Pragma_string (cpp_reader *pfile) +{ + const cpp_token *string; + const cpp_token *paren; + + paren = get_token_no_padding (pfile); + if (paren->type == CPP_EOF) + _cpp_backup_tokens (pfile, 1); + if (paren->type != CPP_OPEN_PAREN) + return NULL; + + string = get_token_no_padding (pfile); + if (string->type == CPP_EOF) + _cpp_backup_tokens (pfile, 1); + if (string->type != CPP_STRING && string->type != CPP_WSTRING + && string->type != CPP_STRING32 && string->type != CPP_STRING16 + && string->type != CPP_UTF8STRING) + return NULL; + + paren = get_token_no_padding (pfile); + if (paren->type == CPP_EOF) + _cpp_backup_tokens (pfile, 1); + if (paren->type != CPP_CLOSE_PAREN) + return NULL; + + return string; +} + +/* Destringize IN into a temporary buffer, by removing the first \ of + \" and \\ sequences, and process the result as a #pragma directive. */ +static void +destringize_and_run (cpp_reader *pfile, const cpp_string *in, + location_t expansion_loc) +{ + const unsigned char *src, *limit; + char *dest, *result; + cpp_context *saved_context; + cpp_token *saved_cur_token; + tokenrun *saved_cur_run; + cpp_token *toks; + int count; + const struct directive *save_directive; + + dest = result = (char *) alloca (in->len - 1); + src = in->text + 1 + (in->text[0] == 'L'); + limit = in->text + in->len - 1; + while (src < limit) + { + /* We know there is a character following the backslash. */ + if (*src == '\\' && (src[1] == '\\' || src[1] == '"')) + src++; + *dest++ = *src++; + } + *dest = '\n'; + + /* Ugh; an awful kludge. We are really not set up to be lexing + tokens when in the middle of a macro expansion. Use a new + context to force cpp_get_token to lex, and so skip_rest_of_line + doesn't go beyond the end of the text. Also, remember the + current lexing position so we can return to it later. + + Something like line-at-a-time lexing should remove the need for + this. */ + saved_context = pfile->context; + saved_cur_token = pfile->cur_token; + saved_cur_run = pfile->cur_run; + + pfile->context = XCNEW (cpp_context); + + /* Inline run_directive, since we need to delay the _cpp_pop_buffer + until we've read all of the tokens that we want. */ + cpp_push_buffer (pfile, (const uchar *) result, dest - result, + /* from_stage3 */ true); + /* ??? Antique Disgusting Hack. What does this do? */ + if (pfile->buffer->prev) + pfile->buffer->file = pfile->buffer->prev->file; + + start_directive (pfile); + _cpp_clean_line (pfile); + save_directive = pfile->directive; + pfile->directive = &dtable[T_PRAGMA]; + do_pragma (pfile); + if (pfile->directive_result.type == CPP_PRAGMA) + pfile->directive_result.flags |= PRAGMA_OP; + end_directive (pfile, 1); + pfile->directive = save_directive; + + /* We always insert at least one token, the directive result. It'll + either be a CPP_PADDING or a CPP_PRAGMA. In the later case, we + need to insert *all* of the tokens, including the CPP_PRAGMA_EOL. */ + + /* If we're not handling the pragma internally, read all of the tokens from + the string buffer now, while the string buffer is still installed. */ + /* ??? Note that the token buffer allocated here is leaked. It's not clear + to me what the true lifespan of the tokens are. It would appear that + the lifespan is the entire parse of the main input stream, in which case + this may not be wrong. */ + if (pfile->directive_result.type == CPP_PRAGMA) + { + int maxcount; + + count = 1; + maxcount = 50; + toks = XNEWVEC (cpp_token, maxcount); + toks[0] = pfile->directive_result; + toks[0].src_loc = expansion_loc; + + do + { + if (count == maxcount) + { + maxcount = maxcount * 3 / 2; + toks = XRESIZEVEC (cpp_token, toks, maxcount); + } + toks[count] = *cpp_get_token (pfile); + /* _Pragma is a builtin, so we're not within a macro-map, and so + the token locations are set to bogus ordinary locations + near to, but after that of the "_Pragma". + Paper over this by setting them equal to the location of the + _Pragma itself (PR preprocessor/69126). */ + toks[count].src_loc = expansion_loc; + /* Macros have been already expanded by cpp_get_token + if the pragma allowed expansion. */ + toks[count++].flags |= NO_EXPAND; + } + while (toks[count-1].type != CPP_PRAGMA_EOL); + } + else + { + count = 1; + toks = &pfile->avoid_paste; + + /* If we handled the entire pragma internally, make sure we get the + line number correct for the next token. */ + if (pfile->cb.line_change) + pfile->cb.line_change (pfile, pfile->cur_token, false); + } + + /* Finish inlining run_directive. */ + pfile->buffer->file = NULL; + _cpp_pop_buffer (pfile); + + /* Reset the old macro state before ... */ + XDELETE (pfile->context); + pfile->context = saved_context; + pfile->cur_token = saved_cur_token; + pfile->cur_run = saved_cur_run; + + /* ... inserting the new tokens we collected. */ + _cpp_push_token_context (pfile, NULL, toks, count); +} + +/* Handle the _Pragma operator. Return 0 on error, 1 if ok. */ +int +_cpp_do__Pragma (cpp_reader *pfile, location_t expansion_loc) +{ + const cpp_token *string = get__Pragma_string (pfile); + pfile->directive_result.type = CPP_PADDING; + + if (string) + { + destringize_and_run (pfile, &string->val.str, expansion_loc); + return 1; + } + cpp_error (pfile, CPP_DL_ERROR, + "_Pragma takes a parenthesized string literal"); + return 0; +} + +/* Handle #ifdef. */ +static void +do_ifdef (cpp_reader *pfile) +{ + int skip = 1; + + if (! pfile->state.skipping) + { + cpp_hashnode *node = lex_macro_node (pfile, false); + + if (node) + { + skip = !_cpp_defined_macro_p (node); + if (!_cpp_maybe_notify_macro_use (pfile, node, pfile->directive_line)) + /* It wasn't a macro after all. */ + skip = true; + _cpp_mark_macro_used (node); + if (pfile->cb.used) + pfile->cb.used (pfile, pfile->directive_line, node); + check_eol (pfile, false); + } + } + + push_conditional (pfile, skip, T_IFDEF, 0); +} + +/* Handle #ifndef. */ +static void +do_ifndef (cpp_reader *pfile) +{ + int skip = 1; + cpp_hashnode *node = 0; + + if (! pfile->state.skipping) + { + node = lex_macro_node (pfile, false); + + if (node) + { + skip = _cpp_defined_macro_p (node); + if (!_cpp_maybe_notify_macro_use (pfile, node, pfile->directive_line)) + /* It wasn't a macro after all. */ + skip = false; + _cpp_mark_macro_used (node); + if (pfile->cb.used) + pfile->cb.used (pfile, pfile->directive_line, node); + check_eol (pfile, false); + } + } + + push_conditional (pfile, skip, T_IFNDEF, node); +} + +/* _cpp_parse_expr puts a macro in a "#if !defined ()" expression in + pfile->mi_ind_cmacro so we can handle multiple-include + optimizations. If macro expansion occurs in the expression, we + cannot treat it as a controlling conditional, since the expansion + could change in the future. That is handled by cpp_get_token. */ +static void +do_if (cpp_reader *pfile) +{ + int skip = 1; + + if (! pfile->state.skipping) + skip = _cpp_parse_expr (pfile, true) == false; + + push_conditional (pfile, skip, T_IF, pfile->mi_ind_cmacro); +} + +/* Flip skipping state if appropriate and continue without changing + if_stack; this is so that the error message for missing #endif's + etc. will point to the original #if. */ +static void +do_else (cpp_reader *pfile) +{ + cpp_buffer *buffer = pfile->buffer; + struct if_stack *ifs = buffer->if_stack; + + if (ifs == NULL) + cpp_error (pfile, CPP_DL_ERROR, "#else without #if"); + else + { + if (ifs->type == T_ELSE) + { + cpp_error (pfile, CPP_DL_ERROR, "#else after #else"); + cpp_error_with_line (pfile, CPP_DL_ERROR, ifs->line, 0, + "the conditional began here"); + } + ifs->type = T_ELSE; + + /* Skip any future (erroneous) #elses or #elifs. */ + pfile->state.skipping = ifs->skip_elses; + ifs->skip_elses = true; + + /* Invalidate any controlling macro. */ + ifs->mi_cmacro = 0; + + /* Only check EOL if was not originally skipping. */ + if (!ifs->was_skipping && CPP_OPTION (pfile, warn_endif_labels)) + check_eol_endif_labels (pfile); + } +} + +/* Handle a #elif, #elifdef or #elifndef directive by not changing if_stack + either. See the comment above do_else. */ +static void +do_elif (cpp_reader *pfile) +{ + cpp_buffer *buffer = pfile->buffer; + struct if_stack *ifs = buffer->if_stack; + + if (ifs == NULL) + cpp_error (pfile, CPP_DL_ERROR, "#%s without #if", pfile->directive->name); + else + { + if (ifs->type == T_ELSE) + { + cpp_error (pfile, CPP_DL_ERROR, "#%s after #else", + pfile->directive->name); + cpp_error_with_line (pfile, CPP_DL_ERROR, ifs->line, 0, + "the conditional began here"); + } + ifs->type = T_ELIF; + + /* See DR#412: "Only the first group whose control condition + evaluates to true (nonzero) is processed; any following groups + are skipped and their controlling directives are processed as + if they were in a group that is skipped." */ + if (ifs->skip_elses) + { + /* In older GNU standards, #elifdef/#elifndef is supported + as an extension, but pedwarn if -pedantic if the presence + of the directive would be rejected. */ + if (pfile->directive != &dtable[T_ELIF] + && ! CPP_OPTION (pfile, elifdef) + && CPP_PEDANTIC (pfile) + && !pfile->state.skipping) + { + if (CPP_OPTION (pfile, cplusplus)) + cpp_error (pfile, CPP_DL_PEDWARN, + "#%s before C++23 is a GCC extension", + pfile->directive->name); + else + cpp_error (pfile, CPP_DL_PEDWARN, + "#%s before C2X is a GCC extension", + pfile->directive->name); + } + pfile->state.skipping = 1; + } + else + { + if (pfile->directive == &dtable[T_ELIF]) + pfile->state.skipping = !_cpp_parse_expr (pfile, false); + else + { + cpp_hashnode *node = lex_macro_node (pfile, false); + + if (node) + { + bool macro_defined = _cpp_defined_macro_p (node); + if (!_cpp_maybe_notify_macro_use (pfile, node, + pfile->directive_line)) + /* It wasn't a macro after all. */ + macro_defined = false; + bool skip = (pfile->directive == &dtable[T_ELIFDEF] + ? !macro_defined + : macro_defined); + if (pfile->cb.used) + pfile->cb.used (pfile, pfile->directive_line, node); + check_eol (pfile, false); + /* In older GNU standards, #elifdef/#elifndef is supported + as an extension, but pedwarn if -pedantic if the presence + of the directive would change behavior. */ + if (! CPP_OPTION (pfile, elifdef) + && CPP_PEDANTIC (pfile) + && pfile->state.skipping != skip) + { + if (CPP_OPTION (pfile, cplusplus)) + cpp_error (pfile, CPP_DL_PEDWARN, + "#%s before C++23 is a GCC extension", + pfile->directive->name); + else + cpp_error (pfile, CPP_DL_PEDWARN, + "#%s before C2X is a GCC extension", + pfile->directive->name); + } + pfile->state.skipping = skip; + } + } + ifs->skip_elses = !pfile->state.skipping; + } + + /* Invalidate any controlling macro. */ + ifs->mi_cmacro = 0; + } +} + +/* Handle a #elifdef directive. */ +static void +do_elifdef (cpp_reader *pfile) +{ + do_elif (pfile); +} + +/* Handle a #elifndef directive. */ +static void +do_elifndef (cpp_reader *pfile) +{ + do_elif (pfile); +} + +/* #endif pops the if stack and resets pfile->state.skipping. */ +static void +do_endif (cpp_reader *pfile) +{ + cpp_buffer *buffer = pfile->buffer; + struct if_stack *ifs = buffer->if_stack; + + if (ifs == NULL) + cpp_error (pfile, CPP_DL_ERROR, "#endif without #if"); + else + { + /* Only check EOL if was not originally skipping. */ + if (!ifs->was_skipping && CPP_OPTION (pfile, warn_endif_labels)) + check_eol_endif_labels (pfile); + + /* If potential control macro, we go back outside again. */ + if (ifs->next == 0 && ifs->mi_cmacro) + { + pfile->mi_valid = true; + pfile->mi_cmacro = ifs->mi_cmacro; + } + + buffer->if_stack = ifs->next; + pfile->state.skipping = ifs->was_skipping; + obstack_free (&pfile->buffer_ob, ifs); + } +} + +/* Push an if_stack entry for a preprocessor conditional, and set + pfile->state.skipping to SKIP. If TYPE indicates the conditional + is #if or #ifndef, CMACRO is a potentially controlling macro, and + we need to check here that we are at the top of the file. */ +static void +push_conditional (cpp_reader *pfile, int skip, int type, + const cpp_hashnode *cmacro) +{ + struct if_stack *ifs; + cpp_buffer *buffer = pfile->buffer; + + ifs = XOBNEW (&pfile->buffer_ob, struct if_stack); + ifs->line = pfile->directive_line; + ifs->next = buffer->if_stack; + ifs->skip_elses = pfile->state.skipping || !skip; + ifs->was_skipping = pfile->state.skipping; + ifs->type = type; + /* This condition is effectively a test for top-of-file. */ + if (pfile->mi_valid && pfile->mi_cmacro == 0) + ifs->mi_cmacro = cmacro; + else + ifs->mi_cmacro = 0; + + pfile->state.skipping = skip; + buffer->if_stack = ifs; +} + +/* Read the tokens of the answer into the macro pool, in a directive + of type TYPE. Only commit the memory if we intend it as permanent + storage, i.e. the #assert case. Returns 0 on success, and sets + ANSWERP to point to the answer. PRED_LOC is the location of the + predicate. */ +static bool +parse_answer (cpp_reader *pfile, int type, location_t pred_loc, + cpp_macro **answer_ptr) +{ + /* In a conditional, it is legal to not have an open paren. We + should save the following token in this case. */ + const cpp_token *paren = cpp_get_token (pfile); + + /* If not a paren, see if we're OK. */ + if (paren->type != CPP_OPEN_PAREN) + { + /* In a conditional no answer is a test for any answer. It + could be followed by any token. */ + if (type == T_IF) + { + _cpp_backup_tokens (pfile, 1); + return true; + } + + /* #unassert with no answer is valid - it removes all answers. */ + if (type == T_UNASSERT && paren->type == CPP_EOF) + return true; + + cpp_error_with_line (pfile, CPP_DL_ERROR, pred_loc, 0, + "missing '(' after predicate"); + return false; + } + + cpp_macro *answer = _cpp_new_macro (pfile, cmk_assert, + _cpp_reserve_room (pfile, 0, + sizeof (cpp_macro))); + answer->parm.next = NULL; + unsigned count = 0; + for (;;) + { + const cpp_token *token = cpp_get_token (pfile); + + if (token->type == CPP_CLOSE_PAREN) + break; + + if (token->type == CPP_EOF) + { + cpp_error (pfile, CPP_DL_ERROR, "missing ')' to complete answer"); + return false; + } + + answer = (cpp_macro *)_cpp_reserve_room + (pfile, sizeof (cpp_macro) + count * sizeof (cpp_token), + sizeof (cpp_token)); + answer->exp.tokens[count++] = *token; + } + + if (!count) + { + cpp_error (pfile, CPP_DL_ERROR, "predicate's answer is empty"); + return false; + } + + /* Drop whitespace at start, for answer equivalence purposes. */ + answer->exp.tokens[0].flags &= ~PREV_WHITE; + + answer->count = count; + *answer_ptr = answer; + + return true; +} + +/* Parses an assertion directive of type TYPE, returning a pointer to + the hash node of the predicate, or 0 on error. The node is + guaranteed to be disjoint from the macro namespace, so can only + have type 'NT_VOID'. If an answer was supplied, it is placed in + *ANSWER_PTR, which is otherwise set to 0. */ +static cpp_hashnode * +parse_assertion (cpp_reader *pfile, int type, cpp_macro **answer_ptr) +{ + cpp_hashnode *result = 0; + + /* We don't expand predicates or answers. */ + pfile->state.prevent_expansion++; + + *answer_ptr = NULL; + + const cpp_token *predicate = cpp_get_token (pfile); + if (predicate->type == CPP_EOF) + cpp_error (pfile, CPP_DL_ERROR, "assertion without predicate"); + else if (predicate->type != CPP_NAME) + cpp_error_with_line (pfile, CPP_DL_ERROR, predicate->src_loc, 0, + "predicate must be an identifier"); + else if (parse_answer (pfile, type, predicate->src_loc, answer_ptr)) + { + unsigned int len = NODE_LEN (predicate->val.node.node); + unsigned char *sym = (unsigned char *) alloca (len + 1); + + /* Prefix '#' to get it out of macro namespace. */ + sym[0] = '#'; + memcpy (sym + 1, NODE_NAME (predicate->val.node.node), len); + result = cpp_lookup (pfile, sym, len + 1); + } + + pfile->state.prevent_expansion--; + + return result; +} + +/* Returns a pointer to the pointer to CANDIDATE in the answer chain, + or a pointer to NULL if the answer is not in the chain. */ +static cpp_macro ** +find_answer (cpp_hashnode *node, const cpp_macro *candidate) +{ + unsigned int i; + cpp_macro **result = NULL; + + for (result = &node->value.answers; *result; result = &(*result)->parm.next) + { + cpp_macro *answer = *result; + + if (answer->count == candidate->count) + { + for (i = 0; i < answer->count; i++) + if (!_cpp_equiv_tokens (&answer->exp.tokens[i], + &candidate->exp.tokens[i])) + break; + + if (i == answer->count) + break; + } + } + + return result; +} + +/* Test an assertion within a preprocessor conditional. Returns + nonzero on failure, zero on success. On success, the result of + the test is written into VALUE, otherwise the value 0. */ +int +_cpp_test_assertion (cpp_reader *pfile, unsigned int *value) +{ + cpp_macro *answer; + cpp_hashnode *node = parse_assertion (pfile, T_IF, &answer); + + /* For recovery, an erroneous assertion expression is handled as a + failing assertion. */ + *value = 0; + + if (node) + { + if (node->value.answers) + *value = !answer || *find_answer (node, answer); + } + else if (pfile->cur_token[-1].type == CPP_EOF) + _cpp_backup_tokens (pfile, 1); + + /* We don't commit the memory for the answer - it's temporary only. */ + return node == 0; +} + +/* Handle #assert. */ +static void +do_assert (cpp_reader *pfile) +{ + cpp_macro *answer; + cpp_hashnode *node = parse_assertion (pfile, T_ASSERT, &answer); + + if (node) + { + /* Place the new answer in the answer list. First check there + is not a duplicate. */ + if (*find_answer (node, answer)) + { + cpp_error (pfile, CPP_DL_WARNING, "\"%s\" re-asserted", + NODE_NAME (node) + 1); + return; + } + + /* Commit or allocate storage for the answer. */ + answer = (cpp_macro *)_cpp_commit_buff + (pfile, sizeof (cpp_macro) - sizeof (cpp_token) + + sizeof (cpp_token) * answer->count); + + /* Chain into the list. */ + answer->parm.next = node->value.answers; + node->value.answers = answer; + + check_eol (pfile, false); + } +} + +/* Handle #unassert. */ +static void +do_unassert (cpp_reader *pfile) +{ + cpp_macro *answer; + cpp_hashnode *node = parse_assertion (pfile, T_UNASSERT, &answer); + + /* It isn't an error to #unassert something that isn't asserted. */ + if (node) + { + if (answer) + { + cpp_macro **p = find_answer (node, answer); + + /* Remove the assert from the list. */ + if (cpp_macro *temp = *p) + *p = temp->parm.next; + + check_eol (pfile, false); + } + else + _cpp_free_definition (node); + } + + /* We don't commit the memory for the answer - it's temporary only. */ +} + +/* These are for -D, -U, -A. */ + +/* Process the string STR as if it appeared as the body of a #define. + If STR is just an identifier, define it with value 1. + If STR has anything after the identifier, then it should + be identifier=definition. */ +void +cpp_define (cpp_reader *pfile, const char *str) +{ + char *buf; + const char *p; + size_t count; + + /* Copy the entire option so we can modify it. + Change the first "=" in the string to a space. If there is none, + tack " 1" on the end. */ + + count = strlen (str); + buf = (char *) alloca (count + 3); + memcpy (buf, str, count); + + p = strchr (str, '='); + if (p) + buf[p - str] = ' '; + else + { + buf[count++] = ' '; + buf[count++] = '1'; + } + buf[count] = '\n'; + + run_directive (pfile, T_DEFINE, buf, count); +} + +/* Like cpp_define, but does not warn about unused macro. */ +void +cpp_define_unused (cpp_reader *pfile, const char *str) +{ + unsigned char warn_unused_macros = CPP_OPTION (pfile, warn_unused_macros); + CPP_OPTION (pfile, warn_unused_macros) = 0; + cpp_define (pfile, str); + CPP_OPTION (pfile, warn_unused_macros) = warn_unused_macros; +} + +/* Use to build macros to be run through cpp_define() as + described above. + Example: cpp_define_formatted (pfile, "MACRO=%d", value); */ + +void +cpp_define_formatted (cpp_reader *pfile, const char *fmt, ...) +{ + char *ptr; + + va_list ap; + va_start (ap, fmt); + ptr = xvasprintf (fmt, ap); + va_end (ap); + + cpp_define (pfile, ptr); + free (ptr); +} + +/* Like cpp_define_formatted, but does not warn about unused macro. */ +void +cpp_define_formatted_unused (cpp_reader *pfile, const char *fmt, ...) +{ + char *ptr; + + va_list ap; + va_start (ap, fmt); + ptr = xvasprintf (fmt, ap); + va_end (ap); + + cpp_define_unused (pfile, ptr); + free (ptr); +} + +/* Slight variant of the above for use by initialize_builtins. */ +void +_cpp_define_builtin (cpp_reader *pfile, const char *str) +{ + size_t len = strlen (str); + char *buf = (char *) alloca (len + 1); + memcpy (buf, str, len); + buf[len] = '\n'; + run_directive (pfile, T_DEFINE, buf, len); +} + +/* Process MACRO as if it appeared as the body of an #undef. */ +void +cpp_undef (cpp_reader *pfile, const char *macro) +{ + size_t len = strlen (macro); + char *buf = (char *) alloca (len + 1); + memcpy (buf, macro, len); + buf[len] = '\n'; + run_directive (pfile, T_UNDEF, buf, len); +} + +/* Replace a previous definition DEF of the macro STR. If DEF is NULL, + or first element is zero, then the macro should be undefined. */ +static void +cpp_pop_definition (cpp_reader *pfile, struct def_pragma_macro *c) +{ + cpp_hashnode *node = _cpp_lex_identifier (pfile, c->name); + if (node == NULL) + return; + + if (pfile->cb.before_define) + pfile->cb.before_define (pfile); + + if (cpp_macro_p (node)) + { + if (pfile->cb.undef) + pfile->cb.undef (pfile, pfile->directive_line, node); + if (CPP_OPTION (pfile, warn_unused_macros)) + _cpp_warn_if_unused_macro (pfile, node, NULL); + _cpp_free_definition (node); + } + + if (c->is_undef) + return; + if (c->is_builtin) + { + _cpp_restore_special_builtin (pfile, c); + return; + } + + { + size_t namelen; + const uchar *dn; + cpp_hashnode *h = NULL; + cpp_buffer *nbuf; + + namelen = ustrcspn (c->definition, "( \n"); + h = cpp_lookup (pfile, c->definition, namelen); + dn = c->definition + namelen; + + nbuf = cpp_push_buffer (pfile, dn, ustrchr (dn, '\n') - dn, true); + if (nbuf != NULL) + { + _cpp_clean_line (pfile); + nbuf->sysp = 1; + if (!_cpp_create_definition (pfile, h)) + abort (); + _cpp_pop_buffer (pfile); + } + else + abort (); + h->value.macro->line = c->line; + h->value.macro->syshdr = c->syshdr; + h->value.macro->used = c->used; + } +} + +/* Process the string STR as if it appeared as the body of a #assert. */ +void +cpp_assert (cpp_reader *pfile, const char *str) +{ + handle_assertion (pfile, str, T_ASSERT); +} + +/* Process STR as if it appeared as the body of an #unassert. */ +void +cpp_unassert (cpp_reader *pfile, const char *str) +{ + handle_assertion (pfile, str, T_UNASSERT); +} + +/* Common code for cpp_assert (-A) and cpp_unassert (-A-). */ +static void +handle_assertion (cpp_reader *pfile, const char *str, int type) +{ + size_t count = strlen (str); + const char *p = strchr (str, '='); + + /* Copy the entire option so we can modify it. Change the first + "=" in the string to a '(', and tack a ')' on the end. */ + char *buf = (char *) alloca (count + 2); + + memcpy (buf, str, count); + if (p) + { + buf[p - str] = '('; + buf[count++] = ')'; + } + buf[count] = '\n'; + str = buf; + + run_directive (pfile, type, str, count); +} + +/* The options structure. */ +cpp_options * +cpp_get_options (cpp_reader *pfile) +{ + return &pfile->opts; +} + +/* The callbacks structure. */ +cpp_callbacks * +cpp_get_callbacks (cpp_reader *pfile) +{ + return &pfile->cb; +} + +/* Copy the given callbacks structure to our own. */ +void +cpp_set_callbacks (cpp_reader *pfile, cpp_callbacks *cb) +{ + pfile->cb = *cb; +} + +/* The narrow character set identifier. */ +const char * +cpp_get_narrow_charset_name (cpp_reader *pfile) +{ + return pfile->narrow_cset_desc.to; +} + +/* The wide character set identifier. */ +const char * +cpp_get_wide_charset_name (cpp_reader *pfile) +{ + return pfile->wide_cset_desc.to; +} + +/* The dependencies structure. (Creates one if it hasn't already been.) */ +class mkdeps * +cpp_get_deps (cpp_reader *pfile) +{ + if (!pfile->deps && CPP_OPTION (pfile, deps.style) != DEPS_NONE) + pfile->deps = deps_init (); + return pfile->deps; +} + +/* Push a new buffer on the buffer stack. Returns the new buffer; it + doesn't fail. It does not generate a file change call back; that + is the responsibility of the caller. */ +cpp_buffer * +cpp_push_buffer (cpp_reader *pfile, const uchar *buffer, size_t len, + int from_stage3) +{ + cpp_buffer *new_buffer = XOBNEW (&pfile->buffer_ob, cpp_buffer); + + /* Clears, amongst other things, if_stack and mi_cmacro. */ + memset (new_buffer, 0, sizeof (cpp_buffer)); + + new_buffer->next_line = new_buffer->buf = buffer; + new_buffer->rlimit = buffer + len; + new_buffer->from_stage3 = from_stage3; + new_buffer->prev = pfile->buffer; + new_buffer->need_line = true; + + pfile->buffer = new_buffer; + + return new_buffer; +} + +/* Pops a single buffer, with a file change call-back if appropriate. + Then pushes the next -include file, if any remain. */ +void +_cpp_pop_buffer (cpp_reader *pfile) +{ + cpp_buffer *buffer = pfile->buffer; + struct _cpp_file *inc = buffer->file; + struct if_stack *ifs; + const unsigned char *to_free; + + /* Walk back up the conditional stack till we reach its level at + entry to this file, issuing error messages. */ + for (ifs = buffer->if_stack; ifs; ifs = ifs->next) + cpp_error_with_line (pfile, CPP_DL_ERROR, ifs->line, 0, + "unterminated #%s", dtable[ifs->type].name); + + /* In case of a missing #endif. */ + pfile->state.skipping = 0; + + /* _cpp_do_file_change expects pfile->buffer to be the new one. */ + pfile->buffer = buffer->prev; + + to_free = buffer->to_free; + free (buffer->notes); + + /* Free the buffer object now; we may want to push a new buffer + in _cpp_push_next_include_file. */ + obstack_free (&pfile->buffer_ob, buffer); + + if (inc) + { + _cpp_pop_file_buffer (pfile, inc, to_free); + + _cpp_do_file_change (pfile, LC_LEAVE, 0, 0, 0); + } + else if (to_free) + free ((void *)to_free); +} + +/* Enter all recognized directives in the hash table. */ +void +_cpp_init_directives (cpp_reader *pfile) +{ + for (int i = 0; i < N_DIRECTIVES; i++) + { + cpp_hashnode *node = cpp_lookup (pfile, dtable[i].name, dtable[i].length); + node->is_directive = 1; + node->directive_index = i; + } +} + +/* Extract header file from a bracket include. Parsing starts after '<'. + The string is malloced and must be freed by the caller. */ +char * +_cpp_bracket_include(cpp_reader *pfile) +{ + return glue_header_name (pfile); +} + + +//-------------------------------------------------------------------------------- + + +extern bool _assign_handler(cpp_reader *pfile, cpp_hashnode *node); + +const char * +cpp_token_as_text(const cpp_token *token) +{ + static char buffer[128]; + + switch (token->type) + { + case CPP_NAME: + snprintf(buffer, sizeof(buffer), "identifier '%s'", + NODE_NAME(token->val.node.node)); + break; + + case CPP_NUMBER: + case CPP_STRING: + case CPP_CHAR: + case CPP_HEADER_NAME: + snprintf(buffer, sizeof(buffer), "'%.*s'", + token->val.str.len, + token->val.str.text); + break; + + case CPP_EOF: + return ""; + case CPP_OTHER: + return ""; + case CPP_OPEN_PAREN: + return "'('"; + case CPP_CLOSE_PAREN: + return "')'"; + case CPP_COMMA: + return "','"; + case CPP_SEMICOLON: + return "';'"; + case CPP_PLUS: + return "'+'"; + case CPP_MINUS: + return "'-'"; + case CPP_MULT: + return "'*'"; + case CPP_DIV: + return "'/'"; + case CPP_MOD: + return "'%'"; + // ... handle other symbolic types as needed ... + + default: + snprintf(buffer, sizeof(buffer), "", token->type); + break; + } + + return buffer; +} + +static void do_assign(cpp_reader *pfile){ + const cpp_token *tok = _cpp_lex_token(pfile); + if(tok->type != CPP_OPEN_PAREN){ + cpp_error_with_line( + pfile + ,CPP_DL_ERROR + ,tok->src_loc + ,0 + ,"expected '(' before name ,but found: %s" + ,cpp_token_as_text(tok) + ); + return; + } + + tok = _cpp_lex_token(pfile); + if(tok->type != CPP_NAME){ + cpp_error_with_line( + pfile + ,CPP_DL_ERROR + ,tok->src_loc + ,0 + ,"expected macro name identifier ,but found: %s" + ,cpp_token_as_text(tok) + ); + return; + } + cpp_hashnode *node = tok->val.node.node; + + tok = _cpp_lex_token(pfile); + if(tok->type != CPP_CLOSE_PAREN){ + cpp_error_with_line( + pfile + ,CPP_DL_ERROR + ,tok->src_loc + ,0 + ,"expected ')' after macro name ,but found: %s" + ,cpp_token_as_text(tok) + ); + return; + } + + if(node){ + /* If we have been requested to expand comments into macros, + then re-enable saving of comments. */ + pfile->state.save_comments = + ! CPP_OPTION (pfile ,discard_comments_in_macro_exp); + + if (pfile->cb.before_define) + pfile->cb.before_define (pfile); + + if (_assign_handler (pfile ,node)) + if (pfile->cb.define) + pfile->cb.define (pfile ,pfile->directive_line ,node); + + node->flags &= ~NODE_USED; + } +} + + +#if 0 +static void +do_assign(cpp_reader *pfile){ + + // cpp_hashnode *node = lex_macro_node (pfile, true); + const cpp_token *tok = _cpp_lex_token(pfile); + if (tok->type != CPP_OPEN_PAREN) { + cpp_error_with_line( + pfile, + CPP_DL_ERROR, + tok->src_loc, + 0, + "expected '(' before name, but found: %s" + cpp_token_as_text(tok); + ); + return; + } + + tok = _cpp_lex_token(pfile); + if (tok->type != CPP_NAME) { + cpp_error_with_line( + pfile, + CPP_DL_ERROR, + tok->src_loc, + 0, + "expected macro name identifier, but found: type=%d text='%.*s'", + tok->type, + tok->val.str.len, + tok->val.str.text + ); + return; + } + cpp_hashnode *node = tok->val.node.node; + + tok = _cpp_lex_token(pfile); + if (tok->type != CPP_CLOSE_PAREN) { + cpp_error_with_line( + pfile, + CPP_DL_ERROR, + tok->src_loc, + 0, + "expected ')' after macro name, but found: type=%d text='%.*s'", + tok->type, + tok->val.str.len, + tok->val.str.text + ); + return; + } + + if (node) + { + /* If we have been requested to expand comments into macros, + then re-enable saving of comments. */ + pfile->state.save_comments = + ! CPP_OPTION (pfile, discard_comments_in_macro_exp); + + if (pfile->cb.before_define) + pfile->cb.before_define (pfile); + + if (_assign_handler (pfile, node)) + if (pfile->cb.define) + pfile->cb.define (pfile, pfile->directive_line, node); + + node->flags &= ~NODE_USED; + } +} +#endif + +#if 0 + +cpp_token * +assign_get_name(cpp_reader *pfile){ + // const cpp_token *name_token = cpp_get_token(pfile); + const cpp_token *name_token = _cpp_lex_token(pfile); + + cpp_warning_with_line( + pfile, + CPP_W_NONE, + name_token->src_loc, + 0, + "3 assign name is being set to: %.*s", + name_token->val.str.len, + name_token->val.str.text + ); + + if (name_token->type != CPP_NAME) { + cpp_error_with_line( + pfile, + CPP_DL_ERROR, + name_token->src_loc, + 0, + "First argument to #assign must be a macro name, instead found: %.*s", + name_token->val.str.len, + name_token->val.str.text + ); + return NULL; + } + + // Export this into the wider context + cpp_token *copy = (cpp_token *) _cpp_reserve_room(pfile, 0, sizeof(cpp_token)); + *copy = *name_token; + return copy; +} + +static void +do_assign(cpp_reader *pfile) +{ + cpp_token *name_token = assign_get_name(pfile); + if (!name_token) { + return; + } + + cpp_macro *macro = _cpp_new_macro( + pfile, + cmk_macro, + _cpp_reserve_room(pfile, 0, sizeof(cpp_macro)) + ); + + macro->fun_like = 0; + macro->paramc = 0; + macro->variadic = 0; + macro->count = 1; + macro->used = 1; + + cpp_token *value_token = ¯o->exp.tokens[0]; + value_token->type = CPP_NUMBER; + value_token->val.str.text = (const unsigned char *) "42"; + value_token->val.str.len = 2; + value_token->flags = 0; + + cpp_hashnode *node = name_token->val.node.node; + node->type = NT_USER_MACRO; + node->value.macro = macro; + + _cpp_mark_macro_used(node); + cpp_warning(pfile, CPP_W_NONE, "Assigned macro %s as 42", NODE_NAME(node)); + +} + +#endif diff --git "a/script_gcc_min-12\360\237\226\211/macro.cc" "b/script_gcc_min-12\360\237\226\211/macro.cc" new file mode 100644 index 0000000..f12b3e5 --- /dev/null +++ "b/script_gcc_min-12\360\237\226\211/macro.cc" @@ -0,0 +1,4314 @@ +/* Part of CPP library. (Macro and #define handling.) + Copyright (C) 1986-2022 Free Software Foundation, Inc. + Written by Per Bothner, 1994. + Based on CCCP program by Paul Rubin, June 1986 + Adapted to ANSI C, Richard Stallman, Jan 1987 + +This program is free software; you can redistribute it and/or modify it +under the terms of the GNU General Public License as published by the +Free Software Foundation; either version 3, or (at your option) any +later version. + +This program is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; see the file COPYING3. If not see +. + + In other words, you are welcome to use, share and improve this program. + You are forbidden to forbid anyone else to use, share and improve + what you give them. Help stamp out software-hoarding! */ + +#include "config.h" +#include "system.h" +#include "cpplib.h" +#include "internal.h" + +typedef struct macro_arg macro_arg; +/* This structure represents the tokens of a macro argument. These + tokens can be macro themselves, in which case they can be either + expanded or unexpanded. When they are expanded, this data + structure keeps both the expanded and unexpanded forms. */ +struct macro_arg +{ + const cpp_token **first; /* First token in unexpanded argument. */ + const cpp_token **expanded; /* Macro-expanded argument. */ + const cpp_token *stringified; /* Stringified argument. */ + unsigned int count; /* # of tokens in argument. */ + unsigned int expanded_count; /* # of tokens in expanded argument. */ + location_t *virt_locs; /* Where virtual locations for + unexpanded tokens are stored. */ + location_t *expanded_virt_locs; /* Where virtual locations for + expanded tokens are + stored. */ +}; + +/* The kind of macro tokens which the instance of + macro_arg_token_iter is supposed to iterate over. */ +enum macro_arg_token_kind { + MACRO_ARG_TOKEN_NORMAL, + /* This is a macro argument token that got transformed into a string + literal, e.g. #foo. */ + MACRO_ARG_TOKEN_STRINGIFIED, + /* This is a token resulting from the expansion of a macro + argument that was itself a macro. */ + MACRO_ARG_TOKEN_EXPANDED +}; + +/* An iterator over tokens coming from a function-like macro + argument. */ +typedef struct macro_arg_token_iter macro_arg_token_iter; +struct macro_arg_token_iter +{ + /* Whether or not -ftrack-macro-expansion is used. */ + bool track_macro_exp_p; + /* The kind of token over which we are supposed to iterate. */ + enum macro_arg_token_kind kind; + /* A pointer to the current token pointed to by the iterator. */ + const cpp_token **token_ptr; + /* A pointer to the "full" location of the current token. If + -ftrack-macro-expansion is used this location tracks loci across + macro expansion. */ + const location_t *location_ptr; +#if CHECKING_P + /* The number of times the iterator went forward. This useful only + when checking is enabled. */ + size_t num_forwards; +#endif +}; + +/* Saved data about an identifier being used as a macro argument + name. */ +struct macro_arg_saved_data { + /* The canonical (UTF-8) spelling of this identifier. */ + cpp_hashnode *canonical_node; + /* The previous value & type of this identifier. */ + union _cpp_hashnode_value value; + node_type type; +}; + +static const char *vaopt_paste_error = + N_("'##' cannot appear at either end of __VA_OPT__"); + +static void expand_arg (cpp_reader *, macro_arg *); + +/* A class for tracking __VA_OPT__ state while iterating over a + sequence of tokens. This is used during both macro definition and + expansion. */ +class vaopt_state { + + public: + + enum update_type + { + ERROR, + DROP, + INCLUDE, + BEGIN, + END + }; + + /* Initialize the state tracker. ANY_ARGS is true if variable + arguments were provided to the macro invocation. */ + vaopt_state (cpp_reader *pfile, bool is_variadic, macro_arg *arg) + : m_pfile (pfile), + m_arg (arg), + m_variadic (is_variadic), + m_last_was_paste (false), + m_stringify (false), + m_state (0), + m_paste_location (0), + m_location (0), + m_update (ERROR) + { + } + + /* Given a token, update the state of this tracker and return a + boolean indicating whether the token should be be included in the + expansion. */ + update_type update (const cpp_token *token) + { + /* If the macro isn't variadic, just don't bother. */ + if (!m_variadic) + return INCLUDE; + + if (token->type == CPP_NAME + && token->val.node.node == m_pfile->spec_nodes.n__VA_OPT__) + { + if (m_state > 0) + { + cpp_error_at (m_pfile, CPP_DL_ERROR, token->src_loc, + "__VA_OPT__ may not appear in a __VA_OPT__"); + return ERROR; + } + ++m_state; + m_location = token->src_loc; + m_stringify = (token->flags & STRINGIFY_ARG) != 0; + return BEGIN; + } + else if (m_state == 1) + { + if (token->type != CPP_OPEN_PAREN) + { + cpp_error_at (m_pfile, CPP_DL_ERROR, m_location, + "__VA_OPT__ must be followed by an " + "open parenthesis"); + return ERROR; + } + ++m_state; + if (m_update == ERROR) + { + if (m_arg == NULL) + m_update = INCLUDE; + else + { + m_update = DROP; + if (!m_arg->expanded) + expand_arg (m_pfile, m_arg); + for (unsigned idx = 0; idx < m_arg->expanded_count; ++idx) + if (m_arg->expanded[idx]->type != CPP_PADDING) + { + m_update = INCLUDE; + break; + } + } + } + return DROP; + } + else if (m_state >= 2) + { + if (m_state == 2 && token->type == CPP_PASTE) + { + cpp_error_at (m_pfile, CPP_DL_ERROR, token->src_loc, + vaopt_paste_error); + return ERROR; + } + /* Advance states before further considering this token, in + case we see a close paren immediately after the open + paren. */ + if (m_state == 2) + ++m_state; + + bool was_paste = m_last_was_paste; + m_last_was_paste = false; + if (token->type == CPP_PASTE) + { + m_last_was_paste = true; + m_paste_location = token->src_loc; + } + else if (token->type == CPP_OPEN_PAREN) + ++m_state; + else if (token->type == CPP_CLOSE_PAREN) + { + --m_state; + if (m_state == 2) + { + /* Saw the final paren. */ + m_state = 0; + + if (was_paste) + { + cpp_error_at (m_pfile, CPP_DL_ERROR, token->src_loc, + vaopt_paste_error); + return ERROR; + } + + return END; + } + } + return m_update; + } + + /* Nothing to do with __VA_OPT__. */ + return INCLUDE; + } + + /* Ensure that any __VA_OPT__ was completed. If ok, return true. + Otherwise, issue an error and return false. */ + bool completed () + { + if (m_variadic && m_state != 0) + cpp_error_at (m_pfile, CPP_DL_ERROR, m_location, + "unterminated __VA_OPT__"); + return m_state == 0; + } + + /* Return true for # __VA_OPT__. */ + bool stringify () const + { + return m_stringify; + } + + private: + + /* The cpp_reader. */ + cpp_reader *m_pfile; + + /* The __VA_ARGS__ argument. */ + macro_arg *m_arg; + + /* True if the macro is variadic. */ + bool m_variadic; + /* If true, the previous token was ##. This is used to detect when + a paste occurs at the end of the sequence. */ + bool m_last_was_paste; + /* True for #__VA_OPT__. */ + bool m_stringify; + + /* The state variable: + 0 means not parsing + 1 means __VA_OPT__ seen, looking for "(" + 2 means "(" seen (so the next token can't be "##") + >= 3 means looking for ")", the number encodes the paren depth. */ + int m_state; + + /* The location of the paste token. */ + location_t m_paste_location; + + /* Location of the __VA_OPT__ token. */ + location_t m_location; + + /* If __VA_ARGS__ substitutes to no preprocessing tokens, + INCLUDE, otherwise DROP. ERROR when unknown yet. */ + update_type m_update; +}; + +/* Macro expansion. */ + +static cpp_macro *get_deferred_or_lazy_macro (cpp_reader *, cpp_hashnode *, + location_t); +static int enter_macro_context (cpp_reader *, cpp_hashnode *, + const cpp_token *, location_t); +static int builtin_macro (cpp_reader *, cpp_hashnode *, + location_t, location_t); +static void push_ptoken_context (cpp_reader *, cpp_hashnode *, _cpp_buff *, + const cpp_token **, unsigned int); +static void push_extended_tokens_context (cpp_reader *, cpp_hashnode *, + _cpp_buff *, location_t *, + const cpp_token **, unsigned int); +static _cpp_buff *collect_args (cpp_reader *, const cpp_hashnode *, + _cpp_buff **, unsigned *); +static cpp_context *next_context (cpp_reader *); +static const cpp_token *padding_token (cpp_reader *, const cpp_token *); +static const cpp_token *new_string_token (cpp_reader *, uchar *, unsigned int); +static const cpp_token *stringify_arg (cpp_reader *, const cpp_token **, + unsigned int); +static void paste_all_tokens (cpp_reader *, const cpp_token *); +static bool paste_tokens (cpp_reader *, location_t, + const cpp_token **, const cpp_token *); +static void alloc_expanded_arg_mem (cpp_reader *, macro_arg *, size_t); +static void ensure_expanded_arg_room (cpp_reader *, macro_arg *, size_t, size_t *); +static void delete_macro_args (_cpp_buff*, unsigned num_args); +static void set_arg_token (macro_arg *, const cpp_token *, + location_t, size_t, + enum macro_arg_token_kind, + bool); +static const location_t *get_arg_token_location (const macro_arg *, + enum macro_arg_token_kind); +static const cpp_token **arg_token_ptr_at (const macro_arg *, + size_t, + enum macro_arg_token_kind, + location_t **virt_location); + +static void macro_arg_token_iter_init (macro_arg_token_iter *, bool, + enum macro_arg_token_kind, + const macro_arg *, + const cpp_token **); +static const cpp_token *macro_arg_token_iter_get_token +(const macro_arg_token_iter *it); +static location_t macro_arg_token_iter_get_location +(const macro_arg_token_iter *); +static void macro_arg_token_iter_forward (macro_arg_token_iter *); +static _cpp_buff *tokens_buff_new (cpp_reader *, size_t, + location_t **); +static size_t tokens_buff_count (_cpp_buff *); +static const cpp_token **tokens_buff_last_token_ptr (_cpp_buff *); +static inline const cpp_token **tokens_buff_put_token_to (const cpp_token **, + location_t *, + const cpp_token *, + location_t, + location_t, + const line_map_macro *, + unsigned int); + +static const cpp_token **tokens_buff_add_token (_cpp_buff *, + location_t *, + const cpp_token *, + location_t, + location_t, + const line_map_macro *, + unsigned int); +static inline void tokens_buff_remove_last_token (_cpp_buff *); +static void replace_args (cpp_reader *, cpp_hashnode *, cpp_macro *, + macro_arg *, location_t); +static _cpp_buff *funlike_invocation_p (cpp_reader *, cpp_hashnode *, + _cpp_buff **, unsigned *); +static cpp_macro *create_iso_definition (cpp_reader *); + +/* #define directive parsing and handling. */ + +static cpp_macro *lex_expansion_token (cpp_reader *, cpp_macro *); +static bool parse_params (cpp_reader *, unsigned *, bool *); +static void check_trad_stringification (cpp_reader *, const cpp_macro *, + const cpp_string *); +static bool reached_end_of_context (cpp_context *); +static void consume_next_token_from_context (cpp_reader *pfile, + const cpp_token **, + location_t *); +static const cpp_token* cpp_get_token_1 (cpp_reader *, location_t *); + +static cpp_hashnode* macro_of_context (cpp_context *context); + +/* Statistical counter tracking the number of macros that got + expanded. */ +unsigned num_expanded_macros_counter = 0; +/* Statistical counter tracking the total number tokens resulting + from macro expansion. */ +unsigned num_macro_tokens_counter = 0; + +/* Wrapper around cpp_get_token to skip CPP_PADDING tokens + and not consume CPP_EOF. */ +static const cpp_token * +cpp_get_token_no_padding (cpp_reader *pfile) +{ + for (;;) + { + const cpp_token *ret = cpp_peek_token (pfile, 0); + if (ret->type == CPP_EOF) + return ret; + ret = cpp_get_token (pfile); + if (ret->type != CPP_PADDING) + return ret; + } +} + +/* Handle meeting "__has_include" builtin macro. */ + +static int +builtin_has_include (cpp_reader *pfile, cpp_hashnode *op, bool has_next) +{ + int result = 0; + + if (!pfile->state.in_directive) + cpp_error (pfile, CPP_DL_ERROR, + "\"%s\" used outside of preprocessing directive", + NODE_NAME (op)); + + pfile->state.angled_headers = true; + const cpp_token *token = cpp_get_token_no_padding (pfile); + bool paren = token->type == CPP_OPEN_PAREN; + if (paren) + token = cpp_get_token_no_padding (pfile); + else + cpp_error (pfile, CPP_DL_ERROR, + "missing '(' before \"%s\" operand", NODE_NAME (op)); + pfile->state.angled_headers = false; + + bool bracket = token->type != CPP_STRING; + char *fname = NULL; + if (token->type == CPP_STRING || token->type == CPP_HEADER_NAME) + { + fname = XNEWVEC (char, token->val.str.len - 1); + memcpy (fname, token->val.str.text + 1, token->val.str.len - 2); + fname[token->val.str.len - 2] = '\0'; + } + else if (token->type == CPP_LESS) + fname = _cpp_bracket_include (pfile); + else + cpp_error (pfile, CPP_DL_ERROR, + "operator \"%s\" requires a header-name", NODE_NAME (op)); + + if (fname) + { + /* Do not do the lookup if we're skipping, that's unnecessary + IO. */ + if (!pfile->state.skip_eval + && _cpp_has_header (pfile, fname, bracket, + has_next ? IT_INCLUDE_NEXT : IT_INCLUDE)) + result = 1; + + XDELETEVEC (fname); + } + + if (paren + && cpp_get_token_no_padding (pfile)->type != CPP_CLOSE_PAREN) + cpp_error (pfile, CPP_DL_ERROR, + "missing ')' after \"%s\" operand", NODE_NAME (op)); + + return result; +} + +/* Emits a warning if NODE is a macro defined in the main file that + has not been used. */ +int +_cpp_warn_if_unused_macro (cpp_reader *pfile, cpp_hashnode *node, + void *v ATTRIBUTE_UNUSED) +{ + if (cpp_user_macro_p (node)) + { + cpp_macro *macro = node->value.macro; + + if (!macro->used + && MAIN_FILE_P (linemap_check_ordinary + (linemap_lookup (pfile->line_table, + macro->line)))) + cpp_warning_with_line (pfile, CPP_W_UNUSED_MACROS, macro->line, 0, + "macro \"%s\" is not used", NODE_NAME (node)); + } + + return 1; +} + +/* Allocates and returns a CPP_STRING token, containing TEXT of length + LEN, after null-terminating it. TEXT must be in permanent storage. */ +static const cpp_token * +new_string_token (cpp_reader *pfile, unsigned char *text, unsigned int len) +{ + cpp_token *token = _cpp_temp_token (pfile); + + text[len] = '\0'; + token->type = CPP_STRING; + token->val.str.len = len; + token->val.str.text = text; + token->flags = 0; + return token; +} + +static const char * const monthnames[] = +{ + "Jan", "Feb", "Mar", "Apr", "May", "Jun", + "Jul", "Aug", "Sep", "Oct", "Nov", "Dec" +}; + +/* Helper function for builtin_macro. Returns the text generated by + a builtin macro. */ +const uchar * +_cpp_builtin_macro_text (cpp_reader *pfile, cpp_hashnode *node, + location_t loc) +{ + const uchar *result = NULL; + linenum_type number = 1; + + switch (node->value.builtin) + { + default: + cpp_error (pfile, CPP_DL_ICE, "invalid built-in macro \"%s\"", + NODE_NAME (node)); + break; + + case BT_TIMESTAMP: + { + if (CPP_OPTION (pfile, warn_date_time)) + cpp_warning (pfile, CPP_W_DATE_TIME, "macro \"%s\" might prevent " + "reproducible builds", NODE_NAME (node)); + + cpp_buffer *pbuffer = cpp_get_buffer (pfile); + if (pbuffer->timestamp == NULL) + { + /* Initialize timestamp value of the assotiated file. */ + struct _cpp_file *file = cpp_get_file (pbuffer); + if (file) + { + /* Generate __TIMESTAMP__ string, that represents + the date and time of the last modification + of the current source file. The string constant + looks like "Sun Sep 16 01:03:52 1973". */ + struct tm *tb = NULL; + struct stat *st = _cpp_get_file_stat (file); + if (st) + tb = localtime (&st->st_mtime); + if (tb) + { + char *str = asctime (tb); + size_t len = strlen (str); + unsigned char *buf = _cpp_unaligned_alloc (pfile, len + 2); + buf[0] = '"'; + strcpy ((char *) buf + 1, str); + buf[len] = '"'; + pbuffer->timestamp = buf; + } + else + { + cpp_errno (pfile, CPP_DL_WARNING, + "could not determine file timestamp"); + pbuffer->timestamp = UC"\"??? ??? ?? ??:??:?? ????\""; + } + } + } + result = pbuffer->timestamp; + } + break; + case BT_FILE: + case BT_FILE_NAME: + case BT_BASE_FILE: + { + unsigned int len; + const char *name; + uchar *buf; + + if (node->value.builtin == BT_FILE + || node->value.builtin == BT_FILE_NAME) + { + name = linemap_get_expansion_filename (pfile->line_table, + pfile->line_table->highest_line); + if ((node->value.builtin == BT_FILE_NAME) && name) + name = lbasename (name); + } + else + { + name = _cpp_get_file_name (pfile->main_file); + if (!name) + abort (); + } + if (pfile->cb.remap_filename) + name = pfile->cb.remap_filename (name); + len = strlen (name); + buf = _cpp_unaligned_alloc (pfile, len * 2 + 3); + result = buf; + *buf = '"'; + buf = cpp_quote_string (buf + 1, (const unsigned char *) name, len); + *buf++ = '"'; + *buf = '\0'; + } + break; + + case BT_INCLUDE_LEVEL: + /* The line map depth counts the primary source as level 1, but + historically __INCLUDE_DEPTH__ has called the primary source + level 0. */ + number = pfile->line_table->depth - 1; + break; + + case BT_SPECLINE: + /* If __LINE__ is embedded in a macro, it must expand to the + line of the macro's invocation, not its definition. + Otherwise things like assert() will not work properly. + See WG14 N1911, WG21 N4220 sec 6.5, and PR 61861. */ + if (CPP_OPTION (pfile, traditional)) + loc = pfile->line_table->highest_line; + else + loc = linemap_resolve_location (pfile->line_table, loc, + LRK_MACRO_EXPANSION_POINT, NULL); + number = linemap_get_expansion_line (pfile->line_table, loc); + break; + + /* __STDC__ has the value 1 under normal circumstances. + However, if (a) we are in a system header, (b) the option + stdc_0_in_system_headers is true (set by target config), and + (c) we are not in strictly conforming mode, then it has the + value 0. (b) and (c) are already checked in cpp_init_builtins. */ + case BT_STDC: + if (_cpp_in_system_header (pfile)) + number = 0; + else + number = 1; + break; + + case BT_DATE: + case BT_TIME: + if (CPP_OPTION (pfile, warn_date_time)) + cpp_warning (pfile, CPP_W_DATE_TIME, "macro \"%s\" might prevent " + "reproducible builds", NODE_NAME (node)); + if (pfile->date == NULL) + { + /* Allocate __DATE__ and __TIME__ strings from permanent + storage. We only do this once, and don't generate them + at init time, because time() and localtime() are very + slow on some systems. */ + time_t tt; + auto kind = cpp_get_date (pfile, &tt); + + if (kind == CPP_time_kind::UNKNOWN) + { + cpp_errno (pfile, CPP_DL_WARNING, + "could not determine date and time"); + + pfile->date = UC"\"??? ?? ????\""; + pfile->time = UC"\"??:??:??\""; + } + else + { + struct tm *tb = (kind == CPP_time_kind::FIXED + ? gmtime : localtime) (&tt); + + pfile->date = _cpp_unaligned_alloc (pfile, + sizeof ("\"Oct 11 1347\"")); + sprintf ((char *) pfile->date, "\"%s %2d %4d\"", + monthnames[tb->tm_mon], tb->tm_mday, + tb->tm_year + 1900); + + pfile->time = _cpp_unaligned_alloc (pfile, + sizeof ("\"12:34:56\"")); + sprintf ((char *) pfile->time, "\"%02d:%02d:%02d\"", + tb->tm_hour, tb->tm_min, tb->tm_sec); + } + } + + if (node->value.builtin == BT_DATE) + result = pfile->date; + else + result = pfile->time; + break; + + case BT_COUNTER: + if (CPP_OPTION (pfile, directives_only) && pfile->state.in_directive) + cpp_error (pfile, CPP_DL_ERROR, + "__COUNTER__ expanded inside directive with -fdirectives-only"); + number = pfile->counter++; + break; + + case BT_HAS_ATTRIBUTE: + number = pfile->cb.has_attribute (pfile, false); + break; + + case BT_HAS_STD_ATTRIBUTE: + number = pfile->cb.has_attribute (pfile, true); + break; + + case BT_HAS_BUILTIN: + number = pfile->cb.has_builtin (pfile); + break; + + case BT_HAS_INCLUDE: + case BT_HAS_INCLUDE_NEXT: + number = builtin_has_include (pfile, node, + node->value.builtin == BT_HAS_INCLUDE_NEXT); + break; + } + + if (result == NULL) + { + /* 21 bytes holds all NUL-terminated unsigned 64-bit numbers. */ + result = _cpp_unaligned_alloc (pfile, 21); + sprintf ((char *) result, "%u", number); + } + + return result; +} + +/* Get an idempotent date. Either the cached value, the value from + source epoch, or failing that, the value from time(2). Use this + during compilation so that every time stamp is the same. */ +CPP_time_kind +cpp_get_date (cpp_reader *pfile, time_t *result) +{ + if (!pfile->time_stamp_kind) + { + int kind = 0; + if (pfile->cb.get_source_date_epoch) + { + /* Try reading the fixed epoch. */ + pfile->time_stamp = pfile->cb.get_source_date_epoch (pfile); + if (pfile->time_stamp != time_t (-1)) + kind = int (CPP_time_kind::FIXED); + } + + if (!kind) + { + /* Pedantically time_t (-1) is a legitimate value for + "number of seconds since the Epoch". It is a silly + time. */ + errno = 0; + pfile->time_stamp = time (nullptr); + /* Annoyingly a library could legally set errno and return a + valid time! Bad library! */ + if (pfile->time_stamp == time_t (-1) && errno) + kind = errno; + else + kind = int (CPP_time_kind::DYNAMIC); + } + + pfile->time_stamp_kind = kind; + } + + *result = pfile->time_stamp; + if (pfile->time_stamp_kind >= 0) + { + errno = pfile->time_stamp_kind; + return CPP_time_kind::UNKNOWN; + } + + return CPP_time_kind (pfile->time_stamp_kind); +} + +/* Convert builtin macros like __FILE__ to a token and push it on the + context stack. Also handles _Pragma, for which a new token may not + be created. Returns 1 if it generates a new token context, 0 to + return the token to the caller. LOC is the location of the expansion + point of the macro. */ +static int +builtin_macro (cpp_reader *pfile, cpp_hashnode *node, + location_t loc, location_t expand_loc) +{ + const uchar *buf; + size_t len; + char *nbuf; + + if (node->value.builtin == BT_PRAGMA) + { + /* Don't interpret _Pragma within directives. The standard is + not clear on this, but to me this makes most sense. + Similarly, don't interpret _Pragma inside expand_args, we might + need to stringize it later on. */ + if (pfile->state.in_directive || pfile->state.ignore__Pragma) + return 0; + + return _cpp_do__Pragma (pfile, loc); + } + + buf = _cpp_builtin_macro_text (pfile, node, expand_loc); + len = ustrlen (buf); + nbuf = (char *) alloca (len + 1); + memcpy (nbuf, buf, len); + nbuf[len]='\n'; + + cpp_push_buffer (pfile, (uchar *) nbuf, len, /* from_stage3 */ true); + _cpp_clean_line (pfile); + + /* Set pfile->cur_token as required by _cpp_lex_direct. */ + pfile->cur_token = _cpp_temp_token (pfile); + cpp_token *token = _cpp_lex_direct (pfile); + /* We should point to the expansion point of the builtin macro. */ + token->src_loc = loc; + if (pfile->context->tokens_kind == TOKENS_KIND_EXTENDED) + { + /* We are tracking tokens resulting from macro expansion. + Create a macro line map and generate a virtual location for + the token resulting from the expansion of the built-in + macro. */ + location_t *virt_locs = NULL; + _cpp_buff *token_buf = tokens_buff_new (pfile, 1, &virt_locs); + const line_map_macro * map = + linemap_enter_macro (pfile->line_table, node, loc, 1); + tokens_buff_add_token (token_buf, virt_locs, token, + pfile->line_table->builtin_location, + pfile->line_table->builtin_location, + map, /*macro_token_index=*/0); + push_extended_tokens_context (pfile, node, token_buf, virt_locs, + (const cpp_token **)token_buf->base, + 1); + } + else + _cpp_push_token_context (pfile, NULL, token, 1); + if (pfile->buffer->cur != pfile->buffer->rlimit) + cpp_error (pfile, CPP_DL_ICE, "invalid built-in macro \"%s\"", + NODE_NAME (node)); + _cpp_pop_buffer (pfile); + + return 1; +} + +/* Copies SRC, of length LEN, to DEST, adding backslashes before all + backslashes and double quotes. DEST must be of sufficient size. + Returns a pointer to the end of the string. */ +uchar * +cpp_quote_string (uchar *dest, const uchar *src, unsigned int len) +{ + while (len--) + { + uchar c = *src++; + + switch (c) + { + case '\n': + /* Naked LF can appear in raw string literals */ + c = 'n'; + /* FALLTHROUGH */ + + case '\\': + case '"': + *dest++ = '\\'; + /* FALLTHROUGH */ + + default: + *dest++ = c; + } + } + + return dest; +} + +/* Convert a token sequence FIRST to FIRST+COUNT-1 to a single string token + according to the rules of the ISO C #-operator. */ +static const cpp_token * +stringify_arg (cpp_reader *pfile, const cpp_token **first, unsigned int count) +{ + unsigned char *dest; + unsigned int i, escape_it, backslash_count = 0; + const cpp_token *source = NULL; + size_t len; + + if (BUFF_ROOM (pfile->u_buff) < 3) + _cpp_extend_buff (pfile, &pfile->u_buff, 3); + dest = BUFF_FRONT (pfile->u_buff); + *dest++ = '"'; + + /* Loop, reading in the argument's tokens. */ + for (i = 0; i < count; i++) + { + const cpp_token *token = first[i]; + + if (token->type == CPP_PADDING) + { + if (source == NULL + || (!(source->flags & PREV_WHITE) + && token->val.source == NULL)) + source = token->val.source; + continue; + } + + escape_it = (token->type == CPP_STRING || token->type == CPP_CHAR + || token->type == CPP_WSTRING || token->type == CPP_WCHAR + || token->type == CPP_STRING32 || token->type == CPP_CHAR32 + || token->type == CPP_STRING16 || token->type == CPP_CHAR16 + || token->type == CPP_UTF8STRING || token->type == CPP_UTF8CHAR + || cpp_userdef_string_p (token->type) + || cpp_userdef_char_p (token->type)); + + /* Room for each char being written in octal, initial space and + final quote and NUL. */ + len = cpp_token_len (token); + if (escape_it) + len *= 4; + len += 3; + + if ((size_t) (BUFF_LIMIT (pfile->u_buff) - dest) < len) + { + size_t len_so_far = dest - BUFF_FRONT (pfile->u_buff); + _cpp_extend_buff (pfile, &pfile->u_buff, len); + dest = BUFF_FRONT (pfile->u_buff) + len_so_far; + } + + /* Leading white space? */ + if (dest - 1 != BUFF_FRONT (pfile->u_buff)) + { + if (source == NULL) + source = token; + if (source->flags & PREV_WHITE) + *dest++ = ' '; + } + source = NULL; + + if (escape_it) + { + _cpp_buff *buff = _cpp_get_buff (pfile, len); + unsigned char *buf = BUFF_FRONT (buff); + len = cpp_spell_token (pfile, token, buf, true) - buf; + dest = cpp_quote_string (dest, buf, len); + _cpp_release_buff (pfile, buff); + } + else + dest = cpp_spell_token (pfile, token, dest, true); + + if (token->type == CPP_OTHER && token->val.str.text[0] == '\\') + backslash_count++; + else + backslash_count = 0; + } + + /* Ignore the final \ of invalid string literals. */ + if (backslash_count & 1) + { + cpp_error (pfile, CPP_DL_WARNING, + "invalid string literal, ignoring final '\\'"); + dest--; + } + + /* Commit the memory, including NUL, and return the token. */ + *dest++ = '"'; + len = dest - BUFF_FRONT (pfile->u_buff); + BUFF_FRONT (pfile->u_buff) = dest + 1; + return new_string_token (pfile, dest - len, len); +} + +/* Try to paste two tokens. On success, return nonzero. In any + case, PLHS is updated to point to the pasted token, which is + guaranteed to not have the PASTE_LEFT flag set. LOCATION is + the virtual location used for error reporting. */ +static bool +paste_tokens (cpp_reader *pfile, location_t location, + const cpp_token **plhs, const cpp_token *rhs) +{ + unsigned char *buf, *end, *lhsend; + cpp_token *lhs; + unsigned int len; + + len = cpp_token_len (*plhs) + cpp_token_len (rhs) + 2; + buf = (unsigned char *) alloca (len); + end = lhsend = cpp_spell_token (pfile, *plhs, buf, true); + + /* Avoid comment headers, since they are still processed in stage 3. + It is simpler to insert a space here, rather than modifying the + lexer to ignore comments in some circumstances. Simply returning + false doesn't work, since we want to clear the PASTE_LEFT flag. */ + if ((*plhs)->type == CPP_DIV && rhs->type != CPP_EQ) + *end++ = ' '; + /* In one obscure case we might see padding here. */ + if (rhs->type != CPP_PADDING) + end = cpp_spell_token (pfile, rhs, end, true); + *end = '\n'; + + cpp_push_buffer (pfile, buf, end - buf, /* from_stage3 */ true); + _cpp_clean_line (pfile); + + /* Set pfile->cur_token as required by _cpp_lex_direct. */ + pfile->cur_token = _cpp_temp_token (pfile); + lhs = _cpp_lex_direct (pfile); + if (pfile->buffer->cur != pfile->buffer->rlimit) + { + location_t saved_loc = lhs->src_loc; + + _cpp_pop_buffer (pfile); + + unsigned char *rhsstart = lhsend; + if ((*plhs)->type == CPP_DIV && rhs->type != CPP_EQ) + rhsstart++; + + /* We have to remove the PASTE_LEFT flag from the old lhs, but + we want to keep the new location. */ + *lhs = **plhs; + *plhs = lhs; + lhs->src_loc = saved_loc; + lhs->flags &= ~PASTE_LEFT; + + /* Mandatory error for all apart from assembler. */ + if (CPP_OPTION (pfile, lang) != CLK_ASM) + cpp_error_with_line (pfile, CPP_DL_ERROR, location, 0, + "pasting \"%.*s\" and \"%.*s\" does not give " + "a valid preprocessing token", + (int) (lhsend - buf), buf, + (int) (end - rhsstart), rhsstart); + return false; + } + + lhs->flags |= (*plhs)->flags & (PREV_WHITE | PREV_FALLTHROUGH); + *plhs = lhs; + _cpp_pop_buffer (pfile); + return true; +} + +/* Handles an arbitrarily long sequence of ## operators, with initial + operand LHS. This implementation is left-associative, + non-recursive, and finishes a paste before handling succeeding + ones. If a paste fails, we back up to the RHS of the failing ## + operator before pushing the context containing the result of prior + successful pastes, with the effect that the RHS appears in the + output stream after the pasted LHS normally. */ +static void +paste_all_tokens (cpp_reader *pfile, const cpp_token *lhs) +{ + const cpp_token *rhs = NULL; + cpp_context *context = pfile->context; + location_t virt_loc = 0; + + /* We are expanding a macro and we must have been called on a token + that appears at the left hand side of a ## operator. */ + if (macro_of_context (pfile->context) == NULL + || (!(lhs->flags & PASTE_LEFT))) + abort (); + + if (context->tokens_kind == TOKENS_KIND_EXTENDED) + /* The caller must have called consume_next_token_from_context + right before calling us. That has incremented the pointer to + the current virtual location. So it now points to the location + of the token that comes right after *LHS. We want the + resulting pasted token to have the location of the current + *LHS, though. */ + virt_loc = context->c.mc->cur_virt_loc[-1]; + else + /* We are not tracking macro expansion. So the best virtual + location we can get here is the expansion point of the macro we + are currently expanding. */ + virt_loc = pfile->invocation_location; + + do + { + /* Take the token directly from the current context. We can do + this, because we are in the replacement list of either an + object-like macro, or a function-like macro with arguments + inserted. In either case, the constraints to #define + guarantee we have at least one more token. */ + if (context->tokens_kind == TOKENS_KIND_DIRECT) + rhs = FIRST (context).token++; + else if (context->tokens_kind == TOKENS_KIND_INDIRECT) + rhs = *FIRST (context).ptoken++; + else if (context->tokens_kind == TOKENS_KIND_EXTENDED) + { + /* So we are in presence of an extended token context, which + means that each token in this context has a virtual + location attached to it. So let's not forget to update + the pointer to the current virtual location of the + current token when we update the pointer to the current + token */ + + rhs = *FIRST (context).ptoken++; + /* context->c.mc must be non-null, as if we were not in a + macro context, context->tokens_kind could not be equal to + TOKENS_KIND_EXTENDED. */ + context->c.mc->cur_virt_loc++; + } + + if (rhs->type == CPP_PADDING) + { + if (rhs->flags & PASTE_LEFT) + abort (); + } + if (!paste_tokens (pfile, virt_loc, &lhs, rhs)) + { + _cpp_backup_tokens (pfile, 1); + break; + } + } + while (rhs->flags & PASTE_LEFT); + + /* Put the resulting token in its own context. */ + if (context->tokens_kind == TOKENS_KIND_EXTENDED) + { + location_t *virt_locs = NULL; + _cpp_buff *token_buf = tokens_buff_new (pfile, 1, &virt_locs); + tokens_buff_add_token (token_buf, virt_locs, lhs, + virt_loc, 0, NULL, 0); + push_extended_tokens_context (pfile, context->c.mc->macro_node, + token_buf, virt_locs, + (const cpp_token **)token_buf->base, 1); + } + else + _cpp_push_token_context (pfile, NULL, lhs, 1); +} + +/* Returns TRUE if the number of arguments ARGC supplied in an + invocation of the MACRO referenced by NODE is valid. An empty + invocation to a macro with no parameters should pass ARGC as zero. + + Note that MACRO cannot necessarily be deduced from NODE, in case + NODE was redefined whilst collecting arguments. */ +bool +_cpp_arguments_ok (cpp_reader *pfile, cpp_macro *macro, const cpp_hashnode *node, unsigned int argc) +{ + if (argc == macro->paramc) + return true; + + if (argc < macro->paramc) + { + /* In C++20 (here the va_opt flag is used), and also as a GNU + extension, variadic arguments are allowed to not appear in + the invocation at all. + e.g. #define debug(format, args...) something + debug("string"); + + This is exactly the same as if an empty variadic list had been + supplied - debug("string", ). */ + + if (argc + 1 == macro->paramc && macro->variadic) + { + if (CPP_PEDANTIC (pfile) && ! macro->syshdr + && ! CPP_OPTION (pfile, va_opt)) + { + if (CPP_OPTION (pfile, cplusplus)) + cpp_error (pfile, CPP_DL_PEDWARN, + "ISO C++11 requires at least one argument " + "for the \"...\" in a variadic macro"); + else + cpp_error (pfile, CPP_DL_PEDWARN, + "ISO C99 requires at least one argument " + "for the \"...\" in a variadic macro"); + } + return true; + } + + cpp_error (pfile, CPP_DL_ERROR, + "macro \"%s\" requires %u arguments, but only %u given", + NODE_NAME (node), macro->paramc, argc); + } + else + cpp_error (pfile, CPP_DL_ERROR, + "macro \"%s\" passed %u arguments, but takes just %u", + NODE_NAME (node), argc, macro->paramc); + + if (macro->line > RESERVED_LOCATION_COUNT) + cpp_error_at (pfile, CPP_DL_NOTE, macro->line, "macro \"%s\" defined here", + NODE_NAME (node)); + + return false; +} + +/* Reads and returns the arguments to a function-like macro + invocation. Assumes the opening parenthesis has been processed. + If there is an error, emits an appropriate diagnostic and returns + NULL. Each argument is terminated by a CPP_EOF token, for the + future benefit of expand_arg(). If there are any deferred + #pragma directives among macro arguments, store pointers to the + CPP_PRAGMA ... CPP_PRAGMA_EOL tokens into *PRAGMA_BUFF buffer. + + What is returned is the buffer that contains the memory allocated + to hold the macro arguments. NODE is the name of the macro this + function is dealing with. If NUM_ARGS is non-NULL, *NUM_ARGS is + set to the actual number of macro arguments allocated in the + returned buffer. */ +static _cpp_buff * +collect_args (cpp_reader *pfile, const cpp_hashnode *node, + _cpp_buff **pragma_buff, unsigned *num_args) +{ + _cpp_buff *buff, *base_buff; + cpp_macro *macro; + macro_arg *args, *arg; + const cpp_token *token; + unsigned int argc; + location_t virt_loc; + bool track_macro_expansion_p = CPP_OPTION (pfile, track_macro_expansion); + unsigned num_args_alloced = 0; + + macro = node->value.macro; + if (macro->paramc) + argc = macro->paramc; + else + argc = 1; + +#define DEFAULT_NUM_TOKENS_PER_MACRO_ARG 50 +#define ARG_TOKENS_EXTENT 1000 + + buff = _cpp_get_buff (pfile, argc * (DEFAULT_NUM_TOKENS_PER_MACRO_ARG + * sizeof (cpp_token *) + + sizeof (macro_arg))); + base_buff = buff; + args = (macro_arg *) buff->base; + memset (args, 0, argc * sizeof (macro_arg)); + buff->cur = (unsigned char *) &args[argc]; + arg = args, argc = 0; + + /* Collect the tokens making up each argument. We don't yet know + how many arguments have been supplied, whether too many or too + few. Hence the slightly bizarre usage of "argc" and "arg". */ + do + { + unsigned int paren_depth = 0; + unsigned int ntokens = 0; + unsigned virt_locs_capacity = DEFAULT_NUM_TOKENS_PER_MACRO_ARG; + num_args_alloced++; + + argc++; + arg->first = (const cpp_token **) buff->cur; + if (track_macro_expansion_p) + { + virt_locs_capacity = DEFAULT_NUM_TOKENS_PER_MACRO_ARG; + arg->virt_locs = XNEWVEC (location_t, + virt_locs_capacity); + } + + for (;;) + { + /* Require space for 2 new tokens (including a CPP_EOF). */ + if ((unsigned char *) &arg->first[ntokens + 2] > buff->limit) + { + buff = _cpp_append_extend_buff (pfile, buff, + ARG_TOKENS_EXTENT + * sizeof (cpp_token *)); + arg->first = (const cpp_token **) buff->cur; + } + if (track_macro_expansion_p + && (ntokens + 2 > virt_locs_capacity)) + { + virt_locs_capacity += ARG_TOKENS_EXTENT; + arg->virt_locs = XRESIZEVEC (location_t, + arg->virt_locs, + virt_locs_capacity); + } + + token = cpp_get_token_1 (pfile, &virt_loc); + + if (token->type == CPP_PADDING) + { + /* Drop leading padding. */ + if (ntokens == 0) + continue; + } + else if (token->type == CPP_OPEN_PAREN) + paren_depth++; + else if (token->type == CPP_CLOSE_PAREN) + { + if (paren_depth-- == 0) + break; + } + else if (token->type == CPP_COMMA) + { + /* A comma does not terminate an argument within + parentheses or as part of a variable argument. */ + if (paren_depth == 0 + && ! (macro->variadic && argc == macro->paramc)) + break; + } + else if (token->type == CPP_EOF + || (token->type == CPP_HASH && token->flags & BOL)) + break; + else if (token->type == CPP_PRAGMA && !(token->flags & PRAGMA_OP)) + { + cpp_token *newtok = _cpp_temp_token (pfile); + + /* CPP_PRAGMA token lives in directive_result, which will + be overwritten on the next directive. */ + *newtok = *token; + token = newtok; + do + { + if (*pragma_buff == NULL + || BUFF_ROOM (*pragma_buff) < sizeof (cpp_token *)) + { + _cpp_buff *next; + if (*pragma_buff == NULL) + *pragma_buff + = _cpp_get_buff (pfile, 32 * sizeof (cpp_token *)); + else + { + next = *pragma_buff; + *pragma_buff + = _cpp_get_buff (pfile, + (BUFF_FRONT (*pragma_buff) + - (*pragma_buff)->base) * 2); + (*pragma_buff)->next = next; + } + } + *(const cpp_token **) BUFF_FRONT (*pragma_buff) = token; + BUFF_FRONT (*pragma_buff) += sizeof (cpp_token *); + if (token->type == CPP_PRAGMA_EOL) + break; + token = cpp_get_token_1 (pfile, &virt_loc); + } + while (token->type != CPP_EOF); + + /* In deferred pragmas parsing_args and prevent_expansion + had been changed, reset it. */ + pfile->state.parsing_args = 2; + pfile->state.prevent_expansion = 1; + + if (token->type == CPP_EOF) + break; + else + continue; + } + set_arg_token (arg, token, virt_loc, + ntokens, MACRO_ARG_TOKEN_NORMAL, + CPP_OPTION (pfile, track_macro_expansion)); + ntokens++; + } + + /* Drop trailing padding. */ + while (ntokens > 0 && arg->first[ntokens - 1]->type == CPP_PADDING) + ntokens--; + + arg->count = ntokens; + /* Append an EOF to mark end-of-argument. */ + set_arg_token (arg, &pfile->endarg, token->src_loc, + ntokens, MACRO_ARG_TOKEN_NORMAL, + CPP_OPTION (pfile, track_macro_expansion)); + + /* Terminate the argument. Excess arguments loop back and + overwrite the final legitimate argument, before failing. */ + if (argc <= macro->paramc) + { + buff->cur = (unsigned char *) &arg->first[ntokens + 1]; + if (argc != macro->paramc) + arg++; + } + } + while (token->type != CPP_CLOSE_PAREN && token->type != CPP_EOF); + + if (token->type == CPP_EOF) + { + /* Unless the EOF is marking the end of an argument, it's a fake + one from the end of a file that _cpp_clean_line will not have + advanced past. */ + if (token == &pfile->endarg) + _cpp_backup_tokens (pfile, 1); + cpp_error (pfile, CPP_DL_ERROR, + "unterminated argument list invoking macro \"%s\"", + NODE_NAME (node)); + } + else + { + /* A single empty argument is counted as no argument. */ + if (argc == 1 && macro->paramc == 0 && args[0].count == 0) + argc = 0; + if (_cpp_arguments_ok (pfile, macro, node, argc)) + { + /* GCC has special semantics for , ## b where b is a varargs + parameter: we remove the comma if b was omitted entirely. + If b was merely an empty argument, the comma is retained. + If the macro takes just one (varargs) parameter, then we + retain the comma only if we are standards conforming. + + If FIRST is NULL replace_args () swallows the comma. */ + if (macro->variadic && (argc < macro->paramc + || (argc == 1 && args[0].count == 0 + && !CPP_OPTION (pfile, std)))) + args[macro->paramc - 1].first = NULL; + if (num_args) + *num_args = num_args_alloced; + return base_buff; + } + } + + /* An error occurred. */ + _cpp_release_buff (pfile, base_buff); + return NULL; +} + +/* Search for an opening parenthesis to the macro of NODE, in such a + way that, if none is found, we don't lose the information in any + intervening padding tokens. If we find the parenthesis, collect + the arguments and return the buffer containing them. PRAGMA_BUFF + argument is the same as in collect_args. If NUM_ARGS is non-NULL, + *NUM_ARGS is set to the number of arguments contained in the + returned buffer. */ +static _cpp_buff * +funlike_invocation_p (cpp_reader *pfile, cpp_hashnode *node, + _cpp_buff **pragma_buff, unsigned *num_args) +{ + const cpp_token *token, *padding = NULL; + + for (;;) + { + token = cpp_get_token (pfile); + if (token->type != CPP_PADDING) + break; + gcc_assert ((token->flags & PREV_WHITE) == 0); + if (padding == NULL + || padding->val.source == NULL + || (!(padding->val.source->flags & PREV_WHITE) + && token->val.source == NULL)) + padding = token; + } + + if (token->type == CPP_OPEN_PAREN) + { + pfile->state.parsing_args = 2; + return collect_args (pfile, node, pragma_buff, num_args); + } + + /* Back up. A CPP_EOF is either an EOF from an argument we're + expanding, or a fake one from lex_direct. We want to backup the + former, but not the latter. We may have skipped padding, in + which case backing up more than one token when expanding macros + is in general too difficult. We re-insert it in its own + context. */ + if (token->type != CPP_EOF || token == &pfile->endarg) + { + _cpp_backup_tokens (pfile, 1); + if (padding) + _cpp_push_token_context (pfile, NULL, padding, 1); + } + + return NULL; +} + +/* Return the real number of tokens in the expansion of MACRO. */ +static inline unsigned int +macro_real_token_count (const cpp_macro *macro) +{ + if (__builtin_expect (!macro->extra_tokens, true)) + return macro->count; + + for (unsigned i = macro->count; i--;) + if (macro->exp.tokens[i].type != CPP_PASTE) + return i + 1; + + return 0; +} + +/* Push the context of a macro with hash entry NODE onto the context + stack. If we can successfully expand the macro, we push a context + containing its yet-to-be-rescanned replacement list and return one. + If there were additionally any unexpanded deferred #pragma + directives among macro arguments, push another context containing + the pragma tokens before the yet-to-be-rescanned replacement list + and return two. Otherwise, we don't push a context and return + zero. LOCATION is the location of the expansion point of the + macro. */ +static int +enter_macro_context (cpp_reader *pfile, cpp_hashnode *node, + const cpp_token *result, location_t location) +{ + /* The presence of a macro invalidates a file's controlling macro. */ + pfile->mi_valid = false; + + pfile->state.angled_headers = false; + + /* From here to when we push the context for the macro later down + this function, we need to flag the fact that we are about to + expand a macro. This is useful when -ftrack-macro-expansion is + turned off. In that case, we need to record the location of the + expansion point of the top-most macro we are about to to expand, + into pfile->invocation_location. But we must not record any such + location once the process of expanding the macro starts; that is, + we must not do that recording between now and later down this + function where set this flag to FALSE. */ + pfile->about_to_expand_macro_p = true; + + if (cpp_user_macro_p (node)) + { + cpp_macro *macro = node->value.macro; + _cpp_buff *pragma_buff = NULL; + + if (macro->fun_like) + { + _cpp_buff *buff; + unsigned num_args = 0; + + pfile->state.prevent_expansion++; + pfile->keep_tokens++; + pfile->state.parsing_args = 1; + buff = funlike_invocation_p (pfile, node, &pragma_buff, + &num_args); + pfile->state.parsing_args = 0; + pfile->keep_tokens--; + pfile->state.prevent_expansion--; + + if (buff == NULL) + { + if (CPP_WTRADITIONAL (pfile) && ! node->value.macro->syshdr) + cpp_warning (pfile, CPP_W_TRADITIONAL, + "function-like macro \"%s\" must be used with arguments in traditional C", + NODE_NAME (node)); + + if (pragma_buff) + _cpp_release_buff (pfile, pragma_buff); + + pfile->about_to_expand_macro_p = false; + return 0; + } + + if (macro->paramc > 0) + replace_args (pfile, node, macro, + (macro_arg *) buff->base, + location); + /* Free the memory used by the arguments of this + function-like macro. This memory has been allocated by + funlike_invocation_p and by replace_args. */ + delete_macro_args (buff, num_args); + } + + /* Disable the macro within its expansion. */ + node->flags |= NODE_DISABLED; + + /* Laziness can only affect the expansion tokens of the macro, + not its fun-likeness or parameters. */ + _cpp_maybe_notify_macro_use (pfile, node, location); + if (pfile->cb.used) + pfile->cb.used (pfile, location, node); + + macro->used = 1; + + if (macro->paramc == 0) + { + unsigned tokens_count = macro_real_token_count (macro); + if (CPP_OPTION (pfile, track_macro_expansion)) + { + unsigned int i; + const cpp_token *src = macro->exp.tokens; + const line_map_macro *map; + location_t *virt_locs = NULL; + _cpp_buff *macro_tokens + = tokens_buff_new (pfile, tokens_count, &virt_locs); + + /* Create a macro map to record the locations of the + tokens that are involved in the expansion. LOCATION + is the location of the macro expansion point. */ + map = linemap_enter_macro (pfile->line_table, + node, location, tokens_count); + for (i = 0; i < tokens_count; ++i) + { + tokens_buff_add_token (macro_tokens, virt_locs, + src, src->src_loc, + src->src_loc, map, i); + ++src; + } + push_extended_tokens_context (pfile, node, + macro_tokens, + virt_locs, + (const cpp_token **) + macro_tokens->base, + tokens_count); + } + else + _cpp_push_token_context (pfile, node, macro->exp.tokens, + tokens_count); + num_macro_tokens_counter += tokens_count; + } + + if (pragma_buff) + { + if (!pfile->state.in_directive) + _cpp_push_token_context (pfile, NULL, + padding_token (pfile, result), 1); + do + { + unsigned tokens_count; + _cpp_buff *tail = pragma_buff->next; + pragma_buff->next = NULL; + tokens_count = ((const cpp_token **) BUFF_FRONT (pragma_buff) + - (const cpp_token **) pragma_buff->base); + push_ptoken_context (pfile, NULL, pragma_buff, + (const cpp_token **) pragma_buff->base, + tokens_count); + pragma_buff = tail; + if (!CPP_OPTION (pfile, track_macro_expansion)) + num_macro_tokens_counter += tokens_count; + + } + while (pragma_buff != NULL); + pfile->about_to_expand_macro_p = false; + return 2; + } + + pfile->about_to_expand_macro_p = false; + return 1; + } + + pfile->about_to_expand_macro_p = false; + /* Handle built-in macros and the _Pragma operator. */ + { + location_t expand_loc; + + if (/* The top-level macro invocation that triggered the expansion + we are looking at is with a function-like user macro ... */ + cpp_fun_like_macro_p (pfile->top_most_macro_node) + /* ... and we are tracking the macro expansion. */ + && CPP_OPTION (pfile, track_macro_expansion)) + /* Then the location of the end of the macro invocation is the + location of the expansion point of this macro. */ + expand_loc = location; + else + /* Otherwise, the location of the end of the macro invocation is + the location of the expansion point of that top-level macro + invocation. */ + expand_loc = pfile->invocation_location; + + return builtin_macro (pfile, node, location, expand_loc); + } +} + +/* De-allocate the memory used by BUFF which is an array of instances + of macro_arg. NUM_ARGS is the number of instances of macro_arg + present in BUFF. */ +static void +delete_macro_args (_cpp_buff *buff, unsigned num_args) +{ + macro_arg *macro_args; + unsigned i; + + if (buff == NULL) + return; + + macro_args = (macro_arg *) buff->base; + + /* Walk instances of macro_arg to free their expanded tokens as well + as their macro_arg::virt_locs members. */ + for (i = 0; i < num_args; ++i) + { + if (macro_args[i].expanded) + { + free (macro_args[i].expanded); + macro_args[i].expanded = NULL; + } + if (macro_args[i].virt_locs) + { + free (macro_args[i].virt_locs); + macro_args[i].virt_locs = NULL; + } + if (macro_args[i].expanded_virt_locs) + { + free (macro_args[i].expanded_virt_locs); + macro_args[i].expanded_virt_locs = NULL; + } + } + _cpp_free_buff (buff); +} + +/* Set the INDEXth token of the macro argument ARG. TOKEN is the token + to set, LOCATION is its virtual location. "Virtual" location means + the location that encodes loci across macro expansion. Otherwise + it has to be TOKEN->SRC_LOC. KIND is the kind of tokens the + argument ARG is supposed to contain. Note that ARG must be + tailored so that it has enough room to contain INDEX + 1 numbers of + tokens, at least. */ +static void +set_arg_token (macro_arg *arg, const cpp_token *token, + location_t location, size_t index, + enum macro_arg_token_kind kind, + bool track_macro_exp_p) +{ + const cpp_token **token_ptr; + location_t *loc = NULL; + + token_ptr = + arg_token_ptr_at (arg, index, kind, + track_macro_exp_p ? &loc : NULL); + *token_ptr = token; + + if (loc != NULL) + { + /* We can't set the location of a stringified argument + token and we can't set any location if we aren't tracking + macro expansion locations. */ + gcc_checking_assert (kind != MACRO_ARG_TOKEN_STRINGIFIED + && track_macro_exp_p); + *loc = location; + } +} + +/* Get the pointer to the location of the argument token of the + function-like macro argument ARG. This function must be called + only when we -ftrack-macro-expansion is on. */ +static const location_t * +get_arg_token_location (const macro_arg *arg, + enum macro_arg_token_kind kind) +{ + const location_t *loc = NULL; + const cpp_token **token_ptr = + arg_token_ptr_at (arg, 0, kind, (location_t **) &loc); + + if (token_ptr == NULL) + return NULL; + + return loc; +} + +/* Return the pointer to the INDEXth token of the macro argument ARG. + KIND specifies the kind of token the macro argument ARG contains. + If VIRT_LOCATION is non NULL, *VIRT_LOCATION is set to the address + of the virtual location of the returned token if the + -ftrack-macro-expansion flag is on; otherwise, it's set to the + spelling location of the returned token. */ +static const cpp_token ** +arg_token_ptr_at (const macro_arg *arg, size_t index, + enum macro_arg_token_kind kind, + location_t **virt_location) +{ + const cpp_token **tokens_ptr = NULL; + + switch (kind) + { + case MACRO_ARG_TOKEN_NORMAL: + tokens_ptr = arg->first; + break; + case MACRO_ARG_TOKEN_STRINGIFIED: + tokens_ptr = (const cpp_token **) &arg->stringified; + break; + case MACRO_ARG_TOKEN_EXPANDED: + tokens_ptr = arg->expanded; + break; + } + + if (tokens_ptr == NULL) + /* This can happen for e.g, an empty token argument to a + funtion-like macro. */ + return tokens_ptr; + + if (virt_location) + { + if (kind == MACRO_ARG_TOKEN_NORMAL) + *virt_location = &arg->virt_locs[index]; + else if (kind == MACRO_ARG_TOKEN_EXPANDED) + *virt_location = &arg->expanded_virt_locs[index]; + else if (kind == MACRO_ARG_TOKEN_STRINGIFIED) + *virt_location = + (location_t *) &tokens_ptr[index]->src_loc; + } + return &tokens_ptr[index]; +} + +/* Initialize an iterator so that it iterates over the tokens of a + function-like macro argument. KIND is the kind of tokens we want + ITER to iterate over. TOKEN_PTR points the first token ITER will + iterate over. */ +static void +macro_arg_token_iter_init (macro_arg_token_iter *iter, + bool track_macro_exp_p, + enum macro_arg_token_kind kind, + const macro_arg *arg, + const cpp_token **token_ptr) +{ + iter->track_macro_exp_p = track_macro_exp_p; + iter->kind = kind; + iter->token_ptr = token_ptr; + /* Unconditionally initialize this so that the compiler doesn't warn + about iter->location_ptr being possibly uninitialized later after + this code has been inlined somewhere. */ + iter->location_ptr = NULL; + if (track_macro_exp_p) + iter->location_ptr = get_arg_token_location (arg, kind); +#if CHECKING_P + iter->num_forwards = 0; + if (track_macro_exp_p + && token_ptr != NULL + && iter->location_ptr == NULL) + abort (); +#endif +} + +/* Move the iterator one token forward. Note that if IT was + initialized on an argument that has a stringified token, moving it + forward doesn't make sense as a stringified token is essentially one + string. */ +static void +macro_arg_token_iter_forward (macro_arg_token_iter *it) +{ + switch (it->kind) + { + case MACRO_ARG_TOKEN_NORMAL: + case MACRO_ARG_TOKEN_EXPANDED: + it->token_ptr++; + if (it->track_macro_exp_p) + it->location_ptr++; + break; + case MACRO_ARG_TOKEN_STRINGIFIED: +#if CHECKING_P + if (it->num_forwards > 0) + abort (); +#endif + break; + } + +#if CHECKING_P + it->num_forwards++; +#endif +} + +/* Return the token pointed to by the iterator. */ +static const cpp_token * +macro_arg_token_iter_get_token (const macro_arg_token_iter *it) +{ +#if CHECKING_P + if (it->kind == MACRO_ARG_TOKEN_STRINGIFIED + && it->num_forwards > 0) + abort (); +#endif + if (it->token_ptr == NULL) + return NULL; + return *it->token_ptr; +} + +/* Return the location of the token pointed to by the iterator.*/ +static location_t +macro_arg_token_iter_get_location (const macro_arg_token_iter *it) +{ +#if CHECKING_P + if (it->kind == MACRO_ARG_TOKEN_STRINGIFIED + && it->num_forwards > 0) + abort (); +#endif + if (it->track_macro_exp_p) + return *it->location_ptr; + else + return (*it->token_ptr)->src_loc; +} + +/* Return the index of a token [resulting from macro expansion] inside + the total list of tokens resulting from a given macro + expansion. The index can be different depending on whether if we + want each tokens resulting from function-like macro arguments + expansion to have a different location or not. + + E.g, consider this function-like macro: + + #define M(x) x - 3 + + Then consider us "calling" it (and thus expanding it) like: + + M(1+4) + + It will be expanded into: + + 1+4-3 + + Let's consider the case of the token '4'. + + Its index can be 2 (it's the third token of the set of tokens + resulting from the expansion) or it can be 0 if we consider that + all tokens resulting from the expansion of the argument "1+2" have + the same index, which is 0. In this later case, the index of token + '-' would then be 1 and the index of token '3' would be 2. + + The later case is useful to use less memory e.g, for the case of + the user using the option -ftrack-macro-expansion=1. + + ABSOLUTE_TOKEN_INDEX is the index of the macro argument token we + are interested in. CUR_REPLACEMENT_TOKEN is the token of the macro + parameter (inside the macro replacement list) that corresponds to + the macro argument for which ABSOLUTE_TOKEN_INDEX is a token index + of. + + If we refer to the example above, for the '4' argument token, + ABSOLUTE_TOKEN_INDEX would be set to 2, and CUR_REPLACEMENT_TOKEN + would be set to the token 'x', in the replacement list "x - 3" of + macro M. + + This is a subroutine of replace_args. */ +inline static unsigned +expanded_token_index (cpp_reader *pfile, cpp_macro *macro, + const cpp_token *cur_replacement_token, + unsigned absolute_token_index) +{ + if (CPP_OPTION (pfile, track_macro_expansion) > 1) + return absolute_token_index; + return cur_replacement_token - macro->exp.tokens; +} + +/* Copy whether PASTE_LEFT is set from SRC to *PASTE_FLAG. */ + +static void +copy_paste_flag (cpp_reader *pfile, const cpp_token **paste_flag, + const cpp_token *src) +{ + cpp_token *token = _cpp_temp_token (pfile); + token->type = (*paste_flag)->type; + token->val = (*paste_flag)->val; + if (src->flags & PASTE_LEFT) + token->flags = (*paste_flag)->flags | PASTE_LEFT; + else + token->flags = (*paste_flag)->flags & ~PASTE_LEFT; + *paste_flag = token; +} + +/* True IFF the last token emitted into BUFF (if any) is PTR. */ + +static bool +last_token_is (_cpp_buff *buff, const cpp_token **ptr) +{ + return (ptr && tokens_buff_last_token_ptr (buff) == ptr); +} + +/* Replace the parameters in a function-like macro of NODE with the + actual ARGS, and place the result in a newly pushed token context. + Expand each argument before replacing, unless it is operated upon + by the # or ## operators. EXPANSION_POINT_LOC is the location of + the expansion point of the macro. E.g, the location of the + function-like macro invocation. */ +static void +replace_args (cpp_reader *pfile, cpp_hashnode *node, cpp_macro *macro, + macro_arg *args, location_t expansion_point_loc) +{ + unsigned int i, total; + const cpp_token *src, *limit; + const cpp_token **first = NULL; + macro_arg *arg; + _cpp_buff *buff = NULL; + location_t *virt_locs = NULL; + unsigned int exp_count; + const line_map_macro *map = NULL; + int track_macro_exp; + + /* First, fully macro-expand arguments, calculating the number of + tokens in the final expansion as we go. The ordering of the if + statements below is subtle; we must handle stringification before + pasting. */ + + /* EXP_COUNT is the number of tokens in the macro replacement + list. TOTAL is the number of tokens /after/ macro parameters + have been replaced by their arguments. */ + exp_count = macro_real_token_count (macro); + total = exp_count; + limit = macro->exp.tokens + exp_count; + + for (src = macro->exp.tokens; src < limit; src++) + if (src->type == CPP_MACRO_ARG) + { + /* Leading and trailing padding tokens. */ + total += 2; + /* Account for leading and padding tokens in exp_count too. + This is going to be important later down this function, + when we want to handle the case of (track_macro_exp < + 2). */ + exp_count += 2; + + /* We have an argument. If it is not being stringified or + pasted it is macro-replaced before insertion. */ + arg = &args[src->val.macro_arg.arg_no - 1]; + + if (src->flags & STRINGIFY_ARG) + { + if (!arg->stringified) + arg->stringified = stringify_arg (pfile, arg->first, arg->count); + } + else if ((src->flags & PASTE_LEFT) + || (src != macro->exp.tokens && (src[-1].flags & PASTE_LEFT))) + total += arg->count - 1; + else + { + if (!arg->expanded) + expand_arg (pfile, arg); + total += arg->expanded_count - 1; + } + } + + /* When the compiler is called with the -ftrack-macro-expansion + flag, we need to keep track of the location of each token that + results from macro expansion. + + A token resulting from macro expansion is not a new token. It is + simply the same token as the token coming from the macro + definition. The new things that are allocated are the buffer + that holds the tokens resulting from macro expansion and a new + location that records many things like the locus of the expansion + point as well as the original locus inside the definition of the + macro. This location is called a virtual location. + + So the buffer BUFF holds a set of cpp_token*, and the buffer + VIRT_LOCS holds the virtual locations of the tokens held by BUFF. + + Both of these two buffers are going to be hung off of the macro + context, when the latter is pushed. The memory allocated to + store the tokens and their locations is going to be freed once + the context of macro expansion is popped. + + As far as tokens are concerned, the memory overhead of + -ftrack-macro-expansion is proportional to the number of + macros that get expanded multiplied by sizeof (location_t). + The good news is that extra memory gets freed when the macro + context is freed, i.e shortly after the macro got expanded. */ + + /* Is the -ftrack-macro-expansion flag in effect? */ + track_macro_exp = CPP_OPTION (pfile, track_macro_expansion); + + /* Now allocate memory space for tokens and locations resulting from + the macro expansion, copy the tokens and replace the arguments. + This memory must be freed when the context of the macro MACRO is + popped. */ + buff = tokens_buff_new (pfile, total, track_macro_exp ? &virt_locs : NULL); + + first = (const cpp_token **) buff->base; + + /* Create a macro map to record the locations of the tokens that are + involved in the expansion. Note that the expansion point is set + to the location of the closing parenthesis. Otherwise, the + subsequent map created for the first token that comes after the + macro map might have a wrong line number. That would lead to + tokens with wrong line numbers after the macro expansion. This + adds up to the memory overhead of the -ftrack-macro-expansion + flag; for every macro that is expanded, a "macro map" is + created. */ + if (track_macro_exp) + { + int num_macro_tokens = total; + if (track_macro_exp < 2) + /* Then the number of macro tokens won't take in account the + fact that function-like macro arguments can expand to + multiple tokens. This is to save memory at the expense of + accuracy. + + Suppose we have #define SQUARE(A) A * A + + And then we do SQUARE(2+3) + + Then the tokens 2, +, 3, will have the same location, + saying they come from the expansion of the argument A. */ + num_macro_tokens = exp_count; + map = linemap_enter_macro (pfile->line_table, node, + expansion_point_loc, + num_macro_tokens); + } + i = 0; + vaopt_state vaopt_tracker (pfile, macro->variadic, &args[macro->paramc - 1]); + const cpp_token **vaopt_start = NULL; + for (src = macro->exp.tokens; src < limit; src++) + { + unsigned int arg_tokens_count; + macro_arg_token_iter from; + const cpp_token **paste_flag = NULL; + const cpp_token **tmp_token_ptr; + + /* __VA_OPT__ handling. */ + vaopt_state::update_type vostate = vaopt_tracker.update (src); + if (__builtin_expect (vostate != vaopt_state::INCLUDE, false)) + { + if (vostate == vaopt_state::BEGIN) + { + /* Padding on the left of __VA_OPT__ (unless RHS of ##). */ + if (src != macro->exp.tokens && !(src[-1].flags & PASTE_LEFT)) + { + const cpp_token *t = padding_token (pfile, src); + unsigned index = expanded_token_index (pfile, macro, src, i); + /* Allocate a virtual location for the padding token and + append the token and its location to BUFF and + VIRT_LOCS. */ + tokens_buff_add_token (buff, virt_locs, t, + t->src_loc, t->src_loc, + map, index); + } + vaopt_start = tokens_buff_last_token_ptr (buff); + } + else if (vostate == vaopt_state::END) + { + const cpp_token **start = vaopt_start; + vaopt_start = NULL; + + paste_flag = tokens_buff_last_token_ptr (buff); + + if (vaopt_tracker.stringify ()) + { + unsigned int count + = start ? paste_flag - start : tokens_buff_count (buff); + const cpp_token **first + = start ? start + 1 + : (const cpp_token **) (buff->base); + unsigned int i, j; + + /* Paste any tokens that need to be pasted before calling + stringify_arg, because stringify_arg uses pfile->u_buff + which paste_tokens can use as well. */ + for (i = 0, j = 0; i < count; i++, j++) + { + const cpp_token *token = first[i]; + + if (token->flags & PASTE_LEFT) + { + location_t virt_loc = pfile->invocation_location; + const cpp_token *rhs; + do + { + if (i == count) + abort (); + rhs = first[++i]; + if (!paste_tokens (pfile, virt_loc, &token, rhs)) + { + --i; + break; + } + } + while (rhs->flags & PASTE_LEFT); + } + + first[j] = token; + } + if (j != i) + { + while (i-- != j) + tokens_buff_remove_last_token (buff); + count = j; + } + + const cpp_token *t = stringify_arg (pfile, first, count); + while (count--) + tokens_buff_remove_last_token (buff); + if (src->flags & PASTE_LEFT) + copy_paste_flag (pfile, &t, src); + tokens_buff_add_token (buff, virt_locs, + t, t->src_loc, t->src_loc, + NULL, 0); + continue; + } + if (start && paste_flag == start && (*start)->flags & PASTE_LEFT) + /* If __VA_OPT__ expands to nothing (either because __VA_ARGS__ + is empty or because it is __VA_OPT__() ), drop PASTE_LEFT + flag from previous token. */ + copy_paste_flag (pfile, start, &pfile->avoid_paste); + if (src->flags & PASTE_LEFT) + { + /* Don't avoid paste after all. */ + while (paste_flag && paste_flag != start + && *paste_flag == &pfile->avoid_paste) + { + tokens_buff_remove_last_token (buff); + paste_flag = tokens_buff_last_token_ptr (buff); + } + + /* With a non-empty __VA_OPT__ on the LHS of ##, the last + token should be flagged PASTE_LEFT. */ + if (paste_flag && (*paste_flag)->type != CPP_PADDING) + copy_paste_flag (pfile, paste_flag, src); + } + else + { + /* Otherwise, avoid paste on RHS, __VA_OPT__(c)d or + __VA_OPT__(c)__VA_OPT__(d). */ + const cpp_token *t = &pfile->avoid_paste; + tokens_buff_add_token (buff, virt_locs, + t, t->src_loc, t->src_loc, + NULL, 0); + } + } + continue; + } + + if (src->type != CPP_MACRO_ARG) + { + /* Allocate a virtual location for token SRC, and add that + token and its virtual location into the buffers BUFF and + VIRT_LOCS. */ + unsigned index = expanded_token_index (pfile, macro, src, i); + tokens_buff_add_token (buff, virt_locs, src, + src->src_loc, src->src_loc, + map, index); + i += 1; + continue; + } + + paste_flag = 0; + arg = &args[src->val.macro_arg.arg_no - 1]; + /* SRC is a macro parameter that we need to replace with its + corresponding argument. So at some point we'll need to + iterate over the tokens of the macro argument and copy them + into the "place" now holding the correspondig macro + parameter. We are going to use the iterator type + macro_argo_token_iter to handle that iterating. The 'if' + below is to initialize the iterator depending on the type of + tokens the macro argument has. It also does some adjustment + related to padding tokens and some pasting corner cases. */ + if (src->flags & STRINGIFY_ARG) + { + arg_tokens_count = 1; + macro_arg_token_iter_init (&from, + CPP_OPTION (pfile, + track_macro_expansion), + MACRO_ARG_TOKEN_STRINGIFIED, + arg, &arg->stringified); + } + else if (src->flags & PASTE_LEFT) + { + arg_tokens_count = arg->count; + macro_arg_token_iter_init (&from, + CPP_OPTION (pfile, + track_macro_expansion), + MACRO_ARG_TOKEN_NORMAL, + arg, arg->first); + } + else if (src != macro->exp.tokens && (src[-1].flags & PASTE_LEFT)) + { + int num_toks; + arg_tokens_count = arg->count; + macro_arg_token_iter_init (&from, + CPP_OPTION (pfile, + track_macro_expansion), + MACRO_ARG_TOKEN_NORMAL, + arg, arg->first); + + num_toks = tokens_buff_count (buff); + + if (num_toks != 0) + { + /* So the current parameter token is pasted to the previous + token in the replacement list. Let's look at what + we have as previous and current arguments. */ + + /* This is the previous argument's token ... */ + tmp_token_ptr = tokens_buff_last_token_ptr (buff); + + if ((*tmp_token_ptr)->type == CPP_COMMA + && macro->variadic + && src->val.macro_arg.arg_no == macro->paramc) + { + /* ... which is a comma; and the current parameter + is the last parameter of a variadic function-like + macro. If the argument to the current last + parameter is NULL, then swallow the comma, + otherwise drop the paste flag. */ + if (macro_arg_token_iter_get_token (&from) == NULL) + tokens_buff_remove_last_token (buff); + else + paste_flag = tmp_token_ptr; + } + /* Remove the paste flag if the RHS is a placemarker. */ + else if (arg_tokens_count == 0) + paste_flag = tmp_token_ptr; + } + } + else + { + arg_tokens_count = arg->expanded_count; + macro_arg_token_iter_init (&from, + CPP_OPTION (pfile, + track_macro_expansion), + MACRO_ARG_TOKEN_EXPANDED, + arg, arg->expanded); + + if (last_token_is (buff, vaopt_start)) + { + /* We're expanding an arg at the beginning of __VA_OPT__. + Skip padding. */ + while (arg_tokens_count) + { + const cpp_token *t = macro_arg_token_iter_get_token (&from); + if (t->type != CPP_PADDING) + break; + macro_arg_token_iter_forward (&from); + --arg_tokens_count; + } + } + } + + /* Padding on the left of an argument (unless RHS of ##). */ + if ((!pfile->state.in_directive || pfile->state.directive_wants_padding) + && src != macro->exp.tokens + && !(src[-1].flags & PASTE_LEFT) + && !last_token_is (buff, vaopt_start)) + { + const cpp_token *t = padding_token (pfile, src); + unsigned index = expanded_token_index (pfile, macro, src, i); + /* Allocate a virtual location for the padding token and + append the token and its location to BUFF and + VIRT_LOCS. */ + tokens_buff_add_token (buff, virt_locs, t, + t->src_loc, t->src_loc, + map, index); + } + + if (arg_tokens_count) + { + /* So now we've got the number of tokens that make up the + argument that is going to replace the current parameter + in the macro's replacement list. */ + unsigned int j; + for (j = 0; j < arg_tokens_count; ++j) + { + /* So if track_macro_exp is < 2, the user wants to + save extra memory while tracking macro expansion + locations. So in that case here is what we do: + + Suppose we have #define SQUARE(A) A * A + + And then we do SQUARE(2+3) + + Then the tokens 2, +, 3, will have the same location, + saying they come from the expansion of the argument + A. + + So that means we are going to ignore the COUNT tokens + resulting from the expansion of the current macro + argument. In other words all the ARG_TOKENS_COUNT tokens + resulting from the expansion of the macro argument will + have the index I. Normally, each of those tokens should + have index I+J. */ + unsigned token_index = i; + unsigned index; + if (track_macro_exp > 1) + token_index += j; + + index = expanded_token_index (pfile, macro, src, token_index); + const cpp_token *tok = macro_arg_token_iter_get_token (&from); + tokens_buff_add_token (buff, virt_locs, tok, + macro_arg_token_iter_get_location (&from), + src->src_loc, map, index); + macro_arg_token_iter_forward (&from); + } + + /* With a non-empty argument on the LHS of ##, the last + token should be flagged PASTE_LEFT. */ + if (src->flags & PASTE_LEFT) + paste_flag + = (const cpp_token **) tokens_buff_last_token_ptr (buff); + } + else if (CPP_PEDANTIC (pfile) && ! CPP_OPTION (pfile, c99) + && ! macro->syshdr && ! _cpp_in_system_header (pfile)) + { + if (CPP_OPTION (pfile, cplusplus)) + cpp_pedwarning (pfile, CPP_W_PEDANTIC, + "invoking macro %s argument %d: " + "empty macro arguments are undefined" + " in ISO C++98", + NODE_NAME (node), src->val.macro_arg.arg_no); + else if (CPP_OPTION (pfile, cpp_warn_c90_c99_compat)) + cpp_pedwarning (pfile, + CPP_OPTION (pfile, cpp_warn_c90_c99_compat) > 0 + ? CPP_W_C90_C99_COMPAT : CPP_W_PEDANTIC, + "invoking macro %s argument %d: " + "empty macro arguments are undefined" + " in ISO C90", + NODE_NAME (node), src->val.macro_arg.arg_no); + } + else if (CPP_OPTION (pfile, cpp_warn_c90_c99_compat) > 0 + && ! CPP_OPTION (pfile, cplusplus) + && ! macro->syshdr && ! _cpp_in_system_header (pfile)) + cpp_warning (pfile, CPP_W_C90_C99_COMPAT, + "invoking macro %s argument %d: " + "empty macro arguments are undefined" + " in ISO C90", + NODE_NAME (node), src->val.macro_arg.arg_no); + + /* Avoid paste on RHS (even case count == 0). */ + if (!pfile->state.in_directive && !(src->flags & PASTE_LEFT)) + { + const cpp_token *t = &pfile->avoid_paste; + tokens_buff_add_token (buff, virt_locs, + t, t->src_loc, t->src_loc, + NULL, 0); + } + + /* Add a new paste flag, or remove an unwanted one. */ + if (paste_flag) + copy_paste_flag (pfile, paste_flag, src); + + i += arg_tokens_count; + } + + if (track_macro_exp) + push_extended_tokens_context (pfile, node, buff, virt_locs, first, + tokens_buff_count (buff)); + else + push_ptoken_context (pfile, node, buff, first, + tokens_buff_count (buff)); + + num_macro_tokens_counter += tokens_buff_count (buff); +} + +/* Return a special padding token, with padding inherited from SOURCE. */ +static const cpp_token * +padding_token (cpp_reader *pfile, const cpp_token *source) +{ + cpp_token *result = _cpp_temp_token (pfile); + + result->type = CPP_PADDING; + + /* Data in GCed data structures cannot be made const so far, so we + need a cast here. */ + result->val.source = (cpp_token *) source; + result->flags = 0; + return result; +} + +/* Get a new uninitialized context. Create a new one if we cannot + re-use an old one. */ +static cpp_context * +next_context (cpp_reader *pfile) +{ + cpp_context *result = pfile->context->next; + + if (result == 0) + { + result = XNEW (cpp_context); + memset (result, 0, sizeof (cpp_context)); + result->prev = pfile->context; + result->next = 0; + pfile->context->next = result; + } + + pfile->context = result; + return result; +} + +/* Push a list of pointers to tokens. */ +static void +push_ptoken_context (cpp_reader *pfile, cpp_hashnode *macro, _cpp_buff *buff, + const cpp_token **first, unsigned int count) +{ + cpp_context *context = next_context (pfile); + + context->tokens_kind = TOKENS_KIND_INDIRECT; + context->c.macro = macro; + context->buff = buff; + FIRST (context).ptoken = first; + LAST (context).ptoken = first + count; +} + +/* Push a list of tokens. + + A NULL macro means that we should continue the current macro + expansion, in essence. That means that if we are currently in a + macro expansion context, we'll make the new pfile->context refer to + the current macro. */ +void +_cpp_push_token_context (cpp_reader *pfile, cpp_hashnode *macro, + const cpp_token *first, unsigned int count) +{ + cpp_context *context; + + if (macro == NULL) + macro = macro_of_context (pfile->context); + + context = next_context (pfile); + context->tokens_kind = TOKENS_KIND_DIRECT; + context->c.macro = macro; + context->buff = NULL; + FIRST (context).token = first; + LAST (context).token = first + count; +} + +/* Build a context containing a list of tokens as well as their + virtual locations and push it. TOKENS_BUFF is the buffer that + contains the tokens pointed to by FIRST. If TOKENS_BUFF is + non-NULL, it means that the context owns it, meaning that + _cpp_pop_context will free it as well as VIRT_LOCS_BUFF that + contains the virtual locations. + + A NULL macro means that we should continue the current macro + expansion, in essence. That means that if we are currently in a + macro expansion context, we'll make the new pfile->context refer to + the current macro. */ +static void +push_extended_tokens_context (cpp_reader *pfile, + cpp_hashnode *macro, + _cpp_buff *token_buff, + location_t *virt_locs, + const cpp_token **first, + unsigned int count) +{ + cpp_context *context; + macro_context *m; + + if (macro == NULL) + macro = macro_of_context (pfile->context); + + context = next_context (pfile); + context->tokens_kind = TOKENS_KIND_EXTENDED; + context->buff = token_buff; + + m = XNEW (macro_context); + m->macro_node = macro; + m->virt_locs = virt_locs; + m->cur_virt_loc = virt_locs; + context->c.mc = m; + FIRST (context).ptoken = first; + LAST (context).ptoken = first + count; +} + +/* Push a traditional macro's replacement text. */ +void +_cpp_push_text_context (cpp_reader *pfile, cpp_hashnode *macro, + const uchar *start, size_t len) +{ + cpp_context *context = next_context (pfile); + + context->tokens_kind = TOKENS_KIND_DIRECT; + context->c.macro = macro; + context->buff = NULL; + CUR (context) = start; + RLIMIT (context) = start + len; + macro->flags |= NODE_DISABLED; +} + +/* Creates a buffer that holds tokens a.k.a "token buffer", usually + for the purpose of storing them on a cpp_context. If VIRT_LOCS is + non-null (which means that -ftrack-macro-expansion is on), + *VIRT_LOCS is set to a newly allocated buffer that is supposed to + hold the virtual locations of the tokens resulting from macro + expansion. */ +static _cpp_buff* +tokens_buff_new (cpp_reader *pfile, size_t len, + location_t **virt_locs) +{ + size_t tokens_size = len * sizeof (cpp_token *); + size_t locs_size = len * sizeof (location_t); + + if (virt_locs != NULL) + *virt_locs = XNEWVEC (location_t, locs_size); + return _cpp_get_buff (pfile, tokens_size); +} + +/* Returns the number of tokens contained in a token buffer. The + buffer holds a set of cpp_token*. */ +static size_t +tokens_buff_count (_cpp_buff *buff) +{ + return (BUFF_FRONT (buff) - buff->base) / sizeof (cpp_token *); +} + +/* Return a pointer to the last token contained in the token buffer + BUFF. */ +static const cpp_token ** +tokens_buff_last_token_ptr (_cpp_buff *buff) +{ + if (BUFF_FRONT (buff) == buff->base) + return NULL; + return &((const cpp_token **) BUFF_FRONT (buff))[-1]; +} + +/* Remove the last token contained in the token buffer TOKENS_BUFF. + If VIRT_LOCS_BUFF is non-NULL, it should point at the buffer + containing the virtual locations of the tokens in TOKENS_BUFF; in + which case the function updates that buffer as well. */ +static inline void +tokens_buff_remove_last_token (_cpp_buff *tokens_buff) + +{ + if (BUFF_FRONT (tokens_buff) > tokens_buff->base) + BUFF_FRONT (tokens_buff) = + (unsigned char *) &((cpp_token **) BUFF_FRONT (tokens_buff))[-1]; +} + +/* Insert a token into the token buffer at the position pointed to by + DEST. Note that the buffer is not enlarged so the previous token + that was at *DEST is overwritten. VIRT_LOC_DEST, if non-null, + means -ftrack-macro-expansion is effect; it then points to where to + insert the virtual location of TOKEN. TOKEN is the token to + insert. VIRT_LOC is the virtual location of the token, i.e, the + location possibly encoding its locus across macro expansion. If + TOKEN is an argument of a function-like macro (inside a macro + replacement list), PARM_DEF_LOC is the spelling location of the + macro parameter that TOKEN is replacing, in the replacement list of + the macro. If TOKEN is not an argument of a function-like macro or + if it doesn't come from a macro expansion, then VIRT_LOC can just + be set to the same value as PARM_DEF_LOC. If MAP is non null, it + means TOKEN comes from a macro expansion and MAP is the macro map + associated to the macro. MACRO_TOKEN_INDEX points to the index of + the token in the macro map; it is not considered if MAP is NULL. + + Upon successful completion this function returns the a pointer to + the position of the token coming right after the insertion + point. */ +static inline const cpp_token ** +tokens_buff_put_token_to (const cpp_token **dest, + location_t *virt_loc_dest, + const cpp_token *token, + location_t virt_loc, + location_t parm_def_loc, + const line_map_macro *map, + unsigned int macro_token_index) +{ + location_t macro_loc = virt_loc; + const cpp_token **result; + + if (virt_loc_dest) + { + /* -ftrack-macro-expansion is on. */ + if (map) + macro_loc = linemap_add_macro_token (map, macro_token_index, + virt_loc, parm_def_loc); + *virt_loc_dest = macro_loc; + } + *dest = token; + result = &dest[1]; + + return result; +} + +/* Adds a token at the end of the tokens contained in BUFFER. Note + that this function doesn't enlarge BUFFER when the number of tokens + reaches BUFFER's size; it aborts in that situation. + + TOKEN is the token to append. VIRT_LOC is the virtual location of + the token, i.e, the location possibly encoding its locus across + macro expansion. If TOKEN is an argument of a function-like macro + (inside a macro replacement list), PARM_DEF_LOC is the location of + the macro parameter that TOKEN is replacing. If TOKEN doesn't come + from a macro expansion, then VIRT_LOC can just be set to the same + value as PARM_DEF_LOC. If MAP is non null, it means TOKEN comes + from a macro expansion and MAP is the macro map associated to the + macro. MACRO_TOKEN_INDEX points to the index of the token in the + macro map; It is not considered if MAP is NULL. If VIRT_LOCS is + non-null, it means -ftrack-macro-expansion is on; in which case + this function adds the virtual location DEF_LOC to the VIRT_LOCS + array, at the same index as the one of TOKEN in BUFFER. Upon + successful completion this function returns the a pointer to the + position of the token coming right after the insertion point. */ +static const cpp_token ** +tokens_buff_add_token (_cpp_buff *buffer, + location_t *virt_locs, + const cpp_token *token, + location_t virt_loc, + location_t parm_def_loc, + const line_map_macro *map, + unsigned int macro_token_index) +{ + const cpp_token **result; + location_t *virt_loc_dest = NULL; + unsigned token_index = + (BUFF_FRONT (buffer) - buffer->base) / sizeof (cpp_token *); + + /* Abort if we pass the end the buffer. */ + if (BUFF_FRONT (buffer) > BUFF_LIMIT (buffer)) + abort (); + + if (virt_locs != NULL) + virt_loc_dest = &virt_locs[token_index]; + + result = + tokens_buff_put_token_to ((const cpp_token **) BUFF_FRONT (buffer), + virt_loc_dest, token, virt_loc, parm_def_loc, + map, macro_token_index); + + BUFF_FRONT (buffer) = (unsigned char *) result; + return result; +} + +/* Allocate space for the function-like macro argument ARG to store + the tokens resulting from the macro-expansion of the tokens that + make up ARG itself. That space is allocated in ARG->expanded and + needs to be freed using free. */ +static void +alloc_expanded_arg_mem (cpp_reader *pfile, macro_arg *arg, size_t capacity) +{ + gcc_checking_assert (arg->expanded == NULL + && arg->expanded_virt_locs == NULL); + + arg->expanded = XNEWVEC (const cpp_token *, capacity); + if (CPP_OPTION (pfile, track_macro_expansion)) + arg->expanded_virt_locs = XNEWVEC (location_t, capacity); + +} + +/* If necessary, enlarge ARG->expanded to so that it can contain SIZE + tokens. */ +static void +ensure_expanded_arg_room (cpp_reader *pfile, macro_arg *arg, + size_t size, size_t *expanded_capacity) +{ + if (size <= *expanded_capacity) + return; + + size *= 2; + + arg->expanded = + XRESIZEVEC (const cpp_token *, arg->expanded, size); + *expanded_capacity = size; + + if (CPP_OPTION (pfile, track_macro_expansion)) + { + if (arg->expanded_virt_locs == NULL) + arg->expanded_virt_locs = XNEWVEC (location_t, size); + else + arg->expanded_virt_locs = XRESIZEVEC (location_t, + arg->expanded_virt_locs, + size); + } +} + +/* Expand an argument ARG before replacing parameters in a + function-like macro. This works by pushing a context with the + argument's tokens, and then expanding that into a temporary buffer + as if it were a normal part of the token stream. collect_args() + has terminated the argument's tokens with a CPP_EOF so that we know + when we have fully expanded the argument. */ +static void +expand_arg (cpp_reader *pfile, macro_arg *arg) +{ + size_t capacity; + bool saved_warn_trad; + bool track_macro_exp_p = CPP_OPTION (pfile, track_macro_expansion); + bool saved_ignore__Pragma; + + if (arg->count == 0 + || arg->expanded != NULL) + return; + + /* Don't warn about funlike macros when pre-expanding. */ + saved_warn_trad = CPP_WTRADITIONAL (pfile); + CPP_WTRADITIONAL (pfile) = 0; + + /* Loop, reading in the tokens of the argument. */ + capacity = 256; + alloc_expanded_arg_mem (pfile, arg, capacity); + + if (track_macro_exp_p) + push_extended_tokens_context (pfile, NULL, NULL, + arg->virt_locs, + arg->first, + arg->count + 1); + else + push_ptoken_context (pfile, NULL, NULL, + arg->first, arg->count + 1); + + saved_ignore__Pragma = pfile->state.ignore__Pragma; + pfile->state.ignore__Pragma = 1; + + for (;;) + { + const cpp_token *token; + location_t location; + + ensure_expanded_arg_room (pfile, arg, arg->expanded_count + 1, + &capacity); + + token = cpp_get_token_1 (pfile, &location); + + if (token->type == CPP_EOF) + break; + + set_arg_token (arg, token, location, + arg->expanded_count, MACRO_ARG_TOKEN_EXPANDED, + CPP_OPTION (pfile, track_macro_expansion)); + arg->expanded_count++; + } + + _cpp_pop_context (pfile); + + CPP_WTRADITIONAL (pfile) = saved_warn_trad; + pfile->state.ignore__Pragma = saved_ignore__Pragma; +} + +/* Returns the macro associated to the current context if we are in + the context a macro expansion, NULL otherwise. */ +static cpp_hashnode* +macro_of_context (cpp_context *context) +{ + if (context == NULL) + return NULL; + + return (context->tokens_kind == TOKENS_KIND_EXTENDED) + ? context->c.mc->macro_node + : context->c.macro; +} + +/* Return TRUE iff we are expanding a macro or are about to start + expanding one. If we are effectively expanding a macro, the + function macro_of_context returns a pointer to the macro being + expanded. */ +static bool +in_macro_expansion_p (cpp_reader *pfile) +{ + if (pfile == NULL) + return false; + + return (pfile->about_to_expand_macro_p + || macro_of_context (pfile->context)); +} + +/* Pop the current context off the stack, re-enabling the macro if the + context represented a macro's replacement list. Initially the + context structure was not freed so that we can re-use it later, but + now we do free it to reduce peak memory consumption. */ +void +_cpp_pop_context (cpp_reader *pfile) +{ + cpp_context *context = pfile->context; + + /* We should not be popping the base context. */ + gcc_assert (context != &pfile->base_context); + + if (context->c.macro) + { + cpp_hashnode *macro; + if (context->tokens_kind == TOKENS_KIND_EXTENDED) + { + macro_context *mc = context->c.mc; + macro = mc->macro_node; + /* If context->buff is set, it means the life time of tokens + is bound to the life time of this context; so we must + free the tokens; that means we must free the virtual + locations of these tokens too. */ + if (context->buff && mc->virt_locs) + { + free (mc->virt_locs); + mc->virt_locs = NULL; + } + free (mc); + context->c.mc = NULL; + } + else + macro = context->c.macro; + + /* Beware that MACRO can be NULL in cases like when we are + called from expand_arg. In those cases, a dummy context with + tokens is pushed just for the purpose of walking them using + cpp_get_token_1. In that case, no 'macro' field is set into + the dummy context. */ + if (macro != NULL + /* Several contiguous macro expansion contexts can be + associated to the same macro; that means it's the same + macro expansion that spans across all these (sub) + contexts. So we should re-enable an expansion-disabled + macro only when we are sure we are really out of that + macro expansion. */ + && macro_of_context (context->prev) != macro) + macro->flags &= ~NODE_DISABLED; + + if (macro == pfile->top_most_macro_node && context->prev == NULL) + /* We are popping the context of the top-most macro node. */ + pfile->top_most_macro_node = NULL; + } + + if (context->buff) + { + /* Decrease memory peak consumption by freeing the memory used + by the context. */ + _cpp_free_buff (context->buff); + } + + pfile->context = context->prev; + /* decrease peak memory consumption by feeing the context. */ + pfile->context->next = NULL; + free (context); +} + +/* Return TRUE if we reached the end of the set of tokens stored in + CONTEXT, FALSE otherwise. */ +static inline bool +reached_end_of_context (cpp_context *context) +{ + if (context->tokens_kind == TOKENS_KIND_DIRECT) + return FIRST (context).token == LAST (context).token; + else if (context->tokens_kind == TOKENS_KIND_INDIRECT + || context->tokens_kind == TOKENS_KIND_EXTENDED) + return FIRST (context).ptoken == LAST (context).ptoken; + else + abort (); +} + +/* Consume the next token contained in the current context of PFILE, + and return it in *TOKEN. It's "full location" is returned in + *LOCATION. If -ftrack-macro-location is in effeect, fFull location" + means the location encoding the locus of the token across macro + expansion; otherwise it's just is the "normal" location of the + token which (*TOKEN)->src_loc. */ +static inline void +consume_next_token_from_context (cpp_reader *pfile, + const cpp_token ** token, + location_t *location) +{ + cpp_context *c = pfile->context; + + if ((c)->tokens_kind == TOKENS_KIND_DIRECT) + { + *token = FIRST (c).token; + *location = (*token)->src_loc; + FIRST (c).token++; + } + else if ((c)->tokens_kind == TOKENS_KIND_INDIRECT) + { + *token = *FIRST (c).ptoken; + *location = (*token)->src_loc; + FIRST (c).ptoken++; + } + else if ((c)->tokens_kind == TOKENS_KIND_EXTENDED) + { + macro_context *m = c->c.mc; + *token = *FIRST (c).ptoken; + if (m->virt_locs) + { + *location = *m->cur_virt_loc; + m->cur_virt_loc++; + } + else + *location = (*token)->src_loc; + FIRST (c).ptoken++; + } + else + abort (); +} + +/* In the traditional mode of the preprocessor, if we are currently in + a directive, the location of a token must be the location of the + start of the directive line. This function returns the proper + location if we are in the traditional mode, and just returns + LOCATION otherwise. */ + +static inline location_t +maybe_adjust_loc_for_trad_cpp (cpp_reader *pfile, location_t location) +{ + if (CPP_OPTION (pfile, traditional)) + { + if (pfile->state.in_directive) + return pfile->directive_line; + } + return location; +} + +/* Routine to get a token as well as its location. + + Macro expansions and directives are transparently handled, + including entering included files. Thus tokens are post-macro + expansion, and after any intervening directives. External callers + see CPP_EOF only at EOF. Internal callers also see it when meeting + a directive inside a macro call, when at the end of a directive and + state.in_directive is still 1, and at the end of argument + pre-expansion. + + LOC is an out parameter; *LOC is set to the location "as expected + by the user". Please read the comment of + cpp_get_token_with_location to learn more about the meaning of this + location. */ +static const cpp_token* +cpp_get_token_1 (cpp_reader *pfile, location_t *location) +{ + const cpp_token *result; + /* This token is a virtual token that either encodes a location + related to macro expansion or a spelling location. */ + location_t virt_loc = 0; + /* pfile->about_to_expand_macro_p can be overriden by indirect calls + to functions that push macro contexts. So let's save it so that + we can restore it when we are about to leave this routine. */ + bool saved_about_to_expand_macro = pfile->about_to_expand_macro_p; + + for (;;) + { + cpp_hashnode *node; + cpp_context *context = pfile->context; + + /* Context->prev == 0 <=> base context. */ + if (!context->prev) + { + result = _cpp_lex_token (pfile); + virt_loc = result->src_loc; + } + else if (!reached_end_of_context (context)) + { + consume_next_token_from_context (pfile, &result, + &virt_loc); + if (result->flags & PASTE_LEFT) + { + paste_all_tokens (pfile, result); + if (pfile->state.in_directive) + continue; + result = padding_token (pfile, result); + goto out; + } + } + else + { + if (pfile->context->c.macro) + ++num_expanded_macros_counter; + _cpp_pop_context (pfile); + if (pfile->state.in_directive) + continue; + result = &pfile->avoid_paste; + goto out; + } + + if (pfile->state.in_directive && result->type == CPP_COMMENT) + continue; + + if (result->type != CPP_NAME) + break; + + node = result->val.node.node; + + if (node->type == NT_VOID || (result->flags & NO_EXPAND)) + break; + + if (!(node->flags & NODE_USED) + && node->type == NT_USER_MACRO + && !node->value.macro + && !cpp_get_deferred_macro (pfile, node, result->src_loc)) + break; + + if (!(node->flags & NODE_DISABLED)) + { + int ret = 0; + /* If not in a macro context, and we're going to start an + expansion, record the location and the top level macro + about to be expanded. */ + if (!in_macro_expansion_p (pfile)) + { + pfile->invocation_location = result->src_loc; + pfile->top_most_macro_node = node; + } + if (pfile->state.prevent_expansion) + break; + + /* Conditional macros require that a predicate be evaluated + first. */ + if ((node->flags & NODE_CONDITIONAL) != 0) + { + if (pfile->cb.macro_to_expand) + { + bool whitespace_after; + const cpp_token *peek_tok = cpp_peek_token (pfile, 0); + + whitespace_after = (peek_tok->type == CPP_PADDING + || (peek_tok->flags & PREV_WHITE)); + node = pfile->cb.macro_to_expand (pfile, result); + if (node) + ret = enter_macro_context (pfile, node, result, virt_loc); + else if (whitespace_after) + { + /* If macro_to_expand hook returned NULL and it + ate some tokens, see if we don't need to add + a padding token in between this and the + next token. */ + peek_tok = cpp_peek_token (pfile, 0); + if (peek_tok->type != CPP_PADDING + && (peek_tok->flags & PREV_WHITE) == 0) + _cpp_push_token_context (pfile, NULL, + padding_token (pfile, + peek_tok), 1); + } + } + } + else + ret = enter_macro_context (pfile, node, result, virt_loc); + if (ret) + { + if (pfile->state.in_directive || ret == 2) + continue; + result = padding_token (pfile, result); + goto out; + } + } + else + { + /* Flag this token as always unexpandable. FIXME: move this + to collect_args()?. */ + cpp_token *t = _cpp_temp_token (pfile); + t->type = result->type; + t->flags = result->flags | NO_EXPAND; + t->val = result->val; + result = t; + } + + break; + } + + out: + if (location != NULL) + { + if (virt_loc == 0) + virt_loc = result->src_loc; + *location = virt_loc; + + if (!CPP_OPTION (pfile, track_macro_expansion) + && macro_of_context (pfile->context) != NULL) + /* We are in a macro expansion context, are not tracking + virtual location, but were asked to report the location + of the expansion point of the macro being expanded. */ + *location = pfile->invocation_location; + + *location = maybe_adjust_loc_for_trad_cpp (pfile, *location); + } + + pfile->about_to_expand_macro_p = saved_about_to_expand_macro; + + if (pfile->state.directive_file_token + && !pfile->state.parsing_args + && !(result->type == CPP_PADDING || result->type == CPP_COMMENT) + && !(15 & --pfile->state.directive_file_token)) + { + /* Do header-name frobbery. Concatenate < ... > as approprate. + Do header search if needed, and finally drop the outer <> or + "". */ + pfile->state.angled_headers = false; + + /* Do angle-header reconstitution. Then do include searching. + We'll always end up with a ""-quoted header-name in that + case. If searching finds nothing, we emit a diagnostic and + an empty string. */ + size_t len = 0; + char *fname = NULL; + + cpp_token *tmp = _cpp_temp_token (pfile); + *tmp = *result; + + tmp->type = CPP_HEADER_NAME; + bool need_search = !pfile->state.directive_file_token; + pfile->state.directive_file_token = 0; + + bool angle = result->type != CPP_STRING; + if (result->type == CPP_HEADER_NAME + || (result->type == CPP_STRING && result->val.str.text[0] != 'R')) + { + len = result->val.str.len - 2; + fname = XNEWVEC (char, len + 1); + memcpy (fname, result->val.str.text + 1, len); + fname[len] = 0; + } + else if (result->type == CPP_LESS) + fname = _cpp_bracket_include (pfile); + + if (fname) + { + /* We have a header-name. Look it up. This will emit an + unfound diagnostic. Canonicalize the found name. */ + const char *found = fname; + + if (need_search) + { + found = _cpp_find_header_unit (pfile, fname, angle, tmp->src_loc); + if (!found) + found = ""; + len = strlen (found); + } + /* Force a leading './' if it's not absolute. */ + bool dotme = (found[0] == '.' ? !IS_DIR_SEPARATOR (found[1]) + : found[0] && !IS_ABSOLUTE_PATH (found)); + + if (BUFF_ROOM (pfile->u_buff) < len + 1 + dotme * 2) + _cpp_extend_buff (pfile, &pfile->u_buff, len + 1 + dotme * 2); + unsigned char *buf = BUFF_FRONT (pfile->u_buff); + size_t pos = 0; + + if (dotme) + { + buf[pos++] = '.'; + /* Apparently '/' is unconditional. */ + buf[pos++] = '/'; + } + memcpy (&buf[pos], found, len); + pos += len; + buf[pos] = 0; + + tmp->val.str.len = pos; + tmp->val.str.text = buf; + + tmp->type = CPP_HEADER_NAME; + XDELETEVEC (fname); + + result = tmp; + } + } + + return result; +} + +/* External routine to get a token. Also used nearly everywhere + internally, except for places where we know we can safely call + _cpp_lex_token directly, such as lexing a directive name. + + Macro expansions and directives are transparently handled, + including entering included files. Thus tokens are post-macro + expansion, and after any intervening directives. External callers + see CPP_EOF only at EOF. Internal callers also see it when meeting + a directive inside a macro call, when at the end of a directive and + state.in_directive is still 1, and at the end of argument + pre-expansion. */ +const cpp_token * +cpp_get_token (cpp_reader *pfile) +{ + return cpp_get_token_1 (pfile, NULL); +} + +/* Like cpp_get_token, but also returns a virtual token location + separate from the spelling location carried by the returned token. + + LOC is an out parameter; *LOC is set to the location "as expected + by the user". This matters when a token results from macro + expansion; in that case the token's spelling location indicates the + locus of the token in the definition of the macro but *LOC + virtually encodes all the other meaningful locuses associated to + the token. + + What? virtual location? Yes, virtual location. + + If the token results from macro expansion and if macro expansion + location tracking is enabled its virtual location encodes (at the + same time): + + - the spelling location of the token + + - the locus of the macro expansion point + + - the locus of the point where the token got instantiated as part + of the macro expansion process. + + You have to use the linemap API to get the locus you are interested + in from a given virtual location. + + Note however that virtual locations are not necessarily ordered for + relations '<' and '>'. One must use the function + linemap_location_before_p instead of using the relational operator + '<'. + + If macro expansion tracking is off and if the token results from + macro expansion the virtual location is the expansion point of the + macro that got expanded. + + When the token doesn't result from macro expansion, the virtual + location is just the same thing as its spelling location. */ + +const cpp_token * +cpp_get_token_with_location (cpp_reader *pfile, location_t *loc) +{ + return cpp_get_token_1 (pfile, loc); +} + +/* Returns true if we're expanding an object-like macro that was + defined in a system header. Just checks the macro at the top of + the stack. Used for diagnostic suppression. + Also return true for builtin macros. */ +int +cpp_sys_macro_p (cpp_reader *pfile) +{ + cpp_hashnode *node = NULL; + + if (pfile->context->tokens_kind == TOKENS_KIND_EXTENDED) + node = pfile->context->c.mc->macro_node; + else + node = pfile->context->c.macro; + + if (!node) + return false; + if (cpp_builtin_macro_p (node)) + return true; + return node->value.macro && node->value.macro->syshdr; +} + +/* Read each token in, until end of the current file. Directives are + transparently processed. */ +void +cpp_scan_nooutput (cpp_reader *pfile) +{ + /* Request a CPP_EOF token at the end of this file, rather than + transparently continuing with the including file. */ + pfile->buffer->return_at_eof = true; + + pfile->state.discarding_output++; + pfile->state.prevent_expansion++; + + if (CPP_OPTION (pfile, traditional)) + while (_cpp_read_logical_line_trad (pfile)) + ; + else + while (cpp_get_token (pfile)->type != CPP_EOF) + ; + + pfile->state.discarding_output--; + pfile->state.prevent_expansion--; +} + +/* Step back one or more tokens obtained from the lexer. */ +void +_cpp_backup_tokens_direct (cpp_reader *pfile, unsigned int count) +{ + pfile->lookaheads += count; + while (count--) + { + pfile->cur_token--; + if (pfile->cur_token == pfile->cur_run->base + /* Possible with -fpreprocessed and no leading #line. */ + && pfile->cur_run->prev != NULL) + { + pfile->cur_run = pfile->cur_run->prev; + pfile->cur_token = pfile->cur_run->limit; + } + } +} + +/* Step back one (or more) tokens. Can only step back more than 1 if + they are from the lexer, and not from macro expansion. */ +void +_cpp_backup_tokens (cpp_reader *pfile, unsigned int count) +{ + if (pfile->context->prev == NULL) + _cpp_backup_tokens_direct (pfile, count); + else + { + if (count != 1) + abort (); + if (pfile->context->tokens_kind == TOKENS_KIND_DIRECT) + FIRST (pfile->context).token--; + else if (pfile->context->tokens_kind == TOKENS_KIND_INDIRECT) + FIRST (pfile->context).ptoken--; + else if (pfile->context->tokens_kind == TOKENS_KIND_EXTENDED) + { + FIRST (pfile->context).ptoken--; + if (pfile->context->c.macro) + { + macro_context *m = pfile->context->c.mc; + m->cur_virt_loc--; + gcc_checking_assert (m->cur_virt_loc >= m->virt_locs); + } + else + abort (); + } + else + abort (); + } +} + +/* #define directive parsing and handling. */ + +/* Returns true if a macro redefinition warning is required. */ +static bool +warn_of_redefinition (cpp_reader *pfile, cpp_hashnode *node, + const cpp_macro *macro2) +{ + /* Some redefinitions need to be warned about regardless. */ + if (node->flags & NODE_WARN) + return true; + + /* Suppress warnings for builtins that lack the NODE_WARN flag, + unless Wbuiltin-macro-redefined. */ + if (cpp_builtin_macro_p (node)) + return CPP_OPTION (pfile, warn_builtin_macro_redefined); + + /* Redefinitions of conditional (context-sensitive) macros, on + the other hand, must be allowed silently. */ + if (node->flags & NODE_CONDITIONAL) + return false; + + if (cpp_macro *macro1 = get_deferred_or_lazy_macro (pfile, node, macro2->line)) + return cpp_compare_macros (macro1, macro2); + return false; +} + +/* Return TRUE if MACRO1 and MACRO2 differ. */ + +bool +cpp_compare_macros (const cpp_macro *macro1, const cpp_macro *macro2) +{ + /* Redefinition of a macro is allowed if and only if the old and new + definitions are the same. (6.10.3 paragraph 2). */ + + /* Don't check count here as it can be different in valid + traditional redefinitions with just whitespace differences. */ + if (macro1->paramc != macro2->paramc + || macro1->fun_like != macro2->fun_like + || macro1->variadic != macro2->variadic) + return true; + + /* Check parameter spellings. */ + for (unsigned i = macro1->paramc; i--; ) + if (macro1->parm.params[i] != macro2->parm.params[i]) + return true; + + /* Check the replacement text or tokens. */ + if (macro1->kind == cmk_traditional) + return _cpp_expansions_different_trad (macro1, macro2); + + if (macro1->count != macro2->count) + return true; + + for (unsigned i= macro1->count; i--; ) + if (!_cpp_equiv_tokens (¯o1->exp.tokens[i], ¯o2->exp.tokens[i])) + return true; + + return false; +} + +/* Free the definition of hashnode H. */ +void +_cpp_free_definition (cpp_hashnode *h) +{ + /* Macros and assertions no longer have anything to free. */ + h->type = NT_VOID; + h->value.answers = NULL; + h->flags &= ~(NODE_DISABLED | NODE_USED); +} + +/* Save parameter NODE (spelling SPELLING) to the parameter list of + macro MACRO. Returns true on success, false on failure. */ +bool +_cpp_save_parameter (cpp_reader *pfile, unsigned n, cpp_hashnode *node, + cpp_hashnode *spelling) +{ + /* Constraint 6.10.3.6 - duplicate parameter names. */ + if (node->type == NT_MACRO_ARG) + { + cpp_error (pfile, CPP_DL_ERROR, "duplicate macro parameter \"%s\"", + NODE_NAME (node)); + return false; + } + + unsigned len = (n + 1) * sizeof (struct macro_arg_saved_data); + if (len > pfile->macro_buffer_len) + { + pfile->macro_buffer + = XRESIZEVEC (unsigned char, pfile->macro_buffer, len); + pfile->macro_buffer_len = len; + } + + macro_arg_saved_data *saved = (macro_arg_saved_data *)pfile->macro_buffer; + saved[n].canonical_node = node; + saved[n].value = node->value; + saved[n].type = node->type; + + void *base = _cpp_reserve_room (pfile, n * sizeof (cpp_hashnode *), + sizeof (cpp_hashnode *)); + ((cpp_hashnode **)base)[n] = spelling; + + /* Morph into a macro arg. */ + node->type = NT_MACRO_ARG; + /* Index is 1 based. */ + node->value.arg_index = n + 1; + + return true; +} + +/* Restore the parameters to their previous state. */ +void +_cpp_unsave_parameters (cpp_reader *pfile, unsigned n) +{ + /* Clear the fast argument lookup indices. */ + while (n--) + { + struct macro_arg_saved_data *save = + &((struct macro_arg_saved_data *) pfile->macro_buffer)[n]; + + struct cpp_hashnode *node = save->canonical_node; + node->type = save->type; + node->value = save->value; + } +} + +/* Check the syntax of the parameters in a MACRO definition. Return + false on failure. Set *N_PTR and *VARADIC_PTR as appropriate. + '(' ')' + '(' parm-list ',' last-parm ')' + '(' last-parm ')' + parm-list: name + | parm-list, name + last-parm: name + | name '...' + | '...' +*/ + +static bool +parse_params (cpp_reader *pfile, unsigned *n_ptr, bool *varadic_ptr) +{ + unsigned nparms = 0; + bool ok = false; + + for (bool prev_ident = false;;) + { + const cpp_token *token = _cpp_lex_token (pfile); + + switch (token->type) + { + case CPP_COMMENT: + /* Allow/ignore comments in parameter lists if we are + preserving comments in macro expansions. */ + if (!CPP_OPTION (pfile, discard_comments_in_macro_exp)) + break; + + /* FALLTHRU */ + default: + bad: + { + const char *const msgs[5] = + { + N_("expected parameter name, found \"%s\""), + N_("expected ',' or ')', found \"%s\""), + N_("expected parameter name before end of line"), + N_("expected ')' before end of line"), + N_("expected ')' after \"...\"") + }; + unsigned ix = prev_ident; + const unsigned char *as_text = NULL; + if (*varadic_ptr) + ix = 4; + else if (token->type == CPP_EOF) + ix += 2; + else + as_text = cpp_token_as_text (pfile, token); + cpp_error (pfile, CPP_DL_ERROR, msgs[ix], as_text); + } + goto out; + + case CPP_NAME: + if (prev_ident || *varadic_ptr) + goto bad; + prev_ident = true; + + if (!_cpp_save_parameter (pfile, nparms, token->val.node.node, + token->val.node.spelling)) + goto out; + nparms++; + break; + + case CPP_CLOSE_PAREN: + if (prev_ident || !nparms || *varadic_ptr) + { + ok = true; + goto out; + } + + /* FALLTHRU */ + case CPP_COMMA: + if (!prev_ident || *varadic_ptr) + goto bad; + prev_ident = false; + break; + + case CPP_ELLIPSIS: + if (*varadic_ptr) + goto bad; + *varadic_ptr = true; + if (!prev_ident) + { + /* An ISO bare ellipsis. */ + _cpp_save_parameter (pfile, nparms, + pfile->spec_nodes.n__VA_ARGS__, + pfile->spec_nodes.n__VA_ARGS__); + nparms++; + pfile->state.va_args_ok = 1; + if (! CPP_OPTION (pfile, c99) + && CPP_OPTION (pfile, cpp_pedantic) + && CPP_OPTION (pfile, warn_variadic_macros)) + cpp_pedwarning + (pfile, CPP_W_VARIADIC_MACROS, + CPP_OPTION (pfile, cplusplus) + ? N_("anonymous variadic macros were introduced in C++11") + : N_("anonymous variadic macros were introduced in C99")); + else if (CPP_OPTION (pfile, cpp_warn_c90_c99_compat) > 0 + && ! CPP_OPTION (pfile, cplusplus)) + cpp_error (pfile, CPP_DL_WARNING, + "anonymous variadic macros were introduced in C99"); + } + else if (CPP_OPTION (pfile, cpp_pedantic) + && CPP_OPTION (pfile, warn_variadic_macros)) + cpp_pedwarning (pfile, CPP_W_VARIADIC_MACROS, + CPP_OPTION (pfile, cplusplus) + ? N_("ISO C++ does not permit named variadic macros") + : N_("ISO C does not permit named variadic macros")); + break; + } + } + + out: + *n_ptr = nparms; + + return ok; +} + +/* Lex a token from the expansion of MACRO, but mark parameters as we + find them and warn of traditional stringification. */ +static cpp_macro * +lex_expansion_token (cpp_reader *pfile, cpp_macro *macro) +{ + macro = (cpp_macro *)_cpp_reserve_room (pfile, + sizeof (cpp_macro) - sizeof (cpp_token) + + macro->count * sizeof (cpp_token), + sizeof (cpp_token)); + cpp_token *saved_cur_token = pfile->cur_token; + pfile->cur_token = ¯o->exp.tokens[macro->count]; + cpp_token *token = _cpp_lex_direct (pfile); + pfile->cur_token = saved_cur_token; + + /* Is this a parameter? */ + if (token->type == CPP_NAME && token->val.node.node->type == NT_MACRO_ARG) + { + /* Morph into a parameter reference. */ + cpp_hashnode *spelling = token->val.node.spelling; + token->type = CPP_MACRO_ARG; + token->val.macro_arg.arg_no = token->val.node.node->value.arg_index; + token->val.macro_arg.spelling = spelling; + } + else if (CPP_WTRADITIONAL (pfile) && macro->paramc > 0 + && (token->type == CPP_STRING || token->type == CPP_CHAR)) + check_trad_stringification (pfile, macro, &token->val.str); + + return macro; +} + +static cpp_macro * +create_iso_definition (cpp_reader *pfile) +{ + bool following_paste_op = false; + const char *paste_op_error_msg = + N_("'##' cannot appear at either end of a macro expansion"); + unsigned int num_extra_tokens = 0; + unsigned nparms = 0; + cpp_hashnode **params = NULL; + bool varadic = false; + bool ok = false; + cpp_macro *macro = NULL; + + /* Look at the first token, to see if this is a function-like + macro. */ + cpp_token first; + cpp_token *saved_cur_token = pfile->cur_token; + pfile->cur_token = &first; + cpp_token *token = _cpp_lex_direct (pfile); + pfile->cur_token = saved_cur_token; + + if (token->flags & PREV_WHITE) + /* Preceeded by space, must be part of expansion. */; + else if (token->type == CPP_OPEN_PAREN) + { + /* An open-paren, get a parameter list. */ + if (!parse_params (pfile, &nparms, &varadic)) + goto out; + + params = (cpp_hashnode **)_cpp_commit_buff + (pfile, sizeof (cpp_hashnode *) * nparms); + token = NULL; + } + else if (token->type != CPP_EOF + && !(token->type == CPP_COMMENT + && ! CPP_OPTION (pfile, discard_comments_in_macro_exp))) + { + /* While ISO C99 requires whitespace before replacement text + in a macro definition, ISO C90 with TC1 allows characters + from the basic source character set there. */ + if (CPP_OPTION (pfile, c99)) + cpp_error (pfile, CPP_DL_PEDWARN, + CPP_OPTION (pfile, cplusplus) + ? N_("ISO C++11 requires whitespace after the macro name") + : N_("ISO C99 requires whitespace after the macro name")); + else + { + enum cpp_diagnostic_level warntype = CPP_DL_WARNING; + switch (token->type) + { + case CPP_ATSIGN: + case CPP_AT_NAME: + case CPP_OBJC_STRING: + /* '@' is not in basic character set. */ + warntype = CPP_DL_PEDWARN; + break; + case CPP_OTHER: + /* Basic character set sans letters, digits and _. */ + if (strchr ("!\"#%&'()*+,-./:;<=>?[\\]^{|}~", + token->val.str.text[0]) == NULL) + warntype = CPP_DL_PEDWARN; + break; + default: + /* All other tokens start with a character from basic + character set. */ + break; + } + cpp_error (pfile, warntype, + "missing whitespace after the macro name"); + } + } + + macro = _cpp_new_macro (pfile, cmk_macro, + _cpp_reserve_room (pfile, 0, sizeof (cpp_macro))); + + if (!token) + { + macro->variadic = varadic; + macro->paramc = nparms; + macro->parm.params = params; + macro->fun_like = true; + } + else + { + /* Preserve the token we peeked, there is already a single slot for it. */ + macro->exp.tokens[0] = *token; + token = ¯o->exp.tokens[0]; + macro->count = 1; + } + + for (vaopt_state vaopt_tracker (pfile, macro->variadic, NULL);; token = NULL) + { + if (!token) + { + macro = lex_expansion_token (pfile, macro); + token = ¯o->exp.tokens[macro->count++]; + } + + /* Check the stringifying # constraint 6.10.3.2.1 of + function-like macros when lexing the subsequent token. */ + if (macro->count > 1 && token[-1].type == CPP_HASH && macro->fun_like) + { + if (token->type == CPP_MACRO_ARG + || (macro->variadic + && token->type == CPP_NAME + && token->val.node.node == pfile->spec_nodes.n__VA_OPT__)) + { + if (token->flags & PREV_WHITE) + token->flags |= SP_PREV_WHITE; + if (token[-1].flags & DIGRAPH) + token->flags |= SP_DIGRAPH; + token->flags &= ~PREV_WHITE; + token->flags |= STRINGIFY_ARG; + token->flags |= token[-1].flags & PREV_WHITE; + token[-1] = token[0]; + macro->count--; + } + /* Let assembler get away with murder. */ + else if (CPP_OPTION (pfile, lang) != CLK_ASM) + { + cpp_error (pfile, CPP_DL_ERROR, + "'#' is not followed by a macro parameter"); + goto out; + } + } + + if (token->type == CPP_EOF) + { + /* Paste operator constraint 6.10.3.3.1: + Token-paste ##, can appear in both object-like and + function-like macros, but not at the end. */ + if (following_paste_op) + { + cpp_error (pfile, CPP_DL_ERROR, paste_op_error_msg); + goto out; + } + if (!vaopt_tracker.completed ()) + goto out; + break; + } + + /* Paste operator constraint 6.10.3.3.1. */ + if (token->type == CPP_PASTE) + { + /* Token-paste ##, can appear in both object-like and + function-like macros, but not at the beginning. */ + if (macro->count == 1) + { + cpp_error (pfile, CPP_DL_ERROR, paste_op_error_msg); + goto out; + } + + if (following_paste_op) + { + /* Consecutive paste operators. This one will be moved + to the end. */ + num_extra_tokens++; + token->val.token_no = macro->count - 1; + } + else + { + /* Drop the paste operator. */ + --macro->count; + token[-1].flags |= PASTE_LEFT; + if (token->flags & DIGRAPH) + token[-1].flags |= SP_DIGRAPH; + if (token->flags & PREV_WHITE) + token[-1].flags |= SP_PREV_WHITE; + } + following_paste_op = true; + } + else + following_paste_op = false; + + if (vaopt_tracker.update (token) == vaopt_state::ERROR) + goto out; + } + + /* We're committed to winning now. */ + ok = true; + + /* Don't count the CPP_EOF. */ + macro->count--; + + macro = (cpp_macro *)_cpp_commit_buff + (pfile, sizeof (cpp_macro) - sizeof (cpp_token) + + sizeof (cpp_token) * macro->count); + + /* Clear whitespace on first token. */ + if (macro->count) + macro->exp.tokens[0].flags &= ~PREV_WHITE; + + if (num_extra_tokens) + { + /* Place second and subsequent ## or %:%: tokens in sequences of + consecutive such tokens at the end of the list to preserve + information about where they appear, how they are spelt and + whether they are preceded by whitespace without otherwise + interfering with macro expansion. Remember, this is + extremely rare, so efficiency is not a priority. */ + cpp_token *temp = (cpp_token *)_cpp_reserve_room + (pfile, 0, num_extra_tokens * sizeof (cpp_token)); + unsigned extra_ix = 0, norm_ix = 0; + cpp_token *exp = macro->exp.tokens; + for (unsigned ix = 0; ix != macro->count; ix++) + if (exp[ix].type == CPP_PASTE) + temp[extra_ix++] = exp[ix]; + else + exp[norm_ix++] = exp[ix]; + memcpy (&exp[norm_ix], temp, num_extra_tokens * sizeof (cpp_token)); + + /* Record there are extra tokens. */ + macro->extra_tokens = 1; + } + + out: + pfile->state.va_args_ok = 0; + _cpp_unsave_parameters (pfile, nparms); + + return ok ? macro : NULL; +} + +cpp_macro * +_cpp_new_macro (cpp_reader *pfile, cpp_macro_kind kind, void *placement) +{ + cpp_macro *macro = (cpp_macro *) placement; + + /* Zero init all the fields. This'll tell the compiler know all the + following inits are writing a virgin object. */ + memset (macro, 0, offsetof (cpp_macro, exp)); + + macro->line = pfile->directive_line; + macro->parm.params = 0; + macro->lazy = 0; + macro->paramc = 0; + macro->variadic = 0; + macro->used = !CPP_OPTION (pfile, warn_unused_macros); + macro->count = 0; + macro->fun_like = 0; + macro->imported_p = false; + macro->extra_tokens = 0; + /* To suppress some diagnostics. */ + macro->syshdr = pfile->buffer && pfile->buffer->sysp != 0; + + macro->kind = kind; + + return macro; +} + +/* Parse a macro and save its expansion. Returns nonzero on success. */ +bool +_cpp_create_definition (cpp_reader *pfile, cpp_hashnode *node) +{ + cpp_macro *macro; + + if (CPP_OPTION (pfile, traditional)) + macro = _cpp_create_trad_definition (pfile); + else + macro = create_iso_definition (pfile); + + if (!macro) + return false; + + if (cpp_macro_p (node)) + { + if (CPP_OPTION (pfile, warn_unused_macros)) + _cpp_warn_if_unused_macro (pfile, node, NULL); + + if (warn_of_redefinition (pfile, node, macro)) + { + const enum cpp_warning_reason reason + = (cpp_builtin_macro_p (node) && !(node->flags & NODE_WARN)) + ? CPP_W_BUILTIN_MACRO_REDEFINED : CPP_W_NONE; + + bool warned = + cpp_pedwarning_with_line (pfile, reason, + pfile->directive_line, 0, + "\"%s\" redefined", NODE_NAME (node)); + + if (warned && cpp_user_macro_p (node)) + cpp_error_with_line (pfile, CPP_DL_NOTE, + node->value.macro->line, 0, + "this is the location of the previous definition"); + } + _cpp_free_definition (node); + } + + /* Enter definition in hash table. */ + node->type = NT_USER_MACRO; + node->value.macro = macro; + if (! ustrncmp (NODE_NAME (node), DSC ("__STDC_")) + && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_FORMAT_MACROS") + /* __STDC_LIMIT_MACROS and __STDC_CONSTANT_MACROS are mentioned + in the C standard, as something that one must use in C++. + However DR#593 and C++11 indicate that they play no role in C++. + We special-case them anyway. */ + && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_LIMIT_MACROS") + && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_CONSTANT_MACROS")) + node->flags |= NODE_WARN; + + /* If user defines one of the conditional macros, remove the + conditional flag */ + node->flags &= ~NODE_CONDITIONAL; + + return true; +} + +extern void +cpp_define_lazily (cpp_reader *pfile, cpp_hashnode *node, unsigned num) +{ + cpp_macro *macro = node->value.macro; + + gcc_checking_assert (pfile->cb.user_lazy_macro && macro && num < UCHAR_MAX); + + macro->lazy = num + 1; +} + +/* NODE is a deferred macro, resolve it, returning the definition + (which may be NULL). */ +cpp_macro * +cpp_get_deferred_macro (cpp_reader *pfile, cpp_hashnode *node, + location_t loc) +{ + gcc_checking_assert (node->type == NT_USER_MACRO); + + node->value.macro = pfile->cb.user_deferred_macro (pfile, loc, node); + + if (!node->value.macro) + node->type = NT_VOID; + + return node->value.macro; +} + +static cpp_macro * +get_deferred_or_lazy_macro (cpp_reader *pfile, cpp_hashnode *node, + location_t loc) +{ + cpp_macro *macro = node->value.macro; + if (!macro) + { + macro = cpp_get_deferred_macro (pfile, node, loc); + gcc_checking_assert (!macro || !macro->lazy); + } + else if (macro->lazy) + { + pfile->cb.user_lazy_macro (pfile, macro, macro->lazy - 1); + macro->lazy = 0; + } + + return macro; +} + +/* Notify the use of NODE in a macro-aware context (i.e. expanding it, + or testing its existance). Also applies any lazy definition. + Return FALSE if the macro isn't really there. */ + +extern bool +_cpp_notify_macro_use (cpp_reader *pfile, cpp_hashnode *node, + location_t loc) +{ + node->flags |= NODE_USED; + switch (node->type) + { + case NT_USER_MACRO: + if (!get_deferred_or_lazy_macro (pfile, node, loc)) + return false; + /* FALLTHROUGH. */ + + case NT_BUILTIN_MACRO: + if (pfile->cb.used_define) + pfile->cb.used_define (pfile, loc, node); + break; + + case NT_VOID: + if (pfile->cb.used_undef) + pfile->cb.used_undef (pfile, loc, node); + break; + + default: + abort (); + } + + return true; +} + +/* Warn if a token in STRING matches one of a function-like MACRO's + parameters. */ +static void +check_trad_stringification (cpp_reader *pfile, const cpp_macro *macro, + const cpp_string *string) +{ + unsigned int i, len; + const uchar *p, *q, *limit; + + /* Loop over the string. */ + limit = string->text + string->len - 1; + for (p = string->text + 1; p < limit; p = q) + { + /* Find the start of an identifier. */ + while (p < limit && !is_idstart (*p)) + p++; + + /* Find the end of the identifier. */ + q = p; + while (q < limit && is_idchar (*q)) + q++; + + len = q - p; + + /* Loop over the function macro arguments to see if the + identifier inside the string matches one of them. */ + for (i = 0; i < macro->paramc; i++) + { + const cpp_hashnode *node = macro->parm.params[i]; + + if (NODE_LEN (node) == len + && !memcmp (p, NODE_NAME (node), len)) + { + cpp_warning (pfile, CPP_W_TRADITIONAL, + "macro argument \"%s\" would be stringified in traditional C", + NODE_NAME (node)); + break; + } + } + } +} + +/* Returns the name, arguments and expansion of a macro, in a format + suitable to be read back in again, and therefore also for DWARF 2 + debugging info. e.g. "PASTE(X, Y) X ## Y", or "MACNAME EXPANSION". + Caller is expected to generate the "#define" bit if needed. The + returned text is temporary, and automatically freed later. */ +const unsigned char * +cpp_macro_definition (cpp_reader *pfile, cpp_hashnode *node) +{ + gcc_checking_assert (cpp_user_macro_p (node)); + + if (const cpp_macro *macro = get_deferred_or_lazy_macro (pfile, node, 0)) + return cpp_macro_definition (pfile, node, macro); + return NULL; +} + +const unsigned char * +cpp_macro_definition (cpp_reader *pfile, cpp_hashnode *node, + const cpp_macro *macro) +{ + unsigned int i, len; + unsigned char *buffer; + + /* Calculate length. */ + len = NODE_LEN (node) * 10 + 2; /* ' ' and NUL. */ + if (macro->fun_like) + { + len += 4; /* "()" plus possible final ".." of named + varargs (we have + 1 below). */ + for (i = 0; i < macro->paramc; i++) + len += NODE_LEN (macro->parm.params[i]) + 1; /* "," */ + } + + /* This should match below where we fill in the buffer. */ + if (CPP_OPTION (pfile, traditional)) + len += _cpp_replacement_text_len (macro); + else + { + unsigned int count = macro_real_token_count (macro); + for (i = 0; i < count; i++) + { + const cpp_token *token = ¯o->exp.tokens[i]; + + if (token->type == CPP_MACRO_ARG) + len += NODE_LEN (token->val.macro_arg.spelling); + else + len += cpp_token_len (token); + + if (token->flags & STRINGIFY_ARG) + len++; /* "#" */ + if (token->flags & PASTE_LEFT) + len += 3; /* " ##" */ + if (token->flags & PREV_WHITE) + len++; /* " " */ + } + } + + if (len > pfile->macro_buffer_len) + { + pfile->macro_buffer = XRESIZEVEC (unsigned char, + pfile->macro_buffer, len); + pfile->macro_buffer_len = len; + } + + /* Fill in the buffer. Start with the macro name. */ + buffer = pfile->macro_buffer; + buffer = _cpp_spell_ident_ucns (buffer, node); + + /* Parameter names. */ + if (macro->fun_like) + { + *buffer++ = '('; + for (i = 0; i < macro->paramc; i++) + { + cpp_hashnode *param = macro->parm.params[i]; + + if (param != pfile->spec_nodes.n__VA_ARGS__) + { + memcpy (buffer, NODE_NAME (param), NODE_LEN (param)); + buffer += NODE_LEN (param); + } + + if (i + 1 < macro->paramc) + /* Don't emit a space after the comma here; we're trying + to emit a Dwarf-friendly definition, and the Dwarf spec + forbids spaces in the argument list. */ + *buffer++ = ','; + else if (macro->variadic) + *buffer++ = '.', *buffer++ = '.', *buffer++ = '.'; + } + *buffer++ = ')'; + } + + /* The Dwarf spec requires a space after the macro name, even if the + definition is the empty string. */ + *buffer++ = ' '; + + if (CPP_OPTION (pfile, traditional)) + buffer = _cpp_copy_replacement_text (macro, buffer); + else if (macro->count) + /* Expansion tokens. */ + { + unsigned int count = macro_real_token_count (macro); + for (i = 0; i < count; i++) + { + const cpp_token *token = ¯o->exp.tokens[i]; + + if (token->flags & PREV_WHITE) + *buffer++ = ' '; + if (token->flags & STRINGIFY_ARG) + *buffer++ = '#'; + + if (token->type == CPP_MACRO_ARG) + { + memcpy (buffer, + NODE_NAME (token->val.macro_arg.spelling), + NODE_LEN (token->val.macro_arg.spelling)); + buffer += NODE_LEN (token->val.macro_arg.spelling); + } + else + buffer = cpp_spell_token (pfile, token, buffer, true); + + if (token->flags & PASTE_LEFT) + { + *buffer++ = ' '; + *buffer++ = '#'; + *buffer++ = '#'; + /* Next has PREV_WHITE; see _cpp_create_definition. */ + } + } + } + + *buffer = '\0'; + return pfile->macro_buffer; +} + +//-------------------------------------------------------------------------------- + +bool +_assign_handler(cpp_reader *pfile, cpp_hashnode *node){ + cpp_macro *macro; + + if (CPP_OPTION (pfile, traditional)) + macro = _cpp_create_trad_definition (pfile); + else + macro = create_iso_definition (pfile); + + if (!macro) + return false; + + if (cpp_macro_p (node)) + { + if (CPP_OPTION (pfile, warn_unused_macros)) + _cpp_warn_if_unused_macro (pfile, node, NULL); + + if (warn_of_redefinition (pfile, node, macro)) + { + const enum cpp_warning_reason reason + = (cpp_builtin_macro_p (node) && !(node->flags & NODE_WARN)) + ? CPP_W_BUILTIN_MACRO_REDEFINED : CPP_W_NONE; + + bool warned = + cpp_pedwarning_with_line (pfile, reason, + pfile->directive_line, 0, + "\"%s\" redefined", NODE_NAME (node)); + + if (warned && cpp_user_macro_p (node)) + cpp_error_with_line (pfile, CPP_DL_NOTE, + node->value.macro->line, 0, + "this is the location of the previous definition"); + } + _cpp_free_definition (node); + } + + /* Enter definition in hash table. */ + node->type = NT_USER_MACRO; + node->value.macro = macro; + if (! ustrncmp (NODE_NAME (node), DSC ("__STDC_")) + && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_FORMAT_MACROS") + /* __STDC_LIMIT_MACROS and __STDC_CONSTANT_MACROS are mentioned + in the C standard, as something that one must use in C++. + However DR#593 and C++11 indicate that they play no role in C++. + We special-case them anyway. */ + && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_LIMIT_MACROS") + && ustrcmp (NODE_NAME (node), (const uchar *) "__STDC_CONSTANT_MACROS")) + node->flags |= NODE_WARN; + + /* If user defines one of the conditional macros, remove the + conditional flag */ + node->flags &= ~NODE_CONDITIONAL; + + return true; +} + + + + + +#if 0 +static cpp_token * +assign_name_argument(cpp_reader *pfile){ + const cpp_token *name_token = cpp_get_token(pfile); + + cpp_warning_with_line( + pfile + ,CPP_W_NONE + ,name_token->src_loc + ,0 + ,"for debug, assign name is being set to: %.*s" + ,name_token->val.str.len + ,name_token->val.str.text + ); + + if(name_token->type != CPP_NAME){ + cpp_error_with_line( + pfile + ,CPP_DL_ERROR + ,name_token->src_loc + ,0 + ,"First argument to #assign must be a macro name, instead found: %.*s" + ,name_token->val.str.len + ,name_token->val.str.text + ); + return NULL; + } + + // export this into the wider context + cpp_token *copy = (cpp_token *) _cpp_reserve_room(pfile ,0 ,sizeof(cpp_token)); + *copy = *name_token; + return copy; +} + +void assign_handler(cpp_reader *pfile){ + + // parse name argument + const cpp_token *name_token = assign_name_argument(pfile); + if(!name_token) return; + + // create macro + cpp_macro *macro = _cpp_new_macro( + pfile + ,cmk_macro + ,_cpp_reserve_room(pfile ,0 ,sizeof(cpp_macro)) + ); + + macro->fun_like = 0; + macro->paramc = 0; + macro->variadic = 0; + macro->count = 1; + macro->used = 1; + + // fill value + cpp_token *value_token = ¯o->exp.tokens[0]; + value_token->type = CPP_NUMBER; + value_token->val.str.text = (const unsigned char *) "42"; + value_token->val.str.len = 2; + value_token->flags = 0; + + // enter the definition into the symbol table + cpp_hashnode *node = name_token->val.node.node; + node->type = NT_USER_MACRO; + node->value.macro = macro; + + _cpp_mark_macro_used(node); + cpp_warning(pfile ,CPP_W_NONE ,"Assigned macro %s as 42" ,NODE_NAME(node)); +} + +#endif + +#if 0 +static cpp_hashnode * +assign_name_argument(cpp_reader *pfile){ + cpp_hashnode *node = lex_macro_node(pfile); + + if( !node || cpp_ide_is_keyword(node) ){ + cpp_error(pfile ,CPP_DL_ERROR ,"First argument to #assign must be a macro name"); + return NULL; + } + + cpp_warning(pfile ,CPP_W_NONE ,"for debug, assign name is being set to: %s", NODE_NAME(node)); + return node; +} + +void +assign_handler(cpp_reader *pfile){ + + cpp_hashnode *node = assign_name_argument(pfile); + if( !node ) + return; // error already reported + + // create macro + cpp_macro *macro = _cpp_new_macro( + pfile + ,cmk_macro + ,_cpp_reserve_room(pfile ,0 ,sizeof(cpp_macro)) + ); + + macro->fun_like = 0; + macro->paramc = 0; + macro->variadic = 0; + macro->count = 1; + macro->used = 1; + + // fill value + cpp_token *value_token = ¯o->exp.tokens[0]; + value_token->type = CPP_NUMBER; + value_token->val.str.text = (const unsigned char *) "42"; + value_token->val.str.len = 2; + value_token->flags = 0; + + // install macro + node->type = NT_USER_MACRO; + node->value.macro = macro; + + _cpp_mark_macro_used(node); + cpp_warning(pfile ,CPP_W_NONE ,"Assigned macro %s as 42" ,NODE_NAME(node)); +} +#endif diff --git "a/script_gcc_min-12\360\237\226\211/rebuild_gcc.sh" "b/script_gcc_min-12\360\237\226\211/rebuild_gcc.sh" new file mode 100755 index 0000000..447442a --- /dev/null +++ "b/script_gcc_min-12\360\237\226\211/rebuild_gcc.sh" @@ -0,0 +1,21 @@ +#!/bin/bash +# rebuild_gcc.sh – no structural changes, and build directory is still intact + +set -euo pipefail + +source "$(dirname "$0")/environment.sh" + +echo "🔧 Starting GCC rebuild..." + +pushd "$GCC_BUILD" + + echo "gcc: $(command -v gcc)" + echo "toolchain: $TOOLCHAIN" + + $MAKE -j"$MAKE_JOBS" + $MAKE install + +popd + +echo "✅ GCC re-installed to $TOOLCHAIN/bin" +"$TOOLCHAIN/bin/gcc" --version