From 375690b61e8ec4735b19bcead756ba526edab57e Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Fri, 26 Oct 2018 12:27:02 +0100 Subject: parse: initial parsing of __attribute__((format)) Add code to parse the __attribute__((format)) used to indicate that a variadic function takes a printf-style format string and where those are. Save the data in ctype ready for checking when such an function is encountered. Signed-off-by: Ben Dooks Signed-off-by: Luc Van Oostenryck --- parse.c | 75 ++++++++++++++++++++++++++++++++++++++++- symbol.h | 17 ++++++++-- validation/varargs-format-bad.c | 1 - 3 files changed, 89 insertions(+), 4 deletions(-) diff --git a/parse.c b/parse.c index 31ecef0f..55cca01e 100644 --- a/parse.c +++ b/parse.c @@ -85,7 +85,7 @@ static attr_t attribute_address_space, attribute_context, attribute_designated_init, attribute_transparent_union, ignore_attribute, - attribute_mode, attribute_force; + attribute_mode, attribute_force, attribute_format; typedef struct symbol *to_mode_t(struct symbol *); @@ -377,6 +377,10 @@ static struct symbol_op attr_force_op = { .attribute = attribute_force, }; +static struct symbol_op attr_format_op = { + .attribute = attribute_format, +}; + static struct symbol_op address_space_op = { .attribute = attribute_address_space, }; @@ -436,6 +440,16 @@ static struct symbol_op mode_word_op = { .to_mode = to_word_mode }; +static struct symbol_op attr_printf_op = { + .type = KW_FORMAT, + .class = FMT_PRINTF, +}; + +static struct symbol_op attr_scanf_op = { + .type = KW_FORMAT, + .class = FMT_SCANF, +}; + /* * Define the keyword and their effects. * The entries in the 'typedef' and put in NS_TYPEDEF and @@ -551,6 +565,9 @@ static struct init_keyword { D("pure", &attr_fun_op, .mods = MOD_PURE), A("const", &attr_fun_op, .mods = MOD_PURE), D("gnu_inline", &attr_fun_op, .mods = MOD_GNU_INLINE), + D("format", &attr_format_op), + D("printf", &attr_printf_op), + D("scanf", &attr_scanf_op), /* Modes */ D("mode", &mode_op), @@ -1188,6 +1205,60 @@ static struct token *attribute_address_space(struct token *token, struct symbol return token; } +static int invalid_format_args(long long start, long long at) +{ + return start < 0 || at < 0 || start > USHRT_MAX || at > USHRT_MAX || + (start == at && start > 0) || + (start == 0 && at == 0); +} + +static struct token *attribute_format(struct token *token, struct symbol *attr, struct decl_state *ctx) +{ + struct expression *args[3]; + int type = FMT_UNKNOWN; + + /* expecting format ( type, start, va_args at) */ + + token = expect(token, '(', "after format attribute"); + if (token_type(token) != TOKEN_IDENT) { + sparse_error(token->pos, "identifier expected for format type"); + } else { + struct symbol *sym = lookup_keyword(token->ident, NS_KEYWORD); + if (sym && sym->op && sym->op->type == KW_FORMAT) + type = sym->op->class; + } + token = conditional_expression(token, &args[0]); + token = expect(token, ',', "format attribute type"); + token = conditional_expression(token, &args[1]); + token = expect(token, ',', "format attribute type position"); + token = conditional_expression(token, &args[2]); + token = expect(token, ')', "format attribute arg position"); + + if (!args[0] || !args[1] || !args[2]) { + // incorrect format attribute + } else if (type != FMT_PRINTF) { + // only printf-style is supported, skip anything else + } else { + long long start, at; + + start = get_expression_value(args[2]); + at = get_expression_value(args[1]); + + if (invalid_format_args(start, at)) { + warning(token->pos, "bad format positions"); + } else if (start == 0) { + /* nothing to do here, is va_list function */ + } else if (start < at) { + warning(token->pos, "format cannot be after va_args"); + } else { + ctx->ctype.format.index = at; + ctx->ctype.format.first = start; + } + } + + return token; +} + static struct symbol *to_QI_mode(struct symbol *ctype) { if (ctype->ctype.base_type != &int_type) @@ -2957,6 +3028,8 @@ struct token *external_declaration(struct token *token, struct symbol_list **lis if (!(decl->ctype.modifiers & MOD_STATIC)) decl->ctype.modifiers |= MOD_EXTERN; + + base_type->ctype.format = decl->ctype.format; } else if (base_type == &void_ctype && !(decl->ctype.modifiers & MOD_EXTERN)) { sparse_error(token->pos, "void declaration"); } diff --git a/symbol.h b/symbol.h index 6d25e7fc..0d5439ee 100644 --- a/symbol.h +++ b/symbol.h @@ -82,8 +82,8 @@ enum keyword { KW_ASM = 1 << 5, KW_MODE = 1 << 6, KW_STATIC = 1 << 7, - // KW UNUSED = 1 << 8, - KW_EXACT = 1 << 9, + KW_EXACT = 1 << 8, + KW_FORMAT = 1 << 9, }; struct context { @@ -95,12 +95,25 @@ extern struct context *alloc_context(void); DECLARE_PTR_LIST(context_list, struct context); +/* the types of formatting from __attribute__((format)) */ +enum { + FMT_UNKNOWN, + FMT_PRINTF, + FMT_SCANF, +}; + +struct attr_format { + unsigned short index; /* index in argument list for format string */ + unsigned short first; /* where first variadic argument is */ +}; + struct ctype { struct symbol *base_type; unsigned long modifiers; unsigned long alignment; struct context_list *contexts; struct ident *as; + struct attr_format format; }; struct decl_state { diff --git a/validation/varargs-format-bad.c b/validation/varargs-format-bad.c index d4635708..82ae357c 100644 --- a/validation/varargs-format-bad.c +++ b/validation/varargs-format-bad.c @@ -9,7 +9,6 @@ static void test(void) { /* * check-name: variadic formatting test with bad formatting parameters * check-command: sparse -Wformat $file - * check-known-to-fail * * check-error-start varargs-format-bad.c:2:73: warning: bad format positions -- cgit 1.2.3-korg